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 Support list of dkim results in the dkim_status ACL condition, making
9 it more usable in the data ACL.
11 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
12 Previously This was only logged, on the assumption that errors would be
13 seen for a previous fflush(). However, a fuse filesystem has been
14 reported as showing this an error for the fclose(). The spool is now in
15 an uncertain state, and we have logged and responded acceptance. Change
16 this to respond with a temp-reject, wipe spoolfiles, and log the error
19 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
20 during a recipient verify callout, a QUIT command was attempted on the
21 now-closed callout channel, causing a paniclog entry.
23 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
24 a connection_reject log_selector, under tls_on_connect. Previously
25 with this combination, when the connect ACL rejected, a spurious
26 paniclog entry was made.
28 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
29 of loadbalancer-detection for resumption, in 4.96 - which tries to
30 use the EHLO response. SMTPS does not have one at the time it is starting
31 TLS. Change the default for the smtp transport host_name_extract option
32 to be a static string, for TLS-on-connect cases; meaning that resumption
33 will always be attempted (unless deliberately overriden).
35 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
36 chunk-separator specification. This was broken by hardening introduced
39 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
40 included the -H with the id; this also messed up exiqgrep.
42 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
43 was not properly skipped and empty elements would cause mis-parsing.
44 Tighten parsing of DKIM header records. Previously, all but lowercase
45 alpha chars would be ignored in potential tag names.
47 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
48 being searched. Previously we only had one for IPv4 addresses. Per the
49 documentation, the error results by default in a no-match result for the
50 list. It is logged if the unknown_in_list log_selector is used.
52 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
53 the routing operation. Previously it would silently stop routing the
56 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
57 info-request event close in time to a scheduled run timer could result in
58 the latter being missed, and no further queue scheduled runs being
59 initiated. This ouwld be more likely on high-load systems.
61 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
62 LF-only mode (as detected from the first header line). Previously we did
63 accept that in (normal) CRLF mode; this has been raised as a possible
64 attack scenario (under the name "smtp smuggling").
66 JH/13 Add an fdatasync call for the received message data file in spool, before
67 loggging reception and sending the SMTP ack. Previously we only flushed
68 the stdio buffer so there was still the possibility of a disk error.
70 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
71 address. Previously, for the last address in a header line (commonly
72 there is only one) the terminating newline was part of the logged
75 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
76 header, when parsing for address-qualification. Previously one was only
77 written if there were rewrite rules.
79 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
82 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
83 the multiple-queue-runners facility for 4.97 broke this, giving only a
84 one-time run of the queue.
86 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
89 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
90 TLS session resulted in a forced-fail, a misleading error was logged for
91 the expansino of tls_certificates. This would affect the common case of
92 that option being set (main-section options) but not having any variable
93 parts. It could also potentially affect tls_privatekeys. The underlyding
94 coding errors go back to 4.90 but were only exposed in 4.97.
96 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
97 pcre2 library starting allocating 20kB rather than 112 bytes per match
98 call, which broke the 2GB total limitation on Exim's memory management
99 when a user had over 104207 messages stored and the appendfile
100 maildir_quota_directory_regex option is in use. Release the allocated
101 memory every thosand files to avoid this.
102 The same issue arises with the ACL regex condition, which is applied
103 to every line of a received message.
105 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
106 which all recipients had been handled (itself an issue) a null-pointer
107 deref was done on trying to write a retry record. Fix that by counting
108 the outstanding recipients before trying to transmit the message.
109 The situation arose for a second MX try within a transport run, when the
110 first had perm-rejected a recipient (the only one for the connection, in
111 the case seen) during pipelining, and then closed the TCP connection.
112 The transport classified that as an I/O error, leaving the message
113 outstanding but having marked up the recipient as dealt-with. It then
114 tried another MX because of the I/O error. Fix this by converting the
115 message-level status to ok if there was a close but all recipients were
116 dealt with. Thanks to Wolfgand Breyha for debug runs.
118 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
119 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
121 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
123 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
124 option specified a fixed address, later rewriting on that address would
125 trip on the configuration data being readonly. Instead of modifying
126 in-place, copy data. Found and fixed by Peter Benie.
128 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
129 data a null pointer was followed. Find and testcase by Sebastian Bugge.
135 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
136 SMTP connection" log lines.
138 JH/02 Option default value updates:
139 - queue_fast_ramp (main) true (was false)
140 - remote_max_parallel (main) 4 (was 2)
142 JH/03 Cache static regex pattern compilations, for use by ACLs.
144 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
145 Make the rewrite never match and keep the logging. Trust the
146 admin to be using verify=header-syntax (to actually reject the message).
148 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
149 (under Linux) we watch the dir containing the final file; previously
150 it would be the dir with the first symlink. We still do not monitor
153 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
154 dn_expand() is happy to pass them through.
156 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
157 the file with the proof had an unchanged name, the new proof(s) were
158 loaded on top of the old ones (and nover used; the old ones were stapled).
160 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
161 more than one message arrived in a single connection a reference from
162 the earlier message could be re-used. Often a sigsegv resulted.
163 These variables were introduced in Exim 4.87.
164 Debug help from Graeme Fowler.
166 JH/09 Fix ${filter } for conditions that modify $value. Previously the
167 modified version would be used in construction the result, and a memory
170 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
171 Find and fix by Jasen Betts.
173 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
174 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
175 the systemwide configuration to override the Exim config.
177 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
178 API changes in libopendmarc.
180 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
181 pid 1, in the normal "background daemon" mode, having to drop process-
182 group leadership also lost track of needing to create listener sockets.
184 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
185 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
187 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
188 a capture group which obtained no text (eg. "(abc)*" matching zero
189 occurrences) could cause a segfault if the corresponding $<n> was
192 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
193 included a close-brace character (eg. it itself used an expansion) an
196 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
197 starting TLS. Previously it was after, meaning that attackers on such
198 ports had to be screened using the host_reject_connection main config
199 option. The new sequence aligns better with the STARTTLS behaviour, and
200 permits defences against crypto-processing load attacks, even though it
201 is strictly an incompatible change.
202 Also, avoid sending any SMTP fail response for either the connect ACL
203 or host_reject_connection, for TLS-on-connect ports.
205 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
206 Previously this was not permitted, but it makes reasonable sense.
207 While there, restore a restriction on using it from a connect ACL; given
208 the change JH/16 it could only return false (and before 4.91 was not
211 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
212 was exactly sized compared to the log buffer, a crash occurred with the
213 misleading message "bad memory reference; pool not found".
214 Found and traced by Jasen Betts.
216 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
217 dns_again_means_nonexist included an element causing a DNS lookup which
218 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
219 included (though probably not limited to) a process crash from stack
220 memory limit, or from excessive open files. Replace this with a paniclog
221 whine (as this is likely a configuration error), and returning
224 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
225 this always failed, probably leading to the usual downgrade to in-clear
228 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
229 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
230 to not checking that list for these lookups.
232 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
235 JH/23 Fix crash in string expansions. Previously, if an empty variable was
236 immediately followed by an expansion operator, a null-indirection read
237 was done, killing the process.
239 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
240 include an SMTP response string which is longer than that supported
241 by the delivering transport. Alleviate by wrapping such lines before
244 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
245 chars (RFC limit). Previously a limit of 12 items was made, which with
246 a not-impossible References: in the message being bounced could still
247 be over-large and get stopped in the transport.
249 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
250 close. Previously a bare socket close was done.
252 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
255 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
256 message being considered for delivery was already being handled by
257 another process, and having an SMTP connection already open, the function
258 to close it tried to use an uninitialized variable. This would afftect
259 high-volume sites more, especially when running mailing-list-style loads.
260 Pollution of logs was the major effect, as the other process delivered
261 the message. Found and partly investigated by Graeme Fowler.
263 JH/29 Change format of the internal ID used for message identification. The old
264 version only supported 31 bits for a PID element; the new 64 (on systems
265 which can use Base-62 encoding, which is all currently supported ones
266 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
267 and must use Base-36). The new ID is 23 characters rather than 16, and is
268 visible in various places - notably logs, message headers, and spool file
269 names. Various of the ancillary utilities also have to know the format.
270 As well as the expanded PID portion, the sub-second part of the time
271 recorded in the ID is expanded to support finer precision. Theoretically
272 this permits a receive rate from a single comms channel of better than the
274 The major timestamp part of the ID is not changed; at 6 characters it is
275 usable until about year 3700.
276 Updating from previously releases is fully supported: old-format spool
277 files are still usable, and the utilities support both formats. New
278 message will use the new format. The one hints-DB file type which uses
279 message-IDs (the transport wait- DB) will be discarded if an old-format ID
280 is seen; new ones will be built with only new-format IDs.
281 Optionally, a utility can be used to convert spool files from old to new,
282 but this is only an efficiency measure not a requirement for operation
283 Downgrading from new to old requires running a provided utility, having
284 first stopped all operations. This will convert any spool files from new
285 back to old (losing time-precision and PID information) and remove any
286 wait- hints databases.
288 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
289 we treated them as item separators when parsing for a list item, but they
290 need to be protected by the doublequotes. While there, add handling for
293 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
294 Found and fixed by Jasen Betts. No testcase for this as my usual text
295 editor insists on emitting only valid UTF-8.
297 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
298 a null-indirection SIGSEGV for the receive process.
300 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
301 -bh use. Internal checking would spot this and take a panic.
303 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
304 In 4.96 this would expand to empty.
306 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
307 certificate. Find and fix by Andreas Metzler.
309 JH/36 Add ARC info to DMARC hostory records.
311 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
312 or fakedefer. Previously the sender could discover that the message
313 had in fact been accepted.
315 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
316 sequences. Previously the input was not noted as being tainted; notably
317 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
318 bad coding of authenticators.
320 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
321 and ${tr...}. Found and diagnosed by Heiko Schlichting.
323 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
324 could be triggered by externally-supplied input. Found by Trend Micro.
327 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
328 be triggered by externally-controlled input. Found by Trend Micro.
331 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
332 be triggered by externally-controlled input. Found by Trend Micro.
335 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
336 Make the rewrite never match and keep the logging. Trust the
337 admin to be using verify=header-syntax (to actually reject the message).
339 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
342 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
348 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
349 after reception to before a subsequent reception. This should
350 mean slightly faster delivery, and also confirmation of reception
353 JH/02 Move from using the pcre library to pcre2. The former is no longer
354 being developed or supported (by the original developer).
356 JH/03 Constification work in the filters module required a major version
357 bump for the local-scan API. Specifically, the "headers_charset"
358 global which is visible via the API is now const and may therefore
359 not be modified by local-scan code.
361 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
362 sendfile() didi not account for the way the ClamAV driver code called it.
364 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
365 time check was being done for every character; replace that with one
368 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
369 sent was prefixed with a length byte.
371 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
372 RFC 5321. Previously Dovecot (at least) would log errors during
375 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
376 in 4.61 without a reason logged. Binaries will be bigger, which might
377 matter on diskspace-constrained systems, but debug is easier.
379 JH/09 Fix macro-definition during "-be" expansion testing. The move to
380 write-protected store for macros had not accounted for these runtime
381 additions; fix by removing this protection for "-be" mode.
383 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
384 handing out large-numbered file descriptors, violating the usual Unix
385 assumption (and required by Posix) that the lowest possible number will be
386 allocated by the kernel when a new one is needed. In the daemon, and any
387 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
388 are not useable for FD_SET() [and hence select()] and overwrite the stack.
389 Assorted crashes happen.
391 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
392 main-section options or in a connect ACL, the value from the first ever
393 connection was never replaced for subsequent connections. Found by
396 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
397 though only once PCRE2 was introduced: the memory accounting used under
398 debug offset allocations by an int, giving a hard trap in early startup.
399 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
401 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
402 with underbars is given. The write-protection of configuration introduced
403 in 4.95 trapped when normalisation was applied to an option not needing
406 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
408 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
409 failing to close the certificates directory, every hour or any time it
412 JH/16 Debugging initiated by an ACL control now continues through into routing
413 and transport processes. Previously debugging stopped any time Exim
414 re-execs, or for processing a queued message.
416 JH/17 The "expand" debug selector now gives more detail, specifically on the
417 result of expansion operators and items.
419 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
420 bad comparison between the option value and the name of the file to
421 be included was done, and a mismatch was wrongly identified.
422 4.88 to 4.95 are affected.
424 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
426 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
427 supplied as the db file-pair basename. Previously, if a directory
428 path was given, for example via the autoreply "once" option, the DB
429 file.pag and file.dir files would be created in that directory's
432 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
433 "taint" log_selector. These were previously deprecated.
435 JH/22 Fix static address-list lookups to properly return the matched item.
436 Previously only the domain part was returned.
438 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
439 the call into OpenSSL to send a TLS Close was being repeated; this
440 resulted in the library waiting for the peer's Close. If that was never
441 sent we waited forever. Fix by tracking send calls.
443 JH/24 The ${run} expansion item now expands its command string elements after
444 splitting. Previously it was before; the new ordering makes handling
445 zero-length arguments simpler. The old ordering can be obtained by
446 appending a new option "preexpand", after a comma, to the "run".
448 JH/25 Taint-check exec arguments for transport-initiated external processes.
449 Previously, tainted values could be used. This affects "pipe", "lmtp" and
450 "queryprogram" transport, transport-filter, and ETRN commands.
451 The ${run} expansion is also affected: in "preexpand" mode no part of
452 the command line may be tainted, in default mode the executable name
455 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
456 the the facility was not passed across execs, and only the first message
457 passed over a connection could use BDAT; any further ones using DATA.
459 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
460 uses $sending_ip_address and an interface is specified.
461 Previously any use of the local address in the EHLO name disabled
462 PIPECONNECT, the common case being to use the rDNS of it.
464 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
465 resumption. Previously verify failed because no certificate status is
466 passed on the wire for the restarted session. Fix by using the recorded
467 ocsp status of the stored session for the new connection.
469 JH/29 TLS resumption: the key for session lookup in the client now includes
470 more info that a server could potentially use in configuring a TLS
471 session, avoiding oferring mismatching sessions to such a server.
472 Previously only the server IP was used.
474 JH/30 Fix string_copyn() for limit greater than actual string length.
475 Previously the copied amount was the limit, which could result in a
476 overlapping memcpy for newly allocated destination soon after a
477 source string shorter than the limit. Found/investigated by KM.
479 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
480 close; it may be needed for a subsequent connection. This caused a
481 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
483 JH/32 Fix CHUNKING for a second message on a connection when the first was
484 rejected. Previously we did not reset the chunking-offered state, and
485 erroneously rejected the BDAT command. Investigation help from
488 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
489 an empty address. Previously the expansion returned an error.
491 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
492 proxy. Previously these were misparsed, leading to paniclog entries.
498 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
499 related applications. Previously an "H" was used where available info
500 says that "M" should be, so change to match.
502 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
503 as arguments, so an implementation trying to copy these into a local
504 buffer was taking a taint-enforcement trap. Fix by using dynamically
505 created buffers. Similar fix for radius expansion condition.
507 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
508 reasonable, eg. to count headers. Fix by using dynamically created
509 buffers rather than a local. Do similar fixes for ACL actions "dcc",
510 "log_reject_target", "malware" and "spam"; the arguments are expanded
511 so could be handling tainted values.
513 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
514 broken the (no-op) support for this sendmail command. Restore it
515 to doing nothing, silently, and returning good status.
517 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
518 record path was given (or the default used) without a leading directory
519 path, an error occurred on trying to open it. Use the transport's working
522 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
523 transport. Previously it was the name on the DNS A-record; use instead
524 the head of the CNAME chain leading there (if there is one). This seems
525 to align better with RFC 6125.
527 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
528 smtp_accept_max_per_host allocated resources which were not released
529 when the limit was exceeded. This eventually crashed the daemon. Fix
530 by adding a release action in that path.
532 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
533 expanded; previously using tainted values was rejected. Fix by using
534 dynamically-created buffers.
536 JH/09 Relax restrictions on ACL verify condition needing access to message
537 headers. Previously they were only permitted in data and non-smtp ACLs;
538 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
539 not_blind, header_sender and header_names_ascii verification.
541 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
542 Previously a macro used one argument twice; when called with the
543 argument as an expression having side-effects, incorrect operation
544 resulted. Use an inlineable function.
546 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
547 held open for a verify callout. Previously this wan not accounted for
548 and a corrupt onward SMTP conversation resulted.
550 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
551 Previously they were embedded naively in the constructed address; when
552 needed, strip the quoting and quote the entire local_part.
553 Also make the inbound_srs expansion condition handle quoting.
555 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
556 excluded, not matching the documentation.
558 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
559 was given for the sqlite_dbfile a trap resulted.
561 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
562 "name" argument resulted in a trap. There is no reason to disallow such;
563 this was a coding error.
565 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
566 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
567 spent suspended, ignoring the POSIX definition. Previously we assumed
568 it did and a constant offset from real time could be used as a correction.
569 Change to using the same clock source for the start-of-message and the
570 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
571 exists, just to get a clock slightly more aligned to reality.
573 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
574 RFC says it is optional some validators care. The missing char was not
575 intended but triggered by a line-wrap alignment. Discovery and fix by
576 Guillaume Outters, hacked on by JH.
578 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
579 name being quoted was tainted a trap would be taken. Fix by using
580 dynamically created buffers. The routine could have been called by a
581 rewrite with the "h" flag, by using the "-F" command-line option, or
582 by using a "name=" option on a control=submission ACL modifier.
584 JH/19 SPF: change the Authentication-Results expansion component to give
585 smtp.helo when the sender domain is empty. Previously it gave
588 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
589 not in 127.0.0.0/8 to help in spotting list domains taken over by a
590 domain-parking registrar.
592 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
593 Previously when a whitespace character was specified it was not inserted
594 after removing the newline.
596 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
597 the domain part of the recipient address. This overrides any tls_sni
598 option set, which was previously used.
600 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
603 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
604 is_tainted() had an off-by-one error in the overenthusiastic direction.
605 Find and fix by Gavan. Although NetBSD is not a supported platform for
606 4.94 this bug could affect other platforms.
608 PP/01 Fix default prime selection to be consistent.
609 One path used ike23 still, instead of exim.dev.20160529.3; now both
610 execution flows will use the same DH primes (currently
611 exim.dev.20160529.3).
613 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
614 option in smtp transport, to match the documentation. Previously
615 verification was not being done in some cases where it should have been.
617 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
618 than one server was defined and depending on the platform memory layout
619 details, an internal consistency trap could be hit while walking the list
622 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
623 files. The value is used by the authresults expansion item. Previously
624 if this was used in a router or transport, a crash could result.
626 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
627 have a sufficient incidence of EINTR returns from select that an
628 interaction with other operations done by the main daemon loop exposed
629 a bug in the error-handling. This was benign apart from the log
632 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
633 for consistency with delivered "=>" and failed "**" lines. While we're
634 there, handle PRX and TFO.
636 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
637 applied. This resulted, if any header-line rewrite rules were configured,
638 in a panic-log triggerable by sending a message with a long address in
639 a header. Fix by increasing the arbitrary limit to larger than a single
640 (dewrapped) 5322 header line maximum size.
642 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
643 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
644 RFC 6648 which deprecates X- options in protocols as a general practice.
645 Changeover between the implementations is handled by the mechanisms
648 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
649 a remote which uses callout/hold. Previously the local was queued.
651 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
654 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
655 log line. The asterisk is now consistently appended to the remote IP
656 (and port, if given), and will also be provided on defer and fail log
657 lines. Previously it could be placed on the local IP if that was being
658 logged, and was only provided on delivery lines.
660 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
662 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
663 authenticator client_send option. Previously the next char, after a pair
664 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
665 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
666 documentation. There is still no way to get a leading ^ immediately
667 after a NUL (ie. for the password of a PLAIN method authenticator.
669 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
670 files. For bad sizes read, delete the record and whine to paniclog.
672 JH/38 When logging an AUTH failure, as server, do not include sensitive
673 information. Previously, the credentials would be included if given
674 as part of the AUTH command line and an ACL denied authentication.
676 JH/39 Bug 2691: fix $local_part_data. When the matching list element
677 referred to a file, bad data was returned. This likely also affected
680 JH/40 The gsasl authenticator now supports caching of the salted password
681 generated by the client-side implementation. This required the addition
682 of a new variable: $auth4.
684 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
685 left undeleted; the attempt to re-create it then failed - resulting in
686 the usual "SIGHUP tp have daemon reload configuration" to not work.
687 This affected any platform not supporting "abstract" Unix-domain
688 sockets (i.e. not Linux).
690 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
691 recipients" response to RCPT in a later response, with a 250. The
692 previous coding assumed this would not happen, and under PIPELINING
693 would result in both lost and duplicate recipients for a message.
695 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
696 Previously the weighting was incorrectly applied. Similar fix for socks
697 proxies. Found and fixed by Heiko Schlichting.
699 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
700 not handle sub-lists included using the +namedlist syntax. While
701 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
702 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
705 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
706 Previously the permanent pool was used, so the store could not be freed.
707 This meant a connection with many messages would use continually-growing
710 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
711 per-pool so as not to waste too much space. Previously a constant size
712 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
713 DOS attacks harder. The cost is wasted memory use in the larger blocks.
715 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
716 same space repeatedly, and a smaller process footprint.
718 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
719 for GnuTLS. Previously the name started "%s" which, while not a bug,
720 looked as if if might be one.
722 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
723 "continued connection") make the $proxy_* variables available. Previously
724 the information was not passed across the exec() call for subsequent
725 transport executions. This also mean that the log lines for the
726 messages can show the proxy information.
728 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
729 receive time. With modern systems the difference is significant.
730 The historical behaviour can be restored by disabling (a new) log_selector
731 "queue_time_exclusive".
733 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
734 content) were specifically tested for. Now, also cover expansions
735 resulting in ACL names and inline ACL content.
737 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
738 making it unusable in complex expressions.
740 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
741 fail with a taint-check on the spoolfile name, and leave the message
744 HS/01 Enforce absolute PID file path name.
746 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
748 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
750 PP/02 Bug 2643: Correct TLS DH constants.
751 A missing NUL termination in our code-generation tool had led to some
752 incorrect Diffie-Hellman constants in the Exim source.
753 Reported by kylon94, code-gen tool fix by Simon Arlott.
755 PP/03 Impose security length checks on various command-line options.
756 Fixes CVE-2020-SPRSS reported by Qualys.
758 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
759 better. Reported by Qualys.
761 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
762 providing a particularly obnoxious sender full name.
765 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
767 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
770 PP/08 Change default for recipients_max from unlimited to 50,000.
772 PP/09 Fix security issue with too many recipients on a message (to remove a
773 known security problem if someone does set recipients_max to unlimited,
774 or if local additions add to the recipient list).
775 Fixes CVE-2020-RCPTL reported by Qualys.
777 PP/10 Fix security issue in SMTP verb option parsing
778 Fixes CVE-2020-EXOPT reported by Qualys.
780 PP/11 Fix security issue in BDAT state confusion.
781 Ensure we reset known-good where we know we need to not be reading BDAT
782 data, as a general case fix, and move the places where we switch to BDAT
783 mode until after various protocol state checks.
784 Fixes CVE-2020-BDATA reported by Qualys.
786 HS/03 Die on "/../" in msglog file names
788 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
789 the Exim runtime user are allowed to create files.
791 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
794 QS/03 When reading the output from interpreted forward files we do not
795 pass the pipe between the parent and the interpreting process to
796 executed child processes (if any).
798 QS/04 Always die if requested from internal logging, even is logging is
801 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
802 the API; compilation noo longer completes with DMARC support included.
803 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
804 been reported on other platforms.
806 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
808 JH/56 Make the majority of info read from config files readonly, for defence-in-
809 depth against exploits. Suggestion by Qualys.
810 Not supported on Solaris 10.
812 JH/57 Fix control=fakreject for a custom message containing tainted data.
813 Previously this resulted in a log complaint, due to a re-expansion present
814 since fakereject was originally introduced.
816 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
817 modifier was given, a loop resulted.
819 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
820 pipelined SMTP command followed the BDAT LAST then it would be
821 incorrectly treated as part of the message body, causing a verification
824 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
825 require looser validation rules than those for 821-level addresses,
826 which only permit IP addresses.
832 JH/01 Avoid costly startup code when not strictly needed. This reduces time
833 for some exim process initialisations. It does mean that the logging
834 of TLS configuration problems is only done for the daemon startup.
836 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
838 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
839 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
842 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
843 DKIM signing is being done. Previously a transport_filter would always
844 disable CHUNKING, falling back to traditional DATA.
846 JH/05 Regard command-line recipients as tainted.
848 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
850 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
851 PAM library frees one of the arguments given to it, despite the
852 documentation. Therefore a plain malloc must be used.
854 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
855 on-stack buffers were used, resulting in a taint trap when DSN information
856 copied from a received message was written into the buffer.
858 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
859 the ordering of its ARC headers. This caused a crash.
861 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
862 a new record was being constructed with information from the peer, a trap
865 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
866 installation would get error messages from DMARC verify, when it hit the
867 nonexistent file indicated by the default. Distros wanting DMARC enabled
868 should both provide the file and set the option.
869 Also enforce no DMARC verification for command-line sourced messages.
871 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
872 could, depending on the platform, hang at the STARTTLS response.
874 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
875 message on a connection. Previously if one message had ARC headers and
876 the following one did not, a crash could result when adding an
877 Authentication-Results: header.
879 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
880 Exim main code and Exim-related utities. The introduction of taint
881 tracking also did many adjustments to string handling. Since then, eximon
882 frequently terminated with an assert failure.
884 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
885 check for 452 responses. This slightly helps the inefficieny of doing
886 a large alias-expansion into a recipient-limited target. The max_rcpt
887 transport option still applies (and at the current default, will override
888 the new feature). The check is done for either cause of synch, and forces
889 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
890 connection. The new facility is not tunable at this time.
892 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
893 library live data was being used, so the results became garbage. Make
894 copies while it is still usable.
896 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
897 on delivery deferred (==) and failed (**) lines (if a delivery was
898 attemtped). Previously it was only on completion (=>) lines.
900 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
901 for the expansion of the server_scram_iter and server_scram_salt options.
903 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
904 are now specifically given a NO_DATA response without hitting the system
905 resolver. The library goes on to do the now-standard TXT lookup.
906 Use of dnsdb lookups is not affected.
908 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
909 only retrieve the errormessage once. Previously two calls to dlerror()
910 were used, and the second one (for mainlog/paniclog) retrieved null
913 JH/20 Taint checking: disallow use of tainted data for
914 - the appendfile transport file and directory options
915 - the pipe transport command
916 - the autoreply transport file, log and once options
917 - file names used by the redirect router (including filter files)
919 - paths used by single-key lookups
920 Previously this was permitted.
922 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
923 adjusted the size of a major service buffer; this failed because the
924 buffer was in use at the time. Change to a compile-time increase in the
925 buffer size, when this authenticator is compiled into exim.
927 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
928 previous fast-mode was untenable in the face of glibs using mmap to
929 support larger malloc requests.
931 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
932 New values supported, if defined on system where compiled:
933 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
934 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
936 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
937 running a limited number of proceses in parallel, a benefit is gained. The
938 amount varies with the platform hardware and load. The use of the option
939 queue_run_in_order means we cannot do this, as ordering becomes
942 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
943 had introduced a string-copy (for ensuring NUL-termination) which was not
944 appropriate for that case, which can include embedded NUL bytes in the
945 block of data. Investigation showed the copy to actually be needless, the
946 data being length-specified.
948 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
949 done during a receiving connection, and both used TLS, global info was
950 used rather than per-connection info for tracking the state of data
951 queued for transmission. This could result in a connection hang.
953 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
954 Previously, when delivering serveral messages down a single connection
955 only the first would provide a SIZE. This was due to the size information
956 not being properly tracked.
958 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
959 TAI (at 37 seconds currently), pretend to be in UTC for time-related
960 expansion and logging. Previously, spurious values such as a future
961 minute could be seen.
963 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
964 it could crash from a null-deref. This could also affect the
965 ${addresses: } operator and ${readsock } item.
967 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
968 message following a mime one, the variable was not reset.
970 JH/30 When an pipelined-connect fails at the first response, assume incorrect
971 cached capability (perhaps the peer reneged?) and immediately retry in
974 JH/31 Fix spurious detection of timeout while writing to transport filter.
976 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
977 an attempt to copy the string was made before checking it.
979 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
980 taint of the lookup key was maintained; we now regard the presence in the
981 filesystem as sufficient validation.
983 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
984 argument is supplied.
986 JH/35 The dsearch lookup now requires that the directory is an absolute path.
987 Previously this was not checked, and nonempty relative paths made an
988 access under Exim's current working directory.
990 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
991 Previously no event was raised.
993 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
994 parameter supplied by the sender MAIL FROM command. Previously it was
995 ignored, and only the check_spool_space option value for the required
998 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
999 the size of the signing public-key. Previously it was instead giving
1000 the size of the signature hash.
1002 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1003 the default. See the (new) dkim_verify_min_keysizes option.
1005 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1006 and an ACL use a lookup for checking either the local_part or domain,
1007 stale data could be accessed. Ensure that variable references are
1008 dropped between messages.
1010 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1011 by the client was not checked as pointing within response data before
1012 being used. A malicious client could thus cause an out-of-bounds read and
1013 possibly gain authentication. Fix by adding the check.
1015 JH/42 Internationalisation: change the default for downconversion in the smtp
1016 transport to be "if needed". Previously it was "as previously set" for
1017 the message, which usually meant "if needed" for message-submission but
1018 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1019 even when the envelope addresses did not need it, resulting in forwarding
1020 failures to non-supporting MTAs. A downconvert in such cases will be
1021 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1022 transport. The change does mean that addresses needing conversion will
1023 be converted when previously a delivery failure would occur.
1025 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1026 response was received it would be used unchecked in a fail-DSN, violating
1027 standards on line-length limits. Truncate if needed.
1029 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1030 doesn't work. (Additionally add a new main config option to configure the
1037 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1038 side, to decode a TLS 1.3 packet capture.
1040 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1041 Previously the default library behaviour applied, sending two, each in
1042 its own TCP segment.
1044 JH/03 Debug output for ACL now gives the config file name and line number for
1047 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1049 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1051 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1052 buffer overrun for (non-chunking) other transports.
1054 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1055 TLS1.3, means that a server rejecting a client certificate is not visible
1056 to the client until the first read of encrypted data (typically the
1057 response to EHLO). Add detection for that case and treat it as a failed
1058 TLS connection attempt, so that the normal retry-in-clear can work (if
1059 suitably configured).
1061 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1062 and/or domain. Found and fixed by Jason Betts.
1064 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1065 configuration). If a CNAME target was not a wellformed name pattern, a
1068 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1069 the OS reports them interleaved with other addresses.
1071 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1072 used both for input and for a verify callout, both encrypted, SMTP
1073 responses being sent by the server could be lost. This resulted in
1074 dropped connections and sometimes bounces generated by a peer sending
1077 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1078 string. Previously it was possible to cause undefined behaviour in a
1079 library routine (usually a crash). Found by "zerons".
1081 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1084 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1085 API was removed, so update to use the newer ones.
1087 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1088 any timeout set, is taking a long time. Previously we would hang on to a
1089 rotated logfile "forever" if the input was arriving with long gaps
1090 (a previous attempt to fix addressed lack, for a long time, of initial
1093 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1094 shared (NFS) environment. The length of the tempfile name is now
1095 4 + 16 ("hdr.$message_exim_id") which might break on file
1096 systems which restrict the file name length to lower values.
1097 (It was "hdr.$pid".)
1099 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1100 shared (NFS) environment.
1102 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1103 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1106 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1107 on some platforms for bit 31.
1109 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1110 to changes apparently associated with TLS1.3 handling some of the APIs
1111 previously used were either nonfunctional or inappropriate. Strings
1112 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1113 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1114 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1115 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1116 and the use of specific cipher names in the encrypted= ACL condition.
1118 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1120 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1121 verification result was not updated unless hosts_require_ocsp applied.
1123 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1124 queue_list_requires_admin set to false, non-admin users were denied the
1127 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1128 directory-of-certs mode. Previously they were advertised despite the
1131 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1132 A single TCP connection by a client will now hold a TLS connection open
1133 for multiple message deliveries, by default. Previously the default was to
1136 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1137 default. If built with the facility, DANE will be used. The facility
1138 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1140 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1141 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1142 must be defined and you must still, unless you define DISABLE_TLS, manage
1143 the the include-dir and library-file requirements that go with that
1144 choice. Non-TLS builds are still supported.
1146 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1149 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1150 default. If the platform supports and has the facility enabled, it will
1151 be requested on all coneections.
1153 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1154 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1156 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1158 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1159 success-DSN messages. Previously the From: header was always the default
1160 one for these; the option was ignored.
1162 JH/28 Fix the timeout on smtp response to apply to the whole response.
1163 Previously it was reset for every read, so a teergrubing peer sending
1164 single bytes within the time limit could extend the connection for a
1165 long time. Credit to Qualsys Security Advisory Team for the discovery.
1167 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1168 delivery address, which leaked information of the results of local
1169 forwarding. Change to the original envelope recipient address, per
1172 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1173 requested. Previously not bounce was generated and a log entry of
1174 error ignored was made.
1176 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1178 JH/32 Introduce a general tainting mechanism for values read from the input
1179 channel, and values derived from them. Refuse to expand any tainted
1180 values, to catch one form of exploit.
1182 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1183 was unused and the unexpanded text used for the test. Found and
1184 fixed by Ruben Jenster.
1186 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1187 an attempt to use a TLS library read routine dereffed a nul pointer,
1190 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1191 them in our smtp response.
1193 JH/36 Have the general router option retry_use_local_part default to true when
1194 any of the restrictive preconditions are set (to anything). Previously it
1195 was only for check_local user. The change removes one item of manual
1196 configuration which is required for proper retries when a remote router
1197 handles a subset of addresses for a domain.
1199 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1200 link count into consideration.
1202 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1203 caused the extension of big_buffer, the following lines were ignored.
1205 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1206 accordance with RFC 2308. Previously there was no expiry, so a longlived
1207 receive process (eg. due to ACL delays) versus a short SOA value could
1210 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1212 JH/39 Promote DMARC support to mainline.
1214 JH/40 Bug 2452: Add a References: header to DSNs.
1216 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1217 parameters. The relevant library call is documented as "Deprecated: This
1218 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1219 3.6.0, DH parameters are negotiated following RFC7919."
1221 HS/06 Change the default of dnssec_request_domains to "*"
1223 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1224 carried on and emitted a BDAT command, even when PIPELINING was not
1227 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1228 buffer was used for the filename, resulting in a trap when tainted
1229 arguments (eg. $domain) were used.
1231 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1232 recommended to avoid a possible server-load attack. The feature can be
1233 re-enabled via the openssl_options main cofiguration option.
1235 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1236 for version 4.90 - adding a "more data" boolean to the arguments.
1237 Bumped the ABI version number also, this having been missed previously;
1238 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1239 respect of usage of smtp_printf() by either local_scan code or libraries
1240 accessed via the ${dlfunc } expansion item. Both will need coding
1241 adjustment for any calls to smtp_printf() to match the new function
1242 signature; a FALSE value for the new argument is always safe.
1244 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1245 the file-offset (which the Linux syscall does, and exim expects); this
1246 resulted in an indefinite loop.
1248 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1249 to do ARC verification. The Authentication-Results: header line added
1250 by the configuration then had no ARC item.
1252 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1253 an incorrect mode was used for reading the result, resulting in it being
1260 JH/01 Remove code calling the customisable local_scan function, unless a new
1261 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1263 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1264 non-signal-safe functions being used.
1266 JH/03 Bug 2269: When presented with a received message having a stupidly large
1267 number of DKIM-Signature headers, disable DKIM verification to avoid
1268 a resource-consumption attack. The limit is set at twenty.
1270 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1271 report of oldest_pass in ${authres } in consequence, and separate out
1272 some descriptions of reasons for verification fail.
1274 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1275 files in the spool were present and unlocked. A queue-runner could spot
1276 them, resulting in a duplicate delivery. Fix that by doing the unlock
1277 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1278 add more error-checking on spoolfile handling while that code is being
1281 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1282 No known attacks, no CVE, this is defensive hardening.
1284 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1285 a queue-runner could start a delivery while other operations were ongoing.
1286 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1287 Found and investigated by Tim Stewart. Fix by using the open message data
1288 file handle rather than opening another, and not locally closing it (which
1289 releases a lock) for that case, while creating the temporary .eml format
1290 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1292 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1293 $sender_verify_failure/$recipient_verify_failure to "random".
1295 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1298 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1299 Previously this would segfault.
1301 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1304 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1305 like zero, since the resolver should be doing this for us, But we need one
1306 as a CNAME but no MX presence gets the CNAME returned; we need to check
1307 that doesn't point to an MX to declare it "no MX returned" rather than
1308 "error, loop". A new main option is added so the older capability of
1309 following some limited number of chain links is maintained.
1311 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1313 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1314 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1315 now-deprecated earlier definitions used only specified the range up to TLS
1316 1.2 (in the older-version library docs).
1318 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1320 JH/15 Rework TLS client-side context management. Stop using a global, and
1321 explicitly pass a context around. This enables future use of TLS for
1322 connections to service-daemons (eg. malware scanning) while a client smtp
1323 connection is using TLS; with cutthrough connections this is quite likely.
1325 JH/16 Fix ARC verification to do AS checks in reverse order.
1327 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1329 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1330 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1331 Previously the "utf8" would be re-prepended for every additional message.
1333 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1334 Previously thery were accepted, resulting in issues when attempting to
1335 forward messages to a non-supporting MTA.
1337 PP/02 Let -n work with printing macros too, not just options.
1339 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1340 one parent address was copied, and bogus data was used at delivery-logging
1341 time. Either a crash (after delivery) or bogus log data could result.
1342 Discovery and analysis by Tim Stewart.
1344 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1345 Previously if the string ended mid-character, we did not insert the
1346 promised '?' replacement.
1348 PP/04 Documentation: current string operators work on bytes, not codepoints.
1350 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1351 should pack well giving a smaller memory footprint so better caching and
1352 therefore performance. Group the declarations where this can't be done so
1353 that the byte-sized flag variables are not interspersed among pointer
1354 variables, giving a better chance of good packing by the compiler.
1356 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1357 non-null, to avoid issues with sites running BATV. Previously reports were
1358 sent with an empty envelope sender so looked like bounces.
1360 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1361 The ignore_error flag wasn't being returned from the filter subprocess so
1362 was not set for later routers. Investigation and fix by Matthias Kurz.
1364 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1365 and a msg:complete for the whole, when a message is manually removed using
1366 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1368 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1369 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1370 hope that is portable enough.
1372 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1373 requiring a known-CA anchor certificate; make it now rely entirely on the
1374 TLSA as an anchor. Checking the name on the leaf cert against the name
1375 on the A-record for the host is still done for TA (but not for EE mode).
1377 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1378 included in delivery lines for non-proxied connections, when compiled with
1379 SUPPORT_SOCKS and running with proxy logging enabled.
1381 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1382 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1383 move the existing event to fire before the normal logging of message
1384 failure so that custom logging is bracketed by normal logging.
1386 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1387 msg:fail:internal event. Developement by Matthias Kurz.
1389 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1390 far too small for todays use of crypto signatures stored there. Go all
1391 the way to the max DNS message size of 64kB, even though this might be
1392 overmuch for IOT constrained device use.
1394 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1395 copy a string over itself. The library routine is documented as not
1396 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1398 JH/32 For main options check_spool_space and check_inode_space, where the
1399 platform supports 64b integers, support more than the previous 2^31 kB
1400 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1401 the previous G, M, k.
1403 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1404 $authenticated_fail_id variable on authentication failure. Previously
1407 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1408 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1409 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1410 GNUTLS_SEC_PARAM_MEDIUM.
1412 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1413 side. Previously we would continue as if no SNI had been received.
1415 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1416 "<" character, which should be a list-separator specification, we walked
1417 off past the nul-terimation.
1419 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1420 causes) even when the retry time is not yet met. Previously they were
1421 not, meaning that when (say) an account was over-quota and temp-rejecting,
1422 and multiple senders' messages were queued, only one sender would get
1423 notified on each configured delay_warning cycle.
1425 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1427 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1428 cipher-suites, an error can be left on the stack even for a succeeding
1429 accept; this results in impossible error messages when a later operation
1432 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1433 return error codes indicating retry. Under TLS1.3 this becomes required.
1435 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1436 it only wrote the new authenticators, resulting in a lack of tracking of
1437 peer changes of ESMTP extensions until the next cache flush.
1439 JH/41 Fix the loop reading a message header line to check for integer overflow,
1440 and more-often against header_maxsize. Previously a crafted message could
1441 induce a crash of the recive process; now the message is cleanly rejected.
1443 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1444 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1450 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1451 When redis_servers is set to a list of > 1 element, and the Redis servers
1452 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1453 case of MOVED into a DEFER case instead, thus moving the query onto the
1454 next server in the list. For a cluster of N elements, all N servers must
1455 be defined in redis_servers.
1457 GF/02 Catch and remove uninitialized value warning in exiqsumm
1458 Check for existence of @ARGV before looking at $ARGV[0]
1460 JH/01 Replace the store_release() internal interface with store_newblock(),
1461 which internalises the check required to safely use the old one, plus
1462 the allocate and data copy operations duplicated in both (!) of the
1463 extant use locations.
1465 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1466 modifier. This matches the restriction on the commandline.
1468 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1469 Previously only the last row was returned.
1471 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1472 we assumed that tags in the header were well-formed, and parsed the
1473 element content after inspecting only the first char of the tag.
1474 Assumptions at that stage could crash the receive process on malformed
1477 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1478 While running the DKIM ACL we operate on the Permanent memory pool so that
1479 variables created with "set" persist to the DATA ACL. Also (at any time)
1480 DNS lookups that fail create cache records using the Permanent pool. But
1481 expansions release any allocations made on the current pool - so a dnsdb
1482 lookup expansion done in the DKIM ACL releases the memory used for the
1483 DNS negative-cache, and bad things result. Solution is to switch to the
1484 Main pool for expansions.
1485 While we're in that code, add checks on the DNS cache during store_reset,
1486 active in the testsuite.
1487 Problem spotted, and debugging aided, by Wolfgang Breyha.
1489 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1490 When none of the hosts presented to a transport match an already-open
1491 connection, close it and proceed with the list. Previously we would
1492 queue the message. Spotted by Lena with Yahoo, probably involving
1495 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1496 Previously a spurious "250 OK id=" response was appended to the proper
1499 JH/08 The "support for" informational output now, which built with Content
1500 Scanning support, has a line for the malware scanner interfaces compiled
1501 in. Interface can be individually included or not at build time.
1503 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1504 by the template makefile "src/EDITME". The "STREAM" support for an older
1505 ClamAV interface method is removed.
1507 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1508 rows affected is given instead).
1510 JH/11 The runtime Berkeley DB library version is now additionally output by
1511 "exim -d -bV". Previously only the compile-time version was shown.
1513 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1514 SMTP connection. Previously, when one had more recipients than the
1515 first, an abortive onward connection was made. Move to full support for
1516 multiple onward connections in sequence, handling cutthrough connection
1517 for all multi-message initiating connections.
1519 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1520 routers. Previously, a multi-recipient message would fail to match the
1521 onward-connection opened for the first recipient, and cause its closure.
1523 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1524 a timeout on read on a GnuTLS initiating connection, resulting in the
1525 initiating connection being dropped. This mattered most when the callout
1526 was marked defer_ok. Fix to keep the two timeout-detection methods
1529 JH/15 Relax results from ACL control request to enable cutthrough, in
1530 unsupported situations, from error to silently (except under debug)
1531 ignoring. This covers use with PRDR, frozen messages, queue-only and
1534 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1536 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1537 metadata, resulting in a crash in free().
1539 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1540 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1541 Broken also in d185889f4, with init system revamp.
1543 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1544 Previously we abruptly closed the connection after reading a malware-
1545 found indication; now we go on to read the "scan ok" response line,
1548 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1549 ACL. Previously, a crash would result.
1551 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1552 macro text after a replacement (previously it was only once per line) and
1553 by skipping builtin macros when searching for an uppercase lead character.
1555 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1556 for the build is renamed.
1558 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1559 was allocated for every new TLS startup, meaning one per message. Fix
1560 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1562 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1563 reported the original. Fix to report (as far as possible) the ACL
1564 result replacing the original.
1566 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1567 OpenSSL. Certificate information is loaded for every new TLS startup,
1568 and the resources needed to be freed.
1570 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1572 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1575 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1576 DATA response info to the (existing) per-recipient response info for
1577 the "C=" log element. It can have useful tracking info from the
1578 destination system. Patch from Simon Arlott.
1580 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1581 length value. Previously this would segfault.
1583 HS/02 Support Avast multiline protoocol, this allows passing flags to
1584 newer versions of the scanner.
1586 JH/28 Ensure that variables possibly set during message acceptance are marked
1587 dead before release of memory in the daemon loop. This stops complaints
1588 about them when the debug_store option is enabled. Discovered specifically
1589 for sender_rate_period, but applies to a whole set of variables.
1590 Do the same for the queue-runner and queue-list loops, for variables set
1591 from spool message files. Do the same for the SMTP per-message loop, for
1592 certain variables indirectly set in ACL operations.
1594 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1595 as a multi-recipient message from a mailinglist manager). The coding had
1596 an arbitrary cutoff number of characters while checking for more input;
1597 enforced by writing a NUL into the buffer. This corrupted long / fast
1598 input. The problem was exposed more widely when more pipelineing of SMTP
1599 responses was introduced, and one Exim system was feeding another.
1600 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1601 receiving system, and refused recipients seen by the sending system
1602 (propating to people being dropped from mailing lists).
1603 Discovered and pinpointed by David Carter.
1605 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1606 replaced by the ${authresults } expansion.
1608 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1610 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1611 allows proper process termination in container environments.
1613 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1614 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1616 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1617 and "err_perm", deprecated since 4.83 when the RFC-defined words
1618 "temperror" and "permerror" were introduced.
1620 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1621 transport-filters or DKIM-signing. The restriction was lost in the
1622 consolidation of verify-callout and delivery SMTP handling.
1623 Extend the restriction to also cover ARC-signing.
1625 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1626 in defer=pass mode supply a 450 to the initiator. Previously the message
1629 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1630 tls_require_ciphers is used as before.
1632 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1633 "pass_unscanned". Only tmpfails from the scanner are written to
1634 the paniclog, as they may require admin intervention (permission
1635 denied, license issues). Other scanner errors (like decompression
1636 bombs) do not cause a paniclog entry.
1638 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1639 Previously it was possible to log spurious information in receive log
1642 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1643 triggered odd behaviour from Outlook Express clients.
1645 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1648 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1649 since the IETF WG has not yet settled on that versus the original
1650 "bare" representation.
1652 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1653 Previously the millisecond value corrupted the output.
1654 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1655 corrupted the output.
1661 JH/01 Rework error string handling in TLS interface so that the caller in
1662 more cases is responsible for logging. This permits library-sourced
1663 string to be attached to addresses during delivery, and collapses
1664 pairs of long lines into single ones.
1666 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1667 during configuration. Wildcards are allowed and expanded.
1669 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1670 This permits better logging.
1672 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1673 do not close it down and have the child transport start it up again on
1674 the passed-on TCP connection. Instead, proxy the child (and any
1675 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1676 affected: the continued delivery log lines do not have any DNSSEC, TLS
1677 Certificate or OCSP information. TLS cipher information is still logged.
1679 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1680 identical IP addresses on different listening ports. Will also affect
1683 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1684 add noisy ifdef guards to special-case this sillyness.
1685 Patch from Bernd Kuhls.
1687 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1688 than 255 are no longer allowed.
1690 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1691 Disable the session-cache too, which might reduce our load. Since we
1692 currrectly use a new context for every connection, both as server and
1693 client, there is no benefit for these.
1694 GnuTLS appears to not support tickets server-side by default (we don't
1695 call gnutls_session_ticket_enable_server()) but client side is enabled
1696 by default on recent versions (3.1.3 +) unless the PFS priority string
1699 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1700 <https://reproducible-builds.org/specs/source-date-epoch/>.
1702 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1703 the check for any unsuccessful recipients did not notice the limit, and
1704 erroneously found still-pending ones.
1706 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1707 MSG_MORE. Only in-clear (not on TLS connections).
1709 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1710 transport-filter is involved we can buffer the headers in memory for
1711 creating the signature, and read the spool data file once for the
1712 signature and again for transmission.
1714 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1715 4.77 as the kernel support then wasn't solid, having issues in 64bit
1716 mode. Now, it's been long enough. Add support for FreeBSD also.
1718 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1719 case where the routing stage had gathered several addresses to send to
1720 a host before calling the transport for the first, we previously failed
1721 to close down TLS in the old transport process before passing the TCP
1722 connection to the new process. The new one sent a STARTTLS command
1723 which naturally failed, giving a failed delivery and bloating the retry
1724 database. Investigation and fix prototype from Wolfgang Breyha.
1726 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1727 false-negatives in the check that the sender had not preempted a response
1728 or prompt from Exim (running as a server), due to that code's lack of
1729 awareness of the SMTP input buffering.
1731 PP/04 Add commandline_checks_require_admin option.
1732 Exim drops privileges sanely, various checks such as -be aren't a
1733 security problem, as long as you trust local users with access to their
1734 own account. When invoked by services which pass untrusted data to
1735 Exim, this might be an issue. Set this option in main configuration
1736 AND make fixes to the calling application, such as using `--` to stop
1739 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1740 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1741 responses to those, into a single TLS record each way (this usually means
1742 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1745 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1746 affects you only if you're dancing at the edge of the param size limits.
1747 If you are, and this message makes sense to you, then: raise the
1748 configured limit or use OpenSSL 1.1. Nothing we can do for older
1751 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1752 cmdline element to get the documented default one. Previously it was
1755 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1756 are unsuable for non-message-specific reasons - eg. connection timeout,
1757 banner-time rejection.
1759 JH/16 Fix logging of delivery remote port, when specified by router, under
1762 PP/06 Repair manualroute's ability to take options in any order, even if one
1763 is the name of a transport.
1766 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1768 JH/17 Change the list-building routines interface to use the expanding-string
1769 triplet model, for better allocation and copying behaviour.
1771 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1772 Previously it was constructed the first time a possibly-matching string
1773 was met in the configuration file input during startup; now it is done
1776 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1777 compatible one, to avoid the (poorly documented) possibility of a config
1778 file in the working directory redirecting the DB files, possibly correpting
1779 some existing file. CVE-2017-10140 assigned for BDB.
1781 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1782 cache-hot. Previously, although the result was properly cached, the
1783 initial verify call returned a defer.
1785 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1786 the main verify for receipient in uncached-mode.
1788 JH/22 Retire historical build files to an "unsupported" subdir. These are
1789 defined as "ones for which we have no current evidence of testing".
1791 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1792 if present. Previously it was ignored.
1794 JH/24 Start using specified-initialisers in C structure init coding. This is
1795 a C99 feature (it's 2017, so now considered safe).
1797 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1798 if was a fixed-sized field and bitmask ops via macros; it is now more
1801 PP/07 GitHub PR 56: Apply MariaDB build fix.
1802 Patch provided by Jaroslav Škarvada.
1804 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1805 during Coverity cleanups [4.87 JH/47]
1806 Diagnosis and fix provided by Michael Fischer v. Mollard.
1808 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1809 the right size to place the terminating semicolon on its own folded
1810 line, the header hash was calculated to an incorrect value thanks to
1811 the (relaxed) space the fold became.
1813 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1814 and confused the parent.
1816 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1817 which could crash as a result. This could lead to undeliverable messages.
1819 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1822 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1823 matters on fast-turnover and PID-randomising systems, which were getting
1824 out-of-order delivery.
1826 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1827 a possibly-overlapping copy. The symptom was that "Remote host closed
1828 connection in response to HELO" was logged instead of the actual 4xx
1831 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1832 Previously only that bufferd was discarded, resulting in SYMTP command
1835 JH/32 DKIM: when a message has multiple signatures matching an identity given
1836 in dkim_verify_signers, run the dkim acl once for each. Previously only
1837 one run was done. Bug 2189.
1839 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1840 "panic the current process" to "deliberately defer". The panic log is
1841 still written with the problem list name; the mail and reject logs now
1842 get a temp-reject line for the message that was being handled, saying
1843 something like "domains check lookup or other defer". The SMTP 451
1844 message is still "Temporary local problem".
1846 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1847 A crafted sequence of BDAT commands could result in in-use memory beeing
1848 freed. CVE-2017-16943.
1850 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1851 from SMTP input. Previously it was always done; now only done for DATA
1852 and not BDAT commands. CVE-2017-16944.
1854 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1855 to the message (such as an overlong header line). Previously this was
1856 not done and we did not exit BDAT mode. Followon from the previous item
1857 though a different problem.
1863 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1864 than -2003 did; needs libidn2 in addition to libidn.
1866 JH/02 The path option on a pipe transport is now expanded before use.
1868 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1869 Patch provided by "Björn", documentation fix added too.
1871 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1872 missing a wire-to-host endian conversion.
1874 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1875 close after a BDAT command line could be taken as a following command,
1876 giving a synch failure. Fix by only checking for synch immediately
1877 before acknowledging the chunk.
1879 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1880 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1881 macro. Patches provided by Josh Soref.
1883 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1884 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1885 by IANA as a service extension. However, John Klensin suggests that we
1888 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1889 the dkim code may be unix-mode line endings rather than smtp wire-format
1890 CRLF, so prepend a CR to any bare LF.
1892 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1893 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1895 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1896 the first were themselves being wrongly included in the feed into dkim
1897 processing; with most chunk sizes in use this resulted in an incorrect
1898 body hash calculated value.
1900 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1901 DKIM signature block, for verification. Although advised against by
1902 standards it is specifically not ruled illegal.
1904 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1906 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1907 missing a body hash (the bh= tag).
1909 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1910 It seems that HAProxy sends the Proxy Protocol information in clear and
1911 only then does a TLS startup, so do the same.
1913 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1914 TCP connections (such as for Spamd) unless the daemon successfully set
1915 Fast Open mode on its listening sockets. This fixes breakage seen on
1916 too-old kernels or those not configured for Fast Open, at the cost of
1917 requiring both directions being enabled for TFO, and TFO never being used
1918 by non-daemon-related Exim processes.
1920 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1921 endings, at least on the first header line. Try to canonify any that get
1922 past that check, despite the cost.
1924 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1925 now limited to an arbitrary five deep, while parsing addresses with the
1926 strip_excess_angle_brackets option enabled.
1928 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1929 instead leave the unprompted TLS handshake in socket buffer for the
1930 TLS library to consume.
1932 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1934 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1936 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1937 region in one operation, for speed, and this leaves assigned pointers
1938 dangling. Add checks run only under the testsuite which checks all
1939 variables at a store-reset and panics on a dangling pointer; add code
1940 explicitly nulling out all the variables discovered. Fixes one known
1941 bug: a transport crash, where a dangling pointer for $sending_ip_address
1942 originally assigned in a verify callout, is re-used.
1944 PP/06 Drop '.' from @INC in various Perl scripts.
1946 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1948 PP/08 Reduce a number of compilation warnings under clang; building with
1949 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1950 should be warning-free.
1952 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1954 HS/01 Fix portability problems introduced by PP/08 for platforms where
1955 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1957 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1958 chunk. This allows us to accept broken chunked messages. We need a more
1959 general solution here.
1961 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1962 already-broken messages in the queue.
1964 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1966 JH/19 Fix reference counting bug in routing-generated-address tracking.
1972 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1973 supports it and a size is available (ie. the sending peer gave us one).
1975 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1976 years of being deprecated). The replacements are the ACLs
1977 acl_smtp_mime and acl_not_smtp_mime.
1979 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1980 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1981 or even an in-clear connection were permitted. Now, if the host lookup
1982 was dnssec and dane was requested then the host is only used if the
1983 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1984 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1985 if one fails this test.
1986 This means that a poorly-configured remote DNS will make it incommunicado;
1987 but it protects against a DNS-interception attack on it.
1989 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1990 non-noisy when a race steals the message being considered.
1992 JH/05 If main configuration option tls_certificate is unset, generate a
1993 self-signed certificate for inbound TLS connections.
1995 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1996 in rewrites and routers.
1998 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1999 and logged a warning sing 4.83; now they are a configuration file error.
2001 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2002 (lacking @domain). Apply the same qualification processing as RCPT.
2004 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2006 JH/10 Support ${sha256:} applied to a string (as well as the previous
2009 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2010 a cutthrough deliver is pending, as we always want to make a connection.
2011 This also avoids re-routing the message when later placing the cutthrough
2012 connection after a verify cache hit.
2013 Do not update it with the verify result either.
2015 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2016 when routing results in more than one destination address.
2018 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2019 signing (which inhibits the cutthrough capability). Previously only
2020 the presence of an option was tested; now an expansion evaluating as
2021 empty is permissible (obviously it should depend only on data available
2022 when the cutthrough connection is made).
2024 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2025 the relevant preceding SMTP command did not note the pipelining mode.
2027 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2028 Previously they were not counted.
2030 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2031 as one having no matching records. Previously we deferred the message
2032 that needed the lookup.
2034 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2035 distinguished as "(=".
2037 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2038 for missing MX records. Previously it only worked for missing A records.
2040 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2042 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2043 after the data-go-ahead and data-ack. Patch from Jason Betts.
2045 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2046 even for a "none" policy. Patch from Tony Meyer.
2048 JH/22 Fix continued use of a connection for further deliveries. If a port was
2049 specified by a router, it must also match for the delivery to be
2052 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2053 When one of the recipients of a message was unsuitable for the connection
2054 (has no matching addresses), we lost track of needing to mark it
2055 deferred. As a result mail would be lost.
2057 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2059 JH/25 Decoding ACL controls is now done using a binary search; the source code
2060 takes up less space and should be simpler to maintain. Merge the ACL
2061 condition decode tables also, with similar effect.
2063 JH/26 Fix problem with one_time used on a redirect router which returned the
2064 parent address unchanged. A retry would see the parent address marked as
2065 delivered, so not attempt the (identical) child. As a result mail would
2068 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2069 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2070 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2073 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2074 Main config options check_{log,spool}_{inodes,space} are now
2075 100 inodes, 10MB unless set otherwise in the configuration.
2077 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2078 Previously it only applied to the main-section connection policy
2081 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2083 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2084 by me. Added RFC7919 DH primes as an alternative.
2086 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2087 are not in the system include path.
2089 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2090 GnuTLS, when a session startup failed (eg because the client disconnected)
2091 Exim did stdio operations after fclose. This was exposed by a recent
2092 change which nulled out the file handle after the fclose.
2094 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2095 signed directly by the cert-signing cert, rather than an intermediate
2096 OCSP-signing cert. This is the model used by LetsEncrypt.
2098 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2100 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2101 an incoming connection.
2103 HS/02 Bug 1802: Do not half-close the connection after sending a request
2106 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2107 fallback to "prime256v1".
2109 JH/34 SECURITY: Use proper copy of DATA command in error message.
2110 Could leak key material. Remotely exploitable. CVE-2016-9963.
2116 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2117 and 3.4.4 - once the server is enabled to respond to an OCSP request
2118 it does even when not requested, resulting in a stapling non-aware
2119 client dropping the TLS connection.
2121 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2122 support variable-length bit vectors. No functional change.
2124 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2125 The I= interface field on outgoing lines is now after the H= remote
2126 host field, same as incoming lines. There is a separate
2127 outgoing_interface log selector which allows you to disable the
2130 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2131 If not running log_selector +smtp_connection the mainlog would be held
2132 open indefinitely after a "too many connections" event, including to a
2133 deleted file after a log rotate. Leave the per net connection logging
2134 leaving it open for efficiency as that will be quickly detected by the
2135 check on the next write.
2137 HS/01 Bug 1671: Fix post transport crash.
2138 Processing the wait-<transport> messages could crash the delivery
2139 process if the message IDs didn't exist for some reason. When
2140 using 'split_spool_directory=yes' the construction of the spool
2141 file name failed already, exposing the same netto behaviour.
2143 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2144 mime_regex ACL conditions.
2146 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2147 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2148 to HELO, local diagnostic string.
2150 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2151 log line to debug. Even when configured with a tls authenticator many
2152 client connections are expected to not authenticate in this way, so
2153 an authenticate fail is not an error.
2155 HS/02 Add the Exim version string to the process info. This way exiwhat
2156 gives some more detail about the running daemon.
2158 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2159 matter for fast-change records such as DNSBLs.
2161 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2162 retry record, even if constant. There may be multiple transports with
2163 different interface settings and the retry behaviour needs to be kept
2166 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2168 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2170 JH/10 Bug 840: fix log_defer_output option of pipe transport
2172 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2173 hosts_randomize. This matters a lot when combined with mua_wrapper.
2175 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2176 ${quote_pgsql:<string>} operator.
2178 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2180 JH/14 Change header file name for freeradius-client. Relevant if compiling
2181 with Radius support; from the Gentoo tree and checked under Fedora.
2183 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2185 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2186 option transport_filter, meaning no filtering.
2188 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2190 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2191 defaults to "*" (all hosts). The variable is now available when not built
2192 with TLS, default unset, mainly to enable keeping the testsuite sane.
2193 If a server certificate is not supplied (via tls_certificate) an error is
2194 logged, and clients will find TLS connections fail on startup. Presumably
2195 they will retry in-clear.
2196 Packagers of Exim are strongly encouraged to create a server certificate
2197 at installation time.
2199 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2200 with the $config_file variable.
2202 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2203 in transport context, after the attempt, and per-recipient. The latter type
2204 is per host attempted. The event data is the error message, and the errno
2205 information encodes the lookup type (A vs. MX) used for the (first) host,
2206 and the trailing two digits of the smtp 4xx response.
2208 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2209 to write to mainlog (or rejectlog, paniclog) in the window between file
2210 creation and permissions/ownership being changed. Particularly affects
2211 installations where exicyclog is run as root, rather than exim user;
2212 result is that the running daemon panics and dies.
2214 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2216 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2217 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2218 "pri" and "weight". Note that the previous implicit priority given by the
2219 list order is no longer honoured.
2221 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2222 for DKIM processing.
2224 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2225 by defining SUPPORT_SOCKS.
2227 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2228 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2229 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2230 variables are renamed to proxy_{local,external}_{address,port}.
2232 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2233 for a build by defining SUPPORT_I18N
2235 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2236 of the query string, and make ${quote_redis:} do that quoting.
2238 JH/29 Move Events support from Experimental to mainline, enabled by default
2239 and removable for a build by defining DISABLE_EVENT.
2241 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2243 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2244 cached by the daemon.
2246 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2247 by defining LOOKUP_REDIS. The libhiredis library is required.
2249 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2250 keys are given for lookup.
2252 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2253 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2254 only supported when built with TLS support. The PolarSSL SHA routines
2255 are still used when the TLS library is too old for convenient support.
2257 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2258 openssl_options), for security. OpenSSL forces this from version 1.1.0
2259 server-side so match that on older versions.
2261 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2262 allocation for $value could be released as the expansion processing
2263 concluded, but leaving the global pointer active for it.
2265 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2266 and to use the domains and local_parts ACL conditions.
2268 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2269 incorrectly not doubled on cutthrough transmission, hence seen as a
2270 body-termination at the receiving system - resulting in truncated mails.
2271 Commonly the sender saw a TCP-level error, and retransmitted the message
2272 via the normal store-and-forward channel. This could result in duplicates
2273 received - but deduplicating mailstores were liable to retain only the
2274 initial truncated version.
2276 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2278 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2280 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2281 we're in there, support oversigning also; bug 1309.
2283 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2285 HS/04 Add support for keep_environment and add_environment options.
2287 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2288 either intentional arithmetic overflow during PRNG, or testing config-
2291 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2292 delivery resulted in actual delivery. Cancel cutthrough before DATA
2295 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2296 rejects a recipient: pass the reject to the originator.
2298 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2299 Many were false-positives and ignorable, but it's worth fixing the
2302 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2303 for the new environment-manipulation done at startup. Move the routines
2304 from being local to tls.c to being global via the os.c file.
2306 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2307 an extract embedded as result-arg for a map, the first arg for extract
2308 is unavailable so we cannot tell if this is a numbered or keyed
2309 extraction. Accept either.
2315 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2318 JH/02 The smtp transport option "multi_domain" is now expanded.
2320 JH/03 The smtp transport now requests PRDR by default, if the server offers
2323 JH/04 Certificate name checking on server certificates, when exim is a client,
2324 is now done by default. The transport option tls_verify_cert_hostnames
2325 can be used to disable this per-host. The build option
2326 EXPERIMENTAL_CERTNAMES is withdrawn.
2328 JH/05 The value of the tls_verify_certificates smtp transport and main options
2329 default to the word "system" to access the system default CA bundle.
2330 For GnuTLS, only version 3.0.20 or later.
2332 JH/06 Verification of the server certificate for a TLS connection is now tried
2333 (but not required) by default. The verification status is now logged by
2334 default, for both outbound TLS and client-certificate supplying inbound
2337 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2340 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2341 Status Notification (bounce) messages are now MIME format per RFC 3464.
2342 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2343 under the control of the dsn_advertise_hosts option, and routers may
2344 have a dsn_lasthop option.
2346 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2347 default, modifiable by a malware= option. The list separator for
2348 the options can now be changed in the usual way. Bug 68.
2350 JH/10 The smtp_receive_timeout main option is now expanded before use.
2352 JH/11 The incoming_interface log option now also enables logging of the
2353 local interface on delivery outgoing connections.
2355 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2356 if the interface and destination host and port all match.
2358 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2361 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2362 Patch from Andrew Lewis.
2364 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2365 now supports optional time-restrictions, weighting, and priority
2366 modifiers per server. Patch originally by <rommer@active.by>.
2368 JH/16 The spamd_address main option now supports a mixed list of local
2369 and remote servers. Remote servers can be IPv6 addresses, and
2370 specify a port-range.
2372 JH/17 Bug 68: The spamd_address main option now supports an optional
2373 timeout value per server.
2375 JH/18 Bug 1581: Router and transport options headers_add/remove can
2376 now have the list separator specified.
2378 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2381 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2384 JH/21 Support for the A6 type of dns record is withdrawn.
2386 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2387 rather than the verbs used.
2389 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2390 from 255 to 1024 chars.
2392 JH/24 Verification callouts now attempt to use TLS by default.
2394 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2395 are generic router options now. The defaults didn't change.
2397 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2398 Original patch from Alexander Shikoff, worked over by JH.
2400 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2401 files if ZCAT_COMMAND is not executable.
2403 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2405 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2407 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2408 Normally benign, it bites when the pair was led to by a CNAME;
2409 modern usage is to not canonicalize the domain to a CNAME target
2410 (and we were inconsistent anyway for A-only vs AAAA+A).
2412 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2414 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2415 when evaluating $sender_host_dnssec.
2417 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2418 $sender_helo_dnssec variable.
2420 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2422 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2424 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2426 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2427 documented as working, but never had. Support all but $spam_report.
2429 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2430 added for tls authenticator.
2432 HS/03 Add perl_taintmode main config option
2438 TL/01 When running the test suite, the README says that variables such as
2439 no_msglog_check are global and can be placed anywhere in a specific
2440 test's script, however it was observed that placement needed to be near
2441 the beginning for it to behave that way. Changed the runtest perl
2442 script to read through the entire script once to detect and set these
2443 variables, reset to the beginning of the script, and then run through
2444 the script parsing/test process like normal.
2446 TL/02 The BSD's have an arc4random API. One of the functions to induce
2447 adding randomness was arc4random_stir(), but it has been removed in
2448 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2449 function when detected.
2451 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2452 cause callback expansion.
2454 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2455 syntax errors in an expansion can be treated as a string instead of
2456 logging or causing an error, due to the internal use of bool_lax
2457 instead of bool when processing it.
2459 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2460 server certificates when making smtp deliveries.
2462 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2464 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2466 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2468 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2469 Merged patch from Sebastian Wiedenroth.
2471 JH/05 Fix results-pipe from transport process. Several recipients, combined
2472 with certificate use, exposed issues where response data items split
2473 over buffer boundaries were not parsed properly. This eventually
2474 resulted in duplicates being sent. This issue only became common enough
2475 to notice due to the introduction of connection certificate information,
2476 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2478 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2479 size buffer was used, resulting in syntax errors when an expansion
2482 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2483 version 3.3.6 or later.
2485 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2486 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2487 both become "event_action", the variables become $event_name, $event_data
2488 and $event_defer_errno. There is a new variable $verify_mode, usable in
2489 routers, transports and related events. The tls:cert event is now also
2490 raised for inbound connections, if the main configuration event_action
2493 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2494 early OCSP support, but no stapling (appears to be less than 1.0.0).
2496 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2497 server certificate names available under the smtp transport option
2498 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2501 JH/10 Time-related extraction expansions from certificates now use the main
2502 option "timezone" setting for output formatting, and are consistent
2503 between OpenSSL and GnuTLS compilations. Bug 1541.
2505 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2506 encoded parameter in the incoming message. Bug 1558.
2508 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2509 include certificate info, eximon was claiming there were spoolfile
2512 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2514 JH/14 Log delivery-related information more consistently, using the sequence
2515 "H=<name> [<ip>]" wherever possible.
2517 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2518 are problematic for Debian distribution, omit them from the release
2521 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2523 JH/16 Fix string representation of time values on 64bit time_t architectures.
2526 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2527 output list separator was used.
2532 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2533 checkers that were complaining about end of non-void function with no
2536 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2537 This was a regression introduced in 4.83 by another bugfix.
2539 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2541 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2542 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2548 TF/01 Correctly close the server side of TLS when forking for delivery.
2550 When a message was received over SMTP with TLS, Exim failed to clear up
2551 the incoming connection properly after forking off the child process to
2552 deliver the message. In some situations the subsequent outgoing
2553 delivery connection happened to have the same fd number as the incoming
2554 connection previously had. Exim would try to use TLS and fail, logging
2555 a "Bad file descriptor" error.
2557 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2558 utilities have not been installed.
2560 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2561 temporary space as the ACL may create new global variables.
2563 TL/01 LDAP support uses per connection or global context settings, depending
2564 upon the detected version of the libraries at build time.
2566 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2567 to extract and use the src ip:port in logging and expansions as if it
2568 were a direct connection from the outside internet. PPv2 support was
2569 updated based on HAProxy spec change in May 2014.
2571 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2573 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2574 Properly escape header and check for NULL return.
2576 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2579 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2581 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2582 characters in header names, implemented as a verify condition.
2583 Contributed by Michael Fischer v. Mollard.
2585 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2586 results permerror and temperror. Previous values are deprecated but
2587 still accepted. In a future release, err_perm and err_temp will be
2588 completely removed, which will be a backward incompatibility if the
2589 ACL tests for either of these two old results. Patch contributed by
2590 user bes-internal on the mailing list.
2592 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2594 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2595 selectors, in both main and reject logs.
2597 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2600 JH/07 Add malware type "sock" for talking to simple daemon.
2602 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2604 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2605 routers/transports under cutthrough routing.
2607 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2608 numbers. Touch up "bool" conditional to keep the same definition.
2610 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2612 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2614 JH/12 Expand items in router/transport headers_add or headers_remove lists
2615 individually rather than the list as a whole. Bug 1452.
2617 Required for reasonable handling of multiple headers_ options when
2618 they may be empty; requires that headers_remove items with embedded
2619 colons must have them doubled (or the list-separator changed).
2621 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2622 view the policy declared in the DMARC record. Currently, $dmarc_status
2623 is a combined value of both the record presence and the result of the
2626 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2628 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2629 dnslookup router and the smtp transport (applying to the forward
2632 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2633 of ldap servers used for a specific lookup. Patch provided by Heiko
2636 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2637 New variable $lookup_dnssec_authenticated for observability.
2639 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2640 Patch submitted by Lars Timman.
2642 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2644 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2645 Requires trusted mode and valid format message id, aborts otherwise.
2646 Patch contributed by Heiko Schlichting.
2648 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2649 certextract with support for various fields. Bug 1358.
2651 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2652 is requested by default, modifiable by smtp transport option
2655 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2656 operate on certificate variables to give certificate fingerprints
2657 Also new ${sha256:cert_variable}.
2659 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2661 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2662 Christian Aistleitner.
2664 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2666 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2667 file. Patch from Wolfgang Breyha.
2669 JH/25 Expand the coverage of the delivery $host and $host_address to
2670 client authenticators run in verify callout. Bug 1476.
2672 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2673 align with daemon_smtp_ports. Bug 72.
2675 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2676 support and error reporting did not work properly.
2678 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2679 and is readable. Patch from Andrew Colin Kissa.
2681 TL/14 Enhance documentation of ${run expansion and how it parses the
2682 commandline after expansion, particularly in the case when an
2683 unquoted variable expansion results in an empty value.
2685 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2687 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2688 by renaming away from it. Addresses GH issue 15, reported by
2691 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2693 TL/15 SECURITY: prevent double expansion in math comparison functions
2694 (can expand unsanitized data). Not remotely exploitable.
2701 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2703 PP/02 Make -n do something, by making it not do something.
2704 When combined with -bP, the name of an option is not output.
2706 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2709 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2710 $sender_host_name and config options to manage this, and basic check
2713 PP/05 DSCP support for outbound connections and control modifier for inbound.
2715 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2716 (Only plugin which currently uses this is kerberos4, which nobody should
2717 be using, but we should make it available and other future plugins might
2718 conceivably use it, even though it would break NAT; stuff *should* be
2719 using channel bindings instead).
2721 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2722 name; added for Sendmail compatibility; requires admin caller.
2723 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2724 just ignore it); requires trusted caller.
2725 Also parse but ignore: -Ac -Am -X<logfile>
2728 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2730 TL/02 Add +smtp_confirmation as a default logging option.
2732 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2733 Patch by Magnus Holmgren from 2007-02-20.
2735 TL/04 Bugzilla 1281 - Spec typo.
2736 Bugzilla 1283 - Spec typo.
2737 Bugzilla 1290 - Spec grammar fixes.
2739 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2741 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2743 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2744 mailing list by by Dmitry Isaikin.
2746 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2748 JH/02 Support "G" suffix to numbers in ${if comparisons.
2750 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2752 NM/01 Bugzilla 1197 - Spec typo
2753 Bugzilla 1196 - Spec examples corrections
2755 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2757 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2758 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2761 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2762 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2764 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2765 "acl {{name}{arg}...}", and optional args on acl condition
2768 JH/05 Permit multiple router/transport headers_add/remove lines.
2770 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2772 JH/07 Avoid using a waiting database for a single-message-only transport.
2773 Performance patch from Paul Fisher. Bugzilla 1262.
2775 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2778 JH/09 Add $headers_added variable, with content from use of ACL modifier
2779 add_header (but not yet added to the message). Bugzilla 199.
2781 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2782 Pulled from Bugzilla 817 by Wolfgang Breyha.
2784 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2786 (nb: this is the same fix as in Exim 4.80.1)
2788 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2791 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2794 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2796 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2797 advertise SMTP AUTH mechanism to us, instead of a generic
2798 protocol violation error. Also, make Exim more robust to bad
2799 data from the Dovecot auth socket.
2801 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2803 When a queue runner is handling a message, Exim first routes the
2804 recipient addresses, during which it prunes them based on the retry
2805 hints database. After that it attempts to deliver the message to
2806 any remaining recipients. It then updates the hints database using
2809 So if a recipient address works intermittently, it can get repeatedly
2810 deferred at routing time. The retry hints record remains fresh so the
2811 address never reaches the final cutoff time.
2813 This is a fairly common occurrence when a user is bumping up against
2814 their storage quota. Exim had some logic in its local delivery code
2815 to deal with this. However it did not apply to per-recipient defers
2816 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2818 This change adds a proper retry rule check during routing so that the
2819 final cutoff time is checked against the message's age. We only do
2820 this check if there is an address retry record and there is not a
2821 domain retry record; this implies that previous attempts to handle
2822 the address had the retry_use_local_parts option turned on. We use
2823 this as an approximation for the destination being like a local
2824 delivery, as in LMTP.
2826 I suspect this new check makes the old local delivery cutoff check
2827 redundant, but I have not verified this so I left the code in place.
2829 TF/02 Correct gecos expansion when From: is a prefix of the username.
2831 Test 0254 submits a message to Exim with the header
2835 When I ran the test suite under the user fanf2, Exim expanded
2836 the header to contain my full name, whereas it should have added
2837 a Resent-Sender: header. It erroneously treats any prefix of the
2838 username as equal to the username.
2840 This change corrects that bug.
2842 GF/01 DCC debug and logging tidyup
2843 Error conditions log to paniclog rather than rejectlog.
2844 Debug lines prefixed by "DCC: " to remove any ambiguity.
2846 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2848 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2849 Bug spotted by Jeremy Harris; was flawed since initial commit.
2850 Would have resulted in OCSP responses post-SNI triggering an Exim
2851 NULL dereference and crash.
2853 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2855 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2856 Bug detection, analysis and fix by Samuel Thibault.
2857 Bugzilla 1331, Debian bug #698092.
2859 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2861 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2862 Server implementation by Todd Lyons, client by JH.
2863 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2864 config variable "prdr_enable" controls whether the server
2865 advertises the facility. If the client requests PRDR a new
2866 acl_data_smtp_prdr ACL is called once for each recipient, after
2867 the body content is received and before the acl_smtp_data ACL.
2868 The client is controlled by both of: a hosts_try_prdr option
2869 on the smtp transport, and the server advertisement.
2870 Default client logging of deliveries and rejections involving
2871 PRDR are flagged with the string "PRDR".
2873 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2874 fclose(). Diagnosis by Todd Lyons.
2876 PP/17 Update configure.default to handle IPv6 localhost better.
2877 Patch by Alain Williams (plus minor tweaks).
2880 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2881 This is now consistent with GnuTLS, and is now documented: the
2882 previous undocumented portable approach to treating the option as
2883 unset was to force an expansion failure. That still works, and
2884 an empty string is now equivalent.
2886 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2887 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2888 not performing validation itself.
2890 PP/20 Added force_command boolean option to pipe transport.
2891 Patch from Nick Koston, of cPanel Inc.
2893 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2896 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2898 PP/21 Fix eximon continuous updating with timestamped log-files.
2899 Broken in a format-string cleanup in 4.80, missed when I repaired the
2900 other false fix of the same issue.
2901 Report and fix from Heiko Schlichting.
2904 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2905 Report from Prashanth Katuri.
2907 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2908 It's SecureTransport, so affects any MacOS clients which use the
2909 system-integrated TLS libraries, including email clients.
2911 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2912 using a MIME ACL for non-SMTP local injection.
2913 Report and assistance in diagnosis by Warren Baker.
2915 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2917 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2919 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2920 last id that failed so it may be referenced in subsequent ACL's.
2922 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2925 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2926 ldap library initialization, allowing self-signed CA's to be
2927 used. Also properly sets require_cert option later in code by
2928 using NULL (global ldap config) instead of ldap handle (per
2929 session). Bug diagnosis and testing by alxgomz.
2931 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2932 the src/util/ subdirectory.
2934 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2935 renamed to Transport Post Delivery Action by Jeremy Harris, as
2938 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2939 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2940 redis_servers = needs to be configured which will be used by the redis
2941 lookup. Patch from Warren Baker, of The Packet Hub.
2943 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2945 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2946 hostname or reverse DNS when processing a host list. Used suggestions
2947 from multiple comments on this bug.
2949 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2951 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2952 Missed a few lines, added it to make the runtest require no keyboard
2955 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2956 contains upper case chars. Make router use caseful_local_part.
2958 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2959 support when GnuTLS has been built with p11-kit.
2965 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2967 This, or similar/improved, will also be change PP/11 of 4.82.
2973 PP/01 Handle short writes when writing local log-files.
2974 In practice, only affects FreeBSD (8 onwards).
2975 Bugzilla 1053, with thanks to Dmitry Isaikin.
2977 NM/01 Bugzilla 949 - Documentation tweak
2979 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2982 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2984 PP/02 Implemented gsasl authenticator.
2986 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2988 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2989 `pkg-config foo` for cflags/libs.
2991 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2992 with rest of GSASL and with heimdal_gssapi.
2994 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2995 `pkg-config foo` for cflags/libs for the TLS implementation.
2997 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2998 properties get this fed in as external SSF. A number of robustness
2999 and debugging improvements to the cyrus_sasl authenticator.
3001 PP/08 cyrus_sasl server now expands the server_realm option.
3003 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3004 Patch by Jeremy Harris.
3006 PP/10 Added dbmjz lookup type.
3008 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3010 PP/12 MAIL args handles TAB as well as SP, for better interop with
3011 non-compliant senders.
3012 Analysis and variant patch by Todd Lyons.
3014 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3015 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3016 Patch from Dirk Mueller <dmueller@suse.com>
3018 PP/13 tls_peerdn now print-escaped for spool files.
3019 Observed some $tls_peerdn in wild which contained \n, which resulted
3020 in spool file corruption.
3022 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3023 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3024 or write after TLS renegotiation, which otherwise led to messages
3027 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3028 as a tracking header (ie: a signed header comes before the signature).
3029 Patch from Wolfgang Breyha.
3031 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3032 comma-sep list; embedded commas doubled.
3034 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3036 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3038 Report and patch from Dmitry Banschikov.
3040 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3041 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3042 failures. We appear to now support TLS1.1+ with Exim.
3044 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3045 lets Exim select keys and certificates based upon TLS SNI from client.
3046 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3047 before an outbound SMTP session. New log_selector, +tls_sni.
3049 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3050 NULL dereference. Report and patch from Alun Jones.
3052 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3053 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3054 Not seeing resolver debug output on NetBSD, but suspect this is a
3055 resolver implementation change.
3057 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3058 Left warnings. Added "eximon gdb" invocation mode.
3060 PP/21 Defaulting "accept_8bitmime" to true, not false.
3062 PP/22 Added -bw for inetd wait mode support.
3064 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3065 locate the relevant includes and libraries. Made this the default.
3067 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3068 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3070 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3071 This may cause build issues on older platforms.
3073 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3074 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3075 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3076 Added SNI support via GnuTLS too.
3077 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3079 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3081 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3082 Applied second patch from Janne, implementing suggestion to default
3083 multiple-strings-in-record handling to match SPF spec.
3085 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3087 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3088 read-only, out of scope).
3089 Patch from Wolfgang Breyha, report from Stuart Northfield.
3091 PP/29 Fix three issues highlighted by clang analyser static analysis.
3092 Only crash-plausible issue would require the Cambridge-specific
3093 iplookup router and a misconfiguration.
3094 Report from Marcin Mirosław.
3096 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3098 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3099 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3100 As part of this, removing so much warning spew let me fix some minor
3101 real issues in debug logging.
3103 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3104 assignment on my part. Fixed.
3106 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3107 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3108 Janne Snabb (who went above and beyond: thank you).
3110 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3111 string otherwise requires a connection and a bunch more work and it's
3112 relatively easy to get wrong. Should also expose TLS library linkage
3115 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3116 64-bit ${eval} (JH/03).
3118 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3119 GNU libc to support some of the 64-bit stuff, should not lead to
3120 conflicts. Defined before os.h is pulled in, so if a given platform
3121 needs to override this, it can.
3123 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3124 protection layer was required, which is not implemented.
3125 Bugzilla 1254, patch from Wolfgang Breyha.
3127 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3128 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3129 tls_dhparam take prime identifiers. Also unbreak combination of
3130 OpenSSL+DH_params+TLSSNI.
3132 PP/39 Disable SSLv2 by default in OpenSSL support.
3138 PP/01 Solaris build fix for Oracle's LDAP libraries.
3139 Bugzilla 1109, patch from Stephen Usher.
3141 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3143 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3146 TF/02 Fix a couple more cases where we did not log the error message
3147 when unlink() failed. See also change 4.74-TF/03.
3149 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3150 lock up or crash if it happened to be inside a call to libc when it
3151 got a SIGUSR1 from exiwhat.
3153 The SIGUSR1 handler appends the current process status to the process
3154 log which is later printed by exiwhat. It used to use the general
3155 purpose logging code to do this, but several functions it calls are
3156 not safe for signals.
3158 The new output code in the SIGUSR1 handler is specific to the process
3159 log, and simple enough that it's easy to inspect for signal safety.
3160 Removing some special cases also simplifies the general logging code.
3161 Removing the spurious timestamps from the process log simplifies
3164 TF/04 Improved ratelimit ACL condition.
3166 The /noupdate option has been deprecated in favour of /readonly which
3167 has clearer semantics. The /leaky, /strict, and /readonly update modes
3168 are mutually exclusive. The update mode is no longer included in the
3169 database key; it just determines when the database is updated. (This
3170 means that when you upgrade Exim will forget old rate measurements.)
3172 Exim now checks that the per_* options are used with an update mode that
3173 makes sense for the current ACL. For example, when Exim is processing a
3174 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3175 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3176 must specify per_mail/readonly. If you omit the update mode it defaults to
3177 /leaky where that makes sense (as before) or /readonly where required.
3179 The /noupdate option is now undocumented but still supported for
3180 backwards compatibility. It is equivalent to /readonly except that in
3181 ACLs where /readonly is required you may specify /leaky/noupdate or
3182 /strict/noupdate which are treated the same as /readonly.
3184 A useful new feature is the /count= option. This is a generalization
3185 of the per_byte option, so that you can measure the throughput of other
3186 aggregate values. For example, the per_byte option is now equivalent
3187 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3189 The per_rcpt option has been generalized using the /count= mechanism
3190 (though it's more complicated than the per_byte equivalence). When it is
3191 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3192 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3193 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3194 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3195 Note that using per_rcpt with a non-readonly update mode in more than
3196 one ACL will cause the recipients to be double-counted. (The per_mail
3197 and per_byte options don't have this problem.)
3199 The handling of very low rates has changed slightly. If the computed rate
3200 is less than the event's count (usually one) then this event is the first
3201 after a long gap. In this case the rate is set to the same as this event's
3202 count, so that the first message of a spam run is counted properly.
3204 The major new feature is a mechanism for counting the rate of unique
3205 events. The new per_addr option counts the number of different
3206 recipients that someone has sent messages to in the last time period. It
3207 behaves like per_rcpt if all the recipient addresses are different, but
3208 duplicate recipient addresses do not increase the measured rate. Like
3209 the /count= option this is a general mechanism, so the per_addr option
3210 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3211 example, measure the rate that a client uses different sender addresses
3212 with the options per_mail/unique=$sender_address. There are further
3213 details in the main documentation.
3215 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3217 TF/06 Removed a few PCRE remnants.
3219 TF/07 Automatically extract Exim's version number from tags in the git
3220 repository when doing development or release builds.
3222 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3223 Bugzilla 879. Patch from Paul Fisher.
3225 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3226 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3229 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3231 PP/05 Variable $av_failed, true if the AV scanner deferred.
3232 Bugzilla 1078. Patch from John Horne.
3234 PP/06 Stop make process more reliably on build failure.
3235 Bugzilla 1087. Patch from Heiko Schlittermann.
3237 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3238 Bugzilla 1089. Patch from Heiko Schlittermann.
3240 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3241 Bugzilla 1131. Patch from Holger Weiß.
3243 PP/09 Handle IPv6 addresses with SPF.
3244 Bugzilla 860. Patch from Wolfgang Breyha.
3246 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3248 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3251 PP/11 match_* no longer expand right-hand-side by default.
3252 New compile-time build option, EXPAND_LISTMATCH_RHS.
3253 New expansion conditions, "inlist", "inlisti".
3255 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3257 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3259 PP/14 fix log_write() format string regression from TF/03.
3260 Bugzilla 1152. Patch from Dmitry Isaikin.
3266 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3268 PP/02 Harmonised TLS library version reporting; only show if debugging.
3269 Layout now matches that introduced for other libraries in 4.74 PP/03.
3271 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3273 PP/04 New "dns_use_edns0" global option.
3275 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3278 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3279 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3281 TK/01 Updated PolarSSL code to 0.14.2.
3282 Bugzilla 1097. Patch from Andreas Metzler.
3284 PP/07 Catch divide-by-zero in ${eval:...}.
3285 Fixes bugzilla 1102.
3287 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3290 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3291 format-string attack -- SECURITY: remote arbitrary code execution.
3293 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3294 time unintentionally subject to list matching rules, letting the header
3295 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3296 arbitrary string expansion). This allowed for information disclosure.
3298 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3299 INT_MIN/-1 -- value coerced to INT_MAX.
3305 NM/01 Workaround for PCRE version dependency in version reporting
3308 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3309 This fixes portability to compilers other than gcc, notably
3310 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3312 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3313 makefiles for portability to HP-UX and POSIX correctness.
3315 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3316 Also via indented variable definition in the Makefile.
3317 (Debugging by Oliver Heesakkers).
3319 PP/02 Restore caching of spamd results with expanded spamd_address.
3320 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3322 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3323 Improves build reliability. Fix from: Frank Elsner
3325 NM/02 Fix wide character breakage in the rfc2047 coding
3326 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3328 NM/03 Allow underscore in dnslist lookups
3329 Fixes bug 1026. Patch from Graeme Fowler
3331 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3332 Code patches from Adam Ciarcinski of NetBSD.
3334 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3337 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3338 is logged, to avoid truncation. Patch from John Horne.
3340 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3341 Patch from Jakob Hirsch.
3343 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3344 SQL string expansion failure details.
3345 Patch from Andrey Oktyabrski.
3347 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3348 Patch from Simon Arlott.
3350 PP/09 New lookups functionality failed to compile on old gcc which rejects
3351 extern declarations in function scope.
3352 Patch from Oliver Fleischmann
3354 PP/10 Use sig_atomic_t for flags set from signal handlers.
3355 Check getgroups() return and improve debugging.
3356 Fixed developed for diagnosis in bug 927 (which turned out to be
3359 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3360 Patch from Mark Zealey.
3362 PP/12 Bugzilla 1056: Improved spamd server selection.
3363 Patch from Mark Zealey.
3365 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3366 Based on patch from Heiko Schlittermann.
3368 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3369 Patch from Uwe Doering, sign-off by Michael Haardt.
3371 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3372 variable declaration deep within a block. Bug and patch from
3375 PP/15 lookups-Makefile IRIX compatibility coercion.
3377 PP/16 Make DISABLE_DKIM build knob functional.
3379 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3380 Patch by Simon Arlott
3382 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3383 variable argument macros. Our copy now differs from upstream.
3389 TF/01 Failure to get a lock on a hints database can have serious
3390 consequences so log it to the panic log.
3392 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3393 controlled using the smtp_confirmation log selector.
3395 TF/03 Include the error message when we fail to unlink a spool file.
3397 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3398 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3399 for maintaining out-of-tree patches for some time.
3401 PP/01 Bugzilla 139: Documentation and portability issues.
3402 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3403 Handle per-OS dynamic-module compilation flags.
3405 PP/02 Let /dev/null have normal permissions.
3406 The 4.73 fixes were a little too stringent and complained about the
3407 permissions on /dev/null. Exempt it from some checks.
3408 Reported by Andreas M. Kirchwitz.
3410 PP/03 Report version information for many libraries, including
3411 Exim version information for dynamically loaded libraries. Created
3412 version.h, now support a version extension string for distributors
3413 who patch heavily. Dynamic module ABI change.
3415 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3416 privilege escalation vulnerability whereby the Exim run-time user
3417 can cause root to append content of the attacker's choosing to
3420 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3423 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3424 If dropping privileges for untrusted macros, we disabled normal logging
3425 on the basis that it would fail; for the Exim run-time user, this is not
3426 the case, and it resulted in successful deliveries going unlogged.
3427 Fixed. Reported by Andreas Metzler.
3433 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3434 only prepend for the Resent-* case. Fixes regression introduced in
3435 Exim 4.70 by NM/22 for Bugzilla 607.
3437 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3438 increasing numbers of administrators be bitten by this.
3440 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3442 PP/03 Bugzilla 994: added openssl_options main configuration option.
3444 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3446 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3448 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3450 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3451 without a peer certificate, leading to a segfault because of an
3452 assumption that peers always have certificates. Be a little more
3453 paranoid. Problem reported by Martin Tscholak.
3455 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3456 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3457 NB: ClamAV planning to remove STREAM in "middle of 2010".
3458 CL also introduces -bmalware, various -d+acl logging additions and
3459 more caution in buffer sizes.
3461 PP/09 Implemented reverse_ip expansion operator.
3463 PP/10 Bugzilla 937: provide a "debug" ACL control.
3465 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3467 PP/12 Bugzilla 973: Implement --version.
3469 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3471 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3473 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3475 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3476 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3477 ignore trailing whitespace.
3479 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3481 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3484 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3485 Notification from Dr Andrew Aitchison.
3487 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3488 ExtendedDetectionInfo response format.
3489 Notification from John Horne.
3491 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3494 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3495 XSL and documented dependency on system catalogs, with examples of how
3498 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3501 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3502 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3503 configuration file which is writeable by the Exim user or group.
3505 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3506 of configuration files to cover files specified with the -C option if
3507 they are going to be used with root privileges, not just the default
3510 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3511 option (effectively making it always true).
3513 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3514 files to be used while preserving root privileges.
3516 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3517 that rogue child processes cannot use them.
3519 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3520 run-time user, instead of root.
3522 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3523 Exim run-time user without dropping privileges.
3525 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3526 result string, instead of calling string_vformat() twice with the same
3529 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3530 for other users. Others should always drop root privileges if they use
3531 -C on the command line, even for a whitelisted configure file.
3533 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3535 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3541 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3542 $data_path, and $header_path variables; fixed documentation bugs and
3545 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3546 exipick to access non-standard spools, including the "frozen" queue
3549 NM/01 Bugzilla 965: Support mysql stored procedures.
3550 Patch from Alain Williams
3552 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3554 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3555 Patch from Andreas Metzler
3557 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3558 Patch from Kirill Miazine
3560 NM/05 Bugzilla 671: Added umask to procmail example.
3562 JJ/03 installed exipick 20100323.0, fixing doc bug
3564 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3565 directory. Notification and patch from Dan Rosenberg.
3567 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3569 TK/02 Improve log output when DKIM signing operation fails.
3571 MH/01 Treat the transport option dkim_domain as a colon separated
3572 list, not as a single string, and sign the message with each element,
3573 omitting multiple occurences of the same signer.
3575 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3576 Bugzilla 985, 986. Patch by Simon Arlott
3578 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3579 Patch by Simon Arlott
3581 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3582 MBX locking. Notification from Dan Rosenberg.
3588 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3590 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3592 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3594 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3596 NM/04 Bugzilla 915: Flag broken perl installation during build.
3602 TK/01 Added patch by Johannes Berg that expands the main option
3603 "spamd_address" if it starts with a dollar sign.
3605 TK/02 Write list of recipients to X-Envelope-Sender header when building
3606 the mbox-format spool file for content scanning (suggested by Jakob
3609 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3610 (http://www.dcc-servers.net/) support via dccifd. Activated by
3611 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3613 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3614 by Mark Daniel Reidel <mr@df.eu>.
3616 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3617 When building exim an external PCRE library is now needed -
3618 PCRE is a system library on the majority of modern systems.
3619 See entry on PCRE_LIBS in EDITME file.
3621 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3622 conversation. Added nologin parameter to request.
3623 Patch contributed by Kirill Miazine.
3625 TF/01 Do not log submission mode rewrites if they do not change the address.
3627 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3629 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3630 log files in place. Contributed by Roberto Lima.
3632 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3634 TF/03 Bugzilla 615: When checking the local_parts router precondition
3635 after a local_part_suffix or local_part_prefix option, Exim now
3636 does not use the address's named list lookup cache, since this
3637 contains cached lookups for the whole local part.
3639 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3640 Robert Millan. Documentation is in experimental-spec.txt.
3642 TF/04 Bugzilla 668: Fix parallel build (make -j).
3644 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3646 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3647 Patch provided by Jan Srzednicki.
3649 TF/05 Leading white space used to be stripped from $spam_report which
3650 wrecked the formatting. Now it is preserved.
3652 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3653 that they are available at delivery time.
3655 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3657 TF/08 TLS error reporting now respects the incoming_interface and
3658 incoming_port log selectors.
3660 TF/09 Produce a more useful error message if an SMTP transport's hosts
3661 setting expands to an empty string.
3663 NM/06 Bugzilla 744: EXPN did not work under TLS.
3664 Patch provided by Phil Pennock.
3666 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3667 Patch provided by Richard Godbee.
3669 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3670 acl_smtp_notquit, added index entry.
3672 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3673 Patch provided by Eugene Bujak.
3675 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3676 accept(). Patch provided by Maxim Dounin.
3678 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3679 Patch provided by Phil Pennock.
3681 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3683 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3684 Patch provided by Brad "anomie" Jorsch.
3686 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3687 Patch provided by Dean Brooks.
3689 TK/05 Add native DKIM support (does not depend on external libraries).
3691 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3692 Patch provided by Graeme Fowler.
3694 NM/16 Bugzilla 851: Documentation example syntax fix.
3696 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3698 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3701 NM/19 Bugzilla 745: TLS version reporting.
3702 Patch provided by Phil Pennock.
3704 NM/20 Bugzilla 167: bool: condition support.
3705 Patch provided by Phil Pennock.
3707 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3708 clients. Patch provided by Phil Pennock.
3710 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3711 Patch provided by Brad "anomie" Jorsch.
3713 NM/23 Bugzilla 687: Fix misparses in eximstats.
3714 Patch provided by Heiko Schlittermann.
3716 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3717 Patch provided by Heiko Schlittermann.
3719 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3720 plus update to original patch.
3722 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3724 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3725 Patch provided by David Brownlee.
3727 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3729 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3731 NM/30 Bugzilla 888: TLS documentation bugfixes.
3733 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3735 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3736 Unlike the original bugzilla I have changed all shell scripts in src tree.
3738 NM/33 Bugzilla 898: Transport filter timeout fix.
3739 Patch by Todd Rinaldo.
3741 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3742 Patch by Serge Demonchaux.
3744 NM/35 Bugzilla 39: Base64 decode bug fixes.
3745 Patch by Jakob Hirsch.
3747 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3749 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3751 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3753 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3759 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3760 ALT-N's libdkim that I have put here:
3761 http://duncanthrax.net/exim-experimental/
3763 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3764 were called 'true' and it seems that C99 defines that as a reserved
3765 keyword to be used with 'bool' variable types. That means you could
3766 not include C99-style headers which use bools without triggering
3767 build errors in sieve.c.
3769 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3770 as mailq or other aliases. Changed the --help handling significantly
3771 to do whats expected. exim_usage() emits usage/help information.
3773 SC/01 Added the -bylocaldomain option to eximstats.
3775 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3777 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3779 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3785 PH/01 Another patch from the Sieve maintainer.
3787 PH/02 When an IPv6 address is converted to a string for single-key lookup
3788 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3789 dots are used instead of colons so that keys in lsearch files need not
3790 contain colons. This was done some time before quoting was made available
3791 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3792 (notated using the quote facility) so as to distinguish them from IPv4
3793 keys. This meant that lookups for IP addresses in host lists did not work
3794 for iplsearch lookups.
3796 This has been fixed by arranging for IPv6 addresses to be expressed with
3797 colons if the lookup type is iplsearch. This is not incompatible, because
3798 previously such lookups could never work.
3800 The situation is now rather anomalous, since one *can* have colons in
3801 ordinary lsearch keys. However, making the change in all cases is
3802 incompatible and would probably break a number of configurations.
3804 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3807 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3808 conversion specification without a maximum field width, thereby enabling
3809 a rogue spamd server to cause a buffer overflow. While nobody in their
3810 right mind would setup Exim to query an untrusted spamd server, an
3811 attacker that gains access to a server running spamd could potentially
3812 exploit this vulnerability to run arbitrary code as the Exim user.
3814 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3815 $primary_hostname instead of what libspf2 thinks the hosts name is.
3817 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3818 a directory entry by the name of the lookup key. Previously, if a
3819 symlink pointed to a non-existing file or a file in a directory that
3820 Exim lacked permissions to read, a lookup for a key matching that
3821 symlink would fail. Now it is enough that a matching directory entry
3822 exists, symlink or not. (Bugzilla 503.)
3824 PH/03 The body_linecount and body_zerocount variables are now exported in the
3827 PH/04 Added the $dnslist_matched variable.
3829 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3830 This means they are set thereafter only if the connection becomes
3833 PH/06 Added the client_condition to authenticators so that some can be skipped
3834 by clients under certain conditions.
3836 PH/07 The error message for a badly-placed control=no_multiline_responses left
3837 "_responses" off the end of the name.
3839 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3841 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3842 (without spaces) instead of just copying the configuration text.
3844 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3846 PH/11 Added $max_received_linelength.
3848 PH/12 Added +ignore_defer and +include_defer to host lists.
3850 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3851 way in which PCRE > 7.0 is built.
3853 PH/14 Implemented queue_only_load_latch.
3855 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3856 MAIL command. The effect was to mangle the value on 64-bit systems.
3858 PH/16 Another patch from the Sieve maintainer.
3860 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3862 PH/18 If a system quota error occurred while trying to create the file for
3863 a maildir delivery, the message "Mailbox is full" was not appended to the
3864 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3865 only to a quota excession during the actual writing of the file.
3867 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3868 characters?) which causes problems in log lines. The DN values are now
3869 passed through string_printing() before being added to log lines.
3871 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3872 and InterBase are left for another time.)
3874 PH/21 Added message_body_newlines option.
3876 PH/22 Guard against possible overflow in moan_check_errorcopy().
3878 PH/23 POSIX allows open() to be a macro; guard against that.
3880 PH/24 If the recipient of an error message contained an @ in the local part
3881 (suitably quoted, of course), incorrect values were put in $domain and
3882 $local_part during the evaluation of errors_copy.
3888 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3889 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3892 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3893 issue a MAIL command.
3895 PH/02 In an ACL statement such as
3897 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3899 if a client was not listed at all, or was listed with a value other than
3900 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3901 the condition was not true (as it should be), so access was not denied.
3902 The bug was that the ! inversion was incorrectly passed on to the second
3903 item. This has been fixed.
3905 PH/03 Added additional dnslists conditions == and =& which are different from
3906 = and & when the dns lookup returns more than one IP address.
3908 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3909 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3911 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3912 FSYNC, which compiles an option called disable_fsync that allows for
3913 bypassing fsync(). The documentation is heavily laced with warnings.
3915 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3917 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3918 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3919 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3920 including adding "make clean"; (3) Added -fPIC when compiling the test
3921 dynamically loaded module, to get rid of a warning.
3923 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3924 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3925 = 0s. The bug is otherwise harmless.
3927 PH/07 There was a bug in the dovecot authenticator such that the value of
3928 $auth1 could be overwritten, and so not correctly preserved, after a
3929 successful authentication. This usually meant that the value preserved by
3930 the server_setid option was incorrect.
3932 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3934 PH/09 Installed PCRE release 7.0.
3936 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3937 run for batched SMTP input. It is now run at the start of every message
3938 in the batch. While fixing this I discovered that the process information
3939 (output by running exiwhat) was not always getting set for -bs and -bS
3940 input. This is fixed, and it now also says "batched" for BSMTP.
3942 PH/11 Added control=no_pipelining.
3944 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3945 patch, slightly modified), and move the expansion of helo_data till after
3946 the connection is made in the smtp transport (so it can use these
3949 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3951 PH/14 Added log_selector = +pid.
3953 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3955 PH/16 Add ${if forany and ${if forall.
3957 PH/17 Added dsn_from option to vary the From: line in DSNs.
3959 PH/18 Flush SMTP output before performing a callout, unless control =
3960 no_callout_flush is set.
3962 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3963 was true (the default) a successful delivery failed to delete the retry
3964 item, thus causing premature timeout of the address. The bug is now
3967 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3969 PH/21 Long custom messages for fakedefer and fakereject are now split up
3970 into multiline responses in the same way that messages for "deny" and
3971 other ACL rejections are.
3973 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3974 with slight modification.
3976 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3977 draft, changing the syntax and factoring some duplicate code".
3979 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3980 for deliveries of the second and subsequent messages over the same SMTP
3983 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3984 ${reduce, with only minor "tidies".
3986 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3988 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3989 expansion side effects.
3991 PH/27 When a message times out after an over-quota error from an Exim-imposed
3992 quota, the bounce message says "mailbox is full". This message was not
3993 being given when it was a system quota that was exceeded. It now should
3996 MH/03 Made $recipients available in local_scan(). local_scan() already has
3997 better access to the recipient list through recipients_list[], but
3998 $recipients can be useful in postmaster-provided expansion strings.
4000 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4001 in the case of a MAIL command with additional options following the
4002 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4003 were accidentally chopped off.
4005 PH/29 SMTP synchronization checks are implemented when a command is read -
4006 there is a check that no more input is waiting when there shouldn't be
4007 any. However, for some commands, a delay in an ACL can mean that it is
4008 some time before the response is written. In this time, more input might
4009 arrive, invalidly. So now there are extra checks after an ACL has run for
4010 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4011 pipelining has not been advertised.
4013 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4015 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4016 returned in the SMTP response when smtp_return_error_details was false.
4017 This has been fixed.
4019 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4020 instead of the C I/O that was originally supplied, because problems were
4021 reported on Solaris.
4023 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4024 Exim which did not show up earlier: it was assuming that a call to
4025 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4026 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4027 was a macro that became an assignment, so it seemed to work. This has
4028 changed to a proper function call with a void return, hence the compile
4029 error. Exim's code has been fixed.
4031 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4034 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4036 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4038 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4039 because it turns out to be too big or there is a timeout) there is no
4040 "Completed" line in the log. When some messages of this type were
4041 selected by exigrep, they were listed as "not completed". Others were
4042 picked up by some special patterns. I have improved the selection
4043 criteria to be more general.
4045 PH/38 The host_find_failed option in the manualroute router can now be set
4046 to "ignore", to completely ignore a host whose IP address cannot be
4047 found. If all hosts are ignored, the behaviour is controlled by the new
4048 host_all_ignored option.
4050 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4051 homing or because of multiple MX records with /mx) generated more than
4052 one IP address, and the following item turned out to be the local host,
4053 all the secondary addresses of the first item were incorrectly removed
4054 from the list, along with the local host and any following hosts (which
4055 is what is supposed to happen).
4057 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4058 whoever called Exim into the -H file. In the case of the daemon it was
4059 behaving confusingly. When first started, it used values for whoever
4060 started the daemon, but after a SIGHUP it used the Exim user (because it
4061 calls itself on a restart). I have changed the code so that it now always
4064 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4065 message are rejected with the same error (e.g. no authentication or bad
4066 sender address), and a DATA command is nevertheless sent (as can happen
4067 with PIPELINING or a stupid MUA), the error message that was given to the
4068 RCPT commands is included in the rejection of the DATA command. This is
4069 intended to be helpful for MUAs that show only the final error to their
4072 PH/42 Another patch from the Sieve maintainer.
4074 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4075 Eximstats - Fixed some broken HTML links and added missing column headers
4077 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4078 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4080 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4082 PH/43 Yet another patch from the Sieve maintainer.
4084 PH/44 I found a way to check for a TCP/IP connection going away before sending
4085 the response to the final '.' that terminates a message, but only in the
4086 case where the client has not sent further data following the '.'
4087 (unfortunately, this is allowed). However, in many cases there won't be
4088 any further data because there won't be any more messages to send. A call
4089 to select() can be used: if it shows that the input is "ready", there is
4090 either input waiting, or the socket has been closed. An attempt to read
4091 the next input character can distinguish the two cases. Previously, Exim
4092 would have sent an OK response which the client would never have see.
4093 This could lead to message repetition. This fix should cure that, at
4094 least in a lot of common cases.
4096 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4097 advertised in response to EHLO.
4103 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4104 fixed by 4.65/MH/01 (is this a record?) are fixed:
4106 (i) An empty string was always treated as zero by the numeric comparison
4107 operators. This behaviour has been restored.
4109 (ii) It is documented that the numeric comparison operators always treat
4110 their arguments as decimal numbers. This was broken in that numbers
4111 starting with 0 were being interpreted as octal.
4113 While fixing these problems I realized that there was another issue that
4114 hadn't been noticed. Values of message_size_limit (both the global option
4115 and the transport option) were treated as octal if they started with 0.
4116 The documentation was vague. These values are now always treated as
4117 decimal, and I will make that clear in the documentation.
4123 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4124 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4127 MH/01 Don't check that the operands of numeric comparison operators are
4128 integers when their expansion is in "skipping" mode (fixes bug
4129 introduced by 4.64-PH/07).
4131 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4132 child addresses, Exim now panics and dies. Previously, because the count
4133 is held in a short int, deliveries were likely to be lost. As such a
4134 large number of recipients for a single message is ridiculous
4135 (performance will be very, very poor), I have chosen to impose a limit
4136 rather than extend the field.
4142 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4143 leftover -K file (the existence of which was triggered by #402).
4144 While we were at it, introduced process PID as part of the -K
4145 filename. This should rule out race conditions when creating
4148 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4149 processing considerably. Previous code took too long for large mails,
4150 triggering a timeout which in turn triggers #401.
4152 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4153 in the DK code in transports.c. sendfile() is not really portable,
4154 hence the _LINUX specificness.
4156 TF/01 In the add_headers option to the mail command in an Exim filter,
4157 there was a bug that Exim would claim a syntax error in any
4158 header after the first one which had an odd number of characters
4161 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4162 callout verification, Exim cached a "reject" for the entire domain. This
4163 is correct for most verifications, but it is not correct for a recipient
4164 verification with use_sender or use_postmaster set, because in that case
4165 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4166 case of MAIL FROM:<> rejection from other early rejections (e.g.
4167 rejection of HELO). When verifying a recipient using a non-null MAIL
4168 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4169 Whatever the result of the callout, the value of the domain cache is
4170 left unchanged (for any other kind of callout, getting as far as trying
4171 RCPT means that the domain itself is ok).
4173 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4176 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4177 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4178 session, the daemon ignores ECONNECT errors and logs others; it now
4179 ignores EPIPE as well.
4181 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4182 (quoted-printable decoding).
4184 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4185 later the small subsequent patch to fix an introduced bug.
4187 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4189 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4191 PH/08 An error is now given if message_size_limit is specified negative.
4193 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4194 to be given (somewhat) arbitrary names.
4196 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4199 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4200 miscellaneous code fixes
4202 PH/10 Added the log_reject_target ACL modifier to specify where to log
4205 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4206 hostname. This is wrong, because it relates to the incoming message (and
4207 probably the interface on which it is arriving) and not to the outgoing
4208 callout (which could be using a different interface). This has been
4209 changed to use the value of the helo_data option from the smtp transport
4210 instead - this is what is used when a message is actually being sent. If
4211 there is no remote transport (possible with a router that sets up host
4212 addresses), $smtp_active_hostname is used.
4214 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4215 tweaks were necessary in order to get it to work (see also 21 below):
4216 (a) The code assumed that strncpy() returns a negative number on buffer
4217 overflow, which isn't the case. Replaced with Exim's string_format()
4219 (b) There were several signed/unsigned issues. I just did the minimum
4220 hacking in of casts. There is scope for a larger refactoring.
4221 (c) The code used strcasecmp() which is not a standard C function.
4222 Replaced with Exim's strcmpic() function.
4223 (d) The code set only $1; it now sets $auth1 as well.
4224 (e) A simple test gave the error "authentication client didn't specify
4225 service in request". It would seem that Dovecot has changed its
4226 interface. Fortunately there's a specification; I followed it and
4227 changed what the client sends and it appears to be working now.
4229 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4232 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4233 address A is aliased to B and C, where B exists and C does not. Without
4234 -v the output is "A verified" because verification stops after a
4235 successful redirection if more than one address is generated. However,
4236 with -v the child addresses are also verified. Exim was outputting "A
4237 failed to verify" and then showing the successful verification for C,
4238 with its parentage. It now outputs "B failed to verify", showing B's
4239 parentage before showing the successful verification of C.
4241 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4242 look up a TXT record in a specific list after matching in a combined
4245 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4246 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4247 they consult the DNS. I had assumed they would set it the way they
4248 wanted; and indeed my experiments on Linux seem to show that in some
4249 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4250 To be on the safe side, however, I have now made the interface to
4251 host_find_byname() similar to host_find_bydns(), with an argument
4252 containing the DNS resolver options. The host_find_byname() function now
4253 sets these options at its start, just as host_find_bydns() does. The smtp
4254 transport options dns_qualify_single and dns_search_parents are passed to
4255 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4256 of host_find_byname() use the default settings of RES_DEFNAMES
4257 (qualify_single) but not RES_DNSRCH (search_parents).
4259 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4260 spool_read_header() do less string testing, by means of a preliminary
4261 switch on the second character of optional "-foo" lines. (This is
4262 overdue, caused by the large number of possibilities that now exist.
4263 Originally there were few.) While I was there, I also converted the
4264 str(n)cmp tests so they don't re-test the leading "-" and the first
4265 character, in the hope this might squeeze out yet more improvement.
4267 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4268 flag allowing group syntax was set by the header_syntax check but not
4269 turned off, possible causing trouble later; (2) The flag was not being
4270 set at all for the header_verify test, causing "group"-style headers to
4271 be rejected. I have now set it in this case, and also caused header_
4272 verify to ignore an empty address taken from a group. While doing this, I
4273 came across some other cases where the code for allowing group syntax
4274 while scanning a header line wasn't quite right (mostly, not resetting
4275 the flag correctly in the right place). These bugs could have caused
4276 trouble for malformed header lines. I hope it is now all correct.
4278 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4279 with the "reply" argument non-NULL. The code, however (which originally
4280 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4281 but it didn't always do it. This confused somebody who was copying the
4282 code for some other use. I have removed all the tests.
4284 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4285 feature that was used to support insecure browsers during the U.S. crypto
4286 embargo. It requires special client support, and Exim is probably the
4287 only MTA that supported it -- and would never use it because real RSA is
4288 always available. This code has been removed, because it had the bad
4289 effect of slowing Exim down by computing (never used) parameters for the
4290 RSA_EXPORT functionality.
4292 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4293 authenticator to fail if there's a tab character in the incoming data
4294 (there should never be unless someone is messing about, as it's supposed
4295 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4296 if the connection is using TLS or if the remote IP is the same as the
4297 local IP, and the "valid-client-cert option" if a client certificate has
4300 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4301 authenticators. This can be used for authorization after authentication
4302 succeeds. (In the case of plaintext, it servers for both authentication
4305 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4306 if any retry times were supplied.
4308 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4309 connection, where there is no client IP address to check. In this
4310 situation, the verify now always succeeds.
4312 PH/25 Applied John Jetmore's -Mset patch.
4314 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4316 PH/27 In a string expansion for a processed (not raw) header when multiple
4317 headers of the same name were present, leading whitespace was being
4318 removed from all of them, but trailing whitespace was being removed only
4319 from the last one. Now trailing whitespace is removed from each header
4320 before concatenation. Completely empty headers in a concatenation (as
4321 before) are ignored.
4323 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4324 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4326 PH/29 [Removed. This was a change that I later backed out, and forgot to
4327 correct the ChangeLog entry (that I had efficiently created) before
4328 committing the later change.]
4330 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4331 address errors (4xx response to RCPT) over the same connection as other
4332 messages routed to the same hosts. Such deliveries are always "forced",
4333 so retry times are not inspected. This resulted in far too many retries
4334 for the affected addresses. The effect occurred only when there were more
4335 hosts than the hosts_max_try setting in the smtp transport when it had
4336 the 4xx errors. Those hosts that it had tried were not added to the list
4337 of hosts for which the message was waiting, so if all were tried, there
4338 was no problem. Two fixes have been applied:
4340 (i) If there are any address or message errors in an SMTP delivery, none
4341 of the hosts (tried or untried) are now added to the list of hosts
4342 for which the message is waiting, so the message should not be a
4343 candidate for sending over the same connection that was used for a
4344 successful delivery of some other message. This seems entirely
4345 reasonable: after all the message is NOT "waiting for some host".
4346 This is so "obvious" that I'm not sure why it wasn't done
4347 previously. Hope I haven't missed anything, but it can't do any
4348 harm, as the worst effect is to miss an optimization.
4350 (ii) If, despite (i), such a delivery is accidentally attempted, the
4351 routing retry time is respected, so at least it doesn't keep
4352 hammering the server.
4354 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4355 in ${readsocket because some servers need this prod.
4357 PH/32 Added some extra debug output when updating a wait-xxx database.
4359 PH/33 The hint "could be header name not terminated by colon", which has been
4360 given for certain expansion errors for a long time, was not being given
4361 for the ${if def:h_colon_omitted{... case.
4363 PH/34 The spec says: "With one important exception, whenever a domain list is
4364 being scanned, $domain contains the subject domain." There was at least
4365 one case where this was not true.
4367 PH/35 The error "getsockname() failed: connection reset by peer" was being
4368 written to the panic log as well as the main log, but it isn't really
4369 panic-worthy as it just means the connection died rather early on. I have
4370 removed the panic log writing for the ECONNRESET error when getsockname()
4373 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4374 runs only) independently of the message's sender address. This meant
4375 that, if the 4xx error was in fact related to the sender, a different
4376 message to the same recipient with a different sender could confuse
4377 things. In particular, this can happen when sending to a greylisting
4378 server, but other circumstances could also provoke similar problems.
4379 I have changed the default so that the retry time for these errors is now
4380 based a combination of the sender and recipient addresses. This change
4381 can be overridden by setting address_retry_include_sender=false in the
4384 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4385 remote server are returned as part of bounce messages. This was not
4386 happening for LMTP over a pipe (the lmtp transport), but now it is the
4387 same for both kinds of LMTP.
4389 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4390 in header lines that were in fact CNAMEs. This is no longer the case.
4392 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4393 and queue runs started by the daemon processed all messages. This has
4394 been fixed so that -R and -S can now usefully be given with -q<time>.
4396 PH/40 Import PCRE release 6.7 (fixes some bugs).
4398 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4400 PH/42 Give an error if -q is specified more than once.
4402 PH/43 Renamed the variables $interface_address and $interface_port as
4403 $received_ip_address and $received_port, to make it clear that these
4404 values apply to message reception, and not to the outgoing interface when
4405 a message is delivered. (The old names remain recognized, of course.)
4407 PH/44 There was no timeout on the connect() call when using a Unix domain
4408 socket in the ${readsocket expansion. There now is.
4410 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4411 be meaningful with "accept".
4413 SC/01 Eximstats V1.43
4414 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4416 SC/02 Eximstats V1.44
4417 Use a glob alias rather than an array ref in the generated
4418 parser. This improves both readability and performance.
4420 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4421 Collect SpamAssassin and rejection statistics.
4422 Don't display local sender or destination tables unless
4423 there is data to show.
4424 Added average volumes into the top table text output.
4426 SC/04 Eximstats V1.46
4427 Collect data on the number of addresses (recipients)
4428 as well as the number of messages.
4430 SC/05 Eximstats V1.47
4431 Added 'Message too big' to the list of mail rejection
4432 reasons (thanks to Marco Gaiarin).
4434 SC/06 Eximstats V1.48
4435 Mainlog lines which have GMT offsets and are too short to
4436 have a flag are now skipped.
4438 SC/07 Eximstats V1.49 (Alain Williams)
4439 Added the -emptyok flag.
4441 SC/08 Eximstats V1.50
4442 Fixes for obtaining the IP address from reject messages.
4444 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4445 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4446 whitespace changes from 4.64-PH/27
4448 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4451 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4452 are found, allow negative numbers in numeric criteria)
4454 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4456 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4457 to match changes made in 4.64-PH/43
4459 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4461 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4462 "domain" is actually an IP address, and force a failure. This locks out
4463 those revolvers/nameservers that support "A-for-A" lookups, in
4464 contravention of the specifications.
4466 PH/48 When a host name was looked up from an IP address, and the subsequent
4467 forward lookup of the name timed out, the host name was left in
4468 $sender_host_name, contrary to the specification.
4470 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4471 restricted to single-key lookups, Exim was not diagnosing an error if
4472 * or *@ was used with a query-style lookup.
4474 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4476 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4477 long ago, but noone interested enough thought of it. Let's just say that
4478 the "1.1" means that there are some new functions that weren't there at
4479 some point in the past.
4481 PH/51 Error processing for expansion failure of helo_data from an smtp
4482 transport during callout processing was broken.
4484 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4485 tested/used via the -bh/-bhc/-bs options.
4487 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4488 bug, fixed in subsequent PCRE releases).
4490 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4491 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4493 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4499 SC/01 Use a glob alias rather than an array ref in eximstats generated
4500 parser. This improves both readability and performance.
4502 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4503 Don't display local sender or destination tables in eximstats unless
4504 there is data to show.
4505 Added average volumes into the eximstats top table text output.
4507 SC/03 Collect data on the number of addresses (recipients) as well
4508 as the number of messages in eximstats.
4510 TF/01 Correct an error in the documentation for the redirect router. Exim
4511 does (usually) call initgroups() when daemonizing.
4513 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4514 with consistent privilege compared to when running as a daemon.
4516 TF/03 Note in the spec that $authenticated_id is not set for local
4517 submissions from trusted users.
4519 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4520 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4522 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4523 by adding some example configuration directives to the default
4524 configuration file. A little bit of work is required to uncomment the
4525 directives and define how usernames and passwords are checked, but
4526 there is now a framework to start from.
4528 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4529 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4530 without this. I don't know how relevant this is to other LDAP libraries.
4532 PH/02 Add the verb name to the "unknown ACL verb" error.
4534 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4536 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4538 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4539 directory not expanded when it should be if an expanded home directory
4540 was set for the address (which is overridden by the transport).
4542 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4545 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4546 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4547 because it is too late at that time, and has no effect.
4549 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4550 security issue with \' (bugzilla #107). I could not use the
4551 PQescapeStringConn() function, because it needs a PGconn value as one of
4554 PH/08 When testing addresses using -bt, indicate those final addresses that
4555 are duplicates that would not cause an additional delivery. At least one
4556 person was confused, thinking that -bt output corresponded to deliveries.
4557 (Suppressing duplicates isn't a good idea as you lose the information
4558 about possibly different redirections that led to the duplicates.)
4560 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4561 systems where poll() doesn't work, in particular OS X.
4563 PH/10 Added more information to debugging output for retry time not reached.
4565 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4566 operations in malware.c.
4568 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4571 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4572 syslog_duplication set false, log lines that would normally be written
4573 both the the main log and to the reject log were not written to syslog at
4576 PH/14 In the default configuration, change the use of "message" in ACL warn
4577 statements to "add_header".
4579 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4580 not followed by a command (e.g. "seen endif").
4582 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4583 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4586 PH/17 Added extra conditions to the default value of delay_warning_condition
4590 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4591 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4592 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4595 The Auto-Submitted: and various List- headers are standardised, whereas I
4596 don't think Precedence: ever was.
4598 PH/18 Refactored debugging code in route_finduser() to show more information,
4599 in particular, the error code if getpwnam() issues one.
4601 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4602 This is apparently needed in addition to the PH/07 change above to avoid
4603 any possible encoding problems.
4605 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4606 but not after initializing Perl.
4608 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4609 output them only if debugging. By default they are written stderr,
4610 apparently, which is not desirable.
4612 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4615 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4618 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4620 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4621 authenticated or an ident call has been made. Suppress the default
4622 values for $authenticated_id and $authenticated_sender (but permit -oMai
4623 and -oMas) when testing with -bh.
4625 PH/24 Re-jigged the order of the tests in the default configuration so that the
4626 tests for valid domains and recipients precede the DNS black list and CSA
4627 tests, on the grounds that those ones are more expensive.
4629 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4630 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4631 This bug exists in every version of Exim that I still have, right back to
4634 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4635 However, an attempt to turn on case-sensitivity in a regex key by
4636 including (?-i) didn't work because the subject string was already
4637 lowercased, and the effects were non-intuitive. It turns out that a
4638 one-line patch can be used to allow (?-i) to work as expected.
4644 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4645 other effects) broke the use of negated acl sub-conditions.
4647 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4650 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4651 "Deny" causes Exim to reject the incoming connection with a 554 error.
4652 Unfortunately, if there is a major crisis, such as a disk failure,
4653 tcp-wrappers gives "deny", whereas what one would like would be some
4654 kind of temporary error. A kludge has been added to help with this.
4655 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4656 554 error is used if errno is still zero or contains ENOENT (which occurs
4657 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4660 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4662 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4663 errors. Otherwise a message that provokes a temporary error (when other
4664 messages do not) can cause a whole host to time out.
4666 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4667 addresses were routed directly to files or pipes from a redirect router.
4668 File deliveries just didn't batch; pipe deliveries might have suffered
4671 PH/06 A failure to get a lock for a hints database would erroneously always say
4672 "Failed to get write lock", even when it was really a read lock.
4674 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4675 of 0600. This has been changed to use the value of the lockfile_mode
4676 option (which defaults to 0600).
4678 PH/08 Applied small patch from the Sieve maintainer.
4680 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4681 folder from quota calculations, a direct delivery into this folder messed
4682 up the contents of the maildirsize file. This was because the regex was
4683 used only to exclude .Trash (or whatever) when the size of the mailbox
4684 was calculated. There was no check that a delivery was happening into an
4685 excluded directory. This bug has been fixed by ignoring all quota
4686 processing for deliveries into excluded directories.
4688 PH/10 Added the maildirfolder_create_regex option to appendfile.
4694 PH/01 The code for finding all the local interface addresses on a FreeBSD
4695 system running IPv6 was broken. This may well have applied to all BSD
4696 systems, as well as to others that have similar system calls. The broken
4697 code found IPv4 interfaces correctly, but gave incorrect values for the
4698 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4699 that it would not match correctly against @[] and not recognize the IPv6
4702 PH/02 The ipliteral router was not recognizing addresses of the form user@
4703 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4705 PH/03 Added disable_ipv6.
4707 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4708 decoded form, because it is most often used to construct To: headers
4709 lines in autoreplies, and the decoded form may well be syntactically
4710 invalid. However, $reply_address has leading white space removed, and all
4711 newlines turned into spaces so that the autoreply transport does not
4714 PH/05 If group was specified without a user on a router, and no group or user
4715 was specified on a transport, the group from the router was ignored.
4717 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4718 for visible compile-time settings that can be used to change these
4719 numbers, for those that want even more. Backwards compatibility with old
4720 spool files has been maintained. However, going back to a previous Exim
4721 release will lost any variables that are in spool files.
4723 PH/07 Two small changes when running in the test harness: increase delay when
4724 passing a TCP/IP connection to a new process, in case the original
4725 process has to generate a bounce, and remove special handling of
4726 127.0.0.2 (sic), which is no longer necessary.
4728 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4729 be the same on different OS.
4731 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4734 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4735 whether --show-vars was specified or not
4737 JJ/02 exipick: Added support for new ACL variable spool format introduced
4740 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4741 syntactically invalid From: or Reply-to: line, and a filter used this to
4742 generate an autoreply, and therefore failed to obtain an address for the
4743 autoreply, Exim could try to deliver to a non-existent relative file
4744 name, causing unrelated and misleading errors. What now happens is that
4745 it logs this as a hard delivery error, but does not attempt to create a
4748 PH/11 The exinext utility has a -C option for testing purposes, but although
4749 the given file was scanned by exinext itself; it wasn't being passed on
4750 when Exim was called.
4752 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4753 an end-of-file indication when reading a command response.
4755 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4756 compiled. In many other places in Exim, IPv6 addresses are always
4757 recognized, so I have changed this. It also means that IPv4 domain
4758 literals of the form [IPV4:n.n.n.n] are now always recognized.
4760 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4761 used if the router is not running as root, for example, when verifying at
4762 ACL time, or when using -bh. The debugging output from this situation was
4763 non-existent - all you got was a failure to exec. I have made two
4766 (a) Failures to set uid/gid, the current directory, or a process leader
4767 in a subprocess such as that created by queryprogram now generate
4768 suitable debugging output when -d is set.
4770 (b) The queryprogram router detects when it is not running as root,
4771 outputs suitable debugging information if -d is set, and then runs
4772 the subprocess without attempting to change uid/gid.
4774 PH/15 Minor change to Makefile for building test_host (undocumented testing
4777 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4778 additional section of a DNS packet that returns MX or SRV records.
4779 Instead, it always explicitly searches for A/AAAA records. This avoids
4780 major problems that occur when a DNS server includes only records of one
4781 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4782 fixed another bug: if SRV records were looked up and the corresponding
4783 address records were *not* found in the additional section, the port
4784 values from the SRV records were lost.
4786 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4787 using the correct key (the original address) when searching the retry
4788 rules in order to find which one to use for generating the retry hint.
4790 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4791 adding the default one. Similarly, if it contains a Reply-To: header, the
4792 errors_reply_to option, if set, is not used.
4794 PH/19 When calculating a retry time, Exim used to measure the "time since
4795 failure" by looking at the "first failed" field in the retry record. Now
4796 it does not use this if it is later than than the arrival time of the
4797 message. Instead it uses the arrival time. This makes for better
4798 behaviour in cases where some deliveries succeed, thus re-setting the
4799 "first failed" field. An example is a quota failure for a huge message
4800 when small messages continue to be delivered. Without this change, the
4801 "time since failure" will always be short, possible causing more frequent
4802 delivery attempts for the huge message than are intended.
4803 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4805 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4806 $1, $2, $3) because the numerical variables can be reset during some
4807 expansion items (e.g. "match"), thereby losing the authentication data.
4809 PH/21 Make -bV show the size of off_t variables so that the test suite can
4810 decide whether to run tests for quotas > 2G.
4812 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4813 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4814 filecount value is greater than 2G or if a quota value is greater than 2G
4815 on a system where the size of off_t is not greater than 4, a panic error
4818 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4819 never match. The debug and -bh output now contains an explicit error
4820 message indicating a malformed IPv4 address or mask.
4822 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4823 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4824 PH/23 above applies.
4826 PH/25 Do not write to syslog when running in the test harness. The only
4827 occasion when this arises is a failure to open the main or panic logs
4828 (for which there is an explicit test).
4830 PH/26 Added the /no_tell option to "control=freeze".
4832 PH/27 If a host name lookup failed very early in a connection, for example, if
4833 the IP address matched host_lookup and the reverse lookup yielded a name
4834 that did not have a forward lookup, an error message of the form "no IP
4835 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4836 could be logged. Now it outputs the IP address instead of "NULL".
4838 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4839 allows the sender and the authenticated sender to be set when
4840 submitting a message from within Exim. Since child_open_exim() is
4841 documented for local_scan(), the new function should be too.
4843 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4844 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4845 results in an empty string is now treated as unset.
4847 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4849 PH/31 Added sender_verify_fail logging option.
4851 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4852 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4853 tidied the source and removed it altogether.
4855 PH/33 When a queue run was abandoned because the load average was too high, a
4856 log line was always written; now it is written only if the queue_run log
4857 selector is set. In addition, the log line for abandonment now contains
4858 information about the queue run such as the pid. This is always present
4859 in "start" and "stop" lines but was omitted from the "abandon" line.
4861 PH/34 Omit spaces between a header name and the colon in the error message that
4862 is given when verify = headers_syntax fails (if there are lots of them,
4863 the message gets confusing).
4865 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4866 names, as there are now some PTR records that contain slashes. This check
4867 is only to protect against broken name servers that fall over on strange
4868 characters, so the fact that it applies to all lookups doesn't matter.
4870 PH/36 Now that the new test suite is complete, we can remove some of the
4871 special code in Exim that was needed for the old test suite. For example,
4872 sorting DNS records because real resolvers return them in an arbitrary
4873 order. The new test suite's fake resolver always returns records in the
4876 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4877 bounces) except when queue_only is set, to avoid logging races between
4878 the different processes.
4880 PH/38 Panic-die if .include specifies a non-absolute path.
4882 PH/39 A tweak to the "H" retry rule from its user.
4884 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4885 a label. They prevented compilation on older perls.
4887 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4888 a warning to be raised on newish perls.
4890 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4891 on queue. Changes to match documented behaviour of showing count of
4892 messages matching specified criteria.
4894 PH/40 Changed the default ident timeout from 30s to 5s.
4896 PH/41 Added support for the use of login_cap features, on those BSD systems
4897 that have them, for controlling the resources used by pipe deliveries.
4899 PH/42 The content-scanning code uses fopen() to create files in which to put
4900 message data. Previously it was not paying any attention to the mode of
4901 the files. Exim runs with umask(0) because the rest of the code creates
4902 files with open(), and sets the required mode explicitly. Thus, these
4903 files were ending up world-writeable. This was not a big issue, because,
4904 being within the spool directory, they were not world-accessible. I have
4905 created a function called modefopen, which takes an additional mode
4906 argument. It sets umask(777), creates the file, chmods it to the required
4907 mode, then resets the umask. All the relevant calls to fopen() in the
4908 content scanning code have been changed to use this function.
4910 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4911 to 24 hours. This avoids potential overflow problems when processing G
4912 and H retry rules. I suspect nobody ever tinkers with this value.
4914 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4916 PH/45 When the plaintext authenticator is running as a client, the server's
4917 challenges are checked to ensure they are valid base64 strings. By
4918 default, the authentication attempt is cancelled if an invalid string is
4919 received. Setting client_ignore_invalid_base64 true ignores these errors.
4920 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4921 they are received. Thus, the responses can be made to depend on the
4922 challenges. If an invalid string is ignored, an empty string is placed in
4925 PH/46 Messages that are created by the autoreply transport now contains a
4926 References: header, in accordance with RFCs 2822 and 3834.
4928 PH/47 Added authenticated_sender_force to the smtp transport.
4930 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4932 PH/49 Installed latest patch from the Sieve maintainer.
4934 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4935 was also set, the appendfile transport was unnecessarily scanning a
4936 directory of message files (e.g. for maildir delivery) to find the count
4937 of files (along with the size), even though it did not need this
4938 information. It now does the scan only if it needs to find either the
4939 size of the count of files.
4941 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4943 PH/52 Two bugs concerned with error handling when the smtp transport is
4946 (i) Exim was not creating retry information for temporary errors given
4947 for individual recipients after the DATA command when the smtp transport
4948 was used in LMTP mode. This meant that they could be retried too
4949 frequently, and not timed out correctly.
4951 (ii) Exim was setting the flag that allows error details to be returned
4952 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4953 recipients that were returned after the DATA command.
4955 PH/53 This is related to PH/52, but is more general: for any failing address,
4956 when detailed error information was permitted to be returned to the
4957 sender, but the error was temporary, then after the final timeout, only
4958 "retry timeout exceeded" was returned. Now it returns the full error as
4959 well as "retry timeout exceeded".
4961 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4962 do this, and (what is worse) MTAs that accept it.
4964 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4965 will now be deprecated.
4967 PH/56 New os.c-cygwin from the Cygwin maintainer.
4969 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4970 formats (previously only available in exim formats via -bpr, -bpru,
4971 and -bpra. Now also available in native and exiqgrep formats)
4973 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4974 with very large, slow to parse queues
4976 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4978 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4980 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4981 responses to authentication challenges, though it was showing the
4982 challenges; (ii) I've removed the CR characters from the debug output for
4985 PH/58 Allow for the insertion of a newline as well as a space when a string
4986 is turned into more than one encoded-word during RFC 2047 encoding. The
4987 Sieve code now uses this.
4989 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4990 data_4xx, lost_connection, tls_required.
4992 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4993 message was being sent as an SMTP response.
4995 PH/61 Add -l and -k options to exicyclog.
4997 PH/62 When verifying, if an address was redirected to one new address, so that
4998 verification continued, and the new address failed or deferred after
4999 having set something in $address_data, the value of $address_data was not
5000 passed back to the ACL. This was different to the case when no
5001 redirection occurred. The value is now passed back in both cases.
5003 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5004 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5005 use login_cap.h, so on its own it isn't the distinguishing feature. The
5006 new name refers directly to the setclassresources() function.
5008 PH/65 Added configuration files for NetBSD3.
5010 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5012 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5013 is preferred over IPv4.
5015 PH/68 The bounce_return_message and bounce_return_body options were not being
5016 honoured for bounces generated during the reception of non-SMTP messages.
5017 In particular, this applied to messages rejected by the ACL. This bug has
5018 been fixed. However, if bounce_return_message is true and bounce_return_
5019 body is false, the headers that are returned for a non-SMTP message
5020 include only those that have been read before the error was detected.
5021 (In the case of an ACL rejection, they have all been read.)
5023 PH/69 The HTML version of the specification is now built in a directory called
5024 spec_html instead of spec.html, because the latter looks like a path with
5025 a MIME-type, and this confuses some software.
5027 PH/70 Catch two compiler warnings in sieve.c.
5029 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5030 function verify_get_ident() calls ip_connect() to connect a socket, but
5031 if the "connect()" function timed out, ip_connect() used to close the
5032 socket. However, verify_get_ident() also closes the socket later, and in
5033 between Exim writes to the log, which may get opened at this point. When
5034 the socket was closed in ip_connect(), the log could get the same file
5035 descriptor number as the socket. This naturally causes chaos. The fix is
5036 not to close the socket in ip_connect(); the socket should be closed by
5037 the function that creates it. There was only one place in the code where
5038 this was missing, in the iplookup router, which I don't think anybody now
5039 uses, but I've fixed it anyway.
5041 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5042 well as to direct DNS lookups. Otherwise the handling of names in host
5043 lists is inconsistent and therefore confusing.
5049 PH/01 Two changes to the default runtime configuration:
5051 (1) Move the checks for relay_from_hosts and authenticated clients from
5052 after to before the (commented out) DNS black list checks.
5054 (2) Add control=submission to the relay_from_hosts and authenticated
5055 clients checks, on the grounds that messages accepted by these
5056 statements are most likely to be submissions.
5058 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5060 (1) Generate an error if the third argument for the ${prvs expansion is
5063 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5066 (3) Reset the variables that are obtained from the first argument of
5067 ${prvscheck and used in the second argument before leaving the code,
5068 because their memory is reclaimed, so using them afterwards may do
5071 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5072 one (it's much easier than Tom thought :-).
5074 (5) Because of (4), we can now allow for the use of $prvscheck_result
5075 inside the third argument.
5077 PH/03 For some reason, the default setting of PATH when running a command from
5078 a pipe transport was just "/usr/bin". I have changed it to
5081 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5082 anything to be listed in the output from -bV.
5084 PH/05 When a filter generated an autoreply, the entire To: header line was
5085 quoted in the delivery log line, like this:
5087 => >A.N.Other <ano@some.domain> <original@ddress> ...
5089 This has been changed so that it extracts the operative address. There
5090 may be more than one such address. If so, they are comma-separated, like
5093 => >ano@some.domain,ona@other.domain <original@ddress> ...
5095 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5096 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5097 not being looked up in the rDNS to get a host name, Exim was showing the
5098 IP address twice in Received: lines, even though the IP addresses were
5099 identical. For example:
5101 Received: from [1.2.3.4] (helo=[1.2.3.4])
5103 However, if the real host name was known, it was omitting the HELO data
5104 if it matched the actual IP address. This has been tidied up so that it
5105 doesn't show the same IP address twice.
5107 PH/07 When both +timestamp and +memory debugging was on, the value given by
5108 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5109 called by the debug printing, thereby overwriting the timestamp buffer.
5110 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5112 PH/08 When the original message was included in an autoreply transport, it
5113 always said "this is a copy of the message, including all the headers",
5114 even if body_only or headers_only was set. It now gives an appropriate
5117 PH/09 Applied a patch from the Sieve maintainer which:
5119 o fixes some comments
5120 o adds the (disabled) notify extension core
5121 o adds some debug output for the result of if/elsif tests
5122 o points to the current vacation draft in the documentation
5123 and documents the missing references header update
5127 o fixes a bug in processing the envelope test (when testing
5128 multiple envelope elements, the last element determined the
5131 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5132 Electronic Mail") by including:
5134 Auto-submitted: auto-generated
5136 in the messages that it generates (bounce messages and others, such as
5137 warnings). In the case of bounce messages for non-SMTP messages, there was
5138 also a typo: it was using "Auto_submitted" (underscore instead of
5139 hyphen). Since every message generated by Exim is necessarily in response
5140 to another message, thes have all been changed to:
5142 Auto-Submitted: auto-replied
5144 in accordance with these statements in the RFC:
5146 The auto-replied keyword:
5148 - SHOULD be used on messages sent in direct response to another
5149 message by an automatic process,
5151 - MUST NOT be used on manually-generated messages,
5153 - MAY be used on Delivery Status Notifications (DSNs) and Message
5154 Disposition Notifications (MDNs),
5156 - MUST NOT be used on messages generated by automatic or periodic
5157 processes, except for messages which are automatic responses to
5160 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5161 to the default Received: header definition.
5163 PH/12 Added log selector acl_warn_skipped (default on).
5165 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5166 variables because (a) they are in the wrong storage pool and (b) even if
5167 they were copied, it wouldn't work properly because of the caching.
5169 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5170 checking when decoding. Apparently there are clients that generate
5171 overlong encoded strings. Why am I not surprised?
5173 PH/15 If the first argument of "${if match_address" was not empty, but did not
5174 contain an "@" character, Exim crashed. Now it writes a panic log message
5175 and treats the condition as false.
5177 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5179 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5180 extension "envelope-auth". The code is finished and in agreement with
5181 other implementations, but there is no documentation so far and in fact,
5182 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5183 not changing the active code.
5185 Print executed "if" and "elsif" statements when debugging is used. This
5186 helps a great deal to understand what a filter does.
5188 Document more things not specified clearly in RFC3028. I had all this
5189 sorted out, when out of a sudden new issues came to my mind. Oops."
5191 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5194 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5195 canonical form (as documented). However, after a host name lookup from
5196 the IP address, check_host() was doing a simple string comparison with
5197 addresses acquired from the DNS when checking that the found name did
5198 have the original IP as one of its addresses. Since any found IPv6
5199 addresses are likely to be in abbreviated form, the comparison could
5200 fail. Luckily, there already exists a function for doing the comparison
5201 by converting both addresses to binary, so now that is used instead of
5202 the text comparison.
5204 PH/20 There was another similar case to PH/19, when a complete host name was
5205 given in a host list; looking up its IP address could give an abbreviated
5206 form, whereas the current host's name might or might not be abbreviated.
5207 The same fix has been applied.
5213 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5214 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5217 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5218 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5220 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5222 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5223 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5224 Also, for "Auto-Submitted", treat anything other than "no" as
5225 non-personal, in accordance with RFC 3834. (Previously it treated
5226 anything starting "auto-" as non-personal.)
5228 TF/01 The control=submission/name=... option had a problem with syntax
5229 errors if the name included a slash character. The /name= option
5230 now slurps the rest of the string, so it can include any characters
5231 but it must come last in the list of options (after /sender_retain
5234 PH/05 Some modifications to the interface to the fake nameserver for the new
5242 TK/01 Added the "success_on_redirect" address verification option. See
5243 NewStuff for rationale and an example.
5245 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5247 PH/02 Patch to exigrep to allow it to work on syslog lines.
5249 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5250 fread() to skip over the body file's header line, because in Cygwin the
5251 header line is locked and is inaccessible.
5253 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5254 co-exist for some time) to make it clear that it is the Exim ID that is
5255 referenced, not the Message-ID: header line.
5257 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5258 string_format() function, because snprintf() does not exist on all
5261 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5262 ${stat: expansion item.
5264 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5265 protocol synchronization error", to keep the pedants happy.
5267 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5268 well as for IRIX systems, when gcc is being used. See the host.c source
5271 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5273 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5276 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5277 transport and to the smtp transport in LMTP mode.
5279 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5281 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5282 run a filter in a subprocess. This could lead to confusion in subsequent
5283 lookups in the parent process. There should also be a search_tidyup() at
5284 the end of the subprocess.
5286 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5287 only if the host matched helo_try_verify_hosts, which caused the
5288 verification to occur when the EHLO/HELO command was issued. The ACL just
5289 tested the remembered result. Now, if a previous verification attempt has
5290 not happened, "verify = helo" does it there and then.
5292 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5294 TK/03 Fix log output including CR from clamd.
5296 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5297 exist provoked a memory error which could cause a segfault.
5299 PH/15 Installed PCRE 6.2
5301 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5303 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5304 of the problem. Specifically, suggested +O2 rather than +O1 for the
5307 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5309 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5310 record, the port was not correctly logged when the outgoing_port log
5311 selector was set (it logged the transort's default port).
5313 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5314 fallback_hosts, and "hosts" in the smtp transport.
5316 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5317 host errors such as "Connection refused".
5319 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5320 authentication with radiusclient 0.4.9:
5322 - Error returned from rc_read_config was caught wrongly
5323 - Username/password not passed on to radius server due to wrong length.
5325 The presumption is that some radiusclient API changes for 4.51/PH/17
5326 were not taken care of correctly. The code is still untested by me (my
5327 Linux distribution still has 0.3.2 of radiusclient), but it was
5328 contributed by a Radius user.
5330 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5331 expanding the "port" option of the smtp transport.
5333 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5334 while reading a MIME header. Thanks to Tom Hughes for a patch.
5336 PH/24 Include config.h inside local_scan.h so that configuration settings are
5339 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5340 that in an ACL for RCPT (for example), you can examine exactly what was
5343 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5344 commands, but it was not correctly comparing the address with the actual
5345 client host address. Thus, it would show the EHLO address in Received:
5346 header lines when this was not necessary.
5348 PH/27 Added the % operator to ${eval:}.
5350 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5351 it should be ignoring failures (because with -C, for example, it has lost
5352 privilege). It wasn't ignoring creation failures other than "already
5355 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5358 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5359 because an input line was too long, either on its own, or by virtue of
5360 too many continuations, the temporary file was not being removed, and the
5361 return code was incorrect.
5363 PH/31 Missing "BOOL" in function definition in filtertest.c.
5365 PH/32 Applied Sieve patches from the maintainer.
5367 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5369 PH/33 Added "verify = not_blind".
5371 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5372 Local/Makefile (with some defaults set). These are used in built scripts
5373 such as exicyclog, but they have never been used in the exim_install
5374 script (though there are many overriding facilities there). I have
5375 arranged that the exim_install script now takes note of these two
5378 PH/35 Installed configuration files for Dragonfly.
5380 PH/36 When a locally submitted message by a trusted user did not contain a
5381 From: header, and the sender address was obtained from -f or from an SMTP
5382 MAIL command, and the trusted user did not use -F to supply a sender
5383 name, $originator_name was incorrectly used when constructing a From:
5384 header. Furthermore, $originator_name was used for submission mode
5385 messages from external hosts without From: headers in a similar way,
5386 which is clearly wrong.
5388 PH/37 Added control=suppress_local_fixups.
5390 PH/38 When log_selector = +received_sender was set, and the addition of the
5391 sender made the log line's construction buffer exactly full, or one byte
5392 less than full, an overflow happened when the terminating "\n" was
5395 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5396 when the result of a list match is failure because a DNS lookup failed.
5398 PH/40 RM_COMMAND is now used in the building process.
5400 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5401 the "build-* directories that it finds.
5403 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5404 address in a domain literal was a prefix of an interface address.
5406 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5407 when verifying a sender address, unless rewrite_headers is false.
5409 PH/44 Wrote a long comment about why errors_to addresses are verified as
5410 recipients, not senders.
5412 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5413 the ratelimit ACL was added.
5415 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5417 PH/46 Added extra information about PostgreSQL errors to the error string.
5419 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5420 avoiding the need to install special zones in a real server. This is
5421 backwards compatible; if it can't find the fake resolver, it drops back.
5422 Thus, both old and new test suites can be run.
5424 TF/02 Added util/ratelimit.pl
5426 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5427 clock is set back in time.
5429 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5430 Candler <B.Candler@pobox.com>.
5432 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5433 OK for addresses that are the result of redirections.
5435 PH/48 A number of further additions for the benefit of the new test suite,
5436 including a fake gethostbyname() that interfaces to the fake DNS resolver
5439 TF/06 The fix for widen_domains has also been applied to qualify_single and
5440 search_parents which are the other dnslookup options that can cause
5443 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5446 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5448 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5449 service" if it encountered an MX record with an empty target hostname.
5450 The message is now "an MX or SRV record indicated no SMTP service".
5452 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5453 if the DNS of the sending site is misconfigured. This is quite a
5454 common situation. This change restores the behaviour of treating a
5455 helo verification defer as a failure.
5457 PH/51 If self=fail was set on a router, the bounce message did not include the
5458 actual error message.
5464 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5466 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5467 command itself ended in error, the underlying message about the transport
5468 filter timeout was being overwritten with the pipe command error. Now the
5469 underlying error message should be appended to the second error message.
5471 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5473 PH/02 Reduce the amount of output that "make" produces by default. Full output
5474 can still be requested.
5476 PH/03 The warning log line about a condition test deferring for a "warn" verb
5477 was being output only once per connection, rather than after each
5478 occurrence (because it was using the same function as for successful
5479 "warn" verbs). This seems wrong, so I have changed it.
5481 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5482 it should not have, which might have caused a crash in the right
5483 circumstances, but probably never did.
5485 PH/04 Installed a modified version of Tony Finch's patch to make submission
5486 mode fix the return path as well as the Sender: header line, and to
5487 add a /name= option so that you can make the user's friendly name appear
5490 TF/03 Added the control = fakedefer ACL modifier.
5492 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5493 Mark Lowes for thorough testing.
5495 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5497 TK/03 Merged latest SRS patch from Miles Wilton.
5499 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5500 with the definition in sysexits.h (which is #included earlier).
5501 Fortunately, Exim does not actually use EX_OK. The code used to try to
5502 preserve the sysexits.h value, by assuming that macro definitions were
5503 scanned for macro replacements. I have been disabused of this notion,
5504 so now the code just undefines EX_OK before #including unistd.h.
5506 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5507 in the smtp transport. When a block could not be written in a single
5508 write() function, the timeout was being re-applied to each part-write.
5509 This seems wrong - if the receiver was accepting one byte at a time it
5510 would take for ever. The timeout is now adjusted when this happens. It
5511 doesn't have to be particularly precise.
5513 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5514 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5516 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5517 without a domain if the check to <postmaster@domain> fails.
5519 SC/01 Eximstats: added -xls and the ability to specify output files
5520 (patch written by Frank Heydlauf).
5522 SC/02 Eximstats: use FileHandles for outputting results.
5524 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5526 SC/04 Eximstats: fixed display of large numbers with -nvr option
5528 SC/05 Eximstats: fixed merging of reports with empty tables.
5530 SC/06 Eximstats: added the -include_original_destination flag
5532 SC/07 Eximstats: removed tabs and trailing whitespace.
5534 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5536 TK/06 MBOX spool code: Add real "From " MBOX separator line
5537 so the .eml file is really in mbox format (even though
5538 most programs do not really care). Patch from Alex Miller.
5540 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5541 The latter is generated from $received_to and is only set if the
5542 message has one envelope recipient. SA can use these headers,
5543 obviously out-of-the-box. Patch from Alex Miller.
5545 PH/08 The ${def test on a variable was returning false if the variable's
5546 value was "0", contrary to what the specification has always said!
5547 The result should be true unless the variable is empty.
5549 PH/09 The syntax error of a character other than { following "${if
5550 def:variable_name" (after optional whitespace) was not being diagnosed.
5551 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5552 accidental colon was present, for example, could give incorrect results.
5554 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5555 result is used (not including appendfile, where other changes are about
5558 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5559 This involved changing a lot of size variables from int to off_t. It
5560 should work with maildirs and everything.
5562 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5563 spamd dying while we are connected to it.
5565 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5568 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5569 function for BDB 4.3.
5571 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5573 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5574 the sources for PCRE have been re-arranged and more files are now
5577 PH/15 The code I had for printing potentially long long variables in PH/11
5578 above was not the best (it lost precision). The length of off_t variables
5579 is now inspected at build time, and an appropriate printing format (%ld
5580 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5581 to be "long long int" or "long int". This is needed for the internal
5582 formatting function string_vformat().
5584 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5585 the configuration file to be ":syslog", then the script "guesses" where
5586 the logs files are, rather than using the compiled in default. In our
5587 case the guess is not the same as the compiled default, so the script
5588 suddenly stopped working when I started to use syslog. The patch checks
5589 to see if log_file_path is "". If so, it attempts to read it from exim
5590 with no configuration file to get the compiled in version, before it
5591 falls back to the previous guessing code."
5593 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5594 implementing BATV in an Exim configuration. See NewStuff for the gory
5597 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5598 Makefile that are specific to HP-UX.
5600 PH/18 If the "use_postmaster" option was set for a recipient callout together
5601 with the "random" option, the postmaster address was used as the MAIL
5602 FROM address for the random test, but not for the subsequent recipient
5603 test. It is now used for both.
5605 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5606 patch removes a few documentation additions to RFC 3028, because the
5607 latest draft now contains them. It adds the new en;ascii-case comparator
5608 and a new error check for 8bit text in MIME parts. Comparator and
5609 require names are now matched exactly. I enabled the subaddress
5610 extension, but it is not well tested yet (read: it works for me)."
5612 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5613 rework some of the code of TK/09 above to avoid the hardwired use of
5614 "%lld" and "long long". Replaced the call to snprintf() with a call to
5617 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5618 records point to non-existent hosts", "retry timeout exceeded", and
5619 "retry time not reached for any host after a long failure period".
5621 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5622 experimental DomainKeys support:
5624 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5625 (2) On an error such as an illegally used "control", the wrong name for
5626 the control was given.
5628 These problems did NOT occur unless DomainKeys support was compiled.
5630 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5632 PH/24 Added ${if match_ip condition.
5634 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5635 databases so that it will be absolutely obvious if a crash occurs in the
5636 DB library. This is a regular occurrence (often caused by mis-matched
5639 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5640 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5641 picked up on a user's system that detects such things. There doesn't seem
5642 to be a gcc warning option for this - only an attribute that has to be
5643 put on the function's prototype. It seems that in Fedora Core 4 they have
5644 set this on a number of new functions. No doubt there will be more in due
5647 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5648 specify a transport. However, if an address that was verified by such a
5649 router was the subject of a callout, Exim crashed because it tried to
5650 read the rcpt_include_affixes from the non-existent transport. Now it
5651 just assumes that the setting of that option is false. This bug was
5652 introduced by 4.51/PH/31.
5654 PH/28 Changed -d+all to exclude +memory, because that information is very
5655 rarely of interest, but it makes the output a lot bigger. People tend to
5656 do -d+all out of habit.
5658 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5659 code in os-type was giving problems when libc.so lives in lib64, like on
5662 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5663 aren't the modern standard, and it seems that some systems' include files
5664 don't always have them. Exim was already checking for some of the newer
5665 ones like T_AAAA, and defining it itself. I've added checks for all the
5666 record types that Exim uses.
5668 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5669 not automatically generating a new one, as it is supposed to. This
5670 prevented TLS from working. If the file did exist, but contained invalid
5671 data, a new version was generated, as expected. It was only the case of a
5672 non-existent file that was broken.
5674 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5675 with a change in libdomainkeys > 0.64.
5677 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5678 from DNS. If the selector record carries the flag, it now has
5679 precedence over the domain-wide flag.
5681 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5683 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5684 the use of a port name that isn't defined in /etc/services) occurred, the
5685 message was deferred as in a normal delivery, and thus remained on the
5686 spool, instead of being failed because of the mua_wrapper setting. This
5687 is now fixed, and I tidied up some of the mua_wrapper messages at the
5690 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5691 the messages in a hash of arrays rather than using individual hashes.
5692 This is a bit cleaner and results in dramatic memory savings, albeit
5693 at a slight CPU cost.
5695 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5696 as requested by Marc Sherman.
5698 SC/10 Eximstats: added histograms for user specified patterns as requested
5701 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5703 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5704 fopen() in the content-scanning modules that did not already have it.
5710 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5711 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5713 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5715 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5717 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5718 to test Sieve filters that use "vacation".
5720 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5721 that changes the way the GnuTLS parameters are stored in the cache file.
5722 The new format can be generated externally. For backward compatibility,
5723 if the data in the cache doesn't make sense, Exim assumes it has read an
5724 old-format file, and it generates new data and writes a new file. This
5725 means that you can't go back to an older release without removing the
5728 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5729 work if there are any delivery delays because "one_time" forces the
5730 parent to be marked "delivered", so its unseen clone is never tried
5731 again. For this reason, Exim now forbids the simultaneous setting of
5734 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5735 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5736 entry if you want to know the details. The fix, however, neglected to
5737 consider the case when local delivery batching is involved. The test for
5738 "previously delivered" was not happening when checking to see if an
5739 address could be batched with a previous (undelivered) one; under
5740 certain circumstances this could lead to multiple deliveries to the same
5743 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5744 in its include files, and this causes problems building Exim.
5746 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5747 header_syntax) but Exim was just ignoring anything given after a slash.
5748 In particular, this caused confusion with an attempt to use "verify =
5749 reverse_host_lookup/defer_ok". An error is now given when options are
5750 supplied for verify items that do not have them. (Maybe reverse_host_
5751 lookup should have a defer_ok option, but that's a different point.)
5753 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5754 defined by RFC 821) to 2048, because there were problems with some AUTH
5755 commands, and RFC 1869 says the size should be increased for extended
5756 SMTP commands that take arguments.
5758 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5761 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5762 "unknown" error; now it says that the functionality isn't in the binary.
5764 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5765 an address' error message when a string expansion fails (syntax or
5766 whatever). Otherwise the password may appear in the log. Following change
5767 PH/42 below, there is no longer a chance of it appearing in a bounce
5770 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5772 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5773 of its addresses was ever tried. (Bugzilla bug #2.)
5775 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5776 the result incorrectly in the debug output. (It correctly added a newline
5777 to what was transported.)
5779 TF/01 Added $received_time.
5781 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5782 commented out examples of how to interface to a virus scanner and to
5783 SpamAssassin. Also added commented examples of av_scanner and
5784 spamd_address settings.
5786 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5787 and controls are allowed in which ACLs. There were a couple of minor
5788 errors. Some of the entries in the conditions table (which is a table of
5789 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5790 negation of where the condition IS allowed.
5792 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5794 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5795 header file does not have a version number, so I've had to invent a new
5796 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5797 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5798 radiusclient), but it was contributed by a Radius user.
5800 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5801 files or directories, for OpenSSL.
5803 PH/19 When an Exim process that is running as root has to create an Exim log
5804 file, it does so in a subprocess that runs as exim:exim so as to get the
5805 ownership right at creation (otherwise, other Exim processes might see
5806 the file with the wrong ownership). There was no test for failure of this
5807 fork() call, which would lead to the process getting stuck as it waited
5808 for a non-existent subprocess. Forks do occasionally fail when resources
5809 run out. I reviewed all the other calls to fork(); they all seem to check
5812 PH/20 When checking for unexpected SMTP input at connect time (before writing
5813 the banner), Exim was not dealing correctly with a non-positive return
5814 from the read() function. If the client had disconnected by this time,
5815 the result was a log entry for a synchronization error with an empty
5816 string after "input=" when read() returned zero. If read() returned -1
5817 (an event I could not check), uninitialized data bytes were printed.
5818 There were reports of junk text (parts of files, etc) appearing after
5821 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5823 PH/22 Added support for macro redefinition, and (re)definition in between
5824 driver and ACL definitions.
5826 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5827 forgetting to use the resulting value; it was using the unexpanded value.
5829 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5830 hadn't been configured. The fix is from Juergen Kreileder, who
5831 understands it better than I do:
5833 "Here's what I see happening with three configured cyrus_sasl
5834 authenticators configured (plain, login, cram-md5):
5836 On startup auth_cyrus_sasl_init() gets called for each of these.
5837 This means three calls to sasl_listmech() without a specified mech_list.
5838 => SASL tests which mechs of all available mechs actually work
5839 => three warnings about OTP not working
5840 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5842 With the patch, sasl_listmech() also gets called three times. But now
5843 SASL's mech_list option is set to the server_mech specified in the the
5844 authenticator. Or in other words, the answer from sasl_listmech()
5845 gets limited to just the mech you're testing for (which is different
5847 => the return list contains just 'plain' or 'login', 'cram-md5' or
5848 nothing depending on the value of ob->server_mech.
5850 I've just tested the patch: Authentication still works fine,
5851 unavailable mechs specified in the exim configuration are still
5852 caught, and the auth.log warnings about OTP are gone."
5854 PH/25 When debugging is enabled, the contents of the command line are added
5855 to the debugging output, even when log_selector=+arguments is not
5858 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5859 answer is "GNU", and only if the return is "GNU/something" is the answer
5862 PH/27 $acl_verify_message is now set immediately after the failure of a
5863 verification in an ACL, and so is available in subsequent modifiers. In
5864 particular, the message can be preserved by coding like this:
5866 warn !verify = sender
5867 set acl_m0 = $acl_verify_message
5869 Previously, $acl_verify_message was set only while expanding "message"
5870 and "log_message" when a very denied access.
5872 PH/28 Modified OS/os.c-Linux with
5874 -#ifndef OS_LOAD_AVERAGE
5875 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5877 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5878 nomenclature these days.)
5880 PH/29 Installed patch from the Sieve maintainer that adds the options
5881 sieve_useraddress and sieve_subaddress to the redirect router.
5883 PH/30 In these circumstances:
5884 . Two addresses routed to the same list of hosts;
5885 . First host does not offer TLS;
5886 . First host accepts first address;
5887 . First host gives temporary error to second address;
5888 . Second host offers TLS and a TLS session is established;
5889 . Second host accepts second address.
5890 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5891 and peerdn, if requested) that were in fact used only for the second
5894 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5895 attention to any local part prefix or suffix that was matched by the
5896 router that accepted the address. It now behaves in the same way as it
5897 does for delivery: the affixes are removed from the local part unless
5898 rcpt_include_affixes is set on the transport.
5900 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5901 timeout during the DATA phase of an incoming message.
5903 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5904 applied a patch sanctioned by the Sieve maintainer.
5906 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5907 the uid or gid is negative. A case of a negative gid caused this to be
5908 noticed. The fix allows for either to be negative.
5910 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5911 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5914 PH/36 The change PH/12 above was broken. Fixed it.
5916 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5917 the grounds that routing the same address twice would always produce the
5918 same answer. This might have been true once, but it is certainly no
5919 longer true now. Routing a child address may depend on the previous
5920 routing that produced that child. Some complicated redirection strategies
5921 went wrong when messages had multiple recipients, and made Exim's
5922 behaviour dependent on the order in which the addresses were given.
5924 I have moved the duplicate checking until after the routing is complete.
5925 Exim scans the addresses that are assigned to local and remote
5926 transports, and removes any duplicates. This means that more work will be
5927 done, as duplicates will always all be routed, but duplicates are
5928 presumably rare, so I don't expect this is of any significance.
5930 For deliveries to pipes, files, and autoreplies, the duplicate checking
5931 still happens during the routing process, since they are not going to be
5934 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5935 It corrects a timeout issue with spamd. This is Ian's comment: "The
5936 background is that sometimes spamd either never reads data from a
5937 connection it has accepted, or it never writes response data. The exiscan
5938 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5939 blindly assumes that writes won't block so it may never time out."
5941 PH/39 Allow G after quota size as well as K and M.
5943 PH/40 The value set for $authenticated_id in an authenticator may not contain
5944 binary zeroes or newlines because the value is written to log lines and
5945 to spool files. There was no check on this. Now the value is run through
5946 the string_printing() function so that such characters are converted to
5947 printable escape sequences.
5949 PH/41 $message_linecount is a new variable that contains the total number of
5950 lines in the message. Compare $body_linecount, which is the count for the
5953 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5954 bounce and delay warning messages, except in certain special cases, which
5957 (a) An SMTP error message from a remote host;
5958 (b) A message specified in a :fail: redirection;
5959 (c) A message specified in a "fail" command in a system filter;
5960 (d) A message specified in a FAIL return from the queryprogram router;
5961 (e) A message specified by the cannot_route_message router option.
5963 In these cases only, Exim does include the error details in bounce and
5964 warning messages. There are also a few cases where bland messages such
5965 as "unrouteable address" or "local delivery error" are given.
5967 PH/43 $value is now also set for the "else" part of a ${run expansion.
5969 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5970 being worked on, but at least Exim now implements the latest version to
5973 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5974 process to complete was treated as a delivery failure, a timeout while
5975 writing the message to the pipe was logged, but erroneously treated as a
5976 successful delivery. Such timeouts include transport filter timeouts. For
5977 consistency with the overall process timeout, these timeouts are now
5978 treated as errors, giving rise to delivery failures by default. However,
5979 there is now a new Boolean option for the pipe transport called
5980 timeout_defer, which, if set TRUE, converts the failures into defers for
5981 both kinds of timeout. A transport filter timeout is now identified in
5984 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5985 systems where "make" and "gmake" are different, calling "gmake" at top
5986 level broke things. I've arranged for the value of $(MAKE) to be passed
5987 from the Makefile to this script so that it can call the same version of
5991 A note about Exim versions 4.44 and 4.50
5992 ----------------------------------------
5994 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5995 changes of various kinds. As a consequence, a big documentation update was
5996 needed. This delayed the release for rather longer than seemed good, especially
5997 in the light of a couple of (minor) security issues. Therefore, the changes
5998 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5999 release. So 4.44 and 4.50 are in effect two different branches that both start
6002 I have left the 4.50 change log unchanged; it contains all the changes since
6003 4.43. The change log for 4.44 is below; many of its items are identical to
6004 those for 4.50. This seems to be the most sensible way to preserve the
6005 historical information.
6011 1. Minor wording change to the doc/README.SIEVE file.
6013 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6014 computation of the current number of files was incorrect.
6016 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6017 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6020 4. Give more explanation in the error message when the command for a transport
6021 filter fails to execute.
6023 5. There are several places where Exim runs a non-Exim command in a
6024 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6025 was being done only for the command run by the queryprogram router. It is
6026 now done for all such subprocesses. The other cases are: ${run, transport
6027 filters, and the commands run by the lmtp and pipe transports.
6029 6. Added CONFIGURE_GROUP build-time option.
6031 7. Some older OS have a limit of 256 on the maximum number of file
6032 descriptors. Exim was using setrlimit() to set 1000 as a large value
6033 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6034 systems. I've change it so that if it can't get 1000, it tries for 256.
6036 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6037 was an oversight, and furthermore, ever since the addition of extra
6038 controls (e.g. 4.43/32), the checks on when to allow different forms of
6039 "control" were broken. There should now be diagnostics for all cases when a
6040 control that does not make sense is encountered.
6042 9. Added the /retain_sender option to "control=submission".
6044 10. $recipients is now available in the predata ACL (oversight).
6046 11. Tidy the search cache before the fork to do a delivery from a message
6047 received from the command line. Otherwise the child will trigger a lookup
6048 failure and thereby defer the delivery if it tries to use (for example) a
6049 cached ldap connection that the parent has called unbind on.
6051 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6052 of $address_data from the recipient verification was clobbered by the
6053 sender verification.
6055 13. The value of address_data from a sender verification is now available in
6056 $sender_address_data in subsequent conditions in the ACL statement.
6058 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6060 15. Added a new option "connect=<time>" to callout options, to set a different
6063 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6064 was its contents. (It was OK if the option was not defined at all.)
6066 17. A "Completed" log line is now written for messages that are removed from
6067 the spool by the -Mrm option.
6069 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6070 information about exactly what failed.
6072 19. Added -dd to debug only the daemon process.
6074 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6075 handles timeouts, both on the server side and network timeouts. Renamed the
6076 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6078 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6079 It is now set to "smtps".
6081 22. $host_address is now set to the target address during the checking of
6082 ignore_target_hosts.
6084 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6085 being passed; this would have caused $sender_host_name to have been used if
6086 matching the list had actually called for a host name (not very likely,
6087 since this list is usually IP addresses). A host name is now passed as
6090 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6091 code that specifies a non-restarting handler (typically sigaction() in
6092 modern systems) in an attempt to fix a rare and obscure crash bug.
6094 25. Narrowed the window for a race in the daemon that could cause it to ignore
6095 SIGCHLD signals. This is not a major problem, because they are used only to
6096 wake it up if nothing else does.
6098 26. A malformed maildirsize file could cause Exim to calculate negative values
6099 for the mailbox size or file count. Odd effects could occur as a result.
6100 The maildirsize information is now recalculated if the size or filecount
6103 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6104 support for a long time. Removed HAVE_SYS_VFS_H.
6106 28. Installed the latest version of exipick from John Jetmore.
6108 29. In an address list, if the pattern was not a regular expression, an empty
6109 subject address (from a bounce message) matched only if the pattern was an
6110 empty string. Non-empty patterns were not even tested. This was the wrong
6111 because it is perfectly reasonable to use an empty address as part of a
6112 database query. An empty address is now tested by patterns that are
6113 lookups. However, all the other forms of pattern expect the subject to
6114 contain a local part and a domain, and therefore, for them, an empty
6115 address still always fails if the pattern is not itself empty.
6117 30. Exim went into a mad DNS loop when attempting to do a callout where the
6118 host was specified on an smtp transport, and looking it up yielded more
6119 than one IP address.
6121 31. Re-factored the code for checking spool and log partition space into a
6122 function that finds that data and another that does the check. The former
6123 is then used to implement four new variables: $spool_space, $log_space,
6124 $spool_inodes, and $log_inodes.
6126 32. The RFC2047 encoding function was originally intended for short strings
6127 such as real names; it was not keeping to the 75-character limit for
6128 encoded words that the RFC imposes. It now respects the limit, and
6129 generates multiple encoded words if necessary. To be on the safe side, I
6130 have increased the buffer size for the ${rfc2047: expansion operator from
6133 33. It is now permitted to omit both strings after an "if" condition; if the
6134 condition is true, the result is "true". As before, when the second string
6135 is omitted, a false condition yields an empty string. This makes it less
6136 cumbersome to write custom ACL and router conditions.
6138 34. Failure to deliver a bounce message always caused it to be frozen, even if
6139 there was an errors_to setting on the router. The errors_to setting is now
6142 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6143 canonical form (fully expanded) before being placed in
6144 $sender_host_address.
6146 36. The table in the code that translates DNS record types into text (T_A to
6147 "A" for instance) was missing entries for NS and CNAME. It is just possible
6148 that this could have caused confusion if both these types were looked up
6149 for the same domain, because the text type is used as part of Exim's
6150 per-process caching. But the chance of anyone hitting this buglet seems
6153 37. The dnsdb lookup has been extended in a number of ways.
6155 (1) There is a new type, "zns", which walks up the domain tree until it
6156 finds some nameserver records. It should be used with care.
6158 (2) There is a new type, "mxh", which is like "mx" except that it returns
6159 just the host names, not the priorities.
6161 (3) It is now possible to give a list of domains (or IP addresses) to be
6162 looked up. The behaviour when one of the lookups defers can be
6163 controlled by a keyword.
6165 (4) It is now possible to specify the separator character for use when
6166 multiple records are returned.
6168 38. The dnslists ACL condition has been extended: it is now possible to supply
6169 a list of IP addresses and/or domains to be looked up in a particular DNS
6172 39. Added log_selector=+queue_time_overall.
6174 40. When running the queue in the test harness, wait just a tad after forking a
6175 delivery process, to get repeatability of debugging output.
6177 41. Include certificate and key file names in error message when GnuTLS fails
6178 to set them up, because the GnuTLS error message doesn't include the name
6179 of the failing file when there is a problem reading it.
6181 42. Allow both -bf and -bF in the same test run.
6183 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6185 44. The "Exiscan patch" is now merged into the mainline Exim source.
6187 45. Sometimes the final signoff response after QUIT could fail to get
6188 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6189 < 0 before doing a fflush(). This bug looks as though it goes back to the
6190 introduction of TLS in release 3.20, but "sometimes" must have been rare
6191 because the tests only now provoked it.
6193 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6194 (this can affect the format of dates).
6196 47. exim_tidydb, when checking for the continued existence of a message for
6197 which it has found a message-specific retry record, was not finding
6198 messages that were in split spool directories. Consequently, it was
6199 deleting retry records that should have stayed in existence.
6201 48. Steve fixed some bugs in eximstats.
6203 49. The SPA authentication driver was not abandoning authentication and moving
6204 on to the next authenticator when an expansion was forced to fail,
6205 contradicting the general specification for all authenticators. Instead it
6206 was generating a temporary error. It now behaves as specified.
6208 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6209 (the order specifies the preference for clients). The order is now AES256,
6210 AES128, 3DES, ARCFOUR128.
6212 51. Small patch to Sieve code - explicitly set From: when generating an
6215 52. Exim crashed if a remote delivery caused a very long error message to be
6216 recorded - for instance if somebody sent an entire SpamAssassin report back
6217 as a large number of 550 error lines. This bug was coincidentally fixed by
6218 increasing the size of one of Exim's internal buffers (big_buffer) that
6219 happened as part of the Exiscan merge. However, to be on the safe side, I
6220 have made the code more robust (and fixed the comments that describe what
6223 53. Now that there can be additional text after "Completed" in log lines (if
6224 the queue_time_overall log selector is set), a one-byte patch to exigrep
6225 was needed to allow it to recognize "Completed" as not the last thing in
6228 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6229 patch that reportedly fixes this has been added. I am not expert enough to
6230 create a test for it. This is what the patch creator wrote:
6232 "I found a little strange behaviour of ldap code when working with
6233 Windows 2003 AD Domain, where users was placed in more than one
6234 Organization Units. When I tried to give exim partial DN, the exit code
6235 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6236 But simultaneously result of request was absolutely normal ldap result,
6237 so I produce this patch..."
6239 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6240 REFERENCE, so I have modified the code to exclude the patch when that macro
6243 55. Some experimental protocols are using DNS PTR records for new purposes. The
6244 keys for these records are domain names, not reversed IP addresses. The
6245 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6246 leaves it alone. Component reversal etc. now happens only for IP addresses.
6249 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6251 57. Double the size of the debug message buffer (to 2048) so that more of very
6252 long debug lines gets shown.
6254 58. The exicyclog utility now does better if the number of log files to keep
6255 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6257 59. Two changes related to the smtp_active_hostname option:
6259 (1) $smtp_active_hostname is now available as a variable.
6260 (2) The default for smtp_banner uses $smtp_active_hostname instead
6261 of $primary_hostname.
6263 60. The host_aton() function is supposed to be passed a string that is known
6264 to be a valid IP address. However, in the case of IPv6 addresses, it was
6265 not checking this. This is a hostage to fortune. Exim now panics and dies
6266 if the condition is not met. A case was found where this could be provoked
6267 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6268 components; fortuitously, this particular loophole had already been fixed
6269 by change 4.50/55 above.
6271 If there are any other similar loopholes, the new check in host_aton()
6272 itself should stop them being exploited. The report I received stated that
6273 data on the command line could provoke the exploit when Exim was running as
6274 exim, but did not say which command line option was involved. All I could
6275 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6276 running as the user.
6279 61. There was a buffer overflow vulnerability in the SPA authentication code
6280 (which came originally from the Samba project). I have added a test to the
6281 spa_base64_to_bits() function which I hope fixes it.
6284 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6285 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6287 63. The daemon start-up calls getloadavg() while still root for those OS that
6288 need the first call to be done as root, but it missed one case: when
6289 deliver_queue_load_max is set with deliver_drop_privilege. This is
6290 necessary for the benefit of the queue runner, because there is no re-exec
6291 when deliver_drop_privilege is set.
6293 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6294 This has been fixed.
6296 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6297 was in use, was not putting the data itself into the right store pool;
6298 consequently, it could be overwritten for a subsequent message in the same
6299 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6302 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6304 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6305 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6306 the function were treating the return as a boolean value, which happened to
6307 work because 0=false and not-0=true, but is not correct code.
6309 68. The host_aton() function was not handling scoped IPv6 addresses (those
6310 with, for example, "%eth0" on the end) correctly.
6312 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6313 negated items (that is, ~something) in unsigned ints. Some compilers
6314 apparently mutter when there is no cast.
6316 70. If an address verification called from an ACL failed, and did not produce a
6317 user-specific message (i.e. there was only a "system" message), nothing was
6318 put in $acl_verify_message. In this situation, it now puts the system
6321 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6322 session; change 4.31/43 added the unwanted input to the log line - except
6323 that it did not do this in the start of session case. It now does.
6325 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6326 This is wrong and can cause the other end to generate a synchronization
6327 error if it is another Exim or anything else that does the synchronization
6328 check. A QUIT command is no longer sent after a timeout.
6330 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6331 during host lookups.
6333 74. The defer_ok option of callout verification was not working if it was used
6334 when verifying addresses in header lines, that is, for this case:
6336 verify = header_sender/callout=defer_ok
6338 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6339 those file descriptors could be used for SMTP connections. If anything
6340 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6341 could be sent to the SMTP client, causing chaos. The daemon now opens
6342 stdin, stdout, and stderr to /dev/null when it puts itself into the
6345 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6346 log by default. The user can override this with suitable Perl magic.
6348 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6349 the log message when discard triggers, was not working for the DATA ACL or
6350 for the non-SMTP ACL.
6352 78. Error message wording change in sieve.c.
6354 79. If smtp_accept_max_per_host was set, the number of connections could be
6355 restricted to fewer than expected, because the daemon was trying to set up
6356 a new connection before checking whether the processes handling previous
6357 connections had finished. The check for completed processes is now done
6358 earlier. On busy systems, this bug wouldn't be noticed because something
6359 else would have woken the daemon, and it would have reaped the completed
6362 80. If a message was submitted locally by a user whose login name contained one
6363 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6364 It caused a spool format error. I have fixed the spool reading code. A
6365 related problem was that the "from" clause in the Received: line became
6366 illegal because of the space(s). It is now covered by ${quote_local_part.
6368 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6371 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6373 83. Patch from Sieve maintainer to fix unterminated string problem in
6374 "vacation" handling.
6376 84. Some minor changes to the Linux configuration files to help with other
6377 OS variants using glibc.
6379 85. One more patch for Sieve to update vacation handling to latest spec.
6382 ----------------------------------------------------
6383 See the note above about the 4.44 and 4.50 releases.
6384 ----------------------------------------------------
6390 1. Change 4.43/35 introduced a bug that caused file counts to be
6391 incorrectly computed when quota_filecount was set in an appendfile
6394 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6395 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6398 3. Give more explanation in the error message when the command for a transport
6399 filter fails to execute.
6401 4. There are several places where Exim runs a non-Exim command in a
6402 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6403 was being done only for the command run by the queryprogram router. It is
6404 now done for all such subprocesses. The other cases are: ${run, transport
6405 filters, and the commands run by the lmtp and pipe transports.
6407 5. Some older OS have a limit of 256 on the maximum number of file
6408 descriptors. Exim was using setrlimit() to set 1000 as a large value
6409 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6410 systems. I've change it so that if it can't get 1000, it tries for 256.
6412 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6413 was an oversight, and furthermore, ever since the addition of extra
6414 controls (e.g. 4.43/32), the checks on when to allow different forms of
6415 "control" were broken. There should now be diagnostics for all cases when a
6416 control that does not make sense is encountered.
6418 7. $recipients is now available in the predata ACL (oversight).
6420 8. Tidy the search cache before the fork to do a delivery from a message
6421 received from the command line. Otherwise the child will trigger a lookup
6422 failure and thereby defer the delivery if it tries to use (for example) a
6423 cached ldap connection that the parent has called unbind on.
6425 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6426 of $address_data from the recipient verification was clobbered by the
6427 sender verification.
6429 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6430 was its contents. (It was OK if the option was not defined at all.)
6432 11. A "Completed" log line is now written for messages that are removed from
6433 the spool by the -Mrm option.
6435 12. $host_address is now set to the target address during the checking of
6436 ignore_target_hosts.
6438 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6439 being passed; this would have caused $sender_host_name to have been used if
6440 matching the list had actually called for a host name (not very likely,
6441 since this list is usually IP addresses). A host name is now passed as
6444 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6445 code that specifies a non-restarting handler (typically sigaction() in
6446 modern systems) in an attempt to fix a rare and obscure crash bug.
6448 15. Narrowed the window for a race in the daemon that could cause it to ignore
6449 SIGCHLD signals. This is not a major problem, because they are used only to
6450 wake it up if nothing else does.
6452 16. A malformed maildirsize file could cause Exim to calculate negative values
6453 for the mailbox size or file count. Odd effects could occur as a result.
6454 The maildirsize information is now recalculated if the size or filecount
6457 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6458 support for a long time. Removed HAVE_SYS_VFS_H.
6460 18. Updated exipick to current release from John Jetmore.
6462 19. Allow an empty sender to be matched against a lookup in an address list.
6463 Previously the only cases considered were a regular expression, or an
6466 20. Exim went into a mad DNS lookup loop when doing a callout where the
6467 host was specified on the transport, if the DNS lookup yielded more than
6470 21. The RFC2047 encoding function was originally intended for short strings
6471 such as real names; it was not keeping to the 75-character limit for
6472 encoded words that the RFC imposes. It now respects the limit, and
6473 generates multiple encoded words if necessary. To be on the safe side, I
6474 have increased the buffer size for the ${rfc2047: expansion operator from
6477 22. Failure to deliver a bounce message always caused it to be frozen, even if
6478 there was an errors_to setting on the router. The errors_to setting is now
6481 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6482 canonical form (fully expanded) before being placed in
6483 $sender_host_address.
6485 24. Updated eximstats to version 1.33
6487 25. Include certificate and key file names in error message when GnuTLS fails
6488 to set them up, because the GnuTLS error message doesn't include the name
6489 of the failing file when there is a problem reading it.
6491 26. Expand error message when OpenSSL has problems setting up cert/key files.
6494 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6495 (this can affect the format of dates).
6497 28. exim_tidydb, when checking for the continued existence of a message for
6498 which it has found a message-specific retry record, was not finding
6499 messages that were in split spool directories. Consequently, it was
6500 deleting retry records that should have stayed in existence.
6502 29. eximstats updated to version 1.35
6503 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6504 1.35 - bugfix such that pie charts by volume are generated correctly
6506 30. The SPA authentication driver was not abandoning authentication and moving
6507 on to the next authenticator when an expansion was forced to fail,
6508 contradicting the general specification for all authenticators. Instead it
6509 was generating a temporary error. It now behaves as specified.
6511 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6512 (the order specifies the preference for clients). The order is now AES256,
6513 AES128, 3DES, ARCFOUR128.
6515 31. Small patch to Sieve code - explicitly set From: when generating an
6518 32. Exim crashed if a remote delivery caused a very long error message to be
6519 recorded - for instance if somebody sent an entire SpamAssassin report back
6520 as a large number of 550 error lines. This bug was coincidentally fixed by
6521 increasing the size of one of Exim's internal buffers (big_buffer) that
6522 happened as part of the Exiscan merge. However, to be on the safe side, I
6523 have made the code more robust (and fixed the comments that describe what
6526 33. Some experimental protocols are using DNS PTR records for new purposes. The
6527 keys for these records are domain names, not reversed IP addresses. The
6528 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6529 leaves it alone. Component reversal etc. now happens only for IP addresses.
6532 34. The host_aton() function is supposed to be passed a string that is known
6533 to be a valid IP address. However, in the case of IPv6 addresses, it was
6534 not checking this. This is a hostage to fortune. Exim now panics and dies
6535 if the condition is not met. A case was found where this could be provoked
6536 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6537 components; fortuitously, this particular loophole had already been fixed
6538 by change 4.50/55 or 4.44/33 above.
6540 If there are any other similar loopholes, the new check in host_aton()
6541 itself should stop them being exploited. The report I received stated that
6542 data on the command line could provoke the exploit when Exim was running as
6543 exim, but did not say which command line option was involved. All I could
6544 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6545 running as the user.
6548 35. There was a buffer overflow vulnerability in the SPA authentication code
6549 (which came originally from the Samba project). I have added a test to the
6550 spa_base64_to_bits() function which I hope fixes it.
6553 36. The daemon start-up calls getloadavg() while still root for those OS that
6554 need the first call to be done as root, but it missed one case: when
6555 deliver_queue_load_max is set with deliver_drop_privilege. This is
6556 necessary for the benefit of the queue runner, because there is no re-exec
6557 when deliver_drop_privilege is set.
6559 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6560 was in use, was not putting the data itself into the right store pool;
6561 consequently, it could be overwritten for a subsequent message in the same
6562 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6565 38. Sometimes the final signoff response after QUIT could fail to get
6566 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6567 < 0 before doing a fflush(). This bug looks as though it goes back to the
6568 introduction of TLS in release 3.20, but "sometimes" must have been rare
6569 because the tests only now provoked it.
6575 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6576 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6577 Now it returns an integer. A number of calls were still expecting a T/F
6578 return. Fortuitously, in all cases, the tests worked in OK situations,
6579 which is the norm. However, things would have gone wrong on any write
6580 failures on the smtp file descriptor. This function is used when sending
6581 messages over SMTP and also when doing verify callouts.
6583 2. When Exim is called to do synchronous delivery of a locally submitted
6584 message (the -odf or -odi options), it no longer closes stderr before doing
6587 3. Implemented the mua_wrapper option.
6589 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6591 5. Implemented the functions header_remove(), header_testname(),
6592 header_add_at_position(), and receive_remove_recipient(), and exported them
6595 6. If an ACL "warn" statement specified the addition of headers, Exim already
6596 inserted X-ACL-Warn: at the start if there was no header name. However, it
6597 was not making this test for the second and subsequent header lines if
6598 there were newlines in the string. This meant that an invalid header could
6599 be inserted if Exim was badly configured.
6601 7. Allow an ACL "warn" statement to add header lines at the start or after all
6602 the Received: headers, as well as at the end.
6604 8. Added the rcpt_4xx retry error code.
6606 9. Added postmaster_mailfrom=xxx to callout verification option.
6608 10. Added mailfrom=xxxx to the callout verification option, for verify=
6611 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6612 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6614 12. Inserted some casts to stop certain compilers warning when using pointer
6615 differences as field lengths or precisions in printf-type calls (mostly
6616 affecting debugging statements).
6618 13. Added optional readline() support for -be (dynamically loaded).
6620 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6621 same clock tick as a message's arrival, so that its received time was the
6622 same as the "first fail" time on the retry record, and that message
6623 remained on the queue past the ultimate address timeout, every queue runner
6624 would try a delivery (because it was past the ultimate address timeout) but
6625 after another failure, the ultimate address timeout, which should have then
6626 bounced the address, did not kick in. This was a "< instead of <=" error;
6627 in most cases the first failure would have been in the next clock tick
6628 after the received time, and all would be well.
6630 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6631 being recognized when the domain list was tested by the match_domain
6632 condition in an expansion string.
6634 16. Added the ${str2b64: operator.
6636 17. Exim was always calling setrlimit() to set a large limit for the number of
6637 processes, without checking whether the existing limit was already
6638 adequate. (It did check for the limit on file descriptors.) Furthermore,
6639 errors from getrlimit() and setrlimit() were being ignored. Now they are
6640 logged to the main and panic logs, but Exim does carry on, to try to do its
6641 job under whatever limits there are.
6643 18. Imported PCRE 5.0.
6645 19. Trivial typo in log message " temporarily refused connection" (the leading
6648 20. If the log selector return_path_on_delivery was set and an address was
6649 redirected to /dev/null, the delivery process crashed because it assumed
6650 that a return path would always be set for a "successful" delivery. In this
6651 case, the whole delivery is bypassed as an optimization, and therefore no
6654 21. Internal re-arrangement: the function for sending a challenge and reading
6655 a response while authentication was assuming a zero-terminated challenge
6656 string. It's now changed to take a pointer and a length, to allow for
6657 binary data in such strings.
6659 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6661 23. Exim was not respecting finduser_retries when seeking the login of the
6662 uid under which it was called; it was always trying 10 times. (The default
6663 setting of finduser_retries is zero.) Also, it was sleeping after the final
6664 failure, which is pointless.
6666 24. Implemented tls_on_connect_ports.
6668 25. Implemented acl_smtp_predata.
6670 26. If the domain in control=submission is set empty, Exim assumes that the
6671 authenticated id is a complete email address when it generates From: or
6672 Sender: header lines.
6674 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6675 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6676 chown and chgrp in /bin and hostname in /usr/bin.
6678 28. Exim was keeping the "process log" file open after each use, just as it
6679 does for the main log. This opens the possibility of it remaining open for
6680 long periods when the USR1 signal hits a daemon. Occasional processlog
6681 errors were reported, that could have been caused by this. Anyway, it seems
6682 much more sensible not to leave this file open at all, so that is what now
6685 29. The long-running daemon process does not normally write to the log once it
6686 has entered its main loop, and it closes the log before doing so. This is
6687 so that log files can straightforwardly be renamed and moved. However,
6688 there are a couple of unusual error situations where the daemon does write
6689 log entries, and I had neglected to close the log afterwards.
6691 30. The text of an SMTP error response that was received during a remote
6692 delivery was being truncated at 512 bytes. This is too short for some of
6693 the long messages that one sometimes sees. I've increased the limit to
6696 31. It is now possible to make retry rules that apply only when a message has a
6697 specific sender, in particular, an empty sender.
6699 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6700 it possible to be selective about when SMTP synchronization is enforced.
6702 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6704 32. Implemented hosts_connection_nolog.
6706 33. Added an ACL for QUIT.
6708 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6711 35. Added mailbox_size and mailbox_filecount to appendfile.
6713 36. Added control = no_multiline_responses to ACLs.
6715 37. There was a bug in the logic of the code that waits for the clock to tick
6716 in the case where the clock went backwards by a substantial amount such
6717 that the microsecond fraction of "now" was more than the microsecond
6718 fraction of "then" (but the whole seconds number was less).
6720 38. Added support for the libradius Radius client library this is found on
6721 FreeBSD (previously only the radiusclient library was supported).
6727 1. When certain lookups returned multiple values in the form name=value, the
6728 quoting of the values was not always being done properly. Specifically:
6729 (a) If the value started with a double quote, but contained no whitespace,
6731 (b) If the value contained whitespace other than a space character (i.e.
6732 tabs or newlines or carriage returns) it was not quoted.
6733 This fix has been applied to the mysql and pgsql lookups by writing a
6734 separate quoting function and calling it from the lookup code. The fix
6735 should probably also be applied to nisplus, ibase and oracle lookups, but
6736 since I cannot test any of those, I have not disturbed their existing code.
6738 2. A hit in the callout cache for a specific address caused a log line with no
6739 reason for rejecting RCPT. Now it says "Previous (cached) callout
6740 verification failure".
6742 3. There was an off-by-one bug in the queryprogram router. An over-long
6743 return line was truncated at 256 instead of 255 characters, thereby
6744 overflowing its buffer with the terminating zero. As well as fixing this, I
6745 have increased the buffer size to 1024 (and made a note to document this).
6747 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6748 when Exim is waiting for an SMTP response from a remote server, Exim
6749 restarts its select() call on the socket, thereby resetting its timeout.
6750 This is not a problem when such interrupts are rare. Somebody set up a cron
6751 job to run exiwhat every 2 minutes, which is less than the normal select()
6752 timeout (5 or 10 minutes). This meant that the select() timeout never
6753 kicked in because it was always reset. I have fixed this by comparing the
6754 time when an interrupt arrives with the time at the start of the first call
6755 to select(). If more time than the timeout has elapsed, the interrupt is
6756 treated as a timeout.
6758 5. Some internal re-factoring in preparation for the addition of Sieve
6759 extensions (by MH). In particular, the "personal" test is moved to a
6760 separate function, and given an option for scanning Cc: and Bcc: (which is
6761 not set for Exim filters).
6763 6. When Exim created an email address using the login of the caller as the
6764 local part (e.g. when creating a From: or Sender: header line), it was not
6765 quoting the local part when it contained special characters such as @.
6767 7. Installed new OpenBSD configuration files.
6769 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6770 try to make them clearer.
6772 9. Callout options, other than the timeout value, were being ignored when
6773 verifying sender addresses in header lines. For example, when using
6775 verify = header_sender/callout=no_cache
6777 the cache was (incorrectly) being used.
6779 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6780 only the Cygwin environment.
6782 11. When return_path_on_delivery was set as a log selector, if different remote
6783 addresses in the same message used different return paths and parallel
6784 remote delivery occurred, the wrong values would sometimes be logged.
6785 (Whenever a remote delivery process finished, the return path value from
6786 the most recently started remote delivery process was logged.)
6788 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6789 lines when AUTH and/or TLS are in use. This is the "received protocol"
6790 field. Exim used to use "asmtp" for authenticated SMTP, without any
6791 indication (in the protocol name) for TLS use. Now it follows the RFC and
6792 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6793 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6794 names appear in log lines as well as in Received: header lines.
6796 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6797 extensions, and comparison tests, and to fix some bugs.
6799 14. Changes to the "personal" filter test:
6801 (1) The test was buggy in that it was just doing the equivalent of
6802 "contains" tests on header lines. For example, if a user's address was
6803 anne@some.where, the "personal" test would incorrectly be true for
6805 To: susanne@some.where
6807 This test is now done by extracting each address from the header in turn,
6808 and checking the entire address. Other tests that are part of "personal"
6809 are now done using regular expressions (for example, to check local parts
6810 of addresses in From: header lines).
6812 (2) The list of non-personal local parts in From: addresses has been
6813 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6814 taken from the Sieve specification recommendations.
6816 (3) If the message contains any header line starting with "List-" it is
6817 treated as non-personal.
6819 (4) The test for "circular" in the Subject: header line has been removed
6820 because it now seems ill-conceived.
6822 15. Minor typos in src/EDITME comments corrected.
6824 16. Installed latest exipick from John Jetmore.
6826 17. If headers_add on a router specified a text string that was too long for
6827 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6828 of string_sprintf() is now avoided.
6830 18. $message_body_size was not set (it was always zero) when running the DATA
6831 ACL and the local_scan() function.
6833 19. For the "mail" command in an Exim filter, no default was being set for
6834 the once_repeat time, causing a random time value to be used if "once" was
6835 specified. (If the value happened to be <= 0, no repeat happened.) The
6836 default is now 0s, meaning "never repeat". The "vacation" command was OK
6837 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6838 (I found it when inspecting the code).
6840 20. There is now an overall timeout for performing a callout verification. It
6841 defaults to 4 times the callout timeout, which applies to individual SMTP
6842 commands during the callout. The overall timeout applies when there is more
6843 than one host that can be tried. The timeout is checked before trying the
6844 next host. This prevents very long delays if there are a large number of
6845 hosts and all are timing out (e.g. when the network connections are timing
6846 out). The value of the overall timeout can be changed by specifying an
6847 additional sub-option for "callout", called "maxwait". For example:
6849 verify = sender/callout=5s,maxwait=20s
6851 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6852 to the end before writing, but this should make it even safer).
6854 22. Exim was forgetting that it had advertised PIPELINING for the second and
6855 subsequent messages on an SMTP connection. It was also not resetting its
6856 memory on STARTTLS and an internal HELO.
6858 23. When Exim logs an SMTP synchronization error within a session, it now
6859 records whether PIPELINING has been advertised or not.
6861 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6862 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6863 rather than long int.
6865 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6867 26. Added the never_mail option to autoreply.
6873 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6874 crash if the getsockname() call failed; this can happen if a connection is
6875 closed very soon after it is established. The problem was simply in the
6876 order in which certain operations were done, causing Exim to try to write
6877 to the SMTP stream before it had set up the file descriptor. The bug has
6878 been fixed by making things happen in the correct order.
6884 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6885 before the connection was closed, thus losing the rejection response.
6887 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6888 some early Solaris releases, but causes trouble in current releases where
6889 socklen_t is defined.
6891 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6894 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6897 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6898 that releasing the top of it at the end releases what was used for sub-
6899 expansions (unless the block got too big). However, discard this block if
6900 the first thing is a variable or header, so that we can use its block when
6901 it is dynamic (useful for very large $message_headers, for example).
6903 6. Lookups now cache *every* query, not just the most recent. A new, separate
6904 store pool is used for this. It can be recovered when all lookup caches are
6905 flushed. Lookups now release memory at the end of their result strings.
6906 This has involved some general refactoring of the lookup sources.
6908 7. Some code has been added to the store_xxx() functions to reduce the amount
6909 of flapping under certain conditions.
6911 8. log_incoming_interface used to affect only the <= reception log lines. Now
6912 it causes the local interface and port to be added to several more SMTP log
6913 lines, for example "SMTP connection from", and rejection lines.
6915 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6917 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6919 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6920 could be overwritten at the end of the current message (or the start of a
6921 new message if it was set in a HELO ACL). The value is now preserved for
6922 the duration of the SMTP connection.
6924 12. If a transport had a headers_rewrite setting, and a matching header line
6925 contained an unqualified address, that address was qualified, even if it
6926 did not match any rewriting rules. The underlying bug was that the values
6927 of the flags that permit the existence of unqualified sender and recipient
6928 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6929 non-local messages, and by -bnq for local messages) were not being
6930 preserved with the message after it was received.
6932 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6933 "next input=" as part of the text comprising the host identity instead of
6934 the correct text. The code was using the same buffer for two different
6935 strings. However, depending on which order the printing function evaluated
6936 its arguments, the bug did not always show up. Under Linux, for example, my
6937 test suite worked just fine.
6939 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6940 doesn't work with some older versions of Perl. It has been changed to "my",
6941 which in any case is probably the better facility to use.
6943 15. A really picky compiler found some instances of statements for creating
6944 error messages that either had too many or two few arguments for the format
6947 16. The size of the buffer for calls to the DNS resolver has been increased
6948 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6949 for addresses that have a lot of PTR records. This alleviates a problem; it
6950 does not fully solve it.
6952 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6953 buffer now truncates the list and logs the incident, which is the same
6954 action as happens when Exim is looking up a host name and its aliases.
6955 Previously in this situation something unpredictable would happen;
6956 sometimes it was "internal error: store_reset failed".
6958 18. If a server dropped the connection unexpectedly when an Exim client was
6959 using GnuTLS and trying to read a response, the client delivery process
6960 crashed while trying to generate an error log message.
6962 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6963 string, for example:
6965 warn message = H1: something\nH2: something
6967 the text was added as a single header line from Exim's point of view
6968 though it ended up OK in the delivered message. However, searching for the
6969 second and subsequent header lines using $h_h2: did not work. This has been
6970 fixed. Similarly, if a system filter added multiple headers in this way,
6971 the routers could not see them.
6973 20. Expanded the error message when iplsearch is called with an invalid key to
6974 suggest using net-iplsearch in a host list.
6976 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6977 ACLs are no longer actually imposed (and a message to that effect is
6980 22. If a "gecos" field in a passwd entry contained escaped characters, in
6981 particular, if it contained a \" sequence, Exim got it wrong when building
6982 a From: or a Sender: header from that name. A second bug also caused
6983 incorrect handling when an unquoted " was present following a character
6984 that needed quoting.
6986 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6987 was not being matched caselessly.
6989 24. Arranged for all hyphens in the exim.8 source to be escaped with
6992 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6993 properly. Recipient callout cache records were still being keyed to include
6994 the sender, even when use_sender was set false. This led to far more
6995 callouts that were necessary. The sender is no longer included in the key
6996 when use_sender is false.
6998 26. Added "control = submission" modifier to ACLs.
7000 27. Added the ${base62d: operator to decode base 62 numbers.
7002 28. dnsdb lookups can now access SRV records.
7004 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7005 the configuration file.
7007 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7008 (-v) mode. This makes the output for a verbose queue run more intelligible.
7010 31. Added a use_postmaster feature to recipient callouts.
7012 32. Added the $body_zerocount variable, containing the number of binary zero
7013 bytes in the message body.
7015 33. The time of last modification of the "new" subdirectory is now used as the
7016 "mailbox time last read" when there is a quota error for a maildir
7019 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7021 35. Added +ignore_unknown as a special item in host lists.
7023 36. Code for decoding IPv6 addresses in host lists is now included, even if
7024 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7025 address was recognized as an IP address, but was then not correctly decoded
7026 into binary, causing unexpected and incorrect effects when compared with
7033 1. Very minor rewording of debugging text in manualroute to say "list of
7034 hosts" instead of "hostlist".
7036 2. If verify=header_syntax was set, and a header line with an unqualified
7037 address (no domain) and a large number of spaces between the end of the
7038 name and the colon was received, the reception process suffered a buffer
7039 overflow, and (when I tested it) crashed. This was caused by some obsolete
7040 code that should have been removed. The fix is to remove it!
7042 3. When running in the test harness, delay a bit after writing a bounce
7043 message to get a bit more predictability in the log output.
7045 4. Added a call to search_tidyup() just before forking a reception process. In
7046 theory, someone could use a lookup in the expansion of smtp_accept_max_
7047 per_host which, without the tidyup, could leave open a database connection.
7049 5. Added the variables $recipient_data and $sender_data which get set from a
7050 lookup success in an ACL "recipients" or "senders" condition, or a router
7051 "senders" option, similar to $domain_data and $local_part_data.
7053 6. Moved the writing of debug_print from before to after the "senders" test
7056 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7057 problems for message scanning, either using a data ACL, or using
7058 local_scan() because the Received: header was not generated till after they
7059 were called (in order to set the time as the time of reception completion).
7060 I have revised the way this works. The header is now generated after the
7061 body is received, but before the ACL or local_scan() are called. After they
7062 are run, the timestamp in the header is updated.
7068 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7069 before starting a queue runner without re-exec. This happened only when
7070 deliver_drop_privilege was set or when the Exim user was set to root. The
7071 effect of the bug was that timeouts during subsequent deliveries caused
7072 crashes instead of being properly handled. The handler is now left at its
7073 default (and expected) setting.
7075 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7076 message, again when deliver_drop_privilege is set or Exim is run as root.
7077 The bug described in (1) was not present in this case, but the tidying up
7078 of the other signals was missing. I have made the two cases consistent.
7080 3. The ignore_target_hosts setting on a manualroute router was being ignored
7081 for hosts that were looked up using the /MX notation.
7083 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7086 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7087 operated on the sender address. After changing the $sender_address to <>
7088 for the sender address verify, Exim was re-instated it as the original
7089 (before rewriting) address, but remembering that it had rewritten it, so it
7090 wasn't rewriting it again. This bug also had the effect of breaking the
7091 sender address verification caching when the sender address was rewritten.
7093 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7094 This has been changed so that if the ip literal address matches
7095 ignore_target_hosts, the router declines.
7097 7. Added expansion conditions match_domain, match_address, and match_local_
7098 part (NOT match_host).
7100 8. The placeholder for the Received: header didn't have a length field set.
7102 9. Added code to Exim itself and to exim_lock to test for a specific race
7103 condition that could lead to file corruption when using MBX delivery. The
7104 issue is with the lockfile that is created in /tmp. If this file is removed
7105 after a process has opened it but before that process has acquired a lock,
7106 there is the potential for a second process to recreate the file and also
7107 acquire a lock. This could lead to two Exim processes writing to the file
7108 at the same time. The added code performs the same test as UW imapd; it
7109 checks after acquiring the lock that its file descriptor still refers to
7110 the same named file.
7112 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7113 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7116 11. Added the smtp_active_hostname option. If used, this will typically be made
7117 to depend on the incoming interface address. Because $interface_address is
7118 not set up until the daemon has forked a reception process, error responses
7119 that can happen earlier (such as "too many connections") no longer contain
7122 12. If an expansion in a condition on a "warn" statement fails because a lookup
7123 defers, the "warn" statement is abandoned, and the next ACL statement is
7124 processed. Previously this caused the whole ACL to be aborted.
7126 13. Added the iplsearch lookup type.
7128 14. Added ident_timeout as a log selector.
7130 15. Added tls_certificate_verified as a log selector.
7132 16. Added a global option tls_require_ciphers (compare the smtp transport
7133 option of the same name). This controls incoming TLS connections.
7135 17. I finally figured out how to make tls_require_ciphers do a similar thing
7136 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7137 before starting the TLS session.
7139 18. Tabs are now shown as \t in -bP output.
7141 19. If the log selector return_path_on_delivery was set, Exim crashed when
7142 bouncing a message because it had too many Received: header lines.
7144 20. If two routers both had headers_remove settings, and the first one included
7145 a superfluous trailing colon, the final name in the first list and the
7146 first name in the second list were incorrectly joined into one item (with a
7147 colon in the middle).
7153 1. Added -C and -D options to the exinext utility, mainly to make it easier
7154 to include in the automated testing, but these could be helpful when
7155 multiple configurations are in use.
7157 2. The exinext utility was not formatting the output nicely when there was
7158 an alternate port involved in the retry record key, nor when there was a
7159 message id as well (for retries that were specific to a specific message
7160 and a specific host). It was also confused by IPv6 addresses, because of
7161 the additional colons they contain. I have fixed the IPv4 problem, and
7162 patched it up to do a reasonable job for IPv6.
7164 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7165 delivery, the log line now contains "pipelined" if PIPELINING was used.
7167 4. An SMTP transport process used to panic and die if the bind() call to set
7168 an explicit outgoing interface failed. This has been changed; it is now
7169 treated in the same way as a connect() failure.
7171 5. A reference to $sender_host_name in the part of a conditional expansion
7172 that was being skipped was still causing a DNS lookup. This no longer
7175 6. The def: expansion condition was not recognizing references to header lines
7176 that used bh_ and bheader_.
7178 7. Added the _cache feature to named lists.
7180 8. The code for checking quota_filecount in the appendfile transport was
7181 allowing one more file than it should have been.
7183 9. For compatibility with Sendmail, the command line option
7191 and sets the incoming protocol and host name (for trusted callers). The
7192 host name and its colon can be omitted when only the protocol is to be set.
7193 Note the Exim already has two private options, -pd and -ps, that refer to
7194 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7195 "s", but I don't think that's a major issue.
7197 10. A number of refactoring changes to the code, none of which should affect
7200 (a) The number of logging options was getting close to filling up the
7201 32-bit word that was used as a bit map. I have split them into two classes:
7202 those that are passed in the argument to log_write(), and those that are
7203 only ever tested independently outside of that function. These are now in
7204 separate 32-bit words, so there is plenty of room for expansion again.
7205 There is no change in the user interface or the logging behaviour.
7207 (b) When building, for example, log lines, the code previously used a
7208 macro that called string_cat() twice, in order to add two strings. This is
7209 not really sufficiently general. Furthermore, there was one instance where
7210 it was actually wrong because one of the argument was used twice, and in
7211 one call a function was used. (As it happened, calling the function twice
7212 did not affect the overall behaviour.) The macro has been replaced by a
7213 function that can join an arbitrary number of extra strings onto a growing
7216 (c) The code for expansion conditions now uses a table and a binary chop
7217 instead of a serial search (which was left over from when there were very
7218 few conditions). Also, it now recognizes conditions like "pam" even when
7219 the relevant support is not compiled in: a suitably worded error message is
7220 given if an attempt is made to use such a condition.
7222 11. Added ${time_interval:xxxxx}.
7224 12. A bug was causing one of the ddress fields not to be passed back correctly
7225 from remote delivery subprocesses. The field in question was not being
7226 subsequently used, so this caused to problems in practice.
7228 13. Added new log selectors queue_time and deliver_time.
7230 14. Might have fixed a bug in maildirsizefile handling that threw up
7231 "unexpected character" debug warnings, and recalculated the data
7232 unnecessarily. In any case, I expanded the warning message to give more
7235 15. Added the message "Restricted characters in address" to the statements in
7236 the default ACL that block characters like @ and % in local parts.
7238 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7239 Three changes have been made:
7241 (a) There was a serious bug; a negative response to MAIL caused the whole
7242 recipient domain to be cached as invalid, thereby blocking all messages
7243 to all local parts at the same domain, from all senders. This bug has
7244 been fixed. The domain is no longer cached after a negative response to
7245 MAIL if the sender used is not empty.
7247 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7250 (c) A new callout option, "use_sender" has been added for people who want
7251 the modified behaviour.
7257 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7260 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7261 indeed breaks things for older releases.
7263 3. Added additional logging to the case where there is a problem reading data
7264 from a filter that is running in a subprocess using a pipe, in order to
7265 try to track down a specific problem.
7267 4. Testing facility fudge: when running in the test harness and attempting
7268 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7269 getting "No route to host". Convert this to a timeout.
7271 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7274 6. Some OS don't have socklen_t but use size_t instead. This affects the
7275 fifth argument of getsockopt() amongst other things. This is now
7276 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7277 can be set for individual OS. I have set it for SunOS5, OSF1, and
7278 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7279 some earlier ones do not.
7281 7. Change 4.30/15 was not doing the test caselessly.
7283 8. The standard form for an IPv6 address literal was being rejected by address
7284 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7285 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7286 this, as well as the form without the "IPv6" on the front (but only when
7287 address literals are enabled, of course).
7289 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7291 10. Exim crashed if a message with an empty sender address specified by -f
7292 encountered a router with an errors_to setting. This could be provoked only
7293 by a command such as
7297 where an empty string was supplied; "<>" did not hit this bug.
7299 11. Installed PCRE release 4.5.
7301 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7302 remained set. It is now erased.
7304 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7305 times from message ids (which are base 36 rather than the normal 62).
7307 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7308 were being counted as actual protocol errors, and logged if the log
7309 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7310 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7311 and DATA following a set of rejected RCPTs do not count as protocol errors.
7312 In other words, Exim assumes they were pipelined, though this may not
7313 actually be the case. Of course, in all cases the client gets an
7314 appropriate error code.
7316 15. If a lookup fails in an ACL condition, a message about the failure may
7317 be available; it is used if testing the ACL cannot continue, because most
7318 such messages specify what the cause of the deferral is. However, some
7319 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7320 that caused an old message to be retained and used if a later statement
7321 caused a defer, replacing the real cause of the deferral.
7323 16. If an IP address had so many PTR records that the DNS lookup buffer
7324 was not large enough to hold them, Exim could crash while trying to process
7325 the truncated data. It now detects and logs this case.
7327 17. Further to 4.21/58, another change has been made: if (and only if) the
7328 first line of a message (the first header line) ends with CRLF, a bare LF
7329 in a subsequent header line has a space inserted after it, so as not to
7330 terminate the header.
7332 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7333 unnecessarily, used atoi() instead of strtol(), and didn't check the
7334 termination when getting file sizes from file names by regex.
7336 19. Completely re-implemented the support for maildirsize files, in the light
7337 of a number of problems with the previous contributed implementation
7338 (4.30/29). In particular:
7340 . If the quota is zero, the maildirsize file is maintained, but no quota is
7343 . If the maildir directory does not exist, it is created before any attempt
7344 to write a maildirsize file.
7346 . The quota value in the file is just a cache; if the quota is changed in
7347 the transport, the new value overrides.
7349 . A regular expression is available for excluding directories from the
7352 20. The autoreply transport checks the characters in options that define the
7353 message's headers; it allows continued headers, but it was checking with
7354 isspace() after an embedded newline instead of explicitly looking for a
7357 21. If all the "regular" hosts to which an address was routed had passed their
7358 expiry times, and had not reached their retry times, the address was
7359 bounced, even if fallback hosts were defined. Now Exim should go on to try
7362 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7363 equivalent code in the SMTP transport. Some hosts send humungous responses
7364 to HELO/EHLO, more than 1024 it seems.
7366 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7367 gives compiler warnings in some environments. I've now done it "properly",
7370 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7371 (because of problems with the built-in one) was declared to return uschar *
7372 instead of char *, causing compiler failure.
7374 25. Fixed a file descriptor leak when processing alias/forward files.
7376 26. Fixed a minor format string issue in dbfn.c.
7378 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7380 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7381 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7384 29. When a sender address is verified, it is cached, to save repeating the test
7385 when there is more than one recipient in a message. However, when the
7386 verification involves a callout, it is possible for different callout
7387 options to be set for different recipients. It is too complicated to keep
7388 track of this in the cache, so now Exim always runs a verification when a
7389 callout is required, relying on the callout cache for the optimization.
7390 The overhead is duplication of the address routing, but this should not be
7393 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7394 to be verified with callout=postmaster, and the main callout worked but the
7395 postmaster check failed, the verification correctly failed. However, if a
7396 subsequent RCPT command asked for sender verification *without* the
7397 postmaster check, incorrect caching caused this verification also to fail,
7400 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7401 it was not caching the DNS options (qualify_single, search_parents) that
7402 were used when the lookup failed. A subsequent lookup with different
7403 options therefore always gave the same answer, though there were cases
7404 where it should not have. (Example: a "domains = !$mx_any" option on a
7405 dnslookup router: the "domains" option is always processed without any
7406 widening, but the router might have qualify_single set.) Now Exim uses the
7407 cached value only when the same options are set.
7409 32. Added John Jetmore's "exipick" utility to the distribution.
7411 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7412 than a timeout (e.g. a certificate is required, and is not provided), an
7413 Exim server now closes the connection immediately. Previously it waited for
7414 the client to close - but if the client is SSL, it seems that they each
7415 wait for each other, leading to a delay before one of them times out.
7417 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7418 maintained 0.8.x compatibility because I don't think many are using it, and
7419 it is clearly obsolete.
7421 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7424 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7425 client certificate was expired. A simple patch fixes this, though I don't
7426 understand the full logic of why the verify callback is called multiple
7429 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7430 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7431 which causes problems with some clients (such as the Certicom SSL Plus
7432 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7433 disables the coutermeasure allowing Eudora to connect."
7435 38. Exim was not checking that a write() to a log file succeeded. This could
7436 lead to Bad Things if a log got too big, in particular if it hit a file
7437 size limit. Exim now panics and dies if it cannot write to a log file, just
7438 as it does if it cannot open a log file.
7440 39. Modified OS/Makefile-Linux so that it now contains
7442 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7444 The two -D definitions ensure that Exim is compiled with large file
7445 support, which makes it possible to handle log files that are bigger than
7448 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7449 instance) a domain was checked against a named list that involved a lookup,
7450 causing $domain_data to be set, then another domain was checked against the
7451 same list, then the first domain was re-checked, the value of $domain_data
7452 after the final check could be wrong. In particular, if the second check
7453 failed, it could be set empty. This bug probably also applied to
7456 41. The strip_trailing_dot option was not being applied to the address given
7457 with the -f command-line option.
7459 42. The code for reading a message's header from the spool was incrementing
7460 $received_count, but never initializing it. This meant that the value was
7461 incorrect (doubled) while delivering a message in the same process in which
7462 it was received. In the most common configuration of Exim, this never
7463 happens - a fresh exec is done - but it can happen when
7464 deliver_drop_privilege is set.
7466 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7467 it now includes up to 150 characters of the unexpected data in the log
7470 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7471 and building data strings. The size of both of these buffers was 10 000
7472 bytes - far larger than anybody would *ever* want, thought I. Needless to
7473 say, somebody hit the limit. I have increased the maximum line length to
7474 20 000 and the maximum data length of concatenated lines to 100 000. I have
7475 also fixed two bugs, because there was no checking on these buffers. Tsk,
7476 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7477 buffer is too small.
7479 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7480 lsearch lookups. Now it does.
7482 46. When parsing a route_list item in a manualroute router, a fixed-length
7483 buffer was used for the list of hosts. I made this 1024 bytes long,
7484 thinking that nobody would ever have a list of hosts that long. Wrong.
7485 Somebody had a whole pile of complicated expansion conditions, and the
7486 string was silently truncated, leading to an expansion error. It turns out
7487 that it is easier to change to an unlimited length (owing to other changes
7488 that have happened since this code was originally written) than to build
7489 structure for giving a limitation error. The length of the item that
7490 expands into the list of hosts is now unlimited.
7492 47. The lsearch lookup could not handle data where the length of text line was
7493 more than 4095 characters. Such lines were truncated, leading to shortened
7494 data being returned. It should now handle lines of any length.
7496 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7497 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7500 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7501 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7502 no longer happens in comment lines. A list of replacements is now placed
7503 at the head of all of the source files, except those whose only change is
7504 to replace PERL_COMMAND in the very first #! line.
7506 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7507 messages on the queue, with a bottom-up merge sort, using code contributed
7508 by Michael Haardt. This should make operations like -bp somewhat faster on
7509 large queues. It won't affect queue runners, except when queue_run_in_order
7512 51. Installed eximstats 1.31 in the distribution.
7514 52. Added support for SRV lookups to the dnslookup router.
7516 53. If an ACL referred to $message_body or $message_body_end, the value was not
7517 reset for any messages that followed in the same SMTP session.
7519 54. The store-handling optimization for building very long strings was not
7520 differentiating between the different store pools. I don't think this
7521 actually made any difference in practice, but I've tidied it.
7523 55. While running the routers to verify a sender address, $sender_address
7524 was still set to the sender address. This is wrong, because when routing to
7525 send a bounce to the sender, it would be empty. Therefore, I have changed
7526 it so that, while verifying a sender address, $sender_address is set to <>.
7527 (There is no change to what happens when verifying a recipient address.)
7529 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7530 target A or AAAA records (if not already returned) without resetting the
7531 qualify_single or search_parents options of the DNS resolver. These are
7532 inappropriate in this case because the targets of MX and SRV records must
7533 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7534 target that, when qualified, matched something in the local domain. These
7535 two options are now turned off when doing these lookups.
7537 57. It seems that at least some releases of Reiserfs (which does not have the
7538 concept of a fixed number of inodes) returns zero and not -1 for the
7539 number of available inodes. This interacted badly with check_spool_inodes,
7540 which assumed that -1 was the "no such thing" setting. What I have done is
7541 to check that the total number of inodes is greater than zero before doing
7542 the test of how many are available.
7544 58. When a "warn" ACL statement has a log_message modifier, the message is
7545 remembered, and not repeated. This is to avoid a lot of repetition when a
7546 message has many recipients that cause the same warning to be written.
7547 However, Exim was preserving the list of already written lines for an
7548 entire SMTP session, which doesn't seem right. The memory is now reset if a
7549 new message is started.
7551 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7552 debugging output unless log_rewrite was also set.
7554 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7555 of (int)(handle) when we know that handle contains (void *)(-1).
7557 61. The Exim daemon panic-logs an error return when it closes the incoming
7558 connection. However "connection reset by peer" seems to be common, and
7559 isn't really an error worthy of noting specially, so that particular error
7562 62. When Exim is trying to find all the local interfaces, it used to panic and
7563 die if the ioctl to get the interface flags failed. However, it seems that
7564 on at least one OS (Solaris 9) it is possible to have an interface that is
7565 included in the list of interfaces, but for which you get a failure error
7566 for this call. This happens when the interface is not "plumbed" into a
7567 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7568 failure of the "get flags" call assumes that the interface is down.
7570 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7571 makes life easier for people who are doing arithmetic on fields extracted
7572 from dates, where you often get leading zeros that should not be
7573 interpreted as octal.
7575 64. Added qualify_domain to the redirect router, to override the global
7578 65. If a pathologically long header line contained very many addresses (the
7579 report of this problem mentioned 10 000) and each of them was rewritten,
7580 Exim could use up a very large amount of memory. (It kept on making new
7581 copies of the header line as it rewrote, and never released the old ones.)
7582 At the expense of a bit more processing, the header rewriting function has
7583 been changed so that it no longer eats memory in this way.
7585 66. The generation of the Received: header has been moved from the time that a
7586 message starts to be received, to the time that it finishes. The timestamp
7587 in the Received: header should now be very close to that of the <= log
7588 line. There are two side-effects of this change:
7590 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7591 logged header lines no longer include the local Received: line, because
7592 it has not yet been created. The same applies to a copy of the message
7593 that is returned to a non-SMTP sender when a message is rejected.
7595 (b) When a filter file is tested using -bf, no additional Received: header
7596 is added to the test message. After some thought, I decided that this
7599 This change does not affect the value of $received_for. It is still set
7600 after address rewriting, but before local_scan() is called.
7602 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7604 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7605 gave an unhelpful panic error message, and a defer error. I have managed to
7606 change this behaviour so that it now rejects any supplied certificate,
7607 which seems right, as the list of acceptable certificates is empty.
7609 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7610 gave an unhelpful defer error. I have not managed to make this reject any
7611 supplied certificates, but the error message it gives is "no certificate
7612 supplied", which is not helpful.
7614 70. exigrep's output now also includes lines that are not associated with any
7615 message, but which match the given pattern. Implemented by a patch from
7616 Martin Sluka, which also tidied up the Perl a bit.
7618 71. Recipient callout verification, like sender verification, was using <> in
7619 the MAIL FROM command. This isn't really the right thing, since the actual
7620 sender may affect whether the remote host accepts the recipient or not. I
7621 have changed it to use the actual sender in the callout; this means that
7622 the cache record is now keyed on a recipient/sender pair, not just the
7623 recipient address. There doesn't seem to be a real danger of callout loops,
7624 since a callout by the remote host to check the sender would use <>.
7625 [SEE ABOVE: changed after hitting problems.]
7627 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7628 temporary errors. However, in the case of such a code being given after
7629 the end of a data transmission (i.e. after ".") Exim was failing to write
7630 a retry record for the message. (Yes, there was some broken host that was
7631 actually sending 8xx at this point.)
7633 73. An unknown lookup type in a host list could cause Exim to panic-die when
7634 the list was checked. (An example that provoked this was putting <; in the
7635 middle of a list instead of at the start.) If this happened during a DATA
7636 ACL check, a -D file could be left lying around. This kind of configuration
7637 error no longer causes Exim to die; instead it causes a defer error. The
7638 incident is still logged to the main and panic logs.
7640 74. Buglet left over from Exim 3 conversion. The message "too many messages
7641 in one connection" was written to the rejectlog but not the mainlog, except
7642 when address rewriting (yes!) was being logged.
7644 75. Added write_rejectlog option.
7646 76. When a system filter was run not as root (that is, when system_filter_user
7647 was set), the values of the $n variables were not being returned to the
7648 main process; thus, they were not subsequently available in the $sn
7651 77. Added +return_path_on_delivery log selector.
7653 78. A connection timeout was being treated differently from recipients deferred
7654 when testing hosts_max_try with a message that was older than the host's
7655 retry timeout. (The host should not be counted, thus allowing all hosts to
7656 be tried at least once before bouncing.) This may have been the cause of an
7657 occasionally reported bug whereby a message would remain on the queue
7658 longer than the retry timeout, but would be bounced if a delivery was
7659 forced. I say "may" because I never totally pinned down the problem;
7660 setting up timeout/retry tests is difficult. See also the next item.
7662 79. The ultimate address timeout was not being applied to errors that involved
7663 a combination of host plus message (for example, a timeout on a MAIL
7664 command). When an address resolved to a number of possible hosts, and they
7665 were not all tried for each delivery (e.g. because of hosts_max_try), a
7666 message could remain on the queue longer than the retry timeout.
7668 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7671 81. Fixed an obscure SMTP outgoing bug which required at least the following
7672 conditions: (a) there was another message waiting for the same server;
7673 (b) the server returned 5xx to all RCPT commands in the first message so
7674 that the message was not completed; (c) the server dropped the connection
7675 or gave a negative response to the RSET that Exim sends to abort the
7676 transaction. The observed case was a dropped connection after DATA that had
7677 been sent in pipelining mode. That is, the server had advertised PIPELINING
7678 but was not implementing it correctly. The effect of the bug was incorrect
7679 behaviour, such as trying another host, and this could lead to a crash.
7685 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7686 and daemon.c were passed as pointers to ints; they should have been
7687 pointers to socklen_t variables (which are typically unsigned ints).
7689 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7692 3. Fixed a really odd bug that affected only the testing scheme; patching a
7693 certain fixed string in the binary changed the value of another string that
7694 happened to be identical to the end of the original first string.
7696 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7697 name", it returns that address as the IP address. On some operating
7698 systems (e.g. Solaris), it also passes back the IP address string as the
7699 "host name". However, on others (e.g. Linux), it passes back an empty
7700 string. Exim wasn't checking for this, and was changing the host name to an
7701 empty string, assuming it had been canonicalized.
7703 5. Although rare, it is permitted to have more than one PTR record for a given
7704 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7705 all the names associated with an address, because they do in Solaris.
7706 However, it seems that they do not in Linux for data that comes from the
7707 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7708 I found this out when I moved to a new Linux workstation and tried to run
7709 the Exim test suite.
7711 To get round this problem I have changed the code so that it now does its
7712 own call to the DNS to look up PTR records when searching for a host name.
7713 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7714 addresses that are only in /etc/hosts are still found.
7716 This behaviour is, however, controlled by an option called host_lookup_
7717 order, which defaults to "bydns:byaddr". If people want to use the other
7718 order, or indeed, just use one or the other means of lookup, they can
7719 specify it in this variable.
7721 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7722 some operating systems, this comes back from gethostbyaddr() as an empty
7723 string, and this is what Exim used to test for. However, it seems that in
7724 other systems, "." is yielded. Exim now tests for this case too.
7726 7. The values of check_spool_space and check_log_space are now held internally
7727 as a number of kilobytes instead of an absolute number of bytes. If a
7728 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7729 kilobyte. This means that much larger values can be stored.
7731 8. Exim monitor: an attempt to get the action menu when not actually pointing
7732 at a message produces an empty menu entitled "No message selected". This
7733 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7734 no entries in it ("Shell widget menu has zero width and/or height"). So I
7735 have added a single, blank menu entry in this case.
7737 9. Added ${quote_local_part.
7739 10. MIME decoding is now applied to the contents of Subject: header lines when
7742 11. Now that a reference to $sender_host_address automatically causes a reverse
7743 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7744 host lookup before query-style lookups in lists that might use this
7745 variable. This has therefore been abolished, and the "net-" prefix is no
7746 longer necessary for query-style lookups.
7748 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7749 have been a typo for LFLAGS, so it has been changed.
7751 13. The install script calls Exim with "-C /dev/null" in order to find the
7752 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7753 to be output. However, since Exim outputs its version number before the
7754 error, it didn't break the script. It just looked ugly. I fixed this by
7755 always allowing "-C /dev/null" if the caller is root.
7757 14. Ignore overlarge ACL variable number when reading spool file - insurance
7758 against a later release with more variables having written the file.
7760 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7761 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7762 this, as well as the form without the "IPv6" on the front.
7764 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7765 OS/Makefile-Darwin file.
7767 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7768 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7770 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7771 variable. However, some OS use ints for inodes. I've added cast to long int
7772 to get rid of the compiler warning.
7774 19. I had forgotten to lock out "/../" in configuration file names when
7775 ALT_CONFIG_PREFIX was set.
7777 20. Routers used for verification do not need to specify transports. However,
7778 if such a router generated a host list, and callout was configured, Exim
7779 crashed, because it could not find a port number from the (non-existent)
7780 transport. It now assumes port 25 in this circumstance.
7782 21. Added the -t option to exigrep.
7784 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7785 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7786 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7787 them is set, LOOKUP_LSEARCH is forced.
7789 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7791 24. Added sender and host information to the "rejected by local_scan()" log
7792 line; previously there was no indication of these.
7794 25. Added .include_if_exists.
7796 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7797 Linux. It turns out that not all file systems support this. Apparently some
7798 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7799 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7800 not supported on the file descriptor, is now ignored when Exim is trying to
7801 sync a directory. This applies only to Linux.
7803 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7805 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7806 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7809 29. Added support for maildirsize files from supplied patch (modified a bit).
7811 30. The use of :fail: followed by an empty string could lead Exim to respond to
7812 sender verification failures with (e.g.):
7814 550 Verification failed for <xxx>
7815 550 Sender verify failed
7817 where the first response line was missing the '-' that indicates it is not
7818 the final line of the response.
7820 31. The loop for finding the name of the user that called Exim had a hardwired
7821 limit of 10; it now uses the value of finduser_retries, which is used for
7822 all other user lookups.
7824 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7827 33. Exim was neglecting to zero errno before one call of strtol() when
7828 expanding a string and expecting an integer value. On some systems this
7829 resulted in spurious "integer overflow" errors. Also, it was casting the
7830 result into an int without checking.
7832 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7833 did not work. The code looks as if it has *never* worked, though it appears
7834 to have been documented since at least release 1.62. I have made it work.
7836 35. The "timeout_DNS" error in retry rules, also documented since at least
7837 1.62, also never worked. As it isn't clear exactly what this means, and
7838 clearly it isn't a major issue, I have abolished the feature by treating it
7839 as "timeout", and writing a warning to the main and panic logs.
7841 36. The display of retry rules for -brt wasn't always showing the error code
7844 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7845 timeout_connect_A, timeout_connect_MX.
7847 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7848 to the empty sender.
7850 39. The daemon was not analysing the content of -oX till after it had closed
7851 stderr and disconnected from the controlling terminal. This meant that any
7852 syntax errors were only noted on the panic log, and the return code from
7853 the command was 0. By re-arranging the code a little, I've made the
7854 decoding happen first, so such errors now appear on stderr, and the return
7855 code is 1. However, the actual setting up of the sockets still happens in
7856 the disconnected process, so errors there are still only recorded on the
7859 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7860 connections (as happens on some IP stacks) was logged at start up time as
7861 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7862 it from "IPv6 and IPv4", which means that two separate sockets are being
7865 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7866 says whether AF_INET or AF_INET6 was passed as an argument.
7868 42. Exiwhat output was messed up when time zones were included in log
7871 43. Exiwhat now gives more information about the daemon's listening ports,
7872 and whether -tls-on-connect was used.
7874 44. The "port" option of the smtp transport is now expanded.
7876 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7877 silently ignored. Now an error message is written to the main and panic
7880 46. There's a new ACL modifier called "logwrite" which writes to a log file
7881 as soon as it is encountered.
7883 47. Added $local_user_uid and $local_user_gid at routing time.
7885 48. Exim crashed when trying to verify a sender address that was being
7888 49. Exim was recognizing only a space character after ".include". It now also
7889 recognizes a tab character.
7891 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7892 extracting the relevant information from the specification. The man page no
7893 longer contains scrambled data for the -d option, and I've added a section
7894 at the front about calling Exim under different names.
7896 51. Added "extra_headers" argument to the "mail" command in filter files.
7898 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7901 53. Installed eximstats 1.29.
7903 54. Added transport_filter_timeout as a generic transport option.
7905 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7906 Cc: header lines. This was required by RFC 822, but it not required by RFC
7909 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7910 incoming messages that did not have them. Now it does so only if the
7911 message originates locally, that is, if there is no associated remote host
7912 address. When Resent- header lines are present, this applies to the Resent-
7913 lines rather than the non-Resent- lines.
7915 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7916 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7918 58. Messages for configuration errors now include the name of the main
7919 configuration file - useful now that there may be more than one file in a
7920 list (.included file names were always shown).
7922 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7923 for those rare installations that do not start the daemon as root or run it
7924 setuid root. I've cut out the call to initgroups() if the daemon is not
7927 60. The Exim user and group can now be bound into the binary as text strings
7928 that are looked up at the start of Exim's processing.
7930 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7932 62. Added $mailstore_basename variable.
7934 63. Installed patch to sieve.c from Michael Haardt.
7936 64. When Exim failed to open the panic log after failing to open the main log,
7937 the original message it was trying to log was written to stderr and debug
7938 output, but if they were not available (the usual case in production), it
7939 was lost. Now it is written to syslog before the two lines that record the
7940 failures to open the logs.
7942 65. Users' Exim filters run in subprocesses under the user's uid. It is
7943 possible for a "deliver" command or an alias in a "personal" command to
7944 provoke an address rewrite. If logging of address rewriting is configured,
7945 this fails because the process is not running as root or exim. There may be
7946 a better way of dealing with this, but for the moment (because 4.30 needs
7947 to be released), I have disabled address rewrite logging when running a
7948 filter in a non-root, non-exim process.
7954 1. The buildconfig auxiliary program wasn't quoting the value set for
7955 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7956 not defined. This bug was masked in 4.22 by the effect that was fixed in
7959 2. Some messages that were rejected after a message id was allocated were
7960 shown as "incomplete" by exigrep. It no longer does this for messages that
7961 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7963 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7964 have allow_domain_literals set, the ID did not get logged in the <= line.
7965 Domain literals are now always recognized in Message-ID: header lines.
7967 4. The first argument for a ${extract expansion item is the key name or field
7968 number. Leading and trailing spaces in this item were not being ignored,
7969 causing some misleading effects.
7971 5. When deliver_drop_privilege was set, single queue runner processes started
7972 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7973 same command in the process it spins off) were not dropping privilege.
7975 6. When the daemon running as "exim" started a queue runner, it always
7976 re-executed Exim in the spun-off process. This is a waste of effort when
7977 deliver_drop_privilege is set. The new process now just calls the
7978 queue-runner function directly.
7984 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7987 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7988 ignored. Though the use of -oP was forcing the writing of a pid file, it
7989 was always written to the default place.
7991 3. If the message "no IP address found for host xxxx" is generated during
7992 incoming verification, it is now followed by identification of the incoming
7993 connection (so you can more easily find what provoked it).
7995 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7997 5. Added some features to "harden" Exim a bit more against certain attacks:
7999 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8000 be put in Local/Makefile. This is like the never_users runtime option,
8001 but it cannot be overridden. The default setting is "root".
8003 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8004 prefix string with which any file named in a -C command line option
8007 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8008 is retained for -C and -D only if the caller of Exim is root. Without
8009 it, the exim user may also use -C and -D and retain privilege.
8011 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8012 command line option is disabled.
8014 6. Macro names set by the -D option must start with an upper case letter, just
8015 like macro names defined in the configuration file.
8017 7. Added "dereference=" facility to LDAP.
8019 8. Two instances of the typo "uknown" in the source files are fixed.
8021 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8022 the Configure-Makefile script screwed up while processing it.
8024 10. Incorporated PCRE 4.4.
8026 11. The SMTP synchronization check was not operating right at the start of an
8027 SMTP session. For example, it could not catch a HELO sent before the client
8028 waited for the greeting. There is now a check for outstanding input at the
8029 point when the greeting is written. Because of the duplex, asynchronous
8030 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8031 way, but not yet received, when the check is performed.
8033 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8034 on TCP/IP sockets, because this apparently causes some broken clients to
8037 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8038 unchanged) from the Cygwin maintainer.
8040 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8041 was supported instead of testing for mbx. Effectively a typo.
8043 15. The spa authenticator server code was not checking that the input it
8044 received was valid base64.
8046 16. The debug output line for the "set" modifier in ACLs was not showing the
8047 name of the variable that was being set.
8049 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8051 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8052 lookup on its own. Something else was needed to trigger the lookup. For
8053 example, a match in host_lookup or the need for a host name in a host list.
8054 Now, if $sender_host_name is referenced and the host name has not yet been
8055 looked up, a lookup is performed. If the lookup fails, the variable remains
8056 empty, and $host_lookup_failed is set to "1".
8058 19. Added "eqi" as a case-independent comparison operator.
8060 20. The saslauthd authentication condition could segfault if neither service
8061 nor realm was specified.
8063 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8064 error message that was logged was misleading, and incoming SMTP
8065 connections were dropped. The message is now more accurate, and temporary
8066 errors are given to SMTP connections.
8068 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8069 (except RSET) with a 421 error, until QUIT is received. However, it was
8070 failing to send a response to QUIT.
8072 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8073 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8074 ACL is now run after the helo_try_verify_hosts code.
8076 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8077 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8078 case-independent, so other case variants are also recognized). Apparently
8079 some systems use these upper case variants.
8081 25. If more than two messages were waiting for the same host, and a transport
8082 filter was specified for the transport, Exim sent two messages over the
8083 same TCP/IP connection, and then failed with "socket operation on non-
8084 socket" when it tried to send the third.
8086 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8088 27. The extern definition of crypt16() in expand.c was not being excluded when
8089 the OS had its own crypt16() function.
8091 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8092 as a preferred synonym for return_size_limit, both as an option and as an
8095 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8097 30. Changed the default configuration ACL to relax the local part checking rule
8098 for addresses that are not in any local domains. For these addresses,
8099 slashes and pipe symbols are allowed within local parts, but the sequence
8100 /../ is explicitly forbidden.
8102 31. SPA server authentication was not clearing the challenge buffer before
8105 32. log_message in a "warn" ACL statement was writing to the reject log as
8106 well as to the main log, which contradicts the documentation and doesn't
8107 seem right (because no rejection is happening). So I have stopped it.
8109 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8110 However, I am unable to do any testing of this.
8112 34. Fixed an infelicity in the appendfile transport. When checking directories
8113 for a mailbox, to see if any needed to be created, it was accidentally
8114 using path names with one or more superfluous leading slashes; tracing
8115 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8117 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8118 contents are added to the log line that is written for every discarded
8119 recipient. (Previously a log_message setting was ignored.)
8121 36. The ${quote: operator now quotes the string if it is empty.
8123 37. The install script runs exim in order to find its version number. If for
8124 some reason other than non-existence or emptiness, which it checks, it
8125 could not run './exim', it was installing it with an empty version number,
8126 i.e. as "exim-". This error state is now caught, and the installation is
8129 38. An argument was missing from the function that creates an error message
8130 when Exim fails to connect to the socket for saslauthd authentication.
8131 This could cause Exim to crash, or give a corrupted message.
8133 39. Added isip, isip4, and isip6 to ${if conditions.
8135 40. The ACL variables $acl_xx are now saved with the message, and can be
8136 accessed later in routers, transports, and filters.
8138 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8139 strings in the file are not string-expanded.
8141 42. If a MAIL command specified a SIZE value that was too large to fit into an
8142 int variable, the check against message_size_limit failed. Such values are
8143 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8144 day this will have to be increased, but I don't think I want to be around
8145 when emails are that large.
8152 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8153 iconv() is not standard in FreeBSD.
8155 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8156 IPv6 enabled. The observed symptom was a segmentation fault on return from
8157 the function os_common_find_running_interfaces() in src/os.c.
8159 3. In the check_special_case() function in daemon.c I had used "errno" as an
8160 argument name, which causes warnings on some systems. This was basically a
8161 typo, since it was named "eno" in the comments!
8163 4. The code that waits for the clock to tick (at a resolution of some fraction
8164 of a second) so as to ensure message-id uniqueness was always waiting for
8165 at least one whole tick, when it could have waited for less. [This is
8166 almost certainly not relevant at current processor speeds, where it is
8167 unlikely to ever wait at all. But we try to future-proof.]
8169 5. The function that sleeps for a time interval that includes fractions of a
8170 second contained a race. It did not block SIGALRM between setting the
8171 timer, and suspending (a couple of lines later). If the interval was short
8172 and the sigsuspend() was delayed until after it had expired, the suspension
8173 never ended. On busy systems this could lead to processes getting stuck for
8176 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8177 process, before it forks any delivery processes. The open lookup caching
8178 mechanism meant that the open file or database connection was passed into
8179 the delivery process. The problem was that delivery processes always tidy
8180 up cached lookup data. This could cause a problem for the next delivery
8181 process started by the queue runner, because the external queue runner
8182 process does not know about the closure. So the next delivery process
8183 still has data in the lookup cache. In the case of a file lookup, there was
8184 no problem because closing a file descriptor in a subprocess doesn't affect
8185 the parent. However, if the lookup was caching a connection to a database,
8186 the connection was closed, and the second delivery process was likely to
8187 see errors such as "PGSQL: query failed: server closed the connection
8188 unexpectedly". The problem has been fixed by closing all cached lookups
8189 in a queue runner before running a delivery process.
8191 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8192 seem to have the "const" qualifier which it has on other OS. I've
8195 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8196 *of the same type* (client or server) with the same public name that an
8197 error should be diagnosed.
8199 9. When Exim looked up a host name for an IP address, but failed to find the
8200 original IP address when looking up the host name (a safety check), it
8201 output the message "<ip address> does not match any IP for NULL", which was
8202 confusing, to say the least. The bug was that the host name should have
8203 appeared instead of "NULL".
8205 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8206 user that is built into the binary, and the -C or -D options is used, root
8207 privilege is dropped before the configuration file is read. In addition,
8208 logging is switched to stderr instead of the normal log files. If the
8209 configuration then re-defines the Exim user, the unprivileged environment
8210 is probably not what is expected, so Exim logs a panic warning message (but
8213 However, if deliver_drop_privilege is set, the unprivileged state may well
8214 be exactly what is intended, so the warning has been cut out in that case,
8215 and Exim is allowed to try to write to its normal log files.
8221 1. smtp_return_error_details was not giving details for temporary sender
8222 or receiver verification errors.
8224 2. Diagnose a configuration error if two authenticators have the same public
8227 3. Exim used not to create the message log file for a message until the first
8228 delivery attempt. This could be confusing when incoming messages were held
8229 for policy or load reasons. The message log file is now created at the time
8230 the message is received, and an initial "Received" line is written to it.
8232 4. The automatically generated man page for command line options had a minor
8233 bug that caused no ill effects; however, a more serious problem was that
8234 the procedure for building the man page automatically didn't always
8235 operate. Consequently, release 4.20 contains an out-of-date version. This
8236 shouldn't happen again.
8238 5. When building Exim with embedded Perl support, the script that builds the
8239 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8240 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8242 6. The freeze_tell option was not being used for messages that were frozen on
8243 arrival, either by an ACL or by local_scan().
8245 7. Added the smtp_incomplete_transaction log selector.
8247 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8248 was accepting AUTH without a new EHLO.
8250 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8251 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8254 10. Logging of TCP/IP connections (when configured) now happens in the main
8255 daemon process instead of the child process, so that the TCP/IP connection
8256 count is more accurate (but it can never be perfect).
8258 11. The use of "drop" in a nested ACL was not being handled correctly in the
8259 outer ACL. Now, if condition failure induced by the nested "drop" causes
8260 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8261 or "require"), the connection is dropped.
8263 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8264 that yield "discard" can now be used with an "accept" or a "discard" verb,
8265 but an error is generated for any others (because I can't see a useful way
8266 to define what should happen).
8268 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8269 are now processed in the same way as lines in the Exim configuration file.
8270 In particular, continuation lines are supported.
8272 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8274 15. Added -ti meaning -t -i.
8276 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8277 domains, and warn by logging if others are found.
8279 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8280 returned from GIFCONF when Exim is trying to find the list of interfaces on
8281 a host. The code in os.c has been modified to copy each ifreq to an aligned
8282 structure in all cases.
8284 Also, in some cases, the returned ifreq's were being copied to a 'struct
8285 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8286 means the last couple of bytes of an IPv6 address could be chopped if the
8287 ifreq contained only a normal sockaddr (14 bytes storage).
8289 18. Named domain lists were not supported in the hosts_treat_as_local option.
8290 An entry such as +xxxx was not recognized, and was treated as a literal
8293 19. Ensure that header lines added by a DATA ACL are included in the reject log
8294 if the ACL subsequently rejects the message.
8296 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8297 MD5 (which is deprecated).
8299 21. When testing a filter file using -bf, Exim was writing a message when it
8300 took the sender from a "From " line in the message, but it was not doing so
8301 when it took $return_path from a Return-Path: header line. It now does.
8303 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8304 with a valid header line field name (a series of printing characters
8305 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8307 23. Changed "disc" in the source to "disk" to conform to the documentation and
8308 the book and for uniformity.
8310 24. Ignore Sendmail's -Ooption=value command line item.
8312 25. When execve() failed while trying to run a command in a pipe transport,
8313 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8314 could be confused with a return value of 69 from the command itself. This
8315 has been changed to 127, the value the shell returns if it is asked to run
8316 a non-existent command. The wording for the related log line suggests a
8317 non-existent command as the problem.
8319 26. If received_header_text expands to an empty string, do not add a Received:
8320 header line to the message. (Well, it adds a token one on the spool, but
8321 marks it "old" so that it doesn't get used or transmitted.)
8323 27. Installed eximstats 1.28 (addition of -nt option).
8325 28. There was no check for failure on the call to getsockname() in the daemon
8326 code. This can fail if there is a shortage of resources on the system, with
8327 ENOMEM, for example. A temporary error is now given on failure.
8329 29. Contrary to the C standard, it seems that in some environments, the
8330 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8331 program. Exim now does this explicitly; it affects the formatting of
8332 timestamps using strftime().
8334 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8335 complaints. I've now initialized all the variables, to avoid this.
8337 32. Header lines added by a system filter were not being "seen" during
8338 transport-time rewrites.
8340 33. The info_callback() function passed to OpenSSL is set up with type void
8341 (*)(SSL *, int, int), as described somewhere. However, when calling the
8342 function (actually a macro) that sets it up, the type void(*)() is
8343 expected. I've put in a cast to prevent warnings from picky compilers.
8345 34. If a DNS black list lookup found a CNAME record, but there were no A
8346 records associated with the domain it pointed at, Exim crashed.
8348 35. If a DNS black list lookup returned more than one A record, Exim ignored
8349 all but the first. It now scans all returned addresses if a particular IP
8350 value is being sought. In this situation, the contents of the
8351 $dnslist_value variable are a list of all the addresses, separated by a
8354 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8355 to accept a host name and all its aliases if the forward lookup for any of
8356 them yielded the IP address of the incoming connection. Now it accepts only
8357 those names whose forward lookup yields the correct IP address. Any other
8358 names are discarded. This closes a loophole whereby a rogue DNS
8359 administrator could create reverse DNS records to break through a
8360 wildcarded host restriction in an ACL.
8362 37. If a user filter or a system filter that ran in a subprocess used any of
8363 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8364 the wrong values were passed to the pipe command ($thisaddress had the
8365 value of $0, $0 had the value of $1, etc). This bug was introduced by
8366 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8368 38. Improved the line breaking for long SMTP error messages from ACLs.
8369 Previously, if there was no break point between 40 and 75 characters, Exim
8370 left the rest of the message alone. Two changes have been made: (a) I've
8371 reduced the minimum length to 35 characters; (b) if it can't find a break
8372 point between 35 and 75 characters, it looks ahead and uses the first one
8373 that it finds. This may give the occasional overlong line, but at least the
8374 remaining text gets split now.
8376 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8377 file descriptors might be low, and that setting 1000 would always raise it.
8378 It turns out that in some environments, the limit is already over 1000 and
8379 that lowering it causes trouble. So now Exim takes care not to decrease it.
8381 40. When delivering a message, the value of $return_path is set to $sender_
8382 address at the start of routing (routers may change the value). By an
8383 oversight, this default was not being set up when an address was tested by
8384 -bt or -bv, which affected the outcome if any router or filter referred to
8387 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8388 log only when "message" or "log_message" are set. However, if one of the
8389 conditions was an address verification, or a call to a nested ACL, the
8390 messages generated by the underlying test were being passed through. This
8391 no longer happens. The underlying message is available in $acl_verify_
8392 message for both "message" and "log_message" expansions, so it can be
8393 passed through if needed.
8395 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8396 new expansion $bh_ to give the encoded byte string without charset
8397 translation. Translation happens only if iconv() is available; HAVE_ICONV
8398 indicates this at build time. HEADERS_CHARSET gives the charset to
8399 translate to; headers_charset can change it in the configuration, and
8400 "headers charset" can change it in an individual filter file.
8402 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8403 that creates RFC 2047 encoded "words" labels them as that charset instead
8404 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8405 expansion operator; (ii) when Exim creates a From: line for a local
8406 message; (iii) when a header line is rewritten to include a "phrase" part.
8408 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8409 buggy, causing it to skip the first lines of messages whose message ID
8410 ended in 'D'. This would not have bitten before Exim release 4.14, because
8411 message IDs were unlikely to end in 'D' before then. The effect was to have
8412 incorrect size information for certain domains.
8414 45. #include "config.h" was missing at the start of the crypt16.c module. This
8415 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8418 46. If there was a timeout during a "random" callout check, Exim treated it as
8419 a failure of the random address, and carried on sending RSET and the real
8420 address. If the delay was just some slowness somewhere, the response to the
8421 original RCPT would be taken as a response to RSET and so on, causing
8422 mayhem of various kinds.
8424 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8425 when I implemented it. It didn't allow for the fact that some option values
8426 may legitimately be negative (e.g. size_addition), and it didn't even do
8427 the right test for positive values.
8429 48. Domain names in DNS records are case-independent. Exim always looks them up
8430 in lower case. Some resolvers return domain names in exactly the case they
8431 appear in the zone file, that is, they may contain uppercase letters. Not
8432 all resolvers do this - some return always lower case. Exim was treating a
8433 change of case by a resolver as a change of domain, similar to a widening
8434 of a domain abbreviation. This triggered its re-routing code and so it was
8435 trying to route what was effectively the same domain again. This normally
8436 caused routing to fail (because the router wouldn't handle the domain
8437 twice). Now Exim checks for this case specially, and just changes the
8438 casing of the domain that it ultimately uses when it transmits the message
8441 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8444 50. If a filter generated a file delivery with a non-absolute name (possible if
8445 no home directory exists for the router), the forbid_file option was not
8448 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8449 the existing equality matching.
8451 52. Exim was using ints instead of ino_t variables in some places where it was
8452 dealing with inode numbers.
8454 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8455 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8456 called TMPDIR, and if it finds it is different, it changes its value.
8458 54. The smtp_printf() function is now made available to local_scan() so
8459 additional output lines can be written before returning. There is also an
8460 smtp_fflush() function to enable the detection of a dropped connection.
8461 The variables smtp_input and smtp_batched_input are exported to
8464 55. Changed the default runtime configuration: the message "Unknown user"
8465 has been removed from the ACL, and instead placed on the localuser router,
8466 using the cannot_route_message feature. This means that any verification
8467 failures that generate their own messages won't get overridden. Similarly,
8468 the "Unrouteable address" message that was in the ACL for unverifiable
8469 relay addresses has also been removed.
8471 56. Added hosts_avoid_esmtp to the smtp transport.
8473 57. The exicyclog script was not checking for the esoteric option
8474 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8475 will work only if exicyclog is run under the appropriate euid.
8477 58. Following a discussion on the list, the rules by which Exim recognises line
8478 endings on incoming messages have been changed. The -dropcr and drop_cr
8479 options are now no-ops, retained only for backwards compatibility. The
8480 following line terminators are recognized: LF CRLF CR. However, special
8481 processing applies to CR:
8483 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8484 nor a local message in the state where . is a terminator.
8486 (ii) If a bare CR is encountered in a header line, an extra space is added
8487 after the line terminator so as not to end the header. The reasoning
8488 behind this is that bare CRs in header lines are most likely either
8489 to be mistakes, or people trying to play silly games.
8491 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8492 was being incorrectly given as 18 bytes larger than it should have been.
8493 This is a VOB (very old bug).
8495 60. This may never have affected anything current, but just in case it has:
8496 When the local host is found other than at the start of a list of hosts,
8497 the local host, those with the same MX, and any that follow, are discarded.
8498 When the list in question was part of a longer list of hosts, the following
8499 hosts (not currently being processed) were also being discarded. This no
8500 longer happens. I'm not sure if this situation could ever has previously
8503 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8506 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8507 header. This is something that is recommended in a new Internet Draft, and
8508 is something that is documented as being done by Sendmail. There are two
8509 possible values. For messages generated by the autoreply transport, Exim
8512 Auto-Submitted: auto-replied
8514 whereas for all other generated messages (e.g. bounces) it adds
8516 Auto-Submitted: auto-generated
8518 63. The "personal" condition in filters now includes a test for the
8519 Auto-Submitted: header. If it contains the string "auto-" the message it
8520 not considered personal.
8522 64. Added rcpt_include_affixes as a generic transport option.
8524 65. Added queue_only_override (default true).
8526 66. Added the syslog_duplication option.
8528 67. If what should have been the first header line of a message consisted of
8529 a space followed by a colon, Exim was mis-interpreting it as a header line.
8530 It isn't of course - it is syntactically invalid and should therefore be
8531 treated as the start of the message body. The misbehaviour could have
8532 caused a number of strange effects, including loss of data in subsequent
8533 header lines, and spool format errors.
8535 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8536 client host had authenticated. This control can now be exercised by an ACL
8537 for more flexibility.
8539 69. By default, callouts do not happen when testing with -bh. There is now a
8540 variant, -bhc, which does actually run the callout code, including
8541 consulting and updating the callout cache.
8543 70. Added support for saslauthd authentication, courtesy of Alexander
8546 71. If statvfs() failed on the spool or log directories while checking their
8547 size for availability, Exim confusingly gave the error "space shortage".
8548 Furthermore, in debugging mode it crashed with a floating point exception.
8549 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8550 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8551 serious problem, Exim now writes to the main and panic logs when this
8552 happens, with details of the failure. It then refuses to accept the
8553 incoming message, giving the message "spool directory problem" or "log
8554 directory problem" with a 421 code for SMTP messages.
8556 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8557 0, 1, and 2 exist, because some OS complain for execs without them (see
8558 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8559 descriptors. However, the code omitted to check that the open succeeded,
8560 causing mysterious errors if for some reason the permissions on /dev/null
8561 got screwed. Now Exim writes a message to the main and panic logs, and
8562 bombs out if it can't open /dev/null.
8564 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8565 interact so that it is all more flexible. It is supposed to remain
8566 backwards compatible. Also added extra_local_interfaces.
8568 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8569 to bomb out with an assertion failure - to the client this appears as a
8570 connection drop. This problem occurs in the part of the code that was taken
8571 from the Samba project. Fortunately, the assertion is in a very simple
8572 function, so I have fixed this by reproducing the function inline in the
8573 one place where it is called, and arranging for authentication to fail
8574 instead of killing the process with assert().
8576 75. The SPA client code was not working when the server requested OEM rather
8577 than Unicode encoding.
8579 76. Added code to make require_files with a specific uid setting more usable in
8580 the case where statting the file as root fails - usually a non-root-mounted
8581 NFS file system. When this happens and the failure is EACCES, Exim now
8582 forks a subprocess and does the per-uid checking as the relevant uid.
8584 77. Added process_log_path.
8586 78. If log_file_path was not explicitly set, a setting of check_log_space or
8587 check_log_inodes was ignored.
8589 79. If a space check for the spool or log partitions fails, the incident is now
8590 logged. Of course, in the latter case the data may get lost...
8592 80. Added the %p formatting code to string_format() so that it can be used to
8593 print addresses in debug_print(). Adjusted all the address printing in the
8594 debugging in store.c to use %p rather than %d.
8596 81. There was a concern that a line of code in smtp_in.c could overflow a
8597 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8598 initially expressed, the concern was not well-founded, because trailing
8599 spaces are removed early. However, if the trailing spaces were followed by
8600 a NULL, they did not get removed, so the overflow was possible. Two fixes
8603 (a) I re-wrote the offending code in a cleaner fashion.
8604 (b) If an incoming SMTP command contains a NULL character, it is rejected
8607 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8608 runs initgroups(), so that if the Exim user is in any additional groups,
8609 they will be used during message reception.
8615 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob