1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Use fewer forks & execs for sending many messages to a single host.
9 By passing back the next message-id from the transport to the delivery
10 process, we can loop there. A two-phase queue run will benefit,
11 particularly for mailinglist and smarthost cases.
16 JH/01 Support list of dkim results in the dkim_status ACL condition, making
17 it more usable in the data ACL.
19 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
20 Previously This was only logged, on the assumption that errors would be
21 seen for a previous fflush(). However, a fuse filesystem has been
22 reported as showing this an error for the fclose(). The spool is now in
23 an uncertain state, and we have logged and responded acceptance. Change
24 this to respond with a temp-reject, wipe spoolfiles, and log the error
27 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
28 during a recipient verify callout, a QUIT command was attempted on the
29 now-closed callout channel, causing a paniclog entry.
31 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
32 a connection_reject log_selector, under tls_on_connect. Previously
33 with this combination, when the connect ACL rejected, a spurious
34 paniclog entry was made.
36 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
37 of loadbalancer-detection for resumption, in 4.96 - which tries to
38 use the EHLO response. SMTPS does not have one at the time it is starting
39 TLS. Change the default for the smtp transport host_name_extract option
40 to be a static string, for TLS-on-connect cases; meaning that resumption
41 will always be attempted (unless deliberately overriden).
43 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
44 chunk-separator specification. This was broken by hardening introduced
47 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
48 included the -H with the id; this also messed up exiqgrep.
50 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
51 was not properly skipped and empty elements would cause mis-parsing.
52 Tighten parsing of DKIM header records. Previously, all but lowercase
53 alpha chars would be ignored in potential tag names.
55 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
56 being searched. Previously we only had one for IPv4 addresses. Per the
57 documentation, the error results by default in a no-match result for the
58 list. It is logged if the unknown_in_list log_selector is used.
60 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
61 the routing operation. Previously it would silently stop routing the
64 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
65 info-request event close in time to a scheduled run timer could result in
66 the latter being missed, and no further queue scheduled runs being
67 initiated. This ouwld be more likely on high-load systems.
69 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
70 LF-only mode (as detected from the first header line). Previously we did
71 accept that in (normal) CRLF mode; this has been raised as a possible
72 attack scenario (under the name "smtp smuggling").
74 JH/13 Add an fdatasync call for the received message data file in spool, before
75 loggging reception and sending the SMTP ack. Previously we only flushed
76 the stdio buffer so there was still the possibility of a disk error.
78 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
79 address. Previously, for the last address in a header line (commonly
80 there is only one) the terminating newline was part of the logged
83 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
84 header, when parsing for address-qualification. Previously one was only
85 written if there were rewrite rules.
87 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
90 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
91 the multiple-queue-runners facility for 4.97 broke this, giving only a
92 one-time run of the queue.
94 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
97 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
98 TLS session resulted in a forced-fail, a misleading error was logged for
99 the expansino of tls_certificates. This would affect the common case of
100 that option being set (main-section options) but not having any variable
101 parts. It could also potentially affect tls_privatekeys. The underlyding
102 coding errors go back to 4.90 but were only exposed in 4.97.
104 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
105 pcre2 library starting allocating 20kB rather than 112 bytes per match
106 call, which broke the 2GB total limitation on Exim's memory management
107 when a user had over 104207 messages stored and the appendfile
108 maildir_quota_directory_regex option is in use. Release the allocated
109 memory every thosand files to avoid this.
110 The same issue arises with the ACL regex condition, which is applied
111 to every line of a received message.
113 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
114 which all recipients had been handled (itself an issue) a null-pointer
115 deref was done on trying to write a retry record. Fix that by counting
116 the outstanding recipients before trying to transmit the message.
117 The situation arose for a second MX try within a transport run, when the
118 first had perm-rejected a recipient (the only one for the connection, in
119 the case seen) during pipelining, and then closed the TCP connection.
120 The transport classified that as an I/O error, leaving the message
121 outstanding but having marked up the recipient as dealt-with. It then
122 tried another MX because of the I/O error. Fix this by converting the
123 message-level status to ok if there was a close but all recipients were
124 dealt with. Thanks to Wolfgand Breyha for debug runs.
126 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
127 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
129 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
131 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
132 option specified a fixed address, later rewriting on that address would
133 trip on the configuration data being readonly. Instead of modifying
134 in-place, copy data. Found and fixed by Peter Benie.
136 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
137 data a null pointer was followed. Find and testcase by Sebastian Bugge.
139 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
140 decimal 127 chars were not encoded, and lowercase hex was used for
141 encoded values. Outstanding since at least 1999.
143 JH/27 Fix crash in logging. When a message with a large number of recipients
144 had been received, and logging of recipients is enabled, the buffer used
145 for logging could reach limit. A read using a null pointer would then
146 be done, resulting in a crash of the receiving process before an SMTP
147 ACK for the message was returned to the sending system. Duplicate
148 messages were created as a result.
149 Find and debug help by Mateusz Krawczyk
151 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
152 DB, wrapping transport record bare-ip "host names" and ipv6
153 "host addresses" in square-brackets. This makes the parsing that
154 exinext does more reliable.
156 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
159 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
161 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
162 ipv6 addresses for the sender. Previously the uncompressed form was used,
163 and if used in textual form this would result in behavior difference
166 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
167 now classed as a protocol error and subject to smtp_max_synprot_errors.
169 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
170 ".." and has following characters.
176 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
177 SMTP connection" log lines.
179 JH/02 Option default value updates:
180 - queue_fast_ramp (main) true (was false)
181 - remote_max_parallel (main) 4 (was 2)
183 JH/03 Cache static regex pattern compilations, for use by ACLs.
185 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
186 Make the rewrite never match and keep the logging. Trust the
187 admin to be using verify=header-syntax (to actually reject the message).
189 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
190 (under Linux) we watch the dir containing the final file; previously
191 it would be the dir with the first symlink. We still do not monitor
194 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
195 dn_expand() is happy to pass them through.
197 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
198 the file with the proof had an unchanged name, the new proof(s) were
199 loaded on top of the old ones (and nover used; the old ones were stapled).
201 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
202 more than one message arrived in a single connection a reference from
203 the earlier message could be re-used. Often a sigsegv resulted.
204 These variables were introduced in Exim 4.87.
205 Debug help from Graeme Fowler.
207 JH/09 Fix ${filter } for conditions that modify $value. Previously the
208 modified version would be used in construction the result, and a memory
211 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
212 Find and fix by Jasen Betts.
214 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
215 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
216 the systemwide configuration to override the Exim config.
218 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
219 API changes in libopendmarc.
221 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
222 pid 1, in the normal "background daemon" mode, having to drop process-
223 group leadership also lost track of needing to create listener sockets.
225 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
226 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
228 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
229 a capture group which obtained no text (eg. "(abc)*" matching zero
230 occurrences) could cause a segfault if the corresponding $<n> was
233 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
234 included a close-brace character (eg. it itself used an expansion) an
237 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
238 starting TLS. Previously it was after, meaning that attackers on such
239 ports had to be screened using the host_reject_connection main config
240 option. The new sequence aligns better with the STARTTLS behaviour, and
241 permits defences against crypto-processing load attacks, even though it
242 is strictly an incompatible change.
243 Also, avoid sending any SMTP fail response for either the connect ACL
244 or host_reject_connection, for TLS-on-connect ports.
246 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
247 Previously this was not permitted, but it makes reasonable sense.
248 While there, restore a restriction on using it from a connect ACL; given
249 the change JH/16 it could only return false (and before 4.91 was not
252 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
253 was exactly sized compared to the log buffer, a crash occurred with the
254 misleading message "bad memory reference; pool not found".
255 Found and traced by Jasen Betts.
257 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
258 dns_again_means_nonexist included an element causing a DNS lookup which
259 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
260 included (though probably not limited to) a process crash from stack
261 memory limit, or from excessive open files. Replace this with a paniclog
262 whine (as this is likely a configuration error), and returning
265 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
266 this always failed, probably leading to the usual downgrade to in-clear
269 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
270 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
271 to not checking that list for these lookups.
273 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
276 JH/23 Fix crash in string expansions. Previously, if an empty variable was
277 immediately followed by an expansion operator, a null-indirection read
278 was done, killing the process.
280 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
281 include an SMTP response string which is longer than that supported
282 by the delivering transport. Alleviate by wrapping such lines before
285 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
286 chars (RFC limit). Previously a limit of 12 items was made, which with
287 a not-impossible References: in the message being bounced could still
288 be over-large and get stopped in the transport.
290 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
291 close. Previously a bare socket close was done.
293 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
296 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
297 message being considered for delivery was already being handled by
298 another process, and having an SMTP connection already open, the function
299 to close it tried to use an uninitialized variable. This would afftect
300 high-volume sites more, especially when running mailing-list-style loads.
301 Pollution of logs was the major effect, as the other process delivered
302 the message. Found and partly investigated by Graeme Fowler.
304 JH/29 Change format of the internal ID used for message identification. The old
305 version only supported 31 bits for a PID element; the new 64 (on systems
306 which can use Base-62 encoding, which is all currently supported ones
307 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
308 and must use Base-36). The new ID is 23 characters rather than 16, and is
309 visible in various places - notably logs, message headers, and spool file
310 names. Various of the ancillary utilities also have to know the format.
311 As well as the expanded PID portion, the sub-second part of the time
312 recorded in the ID is expanded to support finer precision. Theoretically
313 this permits a receive rate from a single comms channel of better than the
315 The major timestamp part of the ID is not changed; at 6 characters it is
316 usable until about year 3700.
317 Updating from previously releases is fully supported: old-format spool
318 files are still usable, and the utilities support both formats. New
319 message will use the new format. The one hints-DB file type which uses
320 message-IDs (the transport wait- DB) will be discarded if an old-format ID
321 is seen; new ones will be built with only new-format IDs.
322 Optionally, a utility can be used to convert spool files from old to new,
323 but this is only an efficiency measure not a requirement for operation
324 Downgrading from new to old requires running a provided utility, having
325 first stopped all operations. This will convert any spool files from new
326 back to old (losing time-precision and PID information) and remove any
327 wait- hints databases.
329 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
330 we treated them as item separators when parsing for a list item, but they
331 need to be protected by the doublequotes. While there, add handling for
334 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
335 Found and fixed by Jasen Betts. No testcase for this as my usual text
336 editor insists on emitting only valid UTF-8.
338 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
339 a null-indirection SIGSEGV for the receive process.
341 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
342 -bh use. Internal checking would spot this and take a panic.
344 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
345 In 4.96 this would expand to empty.
347 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
348 certificate. Find and fix by Andreas Metzler.
350 JH/36 Add ARC info to DMARC hostory records.
352 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
353 or fakedefer. Previously the sender could discover that the message
354 had in fact been accepted.
356 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
357 sequences. Previously the input was not noted as being tainted; notably
358 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
359 bad coding of authenticators.
361 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
362 and ${tr...}. Found and diagnosed by Heiko Schlichting.
364 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
365 could be triggered by externally-supplied input. Found by Trend Micro.
368 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
369 be triggered by externally-controlled input. Found by Trend Micro.
372 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
373 be triggered by externally-controlled input. Found by Trend Micro.
376 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
377 Make the rewrite never match and keep the logging. Trust the
378 admin to be using verify=header-syntax (to actually reject the message).
380 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
383 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
389 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
390 after reception to before a subsequent reception. This should
391 mean slightly faster delivery, and also confirmation of reception
394 JH/02 Move from using the pcre library to pcre2. The former is no longer
395 being developed or supported (by the original developer).
397 JH/03 Constification work in the filters module required a major version
398 bump for the local-scan API. Specifically, the "headers_charset"
399 global which is visible via the API is now const and may therefore
400 not be modified by local-scan code.
402 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
403 sendfile() didi not account for the way the ClamAV driver code called it.
405 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
406 time check was being done for every character; replace that with one
409 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
410 sent was prefixed with a length byte.
412 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
413 RFC 5321. Previously Dovecot (at least) would log errors during
416 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
417 in 4.61 without a reason logged. Binaries will be bigger, which might
418 matter on diskspace-constrained systems, but debug is easier.
420 JH/09 Fix macro-definition during "-be" expansion testing. The move to
421 write-protected store for macros had not accounted for these runtime
422 additions; fix by removing this protection for "-be" mode.
424 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
425 handing out large-numbered file descriptors, violating the usual Unix
426 assumption (and required by Posix) that the lowest possible number will be
427 allocated by the kernel when a new one is needed. In the daemon, and any
428 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
429 are not useable for FD_SET() [and hence select()] and overwrite the stack.
430 Assorted crashes happen.
432 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
433 main-section options or in a connect ACL, the value from the first ever
434 connection was never replaced for subsequent connections. Found by
437 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
438 though only once PCRE2 was introduced: the memory accounting used under
439 debug offset allocations by an int, giving a hard trap in early startup.
440 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
442 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
443 with underbars is given. The write-protection of configuration introduced
444 in 4.95 trapped when normalisation was applied to an option not needing
447 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
449 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
450 failing to close the certificates directory, every hour or any time it
453 JH/16 Debugging initiated by an ACL control now continues through into routing
454 and transport processes. Previously debugging stopped any time Exim
455 re-execs, or for processing a queued message.
457 JH/17 The "expand" debug selector now gives more detail, specifically on the
458 result of expansion operators and items.
460 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
461 bad comparison between the option value and the name of the file to
462 be included was done, and a mismatch was wrongly identified.
463 4.88 to 4.95 are affected.
465 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
467 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
468 supplied as the db file-pair basename. Previously, if a directory
469 path was given, for example via the autoreply "once" option, the DB
470 file.pag and file.dir files would be created in that directory's
473 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
474 "taint" log_selector. These were previously deprecated.
476 JH/22 Fix static address-list lookups to properly return the matched item.
477 Previously only the domain part was returned.
479 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
480 the call into OpenSSL to send a TLS Close was being repeated; this
481 resulted in the library waiting for the peer's Close. If that was never
482 sent we waited forever. Fix by tracking send calls.
484 JH/24 The ${run} expansion item now expands its command string elements after
485 splitting. Previously it was before; the new ordering makes handling
486 zero-length arguments simpler. The old ordering can be obtained by
487 appending a new option "preexpand", after a comma, to the "run".
489 JH/25 Taint-check exec arguments for transport-initiated external processes.
490 Previously, tainted values could be used. This affects "pipe", "lmtp" and
491 "queryprogram" transport, transport-filter, and ETRN commands.
492 The ${run} expansion is also affected: in "preexpand" mode no part of
493 the command line may be tainted, in default mode the executable name
496 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
497 the the facility was not passed across execs, and only the first message
498 passed over a connection could use BDAT; any further ones using DATA.
500 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
501 uses $sending_ip_address and an interface is specified.
502 Previously any use of the local address in the EHLO name disabled
503 PIPECONNECT, the common case being to use the rDNS of it.
505 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
506 resumption. Previously verify failed because no certificate status is
507 passed on the wire for the restarted session. Fix by using the recorded
508 ocsp status of the stored session for the new connection.
510 JH/29 TLS resumption: the key for session lookup in the client now includes
511 more info that a server could potentially use in configuring a TLS
512 session, avoiding oferring mismatching sessions to such a server.
513 Previously only the server IP was used.
515 JH/30 Fix string_copyn() for limit greater than actual string length.
516 Previously the copied amount was the limit, which could result in a
517 overlapping memcpy for newly allocated destination soon after a
518 source string shorter than the limit. Found/investigated by KM.
520 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
521 close; it may be needed for a subsequent connection. This caused a
522 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
524 JH/32 Fix CHUNKING for a second message on a connection when the first was
525 rejected. Previously we did not reset the chunking-offered state, and
526 erroneously rejected the BDAT command. Investigation help from
529 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
530 an empty address. Previously the expansion returned an error.
532 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
533 proxy. Previously these were misparsed, leading to paniclog entries.
539 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
540 related applications. Previously an "H" was used where available info
541 says that "M" should be, so change to match.
543 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
544 as arguments, so an implementation trying to copy these into a local
545 buffer was taking a taint-enforcement trap. Fix by using dynamically
546 created buffers. Similar fix for radius expansion condition.
548 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
549 reasonable, eg. to count headers. Fix by using dynamically created
550 buffers rather than a local. Do similar fixes for ACL actions "dcc",
551 "log_reject_target", "malware" and "spam"; the arguments are expanded
552 so could be handling tainted values.
554 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
555 broken the (no-op) support for this sendmail command. Restore it
556 to doing nothing, silently, and returning good status.
558 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
559 record path was given (or the default used) without a leading directory
560 path, an error occurred on trying to open it. Use the transport's working
563 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
564 transport. Previously it was the name on the DNS A-record; use instead
565 the head of the CNAME chain leading there (if there is one). This seems
566 to align better with RFC 6125.
568 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
569 smtp_accept_max_per_host allocated resources which were not released
570 when the limit was exceeded. This eventually crashed the daemon. Fix
571 by adding a release action in that path.
573 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
574 expanded; previously using tainted values was rejected. Fix by using
575 dynamically-created buffers.
577 JH/09 Relax restrictions on ACL verify condition needing access to message
578 headers. Previously they were only permitted in data and non-smtp ACLs;
579 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
580 not_blind, header_sender and header_names_ascii verification.
582 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
583 Previously a macro used one argument twice; when called with the
584 argument as an expression having side-effects, incorrect operation
585 resulted. Use an inlineable function.
587 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
588 held open for a verify callout. Previously this wan not accounted for
589 and a corrupt onward SMTP conversation resulted.
591 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
592 Previously they were embedded naively in the constructed address; when
593 needed, strip the quoting and quote the entire local_part.
594 Also make the inbound_srs expansion condition handle quoting.
596 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
597 excluded, not matching the documentation.
599 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
600 was given for the sqlite_dbfile a trap resulted.
602 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
603 "name" argument resulted in a trap. There is no reason to disallow such;
604 this was a coding error.
606 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
607 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
608 spent suspended, ignoring the POSIX definition. Previously we assumed
609 it did and a constant offset from real time could be used as a correction.
610 Change to using the same clock source for the start-of-message and the
611 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
612 exists, just to get a clock slightly more aligned to reality.
614 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
615 RFC says it is optional some validators care. The missing char was not
616 intended but triggered by a line-wrap alignment. Discovery and fix by
617 Guillaume Outters, hacked on by JH.
619 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
620 name being quoted was tainted a trap would be taken. Fix by using
621 dynamically created buffers. The routine could have been called by a
622 rewrite with the "h" flag, by using the "-F" command-line option, or
623 by using a "name=" option on a control=submission ACL modifier.
625 JH/19 SPF: change the Authentication-Results expansion component to give
626 smtp.helo when the sender domain is empty. Previously it gave
629 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
630 not in 127.0.0.0/8 to help in spotting list domains taken over by a
631 domain-parking registrar.
633 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
634 Previously when a whitespace character was specified it was not inserted
635 after removing the newline.
637 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
638 the domain part of the recipient address. This overrides any tls_sni
639 option set, which was previously used.
641 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
644 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
645 is_tainted() had an off-by-one error in the overenthusiastic direction.
646 Find and fix by Gavan. Although NetBSD is not a supported platform for
647 4.94 this bug could affect other platforms.
649 PP/01 Fix default prime selection to be consistent.
650 One path used ike23 still, instead of exim.dev.20160529.3; now both
651 execution flows will use the same DH primes (currently
652 exim.dev.20160529.3).
654 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
655 option in smtp transport, to match the documentation. Previously
656 verification was not being done in some cases where it should have been.
658 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
659 than one server was defined and depending on the platform memory layout
660 details, an internal consistency trap could be hit while walking the list
663 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
664 files. The value is used by the authresults expansion item. Previously
665 if this was used in a router or transport, a crash could result.
667 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
668 have a sufficient incidence of EINTR returns from select that an
669 interaction with other operations done by the main daemon loop exposed
670 a bug in the error-handling. This was benign apart from the log
673 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
674 for consistency with delivered "=>" and failed "**" lines. While we're
675 there, handle PRX and TFO.
677 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
678 applied. This resulted, if any header-line rewrite rules were configured,
679 in a panic-log triggerable by sending a message with a long address in
680 a header. Fix by increasing the arbitrary limit to larger than a single
681 (dewrapped) 5322 header line maximum size.
683 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
684 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
685 RFC 6648 which deprecates X- options in protocols as a general practice.
686 Changeover between the implementations is handled by the mechanisms
689 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
690 a remote which uses callout/hold. Previously the local was queued.
692 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
695 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
696 log line. The asterisk is now consistently appended to the remote IP
697 (and port, if given), and will also be provided on defer and fail log
698 lines. Previously it could be placed on the local IP if that was being
699 logged, and was only provided on delivery lines.
701 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
703 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
704 authenticator client_send option. Previously the next char, after a pair
705 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
706 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
707 documentation. There is still no way to get a leading ^ immediately
708 after a NUL (ie. for the password of a PLAIN method authenticator.
710 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
711 files. For bad sizes read, delete the record and whine to paniclog.
713 JH/38 When logging an AUTH failure, as server, do not include sensitive
714 information. Previously, the credentials would be included if given
715 as part of the AUTH command line and an ACL denied authentication.
717 JH/39 Bug 2691: fix $local_part_data. When the matching list element
718 referred to a file, bad data was returned. This likely also affected
721 JH/40 The gsasl authenticator now supports caching of the salted password
722 generated by the client-side implementation. This required the addition
723 of a new variable: $auth4.
725 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
726 left undeleted; the attempt to re-create it then failed - resulting in
727 the usual "SIGHUP tp have daemon reload configuration" to not work.
728 This affected any platform not supporting "abstract" Unix-domain
729 sockets (i.e. not Linux).
731 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
732 recipients" response to RCPT in a later response, with a 250. The
733 previous coding assumed this would not happen, and under PIPELINING
734 would result in both lost and duplicate recipients for a message.
736 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
737 Previously the weighting was incorrectly applied. Similar fix for socks
738 proxies. Found and fixed by Heiko Schlichting.
740 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
741 not handle sub-lists included using the +namedlist syntax. While
742 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
743 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
746 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
747 Previously the permanent pool was used, so the store could not be freed.
748 This meant a connection with many messages would use continually-growing
751 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
752 per-pool so as not to waste too much space. Previously a constant size
753 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
754 DOS attacks harder. The cost is wasted memory use in the larger blocks.
756 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
757 same space repeatedly, and a smaller process footprint.
759 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
760 for GnuTLS. Previously the name started "%s" which, while not a bug,
761 looked as if if might be one.
763 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
764 "continued connection") make the $proxy_* variables available. Previously
765 the information was not passed across the exec() call for subsequent
766 transport executions. This also mean that the log lines for the
767 messages can show the proxy information.
769 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
770 receive time. With modern systems the difference is significant.
771 The historical behaviour can be restored by disabling (a new) log_selector
772 "queue_time_exclusive".
774 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
775 content) were specifically tested for. Now, also cover expansions
776 resulting in ACL names and inline ACL content.
778 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
779 making it unusable in complex expressions.
781 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
782 fail with a taint-check on the spoolfile name, and leave the message
785 HS/01 Enforce absolute PID file path name.
787 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
789 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
791 PP/02 Bug 2643: Correct TLS DH constants.
792 A missing NUL termination in our code-generation tool had led to some
793 incorrect Diffie-Hellman constants in the Exim source.
794 Reported by kylon94, code-gen tool fix by Simon Arlott.
796 PP/03 Impose security length checks on various command-line options.
797 Fixes CVE-2020-SPRSS reported by Qualys.
799 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
800 better. Reported by Qualys.
802 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
803 providing a particularly obnoxious sender full name.
806 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
808 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
811 PP/08 Change default for recipients_max from unlimited to 50,000.
813 PP/09 Fix security issue with too many recipients on a message (to remove a
814 known security problem if someone does set recipients_max to unlimited,
815 or if local additions add to the recipient list).
816 Fixes CVE-2020-RCPTL reported by Qualys.
818 PP/10 Fix security issue in SMTP verb option parsing
819 Fixes CVE-2020-EXOPT reported by Qualys.
821 PP/11 Fix security issue in BDAT state confusion.
822 Ensure we reset known-good where we know we need to not be reading BDAT
823 data, as a general case fix, and move the places where we switch to BDAT
824 mode until after various protocol state checks.
825 Fixes CVE-2020-BDATA reported by Qualys.
827 HS/03 Die on "/../" in msglog file names
829 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
830 the Exim runtime user are allowed to create files.
832 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
835 QS/03 When reading the output from interpreted forward files we do not
836 pass the pipe between the parent and the interpreting process to
837 executed child processes (if any).
839 QS/04 Always die if requested from internal logging, even is logging is
842 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
843 the API; compilation noo longer completes with DMARC support included.
844 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
845 been reported on other platforms.
847 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
849 JH/56 Make the majority of info read from config files readonly, for defence-in-
850 depth against exploits. Suggestion by Qualys.
851 Not supported on Solaris 10.
853 JH/57 Fix control=fakreject for a custom message containing tainted data.
854 Previously this resulted in a log complaint, due to a re-expansion present
855 since fakereject was originally introduced.
857 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
858 modifier was given, a loop resulted.
860 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
861 pipelined SMTP command followed the BDAT LAST then it would be
862 incorrectly treated as part of the message body, causing a verification
865 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
866 require looser validation rules than those for 821-level addresses,
867 which only permit IP addresses.
873 JH/01 Avoid costly startup code when not strictly needed. This reduces time
874 for some exim process initialisations. It does mean that the logging
875 of TLS configuration problems is only done for the daemon startup.
877 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
879 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
880 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
883 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
884 DKIM signing is being done. Previously a transport_filter would always
885 disable CHUNKING, falling back to traditional DATA.
887 JH/05 Regard command-line recipients as tainted.
889 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
891 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
892 PAM library frees one of the arguments given to it, despite the
893 documentation. Therefore a plain malloc must be used.
895 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
896 on-stack buffers were used, resulting in a taint trap when DSN information
897 copied from a received message was written into the buffer.
899 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
900 the ordering of its ARC headers. This caused a crash.
902 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
903 a new record was being constructed with information from the peer, a trap
906 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
907 installation would get error messages from DMARC verify, when it hit the
908 nonexistent file indicated by the default. Distros wanting DMARC enabled
909 should both provide the file and set the option.
910 Also enforce no DMARC verification for command-line sourced messages.
912 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
913 could, depending on the platform, hang at the STARTTLS response.
915 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
916 message on a connection. Previously if one message had ARC headers and
917 the following one did not, a crash could result when adding an
918 Authentication-Results: header.
920 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
921 Exim main code and Exim-related utities. The introduction of taint
922 tracking also did many adjustments to string handling. Since then, eximon
923 frequently terminated with an assert failure.
925 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
926 check for 452 responses. This slightly helps the inefficieny of doing
927 a large alias-expansion into a recipient-limited target. The max_rcpt
928 transport option still applies (and at the current default, will override
929 the new feature). The check is done for either cause of synch, and forces
930 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
931 connection. The new facility is not tunable at this time.
933 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
934 library live data was being used, so the results became garbage. Make
935 copies while it is still usable.
937 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
938 on delivery deferred (==) and failed (**) lines (if a delivery was
939 attemtped). Previously it was only on completion (=>) lines.
941 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
942 for the expansion of the server_scram_iter and server_scram_salt options.
944 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
945 are now specifically given a NO_DATA response without hitting the system
946 resolver. The library goes on to do the now-standard TXT lookup.
947 Use of dnsdb lookups is not affected.
949 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
950 only retrieve the errormessage once. Previously two calls to dlerror()
951 were used, and the second one (for mainlog/paniclog) retrieved null
954 JH/20 Taint checking: disallow use of tainted data for
955 - the appendfile transport file and directory options
956 - the pipe transport command
957 - the autoreply transport file, log and once options
958 - file names used by the redirect router (including filter files)
960 - paths used by single-key lookups
961 Previously this was permitted.
963 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
964 adjusted the size of a major service buffer; this failed because the
965 buffer was in use at the time. Change to a compile-time increase in the
966 buffer size, when this authenticator is compiled into exim.
968 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
969 previous fast-mode was untenable in the face of glibs using mmap to
970 support larger malloc requests.
972 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
973 New values supported, if defined on system where compiled:
974 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
975 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
977 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
978 running a limited number of proceses in parallel, a benefit is gained. The
979 amount varies with the platform hardware and load. The use of the option
980 queue_run_in_order means we cannot do this, as ordering becomes
983 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
984 had introduced a string-copy (for ensuring NUL-termination) which was not
985 appropriate for that case, which can include embedded NUL bytes in the
986 block of data. Investigation showed the copy to actually be needless, the
987 data being length-specified.
989 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
990 done during a receiving connection, and both used TLS, global info was
991 used rather than per-connection info for tracking the state of data
992 queued for transmission. This could result in a connection hang.
994 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
995 Previously, when delivering serveral messages down a single connection
996 only the first would provide a SIZE. This was due to the size information
997 not being properly tracked.
999 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
1000 TAI (at 37 seconds currently), pretend to be in UTC for time-related
1001 expansion and logging. Previously, spurious values such as a future
1002 minute could be seen.
1004 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
1005 it could crash from a null-deref. This could also affect the
1006 ${addresses: } operator and ${readsock } item.
1008 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1009 message following a mime one, the variable was not reset.
1011 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1012 cached capability (perhaps the peer reneged?) and immediately retry in
1015 JH/31 Fix spurious detection of timeout while writing to transport filter.
1017 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1018 an attempt to copy the string was made before checking it.
1020 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1021 taint of the lookup key was maintained; we now regard the presence in the
1022 filesystem as sufficient validation.
1024 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1025 argument is supplied.
1027 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1028 Previously this was not checked, and nonempty relative paths made an
1029 access under Exim's current working directory.
1031 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1032 Previously no event was raised.
1034 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1035 parameter supplied by the sender MAIL FROM command. Previously it was
1036 ignored, and only the check_spool_space option value for the required
1039 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1040 the size of the signing public-key. Previously it was instead giving
1041 the size of the signature hash.
1043 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1044 the default. See the (new) dkim_verify_min_keysizes option.
1046 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1047 and an ACL use a lookup for checking either the local_part or domain,
1048 stale data could be accessed. Ensure that variable references are
1049 dropped between messages.
1051 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1052 by the client was not checked as pointing within response data before
1053 being used. A malicious client could thus cause an out-of-bounds read and
1054 possibly gain authentication. Fix by adding the check.
1056 JH/42 Internationalisation: change the default for downconversion in the smtp
1057 transport to be "if needed". Previously it was "as previously set" for
1058 the message, which usually meant "if needed" for message-submission but
1059 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1060 even when the envelope addresses did not need it, resulting in forwarding
1061 failures to non-supporting MTAs. A downconvert in such cases will be
1062 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1063 transport. The change does mean that addresses needing conversion will
1064 be converted when previously a delivery failure would occur.
1066 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1067 response was received it would be used unchecked in a fail-DSN, violating
1068 standards on line-length limits. Truncate if needed.
1070 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1071 doesn't work. (Additionally add a new main config option to configure the
1078 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1079 side, to decode a TLS 1.3 packet capture.
1081 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1082 Previously the default library behaviour applied, sending two, each in
1083 its own TCP segment.
1085 JH/03 Debug output for ACL now gives the config file name and line number for
1088 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1090 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1092 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1093 buffer overrun for (non-chunking) other transports.
1095 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1096 TLS1.3, means that a server rejecting a client certificate is not visible
1097 to the client until the first read of encrypted data (typically the
1098 response to EHLO). Add detection for that case and treat it as a failed
1099 TLS connection attempt, so that the normal retry-in-clear can work (if
1100 suitably configured).
1102 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1103 and/or domain. Found and fixed by Jason Betts.
1105 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1106 configuration). If a CNAME target was not a wellformed name pattern, a
1109 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1110 the OS reports them interleaved with other addresses.
1112 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1113 used both for input and for a verify callout, both encrypted, SMTP
1114 responses being sent by the server could be lost. This resulted in
1115 dropped connections and sometimes bounces generated by a peer sending
1118 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1119 string. Previously it was possible to cause undefined behaviour in a
1120 library routine (usually a crash). Found by "zerons".
1122 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1125 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1126 API was removed, so update to use the newer ones.
1128 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1129 any timeout set, is taking a long time. Previously we would hang on to a
1130 rotated logfile "forever" if the input was arriving with long gaps
1131 (a previous attempt to fix addressed lack, for a long time, of initial
1134 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1135 shared (NFS) environment. The length of the tempfile name is now
1136 4 + 16 ("hdr.$message_exim_id") which might break on file
1137 systems which restrict the file name length to lower values.
1138 (It was "hdr.$pid".)
1140 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1141 shared (NFS) environment.
1143 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1144 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1147 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1148 on some platforms for bit 31.
1150 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1151 to changes apparently associated with TLS1.3 handling some of the APIs
1152 previously used were either nonfunctional or inappropriate. Strings
1153 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1154 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1155 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1156 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1157 and the use of specific cipher names in the encrypted= ACL condition.
1159 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1161 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1162 verification result was not updated unless hosts_require_ocsp applied.
1164 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1165 queue_list_requires_admin set to false, non-admin users were denied the
1168 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1169 directory-of-certs mode. Previously they were advertised despite the
1172 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1173 A single TCP connection by a client will now hold a TLS connection open
1174 for multiple message deliveries, by default. Previously the default was to
1177 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1178 default. If built with the facility, DANE will be used. The facility
1179 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1181 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1182 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1183 must be defined and you must still, unless you define DISABLE_TLS, manage
1184 the the include-dir and library-file requirements that go with that
1185 choice. Non-TLS builds are still supported.
1187 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1190 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1191 default. If the platform supports and has the facility enabled, it will
1192 be requested on all coneections.
1194 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1195 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1197 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1199 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1200 success-DSN messages. Previously the From: header was always the default
1201 one for these; the option was ignored.
1203 JH/28 Fix the timeout on smtp response to apply to the whole response.
1204 Previously it was reset for every read, so a teergrubing peer sending
1205 single bytes within the time limit could extend the connection for a
1206 long time. Credit to Qualsys Security Advisory Team for the discovery.
1208 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1209 delivery address, which leaked information of the results of local
1210 forwarding. Change to the original envelope recipient address, per
1213 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1214 requested. Previously not bounce was generated and a log entry of
1215 error ignored was made.
1217 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1219 JH/32 Introduce a general tainting mechanism for values read from the input
1220 channel, and values derived from them. Refuse to expand any tainted
1221 values, to catch one form of exploit.
1223 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1224 was unused and the unexpanded text used for the test. Found and
1225 fixed by Ruben Jenster.
1227 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1228 an attempt to use a TLS library read routine dereffed a nul pointer,
1231 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1232 them in our smtp response.
1234 JH/36 Have the general router option retry_use_local_part default to true when
1235 any of the restrictive preconditions are set (to anything). Previously it
1236 was only for check_local user. The change removes one item of manual
1237 configuration which is required for proper retries when a remote router
1238 handles a subset of addresses for a domain.
1240 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1241 link count into consideration.
1243 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1244 caused the extension of big_buffer, the following lines were ignored.
1246 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1247 accordance with RFC 2308. Previously there was no expiry, so a longlived
1248 receive process (eg. due to ACL delays) versus a short SOA value could
1251 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1253 JH/39 Promote DMARC support to mainline.
1255 JH/40 Bug 2452: Add a References: header to DSNs.
1257 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1258 parameters. The relevant library call is documented as "Deprecated: This
1259 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1260 3.6.0, DH parameters are negotiated following RFC7919."
1262 HS/06 Change the default of dnssec_request_domains to "*"
1264 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1265 carried on and emitted a BDAT command, even when PIPELINING was not
1268 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1269 buffer was used for the filename, resulting in a trap when tainted
1270 arguments (eg. $domain) were used.
1272 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1273 recommended to avoid a possible server-load attack. The feature can be
1274 re-enabled via the openssl_options main cofiguration option.
1276 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1277 for version 4.90 - adding a "more data" boolean to the arguments.
1278 Bumped the ABI version number also, this having been missed previously;
1279 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1280 respect of usage of smtp_printf() by either local_scan code or libraries
1281 accessed via the ${dlfunc } expansion item. Both will need coding
1282 adjustment for any calls to smtp_printf() to match the new function
1283 signature; a FALSE value for the new argument is always safe.
1285 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1286 the file-offset (which the Linux syscall does, and exim expects); this
1287 resulted in an indefinite loop.
1289 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1290 to do ARC verification. The Authentication-Results: header line added
1291 by the configuration then had no ARC item.
1293 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1294 an incorrect mode was used for reading the result, resulting in it being
1301 JH/01 Remove code calling the customisable local_scan function, unless a new
1302 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1304 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1305 non-signal-safe functions being used.
1307 JH/03 Bug 2269: When presented with a received message having a stupidly large
1308 number of DKIM-Signature headers, disable DKIM verification to avoid
1309 a resource-consumption attack. The limit is set at twenty.
1311 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1312 report of oldest_pass in ${authres } in consequence, and separate out
1313 some descriptions of reasons for verification fail.
1315 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1316 files in the spool were present and unlocked. A queue-runner could spot
1317 them, resulting in a duplicate delivery. Fix that by doing the unlock
1318 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1319 add more error-checking on spoolfile handling while that code is being
1322 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1323 No known attacks, no CVE, this is defensive hardening.
1325 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1326 a queue-runner could start a delivery while other operations were ongoing.
1327 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1328 Found and investigated by Tim Stewart. Fix by using the open message data
1329 file handle rather than opening another, and not locally closing it (which
1330 releases a lock) for that case, while creating the temporary .eml format
1331 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1333 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1334 $sender_verify_failure/$recipient_verify_failure to "random".
1336 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1339 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1340 Previously this would segfault.
1342 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1345 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1346 like zero, since the resolver should be doing this for us, But we need one
1347 as a CNAME but no MX presence gets the CNAME returned; we need to check
1348 that doesn't point to an MX to declare it "no MX returned" rather than
1349 "error, loop". A new main option is added so the older capability of
1350 following some limited number of chain links is maintained.
1352 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1354 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1355 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1356 now-deprecated earlier definitions used only specified the range up to TLS
1357 1.2 (in the older-version library docs).
1359 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1361 JH/15 Rework TLS client-side context management. Stop using a global, and
1362 explicitly pass a context around. This enables future use of TLS for
1363 connections to service-daemons (eg. malware scanning) while a client smtp
1364 connection is using TLS; with cutthrough connections this is quite likely.
1366 JH/16 Fix ARC verification to do AS checks in reverse order.
1368 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1370 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1371 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1372 Previously the "utf8" would be re-prepended for every additional message.
1374 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1375 Previously thery were accepted, resulting in issues when attempting to
1376 forward messages to a non-supporting MTA.
1378 PP/02 Let -n work with printing macros too, not just options.
1380 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1381 one parent address was copied, and bogus data was used at delivery-logging
1382 time. Either a crash (after delivery) or bogus log data could result.
1383 Discovery and analysis by Tim Stewart.
1385 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1386 Previously if the string ended mid-character, we did not insert the
1387 promised '?' replacement.
1389 PP/04 Documentation: current string operators work on bytes, not codepoints.
1391 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1392 should pack well giving a smaller memory footprint so better caching and
1393 therefore performance. Group the declarations where this can't be done so
1394 that the byte-sized flag variables are not interspersed among pointer
1395 variables, giving a better chance of good packing by the compiler.
1397 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1398 non-null, to avoid issues with sites running BATV. Previously reports were
1399 sent with an empty envelope sender so looked like bounces.
1401 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1402 The ignore_error flag wasn't being returned from the filter subprocess so
1403 was not set for later routers. Investigation and fix by Matthias Kurz.
1405 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1406 and a msg:complete for the whole, when a message is manually removed using
1407 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1409 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1410 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1411 hope that is portable enough.
1413 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1414 requiring a known-CA anchor certificate; make it now rely entirely on the
1415 TLSA as an anchor. Checking the name on the leaf cert against the name
1416 on the A-record for the host is still done for TA (but not for EE mode).
1418 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1419 included in delivery lines for non-proxied connections, when compiled with
1420 SUPPORT_SOCKS and running with proxy logging enabled.
1422 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1423 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1424 move the existing event to fire before the normal logging of message
1425 failure so that custom logging is bracketed by normal logging.
1427 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1428 msg:fail:internal event. Developement by Matthias Kurz.
1430 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1431 far too small for todays use of crypto signatures stored there. Go all
1432 the way to the max DNS message size of 64kB, even though this might be
1433 overmuch for IOT constrained device use.
1435 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1436 copy a string over itself. The library routine is documented as not
1437 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1439 JH/32 For main options check_spool_space and check_inode_space, where the
1440 platform supports 64b integers, support more than the previous 2^31 kB
1441 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1442 the previous G, M, k.
1444 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1445 $authenticated_fail_id variable on authentication failure. Previously
1448 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1449 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1450 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1451 GNUTLS_SEC_PARAM_MEDIUM.
1453 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1454 side. Previously we would continue as if no SNI had been received.
1456 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1457 "<" character, which should be a list-separator specification, we walked
1458 off past the nul-terimation.
1460 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1461 causes) even when the retry time is not yet met. Previously they were
1462 not, meaning that when (say) an account was over-quota and temp-rejecting,
1463 and multiple senders' messages were queued, only one sender would get
1464 notified on each configured delay_warning cycle.
1466 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1468 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1469 cipher-suites, an error can be left on the stack even for a succeeding
1470 accept; this results in impossible error messages when a later operation
1473 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1474 return error codes indicating retry. Under TLS1.3 this becomes required.
1476 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1477 it only wrote the new authenticators, resulting in a lack of tracking of
1478 peer changes of ESMTP extensions until the next cache flush.
1480 JH/41 Fix the loop reading a message header line to check for integer overflow,
1481 and more-often against header_maxsize. Previously a crafted message could
1482 induce a crash of the recive process; now the message is cleanly rejected.
1484 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1485 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1491 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1492 When redis_servers is set to a list of > 1 element, and the Redis servers
1493 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1494 case of MOVED into a DEFER case instead, thus moving the query onto the
1495 next server in the list. For a cluster of N elements, all N servers must
1496 be defined in redis_servers.
1498 GF/02 Catch and remove uninitialized value warning in exiqsumm
1499 Check for existence of @ARGV before looking at $ARGV[0]
1501 JH/01 Replace the store_release() internal interface with store_newblock(),
1502 which internalises the check required to safely use the old one, plus
1503 the allocate and data copy operations duplicated in both (!) of the
1504 extant use locations.
1506 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1507 modifier. This matches the restriction on the commandline.
1509 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1510 Previously only the last row was returned.
1512 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1513 we assumed that tags in the header were well-formed, and parsed the
1514 element content after inspecting only the first char of the tag.
1515 Assumptions at that stage could crash the receive process on malformed
1518 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1519 While running the DKIM ACL we operate on the Permanent memory pool so that
1520 variables created with "set" persist to the DATA ACL. Also (at any time)
1521 DNS lookups that fail create cache records using the Permanent pool. But
1522 expansions release any allocations made on the current pool - so a dnsdb
1523 lookup expansion done in the DKIM ACL releases the memory used for the
1524 DNS negative-cache, and bad things result. Solution is to switch to the
1525 Main pool for expansions.
1526 While we're in that code, add checks on the DNS cache during store_reset,
1527 active in the testsuite.
1528 Problem spotted, and debugging aided, by Wolfgang Breyha.
1530 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1531 When none of the hosts presented to a transport match an already-open
1532 connection, close it and proceed with the list. Previously we would
1533 queue the message. Spotted by Lena with Yahoo, probably involving
1536 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1537 Previously a spurious "250 OK id=" response was appended to the proper
1540 JH/08 The "support for" informational output now, which built with Content
1541 Scanning support, has a line for the malware scanner interfaces compiled
1542 in. Interface can be individually included or not at build time.
1544 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1545 by the template makefile "src/EDITME". The "STREAM" support for an older
1546 ClamAV interface method is removed.
1548 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1549 rows affected is given instead).
1551 JH/11 The runtime Berkeley DB library version is now additionally output by
1552 "exim -d -bV". Previously only the compile-time version was shown.
1554 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1555 SMTP connection. Previously, when one had more recipients than the
1556 first, an abortive onward connection was made. Move to full support for
1557 multiple onward connections in sequence, handling cutthrough connection
1558 for all multi-message initiating connections.
1560 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1561 routers. Previously, a multi-recipient message would fail to match the
1562 onward-connection opened for the first recipient, and cause its closure.
1564 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1565 a timeout on read on a GnuTLS initiating connection, resulting in the
1566 initiating connection being dropped. This mattered most when the callout
1567 was marked defer_ok. Fix to keep the two timeout-detection methods
1570 JH/15 Relax results from ACL control request to enable cutthrough, in
1571 unsupported situations, from error to silently (except under debug)
1572 ignoring. This covers use with PRDR, frozen messages, queue-only and
1575 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1577 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1578 metadata, resulting in a crash in free().
1580 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1581 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1582 Broken also in d185889f4, with init system revamp.
1584 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1585 Previously we abruptly closed the connection after reading a malware-
1586 found indication; now we go on to read the "scan ok" response line,
1589 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1590 ACL. Previously, a crash would result.
1592 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1593 macro text after a replacement (previously it was only once per line) and
1594 by skipping builtin macros when searching for an uppercase lead character.
1596 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1597 for the build is renamed.
1599 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1600 was allocated for every new TLS startup, meaning one per message. Fix
1601 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1603 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1604 reported the original. Fix to report (as far as possible) the ACL
1605 result replacing the original.
1607 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1608 OpenSSL. Certificate information is loaded for every new TLS startup,
1609 and the resources needed to be freed.
1611 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1613 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1616 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1617 DATA response info to the (existing) per-recipient response info for
1618 the "C=" log element. It can have useful tracking info from the
1619 destination system. Patch from Simon Arlott.
1621 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1622 length value. Previously this would segfault.
1624 HS/02 Support Avast multiline protoocol, this allows passing flags to
1625 newer versions of the scanner.
1627 JH/28 Ensure that variables possibly set during message acceptance are marked
1628 dead before release of memory in the daemon loop. This stops complaints
1629 about them when the debug_store option is enabled. Discovered specifically
1630 for sender_rate_period, but applies to a whole set of variables.
1631 Do the same for the queue-runner and queue-list loops, for variables set
1632 from spool message files. Do the same for the SMTP per-message loop, for
1633 certain variables indirectly set in ACL operations.
1635 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1636 as a multi-recipient message from a mailinglist manager). The coding had
1637 an arbitrary cutoff number of characters while checking for more input;
1638 enforced by writing a NUL into the buffer. This corrupted long / fast
1639 input. The problem was exposed more widely when more pipelineing of SMTP
1640 responses was introduced, and one Exim system was feeding another.
1641 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1642 receiving system, and refused recipients seen by the sending system
1643 (propating to people being dropped from mailing lists).
1644 Discovered and pinpointed by David Carter.
1646 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1647 replaced by the ${authresults } expansion.
1649 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1651 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1652 allows proper process termination in container environments.
1654 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1655 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1657 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1658 and "err_perm", deprecated since 4.83 when the RFC-defined words
1659 "temperror" and "permerror" were introduced.
1661 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1662 transport-filters or DKIM-signing. The restriction was lost in the
1663 consolidation of verify-callout and delivery SMTP handling.
1664 Extend the restriction to also cover ARC-signing.
1666 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1667 in defer=pass mode supply a 450 to the initiator. Previously the message
1670 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1671 tls_require_ciphers is used as before.
1673 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1674 "pass_unscanned". Only tmpfails from the scanner are written to
1675 the paniclog, as they may require admin intervention (permission
1676 denied, license issues). Other scanner errors (like decompression
1677 bombs) do not cause a paniclog entry.
1679 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1680 Previously it was possible to log spurious information in receive log
1683 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1684 triggered odd behaviour from Outlook Express clients.
1686 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1689 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1690 since the IETF WG has not yet settled on that versus the original
1691 "bare" representation.
1693 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1694 Previously the millisecond value corrupted the output.
1695 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1696 corrupted the output.
1702 JH/01 Rework error string handling in TLS interface so that the caller in
1703 more cases is responsible for logging. This permits library-sourced
1704 string to be attached to addresses during delivery, and collapses
1705 pairs of long lines into single ones.
1707 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1708 during configuration. Wildcards are allowed and expanded.
1710 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1711 This permits better logging.
1713 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1714 do not close it down and have the child transport start it up again on
1715 the passed-on TCP connection. Instead, proxy the child (and any
1716 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1717 affected: the continued delivery log lines do not have any DNSSEC, TLS
1718 Certificate or OCSP information. TLS cipher information is still logged.
1720 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1721 identical IP addresses on different listening ports. Will also affect
1724 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1725 add noisy ifdef guards to special-case this sillyness.
1726 Patch from Bernd Kuhls.
1728 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1729 than 255 are no longer allowed.
1731 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1732 Disable the session-cache too, which might reduce our load. Since we
1733 currrectly use a new context for every connection, both as server and
1734 client, there is no benefit for these.
1735 GnuTLS appears to not support tickets server-side by default (we don't
1736 call gnutls_session_ticket_enable_server()) but client side is enabled
1737 by default on recent versions (3.1.3 +) unless the PFS priority string
1740 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1741 <https://reproducible-builds.org/specs/source-date-epoch/>.
1743 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1744 the check for any unsuccessful recipients did not notice the limit, and
1745 erroneously found still-pending ones.
1747 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1748 MSG_MORE. Only in-clear (not on TLS connections).
1750 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1751 transport-filter is involved we can buffer the headers in memory for
1752 creating the signature, and read the spool data file once for the
1753 signature and again for transmission.
1755 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1756 4.77 as the kernel support then wasn't solid, having issues in 64bit
1757 mode. Now, it's been long enough. Add support for FreeBSD also.
1759 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1760 case where the routing stage had gathered several addresses to send to
1761 a host before calling the transport for the first, we previously failed
1762 to close down TLS in the old transport process before passing the TCP
1763 connection to the new process. The new one sent a STARTTLS command
1764 which naturally failed, giving a failed delivery and bloating the retry
1765 database. Investigation and fix prototype from Wolfgang Breyha.
1767 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1768 false-negatives in the check that the sender had not preempted a response
1769 or prompt from Exim (running as a server), due to that code's lack of
1770 awareness of the SMTP input buffering.
1772 PP/04 Add commandline_checks_require_admin option.
1773 Exim drops privileges sanely, various checks such as -be aren't a
1774 security problem, as long as you trust local users with access to their
1775 own account. When invoked by services which pass untrusted data to
1776 Exim, this might be an issue. Set this option in main configuration
1777 AND make fixes to the calling application, such as using `--` to stop
1780 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1781 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1782 responses to those, into a single TLS record each way (this usually means
1783 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1786 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1787 affects you only if you're dancing at the edge of the param size limits.
1788 If you are, and this message makes sense to you, then: raise the
1789 configured limit or use OpenSSL 1.1. Nothing we can do for older
1792 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1793 cmdline element to get the documented default one. Previously it was
1796 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1797 are unsuable for non-message-specific reasons - eg. connection timeout,
1798 banner-time rejection.
1800 JH/16 Fix logging of delivery remote port, when specified by router, under
1803 PP/06 Repair manualroute's ability to take options in any order, even if one
1804 is the name of a transport.
1807 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1809 JH/17 Change the list-building routines interface to use the expanding-string
1810 triplet model, for better allocation and copying behaviour.
1812 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1813 Previously it was constructed the first time a possibly-matching string
1814 was met in the configuration file input during startup; now it is done
1817 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1818 compatible one, to avoid the (poorly documented) possibility of a config
1819 file in the working directory redirecting the DB files, possibly correpting
1820 some existing file. CVE-2017-10140 assigned for BDB.
1822 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1823 cache-hot. Previously, although the result was properly cached, the
1824 initial verify call returned a defer.
1826 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1827 the main verify for receipient in uncached-mode.
1829 JH/22 Retire historical build files to an "unsupported" subdir. These are
1830 defined as "ones for which we have no current evidence of testing".
1832 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1833 if present. Previously it was ignored.
1835 JH/24 Start using specified-initialisers in C structure init coding. This is
1836 a C99 feature (it's 2017, so now considered safe).
1838 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1839 if was a fixed-sized field and bitmask ops via macros; it is now more
1842 PP/07 GitHub PR 56: Apply MariaDB build fix.
1843 Patch provided by Jaroslav Å karvada.
1845 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1846 during Coverity cleanups [4.87 JH/47]
1847 Diagnosis and fix provided by Michael Fischer v. Mollard.
1849 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1850 the right size to place the terminating semicolon on its own folded
1851 line, the header hash was calculated to an incorrect value thanks to
1852 the (relaxed) space the fold became.
1854 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1855 and confused the parent.
1857 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1858 which could crash as a result. This could lead to undeliverable messages.
1860 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1863 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1864 matters on fast-turnover and PID-randomising systems, which were getting
1865 out-of-order delivery.
1867 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1868 a possibly-overlapping copy. The symptom was that "Remote host closed
1869 connection in response to HELO" was logged instead of the actual 4xx
1872 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1873 Previously only that bufferd was discarded, resulting in SYMTP command
1876 JH/32 DKIM: when a message has multiple signatures matching an identity given
1877 in dkim_verify_signers, run the dkim acl once for each. Previously only
1878 one run was done. Bug 2189.
1880 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1881 "panic the current process" to "deliberately defer". The panic log is
1882 still written with the problem list name; the mail and reject logs now
1883 get a temp-reject line for the message that was being handled, saying
1884 something like "domains check lookup or other defer". The SMTP 451
1885 message is still "Temporary local problem".
1887 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1888 A crafted sequence of BDAT commands could result in in-use memory beeing
1889 freed. CVE-2017-16943.
1891 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1892 from SMTP input. Previously it was always done; now only done for DATA
1893 and not BDAT commands. CVE-2017-16944.
1895 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1896 to the message (such as an overlong header line). Previously this was
1897 not done and we did not exit BDAT mode. Followon from the previous item
1898 though a different problem.
1904 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1905 than -2003 did; needs libidn2 in addition to libidn.
1907 JH/02 The path option on a pipe transport is now expanded before use.
1909 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1910 Patch provided by "Björn", documentation fix added too.
1912 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1913 missing a wire-to-host endian conversion.
1915 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1916 close after a BDAT command line could be taken as a following command,
1917 giving a synch failure. Fix by only checking for synch immediately
1918 before acknowledging the chunk.
1920 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1921 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1922 macro. Patches provided by Josh Soref.
1924 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1925 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1926 by IANA as a service extension. However, John Klensin suggests that we
1929 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1930 the dkim code may be unix-mode line endings rather than smtp wire-format
1931 CRLF, so prepend a CR to any bare LF.
1933 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1934 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1936 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1937 the first were themselves being wrongly included in the feed into dkim
1938 processing; with most chunk sizes in use this resulted in an incorrect
1939 body hash calculated value.
1941 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1942 DKIM signature block, for verification. Although advised against by
1943 standards it is specifically not ruled illegal.
1945 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1947 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1948 missing a body hash (the bh= tag).
1950 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1951 It seems that HAProxy sends the Proxy Protocol information in clear and
1952 only then does a TLS startup, so do the same.
1954 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1955 TCP connections (such as for Spamd) unless the daemon successfully set
1956 Fast Open mode on its listening sockets. This fixes breakage seen on
1957 too-old kernels or those not configured for Fast Open, at the cost of
1958 requiring both directions being enabled for TFO, and TFO never being used
1959 by non-daemon-related Exim processes.
1961 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1962 endings, at least on the first header line. Try to canonify any that get
1963 past that check, despite the cost.
1965 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1966 now limited to an arbitrary five deep, while parsing addresses with the
1967 strip_excess_angle_brackets option enabled.
1969 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1970 instead leave the unprompted TLS handshake in socket buffer for the
1971 TLS library to consume.
1973 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1975 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1977 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1978 region in one operation, for speed, and this leaves assigned pointers
1979 dangling. Add checks run only under the testsuite which checks all
1980 variables at a store-reset and panics on a dangling pointer; add code
1981 explicitly nulling out all the variables discovered. Fixes one known
1982 bug: a transport crash, where a dangling pointer for $sending_ip_address
1983 originally assigned in a verify callout, is re-used.
1985 PP/06 Drop '.' from @INC in various Perl scripts.
1987 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1989 PP/08 Reduce a number of compilation warnings under clang; building with
1990 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1991 should be warning-free.
1993 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1995 HS/01 Fix portability problems introduced by PP/08 for platforms where
1996 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1998 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1999 chunk. This allows us to accept broken chunked messages. We need a more
2000 general solution here.
2002 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
2003 already-broken messages in the queue.
2005 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
2007 JH/19 Fix reference counting bug in routing-generated-address tracking.
2013 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2014 supports it and a size is available (ie. the sending peer gave us one).
2016 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2017 years of being deprecated). The replacements are the ACLs
2018 acl_smtp_mime and acl_not_smtp_mime.
2020 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2021 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2022 or even an in-clear connection were permitted. Now, if the host lookup
2023 was dnssec and dane was requested then the host is only used if the
2024 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2025 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2026 if one fails this test.
2027 This means that a poorly-configured remote DNS will make it incommunicado;
2028 but it protects against a DNS-interception attack on it.
2030 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2031 non-noisy when a race steals the message being considered.
2033 JH/05 If main configuration option tls_certificate is unset, generate a
2034 self-signed certificate for inbound TLS connections.
2036 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2037 in rewrites and routers.
2039 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2040 and logged a warning sing 4.83; now they are a configuration file error.
2042 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2043 (lacking @domain). Apply the same qualification processing as RCPT.
2045 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2047 JH/10 Support ${sha256:} applied to a string (as well as the previous
2050 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2051 a cutthrough deliver is pending, as we always want to make a connection.
2052 This also avoids re-routing the message when later placing the cutthrough
2053 connection after a verify cache hit.
2054 Do not update it with the verify result either.
2056 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2057 when routing results in more than one destination address.
2059 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2060 signing (which inhibits the cutthrough capability). Previously only
2061 the presence of an option was tested; now an expansion evaluating as
2062 empty is permissible (obviously it should depend only on data available
2063 when the cutthrough connection is made).
2065 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2066 the relevant preceding SMTP command did not note the pipelining mode.
2068 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2069 Previously they were not counted.
2071 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2072 as one having no matching records. Previously we deferred the message
2073 that needed the lookup.
2075 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2076 distinguished as "(=".
2078 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2079 for missing MX records. Previously it only worked for missing A records.
2081 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2083 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2084 after the data-go-ahead and data-ack. Patch from Jason Betts.
2086 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2087 even for a "none" policy. Patch from Tony Meyer.
2089 JH/22 Fix continued use of a connection for further deliveries. If a port was
2090 specified by a router, it must also match for the delivery to be
2093 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2094 When one of the recipients of a message was unsuitable for the connection
2095 (has no matching addresses), we lost track of needing to mark it
2096 deferred. As a result mail would be lost.
2098 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2100 JH/25 Decoding ACL controls is now done using a binary search; the source code
2101 takes up less space and should be simpler to maintain. Merge the ACL
2102 condition decode tables also, with similar effect.
2104 JH/26 Fix problem with one_time used on a redirect router which returned the
2105 parent address unchanged. A retry would see the parent address marked as
2106 delivered, so not attempt the (identical) child. As a result mail would
2109 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2110 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2111 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2114 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2115 Main config options check_{log,spool}_{inodes,space} are now
2116 100 inodes, 10MB unless set otherwise in the configuration.
2118 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2119 Previously it only applied to the main-section connection policy
2122 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2124 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2125 by me. Added RFC7919 DH primes as an alternative.
2127 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2128 are not in the system include path.
2130 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2131 GnuTLS, when a session startup failed (eg because the client disconnected)
2132 Exim did stdio operations after fclose. This was exposed by a recent
2133 change which nulled out the file handle after the fclose.
2135 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2136 signed directly by the cert-signing cert, rather than an intermediate
2137 OCSP-signing cert. This is the model used by LetsEncrypt.
2139 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2141 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2142 an incoming connection.
2144 HS/02 Bug 1802: Do not half-close the connection after sending a request
2147 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2148 fallback to "prime256v1".
2150 JH/34 SECURITY: Use proper copy of DATA command in error message.
2151 Could leak key material. Remotely exploitable. CVE-2016-9963.
2157 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2158 and 3.4.4 - once the server is enabled to respond to an OCSP request
2159 it does even when not requested, resulting in a stapling non-aware
2160 client dropping the TLS connection.
2162 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2163 support variable-length bit vectors. No functional change.
2165 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2166 The I= interface field on outgoing lines is now after the H= remote
2167 host field, same as incoming lines. There is a separate
2168 outgoing_interface log selector which allows you to disable the
2171 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2172 If not running log_selector +smtp_connection the mainlog would be held
2173 open indefinitely after a "too many connections" event, including to a
2174 deleted file after a log rotate. Leave the per net connection logging
2175 leaving it open for efficiency as that will be quickly detected by the
2176 check on the next write.
2178 HS/01 Bug 1671: Fix post transport crash.
2179 Processing the wait-<transport> messages could crash the delivery
2180 process if the message IDs didn't exist for some reason. When
2181 using 'split_spool_directory=yes' the construction of the spool
2182 file name failed already, exposing the same netto behaviour.
2184 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2185 mime_regex ACL conditions.
2187 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2188 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2189 to HELO, local diagnostic string.
2191 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2192 log line to debug. Even when configured with a tls authenticator many
2193 client connections are expected to not authenticate in this way, so
2194 an authenticate fail is not an error.
2196 HS/02 Add the Exim version string to the process info. This way exiwhat
2197 gives some more detail about the running daemon.
2199 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2200 matter for fast-change records such as DNSBLs.
2202 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2203 retry record, even if constant. There may be multiple transports with
2204 different interface settings and the retry behaviour needs to be kept
2207 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2209 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2211 JH/10 Bug 840: fix log_defer_output option of pipe transport
2213 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2214 hosts_randomize. This matters a lot when combined with mua_wrapper.
2216 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2217 ${quote_pgsql:<string>} operator.
2219 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2221 JH/14 Change header file name for freeradius-client. Relevant if compiling
2222 with Radius support; from the Gentoo tree and checked under Fedora.
2224 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2226 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2227 option transport_filter, meaning no filtering.
2229 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2231 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2232 defaults to "*" (all hosts). The variable is now available when not built
2233 with TLS, default unset, mainly to enable keeping the testsuite sane.
2234 If a server certificate is not supplied (via tls_certificate) an error is
2235 logged, and clients will find TLS connections fail on startup. Presumably
2236 they will retry in-clear.
2237 Packagers of Exim are strongly encouraged to create a server certificate
2238 at installation time.
2240 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2241 with the $config_file variable.
2243 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2244 in transport context, after the attempt, and per-recipient. The latter type
2245 is per host attempted. The event data is the error message, and the errno
2246 information encodes the lookup type (A vs. MX) used for the (first) host,
2247 and the trailing two digits of the smtp 4xx response.
2249 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2250 to write to mainlog (or rejectlog, paniclog) in the window between file
2251 creation and permissions/ownership being changed. Particularly affects
2252 installations where exicyclog is run as root, rather than exim user;
2253 result is that the running daemon panics and dies.
2255 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2257 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2258 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2259 "pri" and "weight". Note that the previous implicit priority given by the
2260 list order is no longer honoured.
2262 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2263 for DKIM processing.
2265 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2266 by defining SUPPORT_SOCKS.
2268 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2269 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2270 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2271 variables are renamed to proxy_{local,external}_{address,port}.
2273 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2274 for a build by defining SUPPORT_I18N
2276 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2277 of the query string, and make ${quote_redis:} do that quoting.
2279 JH/29 Move Events support from Experimental to mainline, enabled by default
2280 and removable for a build by defining DISABLE_EVENT.
2282 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2284 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2285 cached by the daemon.
2287 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2288 by defining LOOKUP_REDIS. The libhiredis library is required.
2290 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2291 keys are given for lookup.
2293 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2294 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2295 only supported when built with TLS support. The PolarSSL SHA routines
2296 are still used when the TLS library is too old for convenient support.
2298 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2299 openssl_options), for security. OpenSSL forces this from version 1.1.0
2300 server-side so match that on older versions.
2302 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2303 allocation for $value could be released as the expansion processing
2304 concluded, but leaving the global pointer active for it.
2306 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2307 and to use the domains and local_parts ACL conditions.
2309 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2310 incorrectly not doubled on cutthrough transmission, hence seen as a
2311 body-termination at the receiving system - resulting in truncated mails.
2312 Commonly the sender saw a TCP-level error, and retransmitted the message
2313 via the normal store-and-forward channel. This could result in duplicates
2314 received - but deduplicating mailstores were liable to retain only the
2315 initial truncated version.
2317 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2319 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2321 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2322 we're in there, support oversigning also; bug 1309.
2324 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2326 HS/04 Add support for keep_environment and add_environment options.
2328 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2329 either intentional arithmetic overflow during PRNG, or testing config-
2332 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2333 delivery resulted in actual delivery. Cancel cutthrough before DATA
2336 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2337 rejects a recipient: pass the reject to the originator.
2339 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2340 Many were false-positives and ignorable, but it's worth fixing the
2343 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2344 for the new environment-manipulation done at startup. Move the routines
2345 from being local to tls.c to being global via the os.c file.
2347 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2348 an extract embedded as result-arg for a map, the first arg for extract
2349 is unavailable so we cannot tell if this is a numbered or keyed
2350 extraction. Accept either.
2356 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2359 JH/02 The smtp transport option "multi_domain" is now expanded.
2361 JH/03 The smtp transport now requests PRDR by default, if the server offers
2364 JH/04 Certificate name checking on server certificates, when exim is a client,
2365 is now done by default. The transport option tls_verify_cert_hostnames
2366 can be used to disable this per-host. The build option
2367 EXPERIMENTAL_CERTNAMES is withdrawn.
2369 JH/05 The value of the tls_verify_certificates smtp transport and main options
2370 default to the word "system" to access the system default CA bundle.
2371 For GnuTLS, only version 3.0.20 or later.
2373 JH/06 Verification of the server certificate for a TLS connection is now tried
2374 (but not required) by default. The verification status is now logged by
2375 default, for both outbound TLS and client-certificate supplying inbound
2378 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2381 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2382 Status Notification (bounce) messages are now MIME format per RFC 3464.
2383 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2384 under the control of the dsn_advertise_hosts option, and routers may
2385 have a dsn_lasthop option.
2387 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2388 default, modifiable by a malware= option. The list separator for
2389 the options can now be changed in the usual way. Bug 68.
2391 JH/10 The smtp_receive_timeout main option is now expanded before use.
2393 JH/11 The incoming_interface log option now also enables logging of the
2394 local interface on delivery outgoing connections.
2396 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2397 if the interface and destination host and port all match.
2399 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2402 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2403 Patch from Andrew Lewis.
2405 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2406 now supports optional time-restrictions, weighting, and priority
2407 modifiers per server. Patch originally by <rommer@active.by>.
2409 JH/16 The spamd_address main option now supports a mixed list of local
2410 and remote servers. Remote servers can be IPv6 addresses, and
2411 specify a port-range.
2413 JH/17 Bug 68: The spamd_address main option now supports an optional
2414 timeout value per server.
2416 JH/18 Bug 1581: Router and transport options headers_add/remove can
2417 now have the list separator specified.
2419 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2422 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2425 JH/21 Support for the A6 type of dns record is withdrawn.
2427 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2428 rather than the verbs used.
2430 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2431 from 255 to 1024 chars.
2433 JH/24 Verification callouts now attempt to use TLS by default.
2435 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2436 are generic router options now. The defaults didn't change.
2438 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2439 Original patch from Alexander Shikoff, worked over by JH.
2441 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2442 files if ZCAT_COMMAND is not executable.
2444 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2446 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2448 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2449 Normally benign, it bites when the pair was led to by a CNAME;
2450 modern usage is to not canonicalize the domain to a CNAME target
2451 (and we were inconsistent anyway for A-only vs AAAA+A).
2453 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2455 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2456 when evaluating $sender_host_dnssec.
2458 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2459 $sender_helo_dnssec variable.
2461 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2463 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2465 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2467 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2468 documented as working, but never had. Support all but $spam_report.
2470 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2471 added for tls authenticator.
2473 HS/03 Add perl_taintmode main config option
2479 TL/01 When running the test suite, the README says that variables such as
2480 no_msglog_check are global and can be placed anywhere in a specific
2481 test's script, however it was observed that placement needed to be near
2482 the beginning for it to behave that way. Changed the runtest perl
2483 script to read through the entire script once to detect and set these
2484 variables, reset to the beginning of the script, and then run through
2485 the script parsing/test process like normal.
2487 TL/02 The BSD's have an arc4random API. One of the functions to induce
2488 adding randomness was arc4random_stir(), but it has been removed in
2489 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2490 function when detected.
2492 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2493 cause callback expansion.
2495 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2496 syntax errors in an expansion can be treated as a string instead of
2497 logging or causing an error, due to the internal use of bool_lax
2498 instead of bool when processing it.
2500 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2501 server certificates when making smtp deliveries.
2503 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2505 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2507 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2509 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2510 Merged patch from Sebastian Wiedenroth.
2512 JH/05 Fix results-pipe from transport process. Several recipients, combined
2513 with certificate use, exposed issues where response data items split
2514 over buffer boundaries were not parsed properly. This eventually
2515 resulted in duplicates being sent. This issue only became common enough
2516 to notice due to the introduction of connection certificate information,
2517 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2519 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2520 size buffer was used, resulting in syntax errors when an expansion
2523 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2524 version 3.3.6 or later.
2526 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2527 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2528 both become "event_action", the variables become $event_name, $event_data
2529 and $event_defer_errno. There is a new variable $verify_mode, usable in
2530 routers, transports and related events. The tls:cert event is now also
2531 raised for inbound connections, if the main configuration event_action
2534 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2535 early OCSP support, but no stapling (appears to be less than 1.0.0).
2537 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2538 server certificate names available under the smtp transport option
2539 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2542 JH/10 Time-related extraction expansions from certificates now use the main
2543 option "timezone" setting for output formatting, and are consistent
2544 between OpenSSL and GnuTLS compilations. Bug 1541.
2546 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2547 encoded parameter in the incoming message. Bug 1558.
2549 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2550 include certificate info, eximon was claiming there were spoolfile
2553 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2555 JH/14 Log delivery-related information more consistently, using the sequence
2556 "H=<name> [<ip>]" wherever possible.
2558 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2559 are problematic for Debian distribution, omit them from the release
2562 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2564 JH/16 Fix string representation of time values on 64bit time_t architectures.
2567 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2568 output list separator was used.
2573 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2574 checkers that were complaining about end of non-void function with no
2577 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2578 This was a regression introduced in 4.83 by another bugfix.
2580 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2582 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2583 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2589 TF/01 Correctly close the server side of TLS when forking for delivery.
2591 When a message was received over SMTP with TLS, Exim failed to clear up
2592 the incoming connection properly after forking off the child process to
2593 deliver the message. In some situations the subsequent outgoing
2594 delivery connection happened to have the same fd number as the incoming
2595 connection previously had. Exim would try to use TLS and fail, logging
2596 a "Bad file descriptor" error.
2598 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2599 utilities have not been installed.
2601 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2602 temporary space as the ACL may create new global variables.
2604 TL/01 LDAP support uses per connection or global context settings, depending
2605 upon the detected version of the libraries at build time.
2607 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2608 to extract and use the src ip:port in logging and expansions as if it
2609 were a direct connection from the outside internet. PPv2 support was
2610 updated based on HAProxy spec change in May 2014.
2612 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2614 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2615 Properly escape header and check for NULL return.
2617 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2620 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2622 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2623 characters in header names, implemented as a verify condition.
2624 Contributed by Michael Fischer v. Mollard.
2626 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2627 results permerror and temperror. Previous values are deprecated but
2628 still accepted. In a future release, err_perm and err_temp will be
2629 completely removed, which will be a backward incompatibility if the
2630 ACL tests for either of these two old results. Patch contributed by
2631 user bes-internal on the mailing list.
2633 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2635 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2636 selectors, in both main and reject logs.
2638 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2641 JH/07 Add malware type "sock" for talking to simple daemon.
2643 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2645 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2646 routers/transports under cutthrough routing.
2648 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2649 numbers. Touch up "bool" conditional to keep the same definition.
2651 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2653 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2655 JH/12 Expand items in router/transport headers_add or headers_remove lists
2656 individually rather than the list as a whole. Bug 1452.
2658 Required for reasonable handling of multiple headers_ options when
2659 they may be empty; requires that headers_remove items with embedded
2660 colons must have them doubled (or the list-separator changed).
2662 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2663 view the policy declared in the DMARC record. Currently, $dmarc_status
2664 is a combined value of both the record presence and the result of the
2667 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2669 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2670 dnslookup router and the smtp transport (applying to the forward
2673 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2674 of ldap servers used for a specific lookup. Patch provided by Heiko
2677 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2678 New variable $lookup_dnssec_authenticated for observability.
2680 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2681 Patch submitted by Lars Timman.
2683 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2685 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2686 Requires trusted mode and valid format message id, aborts otherwise.
2687 Patch contributed by Heiko Schlichting.
2689 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2690 certextract with support for various fields. Bug 1358.
2692 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2693 is requested by default, modifiable by smtp transport option
2696 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2697 operate on certificate variables to give certificate fingerprints
2698 Also new ${sha256:cert_variable}.
2700 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2702 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2703 Christian Aistleitner.
2705 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2707 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2708 file. Patch from Wolfgang Breyha.
2710 JH/25 Expand the coverage of the delivery $host and $host_address to
2711 client authenticators run in verify callout. Bug 1476.
2713 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2714 align with daemon_smtp_ports. Bug 72.
2716 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2717 support and error reporting did not work properly.
2719 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2720 and is readable. Patch from Andrew Colin Kissa.
2722 TL/14 Enhance documentation of ${run expansion and how it parses the
2723 commandline after expansion, particularly in the case when an
2724 unquoted variable expansion results in an empty value.
2726 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2728 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2729 by renaming away from it. Addresses GH issue 15, reported by
2732 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2734 TL/15 SECURITY: prevent double expansion in math comparison functions
2735 (can expand unsanitized data). Not remotely exploitable.
2742 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2744 PP/02 Make -n do something, by making it not do something.
2745 When combined with -bP, the name of an option is not output.
2747 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2750 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2751 $sender_host_name and config options to manage this, and basic check
2754 PP/05 DSCP support for outbound connections and control modifier for inbound.
2756 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2757 (Only plugin which currently uses this is kerberos4, which nobody should
2758 be using, but we should make it available and other future plugins might
2759 conceivably use it, even though it would break NAT; stuff *should* be
2760 using channel bindings instead).
2762 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2763 name; added for Sendmail compatibility; requires admin caller.
2764 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2765 just ignore it); requires trusted caller.
2766 Also parse but ignore: -Ac -Am -X<logfile>
2769 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2771 TL/02 Add +smtp_confirmation as a default logging option.
2773 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2774 Patch by Magnus Holmgren from 2007-02-20.
2776 TL/04 Bugzilla 1281 - Spec typo.
2777 Bugzilla 1283 - Spec typo.
2778 Bugzilla 1290 - Spec grammar fixes.
2780 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2782 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2784 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2785 mailing list by by Dmitry Isaikin.
2787 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2789 JH/02 Support "G" suffix to numbers in ${if comparisons.
2791 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2793 NM/01 Bugzilla 1197 - Spec typo
2794 Bugzilla 1196 - Spec examples corrections
2796 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2798 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2799 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2802 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2803 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2805 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2806 "acl {{name}{arg}...}", and optional args on acl condition
2809 JH/05 Permit multiple router/transport headers_add/remove lines.
2811 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2813 JH/07 Avoid using a waiting database for a single-message-only transport.
2814 Performance patch from Paul Fisher. Bugzilla 1262.
2816 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2819 JH/09 Add $headers_added variable, with content from use of ACL modifier
2820 add_header (but not yet added to the message). Bugzilla 199.
2822 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2823 Pulled from Bugzilla 817 by Wolfgang Breyha.
2825 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2827 (nb: this is the same fix as in Exim 4.80.1)
2829 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2832 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2835 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2837 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2838 advertise SMTP AUTH mechanism to us, instead of a generic
2839 protocol violation error. Also, make Exim more robust to bad
2840 data from the Dovecot auth socket.
2842 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2844 When a queue runner is handling a message, Exim first routes the
2845 recipient addresses, during which it prunes them based on the retry
2846 hints database. After that it attempts to deliver the message to
2847 any remaining recipients. It then updates the hints database using
2850 So if a recipient address works intermittently, it can get repeatedly
2851 deferred at routing time. The retry hints record remains fresh so the
2852 address never reaches the final cutoff time.
2854 This is a fairly common occurrence when a user is bumping up against
2855 their storage quota. Exim had some logic in its local delivery code
2856 to deal with this. However it did not apply to per-recipient defers
2857 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2859 This change adds a proper retry rule check during routing so that the
2860 final cutoff time is checked against the message's age. We only do
2861 this check if there is an address retry record and there is not a
2862 domain retry record; this implies that previous attempts to handle
2863 the address had the retry_use_local_parts option turned on. We use
2864 this as an approximation for the destination being like a local
2865 delivery, as in LMTP.
2867 I suspect this new check makes the old local delivery cutoff check
2868 redundant, but I have not verified this so I left the code in place.
2870 TF/02 Correct gecos expansion when From: is a prefix of the username.
2872 Test 0254 submits a message to Exim with the header
2876 When I ran the test suite under the user fanf2, Exim expanded
2877 the header to contain my full name, whereas it should have added
2878 a Resent-Sender: header. It erroneously treats any prefix of the
2879 username as equal to the username.
2881 This change corrects that bug.
2883 GF/01 DCC debug and logging tidyup
2884 Error conditions log to paniclog rather than rejectlog.
2885 Debug lines prefixed by "DCC: " to remove any ambiguity.
2887 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2889 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2890 Bug spotted by Jeremy Harris; was flawed since initial commit.
2891 Would have resulted in OCSP responses post-SNI triggering an Exim
2892 NULL dereference and crash.
2894 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2896 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2897 Bug detection, analysis and fix by Samuel Thibault.
2898 Bugzilla 1331, Debian bug #698092.
2900 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2902 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2903 Server implementation by Todd Lyons, client by JH.
2904 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2905 config variable "prdr_enable" controls whether the server
2906 advertises the facility. If the client requests PRDR a new
2907 acl_data_smtp_prdr ACL is called once for each recipient, after
2908 the body content is received and before the acl_smtp_data ACL.
2909 The client is controlled by both of: a hosts_try_prdr option
2910 on the smtp transport, and the server advertisement.
2911 Default client logging of deliveries and rejections involving
2912 PRDR are flagged with the string "PRDR".
2914 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2915 fclose(). Diagnosis by Todd Lyons.
2917 PP/17 Update configure.default to handle IPv6 localhost better.
2918 Patch by Alain Williams (plus minor tweaks).
2921 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2922 This is now consistent with GnuTLS, and is now documented: the
2923 previous undocumented portable approach to treating the option as
2924 unset was to force an expansion failure. That still works, and
2925 an empty string is now equivalent.
2927 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2928 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2929 not performing validation itself.
2931 PP/20 Added force_command boolean option to pipe transport.
2932 Patch from Nick Koston, of cPanel Inc.
2934 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2937 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2939 PP/21 Fix eximon continuous updating with timestamped log-files.
2940 Broken in a format-string cleanup in 4.80, missed when I repaired the
2941 other false fix of the same issue.
2942 Report and fix from Heiko Schlichting.
2945 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2946 Report from Prashanth Katuri.
2948 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2949 It's SecureTransport, so affects any MacOS clients which use the
2950 system-integrated TLS libraries, including email clients.
2952 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2953 using a MIME ACL for non-SMTP local injection.
2954 Report and assistance in diagnosis by Warren Baker.
2956 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2958 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2960 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2961 last id that failed so it may be referenced in subsequent ACL's.
2963 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2966 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2967 ldap library initialization, allowing self-signed CA's to be
2968 used. Also properly sets require_cert option later in code by
2969 using NULL (global ldap config) instead of ldap handle (per
2970 session). Bug diagnosis and testing by alxgomz.
2972 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2973 the src/util/ subdirectory.
2975 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2976 renamed to Transport Post Delivery Action by Jeremy Harris, as
2979 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2980 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2981 redis_servers = needs to be configured which will be used by the redis
2982 lookup. Patch from Warren Baker, of The Packet Hub.
2984 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2986 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2987 hostname or reverse DNS when processing a host list. Used suggestions
2988 from multiple comments on this bug.
2990 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2992 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2993 Missed a few lines, added it to make the runtest require no keyboard
2996 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2997 contains upper case chars. Make router use caseful_local_part.
2999 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
3000 support when GnuTLS has been built with p11-kit.
3006 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3008 This, or similar/improved, will also be change PP/11 of 4.82.
3014 PP/01 Handle short writes when writing local log-files.
3015 In practice, only affects FreeBSD (8 onwards).
3016 Bugzilla 1053, with thanks to Dmitry Isaikin.
3018 NM/01 Bugzilla 949 - Documentation tweak
3020 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3023 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3025 PP/02 Implemented gsasl authenticator.
3027 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3029 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3030 `pkg-config foo` for cflags/libs.
3032 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3033 with rest of GSASL and with heimdal_gssapi.
3035 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3036 `pkg-config foo` for cflags/libs for the TLS implementation.
3038 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3039 properties get this fed in as external SSF. A number of robustness
3040 and debugging improvements to the cyrus_sasl authenticator.
3042 PP/08 cyrus_sasl server now expands the server_realm option.
3044 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3045 Patch by Jeremy Harris.
3047 PP/10 Added dbmjz lookup type.
3049 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3051 PP/12 MAIL args handles TAB as well as SP, for better interop with
3052 non-compliant senders.
3053 Analysis and variant patch by Todd Lyons.
3055 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3056 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3057 Patch from Dirk Mueller <dmueller@suse.com>
3059 PP/13 tls_peerdn now print-escaped for spool files.
3060 Observed some $tls_peerdn in wild which contained \n, which resulted
3061 in spool file corruption.
3063 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3064 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3065 or write after TLS renegotiation, which otherwise led to messages
3068 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3069 as a tracking header (ie: a signed header comes before the signature).
3070 Patch from Wolfgang Breyha.
3072 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3073 comma-sep list; embedded commas doubled.
3075 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3077 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3079 Report and patch from Dmitry Banschikov.
3081 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3082 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3083 failures. We appear to now support TLS1.1+ with Exim.
3085 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3086 lets Exim select keys and certificates based upon TLS SNI from client.
3087 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3088 before an outbound SMTP session. New log_selector, +tls_sni.
3090 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3091 NULL dereference. Report and patch from Alun Jones.
3093 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3094 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3095 Not seeing resolver debug output on NetBSD, but suspect this is a
3096 resolver implementation change.
3098 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3099 Left warnings. Added "eximon gdb" invocation mode.
3101 PP/21 Defaulting "accept_8bitmime" to true, not false.
3103 PP/22 Added -bw for inetd wait mode support.
3105 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3106 locate the relevant includes and libraries. Made this the default.
3108 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3109 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3111 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3112 This may cause build issues on older platforms.
3114 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3115 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3116 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3117 Added SNI support via GnuTLS too.
3118 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3120 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3122 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3123 Applied second patch from Janne, implementing suggestion to default
3124 multiple-strings-in-record handling to match SPF spec.
3126 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3128 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3129 read-only, out of scope).
3130 Patch from Wolfgang Breyha, report from Stuart Northfield.
3132 PP/29 Fix three issues highlighted by clang analyser static analysis.
3133 Only crash-plausible issue would require the Cambridge-specific
3134 iplookup router and a misconfiguration.
3135 Report from Marcin Mirosław.
3137 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3139 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3140 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3141 As part of this, removing so much warning spew let me fix some minor
3142 real issues in debug logging.
3144 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3145 assignment on my part. Fixed.
3147 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3148 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3149 Janne Snabb (who went above and beyond: thank you).
3151 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3152 string otherwise requires a connection and a bunch more work and it's
3153 relatively easy to get wrong. Should also expose TLS library linkage
3156 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3157 64-bit ${eval} (JH/03).
3159 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3160 GNU libc to support some of the 64-bit stuff, should not lead to
3161 conflicts. Defined before os.h is pulled in, so if a given platform
3162 needs to override this, it can.
3164 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3165 protection layer was required, which is not implemented.
3166 Bugzilla 1254, patch from Wolfgang Breyha.
3168 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3169 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3170 tls_dhparam take prime identifiers. Also unbreak combination of
3171 OpenSSL+DH_params+TLSSNI.
3173 PP/39 Disable SSLv2 by default in OpenSSL support.
3179 PP/01 Solaris build fix for Oracle's LDAP libraries.
3180 Bugzilla 1109, patch from Stephen Usher.
3182 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3184 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3187 TF/02 Fix a couple more cases where we did not log the error message
3188 when unlink() failed. See also change 4.74-TF/03.
3190 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3191 lock up or crash if it happened to be inside a call to libc when it
3192 got a SIGUSR1 from exiwhat.
3194 The SIGUSR1 handler appends the current process status to the process
3195 log which is later printed by exiwhat. It used to use the general
3196 purpose logging code to do this, but several functions it calls are
3197 not safe for signals.
3199 The new output code in the SIGUSR1 handler is specific to the process
3200 log, and simple enough that it's easy to inspect for signal safety.
3201 Removing some special cases also simplifies the general logging code.
3202 Removing the spurious timestamps from the process log simplifies
3205 TF/04 Improved ratelimit ACL condition.
3207 The /noupdate option has been deprecated in favour of /readonly which
3208 has clearer semantics. The /leaky, /strict, and /readonly update modes
3209 are mutually exclusive. The update mode is no longer included in the
3210 database key; it just determines when the database is updated. (This
3211 means that when you upgrade Exim will forget old rate measurements.)
3213 Exim now checks that the per_* options are used with an update mode that
3214 makes sense for the current ACL. For example, when Exim is processing a
3215 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3216 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3217 must specify per_mail/readonly. If you omit the update mode it defaults to
3218 /leaky where that makes sense (as before) or /readonly where required.
3220 The /noupdate option is now undocumented but still supported for
3221 backwards compatibility. It is equivalent to /readonly except that in
3222 ACLs where /readonly is required you may specify /leaky/noupdate or
3223 /strict/noupdate which are treated the same as /readonly.
3225 A useful new feature is the /count= option. This is a generalization
3226 of the per_byte option, so that you can measure the throughput of other
3227 aggregate values. For example, the per_byte option is now equivalent
3228 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3230 The per_rcpt option has been generalized using the /count= mechanism
3231 (though it's more complicated than the per_byte equivalence). When it is
3232 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3233 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3234 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3235 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3236 Note that using per_rcpt with a non-readonly update mode in more than
3237 one ACL will cause the recipients to be double-counted. (The per_mail
3238 and per_byte options don't have this problem.)
3240 The handling of very low rates has changed slightly. If the computed rate
3241 is less than the event's count (usually one) then this event is the first
3242 after a long gap. In this case the rate is set to the same as this event's
3243 count, so that the first message of a spam run is counted properly.
3245 The major new feature is a mechanism for counting the rate of unique
3246 events. The new per_addr option counts the number of different
3247 recipients that someone has sent messages to in the last time period. It
3248 behaves like per_rcpt if all the recipient addresses are different, but
3249 duplicate recipient addresses do not increase the measured rate. Like
3250 the /count= option this is a general mechanism, so the per_addr option
3251 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3252 example, measure the rate that a client uses different sender addresses
3253 with the options per_mail/unique=$sender_address. There are further
3254 details in the main documentation.
3256 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3258 TF/06 Removed a few PCRE remnants.
3260 TF/07 Automatically extract Exim's version number from tags in the git
3261 repository when doing development or release builds.
3263 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3264 Bugzilla 879. Patch from Paul Fisher.
3266 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3267 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3270 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3272 PP/05 Variable $av_failed, true if the AV scanner deferred.
3273 Bugzilla 1078. Patch from John Horne.
3275 PP/06 Stop make process more reliably on build failure.
3276 Bugzilla 1087. Patch from Heiko Schlittermann.
3278 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3279 Bugzilla 1089. Patch from Heiko Schlittermann.
3281 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3282 Bugzilla 1131. Patch from Holger Weiß.
3284 PP/09 Handle IPv6 addresses with SPF.
3285 Bugzilla 860. Patch from Wolfgang Breyha.
3287 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3289 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3292 PP/11 match_* no longer expand right-hand-side by default.
3293 New compile-time build option, EXPAND_LISTMATCH_RHS.
3294 New expansion conditions, "inlist", "inlisti".
3296 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3298 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3300 PP/14 fix log_write() format string regression from TF/03.
3301 Bugzilla 1152. Patch from Dmitry Isaikin.
3307 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3309 PP/02 Harmonised TLS library version reporting; only show if debugging.
3310 Layout now matches that introduced for other libraries in 4.74 PP/03.
3312 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3314 PP/04 New "dns_use_edns0" global option.
3316 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3319 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3320 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3322 TK/01 Updated PolarSSL code to 0.14.2.
3323 Bugzilla 1097. Patch from Andreas Metzler.
3325 PP/07 Catch divide-by-zero in ${eval:...}.
3326 Fixes bugzilla 1102.
3328 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3331 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3332 format-string attack -- SECURITY: remote arbitrary code execution.
3334 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3335 time unintentionally subject to list matching rules, letting the header
3336 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3337 arbitrary string expansion). This allowed for information disclosure.
3339 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3340 INT_MIN/-1 -- value coerced to INT_MAX.
3346 NM/01 Workaround for PCRE version dependency in version reporting
3349 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3350 This fixes portability to compilers other than gcc, notably
3351 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3353 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3354 makefiles for portability to HP-UX and POSIX correctness.
3356 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3357 Also via indented variable definition in the Makefile.
3358 (Debugging by Oliver Heesakkers).
3360 PP/02 Restore caching of spamd results with expanded spamd_address.
3361 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3363 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3364 Improves build reliability. Fix from: Frank Elsner
3366 NM/02 Fix wide character breakage in the rfc2047 coding
3367 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3369 NM/03 Allow underscore in dnslist lookups
3370 Fixes bug 1026. Patch from Graeme Fowler
3372 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3373 Code patches from Adam Ciarcinski of NetBSD.
3375 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3378 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3379 is logged, to avoid truncation. Patch from John Horne.
3381 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3382 Patch from Jakob Hirsch.
3384 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3385 SQL string expansion failure details.
3386 Patch from Andrey Oktyabrski.
3388 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3389 Patch from Simon Arlott.
3391 PP/09 New lookups functionality failed to compile on old gcc which rejects
3392 extern declarations in function scope.
3393 Patch from Oliver Fleischmann
3395 PP/10 Use sig_atomic_t for flags set from signal handlers.
3396 Check getgroups() return and improve debugging.
3397 Fixed developed for diagnosis in bug 927 (which turned out to be
3400 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3401 Patch from Mark Zealey.
3403 PP/12 Bugzilla 1056: Improved spamd server selection.
3404 Patch from Mark Zealey.
3406 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3407 Based on patch from Heiko Schlittermann.
3409 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3410 Patch from Uwe Doering, sign-off by Michael Haardt.
3412 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3413 variable declaration deep within a block. Bug and patch from
3416 PP/15 lookups-Makefile IRIX compatibility coercion.
3418 PP/16 Make DISABLE_DKIM build knob functional.
3420 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3421 Patch by Simon Arlott
3423 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3424 variable argument macros. Our copy now differs from upstream.
3430 TF/01 Failure to get a lock on a hints database can have serious
3431 consequences so log it to the panic log.
3433 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3434 controlled using the smtp_confirmation log selector.
3436 TF/03 Include the error message when we fail to unlink a spool file.
3438 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3439 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3440 for maintaining out-of-tree patches for some time.
3442 PP/01 Bugzilla 139: Documentation and portability issues.
3443 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3444 Handle per-OS dynamic-module compilation flags.
3446 PP/02 Let /dev/null have normal permissions.
3447 The 4.73 fixes were a little too stringent and complained about the
3448 permissions on /dev/null. Exempt it from some checks.
3449 Reported by Andreas M. Kirchwitz.
3451 PP/03 Report version information for many libraries, including
3452 Exim version information for dynamically loaded libraries. Created
3453 version.h, now support a version extension string for distributors
3454 who patch heavily. Dynamic module ABI change.
3456 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3457 privilege escalation vulnerability whereby the Exim run-time user
3458 can cause root to append content of the attacker's choosing to
3461 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3464 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3465 If dropping privileges for untrusted macros, we disabled normal logging
3466 on the basis that it would fail; for the Exim run-time user, this is not
3467 the case, and it resulted in successful deliveries going unlogged.
3468 Fixed. Reported by Andreas Metzler.
3474 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3475 only prepend for the Resent-* case. Fixes regression introduced in
3476 Exim 4.70 by NM/22 for Bugzilla 607.
3478 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3479 increasing numbers of administrators be bitten by this.
3481 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3483 PP/03 Bugzilla 994: added openssl_options main configuration option.
3485 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3487 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3489 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3491 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3492 without a peer certificate, leading to a segfault because of an
3493 assumption that peers always have certificates. Be a little more
3494 paranoid. Problem reported by Martin Tscholak.
3496 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3497 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3498 NB: ClamAV planning to remove STREAM in "middle of 2010".
3499 CL also introduces -bmalware, various -d+acl logging additions and
3500 more caution in buffer sizes.
3502 PP/09 Implemented reverse_ip expansion operator.
3504 PP/10 Bugzilla 937: provide a "debug" ACL control.
3506 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3508 PP/12 Bugzilla 973: Implement --version.
3510 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3512 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3514 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3516 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3517 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3518 ignore trailing whitespace.
3520 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3522 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3525 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3526 Notification from Dr Andrew Aitchison.
3528 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3529 ExtendedDetectionInfo response format.
3530 Notification from John Horne.
3532 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3535 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3536 XSL and documented dependency on system catalogs, with examples of how
3539 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3542 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3543 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3544 configuration file which is writeable by the Exim user or group.
3546 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3547 of configuration files to cover files specified with the -C option if
3548 they are going to be used with root privileges, not just the default
3551 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3552 option (effectively making it always true).
3554 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3555 files to be used while preserving root privileges.
3557 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3558 that rogue child processes cannot use them.
3560 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3561 run-time user, instead of root.
3563 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3564 Exim run-time user without dropping privileges.
3566 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3567 result string, instead of calling string_vformat() twice with the same
3570 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3571 for other users. Others should always drop root privileges if they use
3572 -C on the command line, even for a whitelisted configure file.
3574 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3576 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3582 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3583 $data_path, and $header_path variables; fixed documentation bugs and
3586 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3587 exipick to access non-standard spools, including the "frozen" queue
3590 NM/01 Bugzilla 965: Support mysql stored procedures.
3591 Patch from Alain Williams
3593 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3595 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3596 Patch from Andreas Metzler
3598 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3599 Patch from Kirill Miazine
3601 NM/05 Bugzilla 671: Added umask to procmail example.
3603 JJ/03 installed exipick 20100323.0, fixing doc bug
3605 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3606 directory. Notification and patch from Dan Rosenberg.
3608 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3610 TK/02 Improve log output when DKIM signing operation fails.
3612 MH/01 Treat the transport option dkim_domain as a colon separated
3613 list, not as a single string, and sign the message with each element,
3614 omitting multiple occurences of the same signer.
3616 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3617 Bugzilla 985, 986. Patch by Simon Arlott
3619 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3620 Patch by Simon Arlott
3622 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3623 MBX locking. Notification from Dan Rosenberg.
3629 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3631 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3633 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3635 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3637 NM/04 Bugzilla 915: Flag broken perl installation during build.
3643 TK/01 Added patch by Johannes Berg that expands the main option
3644 "spamd_address" if it starts with a dollar sign.
3646 TK/02 Write list of recipients to X-Envelope-Sender header when building
3647 the mbox-format spool file for content scanning (suggested by Jakob
3650 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3651 (http://www.dcc-servers.net/) support via dccifd. Activated by
3652 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3654 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3655 by Mark Daniel Reidel <mr@df.eu>.
3657 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3658 When building exim an external PCRE library is now needed -
3659 PCRE is a system library on the majority of modern systems.
3660 See entry on PCRE_LIBS in EDITME file.
3662 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3663 conversation. Added nologin parameter to request.
3664 Patch contributed by Kirill Miazine.
3666 TF/01 Do not log submission mode rewrites if they do not change the address.
3668 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3670 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3671 log files in place. Contributed by Roberto Lima.
3673 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3675 TF/03 Bugzilla 615: When checking the local_parts router precondition
3676 after a local_part_suffix or local_part_prefix option, Exim now
3677 does not use the address's named list lookup cache, since this
3678 contains cached lookups for the whole local part.
3680 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3681 Robert Millan. Documentation is in experimental-spec.txt.
3683 TF/04 Bugzilla 668: Fix parallel build (make -j).
3685 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3687 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3688 Patch provided by Jan Srzednicki.
3690 TF/05 Leading white space used to be stripped from $spam_report which
3691 wrecked the formatting. Now it is preserved.
3693 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3694 that they are available at delivery time.
3696 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3698 TF/08 TLS error reporting now respects the incoming_interface and
3699 incoming_port log selectors.
3701 TF/09 Produce a more useful error message if an SMTP transport's hosts
3702 setting expands to an empty string.
3704 NM/06 Bugzilla 744: EXPN did not work under TLS.
3705 Patch provided by Phil Pennock.
3707 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3708 Patch provided by Richard Godbee.
3710 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3711 acl_smtp_notquit, added index entry.
3713 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3714 Patch provided by Eugene Bujak.
3716 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3717 accept(). Patch provided by Maxim Dounin.
3719 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3720 Patch provided by Phil Pennock.
3722 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3724 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3725 Patch provided by Brad "anomie" Jorsch.
3727 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3728 Patch provided by Dean Brooks.
3730 TK/05 Add native DKIM support (does not depend on external libraries).
3732 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3733 Patch provided by Graeme Fowler.
3735 NM/16 Bugzilla 851: Documentation example syntax fix.
3737 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3739 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3742 NM/19 Bugzilla 745: TLS version reporting.
3743 Patch provided by Phil Pennock.
3745 NM/20 Bugzilla 167: bool: condition support.
3746 Patch provided by Phil Pennock.
3748 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3749 clients. Patch provided by Phil Pennock.
3751 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3752 Patch provided by Brad "anomie" Jorsch.
3754 NM/23 Bugzilla 687: Fix misparses in eximstats.
3755 Patch provided by Heiko Schlittermann.
3757 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3758 Patch provided by Heiko Schlittermann.
3760 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3761 plus update to original patch.
3763 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3765 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3766 Patch provided by David Brownlee.
3768 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3770 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3772 NM/30 Bugzilla 888: TLS documentation bugfixes.
3774 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3776 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3777 Unlike the original bugzilla I have changed all shell scripts in src tree.
3779 NM/33 Bugzilla 898: Transport filter timeout fix.
3780 Patch by Todd Rinaldo.
3782 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3783 Patch by Serge Demonchaux.
3785 NM/35 Bugzilla 39: Base64 decode bug fixes.
3786 Patch by Jakob Hirsch.
3788 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3790 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3792 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3794 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3800 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3801 ALT-N's libdkim that I have put here:
3802 http://duncanthrax.net/exim-experimental/
3804 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3805 were called 'true' and it seems that C99 defines that as a reserved
3806 keyword to be used with 'bool' variable types. That means you could
3807 not include C99-style headers which use bools without triggering
3808 build errors in sieve.c.
3810 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3811 as mailq or other aliases. Changed the --help handling significantly
3812 to do whats expected. exim_usage() emits usage/help information.
3814 SC/01 Added the -bylocaldomain option to eximstats.
3816 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3818 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3820 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3826 PH/01 Another patch from the Sieve maintainer.
3828 PH/02 When an IPv6 address is converted to a string for single-key lookup
3829 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3830 dots are used instead of colons so that keys in lsearch files need not
3831 contain colons. This was done some time before quoting was made available
3832 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3833 (notated using the quote facility) so as to distinguish them from IPv4
3834 keys. This meant that lookups for IP addresses in host lists did not work
3835 for iplsearch lookups.
3837 This has been fixed by arranging for IPv6 addresses to be expressed with
3838 colons if the lookup type is iplsearch. This is not incompatible, because
3839 previously such lookups could never work.
3841 The situation is now rather anomalous, since one *can* have colons in
3842 ordinary lsearch keys. However, making the change in all cases is
3843 incompatible and would probably break a number of configurations.
3845 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3848 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3849 conversion specification without a maximum field width, thereby enabling
3850 a rogue spamd server to cause a buffer overflow. While nobody in their
3851 right mind would setup Exim to query an untrusted spamd server, an
3852 attacker that gains access to a server running spamd could potentially
3853 exploit this vulnerability to run arbitrary code as the Exim user.
3855 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3856 $primary_hostname instead of what libspf2 thinks the hosts name is.
3858 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3859 a directory entry by the name of the lookup key. Previously, if a
3860 symlink pointed to a non-existing file or a file in a directory that
3861 Exim lacked permissions to read, a lookup for a key matching that
3862 symlink would fail. Now it is enough that a matching directory entry
3863 exists, symlink or not. (Bugzilla 503.)
3865 PH/03 The body_linecount and body_zerocount variables are now exported in the
3868 PH/04 Added the $dnslist_matched variable.
3870 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3871 This means they are set thereafter only if the connection becomes
3874 PH/06 Added the client_condition to authenticators so that some can be skipped
3875 by clients under certain conditions.
3877 PH/07 The error message for a badly-placed control=no_multiline_responses left
3878 "_responses" off the end of the name.
3880 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3882 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3883 (without spaces) instead of just copying the configuration text.
3885 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3887 PH/11 Added $max_received_linelength.
3889 PH/12 Added +ignore_defer and +include_defer to host lists.
3891 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3892 way in which PCRE > 7.0 is built.
3894 PH/14 Implemented queue_only_load_latch.
3896 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3897 MAIL command. The effect was to mangle the value on 64-bit systems.
3899 PH/16 Another patch from the Sieve maintainer.
3901 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3903 PH/18 If a system quota error occurred while trying to create the file for
3904 a maildir delivery, the message "Mailbox is full" was not appended to the
3905 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3906 only to a quota excession during the actual writing of the file.
3908 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3909 characters?) which causes problems in log lines. The DN values are now
3910 passed through string_printing() before being added to log lines.
3912 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3913 and InterBase are left for another time.)
3915 PH/21 Added message_body_newlines option.
3917 PH/22 Guard against possible overflow in moan_check_errorcopy().
3919 PH/23 POSIX allows open() to be a macro; guard against that.
3921 PH/24 If the recipient of an error message contained an @ in the local part
3922 (suitably quoted, of course), incorrect values were put in $domain and
3923 $local_part during the evaluation of errors_copy.
3929 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3930 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3933 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3934 issue a MAIL command.
3936 PH/02 In an ACL statement such as
3938 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3940 if a client was not listed at all, or was listed with a value other than
3941 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3942 the condition was not true (as it should be), so access was not denied.
3943 The bug was that the ! inversion was incorrectly passed on to the second
3944 item. This has been fixed.
3946 PH/03 Added additional dnslists conditions == and =& which are different from
3947 = and & when the dns lookup returns more than one IP address.
3949 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3950 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3952 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3953 FSYNC, which compiles an option called disable_fsync that allows for
3954 bypassing fsync(). The documentation is heavily laced with warnings.
3956 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3958 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3959 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3960 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3961 including adding "make clean"; (3) Added -fPIC when compiling the test
3962 dynamically loaded module, to get rid of a warning.
3964 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3965 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3966 = 0s. The bug is otherwise harmless.
3968 PH/07 There was a bug in the dovecot authenticator such that the value of
3969 $auth1 could be overwritten, and so not correctly preserved, after a
3970 successful authentication. This usually meant that the value preserved by
3971 the server_setid option was incorrect.
3973 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3975 PH/09 Installed PCRE release 7.0.
3977 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3978 run for batched SMTP input. It is now run at the start of every message
3979 in the batch. While fixing this I discovered that the process information
3980 (output by running exiwhat) was not always getting set for -bs and -bS
3981 input. This is fixed, and it now also says "batched" for BSMTP.
3983 PH/11 Added control=no_pipelining.
3985 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3986 patch, slightly modified), and move the expansion of helo_data till after
3987 the connection is made in the smtp transport (so it can use these
3990 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3992 PH/14 Added log_selector = +pid.
3994 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3996 PH/16 Add ${if forany and ${if forall.
3998 PH/17 Added dsn_from option to vary the From: line in DSNs.
4000 PH/18 Flush SMTP output before performing a callout, unless control =
4001 no_callout_flush is set.
4003 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
4004 was true (the default) a successful delivery failed to delete the retry
4005 item, thus causing premature timeout of the address. The bug is now
4008 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4010 PH/21 Long custom messages for fakedefer and fakereject are now split up
4011 into multiline responses in the same way that messages for "deny" and
4012 other ACL rejections are.
4014 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4015 with slight modification.
4017 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4018 draft, changing the syntax and factoring some duplicate code".
4020 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4021 for deliveries of the second and subsequent messages over the same SMTP
4024 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4025 ${reduce, with only minor "tidies".
4027 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4029 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4030 expansion side effects.
4032 PH/27 When a message times out after an over-quota error from an Exim-imposed
4033 quota, the bounce message says "mailbox is full". This message was not
4034 being given when it was a system quota that was exceeded. It now should
4037 MH/03 Made $recipients available in local_scan(). local_scan() already has
4038 better access to the recipient list through recipients_list[], but
4039 $recipients can be useful in postmaster-provided expansion strings.
4041 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4042 in the case of a MAIL command with additional options following the
4043 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4044 were accidentally chopped off.
4046 PH/29 SMTP synchronization checks are implemented when a command is read -
4047 there is a check that no more input is waiting when there shouldn't be
4048 any. However, for some commands, a delay in an ACL can mean that it is
4049 some time before the response is written. In this time, more input might
4050 arrive, invalidly. So now there are extra checks after an ACL has run for
4051 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4052 pipelining has not been advertised.
4054 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4056 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4057 returned in the SMTP response when smtp_return_error_details was false.
4058 This has been fixed.
4060 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4061 instead of the C I/O that was originally supplied, because problems were
4062 reported on Solaris.
4064 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4065 Exim which did not show up earlier: it was assuming that a call to
4066 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4067 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4068 was a macro that became an assignment, so it seemed to work. This has
4069 changed to a proper function call with a void return, hence the compile
4070 error. Exim's code has been fixed.
4072 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4075 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4077 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4079 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4080 because it turns out to be too big or there is a timeout) there is no
4081 "Completed" line in the log. When some messages of this type were
4082 selected by exigrep, they were listed as "not completed". Others were
4083 picked up by some special patterns. I have improved the selection
4084 criteria to be more general.
4086 PH/38 The host_find_failed option in the manualroute router can now be set
4087 to "ignore", to completely ignore a host whose IP address cannot be
4088 found. If all hosts are ignored, the behaviour is controlled by the new
4089 host_all_ignored option.
4091 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4092 homing or because of multiple MX records with /mx) generated more than
4093 one IP address, and the following item turned out to be the local host,
4094 all the secondary addresses of the first item were incorrectly removed
4095 from the list, along with the local host and any following hosts (which
4096 is what is supposed to happen).
4098 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4099 whoever called Exim into the -H file. In the case of the daemon it was
4100 behaving confusingly. When first started, it used values for whoever
4101 started the daemon, but after a SIGHUP it used the Exim user (because it
4102 calls itself on a restart). I have changed the code so that it now always
4105 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4106 message are rejected with the same error (e.g. no authentication or bad
4107 sender address), and a DATA command is nevertheless sent (as can happen
4108 with PIPELINING or a stupid MUA), the error message that was given to the
4109 RCPT commands is included in the rejection of the DATA command. This is
4110 intended to be helpful for MUAs that show only the final error to their
4113 PH/42 Another patch from the Sieve maintainer.
4115 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4116 Eximstats - Fixed some broken HTML links and added missing column headers
4118 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4119 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4121 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4123 PH/43 Yet another patch from the Sieve maintainer.
4125 PH/44 I found a way to check for a TCP/IP connection going away before sending
4126 the response to the final '.' that terminates a message, but only in the
4127 case where the client has not sent further data following the '.'
4128 (unfortunately, this is allowed). However, in many cases there won't be
4129 any further data because there won't be any more messages to send. A call
4130 to select() can be used: if it shows that the input is "ready", there is
4131 either input waiting, or the socket has been closed. An attempt to read
4132 the next input character can distinguish the two cases. Previously, Exim
4133 would have sent an OK response which the client would never have see.
4134 This could lead to message repetition. This fix should cure that, at
4135 least in a lot of common cases.
4137 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4138 advertised in response to EHLO.
4144 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4145 fixed by 4.65/MH/01 (is this a record?) are fixed:
4147 (i) An empty string was always treated as zero by the numeric comparison
4148 operators. This behaviour has been restored.
4150 (ii) It is documented that the numeric comparison operators always treat
4151 their arguments as decimal numbers. This was broken in that numbers
4152 starting with 0 were being interpreted as octal.
4154 While fixing these problems I realized that there was another issue that
4155 hadn't been noticed. Values of message_size_limit (both the global option
4156 and the transport option) were treated as octal if they started with 0.
4157 The documentation was vague. These values are now always treated as
4158 decimal, and I will make that clear in the documentation.
4164 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4165 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4168 MH/01 Don't check that the operands of numeric comparison operators are
4169 integers when their expansion is in "skipping" mode (fixes bug
4170 introduced by 4.64-PH/07).
4172 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4173 child addresses, Exim now panics and dies. Previously, because the count
4174 is held in a short int, deliveries were likely to be lost. As such a
4175 large number of recipients for a single message is ridiculous
4176 (performance will be very, very poor), I have chosen to impose a limit
4177 rather than extend the field.
4183 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4184 leftover -K file (the existence of which was triggered by #402).
4185 While we were at it, introduced process PID as part of the -K
4186 filename. This should rule out race conditions when creating
4189 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4190 processing considerably. Previous code took too long for large mails,
4191 triggering a timeout which in turn triggers #401.
4193 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4194 in the DK code in transports.c. sendfile() is not really portable,
4195 hence the _LINUX specificness.
4197 TF/01 In the add_headers option to the mail command in an Exim filter,
4198 there was a bug that Exim would claim a syntax error in any
4199 header after the first one which had an odd number of characters
4202 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4203 callout verification, Exim cached a "reject" for the entire domain. This
4204 is correct for most verifications, but it is not correct for a recipient
4205 verification with use_sender or use_postmaster set, because in that case
4206 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4207 case of MAIL FROM:<> rejection from other early rejections (e.g.
4208 rejection of HELO). When verifying a recipient using a non-null MAIL
4209 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4210 Whatever the result of the callout, the value of the domain cache is
4211 left unchanged (for any other kind of callout, getting as far as trying
4212 RCPT means that the domain itself is ok).
4214 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4217 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4218 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4219 session, the daemon ignores ECONNECT errors and logs others; it now
4220 ignores EPIPE as well.
4222 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4223 (quoted-printable decoding).
4225 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4226 later the small subsequent patch to fix an introduced bug.
4228 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4230 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4232 PH/08 An error is now given if message_size_limit is specified negative.
4234 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4235 to be given (somewhat) arbitrary names.
4237 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4240 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4241 miscellaneous code fixes
4243 PH/10 Added the log_reject_target ACL modifier to specify where to log
4246 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4247 hostname. This is wrong, because it relates to the incoming message (and
4248 probably the interface on which it is arriving) and not to the outgoing
4249 callout (which could be using a different interface). This has been
4250 changed to use the value of the helo_data option from the smtp transport
4251 instead - this is what is used when a message is actually being sent. If
4252 there is no remote transport (possible with a router that sets up host
4253 addresses), $smtp_active_hostname is used.
4255 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4256 tweaks were necessary in order to get it to work (see also 21 below):
4257 (a) The code assumed that strncpy() returns a negative number on buffer
4258 overflow, which isn't the case. Replaced with Exim's string_format()
4260 (b) There were several signed/unsigned issues. I just did the minimum
4261 hacking in of casts. There is scope for a larger refactoring.
4262 (c) The code used strcasecmp() which is not a standard C function.
4263 Replaced with Exim's strcmpic() function.
4264 (d) The code set only $1; it now sets $auth1 as well.
4265 (e) A simple test gave the error "authentication client didn't specify
4266 service in request". It would seem that Dovecot has changed its
4267 interface. Fortunately there's a specification; I followed it and
4268 changed what the client sends and it appears to be working now.
4270 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4273 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4274 address A is aliased to B and C, where B exists and C does not. Without
4275 -v the output is "A verified" because verification stops after a
4276 successful redirection if more than one address is generated. However,
4277 with -v the child addresses are also verified. Exim was outputting "A
4278 failed to verify" and then showing the successful verification for C,
4279 with its parentage. It now outputs "B failed to verify", showing B's
4280 parentage before showing the successful verification of C.
4282 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4283 look up a TXT record in a specific list after matching in a combined
4286 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4287 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4288 they consult the DNS. I had assumed they would set it the way they
4289 wanted; and indeed my experiments on Linux seem to show that in some
4290 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4291 To be on the safe side, however, I have now made the interface to
4292 host_find_byname() similar to host_find_bydns(), with an argument
4293 containing the DNS resolver options. The host_find_byname() function now
4294 sets these options at its start, just as host_find_bydns() does. The smtp
4295 transport options dns_qualify_single and dns_search_parents are passed to
4296 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4297 of host_find_byname() use the default settings of RES_DEFNAMES
4298 (qualify_single) but not RES_DNSRCH (search_parents).
4300 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4301 spool_read_header() do less string testing, by means of a preliminary
4302 switch on the second character of optional "-foo" lines. (This is
4303 overdue, caused by the large number of possibilities that now exist.
4304 Originally there were few.) While I was there, I also converted the
4305 str(n)cmp tests so they don't re-test the leading "-" and the first
4306 character, in the hope this might squeeze out yet more improvement.
4308 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4309 flag allowing group syntax was set by the header_syntax check but not
4310 turned off, possible causing trouble later; (2) The flag was not being
4311 set at all for the header_verify test, causing "group"-style headers to
4312 be rejected. I have now set it in this case, and also caused header_
4313 verify to ignore an empty address taken from a group. While doing this, I
4314 came across some other cases where the code for allowing group syntax
4315 while scanning a header line wasn't quite right (mostly, not resetting
4316 the flag correctly in the right place). These bugs could have caused
4317 trouble for malformed header lines. I hope it is now all correct.
4319 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4320 with the "reply" argument non-NULL. The code, however (which originally
4321 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4322 but it didn't always do it. This confused somebody who was copying the
4323 code for some other use. I have removed all the tests.
4325 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4326 feature that was used to support insecure browsers during the U.S. crypto
4327 embargo. It requires special client support, and Exim is probably the
4328 only MTA that supported it -- and would never use it because real RSA is
4329 always available. This code has been removed, because it had the bad
4330 effect of slowing Exim down by computing (never used) parameters for the
4331 RSA_EXPORT functionality.
4333 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4334 authenticator to fail if there's a tab character in the incoming data
4335 (there should never be unless someone is messing about, as it's supposed
4336 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4337 if the connection is using TLS or if the remote IP is the same as the
4338 local IP, and the "valid-client-cert option" if a client certificate has
4341 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4342 authenticators. This can be used for authorization after authentication
4343 succeeds. (In the case of plaintext, it servers for both authentication
4346 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4347 if any retry times were supplied.
4349 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4350 connection, where there is no client IP address to check. In this
4351 situation, the verify now always succeeds.
4353 PH/25 Applied John Jetmore's -Mset patch.
4355 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4357 PH/27 In a string expansion for a processed (not raw) header when multiple
4358 headers of the same name were present, leading whitespace was being
4359 removed from all of them, but trailing whitespace was being removed only
4360 from the last one. Now trailing whitespace is removed from each header
4361 before concatenation. Completely empty headers in a concatenation (as
4362 before) are ignored.
4364 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4365 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4367 PH/29 [Removed. This was a change that I later backed out, and forgot to
4368 correct the ChangeLog entry (that I had efficiently created) before
4369 committing the later change.]
4371 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4372 address errors (4xx response to RCPT) over the same connection as other
4373 messages routed to the same hosts. Such deliveries are always "forced",
4374 so retry times are not inspected. This resulted in far too many retries
4375 for the affected addresses. The effect occurred only when there were more
4376 hosts than the hosts_max_try setting in the smtp transport when it had
4377 the 4xx errors. Those hosts that it had tried were not added to the list
4378 of hosts for which the message was waiting, so if all were tried, there
4379 was no problem. Two fixes have been applied:
4381 (i) If there are any address or message errors in an SMTP delivery, none
4382 of the hosts (tried or untried) are now added to the list of hosts
4383 for which the message is waiting, so the message should not be a
4384 candidate for sending over the same connection that was used for a
4385 successful delivery of some other message. This seems entirely
4386 reasonable: after all the message is NOT "waiting for some host".
4387 This is so "obvious" that I'm not sure why it wasn't done
4388 previously. Hope I haven't missed anything, but it can't do any
4389 harm, as the worst effect is to miss an optimization.
4391 (ii) If, despite (i), such a delivery is accidentally attempted, the
4392 routing retry time is respected, so at least it doesn't keep
4393 hammering the server.
4395 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4396 in ${readsocket because some servers need this prod.
4398 PH/32 Added some extra debug output when updating a wait-xxx database.
4400 PH/33 The hint "could be header name not terminated by colon", which has been
4401 given for certain expansion errors for a long time, was not being given
4402 for the ${if def:h_colon_omitted{... case.
4404 PH/34 The spec says: "With one important exception, whenever a domain list is
4405 being scanned, $domain contains the subject domain." There was at least
4406 one case where this was not true.
4408 PH/35 The error "getsockname() failed: connection reset by peer" was being
4409 written to the panic log as well as the main log, but it isn't really
4410 panic-worthy as it just means the connection died rather early on. I have
4411 removed the panic log writing for the ECONNRESET error when getsockname()
4414 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4415 runs only) independently of the message's sender address. This meant
4416 that, if the 4xx error was in fact related to the sender, a different
4417 message to the same recipient with a different sender could confuse
4418 things. In particular, this can happen when sending to a greylisting
4419 server, but other circumstances could also provoke similar problems.
4420 I have changed the default so that the retry time for these errors is now
4421 based a combination of the sender and recipient addresses. This change
4422 can be overridden by setting address_retry_include_sender=false in the
4425 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4426 remote server are returned as part of bounce messages. This was not
4427 happening for LMTP over a pipe (the lmtp transport), but now it is the
4428 same for both kinds of LMTP.
4430 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4431 in header lines that were in fact CNAMEs. This is no longer the case.
4433 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4434 and queue runs started by the daemon processed all messages. This has
4435 been fixed so that -R and -S can now usefully be given with -q<time>.
4437 PH/40 Import PCRE release 6.7 (fixes some bugs).
4439 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4441 PH/42 Give an error if -q is specified more than once.
4443 PH/43 Renamed the variables $interface_address and $interface_port as
4444 $received_ip_address and $received_port, to make it clear that these
4445 values apply to message reception, and not to the outgoing interface when
4446 a message is delivered. (The old names remain recognized, of course.)
4448 PH/44 There was no timeout on the connect() call when using a Unix domain
4449 socket in the ${readsocket expansion. There now is.
4451 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4452 be meaningful with "accept".
4454 SC/01 Eximstats V1.43
4455 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4457 SC/02 Eximstats V1.44
4458 Use a glob alias rather than an array ref in the generated
4459 parser. This improves both readability and performance.
4461 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4462 Collect SpamAssassin and rejection statistics.
4463 Don't display local sender or destination tables unless
4464 there is data to show.
4465 Added average volumes into the top table text output.
4467 SC/04 Eximstats V1.46
4468 Collect data on the number of addresses (recipients)
4469 as well as the number of messages.
4471 SC/05 Eximstats V1.47
4472 Added 'Message too big' to the list of mail rejection
4473 reasons (thanks to Marco Gaiarin).
4475 SC/06 Eximstats V1.48
4476 Mainlog lines which have GMT offsets and are too short to
4477 have a flag are now skipped.
4479 SC/07 Eximstats V1.49 (Alain Williams)
4480 Added the -emptyok flag.
4482 SC/08 Eximstats V1.50
4483 Fixes for obtaining the IP address from reject messages.
4485 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4486 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4487 whitespace changes from 4.64-PH/27
4489 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4492 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4493 are found, allow negative numbers in numeric criteria)
4495 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4497 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4498 to match changes made in 4.64-PH/43
4500 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4502 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4503 "domain" is actually an IP address, and force a failure. This locks out
4504 those revolvers/nameservers that support "A-for-A" lookups, in
4505 contravention of the specifications.
4507 PH/48 When a host name was looked up from an IP address, and the subsequent
4508 forward lookup of the name timed out, the host name was left in
4509 $sender_host_name, contrary to the specification.
4511 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4512 restricted to single-key lookups, Exim was not diagnosing an error if
4513 * or *@ was used with a query-style lookup.
4515 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4517 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4518 long ago, but noone interested enough thought of it. Let's just say that
4519 the "1.1" means that there are some new functions that weren't there at
4520 some point in the past.
4522 PH/51 Error processing for expansion failure of helo_data from an smtp
4523 transport during callout processing was broken.
4525 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4526 tested/used via the -bh/-bhc/-bs options.
4528 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4529 bug, fixed in subsequent PCRE releases).
4531 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4532 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4534 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4540 SC/01 Use a glob alias rather than an array ref in eximstats generated
4541 parser. This improves both readability and performance.
4543 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4544 Don't display local sender or destination tables in eximstats unless
4545 there is data to show.
4546 Added average volumes into the eximstats top table text output.
4548 SC/03 Collect data on the number of addresses (recipients) as well
4549 as the number of messages in eximstats.
4551 TF/01 Correct an error in the documentation for the redirect router. Exim
4552 does (usually) call initgroups() when daemonizing.
4554 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4555 with consistent privilege compared to when running as a daemon.
4557 TF/03 Note in the spec that $authenticated_id is not set for local
4558 submissions from trusted users.
4560 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4561 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4563 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4564 by adding some example configuration directives to the default
4565 configuration file. A little bit of work is required to uncomment the
4566 directives and define how usernames and passwords are checked, but
4567 there is now a framework to start from.
4569 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4570 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4571 without this. I don't know how relevant this is to other LDAP libraries.
4573 PH/02 Add the verb name to the "unknown ACL verb" error.
4575 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4577 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4579 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4580 directory not expanded when it should be if an expanded home directory
4581 was set for the address (which is overridden by the transport).
4583 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4586 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4587 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4588 because it is too late at that time, and has no effect.
4590 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4591 security issue with \' (bugzilla #107). I could not use the
4592 PQescapeStringConn() function, because it needs a PGconn value as one of
4595 PH/08 When testing addresses using -bt, indicate those final addresses that
4596 are duplicates that would not cause an additional delivery. At least one
4597 person was confused, thinking that -bt output corresponded to deliveries.
4598 (Suppressing duplicates isn't a good idea as you lose the information
4599 about possibly different redirections that led to the duplicates.)
4601 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4602 systems where poll() doesn't work, in particular OS X.
4604 PH/10 Added more information to debugging output for retry time not reached.
4606 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4607 operations in malware.c.
4609 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4612 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4613 syslog_duplication set false, log lines that would normally be written
4614 both the the main log and to the reject log were not written to syslog at
4617 PH/14 In the default configuration, change the use of "message" in ACL warn
4618 statements to "add_header".
4620 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4621 not followed by a command (e.g. "seen endif").
4623 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4624 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4627 PH/17 Added extra conditions to the default value of delay_warning_condition
4631 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4632 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4633 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4636 The Auto-Submitted: and various List- headers are standardised, whereas I
4637 don't think Precedence: ever was.
4639 PH/18 Refactored debugging code in route_finduser() to show more information,
4640 in particular, the error code if getpwnam() issues one.
4642 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4643 This is apparently needed in addition to the PH/07 change above to avoid
4644 any possible encoding problems.
4646 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4647 but not after initializing Perl.
4649 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4650 output them only if debugging. By default they are written stderr,
4651 apparently, which is not desirable.
4653 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4656 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4659 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4661 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4662 authenticated or an ident call has been made. Suppress the default
4663 values for $authenticated_id and $authenticated_sender (but permit -oMai
4664 and -oMas) when testing with -bh.
4666 PH/24 Re-jigged the order of the tests in the default configuration so that the
4667 tests for valid domains and recipients precede the DNS black list and CSA
4668 tests, on the grounds that those ones are more expensive.
4670 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4671 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4672 This bug exists in every version of Exim that I still have, right back to
4675 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4676 However, an attempt to turn on case-sensitivity in a regex key by
4677 including (?-i) didn't work because the subject string was already
4678 lowercased, and the effects were non-intuitive. It turns out that a
4679 one-line patch can be used to allow (?-i) to work as expected.
4685 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4686 other effects) broke the use of negated acl sub-conditions.
4688 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4691 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4692 "Deny" causes Exim to reject the incoming connection with a 554 error.
4693 Unfortunately, if there is a major crisis, such as a disk failure,
4694 tcp-wrappers gives "deny", whereas what one would like would be some
4695 kind of temporary error. A kludge has been added to help with this.
4696 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4697 554 error is used if errno is still zero or contains ENOENT (which occurs
4698 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4701 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4703 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4704 errors. Otherwise a message that provokes a temporary error (when other
4705 messages do not) can cause a whole host to time out.
4707 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4708 addresses were routed directly to files or pipes from a redirect router.
4709 File deliveries just didn't batch; pipe deliveries might have suffered
4712 PH/06 A failure to get a lock for a hints database would erroneously always say
4713 "Failed to get write lock", even when it was really a read lock.
4715 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4716 of 0600. This has been changed to use the value of the lockfile_mode
4717 option (which defaults to 0600).
4719 PH/08 Applied small patch from the Sieve maintainer.
4721 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4722 folder from quota calculations, a direct delivery into this folder messed
4723 up the contents of the maildirsize file. This was because the regex was
4724 used only to exclude .Trash (or whatever) when the size of the mailbox
4725 was calculated. There was no check that a delivery was happening into an
4726 excluded directory. This bug has been fixed by ignoring all quota
4727 processing for deliveries into excluded directories.
4729 PH/10 Added the maildirfolder_create_regex option to appendfile.
4735 PH/01 The code for finding all the local interface addresses on a FreeBSD
4736 system running IPv6 was broken. This may well have applied to all BSD
4737 systems, as well as to others that have similar system calls. The broken
4738 code found IPv4 interfaces correctly, but gave incorrect values for the
4739 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4740 that it would not match correctly against @[] and not recognize the IPv6
4743 PH/02 The ipliteral router was not recognizing addresses of the form user@
4744 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4746 PH/03 Added disable_ipv6.
4748 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4749 decoded form, because it is most often used to construct To: headers
4750 lines in autoreplies, and the decoded form may well be syntactically
4751 invalid. However, $reply_address has leading white space removed, and all
4752 newlines turned into spaces so that the autoreply transport does not
4755 PH/05 If group was specified without a user on a router, and no group or user
4756 was specified on a transport, the group from the router was ignored.
4758 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4759 for visible compile-time settings that can be used to change these
4760 numbers, for those that want even more. Backwards compatibility with old
4761 spool files has been maintained. However, going back to a previous Exim
4762 release will lost any variables that are in spool files.
4764 PH/07 Two small changes when running in the test harness: increase delay when
4765 passing a TCP/IP connection to a new process, in case the original
4766 process has to generate a bounce, and remove special handling of
4767 127.0.0.2 (sic), which is no longer necessary.
4769 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4770 be the same on different OS.
4772 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4775 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4776 whether --show-vars was specified or not
4778 JJ/02 exipick: Added support for new ACL variable spool format introduced
4781 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4782 syntactically invalid From: or Reply-to: line, and a filter used this to
4783 generate an autoreply, and therefore failed to obtain an address for the
4784 autoreply, Exim could try to deliver to a non-existent relative file
4785 name, causing unrelated and misleading errors. What now happens is that
4786 it logs this as a hard delivery error, but does not attempt to create a
4789 PH/11 The exinext utility has a -C option for testing purposes, but although
4790 the given file was scanned by exinext itself; it wasn't being passed on
4791 when Exim was called.
4793 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4794 an end-of-file indication when reading a command response.
4796 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4797 compiled. In many other places in Exim, IPv6 addresses are always
4798 recognized, so I have changed this. It also means that IPv4 domain
4799 literals of the form [IPV4:n.n.n.n] are now always recognized.
4801 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4802 used if the router is not running as root, for example, when verifying at
4803 ACL time, or when using -bh. The debugging output from this situation was
4804 non-existent - all you got was a failure to exec. I have made two
4807 (a) Failures to set uid/gid, the current directory, or a process leader
4808 in a subprocess such as that created by queryprogram now generate
4809 suitable debugging output when -d is set.
4811 (b) The queryprogram router detects when it is not running as root,
4812 outputs suitable debugging information if -d is set, and then runs
4813 the subprocess without attempting to change uid/gid.
4815 PH/15 Minor change to Makefile for building test_host (undocumented testing
4818 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4819 additional section of a DNS packet that returns MX or SRV records.
4820 Instead, it always explicitly searches for A/AAAA records. This avoids
4821 major problems that occur when a DNS server includes only records of one
4822 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4823 fixed another bug: if SRV records were looked up and the corresponding
4824 address records were *not* found in the additional section, the port
4825 values from the SRV records were lost.
4827 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4828 using the correct key (the original address) when searching the retry
4829 rules in order to find which one to use for generating the retry hint.
4831 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4832 adding the default one. Similarly, if it contains a Reply-To: header, the
4833 errors_reply_to option, if set, is not used.
4835 PH/19 When calculating a retry time, Exim used to measure the "time since
4836 failure" by looking at the "first failed" field in the retry record. Now
4837 it does not use this if it is later than than the arrival time of the
4838 message. Instead it uses the arrival time. This makes for better
4839 behaviour in cases where some deliveries succeed, thus re-setting the
4840 "first failed" field. An example is a quota failure for a huge message
4841 when small messages continue to be delivered. Without this change, the
4842 "time since failure" will always be short, possible causing more frequent
4843 delivery attempts for the huge message than are intended.
4844 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4846 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4847 $1, $2, $3) because the numerical variables can be reset during some
4848 expansion items (e.g. "match"), thereby losing the authentication data.
4850 PH/21 Make -bV show the size of off_t variables so that the test suite can
4851 decide whether to run tests for quotas > 2G.
4853 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4854 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4855 filecount value is greater than 2G or if a quota value is greater than 2G
4856 on a system where the size of off_t is not greater than 4, a panic error
4859 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4860 never match. The debug and -bh output now contains an explicit error
4861 message indicating a malformed IPv4 address or mask.
4863 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4864 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4865 PH/23 above applies.
4867 PH/25 Do not write to syslog when running in the test harness. The only
4868 occasion when this arises is a failure to open the main or panic logs
4869 (for which there is an explicit test).
4871 PH/26 Added the /no_tell option to "control=freeze".
4873 PH/27 If a host name lookup failed very early in a connection, for example, if
4874 the IP address matched host_lookup and the reverse lookup yielded a name
4875 that did not have a forward lookup, an error message of the form "no IP
4876 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4877 could be logged. Now it outputs the IP address instead of "NULL".
4879 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4880 allows the sender and the authenticated sender to be set when
4881 submitting a message from within Exim. Since child_open_exim() is
4882 documented for local_scan(), the new function should be too.
4884 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4885 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4886 results in an empty string is now treated as unset.
4888 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4890 PH/31 Added sender_verify_fail logging option.
4892 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4893 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4894 tidied the source and removed it altogether.
4896 PH/33 When a queue run was abandoned because the load average was too high, a
4897 log line was always written; now it is written only if the queue_run log
4898 selector is set. In addition, the log line for abandonment now contains
4899 information about the queue run such as the pid. This is always present
4900 in "start" and "stop" lines but was omitted from the "abandon" line.
4902 PH/34 Omit spaces between a header name and the colon in the error message that
4903 is given when verify = headers_syntax fails (if there are lots of them,
4904 the message gets confusing).
4906 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4907 names, as there are now some PTR records that contain slashes. This check
4908 is only to protect against broken name servers that fall over on strange
4909 characters, so the fact that it applies to all lookups doesn't matter.
4911 PH/36 Now that the new test suite is complete, we can remove some of the
4912 special code in Exim that was needed for the old test suite. For example,
4913 sorting DNS records because real resolvers return them in an arbitrary
4914 order. The new test suite's fake resolver always returns records in the
4917 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4918 bounces) except when queue_only is set, to avoid logging races between
4919 the different processes.
4921 PH/38 Panic-die if .include specifies a non-absolute path.
4923 PH/39 A tweak to the "H" retry rule from its user.
4925 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4926 a label. They prevented compilation on older perls.
4928 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4929 a warning to be raised on newish perls.
4931 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4932 on queue. Changes to match documented behaviour of showing count of
4933 messages matching specified criteria.
4935 PH/40 Changed the default ident timeout from 30s to 5s.
4937 PH/41 Added support for the use of login_cap features, on those BSD systems
4938 that have them, for controlling the resources used by pipe deliveries.
4940 PH/42 The content-scanning code uses fopen() to create files in which to put
4941 message data. Previously it was not paying any attention to the mode of
4942 the files. Exim runs with umask(0) because the rest of the code creates
4943 files with open(), and sets the required mode explicitly. Thus, these
4944 files were ending up world-writeable. This was not a big issue, because,
4945 being within the spool directory, they were not world-accessible. I have
4946 created a function called modefopen, which takes an additional mode
4947 argument. It sets umask(777), creates the file, chmods it to the required
4948 mode, then resets the umask. All the relevant calls to fopen() in the
4949 content scanning code have been changed to use this function.
4951 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4952 to 24 hours. This avoids potential overflow problems when processing G
4953 and H retry rules. I suspect nobody ever tinkers with this value.
4955 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4957 PH/45 When the plaintext authenticator is running as a client, the server's
4958 challenges are checked to ensure they are valid base64 strings. By
4959 default, the authentication attempt is cancelled if an invalid string is
4960 received. Setting client_ignore_invalid_base64 true ignores these errors.
4961 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4962 they are received. Thus, the responses can be made to depend on the
4963 challenges. If an invalid string is ignored, an empty string is placed in
4966 PH/46 Messages that are created by the autoreply transport now contains a
4967 References: header, in accordance with RFCs 2822 and 3834.
4969 PH/47 Added authenticated_sender_force to the smtp transport.
4971 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4973 PH/49 Installed latest patch from the Sieve maintainer.
4975 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4976 was also set, the appendfile transport was unnecessarily scanning a
4977 directory of message files (e.g. for maildir delivery) to find the count
4978 of files (along with the size), even though it did not need this
4979 information. It now does the scan only if it needs to find either the
4980 size of the count of files.
4982 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4984 PH/52 Two bugs concerned with error handling when the smtp transport is
4987 (i) Exim was not creating retry information for temporary errors given
4988 for individual recipients after the DATA command when the smtp transport
4989 was used in LMTP mode. This meant that they could be retried too
4990 frequently, and not timed out correctly.
4992 (ii) Exim was setting the flag that allows error details to be returned
4993 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4994 recipients that were returned after the DATA command.
4996 PH/53 This is related to PH/52, but is more general: for any failing address,
4997 when detailed error information was permitted to be returned to the
4998 sender, but the error was temporary, then after the final timeout, only
4999 "retry timeout exceeded" was returned. Now it returns the full error as
5000 well as "retry timeout exceeded".
5002 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
5003 do this, and (what is worse) MTAs that accept it.
5005 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
5006 will now be deprecated.
5008 PH/56 New os.c-cygwin from the Cygwin maintainer.
5010 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5011 formats (previously only available in exim formats via -bpr, -bpru,
5012 and -bpra. Now also available in native and exiqgrep formats)
5014 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5015 with very large, slow to parse queues
5017 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5019 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5021 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5022 responses to authentication challenges, though it was showing the
5023 challenges; (ii) I've removed the CR characters from the debug output for
5026 PH/58 Allow for the insertion of a newline as well as a space when a string
5027 is turned into more than one encoded-word during RFC 2047 encoding. The
5028 Sieve code now uses this.
5030 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5031 data_4xx, lost_connection, tls_required.
5033 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5034 message was being sent as an SMTP response.
5036 PH/61 Add -l and -k options to exicyclog.
5038 PH/62 When verifying, if an address was redirected to one new address, so that
5039 verification continued, and the new address failed or deferred after
5040 having set something in $address_data, the value of $address_data was not
5041 passed back to the ACL. This was different to the case when no
5042 redirection occurred. The value is now passed back in both cases.
5044 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5045 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5046 use login_cap.h, so on its own it isn't the distinguishing feature. The
5047 new name refers directly to the setclassresources() function.
5049 PH/65 Added configuration files for NetBSD3.
5051 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5053 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5054 is preferred over IPv4.
5056 PH/68 The bounce_return_message and bounce_return_body options were not being
5057 honoured for bounces generated during the reception of non-SMTP messages.
5058 In particular, this applied to messages rejected by the ACL. This bug has
5059 been fixed. However, if bounce_return_message is true and bounce_return_
5060 body is false, the headers that are returned for a non-SMTP message
5061 include only those that have been read before the error was detected.
5062 (In the case of an ACL rejection, they have all been read.)
5064 PH/69 The HTML version of the specification is now built in a directory called
5065 spec_html instead of spec.html, because the latter looks like a path with
5066 a MIME-type, and this confuses some software.
5068 PH/70 Catch two compiler warnings in sieve.c.
5070 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5071 function verify_get_ident() calls ip_connect() to connect a socket, but
5072 if the "connect()" function timed out, ip_connect() used to close the
5073 socket. However, verify_get_ident() also closes the socket later, and in
5074 between Exim writes to the log, which may get opened at this point. When
5075 the socket was closed in ip_connect(), the log could get the same file
5076 descriptor number as the socket. This naturally causes chaos. The fix is
5077 not to close the socket in ip_connect(); the socket should be closed by
5078 the function that creates it. There was only one place in the code where
5079 this was missing, in the iplookup router, which I don't think anybody now
5080 uses, but I've fixed it anyway.
5082 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5083 well as to direct DNS lookups. Otherwise the handling of names in host
5084 lists is inconsistent and therefore confusing.
5090 PH/01 Two changes to the default runtime configuration:
5092 (1) Move the checks for relay_from_hosts and authenticated clients from
5093 after to before the (commented out) DNS black list checks.
5095 (2) Add control=submission to the relay_from_hosts and authenticated
5096 clients checks, on the grounds that messages accepted by these
5097 statements are most likely to be submissions.
5099 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5101 (1) Generate an error if the third argument for the ${prvs expansion is
5104 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5107 (3) Reset the variables that are obtained from the first argument of
5108 ${prvscheck and used in the second argument before leaving the code,
5109 because their memory is reclaimed, so using them afterwards may do
5112 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5113 one (it's much easier than Tom thought :-).
5115 (5) Because of (4), we can now allow for the use of $prvscheck_result
5116 inside the third argument.
5118 PH/03 For some reason, the default setting of PATH when running a command from
5119 a pipe transport was just "/usr/bin". I have changed it to
5122 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5123 anything to be listed in the output from -bV.
5125 PH/05 When a filter generated an autoreply, the entire To: header line was
5126 quoted in the delivery log line, like this:
5128 => >A.N.Other <ano@some.domain> <original@ddress> ...
5130 This has been changed so that it extracts the operative address. There
5131 may be more than one such address. If so, they are comma-separated, like
5134 => >ano@some.domain,ona@other.domain <original@ddress> ...
5136 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5137 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5138 not being looked up in the rDNS to get a host name, Exim was showing the
5139 IP address twice in Received: lines, even though the IP addresses were
5140 identical. For example:
5142 Received: from [1.2.3.4] (helo=[1.2.3.4])
5144 However, if the real host name was known, it was omitting the HELO data
5145 if it matched the actual IP address. This has been tidied up so that it
5146 doesn't show the same IP address twice.
5148 PH/07 When both +timestamp and +memory debugging was on, the value given by
5149 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5150 called by the debug printing, thereby overwriting the timestamp buffer.
5151 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5153 PH/08 When the original message was included in an autoreply transport, it
5154 always said "this is a copy of the message, including all the headers",
5155 even if body_only or headers_only was set. It now gives an appropriate
5158 PH/09 Applied a patch from the Sieve maintainer which:
5160 o fixes some comments
5161 o adds the (disabled) notify extension core
5162 o adds some debug output for the result of if/elsif tests
5163 o points to the current vacation draft in the documentation
5164 and documents the missing references header update
5168 o fixes a bug in processing the envelope test (when testing
5169 multiple envelope elements, the last element determined the
5172 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5173 Electronic Mail") by including:
5175 Auto-submitted: auto-generated
5177 in the messages that it generates (bounce messages and others, such as
5178 warnings). In the case of bounce messages for non-SMTP messages, there was
5179 also a typo: it was using "Auto_submitted" (underscore instead of
5180 hyphen). Since every message generated by Exim is necessarily in response
5181 to another message, thes have all been changed to:
5183 Auto-Submitted: auto-replied
5185 in accordance with these statements in the RFC:
5187 The auto-replied keyword:
5189 - SHOULD be used on messages sent in direct response to another
5190 message by an automatic process,
5192 - MUST NOT be used on manually-generated messages,
5194 - MAY be used on Delivery Status Notifications (DSNs) and Message
5195 Disposition Notifications (MDNs),
5197 - MUST NOT be used on messages generated by automatic or periodic
5198 processes, except for messages which are automatic responses to
5201 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5202 to the default Received: header definition.
5204 PH/12 Added log selector acl_warn_skipped (default on).
5206 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5207 variables because (a) they are in the wrong storage pool and (b) even if
5208 they were copied, it wouldn't work properly because of the caching.
5210 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5211 checking when decoding. Apparently there are clients that generate
5212 overlong encoded strings. Why am I not surprised?
5214 PH/15 If the first argument of "${if match_address" was not empty, but did not
5215 contain an "@" character, Exim crashed. Now it writes a panic log message
5216 and treats the condition as false.
5218 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5220 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5221 extension "envelope-auth". The code is finished and in agreement with
5222 other implementations, but there is no documentation so far and in fact,
5223 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5224 not changing the active code.
5226 Print executed "if" and "elsif" statements when debugging is used. This
5227 helps a great deal to understand what a filter does.
5229 Document more things not specified clearly in RFC3028. I had all this
5230 sorted out, when out of a sudden new issues came to my mind. Oops."
5232 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5235 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5236 canonical form (as documented). However, after a host name lookup from
5237 the IP address, check_host() was doing a simple string comparison with
5238 addresses acquired from the DNS when checking that the found name did
5239 have the original IP as one of its addresses. Since any found IPv6
5240 addresses are likely to be in abbreviated form, the comparison could
5241 fail. Luckily, there already exists a function for doing the comparison
5242 by converting both addresses to binary, so now that is used instead of
5243 the text comparison.
5245 PH/20 There was another similar case to PH/19, when a complete host name was
5246 given in a host list; looking up its IP address could give an abbreviated
5247 form, whereas the current host's name might or might not be abbreviated.
5248 The same fix has been applied.
5254 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5255 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5258 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5259 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5261 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5263 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5264 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5265 Also, for "Auto-Submitted", treat anything other than "no" as
5266 non-personal, in accordance with RFC 3834. (Previously it treated
5267 anything starting "auto-" as non-personal.)
5269 TF/01 The control=submission/name=... option had a problem with syntax
5270 errors if the name included a slash character. The /name= option
5271 now slurps the rest of the string, so it can include any characters
5272 but it must come last in the list of options (after /sender_retain
5275 PH/05 Some modifications to the interface to the fake nameserver for the new
5283 TK/01 Added the "success_on_redirect" address verification option. See
5284 NewStuff for rationale and an example.
5286 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5288 PH/02 Patch to exigrep to allow it to work on syslog lines.
5290 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5291 fread() to skip over the body file's header line, because in Cygwin the
5292 header line is locked and is inaccessible.
5294 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5295 co-exist for some time) to make it clear that it is the Exim ID that is
5296 referenced, not the Message-ID: header line.
5298 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5299 string_format() function, because snprintf() does not exist on all
5302 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5303 ${stat: expansion item.
5305 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5306 protocol synchronization error", to keep the pedants happy.
5308 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5309 well as for IRIX systems, when gcc is being used. See the host.c source
5312 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5314 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5317 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5318 transport and to the smtp transport in LMTP mode.
5320 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5322 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5323 run a filter in a subprocess. This could lead to confusion in subsequent
5324 lookups in the parent process. There should also be a search_tidyup() at
5325 the end of the subprocess.
5327 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5328 only if the host matched helo_try_verify_hosts, which caused the
5329 verification to occur when the EHLO/HELO command was issued. The ACL just
5330 tested the remembered result. Now, if a previous verification attempt has
5331 not happened, "verify = helo" does it there and then.
5333 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5335 TK/03 Fix log output including CR from clamd.
5337 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5338 exist provoked a memory error which could cause a segfault.
5340 PH/15 Installed PCRE 6.2
5342 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5344 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5345 of the problem. Specifically, suggested +O2 rather than +O1 for the
5348 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5350 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5351 record, the port was not correctly logged when the outgoing_port log
5352 selector was set (it logged the transort's default port).
5354 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5355 fallback_hosts, and "hosts" in the smtp transport.
5357 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5358 host errors such as "Connection refused".
5360 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5361 authentication with radiusclient 0.4.9:
5363 - Error returned from rc_read_config was caught wrongly
5364 - Username/password not passed on to radius server due to wrong length.
5366 The presumption is that some radiusclient API changes for 4.51/PH/17
5367 were not taken care of correctly. The code is still untested by me (my
5368 Linux distribution still has 0.3.2 of radiusclient), but it was
5369 contributed by a Radius user.
5371 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5372 expanding the "port" option of the smtp transport.
5374 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5375 while reading a MIME header. Thanks to Tom Hughes for a patch.
5377 PH/24 Include config.h inside local_scan.h so that configuration settings are
5380 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5381 that in an ACL for RCPT (for example), you can examine exactly what was
5384 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5385 commands, but it was not correctly comparing the address with the actual
5386 client host address. Thus, it would show the EHLO address in Received:
5387 header lines when this was not necessary.
5389 PH/27 Added the % operator to ${eval:}.
5391 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5392 it should be ignoring failures (because with -C, for example, it has lost
5393 privilege). It wasn't ignoring creation failures other than "already
5396 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5399 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5400 because an input line was too long, either on its own, or by virtue of
5401 too many continuations, the temporary file was not being removed, and the
5402 return code was incorrect.
5404 PH/31 Missing "BOOL" in function definition in filtertest.c.
5406 PH/32 Applied Sieve patches from the maintainer.
5408 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5410 PH/33 Added "verify = not_blind".
5412 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5413 Local/Makefile (with some defaults set). These are used in built scripts
5414 such as exicyclog, but they have never been used in the exim_install
5415 script (though there are many overriding facilities there). I have
5416 arranged that the exim_install script now takes note of these two
5419 PH/35 Installed configuration files for Dragonfly.
5421 PH/36 When a locally submitted message by a trusted user did not contain a
5422 From: header, and the sender address was obtained from -f or from an SMTP
5423 MAIL command, and the trusted user did not use -F to supply a sender
5424 name, $originator_name was incorrectly used when constructing a From:
5425 header. Furthermore, $originator_name was used for submission mode
5426 messages from external hosts without From: headers in a similar way,
5427 which is clearly wrong.
5429 PH/37 Added control=suppress_local_fixups.
5431 PH/38 When log_selector = +received_sender was set, and the addition of the
5432 sender made the log line's construction buffer exactly full, or one byte
5433 less than full, an overflow happened when the terminating "\n" was
5436 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5437 when the result of a list match is failure because a DNS lookup failed.
5439 PH/40 RM_COMMAND is now used in the building process.
5441 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5442 the "build-* directories that it finds.
5444 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5445 address in a domain literal was a prefix of an interface address.
5447 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5448 when verifying a sender address, unless rewrite_headers is false.
5450 PH/44 Wrote a long comment about why errors_to addresses are verified as
5451 recipients, not senders.
5453 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5454 the ratelimit ACL was added.
5456 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5458 PH/46 Added extra information about PostgreSQL errors to the error string.
5460 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5461 avoiding the need to install special zones in a real server. This is
5462 backwards compatible; if it can't find the fake resolver, it drops back.
5463 Thus, both old and new test suites can be run.
5465 TF/02 Added util/ratelimit.pl
5467 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5468 clock is set back in time.
5470 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5471 Candler <B.Candler@pobox.com>.
5473 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5474 OK for addresses that are the result of redirections.
5476 PH/48 A number of further additions for the benefit of the new test suite,
5477 including a fake gethostbyname() that interfaces to the fake DNS resolver
5480 TF/06 The fix for widen_domains has also been applied to qualify_single and
5481 search_parents which are the other dnslookup options that can cause
5484 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5487 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5489 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5490 service" if it encountered an MX record with an empty target hostname.
5491 The message is now "an MX or SRV record indicated no SMTP service".
5493 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5494 if the DNS of the sending site is misconfigured. This is quite a
5495 common situation. This change restores the behaviour of treating a
5496 helo verification defer as a failure.
5498 PH/51 If self=fail was set on a router, the bounce message did not include the
5499 actual error message.
5505 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5507 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5508 command itself ended in error, the underlying message about the transport
5509 filter timeout was being overwritten with the pipe command error. Now the
5510 underlying error message should be appended to the second error message.
5512 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5514 PH/02 Reduce the amount of output that "make" produces by default. Full output
5515 can still be requested.
5517 PH/03 The warning log line about a condition test deferring for a "warn" verb
5518 was being output only once per connection, rather than after each
5519 occurrence (because it was using the same function as for successful
5520 "warn" verbs). This seems wrong, so I have changed it.
5522 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5523 it should not have, which might have caused a crash in the right
5524 circumstances, but probably never did.
5526 PH/04 Installed a modified version of Tony Finch's patch to make submission
5527 mode fix the return path as well as the Sender: header line, and to
5528 add a /name= option so that you can make the user's friendly name appear
5531 TF/03 Added the control = fakedefer ACL modifier.
5533 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5534 Mark Lowes for thorough testing.
5536 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5538 TK/03 Merged latest SRS patch from Miles Wilton.
5540 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5541 with the definition in sysexits.h (which is #included earlier).
5542 Fortunately, Exim does not actually use EX_OK. The code used to try to
5543 preserve the sysexits.h value, by assuming that macro definitions were
5544 scanned for macro replacements. I have been disabused of this notion,
5545 so now the code just undefines EX_OK before #including unistd.h.
5547 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5548 in the smtp transport. When a block could not be written in a single
5549 write() function, the timeout was being re-applied to each part-write.
5550 This seems wrong - if the receiver was accepting one byte at a time it
5551 would take for ever. The timeout is now adjusted when this happens. It
5552 doesn't have to be particularly precise.
5554 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5555 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5557 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5558 without a domain if the check to <postmaster@domain> fails.
5560 SC/01 Eximstats: added -xls and the ability to specify output files
5561 (patch written by Frank Heydlauf).
5563 SC/02 Eximstats: use FileHandles for outputting results.
5565 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5567 SC/04 Eximstats: fixed display of large numbers with -nvr option
5569 SC/05 Eximstats: fixed merging of reports with empty tables.
5571 SC/06 Eximstats: added the -include_original_destination flag
5573 SC/07 Eximstats: removed tabs and trailing whitespace.
5575 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5577 TK/06 MBOX spool code: Add real "From " MBOX separator line
5578 so the .eml file is really in mbox format (even though
5579 most programs do not really care). Patch from Alex Miller.
5581 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5582 The latter is generated from $received_to and is only set if the
5583 message has one envelope recipient. SA can use these headers,
5584 obviously out-of-the-box. Patch from Alex Miller.
5586 PH/08 The ${def test on a variable was returning false if the variable's
5587 value was "0", contrary to what the specification has always said!
5588 The result should be true unless the variable is empty.
5590 PH/09 The syntax error of a character other than { following "${if
5591 def:variable_name" (after optional whitespace) was not being diagnosed.
5592 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5593 accidental colon was present, for example, could give incorrect results.
5595 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5596 result is used (not including appendfile, where other changes are about
5599 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5600 This involved changing a lot of size variables from int to off_t. It
5601 should work with maildirs and everything.
5603 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5604 spamd dying while we are connected to it.
5606 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5609 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5610 function for BDB 4.3.
5612 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5614 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5615 the sources for PCRE have been re-arranged and more files are now
5618 PH/15 The code I had for printing potentially long long variables in PH/11
5619 above was not the best (it lost precision). The length of off_t variables
5620 is now inspected at build time, and an appropriate printing format (%ld
5621 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5622 to be "long long int" or "long int". This is needed for the internal
5623 formatting function string_vformat().
5625 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5626 the configuration file to be ":syslog", then the script "guesses" where
5627 the logs files are, rather than using the compiled in default. In our
5628 case the guess is not the same as the compiled default, so the script
5629 suddenly stopped working when I started to use syslog. The patch checks
5630 to see if log_file_path is "". If so, it attempts to read it from exim
5631 with no configuration file to get the compiled in version, before it
5632 falls back to the previous guessing code."
5634 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5635 implementing BATV in an Exim configuration. See NewStuff for the gory
5638 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5639 Makefile that are specific to HP-UX.
5641 PH/18 If the "use_postmaster" option was set for a recipient callout together
5642 with the "random" option, the postmaster address was used as the MAIL
5643 FROM address for the random test, but not for the subsequent recipient
5644 test. It is now used for both.
5646 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5647 patch removes a few documentation additions to RFC 3028, because the
5648 latest draft now contains them. It adds the new en;ascii-case comparator
5649 and a new error check for 8bit text in MIME parts. Comparator and
5650 require names are now matched exactly. I enabled the subaddress
5651 extension, but it is not well tested yet (read: it works for me)."
5653 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5654 rework some of the code of TK/09 above to avoid the hardwired use of
5655 "%lld" and "long long". Replaced the call to snprintf() with a call to
5658 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5659 records point to non-existent hosts", "retry timeout exceeded", and
5660 "retry time not reached for any host after a long failure period".
5662 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5663 experimental DomainKeys support:
5665 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5666 (2) On an error such as an illegally used "control", the wrong name for
5667 the control was given.
5669 These problems did NOT occur unless DomainKeys support was compiled.
5671 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5673 PH/24 Added ${if match_ip condition.
5675 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5676 databases so that it will be absolutely obvious if a crash occurs in the
5677 DB library. This is a regular occurrence (often caused by mis-matched
5680 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5681 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5682 picked up on a user's system that detects such things. There doesn't seem
5683 to be a gcc warning option for this - only an attribute that has to be
5684 put on the function's prototype. It seems that in Fedora Core 4 they have
5685 set this on a number of new functions. No doubt there will be more in due
5688 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5689 specify a transport. However, if an address that was verified by such a
5690 router was the subject of a callout, Exim crashed because it tried to
5691 read the rcpt_include_affixes from the non-existent transport. Now it
5692 just assumes that the setting of that option is false. This bug was
5693 introduced by 4.51/PH/31.
5695 PH/28 Changed -d+all to exclude +memory, because that information is very
5696 rarely of interest, but it makes the output a lot bigger. People tend to
5697 do -d+all out of habit.
5699 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5700 code in os-type was giving problems when libc.so lives in lib64, like on
5703 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5704 aren't the modern standard, and it seems that some systems' include files
5705 don't always have them. Exim was already checking for some of the newer
5706 ones like T_AAAA, and defining it itself. I've added checks for all the
5707 record types that Exim uses.
5709 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5710 not automatically generating a new one, as it is supposed to. This
5711 prevented TLS from working. If the file did exist, but contained invalid
5712 data, a new version was generated, as expected. It was only the case of a
5713 non-existent file that was broken.
5715 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5716 with a change in libdomainkeys > 0.64.
5718 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5719 from DNS. If the selector record carries the flag, it now has
5720 precedence over the domain-wide flag.
5722 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5724 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5725 the use of a port name that isn't defined in /etc/services) occurred, the
5726 message was deferred as in a normal delivery, and thus remained on the
5727 spool, instead of being failed because of the mua_wrapper setting. This
5728 is now fixed, and I tidied up some of the mua_wrapper messages at the
5731 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5732 the messages in a hash of arrays rather than using individual hashes.
5733 This is a bit cleaner and results in dramatic memory savings, albeit
5734 at a slight CPU cost.
5736 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5737 as requested by Marc Sherman.
5739 SC/10 Eximstats: added histograms for user specified patterns as requested
5742 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5744 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5745 fopen() in the content-scanning modules that did not already have it.
5751 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5752 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5754 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5756 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5758 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5759 to test Sieve filters that use "vacation".
5761 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5762 that changes the way the GnuTLS parameters are stored in the cache file.
5763 The new format can be generated externally. For backward compatibility,
5764 if the data in the cache doesn't make sense, Exim assumes it has read an
5765 old-format file, and it generates new data and writes a new file. This
5766 means that you can't go back to an older release without removing the
5769 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5770 work if there are any delivery delays because "one_time" forces the
5771 parent to be marked "delivered", so its unseen clone is never tried
5772 again. For this reason, Exim now forbids the simultaneous setting of
5775 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5776 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5777 entry if you want to know the details. The fix, however, neglected to
5778 consider the case when local delivery batching is involved. The test for
5779 "previously delivered" was not happening when checking to see if an
5780 address could be batched with a previous (undelivered) one; under
5781 certain circumstances this could lead to multiple deliveries to the same
5784 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5785 in its include files, and this causes problems building Exim.
5787 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5788 header_syntax) but Exim was just ignoring anything given after a slash.
5789 In particular, this caused confusion with an attempt to use "verify =
5790 reverse_host_lookup/defer_ok". An error is now given when options are
5791 supplied for verify items that do not have them. (Maybe reverse_host_
5792 lookup should have a defer_ok option, but that's a different point.)
5794 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5795 defined by RFC 821) to 2048, because there were problems with some AUTH
5796 commands, and RFC 1869 says the size should be increased for extended
5797 SMTP commands that take arguments.
5799 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5802 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5803 "unknown" error; now it says that the functionality isn't in the binary.
5805 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5806 an address' error message when a string expansion fails (syntax or
5807 whatever). Otherwise the password may appear in the log. Following change
5808 PH/42 below, there is no longer a chance of it appearing in a bounce
5811 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5813 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5814 of its addresses was ever tried. (Bugzilla bug #2.)
5816 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5817 the result incorrectly in the debug output. (It correctly added a newline
5818 to what was transported.)
5820 TF/01 Added $received_time.
5822 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5823 commented out examples of how to interface to a virus scanner and to
5824 SpamAssassin. Also added commented examples of av_scanner and
5825 spamd_address settings.
5827 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5828 and controls are allowed in which ACLs. There were a couple of minor
5829 errors. Some of the entries in the conditions table (which is a table of
5830 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5831 negation of where the condition IS allowed.
5833 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5835 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5836 header file does not have a version number, so I've had to invent a new
5837 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5838 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5839 radiusclient), but it was contributed by a Radius user.
5841 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5842 files or directories, for OpenSSL.
5844 PH/19 When an Exim process that is running as root has to create an Exim log
5845 file, it does so in a subprocess that runs as exim:exim so as to get the
5846 ownership right at creation (otherwise, other Exim processes might see
5847 the file with the wrong ownership). There was no test for failure of this
5848 fork() call, which would lead to the process getting stuck as it waited
5849 for a non-existent subprocess. Forks do occasionally fail when resources
5850 run out. I reviewed all the other calls to fork(); they all seem to check
5853 PH/20 When checking for unexpected SMTP input at connect time (before writing
5854 the banner), Exim was not dealing correctly with a non-positive return
5855 from the read() function. If the client had disconnected by this time,
5856 the result was a log entry for a synchronization error with an empty
5857 string after "input=" when read() returned zero. If read() returned -1
5858 (an event I could not check), uninitialized data bytes were printed.
5859 There were reports of junk text (parts of files, etc) appearing after
5862 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5864 PH/22 Added support for macro redefinition, and (re)definition in between
5865 driver and ACL definitions.
5867 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5868 forgetting to use the resulting value; it was using the unexpanded value.
5870 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5871 hadn't been configured. The fix is from Juergen Kreileder, who
5872 understands it better than I do:
5874 "Here's what I see happening with three configured cyrus_sasl
5875 authenticators configured (plain, login, cram-md5):
5877 On startup auth_cyrus_sasl_init() gets called for each of these.
5878 This means three calls to sasl_listmech() without a specified mech_list.
5879 => SASL tests which mechs of all available mechs actually work
5880 => three warnings about OTP not working
5881 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5883 With the patch, sasl_listmech() also gets called three times. But now
5884 SASL's mech_list option is set to the server_mech specified in the the
5885 authenticator. Or in other words, the answer from sasl_listmech()
5886 gets limited to just the mech you're testing for (which is different
5888 => the return list contains just 'plain' or 'login', 'cram-md5' or
5889 nothing depending on the value of ob->server_mech.
5891 I've just tested the patch: Authentication still works fine,
5892 unavailable mechs specified in the exim configuration are still
5893 caught, and the auth.log warnings about OTP are gone."
5895 PH/25 When debugging is enabled, the contents of the command line are added
5896 to the debugging output, even when log_selector=+arguments is not
5899 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5900 answer is "GNU", and only if the return is "GNU/something" is the answer
5903 PH/27 $acl_verify_message is now set immediately after the failure of a
5904 verification in an ACL, and so is available in subsequent modifiers. In
5905 particular, the message can be preserved by coding like this:
5907 warn !verify = sender
5908 set acl_m0 = $acl_verify_message
5910 Previously, $acl_verify_message was set only while expanding "message"
5911 and "log_message" when a very denied access.
5913 PH/28 Modified OS/os.c-Linux with
5915 -#ifndef OS_LOAD_AVERAGE
5916 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5918 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5919 nomenclature these days.)
5921 PH/29 Installed patch from the Sieve maintainer that adds the options
5922 sieve_useraddress and sieve_subaddress to the redirect router.
5924 PH/30 In these circumstances:
5925 . Two addresses routed to the same list of hosts;
5926 . First host does not offer TLS;
5927 . First host accepts first address;
5928 . First host gives temporary error to second address;
5929 . Second host offers TLS and a TLS session is established;
5930 . Second host accepts second address.
5931 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5932 and peerdn, if requested) that were in fact used only for the second
5935 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5936 attention to any local part prefix or suffix that was matched by the
5937 router that accepted the address. It now behaves in the same way as it
5938 does for delivery: the affixes are removed from the local part unless
5939 rcpt_include_affixes is set on the transport.
5941 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5942 timeout during the DATA phase of an incoming message.
5944 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5945 applied a patch sanctioned by the Sieve maintainer.
5947 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5948 the uid or gid is negative. A case of a negative gid caused this to be
5949 noticed. The fix allows for either to be negative.
5951 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5952 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5955 PH/36 The change PH/12 above was broken. Fixed it.
5957 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5958 the grounds that routing the same address twice would always produce the
5959 same answer. This might have been true once, but it is certainly no
5960 longer true now. Routing a child address may depend on the previous
5961 routing that produced that child. Some complicated redirection strategies
5962 went wrong when messages had multiple recipients, and made Exim's
5963 behaviour dependent on the order in which the addresses were given.
5965 I have moved the duplicate checking until after the routing is complete.
5966 Exim scans the addresses that are assigned to local and remote
5967 transports, and removes any duplicates. This means that more work will be
5968 done, as duplicates will always all be routed, but duplicates are
5969 presumably rare, so I don't expect this is of any significance.
5971 For deliveries to pipes, files, and autoreplies, the duplicate checking
5972 still happens during the routing process, since they are not going to be
5975 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5976 It corrects a timeout issue with spamd. This is Ian's comment: "The
5977 background is that sometimes spamd either never reads data from a
5978 connection it has accepted, or it never writes response data. The exiscan
5979 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5980 blindly assumes that writes won't block so it may never time out."
5982 PH/39 Allow G after quota size as well as K and M.
5984 PH/40 The value set for $authenticated_id in an authenticator may not contain
5985 binary zeroes or newlines because the value is written to log lines and
5986 to spool files. There was no check on this. Now the value is run through
5987 the string_printing() function so that such characters are converted to
5988 printable escape sequences.
5990 PH/41 $message_linecount is a new variable that contains the total number of
5991 lines in the message. Compare $body_linecount, which is the count for the
5994 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5995 bounce and delay warning messages, except in certain special cases, which
5998 (a) An SMTP error message from a remote host;
5999 (b) A message specified in a :fail: redirection;
6000 (c) A message specified in a "fail" command in a system filter;
6001 (d) A message specified in a FAIL return from the queryprogram router;
6002 (e) A message specified by the cannot_route_message router option.
6004 In these cases only, Exim does include the error details in bounce and
6005 warning messages. There are also a few cases where bland messages such
6006 as "unrouteable address" or "local delivery error" are given.
6008 PH/43 $value is now also set for the "else" part of a ${run expansion.
6010 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6011 being worked on, but at least Exim now implements the latest version to
6014 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6015 process to complete was treated as a delivery failure, a timeout while
6016 writing the message to the pipe was logged, but erroneously treated as a
6017 successful delivery. Such timeouts include transport filter timeouts. For
6018 consistency with the overall process timeout, these timeouts are now
6019 treated as errors, giving rise to delivery failures by default. However,
6020 there is now a new Boolean option for the pipe transport called
6021 timeout_defer, which, if set TRUE, converts the failures into defers for
6022 both kinds of timeout. A transport filter timeout is now identified in
6025 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6026 systems where "make" and "gmake" are different, calling "gmake" at top
6027 level broke things. I've arranged for the value of $(MAKE) to be passed
6028 from the Makefile to this script so that it can call the same version of
6032 A note about Exim versions 4.44 and 4.50
6033 ----------------------------------------
6035 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6036 changes of various kinds. As a consequence, a big documentation update was
6037 needed. This delayed the release for rather longer than seemed good, especially
6038 in the light of a couple of (minor) security issues. Therefore, the changes
6039 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6040 release. So 4.44 and 4.50 are in effect two different branches that both start
6043 I have left the 4.50 change log unchanged; it contains all the changes since
6044 4.43. The change log for 4.44 is below; many of its items are identical to
6045 those for 4.50. This seems to be the most sensible way to preserve the
6046 historical information.
6052 1. Minor wording change to the doc/README.SIEVE file.
6054 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6055 computation of the current number of files was incorrect.
6057 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6058 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6061 4. Give more explanation in the error message when the command for a transport
6062 filter fails to execute.
6064 5. There are several places where Exim runs a non-Exim command in a
6065 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6066 was being done only for the command run by the queryprogram router. It is
6067 now done for all such subprocesses. The other cases are: ${run, transport
6068 filters, and the commands run by the lmtp and pipe transports.
6070 6. Added CONFIGURE_GROUP build-time option.
6072 7. Some older OS have a limit of 256 on the maximum number of file
6073 descriptors. Exim was using setrlimit() to set 1000 as a large value
6074 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6075 systems. I've change it so that if it can't get 1000, it tries for 256.
6077 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6078 was an oversight, and furthermore, ever since the addition of extra
6079 controls (e.g. 4.43/32), the checks on when to allow different forms of
6080 "control" were broken. There should now be diagnostics for all cases when a
6081 control that does not make sense is encountered.
6083 9. Added the /retain_sender option to "control=submission".
6085 10. $recipients is now available in the predata ACL (oversight).
6087 11. Tidy the search cache before the fork to do a delivery from a message
6088 received from the command line. Otherwise the child will trigger a lookup
6089 failure and thereby defer the delivery if it tries to use (for example) a
6090 cached ldap connection that the parent has called unbind on.
6092 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6093 of $address_data from the recipient verification was clobbered by the
6094 sender verification.
6096 13. The value of address_data from a sender verification is now available in
6097 $sender_address_data in subsequent conditions in the ACL statement.
6099 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6101 15. Added a new option "connect=<time>" to callout options, to set a different
6104 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6105 was its contents. (It was OK if the option was not defined at all.)
6107 17. A "Completed" log line is now written for messages that are removed from
6108 the spool by the -Mrm option.
6110 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6111 information about exactly what failed.
6113 19. Added -dd to debug only the daemon process.
6115 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6116 handles timeouts, both on the server side and network timeouts. Renamed the
6117 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6119 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6120 It is now set to "smtps".
6122 22. $host_address is now set to the target address during the checking of
6123 ignore_target_hosts.
6125 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6126 being passed; this would have caused $sender_host_name to have been used if
6127 matching the list had actually called for a host name (not very likely,
6128 since this list is usually IP addresses). A host name is now passed as
6131 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6132 code that specifies a non-restarting handler (typically sigaction() in
6133 modern systems) in an attempt to fix a rare and obscure crash bug.
6135 25. Narrowed the window for a race in the daemon that could cause it to ignore
6136 SIGCHLD signals. This is not a major problem, because they are used only to
6137 wake it up if nothing else does.
6139 26. A malformed maildirsize file could cause Exim to calculate negative values
6140 for the mailbox size or file count. Odd effects could occur as a result.
6141 The maildirsize information is now recalculated if the size or filecount
6144 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6145 support for a long time. Removed HAVE_SYS_VFS_H.
6147 28. Installed the latest version of exipick from John Jetmore.
6149 29. In an address list, if the pattern was not a regular expression, an empty
6150 subject address (from a bounce message) matched only if the pattern was an
6151 empty string. Non-empty patterns were not even tested. This was the wrong
6152 because it is perfectly reasonable to use an empty address as part of a
6153 database query. An empty address is now tested by patterns that are
6154 lookups. However, all the other forms of pattern expect the subject to
6155 contain a local part and a domain, and therefore, for them, an empty
6156 address still always fails if the pattern is not itself empty.
6158 30. Exim went into a mad DNS loop when attempting to do a callout where the
6159 host was specified on an smtp transport, and looking it up yielded more
6160 than one IP address.
6162 31. Re-factored the code for checking spool and log partition space into a
6163 function that finds that data and another that does the check. The former
6164 is then used to implement four new variables: $spool_space, $log_space,
6165 $spool_inodes, and $log_inodes.
6167 32. The RFC2047 encoding function was originally intended for short strings
6168 such as real names; it was not keeping to the 75-character limit for
6169 encoded words that the RFC imposes. It now respects the limit, and
6170 generates multiple encoded words if necessary. To be on the safe side, I
6171 have increased the buffer size for the ${rfc2047: expansion operator from
6174 33. It is now permitted to omit both strings after an "if" condition; if the
6175 condition is true, the result is "true". As before, when the second string
6176 is omitted, a false condition yields an empty string. This makes it less
6177 cumbersome to write custom ACL and router conditions.
6179 34. Failure to deliver a bounce message always caused it to be frozen, even if
6180 there was an errors_to setting on the router. The errors_to setting is now
6183 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6184 canonical form (fully expanded) before being placed in
6185 $sender_host_address.
6187 36. The table in the code that translates DNS record types into text (T_A to
6188 "A" for instance) was missing entries for NS and CNAME. It is just possible
6189 that this could have caused confusion if both these types were looked up
6190 for the same domain, because the text type is used as part of Exim's
6191 per-process caching. But the chance of anyone hitting this buglet seems
6194 37. The dnsdb lookup has been extended in a number of ways.
6196 (1) There is a new type, "zns", which walks up the domain tree until it
6197 finds some nameserver records. It should be used with care.
6199 (2) There is a new type, "mxh", which is like "mx" except that it returns
6200 just the host names, not the priorities.
6202 (3) It is now possible to give a list of domains (or IP addresses) to be
6203 looked up. The behaviour when one of the lookups defers can be
6204 controlled by a keyword.
6206 (4) It is now possible to specify the separator character for use when
6207 multiple records are returned.
6209 38. The dnslists ACL condition has been extended: it is now possible to supply
6210 a list of IP addresses and/or domains to be looked up in a particular DNS
6213 39. Added log_selector=+queue_time_overall.
6215 40. When running the queue in the test harness, wait just a tad after forking a
6216 delivery process, to get repeatability of debugging output.
6218 41. Include certificate and key file names in error message when GnuTLS fails
6219 to set them up, because the GnuTLS error message doesn't include the name
6220 of the failing file when there is a problem reading it.
6222 42. Allow both -bf and -bF in the same test run.
6224 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6226 44. The "Exiscan patch" is now merged into the mainline Exim source.
6228 45. Sometimes the final signoff response after QUIT could fail to get
6229 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6230 < 0 before doing a fflush(). This bug looks as though it goes back to the
6231 introduction of TLS in release 3.20, but "sometimes" must have been rare
6232 because the tests only now provoked it.
6234 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6235 (this can affect the format of dates).
6237 47. exim_tidydb, when checking for the continued existence of a message for
6238 which it has found a message-specific retry record, was not finding
6239 messages that were in split spool directories. Consequently, it was
6240 deleting retry records that should have stayed in existence.
6242 48. Steve fixed some bugs in eximstats.
6244 49. The SPA authentication driver was not abandoning authentication and moving
6245 on to the next authenticator when an expansion was forced to fail,
6246 contradicting the general specification for all authenticators. Instead it
6247 was generating a temporary error. It now behaves as specified.
6249 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6250 (the order specifies the preference for clients). The order is now AES256,
6251 AES128, 3DES, ARCFOUR128.
6253 51. Small patch to Sieve code - explicitly set From: when generating an
6256 52. Exim crashed if a remote delivery caused a very long error message to be
6257 recorded - for instance if somebody sent an entire SpamAssassin report back
6258 as a large number of 550 error lines. This bug was coincidentally fixed by
6259 increasing the size of one of Exim's internal buffers (big_buffer) that
6260 happened as part of the Exiscan merge. However, to be on the safe side, I
6261 have made the code more robust (and fixed the comments that describe what
6264 53. Now that there can be additional text after "Completed" in log lines (if
6265 the queue_time_overall log selector is set), a one-byte patch to exigrep
6266 was needed to allow it to recognize "Completed" as not the last thing in
6269 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6270 patch that reportedly fixes this has been added. I am not expert enough to
6271 create a test for it. This is what the patch creator wrote:
6273 "I found a little strange behaviour of ldap code when working with
6274 Windows 2003 AD Domain, where users was placed in more than one
6275 Organization Units. When I tried to give exim partial DN, the exit code
6276 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6277 But simultaneously result of request was absolutely normal ldap result,
6278 so I produce this patch..."
6280 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6281 REFERENCE, so I have modified the code to exclude the patch when that macro
6284 55. Some experimental protocols are using DNS PTR records for new purposes. The
6285 keys for these records are domain names, not reversed IP addresses. The
6286 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6287 leaves it alone. Component reversal etc. now happens only for IP addresses.
6290 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6292 57. Double the size of the debug message buffer (to 2048) so that more of very
6293 long debug lines gets shown.
6295 58. The exicyclog utility now does better if the number of log files to keep
6296 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6298 59. Two changes related to the smtp_active_hostname option:
6300 (1) $smtp_active_hostname is now available as a variable.
6301 (2) The default for smtp_banner uses $smtp_active_hostname instead
6302 of $primary_hostname.
6304 60. The host_aton() function is supposed to be passed a string that is known
6305 to be a valid IP address. However, in the case of IPv6 addresses, it was
6306 not checking this. This is a hostage to fortune. Exim now panics and dies
6307 if the condition is not met. A case was found where this could be provoked
6308 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6309 components; fortuitously, this particular loophole had already been fixed
6310 by change 4.50/55 above.
6312 If there are any other similar loopholes, the new check in host_aton()
6313 itself should stop them being exploited. The report I received stated that
6314 data on the command line could provoke the exploit when Exim was running as
6315 exim, but did not say which command line option was involved. All I could
6316 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6317 running as the user.
6320 61. There was a buffer overflow vulnerability in the SPA authentication code
6321 (which came originally from the Samba project). I have added a test to the
6322 spa_base64_to_bits() function which I hope fixes it.
6325 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6326 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6328 63. The daemon start-up calls getloadavg() while still root for those OS that
6329 need the first call to be done as root, but it missed one case: when
6330 deliver_queue_load_max is set with deliver_drop_privilege. This is
6331 necessary for the benefit of the queue runner, because there is no re-exec
6332 when deliver_drop_privilege is set.
6334 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6335 This has been fixed.
6337 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6338 was in use, was not putting the data itself into the right store pool;
6339 consequently, it could be overwritten for a subsequent message in the same
6340 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6343 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6345 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6346 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6347 the function were treating the return as a boolean value, which happened to
6348 work because 0=false and not-0=true, but is not correct code.
6350 68. The host_aton() function was not handling scoped IPv6 addresses (those
6351 with, for example, "%eth0" on the end) correctly.
6353 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6354 negated items (that is, ~something) in unsigned ints. Some compilers
6355 apparently mutter when there is no cast.
6357 70. If an address verification called from an ACL failed, and did not produce a
6358 user-specific message (i.e. there was only a "system" message), nothing was
6359 put in $acl_verify_message. In this situation, it now puts the system
6362 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6363 session; change 4.31/43 added the unwanted input to the log line - except
6364 that it did not do this in the start of session case. It now does.
6366 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6367 This is wrong and can cause the other end to generate a synchronization
6368 error if it is another Exim or anything else that does the synchronization
6369 check. A QUIT command is no longer sent after a timeout.
6371 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6372 during host lookups.
6374 74. The defer_ok option of callout verification was not working if it was used
6375 when verifying addresses in header lines, that is, for this case:
6377 verify = header_sender/callout=defer_ok
6379 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6380 those file descriptors could be used for SMTP connections. If anything
6381 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6382 could be sent to the SMTP client, causing chaos. The daemon now opens
6383 stdin, stdout, and stderr to /dev/null when it puts itself into the
6386 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6387 log by default. The user can override this with suitable Perl magic.
6389 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6390 the log message when discard triggers, was not working for the DATA ACL or
6391 for the non-SMTP ACL.
6393 78. Error message wording change in sieve.c.
6395 79. If smtp_accept_max_per_host was set, the number of connections could be
6396 restricted to fewer than expected, because the daemon was trying to set up
6397 a new connection before checking whether the processes handling previous
6398 connections had finished. The check for completed processes is now done
6399 earlier. On busy systems, this bug wouldn't be noticed because something
6400 else would have woken the daemon, and it would have reaped the completed
6403 80. If a message was submitted locally by a user whose login name contained one
6404 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6405 It caused a spool format error. I have fixed the spool reading code. A
6406 related problem was that the "from" clause in the Received: line became
6407 illegal because of the space(s). It is now covered by ${quote_local_part.
6409 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6412 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6414 83. Patch from Sieve maintainer to fix unterminated string problem in
6415 "vacation" handling.
6417 84. Some minor changes to the Linux configuration files to help with other
6418 OS variants using glibc.
6420 85. One more patch for Sieve to update vacation handling to latest spec.
6423 ----------------------------------------------------
6424 See the note above about the 4.44 and 4.50 releases.
6425 ----------------------------------------------------
6431 1. Change 4.43/35 introduced a bug that caused file counts to be
6432 incorrectly computed when quota_filecount was set in an appendfile
6435 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6436 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6439 3. Give more explanation in the error message when the command for a transport
6440 filter fails to execute.
6442 4. There are several places where Exim runs a non-Exim command in a
6443 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6444 was being done only for the command run by the queryprogram router. It is
6445 now done for all such subprocesses. The other cases are: ${run, transport
6446 filters, and the commands run by the lmtp and pipe transports.
6448 5. Some older OS have a limit of 256 on the maximum number of file
6449 descriptors. Exim was using setrlimit() to set 1000 as a large value
6450 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6451 systems. I've change it so that if it can't get 1000, it tries for 256.
6453 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6454 was an oversight, and furthermore, ever since the addition of extra
6455 controls (e.g. 4.43/32), the checks on when to allow different forms of
6456 "control" were broken. There should now be diagnostics for all cases when a
6457 control that does not make sense is encountered.
6459 7. $recipients is now available in the predata ACL (oversight).
6461 8. Tidy the search cache before the fork to do a delivery from a message
6462 received from the command line. Otherwise the child will trigger a lookup
6463 failure and thereby defer the delivery if it tries to use (for example) a
6464 cached ldap connection that the parent has called unbind on.
6466 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6467 of $address_data from the recipient verification was clobbered by the
6468 sender verification.
6470 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6471 was its contents. (It was OK if the option was not defined at all.)
6473 11. A "Completed" log line is now written for messages that are removed from
6474 the spool by the -Mrm option.
6476 12. $host_address is now set to the target address during the checking of
6477 ignore_target_hosts.
6479 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6480 being passed; this would have caused $sender_host_name to have been used if
6481 matching the list had actually called for a host name (not very likely,
6482 since this list is usually IP addresses). A host name is now passed as
6485 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6486 code that specifies a non-restarting handler (typically sigaction() in
6487 modern systems) in an attempt to fix a rare and obscure crash bug.
6489 15. Narrowed the window for a race in the daemon that could cause it to ignore
6490 SIGCHLD signals. This is not a major problem, because they are used only to
6491 wake it up if nothing else does.
6493 16. A malformed maildirsize file could cause Exim to calculate negative values
6494 for the mailbox size or file count. Odd effects could occur as a result.
6495 The maildirsize information is now recalculated if the size or filecount
6498 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6499 support for a long time. Removed HAVE_SYS_VFS_H.
6501 18. Updated exipick to current release from John Jetmore.
6503 19. Allow an empty sender to be matched against a lookup in an address list.
6504 Previously the only cases considered were a regular expression, or an
6507 20. Exim went into a mad DNS lookup loop when doing a callout where the
6508 host was specified on the transport, if the DNS lookup yielded more than
6511 21. The RFC2047 encoding function was originally intended for short strings
6512 such as real names; it was not keeping to the 75-character limit for
6513 encoded words that the RFC imposes. It now respects the limit, and
6514 generates multiple encoded words if necessary. To be on the safe side, I
6515 have increased the buffer size for the ${rfc2047: expansion operator from
6518 22. Failure to deliver a bounce message always caused it to be frozen, even if
6519 there was an errors_to setting on the router. The errors_to setting is now
6522 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6523 canonical form (fully expanded) before being placed in
6524 $sender_host_address.
6526 24. Updated eximstats to version 1.33
6528 25. Include certificate and key file names in error message when GnuTLS fails
6529 to set them up, because the GnuTLS error message doesn't include the name
6530 of the failing file when there is a problem reading it.
6532 26. Expand error message when OpenSSL has problems setting up cert/key files.
6535 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6536 (this can affect the format of dates).
6538 28. exim_tidydb, when checking for the continued existence of a message for
6539 which it has found a message-specific retry record, was not finding
6540 messages that were in split spool directories. Consequently, it was
6541 deleting retry records that should have stayed in existence.
6543 29. eximstats updated to version 1.35
6544 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6545 1.35 - bugfix such that pie charts by volume are generated correctly
6547 30. The SPA authentication driver was not abandoning authentication and moving
6548 on to the next authenticator when an expansion was forced to fail,
6549 contradicting the general specification for all authenticators. Instead it
6550 was generating a temporary error. It now behaves as specified.
6552 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6553 (the order specifies the preference for clients). The order is now AES256,
6554 AES128, 3DES, ARCFOUR128.
6556 31. Small patch to Sieve code - explicitly set From: when generating an
6559 32. Exim crashed if a remote delivery caused a very long error message to be
6560 recorded - for instance if somebody sent an entire SpamAssassin report back
6561 as a large number of 550 error lines. This bug was coincidentally fixed by
6562 increasing the size of one of Exim's internal buffers (big_buffer) that
6563 happened as part of the Exiscan merge. However, to be on the safe side, I
6564 have made the code more robust (and fixed the comments that describe what
6567 33. Some experimental protocols are using DNS PTR records for new purposes. The
6568 keys for these records are domain names, not reversed IP addresses. The
6569 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6570 leaves it alone. Component reversal etc. now happens only for IP addresses.
6573 34. The host_aton() function is supposed to be passed a string that is known
6574 to be a valid IP address. However, in the case of IPv6 addresses, it was
6575 not checking this. This is a hostage to fortune. Exim now panics and dies
6576 if the condition is not met. A case was found where this could be provoked
6577 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6578 components; fortuitously, this particular loophole had already been fixed
6579 by change 4.50/55 or 4.44/33 above.
6581 If there are any other similar loopholes, the new check in host_aton()
6582 itself should stop them being exploited. The report I received stated that
6583 data on the command line could provoke the exploit when Exim was running as
6584 exim, but did not say which command line option was involved. All I could
6585 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6586 running as the user.
6589 35. There was a buffer overflow vulnerability in the SPA authentication code
6590 (which came originally from the Samba project). I have added a test to the
6591 spa_base64_to_bits() function which I hope fixes it.
6594 36. The daemon start-up calls getloadavg() while still root for those OS that
6595 need the first call to be done as root, but it missed one case: when
6596 deliver_queue_load_max is set with deliver_drop_privilege. This is
6597 necessary for the benefit of the queue runner, because there is no re-exec
6598 when deliver_drop_privilege is set.
6600 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6601 was in use, was not putting the data itself into the right store pool;
6602 consequently, it could be overwritten for a subsequent message in the same
6603 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6606 38. Sometimes the final signoff response after QUIT could fail to get
6607 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6608 < 0 before doing a fflush(). This bug looks as though it goes back to the
6609 introduction of TLS in release 3.20, but "sometimes" must have been rare
6610 because the tests only now provoked it.
6616 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6617 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6618 Now it returns an integer. A number of calls were still expecting a T/F
6619 return. Fortuitously, in all cases, the tests worked in OK situations,
6620 which is the norm. However, things would have gone wrong on any write
6621 failures on the smtp file descriptor. This function is used when sending
6622 messages over SMTP and also when doing verify callouts.
6624 2. When Exim is called to do synchronous delivery of a locally submitted
6625 message (the -odf or -odi options), it no longer closes stderr before doing
6628 3. Implemented the mua_wrapper option.
6630 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6632 5. Implemented the functions header_remove(), header_testname(),
6633 header_add_at_position(), and receive_remove_recipient(), and exported them
6636 6. If an ACL "warn" statement specified the addition of headers, Exim already
6637 inserted X-ACL-Warn: at the start if there was no header name. However, it
6638 was not making this test for the second and subsequent header lines if
6639 there were newlines in the string. This meant that an invalid header could
6640 be inserted if Exim was badly configured.
6642 7. Allow an ACL "warn" statement to add header lines at the start or after all
6643 the Received: headers, as well as at the end.
6645 8. Added the rcpt_4xx retry error code.
6647 9. Added postmaster_mailfrom=xxx to callout verification option.
6649 10. Added mailfrom=xxxx to the callout verification option, for verify=
6652 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6653 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6655 12. Inserted some casts to stop certain compilers warning when using pointer
6656 differences as field lengths or precisions in printf-type calls (mostly
6657 affecting debugging statements).
6659 13. Added optional readline() support for -be (dynamically loaded).
6661 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6662 same clock tick as a message's arrival, so that its received time was the
6663 same as the "first fail" time on the retry record, and that message
6664 remained on the queue past the ultimate address timeout, every queue runner
6665 would try a delivery (because it was past the ultimate address timeout) but
6666 after another failure, the ultimate address timeout, which should have then
6667 bounced the address, did not kick in. This was a "< instead of <=" error;
6668 in most cases the first failure would have been in the next clock tick
6669 after the received time, and all would be well.
6671 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6672 being recognized when the domain list was tested by the match_domain
6673 condition in an expansion string.
6675 16. Added the ${str2b64: operator.
6677 17. Exim was always calling setrlimit() to set a large limit for the number of
6678 processes, without checking whether the existing limit was already
6679 adequate. (It did check for the limit on file descriptors.) Furthermore,
6680 errors from getrlimit() and setrlimit() were being ignored. Now they are
6681 logged to the main and panic logs, but Exim does carry on, to try to do its
6682 job under whatever limits there are.
6684 18. Imported PCRE 5.0.
6686 19. Trivial typo in log message " temporarily refused connection" (the leading
6689 20. If the log selector return_path_on_delivery was set and an address was
6690 redirected to /dev/null, the delivery process crashed because it assumed
6691 that a return path would always be set for a "successful" delivery. In this
6692 case, the whole delivery is bypassed as an optimization, and therefore no
6695 21. Internal re-arrangement: the function for sending a challenge and reading
6696 a response while authentication was assuming a zero-terminated challenge
6697 string. It's now changed to take a pointer and a length, to allow for
6698 binary data in such strings.
6700 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6702 23. Exim was not respecting finduser_retries when seeking the login of the
6703 uid under which it was called; it was always trying 10 times. (The default
6704 setting of finduser_retries is zero.) Also, it was sleeping after the final
6705 failure, which is pointless.
6707 24. Implemented tls_on_connect_ports.
6709 25. Implemented acl_smtp_predata.
6711 26. If the domain in control=submission is set empty, Exim assumes that the
6712 authenticated id is a complete email address when it generates From: or
6713 Sender: header lines.
6715 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6716 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6717 chown and chgrp in /bin and hostname in /usr/bin.
6719 28. Exim was keeping the "process log" file open after each use, just as it
6720 does for the main log. This opens the possibility of it remaining open for
6721 long periods when the USR1 signal hits a daemon. Occasional processlog
6722 errors were reported, that could have been caused by this. Anyway, it seems
6723 much more sensible not to leave this file open at all, so that is what now
6726 29. The long-running daemon process does not normally write to the log once it
6727 has entered its main loop, and it closes the log before doing so. This is
6728 so that log files can straightforwardly be renamed and moved. However,
6729 there are a couple of unusual error situations where the daemon does write
6730 log entries, and I had neglected to close the log afterwards.
6732 30. The text of an SMTP error response that was received during a remote
6733 delivery was being truncated at 512 bytes. This is too short for some of
6734 the long messages that one sometimes sees. I've increased the limit to
6737 31. It is now possible to make retry rules that apply only when a message has a
6738 specific sender, in particular, an empty sender.
6740 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6741 it possible to be selective about when SMTP synchronization is enforced.
6743 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6745 32. Implemented hosts_connection_nolog.
6747 33. Added an ACL for QUIT.
6749 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6752 35. Added mailbox_size and mailbox_filecount to appendfile.
6754 36. Added control = no_multiline_responses to ACLs.
6756 37. There was a bug in the logic of the code that waits for the clock to tick
6757 in the case where the clock went backwards by a substantial amount such
6758 that the microsecond fraction of "now" was more than the microsecond
6759 fraction of "then" (but the whole seconds number was less).
6761 38. Added support for the libradius Radius client library this is found on
6762 FreeBSD (previously only the radiusclient library was supported).
6768 1. When certain lookups returned multiple values in the form name=value, the
6769 quoting of the values was not always being done properly. Specifically:
6770 (a) If the value started with a double quote, but contained no whitespace,
6772 (b) If the value contained whitespace other than a space character (i.e.
6773 tabs or newlines or carriage returns) it was not quoted.
6774 This fix has been applied to the mysql and pgsql lookups by writing a
6775 separate quoting function and calling it from the lookup code. The fix
6776 should probably also be applied to nisplus, ibase and oracle lookups, but
6777 since I cannot test any of those, I have not disturbed their existing code.
6779 2. A hit in the callout cache for a specific address caused a log line with no
6780 reason for rejecting RCPT. Now it says "Previous (cached) callout
6781 verification failure".
6783 3. There was an off-by-one bug in the queryprogram router. An over-long
6784 return line was truncated at 256 instead of 255 characters, thereby
6785 overflowing its buffer with the terminating zero. As well as fixing this, I
6786 have increased the buffer size to 1024 (and made a note to document this).
6788 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6789 when Exim is waiting for an SMTP response from a remote server, Exim
6790 restarts its select() call on the socket, thereby resetting its timeout.
6791 This is not a problem when such interrupts are rare. Somebody set up a cron
6792 job to run exiwhat every 2 minutes, which is less than the normal select()
6793 timeout (5 or 10 minutes). This meant that the select() timeout never
6794 kicked in because it was always reset. I have fixed this by comparing the
6795 time when an interrupt arrives with the time at the start of the first call
6796 to select(). If more time than the timeout has elapsed, the interrupt is
6797 treated as a timeout.
6799 5. Some internal re-factoring in preparation for the addition of Sieve
6800 extensions (by MH). In particular, the "personal" test is moved to a
6801 separate function, and given an option for scanning Cc: and Bcc: (which is
6802 not set for Exim filters).
6804 6. When Exim created an email address using the login of the caller as the
6805 local part (e.g. when creating a From: or Sender: header line), it was not
6806 quoting the local part when it contained special characters such as @.
6808 7. Installed new OpenBSD configuration files.
6810 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6811 try to make them clearer.
6813 9. Callout options, other than the timeout value, were being ignored when
6814 verifying sender addresses in header lines. For example, when using
6816 verify = header_sender/callout=no_cache
6818 the cache was (incorrectly) being used.
6820 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6821 only the Cygwin environment.
6823 11. When return_path_on_delivery was set as a log selector, if different remote
6824 addresses in the same message used different return paths and parallel
6825 remote delivery occurred, the wrong values would sometimes be logged.
6826 (Whenever a remote delivery process finished, the return path value from
6827 the most recently started remote delivery process was logged.)
6829 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6830 lines when AUTH and/or TLS are in use. This is the "received protocol"
6831 field. Exim used to use "asmtp" for authenticated SMTP, without any
6832 indication (in the protocol name) for TLS use. Now it follows the RFC and
6833 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6834 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6835 names appear in log lines as well as in Received: header lines.
6837 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6838 extensions, and comparison tests, and to fix some bugs.
6840 14. Changes to the "personal" filter test:
6842 (1) The test was buggy in that it was just doing the equivalent of
6843 "contains" tests on header lines. For example, if a user's address was
6844 anne@some.where, the "personal" test would incorrectly be true for
6846 To: susanne@some.where
6848 This test is now done by extracting each address from the header in turn,
6849 and checking the entire address. Other tests that are part of "personal"
6850 are now done using regular expressions (for example, to check local parts
6851 of addresses in From: header lines).
6853 (2) The list of non-personal local parts in From: addresses has been
6854 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6855 taken from the Sieve specification recommendations.
6857 (3) If the message contains any header line starting with "List-" it is
6858 treated as non-personal.
6860 (4) The test for "circular" in the Subject: header line has been removed
6861 because it now seems ill-conceived.
6863 15. Minor typos in src/EDITME comments corrected.
6865 16. Installed latest exipick from John Jetmore.
6867 17. If headers_add on a router specified a text string that was too long for
6868 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6869 of string_sprintf() is now avoided.
6871 18. $message_body_size was not set (it was always zero) when running the DATA
6872 ACL and the local_scan() function.
6874 19. For the "mail" command in an Exim filter, no default was being set for
6875 the once_repeat time, causing a random time value to be used if "once" was
6876 specified. (If the value happened to be <= 0, no repeat happened.) The
6877 default is now 0s, meaning "never repeat". The "vacation" command was OK
6878 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6879 (I found it when inspecting the code).
6881 20. There is now an overall timeout for performing a callout verification. It
6882 defaults to 4 times the callout timeout, which applies to individual SMTP
6883 commands during the callout. The overall timeout applies when there is more
6884 than one host that can be tried. The timeout is checked before trying the
6885 next host. This prevents very long delays if there are a large number of
6886 hosts and all are timing out (e.g. when the network connections are timing
6887 out). The value of the overall timeout can be changed by specifying an
6888 additional sub-option for "callout", called "maxwait". For example:
6890 verify = sender/callout=5s,maxwait=20s
6892 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6893 to the end before writing, but this should make it even safer).
6895 22. Exim was forgetting that it had advertised PIPELINING for the second and
6896 subsequent messages on an SMTP connection. It was also not resetting its
6897 memory on STARTTLS and an internal HELO.
6899 23. When Exim logs an SMTP synchronization error within a session, it now
6900 records whether PIPELINING has been advertised or not.
6902 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6903 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6904 rather than long int.
6906 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6908 26. Added the never_mail option to autoreply.
6914 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6915 crash if the getsockname() call failed; this can happen if a connection is
6916 closed very soon after it is established. The problem was simply in the
6917 order in which certain operations were done, causing Exim to try to write
6918 to the SMTP stream before it had set up the file descriptor. The bug has
6919 been fixed by making things happen in the correct order.
6925 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6926 before the connection was closed, thus losing the rejection response.
6928 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6929 some early Solaris releases, but causes trouble in current releases where
6930 socklen_t is defined.
6932 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6935 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6938 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6939 that releasing the top of it at the end releases what was used for sub-
6940 expansions (unless the block got too big). However, discard this block if
6941 the first thing is a variable or header, so that we can use its block when
6942 it is dynamic (useful for very large $message_headers, for example).
6944 6. Lookups now cache *every* query, not just the most recent. A new, separate
6945 store pool is used for this. It can be recovered when all lookup caches are
6946 flushed. Lookups now release memory at the end of their result strings.
6947 This has involved some general refactoring of the lookup sources.
6949 7. Some code has been added to the store_xxx() functions to reduce the amount
6950 of flapping under certain conditions.
6952 8. log_incoming_interface used to affect only the <= reception log lines. Now
6953 it causes the local interface and port to be added to several more SMTP log
6954 lines, for example "SMTP connection from", and rejection lines.
6956 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6958 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6960 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6961 could be overwritten at the end of the current message (or the start of a
6962 new message if it was set in a HELO ACL). The value is now preserved for
6963 the duration of the SMTP connection.
6965 12. If a transport had a headers_rewrite setting, and a matching header line
6966 contained an unqualified address, that address was qualified, even if it
6967 did not match any rewriting rules. The underlying bug was that the values
6968 of the flags that permit the existence of unqualified sender and recipient
6969 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6970 non-local messages, and by -bnq for local messages) were not being
6971 preserved with the message after it was received.
6973 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6974 "next input=" as part of the text comprising the host identity instead of
6975 the correct text. The code was using the same buffer for two different
6976 strings. However, depending on which order the printing function evaluated
6977 its arguments, the bug did not always show up. Under Linux, for example, my
6978 test suite worked just fine.
6980 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6981 doesn't work with some older versions of Perl. It has been changed to "my",
6982 which in any case is probably the better facility to use.
6984 15. A really picky compiler found some instances of statements for creating
6985 error messages that either had too many or two few arguments for the format
6988 16. The size of the buffer for calls to the DNS resolver has been increased
6989 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6990 for addresses that have a lot of PTR records. This alleviates a problem; it
6991 does not fully solve it.
6993 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6994 buffer now truncates the list and logs the incident, which is the same
6995 action as happens when Exim is looking up a host name and its aliases.
6996 Previously in this situation something unpredictable would happen;
6997 sometimes it was "internal error: store_reset failed".
6999 18. If a server dropped the connection unexpectedly when an Exim client was
7000 using GnuTLS and trying to read a response, the client delivery process
7001 crashed while trying to generate an error log message.
7003 19. If a "warn" verb in an ACL added multiple headers to a message in a single
7004 string, for example:
7006 warn message = H1: something\nH2: something
7008 the text was added as a single header line from Exim's point of view
7009 though it ended up OK in the delivered message. However, searching for the
7010 second and subsequent header lines using $h_h2: did not work. This has been
7011 fixed. Similarly, if a system filter added multiple headers in this way,
7012 the routers could not see them.
7014 20. Expanded the error message when iplsearch is called with an invalid key to
7015 suggest using net-iplsearch in a host list.
7017 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7018 ACLs are no longer actually imposed (and a message to that effect is
7021 22. If a "gecos" field in a passwd entry contained escaped characters, in
7022 particular, if it contained a \" sequence, Exim got it wrong when building
7023 a From: or a Sender: header from that name. A second bug also caused
7024 incorrect handling when an unquoted " was present following a character
7025 that needed quoting.
7027 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7028 was not being matched caselessly.
7030 24. Arranged for all hyphens in the exim.8 source to be escaped with
7033 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7034 properly. Recipient callout cache records were still being keyed to include
7035 the sender, even when use_sender was set false. This led to far more
7036 callouts that were necessary. The sender is no longer included in the key
7037 when use_sender is false.
7039 26. Added "control = submission" modifier to ACLs.
7041 27. Added the ${base62d: operator to decode base 62 numbers.
7043 28. dnsdb lookups can now access SRV records.
7045 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7046 the configuration file.
7048 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7049 (-v) mode. This makes the output for a verbose queue run more intelligible.
7051 31. Added a use_postmaster feature to recipient callouts.
7053 32. Added the $body_zerocount variable, containing the number of binary zero
7054 bytes in the message body.
7056 33. The time of last modification of the "new" subdirectory is now used as the
7057 "mailbox time last read" when there is a quota error for a maildir
7060 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7062 35. Added +ignore_unknown as a special item in host lists.
7064 36. Code for decoding IPv6 addresses in host lists is now included, even if
7065 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7066 address was recognized as an IP address, but was then not correctly decoded
7067 into binary, causing unexpected and incorrect effects when compared with
7074 1. Very minor rewording of debugging text in manualroute to say "list of
7075 hosts" instead of "hostlist".
7077 2. If verify=header_syntax was set, and a header line with an unqualified
7078 address (no domain) and a large number of spaces between the end of the
7079 name and the colon was received, the reception process suffered a buffer
7080 overflow, and (when I tested it) crashed. This was caused by some obsolete
7081 code that should have been removed. The fix is to remove it!
7083 3. When running in the test harness, delay a bit after writing a bounce
7084 message to get a bit more predictability in the log output.
7086 4. Added a call to search_tidyup() just before forking a reception process. In
7087 theory, someone could use a lookup in the expansion of smtp_accept_max_
7088 per_host which, without the tidyup, could leave open a database connection.
7090 5. Added the variables $recipient_data and $sender_data which get set from a
7091 lookup success in an ACL "recipients" or "senders" condition, or a router
7092 "senders" option, similar to $domain_data and $local_part_data.
7094 6. Moved the writing of debug_print from before to after the "senders" test
7097 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7098 problems for message scanning, either using a data ACL, or using
7099 local_scan() because the Received: header was not generated till after they
7100 were called (in order to set the time as the time of reception completion).
7101 I have revised the way this works. The header is now generated after the
7102 body is received, but before the ACL or local_scan() are called. After they
7103 are run, the timestamp in the header is updated.
7109 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7110 before starting a queue runner without re-exec. This happened only when
7111 deliver_drop_privilege was set or when the Exim user was set to root. The
7112 effect of the bug was that timeouts during subsequent deliveries caused
7113 crashes instead of being properly handled. The handler is now left at its
7114 default (and expected) setting.
7116 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7117 message, again when deliver_drop_privilege is set or Exim is run as root.
7118 The bug described in (1) was not present in this case, but the tidying up
7119 of the other signals was missing. I have made the two cases consistent.
7121 3. The ignore_target_hosts setting on a manualroute router was being ignored
7122 for hosts that were looked up using the /MX notation.
7124 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7127 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7128 operated on the sender address. After changing the $sender_address to <>
7129 for the sender address verify, Exim was re-instated it as the original
7130 (before rewriting) address, but remembering that it had rewritten it, so it
7131 wasn't rewriting it again. This bug also had the effect of breaking the
7132 sender address verification caching when the sender address was rewritten.
7134 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7135 This has been changed so that if the ip literal address matches
7136 ignore_target_hosts, the router declines.
7138 7. Added expansion conditions match_domain, match_address, and match_local_
7139 part (NOT match_host).
7141 8. The placeholder for the Received: header didn't have a length field set.
7143 9. Added code to Exim itself and to exim_lock to test for a specific race
7144 condition that could lead to file corruption when using MBX delivery. The
7145 issue is with the lockfile that is created in /tmp. If this file is removed
7146 after a process has opened it but before that process has acquired a lock,
7147 there is the potential for a second process to recreate the file and also
7148 acquire a lock. This could lead to two Exim processes writing to the file
7149 at the same time. The added code performs the same test as UW imapd; it
7150 checks after acquiring the lock that its file descriptor still refers to
7151 the same named file.
7153 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7154 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7157 11. Added the smtp_active_hostname option. If used, this will typically be made
7158 to depend on the incoming interface address. Because $interface_address is
7159 not set up until the daemon has forked a reception process, error responses
7160 that can happen earlier (such as "too many connections") no longer contain
7163 12. If an expansion in a condition on a "warn" statement fails because a lookup
7164 defers, the "warn" statement is abandoned, and the next ACL statement is
7165 processed. Previously this caused the whole ACL to be aborted.
7167 13. Added the iplsearch lookup type.
7169 14. Added ident_timeout as a log selector.
7171 15. Added tls_certificate_verified as a log selector.
7173 16. Added a global option tls_require_ciphers (compare the smtp transport
7174 option of the same name). This controls incoming TLS connections.
7176 17. I finally figured out how to make tls_require_ciphers do a similar thing
7177 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7178 before starting the TLS session.
7180 18. Tabs are now shown as \t in -bP output.
7182 19. If the log selector return_path_on_delivery was set, Exim crashed when
7183 bouncing a message because it had too many Received: header lines.
7185 20. If two routers both had headers_remove settings, and the first one included
7186 a superfluous trailing colon, the final name in the first list and the
7187 first name in the second list were incorrectly joined into one item (with a
7188 colon in the middle).
7194 1. Added -C and -D options to the exinext utility, mainly to make it easier
7195 to include in the automated testing, but these could be helpful when
7196 multiple configurations are in use.
7198 2. The exinext utility was not formatting the output nicely when there was
7199 an alternate port involved in the retry record key, nor when there was a
7200 message id as well (for retries that were specific to a specific message
7201 and a specific host). It was also confused by IPv6 addresses, because of
7202 the additional colons they contain. I have fixed the IPv4 problem, and
7203 patched it up to do a reasonable job for IPv6.
7205 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7206 delivery, the log line now contains "pipelined" if PIPELINING was used.
7208 4. An SMTP transport process used to panic and die if the bind() call to set
7209 an explicit outgoing interface failed. This has been changed; it is now
7210 treated in the same way as a connect() failure.
7212 5. A reference to $sender_host_name in the part of a conditional expansion
7213 that was being skipped was still causing a DNS lookup. This no longer
7216 6. The def: expansion condition was not recognizing references to header lines
7217 that used bh_ and bheader_.
7219 7. Added the _cache feature to named lists.
7221 8. The code for checking quota_filecount in the appendfile transport was
7222 allowing one more file than it should have been.
7224 9. For compatibility with Sendmail, the command line option
7232 and sets the incoming protocol and host name (for trusted callers). The
7233 host name and its colon can be omitted when only the protocol is to be set.
7234 Note the Exim already has two private options, -pd and -ps, that refer to
7235 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7236 "s", but I don't think that's a major issue.
7238 10. A number of refactoring changes to the code, none of which should affect
7241 (a) The number of logging options was getting close to filling up the
7242 32-bit word that was used as a bit map. I have split them into two classes:
7243 those that are passed in the argument to log_write(), and those that are
7244 only ever tested independently outside of that function. These are now in
7245 separate 32-bit words, so there is plenty of room for expansion again.
7246 There is no change in the user interface or the logging behaviour.
7248 (b) When building, for example, log lines, the code previously used a
7249 macro that called string_cat() twice, in order to add two strings. This is
7250 not really sufficiently general. Furthermore, there was one instance where
7251 it was actually wrong because one of the argument was used twice, and in
7252 one call a function was used. (As it happened, calling the function twice
7253 did not affect the overall behaviour.) The macro has been replaced by a
7254 function that can join an arbitrary number of extra strings onto a growing
7257 (c) The code for expansion conditions now uses a table and a binary chop
7258 instead of a serial search (which was left over from when there were very
7259 few conditions). Also, it now recognizes conditions like "pam" even when
7260 the relevant support is not compiled in: a suitably worded error message is
7261 given if an attempt is made to use such a condition.
7263 11. Added ${time_interval:xxxxx}.
7265 12. A bug was causing one of the ddress fields not to be passed back correctly
7266 from remote delivery subprocesses. The field in question was not being
7267 subsequently used, so this caused to problems in practice.
7269 13. Added new log selectors queue_time and deliver_time.
7271 14. Might have fixed a bug in maildirsizefile handling that threw up
7272 "unexpected character" debug warnings, and recalculated the data
7273 unnecessarily. In any case, I expanded the warning message to give more
7276 15. Added the message "Restricted characters in address" to the statements in
7277 the default ACL that block characters like @ and % in local parts.
7279 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7280 Three changes have been made:
7282 (a) There was a serious bug; a negative response to MAIL caused the whole
7283 recipient domain to be cached as invalid, thereby blocking all messages
7284 to all local parts at the same domain, from all senders. This bug has
7285 been fixed. The domain is no longer cached after a negative response to
7286 MAIL if the sender used is not empty.
7288 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7291 (c) A new callout option, "use_sender" has been added for people who want
7292 the modified behaviour.
7298 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7301 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7302 indeed breaks things for older releases.
7304 3. Added additional logging to the case where there is a problem reading data
7305 from a filter that is running in a subprocess using a pipe, in order to
7306 try to track down a specific problem.
7308 4. Testing facility fudge: when running in the test harness and attempting
7309 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7310 getting "No route to host". Convert this to a timeout.
7312 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7315 6. Some OS don't have socklen_t but use size_t instead. This affects the
7316 fifth argument of getsockopt() amongst other things. This is now
7317 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7318 can be set for individual OS. I have set it for SunOS5, OSF1, and
7319 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7320 some earlier ones do not.
7322 7. Change 4.30/15 was not doing the test caselessly.
7324 8. The standard form for an IPv6 address literal was being rejected by address
7325 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7326 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7327 this, as well as the form without the "IPv6" on the front (but only when
7328 address literals are enabled, of course).
7330 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7332 10. Exim crashed if a message with an empty sender address specified by -f
7333 encountered a router with an errors_to setting. This could be provoked only
7334 by a command such as
7338 where an empty string was supplied; "<>" did not hit this bug.
7340 11. Installed PCRE release 4.5.
7342 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7343 remained set. It is now erased.
7345 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7346 times from message ids (which are base 36 rather than the normal 62).
7348 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7349 were being counted as actual protocol errors, and logged if the log
7350 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7351 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7352 and DATA following a set of rejected RCPTs do not count as protocol errors.
7353 In other words, Exim assumes they were pipelined, though this may not
7354 actually be the case. Of course, in all cases the client gets an
7355 appropriate error code.
7357 15. If a lookup fails in an ACL condition, a message about the failure may
7358 be available; it is used if testing the ACL cannot continue, because most
7359 such messages specify what the cause of the deferral is. However, some
7360 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7361 that caused an old message to be retained and used if a later statement
7362 caused a defer, replacing the real cause of the deferral.
7364 16. If an IP address had so many PTR records that the DNS lookup buffer
7365 was not large enough to hold them, Exim could crash while trying to process
7366 the truncated data. It now detects and logs this case.
7368 17. Further to 4.21/58, another change has been made: if (and only if) the
7369 first line of a message (the first header line) ends with CRLF, a bare LF
7370 in a subsequent header line has a space inserted after it, so as not to
7371 terminate the header.
7373 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7374 unnecessarily, used atoi() instead of strtol(), and didn't check the
7375 termination when getting file sizes from file names by regex.
7377 19. Completely re-implemented the support for maildirsize files, in the light
7378 of a number of problems with the previous contributed implementation
7379 (4.30/29). In particular:
7381 . If the quota is zero, the maildirsize file is maintained, but no quota is
7384 . If the maildir directory does not exist, it is created before any attempt
7385 to write a maildirsize file.
7387 . The quota value in the file is just a cache; if the quota is changed in
7388 the transport, the new value overrides.
7390 . A regular expression is available for excluding directories from the
7393 20. The autoreply transport checks the characters in options that define the
7394 message's headers; it allows continued headers, but it was checking with
7395 isspace() after an embedded newline instead of explicitly looking for a
7398 21. If all the "regular" hosts to which an address was routed had passed their
7399 expiry times, and had not reached their retry times, the address was
7400 bounced, even if fallback hosts were defined. Now Exim should go on to try
7403 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7404 equivalent code in the SMTP transport. Some hosts send humungous responses
7405 to HELO/EHLO, more than 1024 it seems.
7407 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7408 gives compiler warnings in some environments. I've now done it "properly",
7411 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7412 (because of problems with the built-in one) was declared to return uschar *
7413 instead of char *, causing compiler failure.
7415 25. Fixed a file descriptor leak when processing alias/forward files.
7417 26. Fixed a minor format string issue in dbfn.c.
7419 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7421 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7422 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7425 29. When a sender address is verified, it is cached, to save repeating the test
7426 when there is more than one recipient in a message. However, when the
7427 verification involves a callout, it is possible for different callout
7428 options to be set for different recipients. It is too complicated to keep
7429 track of this in the cache, so now Exim always runs a verification when a
7430 callout is required, relying on the callout cache for the optimization.
7431 The overhead is duplication of the address routing, but this should not be
7434 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7435 to be verified with callout=postmaster, and the main callout worked but the
7436 postmaster check failed, the verification correctly failed. However, if a
7437 subsequent RCPT command asked for sender verification *without* the
7438 postmaster check, incorrect caching caused this verification also to fail,
7441 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7442 it was not caching the DNS options (qualify_single, search_parents) that
7443 were used when the lookup failed. A subsequent lookup with different
7444 options therefore always gave the same answer, though there were cases
7445 where it should not have. (Example: a "domains = !$mx_any" option on a
7446 dnslookup router: the "domains" option is always processed without any
7447 widening, but the router might have qualify_single set.) Now Exim uses the
7448 cached value only when the same options are set.
7450 32. Added John Jetmore's "exipick" utility to the distribution.
7452 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7453 than a timeout (e.g. a certificate is required, and is not provided), an
7454 Exim server now closes the connection immediately. Previously it waited for
7455 the client to close - but if the client is SSL, it seems that they each
7456 wait for each other, leading to a delay before one of them times out.
7458 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7459 maintained 0.8.x compatibility because I don't think many are using it, and
7460 it is clearly obsolete.
7462 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7465 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7466 client certificate was expired. A simple patch fixes this, though I don't
7467 understand the full logic of why the verify callback is called multiple
7470 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7471 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7472 which causes problems with some clients (such as the Certicom SSL Plus
7473 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7474 disables the coutermeasure allowing Eudora to connect."
7476 38. Exim was not checking that a write() to a log file succeeded. This could
7477 lead to Bad Things if a log got too big, in particular if it hit a file
7478 size limit. Exim now panics and dies if it cannot write to a log file, just
7479 as it does if it cannot open a log file.
7481 39. Modified OS/Makefile-Linux so that it now contains
7483 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7485 The two -D definitions ensure that Exim is compiled with large file
7486 support, which makes it possible to handle log files that are bigger than
7489 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7490 instance) a domain was checked against a named list that involved a lookup,
7491 causing $domain_data to be set, then another domain was checked against the
7492 same list, then the first domain was re-checked, the value of $domain_data
7493 after the final check could be wrong. In particular, if the second check
7494 failed, it could be set empty. This bug probably also applied to
7497 41. The strip_trailing_dot option was not being applied to the address given
7498 with the -f command-line option.
7500 42. The code for reading a message's header from the spool was incrementing
7501 $received_count, but never initializing it. This meant that the value was
7502 incorrect (doubled) while delivering a message in the same process in which
7503 it was received. In the most common configuration of Exim, this never
7504 happens - a fresh exec is done - but it can happen when
7505 deliver_drop_privilege is set.
7507 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7508 it now includes up to 150 characters of the unexpected data in the log
7511 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7512 and building data strings. The size of both of these buffers was 10 000
7513 bytes - far larger than anybody would *ever* want, thought I. Needless to
7514 say, somebody hit the limit. I have increased the maximum line length to
7515 20 000 and the maximum data length of concatenated lines to 100 000. I have
7516 also fixed two bugs, because there was no checking on these buffers. Tsk,
7517 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7518 buffer is too small.
7520 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7521 lsearch lookups. Now it does.
7523 46. When parsing a route_list item in a manualroute router, a fixed-length
7524 buffer was used for the list of hosts. I made this 1024 bytes long,
7525 thinking that nobody would ever have a list of hosts that long. Wrong.
7526 Somebody had a whole pile of complicated expansion conditions, and the
7527 string was silently truncated, leading to an expansion error. It turns out
7528 that it is easier to change to an unlimited length (owing to other changes
7529 that have happened since this code was originally written) than to build
7530 structure for giving a limitation error. The length of the item that
7531 expands into the list of hosts is now unlimited.
7533 47. The lsearch lookup could not handle data where the length of text line was
7534 more than 4095 characters. Such lines were truncated, leading to shortened
7535 data being returned. It should now handle lines of any length.
7537 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7538 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7541 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7542 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7543 no longer happens in comment lines. A list of replacements is now placed
7544 at the head of all of the source files, except those whose only change is
7545 to replace PERL_COMMAND in the very first #! line.
7547 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7548 messages on the queue, with a bottom-up merge sort, using code contributed
7549 by Michael Haardt. This should make operations like -bp somewhat faster on
7550 large queues. It won't affect queue runners, except when queue_run_in_order
7553 51. Installed eximstats 1.31 in the distribution.
7555 52. Added support for SRV lookups to the dnslookup router.
7557 53. If an ACL referred to $message_body or $message_body_end, the value was not
7558 reset for any messages that followed in the same SMTP session.
7560 54. The store-handling optimization for building very long strings was not
7561 differentiating between the different store pools. I don't think this
7562 actually made any difference in practice, but I've tidied it.
7564 55. While running the routers to verify a sender address, $sender_address
7565 was still set to the sender address. This is wrong, because when routing to
7566 send a bounce to the sender, it would be empty. Therefore, I have changed
7567 it so that, while verifying a sender address, $sender_address is set to <>.
7568 (There is no change to what happens when verifying a recipient address.)
7570 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7571 target A or AAAA records (if not already returned) without resetting the
7572 qualify_single or search_parents options of the DNS resolver. These are
7573 inappropriate in this case because the targets of MX and SRV records must
7574 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7575 target that, when qualified, matched something in the local domain. These
7576 two options are now turned off when doing these lookups.
7578 57. It seems that at least some releases of Reiserfs (which does not have the
7579 concept of a fixed number of inodes) returns zero and not -1 for the
7580 number of available inodes. This interacted badly with check_spool_inodes,
7581 which assumed that -1 was the "no such thing" setting. What I have done is
7582 to check that the total number of inodes is greater than zero before doing
7583 the test of how many are available.
7585 58. When a "warn" ACL statement has a log_message modifier, the message is
7586 remembered, and not repeated. This is to avoid a lot of repetition when a
7587 message has many recipients that cause the same warning to be written.
7588 However, Exim was preserving the list of already written lines for an
7589 entire SMTP session, which doesn't seem right. The memory is now reset if a
7590 new message is started.
7592 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7593 debugging output unless log_rewrite was also set.
7595 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7596 of (int)(handle) when we know that handle contains (void *)(-1).
7598 61. The Exim daemon panic-logs an error return when it closes the incoming
7599 connection. However "connection reset by peer" seems to be common, and
7600 isn't really an error worthy of noting specially, so that particular error
7603 62. When Exim is trying to find all the local interfaces, it used to panic and
7604 die if the ioctl to get the interface flags failed. However, it seems that
7605 on at least one OS (Solaris 9) it is possible to have an interface that is
7606 included in the list of interfaces, but for which you get a failure error
7607 for this call. This happens when the interface is not "plumbed" into a
7608 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7609 failure of the "get flags" call assumes that the interface is down.
7611 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7612 makes life easier for people who are doing arithmetic on fields extracted
7613 from dates, where you often get leading zeros that should not be
7614 interpreted as octal.
7616 64. Added qualify_domain to the redirect router, to override the global
7619 65. If a pathologically long header line contained very many addresses (the
7620 report of this problem mentioned 10 000) and each of them was rewritten,
7621 Exim could use up a very large amount of memory. (It kept on making new
7622 copies of the header line as it rewrote, and never released the old ones.)
7623 At the expense of a bit more processing, the header rewriting function has
7624 been changed so that it no longer eats memory in this way.
7626 66. The generation of the Received: header has been moved from the time that a
7627 message starts to be received, to the time that it finishes. The timestamp
7628 in the Received: header should now be very close to that of the <= log
7629 line. There are two side-effects of this change:
7631 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7632 logged header lines no longer include the local Received: line, because
7633 it has not yet been created. The same applies to a copy of the message
7634 that is returned to a non-SMTP sender when a message is rejected.
7636 (b) When a filter file is tested using -bf, no additional Received: header
7637 is added to the test message. After some thought, I decided that this
7640 This change does not affect the value of $received_for. It is still set
7641 after address rewriting, but before local_scan() is called.
7643 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7645 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7646 gave an unhelpful panic error message, and a defer error. I have managed to
7647 change this behaviour so that it now rejects any supplied certificate,
7648 which seems right, as the list of acceptable certificates is empty.
7650 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7651 gave an unhelpful defer error. I have not managed to make this reject any
7652 supplied certificates, but the error message it gives is "no certificate
7653 supplied", which is not helpful.
7655 70. exigrep's output now also includes lines that are not associated with any
7656 message, but which match the given pattern. Implemented by a patch from
7657 Martin Sluka, which also tidied up the Perl a bit.
7659 71. Recipient callout verification, like sender verification, was using <> in
7660 the MAIL FROM command. This isn't really the right thing, since the actual
7661 sender may affect whether the remote host accepts the recipient or not. I
7662 have changed it to use the actual sender in the callout; this means that
7663 the cache record is now keyed on a recipient/sender pair, not just the
7664 recipient address. There doesn't seem to be a real danger of callout loops,
7665 since a callout by the remote host to check the sender would use <>.
7666 [SEE ABOVE: changed after hitting problems.]
7668 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7669 temporary errors. However, in the case of such a code being given after
7670 the end of a data transmission (i.e. after ".") Exim was failing to write
7671 a retry record for the message. (Yes, there was some broken host that was
7672 actually sending 8xx at this point.)
7674 73. An unknown lookup type in a host list could cause Exim to panic-die when
7675 the list was checked. (An example that provoked this was putting <; in the
7676 middle of a list instead of at the start.) If this happened during a DATA
7677 ACL check, a -D file could be left lying around. This kind of configuration
7678 error no longer causes Exim to die; instead it causes a defer error. The
7679 incident is still logged to the main and panic logs.
7681 74. Buglet left over from Exim 3 conversion. The message "too many messages
7682 in one connection" was written to the rejectlog but not the mainlog, except
7683 when address rewriting (yes!) was being logged.
7685 75. Added write_rejectlog option.
7687 76. When a system filter was run not as root (that is, when system_filter_user
7688 was set), the values of the $n variables were not being returned to the
7689 main process; thus, they were not subsequently available in the $sn
7692 77. Added +return_path_on_delivery log selector.
7694 78. A connection timeout was being treated differently from recipients deferred
7695 when testing hosts_max_try with a message that was older than the host's
7696 retry timeout. (The host should not be counted, thus allowing all hosts to
7697 be tried at least once before bouncing.) This may have been the cause of an
7698 occasionally reported bug whereby a message would remain on the queue
7699 longer than the retry timeout, but would be bounced if a delivery was
7700 forced. I say "may" because I never totally pinned down the problem;
7701 setting up timeout/retry tests is difficult. See also the next item.
7703 79. The ultimate address timeout was not being applied to errors that involved
7704 a combination of host plus message (for example, a timeout on a MAIL
7705 command). When an address resolved to a number of possible hosts, and they
7706 were not all tried for each delivery (e.g. because of hosts_max_try), a
7707 message could remain on the queue longer than the retry timeout.
7709 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7712 81. Fixed an obscure SMTP outgoing bug which required at least the following
7713 conditions: (a) there was another message waiting for the same server;
7714 (b) the server returned 5xx to all RCPT commands in the first message so
7715 that the message was not completed; (c) the server dropped the connection
7716 or gave a negative response to the RSET that Exim sends to abort the
7717 transaction. The observed case was a dropped connection after DATA that had
7718 been sent in pipelining mode. That is, the server had advertised PIPELINING
7719 but was not implementing it correctly. The effect of the bug was incorrect
7720 behaviour, such as trying another host, and this could lead to a crash.
7726 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7727 and daemon.c were passed as pointers to ints; they should have been
7728 pointers to socklen_t variables (which are typically unsigned ints).
7730 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7733 3. Fixed a really odd bug that affected only the testing scheme; patching a
7734 certain fixed string in the binary changed the value of another string that
7735 happened to be identical to the end of the original first string.
7737 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7738 name", it returns that address as the IP address. On some operating
7739 systems (e.g. Solaris), it also passes back the IP address string as the
7740 "host name". However, on others (e.g. Linux), it passes back an empty
7741 string. Exim wasn't checking for this, and was changing the host name to an
7742 empty string, assuming it had been canonicalized.
7744 5. Although rare, it is permitted to have more than one PTR record for a given
7745 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7746 all the names associated with an address, because they do in Solaris.
7747 However, it seems that they do not in Linux for data that comes from the
7748 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7749 I found this out when I moved to a new Linux workstation and tried to run
7750 the Exim test suite.
7752 To get round this problem I have changed the code so that it now does its
7753 own call to the DNS to look up PTR records when searching for a host name.
7754 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7755 addresses that are only in /etc/hosts are still found.
7757 This behaviour is, however, controlled by an option called host_lookup_
7758 order, which defaults to "bydns:byaddr". If people want to use the other
7759 order, or indeed, just use one or the other means of lookup, they can
7760 specify it in this variable.
7762 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7763 some operating systems, this comes back from gethostbyaddr() as an empty
7764 string, and this is what Exim used to test for. However, it seems that in
7765 other systems, "." is yielded. Exim now tests for this case too.
7767 7. The values of check_spool_space and check_log_space are now held internally
7768 as a number of kilobytes instead of an absolute number of bytes. If a
7769 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7770 kilobyte. This means that much larger values can be stored.
7772 8. Exim monitor: an attempt to get the action menu when not actually pointing
7773 at a message produces an empty menu entitled "No message selected". This
7774 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7775 no entries in it ("Shell widget menu has zero width and/or height"). So I
7776 have added a single, blank menu entry in this case.
7778 9. Added ${quote_local_part.
7780 10. MIME decoding is now applied to the contents of Subject: header lines when
7783 11. Now that a reference to $sender_host_address automatically causes a reverse
7784 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7785 host lookup before query-style lookups in lists that might use this
7786 variable. This has therefore been abolished, and the "net-" prefix is no
7787 longer necessary for query-style lookups.
7789 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7790 have been a typo for LFLAGS, so it has been changed.
7792 13. The install script calls Exim with "-C /dev/null" in order to find the
7793 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7794 to be output. However, since Exim outputs its version number before the
7795 error, it didn't break the script. It just looked ugly. I fixed this by
7796 always allowing "-C /dev/null" if the caller is root.
7798 14. Ignore overlarge ACL variable number when reading spool file - insurance
7799 against a later release with more variables having written the file.
7801 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7802 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7803 this, as well as the form without the "IPv6" on the front.
7805 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7806 OS/Makefile-Darwin file.
7808 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7809 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7811 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7812 variable. However, some OS use ints for inodes. I've added cast to long int
7813 to get rid of the compiler warning.
7815 19. I had forgotten to lock out "/../" in configuration file names when
7816 ALT_CONFIG_PREFIX was set.
7818 20. Routers used for verification do not need to specify transports. However,
7819 if such a router generated a host list, and callout was configured, Exim
7820 crashed, because it could not find a port number from the (non-existent)
7821 transport. It now assumes port 25 in this circumstance.
7823 21. Added the -t option to exigrep.
7825 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7826 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7827 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7828 them is set, LOOKUP_LSEARCH is forced.
7830 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7832 24. Added sender and host information to the "rejected by local_scan()" log
7833 line; previously there was no indication of these.
7835 25. Added .include_if_exists.
7837 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7838 Linux. It turns out that not all file systems support this. Apparently some
7839 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7840 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7841 not supported on the file descriptor, is now ignored when Exim is trying to
7842 sync a directory. This applies only to Linux.
7844 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7846 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7847 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7850 29. Added support for maildirsize files from supplied patch (modified a bit).
7852 30. The use of :fail: followed by an empty string could lead Exim to respond to
7853 sender verification failures with (e.g.):
7855 550 Verification failed for <xxx>
7856 550 Sender verify failed
7858 where the first response line was missing the '-' that indicates it is not
7859 the final line of the response.
7861 31. The loop for finding the name of the user that called Exim had a hardwired
7862 limit of 10; it now uses the value of finduser_retries, which is used for
7863 all other user lookups.
7865 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7868 33. Exim was neglecting to zero errno before one call of strtol() when
7869 expanding a string and expecting an integer value. On some systems this
7870 resulted in spurious "integer overflow" errors. Also, it was casting the
7871 result into an int without checking.
7873 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7874 did not work. The code looks as if it has *never* worked, though it appears
7875 to have been documented since at least release 1.62. I have made it work.
7877 35. The "timeout_DNS" error in retry rules, also documented since at least
7878 1.62, also never worked. As it isn't clear exactly what this means, and
7879 clearly it isn't a major issue, I have abolished the feature by treating it
7880 as "timeout", and writing a warning to the main and panic logs.
7882 36. The display of retry rules for -brt wasn't always showing the error code
7885 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7886 timeout_connect_A, timeout_connect_MX.
7888 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7889 to the empty sender.
7891 39. The daemon was not analysing the content of -oX till after it had closed
7892 stderr and disconnected from the controlling terminal. This meant that any
7893 syntax errors were only noted on the panic log, and the return code from
7894 the command was 0. By re-arranging the code a little, I've made the
7895 decoding happen first, so such errors now appear on stderr, and the return
7896 code is 1. However, the actual setting up of the sockets still happens in
7897 the disconnected process, so errors there are still only recorded on the
7900 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7901 connections (as happens on some IP stacks) was logged at start up time as
7902 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7903 it from "IPv6 and IPv4", which means that two separate sockets are being
7906 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7907 says whether AF_INET or AF_INET6 was passed as an argument.
7909 42. Exiwhat output was messed up when time zones were included in log
7912 43. Exiwhat now gives more information about the daemon's listening ports,
7913 and whether -tls-on-connect was used.
7915 44. The "port" option of the smtp transport is now expanded.
7917 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7918 silently ignored. Now an error message is written to the main and panic
7921 46. There's a new ACL modifier called "logwrite" which writes to a log file
7922 as soon as it is encountered.
7924 47. Added $local_user_uid and $local_user_gid at routing time.
7926 48. Exim crashed when trying to verify a sender address that was being
7929 49. Exim was recognizing only a space character after ".include". It now also
7930 recognizes a tab character.
7932 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7933 extracting the relevant information from the specification. The man page no
7934 longer contains scrambled data for the -d option, and I've added a section
7935 at the front about calling Exim under different names.
7937 51. Added "extra_headers" argument to the "mail" command in filter files.
7939 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7942 53. Installed eximstats 1.29.
7944 54. Added transport_filter_timeout as a generic transport option.
7946 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7947 Cc: header lines. This was required by RFC 822, but it not required by RFC
7950 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7951 incoming messages that did not have them. Now it does so only if the
7952 message originates locally, that is, if there is no associated remote host
7953 address. When Resent- header lines are present, this applies to the Resent-
7954 lines rather than the non-Resent- lines.
7956 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7957 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7959 58. Messages for configuration errors now include the name of the main
7960 configuration file - useful now that there may be more than one file in a
7961 list (.included file names were always shown).
7963 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7964 for those rare installations that do not start the daemon as root or run it
7965 setuid root. I've cut out the call to initgroups() if the daemon is not
7968 60. The Exim user and group can now be bound into the binary as text strings
7969 that are looked up at the start of Exim's processing.
7971 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7973 62. Added $mailstore_basename variable.
7975 63. Installed patch to sieve.c from Michael Haardt.
7977 64. When Exim failed to open the panic log after failing to open the main log,
7978 the original message it was trying to log was written to stderr and debug
7979 output, but if they were not available (the usual case in production), it
7980 was lost. Now it is written to syslog before the two lines that record the
7981 failures to open the logs.
7983 65. Users' Exim filters run in subprocesses under the user's uid. It is
7984 possible for a "deliver" command or an alias in a "personal" command to
7985 provoke an address rewrite. If logging of address rewriting is configured,
7986 this fails because the process is not running as root or exim. There may be
7987 a better way of dealing with this, but for the moment (because 4.30 needs
7988 to be released), I have disabled address rewrite logging when running a
7989 filter in a non-root, non-exim process.
7995 1. The buildconfig auxiliary program wasn't quoting the value set for
7996 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7997 not defined. This bug was masked in 4.22 by the effect that was fixed in
8000 2. Some messages that were rejected after a message id was allocated were
8001 shown as "incomplete" by exigrep. It no longer does this for messages that
8002 are rejected by local_scan() or the DATA or non-SMTP ACLs.
8004 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
8005 have allow_domain_literals set, the ID did not get logged in the <= line.
8006 Domain literals are now always recognized in Message-ID: header lines.
8008 4. The first argument for a ${extract expansion item is the key name or field
8009 number. Leading and trailing spaces in this item were not being ignored,
8010 causing some misleading effects.
8012 5. When deliver_drop_privilege was set, single queue runner processes started
8013 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8014 same command in the process it spins off) were not dropping privilege.
8016 6. When the daemon running as "exim" started a queue runner, it always
8017 re-executed Exim in the spun-off process. This is a waste of effort when
8018 deliver_drop_privilege is set. The new process now just calls the
8019 queue-runner function directly.
8025 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8028 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8029 ignored. Though the use of -oP was forcing the writing of a pid file, it
8030 was always written to the default place.
8032 3. If the message "no IP address found for host xxxx" is generated during
8033 incoming verification, it is now followed by identification of the incoming
8034 connection (so you can more easily find what provoked it).
8036 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8038 5. Added some features to "harden" Exim a bit more against certain attacks:
8040 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8041 be put in Local/Makefile. This is like the never_users runtime option,
8042 but it cannot be overridden. The default setting is "root".
8044 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8045 prefix string with which any file named in a -C command line option
8048 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8049 is retained for -C and -D only if the caller of Exim is root. Without
8050 it, the exim user may also use -C and -D and retain privilege.
8052 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8053 command line option is disabled.
8055 6. Macro names set by the -D option must start with an upper case letter, just
8056 like macro names defined in the configuration file.
8058 7. Added "dereference=" facility to LDAP.
8060 8. Two instances of the typo "uknown" in the source files are fixed.
8062 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8063 the Configure-Makefile script screwed up while processing it.
8065 10. Incorporated PCRE 4.4.
8067 11. The SMTP synchronization check was not operating right at the start of an
8068 SMTP session. For example, it could not catch a HELO sent before the client
8069 waited for the greeting. There is now a check for outstanding input at the
8070 point when the greeting is written. Because of the duplex, asynchronous
8071 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8072 way, but not yet received, when the check is performed.
8074 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8075 on TCP/IP sockets, because this apparently causes some broken clients to
8078 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8079 unchanged) from the Cygwin maintainer.
8081 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8082 was supported instead of testing for mbx. Effectively a typo.
8084 15. The spa authenticator server code was not checking that the input it
8085 received was valid base64.
8087 16. The debug output line for the "set" modifier in ACLs was not showing the
8088 name of the variable that was being set.
8090 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8092 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8093 lookup on its own. Something else was needed to trigger the lookup. For
8094 example, a match in host_lookup or the need for a host name in a host list.
8095 Now, if $sender_host_name is referenced and the host name has not yet been
8096 looked up, a lookup is performed. If the lookup fails, the variable remains
8097 empty, and $host_lookup_failed is set to "1".
8099 19. Added "eqi" as a case-independent comparison operator.
8101 20. The saslauthd authentication condition could segfault if neither service
8102 nor realm was specified.
8104 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8105 error message that was logged was misleading, and incoming SMTP
8106 connections were dropped. The message is now more accurate, and temporary
8107 errors are given to SMTP connections.
8109 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8110 (except RSET) with a 421 error, until QUIT is received. However, it was
8111 failing to send a response to QUIT.
8113 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8114 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8115 ACL is now run after the helo_try_verify_hosts code.
8117 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8118 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8119 case-independent, so other case variants are also recognized). Apparently
8120 some systems use these upper case variants.
8122 25. If more than two messages were waiting for the same host, and a transport
8123 filter was specified for the transport, Exim sent two messages over the
8124 same TCP/IP connection, and then failed with "socket operation on non-
8125 socket" when it tried to send the third.
8127 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8129 27. The extern definition of crypt16() in expand.c was not being excluded when
8130 the OS had its own crypt16() function.
8132 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8133 as a preferred synonym for return_size_limit, both as an option and as an
8136 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8138 30. Changed the default configuration ACL to relax the local part checking rule
8139 for addresses that are not in any local domains. For these addresses,
8140 slashes and pipe symbols are allowed within local parts, but the sequence
8141 /../ is explicitly forbidden.
8143 31. SPA server authentication was not clearing the challenge buffer before
8146 32. log_message in a "warn" ACL statement was writing to the reject log as
8147 well as to the main log, which contradicts the documentation and doesn't
8148 seem right (because no rejection is happening). So I have stopped it.
8150 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8151 However, I am unable to do any testing of this.
8153 34. Fixed an infelicity in the appendfile transport. When checking directories
8154 for a mailbox, to see if any needed to be created, it was accidentally
8155 using path names with one or more superfluous leading slashes; tracing
8156 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8158 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8159 contents are added to the log line that is written for every discarded
8160 recipient. (Previously a log_message setting was ignored.)
8162 36. The ${quote: operator now quotes the string if it is empty.
8164 37. The install script runs exim in order to find its version number. If for
8165 some reason other than non-existence or emptiness, which it checks, it
8166 could not run './exim', it was installing it with an empty version number,
8167 i.e. as "exim-". This error state is now caught, and the installation is
8170 38. An argument was missing from the function that creates an error message
8171 when Exim fails to connect to the socket for saslauthd authentication.
8172 This could cause Exim to crash, or give a corrupted message.
8174 39. Added isip, isip4, and isip6 to ${if conditions.
8176 40. The ACL variables $acl_xx are now saved with the message, and can be
8177 accessed later in routers, transports, and filters.
8179 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8180 strings in the file are not string-expanded.
8182 42. If a MAIL command specified a SIZE value that was too large to fit into an
8183 int variable, the check against message_size_limit failed. Such values are
8184 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8185 day this will have to be increased, but I don't think I want to be around
8186 when emails are that large.
8193 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8194 iconv() is not standard in FreeBSD.
8196 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8197 IPv6 enabled. The observed symptom was a segmentation fault on return from
8198 the function os_common_find_running_interfaces() in src/os.c.
8200 3. In the check_special_case() function in daemon.c I had used "errno" as an
8201 argument name, which causes warnings on some systems. This was basically a
8202 typo, since it was named "eno" in the comments!
8204 4. The code that waits for the clock to tick (at a resolution of some fraction
8205 of a second) so as to ensure message-id uniqueness was always waiting for
8206 at least one whole tick, when it could have waited for less. [This is
8207 almost certainly not relevant at current processor speeds, where it is
8208 unlikely to ever wait at all. But we try to future-proof.]
8210 5. The function that sleeps for a time interval that includes fractions of a
8211 second contained a race. It did not block SIGALRM between setting the
8212 timer, and suspending (a couple of lines later). If the interval was short
8213 and the sigsuspend() was delayed until after it had expired, the suspension
8214 never ended. On busy systems this could lead to processes getting stuck for
8217 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8218 process, before it forks any delivery processes. The open lookup caching
8219 mechanism meant that the open file or database connection was passed into
8220 the delivery process. The problem was that delivery processes always tidy
8221 up cached lookup data. This could cause a problem for the next delivery
8222 process started by the queue runner, because the external queue runner
8223 process does not know about the closure. So the next delivery process
8224 still has data in the lookup cache. In the case of a file lookup, there was
8225 no problem because closing a file descriptor in a subprocess doesn't affect
8226 the parent. However, if the lookup was caching a connection to a database,
8227 the connection was closed, and the second delivery process was likely to
8228 see errors such as "PGSQL: query failed: server closed the connection
8229 unexpectedly". The problem has been fixed by closing all cached lookups
8230 in a queue runner before running a delivery process.
8232 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8233 seem to have the "const" qualifier which it has on other OS. I've
8236 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8237 *of the same type* (client or server) with the same public name that an
8238 error should be diagnosed.
8240 9. When Exim looked up a host name for an IP address, but failed to find the
8241 original IP address when looking up the host name (a safety check), it
8242 output the message "<ip address> does not match any IP for NULL", which was
8243 confusing, to say the least. The bug was that the host name should have
8244 appeared instead of "NULL".
8246 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8247 user that is built into the binary, and the -C or -D options is used, root
8248 privilege is dropped before the configuration file is read. In addition,
8249 logging is switched to stderr instead of the normal log files. If the
8250 configuration then re-defines the Exim user, the unprivileged environment
8251 is probably not what is expected, so Exim logs a panic warning message (but
8254 However, if deliver_drop_privilege is set, the unprivileged state may well
8255 be exactly what is intended, so the warning has been cut out in that case,
8256 and Exim is allowed to try to write to its normal log files.
8262 1. smtp_return_error_details was not giving details for temporary sender
8263 or receiver verification errors.
8265 2. Diagnose a configuration error if two authenticators have the same public
8268 3. Exim used not to create the message log file for a message until the first
8269 delivery attempt. This could be confusing when incoming messages were held
8270 for policy or load reasons. The message log file is now created at the time
8271 the message is received, and an initial "Received" line is written to it.
8273 4. The automatically generated man page for command line options had a minor
8274 bug that caused no ill effects; however, a more serious problem was that
8275 the procedure for building the man page automatically didn't always
8276 operate. Consequently, release 4.20 contains an out-of-date version. This
8277 shouldn't happen again.
8279 5. When building Exim with embedded Perl support, the script that builds the
8280 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8281 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8283 6. The freeze_tell option was not being used for messages that were frozen on
8284 arrival, either by an ACL or by local_scan().
8286 7. Added the smtp_incomplete_transaction log selector.
8288 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8289 was accepting AUTH without a new EHLO.
8291 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8292 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8295 10. Logging of TCP/IP connections (when configured) now happens in the main
8296 daemon process instead of the child process, so that the TCP/IP connection
8297 count is more accurate (but it can never be perfect).
8299 11. The use of "drop" in a nested ACL was not being handled correctly in the
8300 outer ACL. Now, if condition failure induced by the nested "drop" causes
8301 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8302 or "require"), the connection is dropped.
8304 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8305 that yield "discard" can now be used with an "accept" or a "discard" verb,
8306 but an error is generated for any others (because I can't see a useful way
8307 to define what should happen).
8309 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8310 are now processed in the same way as lines in the Exim configuration file.
8311 In particular, continuation lines are supported.
8313 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8315 15. Added -ti meaning -t -i.
8317 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8318 domains, and warn by logging if others are found.
8320 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8321 returned from GIFCONF when Exim is trying to find the list of interfaces on
8322 a host. The code in os.c has been modified to copy each ifreq to an aligned
8323 structure in all cases.
8325 Also, in some cases, the returned ifreq's were being copied to a 'struct
8326 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8327 means the last couple of bytes of an IPv6 address could be chopped if the
8328 ifreq contained only a normal sockaddr (14 bytes storage).
8330 18. Named domain lists were not supported in the hosts_treat_as_local option.
8331 An entry such as +xxxx was not recognized, and was treated as a literal
8334 19. Ensure that header lines added by a DATA ACL are included in the reject log
8335 if the ACL subsequently rejects the message.
8337 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8338 MD5 (which is deprecated).
8340 21. When testing a filter file using -bf, Exim was writing a message when it
8341 took the sender from a "From " line in the message, but it was not doing so
8342 when it took $return_path from a Return-Path: header line. It now does.
8344 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8345 with a valid header line field name (a series of printing characters
8346 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8348 23. Changed "disc" in the source to "disk" to conform to the documentation and
8349 the book and for uniformity.
8351 24. Ignore Sendmail's -Ooption=value command line item.
8353 25. When execve() failed while trying to run a command in a pipe transport,
8354 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8355 could be confused with a return value of 69 from the command itself. This
8356 has been changed to 127, the value the shell returns if it is asked to run
8357 a non-existent command. The wording for the related log line suggests a
8358 non-existent command as the problem.
8360 26. If received_header_text expands to an empty string, do not add a Received:
8361 header line to the message. (Well, it adds a token one on the spool, but
8362 marks it "old" so that it doesn't get used or transmitted.)
8364 27. Installed eximstats 1.28 (addition of -nt option).
8366 28. There was no check for failure on the call to getsockname() in the daemon
8367 code. This can fail if there is a shortage of resources on the system, with
8368 ENOMEM, for example. A temporary error is now given on failure.
8370 29. Contrary to the C standard, it seems that in some environments, the
8371 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8372 program. Exim now does this explicitly; it affects the formatting of
8373 timestamps using strftime().
8375 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8376 complaints. I've now initialized all the variables, to avoid this.
8378 32. Header lines added by a system filter were not being "seen" during
8379 transport-time rewrites.
8381 33. The info_callback() function passed to OpenSSL is set up with type void
8382 (*)(SSL *, int, int), as described somewhere. However, when calling the
8383 function (actually a macro) that sets it up, the type void(*)() is
8384 expected. I've put in a cast to prevent warnings from picky compilers.
8386 34. If a DNS black list lookup found a CNAME record, but there were no A
8387 records associated with the domain it pointed at, Exim crashed.
8389 35. If a DNS black list lookup returned more than one A record, Exim ignored
8390 all but the first. It now scans all returned addresses if a particular IP
8391 value is being sought. In this situation, the contents of the
8392 $dnslist_value variable are a list of all the addresses, separated by a
8395 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8396 to accept a host name and all its aliases if the forward lookup for any of
8397 them yielded the IP address of the incoming connection. Now it accepts only
8398 those names whose forward lookup yields the correct IP address. Any other
8399 names are discarded. This closes a loophole whereby a rogue DNS
8400 administrator could create reverse DNS records to break through a
8401 wildcarded host restriction in an ACL.
8403 37. If a user filter or a system filter that ran in a subprocess used any of
8404 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8405 the wrong values were passed to the pipe command ($thisaddress had the
8406 value of $0, $0 had the value of $1, etc). This bug was introduced by
8407 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8409 38. Improved the line breaking for long SMTP error messages from ACLs.
8410 Previously, if there was no break point between 40 and 75 characters, Exim
8411 left the rest of the message alone. Two changes have been made: (a) I've
8412 reduced the minimum length to 35 characters; (b) if it can't find a break
8413 point between 35 and 75 characters, it looks ahead and uses the first one
8414 that it finds. This may give the occasional overlong line, but at least the
8415 remaining text gets split now.
8417 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8418 file descriptors might be low, and that setting 1000 would always raise it.
8419 It turns out that in some environments, the limit is already over 1000 and
8420 that lowering it causes trouble. So now Exim takes care not to decrease it.
8422 40. When delivering a message, the value of $return_path is set to $sender_
8423 address at the start of routing (routers may change the value). By an
8424 oversight, this default was not being set up when an address was tested by
8425 -bt or -bv, which affected the outcome if any router or filter referred to
8428 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8429 log only when "message" or "log_message" are set. However, if one of the
8430 conditions was an address verification, or a call to a nested ACL, the
8431 messages generated by the underlying test were being passed through. This
8432 no longer happens. The underlying message is available in $acl_verify_
8433 message for both "message" and "log_message" expansions, so it can be
8434 passed through if needed.
8436 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8437 new expansion $bh_ to give the encoded byte string without charset
8438 translation. Translation happens only if iconv() is available; HAVE_ICONV
8439 indicates this at build time. HEADERS_CHARSET gives the charset to
8440 translate to; headers_charset can change it in the configuration, and
8441 "headers charset" can change it in an individual filter file.
8443 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8444 that creates RFC 2047 encoded "words" labels them as that charset instead
8445 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8446 expansion operator; (ii) when Exim creates a From: line for a local
8447 message; (iii) when a header line is rewritten to include a "phrase" part.
8449 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8450 buggy, causing it to skip the first lines of messages whose message ID
8451 ended in 'D'. This would not have bitten before Exim release 4.14, because
8452 message IDs were unlikely to end in 'D' before then. The effect was to have
8453 incorrect size information for certain domains.
8455 45. #include "config.h" was missing at the start of the crypt16.c module. This
8456 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8459 46. If there was a timeout during a "random" callout check, Exim treated it as
8460 a failure of the random address, and carried on sending RSET and the real
8461 address. If the delay was just some slowness somewhere, the response to the
8462 original RCPT would be taken as a response to RSET and so on, causing
8463 mayhem of various kinds.
8465 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8466 when I implemented it. It didn't allow for the fact that some option values
8467 may legitimately be negative (e.g. size_addition), and it didn't even do
8468 the right test for positive values.
8470 48. Domain names in DNS records are case-independent. Exim always looks them up
8471 in lower case. Some resolvers return domain names in exactly the case they
8472 appear in the zone file, that is, they may contain uppercase letters. Not
8473 all resolvers do this - some return always lower case. Exim was treating a
8474 change of case by a resolver as a change of domain, similar to a widening
8475 of a domain abbreviation. This triggered its re-routing code and so it was
8476 trying to route what was effectively the same domain again. This normally
8477 caused routing to fail (because the router wouldn't handle the domain
8478 twice). Now Exim checks for this case specially, and just changes the
8479 casing of the domain that it ultimately uses when it transmits the message
8482 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8485 50. If a filter generated a file delivery with a non-absolute name (possible if
8486 no home directory exists for the router), the forbid_file option was not
8489 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8490 the existing equality matching.
8492 52. Exim was using ints instead of ino_t variables in some places where it was
8493 dealing with inode numbers.
8495 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8496 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8497 called TMPDIR, and if it finds it is different, it changes its value.
8499 54. The smtp_printf() function is now made available to local_scan() so
8500 additional output lines can be written before returning. There is also an
8501 smtp_fflush() function to enable the detection of a dropped connection.
8502 The variables smtp_input and smtp_batched_input are exported to
8505 55. Changed the default runtime configuration: the message "Unknown user"
8506 has been removed from the ACL, and instead placed on the localuser router,
8507 using the cannot_route_message feature. This means that any verification
8508 failures that generate their own messages won't get overridden. Similarly,
8509 the "Unrouteable address" message that was in the ACL for unverifiable
8510 relay addresses has also been removed.
8512 56. Added hosts_avoid_esmtp to the smtp transport.
8514 57. The exicyclog script was not checking for the esoteric option
8515 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8516 will work only if exicyclog is run under the appropriate euid.
8518 58. Following a discussion on the list, the rules by which Exim recognises line
8519 endings on incoming messages have been changed. The -dropcr and drop_cr
8520 options are now no-ops, retained only for backwards compatibility. The
8521 following line terminators are recognized: LF CRLF CR. However, special
8522 processing applies to CR:
8524 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8525 nor a local message in the state where . is a terminator.
8527 (ii) If a bare CR is encountered in a header line, an extra space is added
8528 after the line terminator so as not to end the header. The reasoning
8529 behind this is that bare CRs in header lines are most likely either
8530 to be mistakes, or people trying to play silly games.
8532 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8533 was being incorrectly given as 18 bytes larger than it should have been.
8534 This is a VOB (very old bug).
8536 60. This may never have affected anything current, but just in case it has:
8537 When the local host is found other than at the start of a list of hosts,
8538 the local host, those with the same MX, and any that follow, are discarded.
8539 When the list in question was part of a longer list of hosts, the following
8540 hosts (not currently being processed) were also being discarded. This no
8541 longer happens. I'm not sure if this situation could ever has previously
8544 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8547 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8548 header. This is something that is recommended in a new Internet Draft, and
8549 is something that is documented as being done by Sendmail. There are two
8550 possible values. For messages generated by the autoreply transport, Exim
8553 Auto-Submitted: auto-replied
8555 whereas for all other generated messages (e.g. bounces) it adds
8557 Auto-Submitted: auto-generated
8559 63. The "personal" condition in filters now includes a test for the
8560 Auto-Submitted: header. If it contains the string "auto-" the message it
8561 not considered personal.
8563 64. Added rcpt_include_affixes as a generic transport option.
8565 65. Added queue_only_override (default true).
8567 66. Added the syslog_duplication option.
8569 67. If what should have been the first header line of a message consisted of
8570 a space followed by a colon, Exim was mis-interpreting it as a header line.
8571 It isn't of course - it is syntactically invalid and should therefore be
8572 treated as the start of the message body. The misbehaviour could have
8573 caused a number of strange effects, including loss of data in subsequent
8574 header lines, and spool format errors.
8576 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8577 client host had authenticated. This control can now be exercised by an ACL
8578 for more flexibility.
8580 69. By default, callouts do not happen when testing with -bh. There is now a
8581 variant, -bhc, which does actually run the callout code, including
8582 consulting and updating the callout cache.
8584 70. Added support for saslauthd authentication, courtesy of Alexander
8587 71. If statvfs() failed on the spool or log directories while checking their
8588 size for availability, Exim confusingly gave the error "space shortage".
8589 Furthermore, in debugging mode it crashed with a floating point exception.
8590 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8591 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8592 serious problem, Exim now writes to the main and panic logs when this
8593 happens, with details of the failure. It then refuses to accept the
8594 incoming message, giving the message "spool directory problem" or "log
8595 directory problem" with a 421 code for SMTP messages.
8597 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8598 0, 1, and 2 exist, because some OS complain for execs without them (see
8599 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8600 descriptors. However, the code omitted to check that the open succeeded,
8601 causing mysterious errors if for some reason the permissions on /dev/null
8602 got screwed. Now Exim writes a message to the main and panic logs, and
8603 bombs out if it can't open /dev/null.
8605 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8606 interact so that it is all more flexible. It is supposed to remain
8607 backwards compatible. Also added extra_local_interfaces.
8609 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8610 to bomb out with an assertion failure - to the client this appears as a
8611 connection drop. This problem occurs in the part of the code that was taken
8612 from the Samba project. Fortunately, the assertion is in a very simple
8613 function, so I have fixed this by reproducing the function inline in the
8614 one place where it is called, and arranging for authentication to fail
8615 instead of killing the process with assert().
8617 75. The SPA client code was not working when the server requested OEM rather
8618 than Unicode encoding.
8620 76. Added code to make require_files with a specific uid setting more usable in
8621 the case where statting the file as root fails - usually a non-root-mounted
8622 NFS file system. When this happens and the failure is EACCES, Exim now
8623 forks a subprocess and does the per-uid checking as the relevant uid.
8625 77. Added process_log_path.
8627 78. If log_file_path was not explicitly set, a setting of check_log_space or
8628 check_log_inodes was ignored.
8630 79. If a space check for the spool or log partitions fails, the incident is now
8631 logged. Of course, in the latter case the data may get lost...
8633 80. Added the %p formatting code to string_format() so that it can be used to
8634 print addresses in debug_print(). Adjusted all the address printing in the
8635 debugging in store.c to use %p rather than %d.
8637 81. There was a concern that a line of code in smtp_in.c could overflow a
8638 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8639 initially expressed, the concern was not well-founded, because trailing
8640 spaces are removed early. However, if the trailing spaces were followed by
8641 a NULL, they did not get removed, so the overflow was possible. Two fixes
8644 (a) I re-wrote the offending code in a cleaner fashion.
8645 (b) If an incoming SMTP command contains a NULL character, it is rejected
8648 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8649 runs initgroups(), so that if the Exim user is in any additional groups,
8650 they will be used during message reception.
8656 The change log for 4.20 and earlier releases has been archived.