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.
132 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
133 SMTP connection" log lines.
135 JH/02 Option default value updates:
136 - queue_fast_ramp (main) true (was false)
137 - remote_max_parallel (main) 4 (was 2)
139 JH/03 Cache static regex pattern compilations, for use by ACLs.
141 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
142 Make the rewrite never match and keep the logging. Trust the
143 admin to be using verify=header-syntax (to actually reject the message).
145 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
146 (under Linux) we watch the dir containing the final file; previously
147 it would be the dir with the first symlink. We still do not monitor
150 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
151 dn_expand() is happy to pass them through.
153 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
154 the file with the proof had an unchanged name, the new proof(s) were
155 loaded on top of the old ones (and nover used; the old ones were stapled).
157 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
158 more than one message arrived in a single connection a reference from
159 the earlier message could be re-used. Often a sigsegv resulted.
160 These variables were introduced in Exim 4.87.
161 Debug help from Graeme Fowler.
163 JH/09 Fix ${filter } for conditions that modify $value. Previously the
164 modified version would be used in construction the result, and a memory
167 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
168 Find and fix by Jasen Betts.
170 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
171 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
172 the systemwide configuration to override the Exim config.
174 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
175 API changes in libopendmarc.
177 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
178 pid 1, in the normal "background daemon" mode, having to drop process-
179 group leadership also lost track of needing to create listener sockets.
181 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
182 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
184 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
185 a capture group which obtained no text (eg. "(abc)*" matching zero
186 occurrences) could cause a segfault if the corresponding $<n> was
189 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
190 included a close-brace character (eg. it itself used an expansion) an
193 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
194 starting TLS. Previously it was after, meaning that attackers on such
195 ports had to be screened using the host_reject_connection main config
196 option. The new sequence aligns better with the STARTTLS behaviour, and
197 permits defences against crypto-processing load attacks, even though it
198 is strictly an incompatible change.
199 Also, avoid sending any SMTP fail response for either the connect ACL
200 or host_reject_connection, for TLS-on-connect ports.
202 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
203 Previously this was not permitted, but it makes reasonable sense.
204 While there, restore a restriction on using it from a connect ACL; given
205 the change JH/16 it could only return false (and before 4.91 was not
208 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
209 was exactly sized compared to the log buffer, a crash occurred with the
210 misleading message "bad memory reference; pool not found".
211 Found and traced by Jasen Betts.
213 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
214 dns_again_means_nonexist included an element causing a DNS lookup which
215 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
216 included (though probably not limited to) a process crash from stack
217 memory limit, or from excessive open files. Replace this with a paniclog
218 whine (as this is likely a configuration error), and returning
221 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
222 this always failed, probably leading to the usual downgrade to in-clear
225 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
226 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
227 to not checking that list for these lookups.
229 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
232 JH/23 Fix crash in string expansions. Previously, if an empty variable was
233 immediately followed by an expansion operator, a null-indirection read
234 was done, killing the process.
236 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
237 include an SMTP response string which is longer than that supported
238 by the delivering transport. Alleviate by wrapping such lines before
241 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
242 chars (RFC limit). Previously a limit of 12 items was made, which with
243 a not-impossible References: in the message being bounced could still
244 be over-large and get stopped in the transport.
246 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
247 close. Previously a bare socket close was done.
249 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
252 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
253 message being considered for delivery was already being handled by
254 another process, and having an SMTP connection already open, the function
255 to close it tried to use an uninitialized variable. This would afftect
256 high-volume sites more, especially when running mailing-list-style loads.
257 Pollution of logs was the major effect, as the other process delivered
258 the message. Found and partly investigated by Graeme Fowler.
260 JH/29 Change format of the internal ID used for message identification. The old
261 version only supported 31 bits for a PID element; the new 64 (on systems
262 which can use Base-62 encoding, which is all currently supported ones
263 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
264 and must use Base-36). The new ID is 23 characters rather than 16, and is
265 visible in various places - notably logs, message headers, and spool file
266 names. Various of the ancillary utilities also have to know the format.
267 As well as the expanded PID portion, the sub-second part of the time
268 recorded in the ID is expanded to support finer precision. Theoretically
269 this permits a receive rate from a single comms channel of better than the
271 The major timestamp part of the ID is not changed; at 6 characters it is
272 usable until about year 3700.
273 Updating from previously releases is fully supported: old-format spool
274 files are still usable, and the utilities support both formats. New
275 message will use the new format. The one hints-DB file type which uses
276 message-IDs (the transport wait- DB) will be discarded if an old-format ID
277 is seen; new ones will be built with only new-format IDs.
278 Optionally, a utility can be used to convert spool files from old to new,
279 but this is only an efficiency measure not a requirement for operation
280 Downgrading from new to old requires running a provided utility, having
281 first stopped all operations. This will convert any spool files from new
282 back to old (losing time-precision and PID information) and remove any
283 wait- hints databases.
285 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
286 we treated them as item separators when parsing for a list item, but they
287 need to be protected by the doublequotes. While there, add handling for
290 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
291 Found and fixed by Jasen Betts. No testcase for this as my usual text
292 editor insists on emitting only valid UTF-8.
294 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
295 a null-indirection SIGSEGV for the receive process.
297 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
298 -bh use. Internal checking would spot this and take a panic.
300 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
301 In 4.96 this would expand to empty.
303 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
304 certificate. Find and fix by Andreas Metzler.
306 JH/36 Add ARC info to DMARC hostory records.
308 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
309 or fakedefer. Previously the sender could discover that the message
310 had in fact been accepted.
312 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
313 sequences. Previously the input was not noted as being tainted; notably
314 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
315 bad coding of authenticators.
317 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
318 and ${tr...}. Found and diagnosed by Heiko Schlichting.
320 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
321 could be triggered by externally-supplied input. Found by Trend Micro.
324 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
325 be triggered by externally-controlled input. Found by Trend Micro.
328 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
329 be triggered by externally-controlled input. Found by Trend Micro.
332 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
333 Make the rewrite never match and keep the logging. Trust the
334 admin to be using verify=header-syntax (to actually reject the message).
336 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
339 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
345 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
346 after reception to before a subsequent reception. This should
347 mean slightly faster delivery, and also confirmation of reception
350 JH/02 Move from using the pcre library to pcre2. The former is no longer
351 being developed or supported (by the original developer).
353 JH/03 Constification work in the filters module required a major version
354 bump for the local-scan API. Specifically, the "headers_charset"
355 global which is visible via the API is now const and may therefore
356 not be modified by local-scan code.
358 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
359 sendfile() didi not account for the way the ClamAV driver code called it.
361 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
362 time check was being done for every character; replace that with one
365 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
366 sent was prefixed with a length byte.
368 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
369 RFC 5321. Previously Dovecot (at least) would log errors during
372 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
373 in 4.61 without a reason logged. Binaries will be bigger, which might
374 matter on diskspace-constrained systems, but debug is easier.
376 JH/09 Fix macro-definition during "-be" expansion testing. The move to
377 write-protected store for macros had not accounted for these runtime
378 additions; fix by removing this protection for "-be" mode.
380 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
381 handing out large-numbered file descriptors, violating the usual Unix
382 assumption (and required by Posix) that the lowest possible number will be
383 allocated by the kernel when a new one is needed. In the daemon, and any
384 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
385 are not useable for FD_SET() [and hence select()] and overwrite the stack.
386 Assorted crashes happen.
388 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
389 main-section options or in a connect ACL, the value from the first ever
390 connection was never replaced for subsequent connections. Found by
393 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
394 though only once PCRE2 was introduced: the memory accounting used under
395 debug offset allocations by an int, giving a hard trap in early startup.
396 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
398 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
399 with underbars is given. The write-protection of configuration introduced
400 in 4.95 trapped when normalisation was applied to an option not needing
403 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
405 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
406 failing to close the certificates directory, every hour or any time it
409 JH/16 Debugging initiated by an ACL control now continues through into routing
410 and transport processes. Previously debugging stopped any time Exim
411 re-execs, or for processing a queued message.
413 JH/17 The "expand" debug selector now gives more detail, specifically on the
414 result of expansion operators and items.
416 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
417 bad comparison between the option value and the name of the file to
418 be included was done, and a mismatch was wrongly identified.
419 4.88 to 4.95 are affected.
421 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
423 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
424 supplied as the db file-pair basename. Previously, if a directory
425 path was given, for example via the autoreply "once" option, the DB
426 file.pag and file.dir files would be created in that directory's
429 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
430 "taint" log_selector. These were previously deprecated.
432 JH/22 Fix static address-list lookups to properly return the matched item.
433 Previously only the domain part was returned.
435 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
436 the call into OpenSSL to send a TLS Close was being repeated; this
437 resulted in the library waiting for the peer's Close. If that was never
438 sent we waited forever. Fix by tracking send calls.
440 JH/24 The ${run} expansion item now expands its command string elements after
441 splitting. Previously it was before; the new ordering makes handling
442 zero-length arguments simpler. The old ordering can be obtained by
443 appending a new option "preexpand", after a comma, to the "run".
445 JH/25 Taint-check exec arguments for transport-initiated external processes.
446 Previously, tainted values could be used. This affects "pipe", "lmtp" and
447 "queryprogram" transport, transport-filter, and ETRN commands.
448 The ${run} expansion is also affected: in "preexpand" mode no part of
449 the command line may be tainted, in default mode the executable name
452 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
453 the the facility was not passed across execs, and only the first message
454 passed over a connection could use BDAT; any further ones using DATA.
456 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
457 uses $sending_ip_address and an interface is specified.
458 Previously any use of the local address in the EHLO name disabled
459 PIPECONNECT, the common case being to use the rDNS of it.
461 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
462 resumption. Previously verify failed because no certificate status is
463 passed on the wire for the restarted session. Fix by using the recorded
464 ocsp status of the stored session for the new connection.
466 JH/29 TLS resumption: the key for session lookup in the client now includes
467 more info that a server could potentially use in configuring a TLS
468 session, avoiding oferring mismatching sessions to such a server.
469 Previously only the server IP was used.
471 JH/30 Fix string_copyn() for limit greater than actual string length.
472 Previously the copied amount was the limit, which could result in a
473 overlapping memcpy for newly allocated destination soon after a
474 source string shorter than the limit. Found/investigated by KM.
476 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
477 close; it may be needed for a subsequent connection. This caused a
478 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
480 JH/32 Fix CHUNKING for a second message on a connection when the first was
481 rejected. Previously we did not reset the chunking-offered state, and
482 erroneously rejected the BDAT command. Investigation help from
485 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
486 an empty address. Previously the expansion returned an error.
488 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
489 proxy. Previously these were misparsed, leading to paniclog entries.
495 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
496 related applications. Previously an "H" was used where available info
497 says that "M" should be, so change to match.
499 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
500 as arguments, so an implementation trying to copy these into a local
501 buffer was taking a taint-enforcement trap. Fix by using dynamically
502 created buffers. Similar fix for radius expansion condition.
504 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
505 reasonable, eg. to count headers. Fix by using dynamically created
506 buffers rather than a local. Do similar fixes for ACL actions "dcc",
507 "log_reject_target", "malware" and "spam"; the arguments are expanded
508 so could be handling tainted values.
510 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
511 broken the (no-op) support for this sendmail command. Restore it
512 to doing nothing, silently, and returning good status.
514 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
515 record path was given (or the default used) without a leading directory
516 path, an error occurred on trying to open it. Use the transport's working
519 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
520 transport. Previously it was the name on the DNS A-record; use instead
521 the head of the CNAME chain leading there (if there is one). This seems
522 to align better with RFC 6125.
524 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
525 smtp_accept_max_per_host allocated resources which were not released
526 when the limit was exceeded. This eventually crashed the daemon. Fix
527 by adding a release action in that path.
529 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
530 expanded; previously using tainted values was rejected. Fix by using
531 dynamically-created buffers.
533 JH/09 Relax restrictions on ACL verify condition needing access to message
534 headers. Previously they were only permitted in data and non-smtp ACLs;
535 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
536 not_blind, header_sender and header_names_ascii verification.
538 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
539 Previously a macro used one argument twice; when called with the
540 argument as an expression having side-effects, incorrect operation
541 resulted. Use an inlineable function.
543 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
544 held open for a verify callout. Previously this wan not accounted for
545 and a corrupt onward SMTP conversation resulted.
547 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
548 Previously they were embedded naively in the constructed address; when
549 needed, strip the quoting and quote the entire local_part.
550 Also make the inbound_srs expansion condition handle quoting.
552 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
553 excluded, not matching the documentation.
555 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
556 was given for the sqlite_dbfile a trap resulted.
558 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
559 "name" argument resulted in a trap. There is no reason to disallow such;
560 this was a coding error.
562 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
563 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
564 spent suspended, ignoring the POSIX definition. Previously we assumed
565 it did and a constant offset from real time could be used as a correction.
566 Change to using the same clock source for the start-of-message and the
567 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
568 exists, just to get a clock slightly more aligned to reality.
570 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
571 RFC says it is optional some validators care. The missing char was not
572 intended but triggered by a line-wrap alignment. Discovery and fix by
573 Guillaume Outters, hacked on by JH.
575 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
576 name being quoted was tainted a trap would be taken. Fix by using
577 dynamically created buffers. The routine could have been called by a
578 rewrite with the "h" flag, by using the "-F" command-line option, or
579 by using a "name=" option on a control=submission ACL modifier.
581 JH/19 SPF: change the Authentication-Results expansion component to give
582 smtp.helo when the sender domain is empty. Previously it gave
585 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
586 not in 127.0.0.0/8 to help in spotting list domains taken over by a
587 domain-parking registrar.
589 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
590 Previously when a whitespace character was specified it was not inserted
591 after removing the newline.
593 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
594 the domain part of the recipient address. This overrides any tls_sni
595 option set, which was previously used.
597 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
600 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
601 is_tainted() had an off-by-one error in the overenthusiastic direction.
602 Find and fix by Gavan. Although NetBSD is not a supported platform for
603 4.94 this bug could affect other platforms.
605 PP/01 Fix default prime selection to be consistent.
606 One path used ike23 still, instead of exim.dev.20160529.3; now both
607 execution flows will use the same DH primes (currently
608 exim.dev.20160529.3).
610 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
611 option in smtp transport, to match the documentation. Previously
612 verification was not being done in some cases where it should have been.
614 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
615 than one server was defined and depending on the platform memory layout
616 details, an internal consistency trap could be hit while walking the list
619 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
620 files. The value is used by the authresults expansion item. Previously
621 if this was used in a router or transport, a crash could result.
623 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
624 have a sufficient incidence of EINTR returns from select that an
625 interaction with other operations done by the main daemon loop exposed
626 a bug in the error-handling. This was benign apart from the log
629 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
630 for consistency with delivered "=>" and failed "**" lines. While we're
631 there, handle PRX and TFO.
633 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
634 applied. This resulted, if any header-line rewrite rules were configured,
635 in a panic-log triggerable by sending a message with a long address in
636 a header. Fix by increasing the arbitrary limit to larger than a single
637 (dewrapped) 5322 header line maximum size.
639 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
640 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
641 RFC 6648 which deprecates X- options in protocols as a general practice.
642 Changeover between the implementations is handled by the mechanisms
645 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
646 a remote which uses callout/hold. Previously the local was queued.
648 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
651 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
652 log line. The asterisk is now consistently appended to the remote IP
653 (and port, if given), and will also be provided on defer and fail log
654 lines. Previously it could be placed on the local IP if that was being
655 logged, and was only provided on delivery lines.
657 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
659 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
660 authenticator client_send option. Previously the next char, after a pair
661 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
662 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
663 documentation. There is still no way to get a leading ^ immediately
664 after a NUL (ie. for the password of a PLAIN method authenticator.
666 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
667 files. For bad sizes read, delete the record and whine to paniclog.
669 JH/38 When logging an AUTH failure, as server, do not include sensitive
670 information. Previously, the credentials would be included if given
671 as part of the AUTH command line and an ACL denied authentication.
673 JH/39 Bug 2691: fix $local_part_data. When the matching list element
674 referred to a file, bad data was returned. This likely also affected
677 JH/40 The gsasl authenticator now supports caching of the salted password
678 generated by the client-side implementation. This required the addition
679 of a new variable: $auth4.
681 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
682 left undeleted; the attempt to re-create it then failed - resulting in
683 the usual "SIGHUP tp have daemon reload configuration" to not work.
684 This affected any platform not supporting "abstract" Unix-domain
685 sockets (i.e. not Linux).
687 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
688 recipients" response to RCPT in a later response, with a 250. The
689 previous coding assumed this would not happen, and under PIPELINING
690 would result in both lost and duplicate recipients for a message.
692 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
693 Previously the weighting was incorrectly applied. Similar fix for socks
694 proxies. Found and fixed by Heiko Schlichting.
696 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
697 not handle sub-lists included using the +namedlist syntax. While
698 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
699 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
702 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
703 Previously the permanent pool was used, so the store could not be freed.
704 This meant a connection with many messages would use continually-growing
707 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
708 per-pool so as not to waste too much space. Previously a constant size
709 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
710 DOS attacks harder. The cost is wasted memory use in the larger blocks.
712 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
713 same space repeatedly, and a smaller process footprint.
715 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
716 for GnuTLS. Previously the name started "%s" which, while not a bug,
717 looked as if if might be one.
719 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
720 "continued connection") make the $proxy_* variables available. Previously
721 the information was not passed across the exec() call for subsequent
722 transport executions. This also mean that the log lines for the
723 messages can show the proxy information.
725 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
726 receive time. With modern systems the difference is significant.
727 The historical behaviour can be restored by disabling (a new) log_selector
728 "queue_time_exclusive".
730 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
731 content) were specifically tested for. Now, also cover expansions
732 resulting in ACL names and inline ACL content.
734 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
735 making it unusable in complex expressions.
737 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
738 fail with a taint-check on the spoolfile name, and leave the message
741 HS/01 Enforce absolute PID file path name.
743 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
745 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
747 PP/02 Bug 2643: Correct TLS DH constants.
748 A missing NUL termination in our code-generation tool had led to some
749 incorrect Diffie-Hellman constants in the Exim source.
750 Reported by kylon94, code-gen tool fix by Simon Arlott.
752 PP/03 Impose security length checks on various command-line options.
753 Fixes CVE-2020-SPRSS reported by Qualys.
755 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
756 better. Reported by Qualys.
758 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
759 providing a particularly obnoxious sender full name.
762 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
764 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
767 PP/08 Change default for recipients_max from unlimited to 50,000.
769 PP/09 Fix security issue with too many recipients on a message (to remove a
770 known security problem if someone does set recipients_max to unlimited,
771 or if local additions add to the recipient list).
772 Fixes CVE-2020-RCPTL reported by Qualys.
774 PP/10 Fix security issue in SMTP verb option parsing
775 Fixes CVE-2020-EXOPT reported by Qualys.
777 PP/11 Fix security issue in BDAT state confusion.
778 Ensure we reset known-good where we know we need to not be reading BDAT
779 data, as a general case fix, and move the places where we switch to BDAT
780 mode until after various protocol state checks.
781 Fixes CVE-2020-BDATA reported by Qualys.
783 HS/03 Die on "/../" in msglog file names
785 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
786 the Exim runtime user are allowed to create files.
788 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
791 QS/03 When reading the output from interpreted forward files we do not
792 pass the pipe between the parent and the interpreting process to
793 executed child processes (if any).
795 QS/04 Always die if requested from internal logging, even is logging is
798 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
799 the API; compilation noo longer completes with DMARC support included.
800 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
801 been reported on other platforms.
803 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
805 JH/56 Make the majority of info read from config files readonly, for defence-in-
806 depth against exploits. Suggestion by Qualys.
807 Not supported on Solaris 10.
809 JH/57 Fix control=fakreject for a custom message containing tainted data.
810 Previously this resulted in a log complaint, due to a re-expansion present
811 since fakereject was originally introduced.
813 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
814 modifier was given, a loop resulted.
816 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
817 pipelined SMTP command followed the BDAT LAST then it would be
818 incorrectly treated as part of the message body, causing a verification
821 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
822 require looser validation rules than those for 821-level addresses,
823 which only permit IP addresses.
829 JH/01 Avoid costly startup code when not strictly needed. This reduces time
830 for some exim process initialisations. It does mean that the logging
831 of TLS configuration problems is only done for the daemon startup.
833 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
835 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
836 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
839 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
840 DKIM signing is being done. Previously a transport_filter would always
841 disable CHUNKING, falling back to traditional DATA.
843 JH/05 Regard command-line recipients as tainted.
845 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
847 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
848 PAM library frees one of the arguments given to it, despite the
849 documentation. Therefore a plain malloc must be used.
851 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
852 on-stack buffers were used, resulting in a taint trap when DSN information
853 copied from a received message was written into the buffer.
855 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
856 the ordering of its ARC headers. This caused a crash.
858 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
859 a new record was being constructed with information from the peer, a trap
862 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
863 installation would get error messages from DMARC verify, when it hit the
864 nonexistent file indicated by the default. Distros wanting DMARC enabled
865 should both provide the file and set the option.
866 Also enforce no DMARC verification for command-line sourced messages.
868 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
869 could, depending on the platform, hang at the STARTTLS response.
871 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
872 message on a connection. Previously if one message had ARC headers and
873 the following one did not, a crash could result when adding an
874 Authentication-Results: header.
876 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
877 Exim main code and Exim-related utities. The introduction of taint
878 tracking also did many adjustments to string handling. Since then, eximon
879 frequently terminated with an assert failure.
881 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
882 check for 452 responses. This slightly helps the inefficieny of doing
883 a large alias-expansion into a recipient-limited target. The max_rcpt
884 transport option still applies (and at the current default, will override
885 the new feature). The check is done for either cause of synch, and forces
886 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
887 connection. The new facility is not tunable at this time.
889 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
890 library live data was being used, so the results became garbage. Make
891 copies while it is still usable.
893 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
894 on delivery deferred (==) and failed (**) lines (if a delivery was
895 attemtped). Previously it was only on completion (=>) lines.
897 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
898 for the expansion of the server_scram_iter and server_scram_salt options.
900 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
901 are now specifically given a NO_DATA response without hitting the system
902 resolver. The library goes on to do the now-standard TXT lookup.
903 Use of dnsdb lookups is not affected.
905 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
906 only retrieve the errormessage once. Previously two calls to dlerror()
907 were used, and the second one (for mainlog/paniclog) retrieved null
910 JH/20 Taint checking: disallow use of tainted data for
911 - the appendfile transport file and directory options
912 - the pipe transport command
913 - the autoreply transport file, log and once options
914 - file names used by the redirect router (including filter files)
916 - paths used by single-key lookups
917 Previously this was permitted.
919 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
920 adjusted the size of a major service buffer; this failed because the
921 buffer was in use at the time. Change to a compile-time increase in the
922 buffer size, when this authenticator is compiled into exim.
924 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
925 previous fast-mode was untenable in the face of glibs using mmap to
926 support larger malloc requests.
928 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
929 New values supported, if defined on system where compiled:
930 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
931 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
933 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
934 running a limited number of proceses in parallel, a benefit is gained. The
935 amount varies with the platform hardware and load. The use of the option
936 queue_run_in_order means we cannot do this, as ordering becomes
939 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
940 had introduced a string-copy (for ensuring NUL-termination) which was not
941 appropriate for that case, which can include embedded NUL bytes in the
942 block of data. Investigation showed the copy to actually be needless, the
943 data being length-specified.
945 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
946 done during a receiving connection, and both used TLS, global info was
947 used rather than per-connection info for tracking the state of data
948 queued for transmission. This could result in a connection hang.
950 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
951 Previously, when delivering serveral messages down a single connection
952 only the first would provide a SIZE. This was due to the size information
953 not being properly tracked.
955 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
956 TAI (at 37 seconds currently), pretend to be in UTC for time-related
957 expansion and logging. Previously, spurious values such as a future
958 minute could be seen.
960 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
961 it could crash from a null-deref. This could also affect the
962 ${addresses: } operator and ${readsock } item.
964 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
965 message following a mime one, the variable was not reset.
967 JH/30 When an pipelined-connect fails at the first response, assume incorrect
968 cached capability (perhaps the peer reneged?) and immediately retry in
971 JH/31 Fix spurious detection of timeout while writing to transport filter.
973 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
974 an attempt to copy the string was made before checking it.
976 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
977 taint of the lookup key was maintained; we now regard the presence in the
978 filesystem as sufficient validation.
980 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
981 argument is supplied.
983 JH/35 The dsearch lookup now requires that the directory is an absolute path.
984 Previously this was not checked, and nonempty relative paths made an
985 access under Exim's current working directory.
987 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
988 Previously no event was raised.
990 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
991 parameter supplied by the sender MAIL FROM command. Previously it was
992 ignored, and only the check_spool_space option value for the required
995 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
996 the size of the signing public-key. Previously it was instead giving
997 the size of the signature hash.
999 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1000 the default. See the (new) dkim_verify_min_keysizes option.
1002 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1003 and an ACL use a lookup for checking either the local_part or domain,
1004 stale data could be accessed. Ensure that variable references are
1005 dropped between messages.
1007 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1008 by the client was not checked as pointing within response data before
1009 being used. A malicious client could thus cause an out-of-bounds read and
1010 possibly gain authentication. Fix by adding the check.
1012 JH/42 Internationalisation: change the default for downconversion in the smtp
1013 transport to be "if needed". Previously it was "as previously set" for
1014 the message, which usually meant "if needed" for message-submission but
1015 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1016 even when the envelope addresses did not need it, resulting in forwarding
1017 failures to non-supporting MTAs. A downconvert in such cases will be
1018 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1019 transport. The change does mean that addresses needing conversion will
1020 be converted when previously a delivery failure would occur.
1022 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1023 response was received it would be used unchecked in a fail-DSN, violating
1024 standards on line-length limits. Truncate if needed.
1026 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1027 doesn't work. (Additionally add a new main config option to configure the
1034 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1035 side, to decode a TLS 1.3 packet capture.
1037 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1038 Previously the default library behaviour applied, sending two, each in
1039 its own TCP segment.
1041 JH/03 Debug output for ACL now gives the config file name and line number for
1044 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1046 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1048 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1049 buffer overrun for (non-chunking) other transports.
1051 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1052 TLS1.3, means that a server rejecting a client certificate is not visible
1053 to the client until the first read of encrypted data (typically the
1054 response to EHLO). Add detection for that case and treat it as a failed
1055 TLS connection attempt, so that the normal retry-in-clear can work (if
1056 suitably configured).
1058 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1059 and/or domain. Found and fixed by Jason Betts.
1061 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1062 configuration). If a CNAME target was not a wellformed name pattern, a
1065 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1066 the OS reports them interleaved with other addresses.
1068 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1069 used both for input and for a verify callout, both encrypted, SMTP
1070 responses being sent by the server could be lost. This resulted in
1071 dropped connections and sometimes bounces generated by a peer sending
1074 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1075 string. Previously it was possible to cause undefined behaviour in a
1076 library routine (usually a crash). Found by "zerons".
1078 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1081 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1082 API was removed, so update to use the newer ones.
1084 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1085 any timeout set, is taking a long time. Previously we would hang on to a
1086 rotated logfile "forever" if the input was arriving with long gaps
1087 (a previous attempt to fix addressed lack, for a long time, of initial
1090 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1091 shared (NFS) environment. The length of the tempfile name is now
1092 4 + 16 ("hdr.$message_exim_id") which might break on file
1093 systems which restrict the file name length to lower values.
1094 (It was "hdr.$pid".)
1096 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1097 shared (NFS) environment.
1099 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1100 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1103 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1104 on some platforms for bit 31.
1106 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1107 to changes apparently associated with TLS1.3 handling some of the APIs
1108 previously used were either nonfunctional or inappropriate. Strings
1109 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1110 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1111 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1112 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1113 and the use of specific cipher names in the encrypted= ACL condition.
1115 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1117 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1118 verification result was not updated unless hosts_require_ocsp applied.
1120 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1121 queue_list_requires_admin set to false, non-admin users were denied the
1124 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1125 directory-of-certs mode. Previously they were advertised despite the
1128 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1129 A single TCP connection by a client will now hold a TLS connection open
1130 for multiple message deliveries, by default. Previously the default was to
1133 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1134 default. If built with the facility, DANE will be used. The facility
1135 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1137 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1138 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1139 must be defined and you must still, unless you define DISABLE_TLS, manage
1140 the the include-dir and library-file requirements that go with that
1141 choice. Non-TLS builds are still supported.
1143 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1146 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1147 default. If the platform supports and has the facility enabled, it will
1148 be requested on all coneections.
1150 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1151 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1153 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1155 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1156 success-DSN messages. Previously the From: header was always the default
1157 one for these; the option was ignored.
1159 JH/28 Fix the timeout on smtp response to apply to the whole response.
1160 Previously it was reset for every read, so a teergrubing peer sending
1161 single bytes within the time limit could extend the connection for a
1162 long time. Credit to Qualsys Security Advisory Team for the discovery.
1164 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1165 delivery address, which leaked information of the results of local
1166 forwarding. Change to the original envelope recipient address, per
1169 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1170 requested. Previously not bounce was generated and a log entry of
1171 error ignored was made.
1173 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1175 JH/32 Introduce a general tainting mechanism for values read from the input
1176 channel, and values derived from them. Refuse to expand any tainted
1177 values, to catch one form of exploit.
1179 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1180 was unused and the unexpanded text used for the test. Found and
1181 fixed by Ruben Jenster.
1183 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1184 an attempt to use a TLS library read routine dereffed a nul pointer,
1187 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1188 them in our smtp response.
1190 JH/36 Have the general router option retry_use_local_part default to true when
1191 any of the restrictive preconditions are set (to anything). Previously it
1192 was only for check_local user. The change removes one item of manual
1193 configuration which is required for proper retries when a remote router
1194 handles a subset of addresses for a domain.
1196 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1197 link count into consideration.
1199 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1200 caused the extension of big_buffer, the following lines were ignored.
1202 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1203 accordance with RFC 2308. Previously there was no expiry, so a longlived
1204 receive process (eg. due to ACL delays) versus a short SOA value could
1207 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1209 JH/39 Promote DMARC support to mainline.
1211 JH/40 Bug 2452: Add a References: header to DSNs.
1213 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1214 parameters. The relevant library call is documented as "Deprecated: This
1215 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1216 3.6.0, DH parameters are negotiated following RFC7919."
1218 HS/06 Change the default of dnssec_request_domains to "*"
1220 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1221 carried on and emitted a BDAT command, even when PIPELINING was not
1224 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1225 buffer was used for the filename, resulting in a trap when tainted
1226 arguments (eg. $domain) were used.
1228 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1229 recommended to avoid a possible server-load attack. The feature can be
1230 re-enabled via the openssl_options main cofiguration option.
1232 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1233 for version 4.90 - adding a "more data" boolean to the arguments.
1234 Bumped the ABI version number also, this having been missed previously;
1235 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1236 respect of usage of smtp_printf() by either local_scan code or libraries
1237 accessed via the ${dlfunc } expansion item. Both will need coding
1238 adjustment for any calls to smtp_printf() to match the new function
1239 signature; a FALSE value for the new argument is always safe.
1241 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1242 the file-offset (which the Linux syscall does, and exim expects); this
1243 resulted in an indefinite loop.
1245 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1246 to do ARC verification. The Authentication-Results: header line added
1247 by the configuration then had no ARC item.
1249 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1250 an incorrect mode was used for reading the result, resulting in it being
1257 JH/01 Remove code calling the customisable local_scan function, unless a new
1258 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1260 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1261 non-signal-safe functions being used.
1263 JH/03 Bug 2269: When presented with a received message having a stupidly large
1264 number of DKIM-Signature headers, disable DKIM verification to avoid
1265 a resource-consumption attack. The limit is set at twenty.
1267 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1268 report of oldest_pass in ${authres } in consequence, and separate out
1269 some descriptions of reasons for verification fail.
1271 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1272 files in the spool were present and unlocked. A queue-runner could spot
1273 them, resulting in a duplicate delivery. Fix that by doing the unlock
1274 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1275 add more error-checking on spoolfile handling while that code is being
1278 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1279 No known attacks, no CVE, this is defensive hardening.
1281 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1282 a queue-runner could start a delivery while other operations were ongoing.
1283 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1284 Found and investigated by Tim Stewart. Fix by using the open message data
1285 file handle rather than opening another, and not locally closing it (which
1286 releases a lock) for that case, while creating the temporary .eml format
1287 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1289 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1290 $sender_verify_failure/$recipient_verify_failure to "random".
1292 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1295 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1296 Previously this would segfault.
1298 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1301 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1302 like zero, since the resolver should be doing this for us, But we need one
1303 as a CNAME but no MX presence gets the CNAME returned; we need to check
1304 that doesn't point to an MX to declare it "no MX returned" rather than
1305 "error, loop". A new main option is added so the older capability of
1306 following some limited number of chain links is maintained.
1308 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1310 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1311 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1312 now-deprecated earlier definitions used only specified the range up to TLS
1313 1.2 (in the older-version library docs).
1315 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1317 JH/15 Rework TLS client-side context management. Stop using a global, and
1318 explicitly pass a context around. This enables future use of TLS for
1319 connections to service-daemons (eg. malware scanning) while a client smtp
1320 connection is using TLS; with cutthrough connections this is quite likely.
1322 JH/16 Fix ARC verification to do AS checks in reverse order.
1324 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1326 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1327 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1328 Previously the "utf8" would be re-prepended for every additional message.
1330 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1331 Previously thery were accepted, resulting in issues when attempting to
1332 forward messages to a non-supporting MTA.
1334 PP/02 Let -n work with printing macros too, not just options.
1336 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1337 one parent address was copied, and bogus data was used at delivery-logging
1338 time. Either a crash (after delivery) or bogus log data could result.
1339 Discovery and analysis by Tim Stewart.
1341 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1342 Previously if the string ended mid-character, we did not insert the
1343 promised '?' replacement.
1345 PP/04 Documentation: current string operators work on bytes, not codepoints.
1347 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1348 should pack well giving a smaller memory footprint so better caching and
1349 therefore performance. Group the declarations where this can't be done so
1350 that the byte-sized flag variables are not interspersed among pointer
1351 variables, giving a better chance of good packing by the compiler.
1353 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1354 non-null, to avoid issues with sites running BATV. Previously reports were
1355 sent with an empty envelope sender so looked like bounces.
1357 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1358 The ignore_error flag wasn't being returned from the filter subprocess so
1359 was not set for later routers. Investigation and fix by Matthias Kurz.
1361 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1362 and a msg:complete for the whole, when a message is manually removed using
1363 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1365 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1366 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1367 hope that is portable enough.
1369 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1370 requiring a known-CA anchor certificate; make it now rely entirely on the
1371 TLSA as an anchor. Checking the name on the leaf cert against the name
1372 on the A-record for the host is still done for TA (but not for EE mode).
1374 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1375 included in delivery lines for non-proxied connections, when compiled with
1376 SUPPORT_SOCKS and running with proxy logging enabled.
1378 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1379 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1380 move the existing event to fire before the normal logging of message
1381 failure so that custom logging is bracketed by normal logging.
1383 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1384 msg:fail:internal event. Developement by Matthias Kurz.
1386 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1387 far too small for todays use of crypto signatures stored there. Go all
1388 the way to the max DNS message size of 64kB, even though this might be
1389 overmuch for IOT constrained device use.
1391 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1392 copy a string over itself. The library routine is documented as not
1393 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1395 JH/32 For main options check_spool_space and check_inode_space, where the
1396 platform supports 64b integers, support more than the previous 2^31 kB
1397 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1398 the previous G, M, k.
1400 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1401 $authenticated_fail_id variable on authentication failure. Previously
1404 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1405 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1406 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1407 GNUTLS_SEC_PARAM_MEDIUM.
1409 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1410 side. Previously we would continue as if no SNI had been received.
1412 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1413 "<" character, which should be a list-separator specification, we walked
1414 off past the nul-terimation.
1416 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1417 causes) even when the retry time is not yet met. Previously they were
1418 not, meaning that when (say) an account was over-quota and temp-rejecting,
1419 and multiple senders' messages were queued, only one sender would get
1420 notified on each configured delay_warning cycle.
1422 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1424 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1425 cipher-suites, an error can be left on the stack even for a succeeding
1426 accept; this results in impossible error messages when a later operation
1429 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1430 return error codes indicating retry. Under TLS1.3 this becomes required.
1432 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1433 it only wrote the new authenticators, resulting in a lack of tracking of
1434 peer changes of ESMTP extensions until the next cache flush.
1436 JH/41 Fix the loop reading a message header line to check for integer overflow,
1437 and more-often against header_maxsize. Previously a crafted message could
1438 induce a crash of the recive process; now the message is cleanly rejected.
1440 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1441 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1447 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1448 When redis_servers is set to a list of > 1 element, and the Redis servers
1449 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1450 case of MOVED into a DEFER case instead, thus moving the query onto the
1451 next server in the list. For a cluster of N elements, all N servers must
1452 be defined in redis_servers.
1454 GF/02 Catch and remove uninitialized value warning in exiqsumm
1455 Check for existence of @ARGV before looking at $ARGV[0]
1457 JH/01 Replace the store_release() internal interface with store_newblock(),
1458 which internalises the check required to safely use the old one, plus
1459 the allocate and data copy operations duplicated in both (!) of the
1460 extant use locations.
1462 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1463 modifier. This matches the restriction on the commandline.
1465 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1466 Previously only the last row was returned.
1468 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1469 we assumed that tags in the header were well-formed, and parsed the
1470 element content after inspecting only the first char of the tag.
1471 Assumptions at that stage could crash the receive process on malformed
1474 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1475 While running the DKIM ACL we operate on the Permanent memory pool so that
1476 variables created with "set" persist to the DATA ACL. Also (at any time)
1477 DNS lookups that fail create cache records using the Permanent pool. But
1478 expansions release any allocations made on the current pool - so a dnsdb
1479 lookup expansion done in the DKIM ACL releases the memory used for the
1480 DNS negative-cache, and bad things result. Solution is to switch to the
1481 Main pool for expansions.
1482 While we're in that code, add checks on the DNS cache during store_reset,
1483 active in the testsuite.
1484 Problem spotted, and debugging aided, by Wolfgang Breyha.
1486 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1487 When none of the hosts presented to a transport match an already-open
1488 connection, close it and proceed with the list. Previously we would
1489 queue the message. Spotted by Lena with Yahoo, probably involving
1492 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1493 Previously a spurious "250 OK id=" response was appended to the proper
1496 JH/08 The "support for" informational output now, which built with Content
1497 Scanning support, has a line for the malware scanner interfaces compiled
1498 in. Interface can be individually included or not at build time.
1500 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1501 by the template makefile "src/EDITME". The "STREAM" support for an older
1502 ClamAV interface method is removed.
1504 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1505 rows affected is given instead).
1507 JH/11 The runtime Berkeley DB library version is now additionally output by
1508 "exim -d -bV". Previously only the compile-time version was shown.
1510 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1511 SMTP connection. Previously, when one had more recipients than the
1512 first, an abortive onward connection was made. Move to full support for
1513 multiple onward connections in sequence, handling cutthrough connection
1514 for all multi-message initiating connections.
1516 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1517 routers. Previously, a multi-recipient message would fail to match the
1518 onward-connection opened for the first recipient, and cause its closure.
1520 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1521 a timeout on read on a GnuTLS initiating connection, resulting in the
1522 initiating connection being dropped. This mattered most when the callout
1523 was marked defer_ok. Fix to keep the two timeout-detection methods
1526 JH/15 Relax results from ACL control request to enable cutthrough, in
1527 unsupported situations, from error to silently (except under debug)
1528 ignoring. This covers use with PRDR, frozen messages, queue-only and
1531 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1533 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1534 metadata, resulting in a crash in free().
1536 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1537 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1538 Broken also in d185889f4, with init system revamp.
1540 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1541 Previously we abruptly closed the connection after reading a malware-
1542 found indication; now we go on to read the "scan ok" response line,
1545 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1546 ACL. Previously, a crash would result.
1548 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1549 macro text after a replacement (previously it was only once per line) and
1550 by skipping builtin macros when searching for an uppercase lead character.
1552 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1553 for the build is renamed.
1555 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1556 was allocated for every new TLS startup, meaning one per message. Fix
1557 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1559 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1560 reported the original. Fix to report (as far as possible) the ACL
1561 result replacing the original.
1563 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1564 OpenSSL. Certificate information is loaded for every new TLS startup,
1565 and the resources needed to be freed.
1567 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1569 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1572 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1573 DATA response info to the (existing) per-recipient response info for
1574 the "C=" log element. It can have useful tracking info from the
1575 destination system. Patch from Simon Arlott.
1577 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1578 length value. Previously this would segfault.
1580 HS/02 Support Avast multiline protoocol, this allows passing flags to
1581 newer versions of the scanner.
1583 JH/28 Ensure that variables possibly set during message acceptance are marked
1584 dead before release of memory in the daemon loop. This stops complaints
1585 about them when the debug_store option is enabled. Discovered specifically
1586 for sender_rate_period, but applies to a whole set of variables.
1587 Do the same for the queue-runner and queue-list loops, for variables set
1588 from spool message files. Do the same for the SMTP per-message loop, for
1589 certain variables indirectly set in ACL operations.
1591 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1592 as a multi-recipient message from a mailinglist manager). The coding had
1593 an arbitrary cutoff number of characters while checking for more input;
1594 enforced by writing a NUL into the buffer. This corrupted long / fast
1595 input. The problem was exposed more widely when more pipelineing of SMTP
1596 responses was introduced, and one Exim system was feeding another.
1597 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1598 receiving system, and refused recipients seen by the sending system
1599 (propating to people being dropped from mailing lists).
1600 Discovered and pinpointed by David Carter.
1602 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1603 replaced by the ${authresults } expansion.
1605 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1607 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1608 allows proper process termination in container environments.
1610 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1611 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1613 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1614 and "err_perm", deprecated since 4.83 when the RFC-defined words
1615 "temperror" and "permerror" were introduced.
1617 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1618 transport-filters or DKIM-signing. The restriction was lost in the
1619 consolidation of verify-callout and delivery SMTP handling.
1620 Extend the restriction to also cover ARC-signing.
1622 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1623 in defer=pass mode supply a 450 to the initiator. Previously the message
1626 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1627 tls_require_ciphers is used as before.
1629 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1630 "pass_unscanned". Only tmpfails from the scanner are written to
1631 the paniclog, as they may require admin intervention (permission
1632 denied, license issues). Other scanner errors (like decompression
1633 bombs) do not cause a paniclog entry.
1635 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1636 Previously it was possible to log spurious information in receive log
1639 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1640 triggered odd behaviour from Outlook Express clients.
1642 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1645 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1646 since the IETF WG has not yet settled on that versus the original
1647 "bare" representation.
1649 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1650 Previously the millisecond value corrupted the output.
1651 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1652 corrupted the output.
1658 JH/01 Rework error string handling in TLS interface so that the caller in
1659 more cases is responsible for logging. This permits library-sourced
1660 string to be attached to addresses during delivery, and collapses
1661 pairs of long lines into single ones.
1663 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1664 during configuration. Wildcards are allowed and expanded.
1666 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1667 This permits better logging.
1669 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1670 do not close it down and have the child transport start it up again on
1671 the passed-on TCP connection. Instead, proxy the child (and any
1672 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1673 affected: the continued delivery log lines do not have any DNSSEC, TLS
1674 Certificate or OCSP information. TLS cipher information is still logged.
1676 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1677 identical IP addresses on different listening ports. Will also affect
1680 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1681 add noisy ifdef guards to special-case this sillyness.
1682 Patch from Bernd Kuhls.
1684 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1685 than 255 are no longer allowed.
1687 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1688 Disable the session-cache too, which might reduce our load. Since we
1689 currrectly use a new context for every connection, both as server and
1690 client, there is no benefit for these.
1691 GnuTLS appears to not support tickets server-side by default (we don't
1692 call gnutls_session_ticket_enable_server()) but client side is enabled
1693 by default on recent versions (3.1.3 +) unless the PFS priority string
1696 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1697 <https://reproducible-builds.org/specs/source-date-epoch/>.
1699 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1700 the check for any unsuccessful recipients did not notice the limit, and
1701 erroneously found still-pending ones.
1703 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1704 MSG_MORE. Only in-clear (not on TLS connections).
1706 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1707 transport-filter is involved we can buffer the headers in memory for
1708 creating the signature, and read the spool data file once for the
1709 signature and again for transmission.
1711 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1712 4.77 as the kernel support then wasn't solid, having issues in 64bit
1713 mode. Now, it's been long enough. Add support for FreeBSD also.
1715 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1716 case where the routing stage had gathered several addresses to send to
1717 a host before calling the transport for the first, we previously failed
1718 to close down TLS in the old transport process before passing the TCP
1719 connection to the new process. The new one sent a STARTTLS command
1720 which naturally failed, giving a failed delivery and bloating the retry
1721 database. Investigation and fix prototype from Wolfgang Breyha.
1723 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1724 false-negatives in the check that the sender had not preempted a response
1725 or prompt from Exim (running as a server), due to that code's lack of
1726 awareness of the SMTP input buffering.
1728 PP/04 Add commandline_checks_require_admin option.
1729 Exim drops privileges sanely, various checks such as -be aren't a
1730 security problem, as long as you trust local users with access to their
1731 own account. When invoked by services which pass untrusted data to
1732 Exim, this might be an issue. Set this option in main configuration
1733 AND make fixes to the calling application, such as using `--` to stop
1736 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1737 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1738 responses to those, into a single TLS record each way (this usually means
1739 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1742 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1743 affects you only if you're dancing at the edge of the param size limits.
1744 If you are, and this message makes sense to you, then: raise the
1745 configured limit or use OpenSSL 1.1. Nothing we can do for older
1748 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1749 cmdline element to get the documented default one. Previously it was
1752 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1753 are unsuable for non-message-specific reasons - eg. connection timeout,
1754 banner-time rejection.
1756 JH/16 Fix logging of delivery remote port, when specified by router, under
1759 PP/06 Repair manualroute's ability to take options in any order, even if one
1760 is the name of a transport.
1763 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1765 JH/17 Change the list-building routines interface to use the expanding-string
1766 triplet model, for better allocation and copying behaviour.
1768 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1769 Previously it was constructed the first time a possibly-matching string
1770 was met in the configuration file input during startup; now it is done
1773 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1774 compatible one, to avoid the (poorly documented) possibility of a config
1775 file in the working directory redirecting the DB files, possibly correpting
1776 some existing file. CVE-2017-10140 assigned for BDB.
1778 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1779 cache-hot. Previously, although the result was properly cached, the
1780 initial verify call returned a defer.
1782 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1783 the main verify for receipient in uncached-mode.
1785 JH/22 Retire historical build files to an "unsupported" subdir. These are
1786 defined as "ones for which we have no current evidence of testing".
1788 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1789 if present. Previously it was ignored.
1791 JH/24 Start using specified-initialisers in C structure init coding. This is
1792 a C99 feature (it's 2017, so now considered safe).
1794 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1795 if was a fixed-sized field and bitmask ops via macros; it is now more
1798 PP/07 GitHub PR 56: Apply MariaDB build fix.
1799 Patch provided by Jaroslav Škarvada.
1801 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1802 during Coverity cleanups [4.87 JH/47]
1803 Diagnosis and fix provided by Michael Fischer v. Mollard.
1805 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1806 the right size to place the terminating semicolon on its own folded
1807 line, the header hash was calculated to an incorrect value thanks to
1808 the (relaxed) space the fold became.
1810 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1811 and confused the parent.
1813 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1814 which could crash as a result. This could lead to undeliverable messages.
1816 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1819 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1820 matters on fast-turnover and PID-randomising systems, which were getting
1821 out-of-order delivery.
1823 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1824 a possibly-overlapping copy. The symptom was that "Remote host closed
1825 connection in response to HELO" was logged instead of the actual 4xx
1828 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1829 Previously only that bufferd was discarded, resulting in SYMTP command
1832 JH/32 DKIM: when a message has multiple signatures matching an identity given
1833 in dkim_verify_signers, run the dkim acl once for each. Previously only
1834 one run was done. Bug 2189.
1836 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1837 "panic the current process" to "deliberately defer". The panic log is
1838 still written with the problem list name; the mail and reject logs now
1839 get a temp-reject line for the message that was being handled, saying
1840 something like "domains check lookup or other defer". The SMTP 451
1841 message is still "Temporary local problem".
1843 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1844 A crafted sequence of BDAT commands could result in in-use memory beeing
1845 freed. CVE-2017-16943.
1847 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1848 from SMTP input. Previously it was always done; now only done for DATA
1849 and not BDAT commands. CVE-2017-16944.
1851 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1852 to the message (such as an overlong header line). Previously this was
1853 not done and we did not exit BDAT mode. Followon from the previous item
1854 though a different problem.
1860 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1861 than -2003 did; needs libidn2 in addition to libidn.
1863 JH/02 The path option on a pipe transport is now expanded before use.
1865 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1866 Patch provided by "Björn", documentation fix added too.
1868 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1869 missing a wire-to-host endian conversion.
1871 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1872 close after a BDAT command line could be taken as a following command,
1873 giving a synch failure. Fix by only checking for synch immediately
1874 before acknowledging the chunk.
1876 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1877 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1878 macro. Patches provided by Josh Soref.
1880 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1881 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1882 by IANA as a service extension. However, John Klensin suggests that we
1885 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1886 the dkim code may be unix-mode line endings rather than smtp wire-format
1887 CRLF, so prepend a CR to any bare LF.
1889 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1890 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1892 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1893 the first were themselves being wrongly included in the feed into dkim
1894 processing; with most chunk sizes in use this resulted in an incorrect
1895 body hash calculated value.
1897 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1898 DKIM signature block, for verification. Although advised against by
1899 standards it is specifically not ruled illegal.
1901 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1903 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1904 missing a body hash (the bh= tag).
1906 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1907 It seems that HAProxy sends the Proxy Protocol information in clear and
1908 only then does a TLS startup, so do the same.
1910 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1911 TCP connections (such as for Spamd) unless the daemon successfully set
1912 Fast Open mode on its listening sockets. This fixes breakage seen on
1913 too-old kernels or those not configured for Fast Open, at the cost of
1914 requiring both directions being enabled for TFO, and TFO never being used
1915 by non-daemon-related Exim processes.
1917 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1918 endings, at least on the first header line. Try to canonify any that get
1919 past that check, despite the cost.
1921 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1922 now limited to an arbitrary five deep, while parsing addresses with the
1923 strip_excess_angle_brackets option enabled.
1925 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1926 instead leave the unprompted TLS handshake in socket buffer for the
1927 TLS library to consume.
1929 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1931 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1933 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1934 region in one operation, for speed, and this leaves assigned pointers
1935 dangling. Add checks run only under the testsuite which checks all
1936 variables at a store-reset and panics on a dangling pointer; add code
1937 explicitly nulling out all the variables discovered. Fixes one known
1938 bug: a transport crash, where a dangling pointer for $sending_ip_address
1939 originally assigned in a verify callout, is re-used.
1941 PP/06 Drop '.' from @INC in various Perl scripts.
1943 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1945 PP/08 Reduce a number of compilation warnings under clang; building with
1946 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1947 should be warning-free.
1949 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1951 HS/01 Fix portability problems introduced by PP/08 for platforms where
1952 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1954 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1955 chunk. This allows us to accept broken chunked messages. We need a more
1956 general solution here.
1958 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1959 already-broken messages in the queue.
1961 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1963 JH/19 Fix reference counting bug in routing-generated-address tracking.
1969 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1970 supports it and a size is available (ie. the sending peer gave us one).
1972 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1973 years of being deprecated). The replacements are the ACLs
1974 acl_smtp_mime and acl_not_smtp_mime.
1976 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1977 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1978 or even an in-clear connection were permitted. Now, if the host lookup
1979 was dnssec and dane was requested then the host is only used if the
1980 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1981 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1982 if one fails this test.
1983 This means that a poorly-configured remote DNS will make it incommunicado;
1984 but it protects against a DNS-interception attack on it.
1986 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1987 non-noisy when a race steals the message being considered.
1989 JH/05 If main configuration option tls_certificate is unset, generate a
1990 self-signed certificate for inbound TLS connections.
1992 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1993 in rewrites and routers.
1995 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1996 and logged a warning sing 4.83; now they are a configuration file error.
1998 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1999 (lacking @domain). Apply the same qualification processing as RCPT.
2001 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2003 JH/10 Support ${sha256:} applied to a string (as well as the previous
2006 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2007 a cutthrough deliver is pending, as we always want to make a connection.
2008 This also avoids re-routing the message when later placing the cutthrough
2009 connection after a verify cache hit.
2010 Do not update it with the verify result either.
2012 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2013 when routing results in more than one destination address.
2015 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2016 signing (which inhibits the cutthrough capability). Previously only
2017 the presence of an option was tested; now an expansion evaluating as
2018 empty is permissible (obviously it should depend only on data available
2019 when the cutthrough connection is made).
2021 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2022 the relevant preceding SMTP command did not note the pipelining mode.
2024 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2025 Previously they were not counted.
2027 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2028 as one having no matching records. Previously we deferred the message
2029 that needed the lookup.
2031 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2032 distinguished as "(=".
2034 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2035 for missing MX records. Previously it only worked for missing A records.
2037 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2039 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2040 after the data-go-ahead and data-ack. Patch from Jason Betts.
2042 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2043 even for a "none" policy. Patch from Tony Meyer.
2045 JH/22 Fix continued use of a connection for further deliveries. If a port was
2046 specified by a router, it must also match for the delivery to be
2049 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2050 When one of the recipients of a message was unsuitable for the connection
2051 (has no matching addresses), we lost track of needing to mark it
2052 deferred. As a result mail would be lost.
2054 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2056 JH/25 Decoding ACL controls is now done using a binary search; the source code
2057 takes up less space and should be simpler to maintain. Merge the ACL
2058 condition decode tables also, with similar effect.
2060 JH/26 Fix problem with one_time used on a redirect router which returned the
2061 parent address unchanged. A retry would see the parent address marked as
2062 delivered, so not attempt the (identical) child. As a result mail would
2065 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2066 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2067 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2070 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2071 Main config options check_{log,spool}_{inodes,space} are now
2072 100 inodes, 10MB unless set otherwise in the configuration.
2074 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2075 Previously it only applied to the main-section connection policy
2078 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2080 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2081 by me. Added RFC7919 DH primes as an alternative.
2083 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2084 are not in the system include path.
2086 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2087 GnuTLS, when a session startup failed (eg because the client disconnected)
2088 Exim did stdio operations after fclose. This was exposed by a recent
2089 change which nulled out the file handle after the fclose.
2091 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2092 signed directly by the cert-signing cert, rather than an intermediate
2093 OCSP-signing cert. This is the model used by LetsEncrypt.
2095 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2097 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2098 an incoming connection.
2100 HS/02 Bug 1802: Do not half-close the connection after sending a request
2103 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2104 fallback to "prime256v1".
2106 JH/34 SECURITY: Use proper copy of DATA command in error message.
2107 Could leak key material. Remotely exploitable. CVE-2016-9963.
2113 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2114 and 3.4.4 - once the server is enabled to respond to an OCSP request
2115 it does even when not requested, resulting in a stapling non-aware
2116 client dropping the TLS connection.
2118 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2119 support variable-length bit vectors. No functional change.
2121 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2122 The I= interface field on outgoing lines is now after the H= remote
2123 host field, same as incoming lines. There is a separate
2124 outgoing_interface log selector which allows you to disable the
2127 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2128 If not running log_selector +smtp_connection the mainlog would be held
2129 open indefinitely after a "too many connections" event, including to a
2130 deleted file after a log rotate. Leave the per net connection logging
2131 leaving it open for efficiency as that will be quickly detected by the
2132 check on the next write.
2134 HS/01 Bug 1671: Fix post transport crash.
2135 Processing the wait-<transport> messages could crash the delivery
2136 process if the message IDs didn't exist for some reason. When
2137 using 'split_spool_directory=yes' the construction of the spool
2138 file name failed already, exposing the same netto behaviour.
2140 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2141 mime_regex ACL conditions.
2143 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2144 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2145 to HELO, local diagnostic string.
2147 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2148 log line to debug. Even when configured with a tls authenticator many
2149 client connections are expected to not authenticate in this way, so
2150 an authenticate fail is not an error.
2152 HS/02 Add the Exim version string to the process info. This way exiwhat
2153 gives some more detail about the running daemon.
2155 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2156 matter for fast-change records such as DNSBLs.
2158 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2159 retry record, even if constant. There may be multiple transports with
2160 different interface settings and the retry behaviour needs to be kept
2163 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2165 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2167 JH/10 Bug 840: fix log_defer_output option of pipe transport
2169 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2170 hosts_randomize. This matters a lot when combined with mua_wrapper.
2172 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2173 ${quote_pgsql:<string>} operator.
2175 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2177 JH/14 Change header file name for freeradius-client. Relevant if compiling
2178 with Radius support; from the Gentoo tree and checked under Fedora.
2180 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2182 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2183 option transport_filter, meaning no filtering.
2185 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2187 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2188 defaults to "*" (all hosts). The variable is now available when not built
2189 with TLS, default unset, mainly to enable keeping the testsuite sane.
2190 If a server certificate is not supplied (via tls_certificate) an error is
2191 logged, and clients will find TLS connections fail on startup. Presumably
2192 they will retry in-clear.
2193 Packagers of Exim are strongly encouraged to create a server certificate
2194 at installation time.
2196 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2197 with the $config_file variable.
2199 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2200 in transport context, after the attempt, and per-recipient. The latter type
2201 is per host attempted. The event data is the error message, and the errno
2202 information encodes the lookup type (A vs. MX) used for the (first) host,
2203 and the trailing two digits of the smtp 4xx response.
2205 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2206 to write to mainlog (or rejectlog, paniclog) in the window between file
2207 creation and permissions/ownership being changed. Particularly affects
2208 installations where exicyclog is run as root, rather than exim user;
2209 result is that the running daemon panics and dies.
2211 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2213 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2214 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2215 "pri" and "weight". Note that the previous implicit priority given by the
2216 list order is no longer honoured.
2218 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2219 for DKIM processing.
2221 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2222 by defining SUPPORT_SOCKS.
2224 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2225 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2226 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2227 variables are renamed to proxy_{local,external}_{address,port}.
2229 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2230 for a build by defining SUPPORT_I18N
2232 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2233 of the query string, and make ${quote_redis:} do that quoting.
2235 JH/29 Move Events support from Experimental to mainline, enabled by default
2236 and removable for a build by defining DISABLE_EVENT.
2238 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2240 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2241 cached by the daemon.
2243 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2244 by defining LOOKUP_REDIS. The libhiredis library is required.
2246 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2247 keys are given for lookup.
2249 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2250 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2251 only supported when built with TLS support. The PolarSSL SHA routines
2252 are still used when the TLS library is too old for convenient support.
2254 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2255 openssl_options), for security. OpenSSL forces this from version 1.1.0
2256 server-side so match that on older versions.
2258 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2259 allocation for $value could be released as the expansion processing
2260 concluded, but leaving the global pointer active for it.
2262 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2263 and to use the domains and local_parts ACL conditions.
2265 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2266 incorrectly not doubled on cutthrough transmission, hence seen as a
2267 body-termination at the receiving system - resulting in truncated mails.
2268 Commonly the sender saw a TCP-level error, and retransmitted the message
2269 via the normal store-and-forward channel. This could result in duplicates
2270 received - but deduplicating mailstores were liable to retain only the
2271 initial truncated version.
2273 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2275 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2277 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2278 we're in there, support oversigning also; bug 1309.
2280 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2282 HS/04 Add support for keep_environment and add_environment options.
2284 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2285 either intentional arithmetic overflow during PRNG, or testing config-
2288 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2289 delivery resulted in actual delivery. Cancel cutthrough before DATA
2292 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2293 rejects a recipient: pass the reject to the originator.
2295 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2296 Many were false-positives and ignorable, but it's worth fixing the
2299 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2300 for the new environment-manipulation done at startup. Move the routines
2301 from being local to tls.c to being global via the os.c file.
2303 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2304 an extract embedded as result-arg for a map, the first arg for extract
2305 is unavailable so we cannot tell if this is a numbered or keyed
2306 extraction. Accept either.
2312 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2315 JH/02 The smtp transport option "multi_domain" is now expanded.
2317 JH/03 The smtp transport now requests PRDR by default, if the server offers
2320 JH/04 Certificate name checking on server certificates, when exim is a client,
2321 is now done by default. The transport option tls_verify_cert_hostnames
2322 can be used to disable this per-host. The build option
2323 EXPERIMENTAL_CERTNAMES is withdrawn.
2325 JH/05 The value of the tls_verify_certificates smtp transport and main options
2326 default to the word "system" to access the system default CA bundle.
2327 For GnuTLS, only version 3.0.20 or later.
2329 JH/06 Verification of the server certificate for a TLS connection is now tried
2330 (but not required) by default. The verification status is now logged by
2331 default, for both outbound TLS and client-certificate supplying inbound
2334 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2337 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2338 Status Notification (bounce) messages are now MIME format per RFC 3464.
2339 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2340 under the control of the dsn_advertise_hosts option, and routers may
2341 have a dsn_lasthop option.
2343 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2344 default, modifiable by a malware= option. The list separator for
2345 the options can now be changed in the usual way. Bug 68.
2347 JH/10 The smtp_receive_timeout main option is now expanded before use.
2349 JH/11 The incoming_interface log option now also enables logging of the
2350 local interface on delivery outgoing connections.
2352 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2353 if the interface and destination host and port all match.
2355 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2358 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2359 Patch from Andrew Lewis.
2361 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2362 now supports optional time-restrictions, weighting, and priority
2363 modifiers per server. Patch originally by <rommer@active.by>.
2365 JH/16 The spamd_address main option now supports a mixed list of local
2366 and remote servers. Remote servers can be IPv6 addresses, and
2367 specify a port-range.
2369 JH/17 Bug 68: The spamd_address main option now supports an optional
2370 timeout value per server.
2372 JH/18 Bug 1581: Router and transport options headers_add/remove can
2373 now have the list separator specified.
2375 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2378 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2381 JH/21 Support for the A6 type of dns record is withdrawn.
2383 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2384 rather than the verbs used.
2386 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2387 from 255 to 1024 chars.
2389 JH/24 Verification callouts now attempt to use TLS by default.
2391 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2392 are generic router options now. The defaults didn't change.
2394 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2395 Original patch from Alexander Shikoff, worked over by JH.
2397 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2398 files if ZCAT_COMMAND is not executable.
2400 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2402 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2404 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2405 Normally benign, it bites when the pair was led to by a CNAME;
2406 modern usage is to not canonicalize the domain to a CNAME target
2407 (and we were inconsistent anyway for A-only vs AAAA+A).
2409 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2411 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2412 when evaluating $sender_host_dnssec.
2414 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2415 $sender_helo_dnssec variable.
2417 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2419 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2421 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2423 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2424 documented as working, but never had. Support all but $spam_report.
2426 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2427 added for tls authenticator.
2429 HS/03 Add perl_taintmode main config option
2435 TL/01 When running the test suite, the README says that variables such as
2436 no_msglog_check are global and can be placed anywhere in a specific
2437 test's script, however it was observed that placement needed to be near
2438 the beginning for it to behave that way. Changed the runtest perl
2439 script to read through the entire script once to detect and set these
2440 variables, reset to the beginning of the script, and then run through
2441 the script parsing/test process like normal.
2443 TL/02 The BSD's have an arc4random API. One of the functions to induce
2444 adding randomness was arc4random_stir(), but it has been removed in
2445 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2446 function when detected.
2448 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2449 cause callback expansion.
2451 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2452 syntax errors in an expansion can be treated as a string instead of
2453 logging or causing an error, due to the internal use of bool_lax
2454 instead of bool when processing it.
2456 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2457 server certificates when making smtp deliveries.
2459 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2461 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2463 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2465 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2466 Merged patch from Sebastian Wiedenroth.
2468 JH/05 Fix results-pipe from transport process. Several recipients, combined
2469 with certificate use, exposed issues where response data items split
2470 over buffer boundaries were not parsed properly. This eventually
2471 resulted in duplicates being sent. This issue only became common enough
2472 to notice due to the introduction of connection certificate information,
2473 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2475 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2476 size buffer was used, resulting in syntax errors when an expansion
2479 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2480 version 3.3.6 or later.
2482 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2483 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2484 both become "event_action", the variables become $event_name, $event_data
2485 and $event_defer_errno. There is a new variable $verify_mode, usable in
2486 routers, transports and related events. The tls:cert event is now also
2487 raised for inbound connections, if the main configuration event_action
2490 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2491 early OCSP support, but no stapling (appears to be less than 1.0.0).
2493 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2494 server certificate names available under the smtp transport option
2495 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2498 JH/10 Time-related extraction expansions from certificates now use the main
2499 option "timezone" setting for output formatting, and are consistent
2500 between OpenSSL and GnuTLS compilations. Bug 1541.
2502 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2503 encoded parameter in the incoming message. Bug 1558.
2505 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2506 include certificate info, eximon was claiming there were spoolfile
2509 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2511 JH/14 Log delivery-related information more consistently, using the sequence
2512 "H=<name> [<ip>]" wherever possible.
2514 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2515 are problematic for Debian distribution, omit them from the release
2518 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2520 JH/16 Fix string representation of time values on 64bit time_t architectures.
2523 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2524 output list separator was used.
2529 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2530 checkers that were complaining about end of non-void function with no
2533 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2534 This was a regression introduced in 4.83 by another bugfix.
2536 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2538 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2539 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2545 TF/01 Correctly close the server side of TLS when forking for delivery.
2547 When a message was received over SMTP with TLS, Exim failed to clear up
2548 the incoming connection properly after forking off the child process to
2549 deliver the message. In some situations the subsequent outgoing
2550 delivery connection happened to have the same fd number as the incoming
2551 connection previously had. Exim would try to use TLS and fail, logging
2552 a "Bad file descriptor" error.
2554 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2555 utilities have not been installed.
2557 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2558 temporary space as the ACL may create new global variables.
2560 TL/01 LDAP support uses per connection or global context settings, depending
2561 upon the detected version of the libraries at build time.
2563 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2564 to extract and use the src ip:port in logging and expansions as if it
2565 were a direct connection from the outside internet. PPv2 support was
2566 updated based on HAProxy spec change in May 2014.
2568 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2570 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2571 Properly escape header and check for NULL return.
2573 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2576 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2578 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2579 characters in header names, implemented as a verify condition.
2580 Contributed by Michael Fischer v. Mollard.
2582 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2583 results permerror and temperror. Previous values are deprecated but
2584 still accepted. In a future release, err_perm and err_temp will be
2585 completely removed, which will be a backward incompatibility if the
2586 ACL tests for either of these two old results. Patch contributed by
2587 user bes-internal on the mailing list.
2589 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2591 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2592 selectors, in both main and reject logs.
2594 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2597 JH/07 Add malware type "sock" for talking to simple daemon.
2599 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2601 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2602 routers/transports under cutthrough routing.
2604 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2605 numbers. Touch up "bool" conditional to keep the same definition.
2607 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2609 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2611 JH/12 Expand items in router/transport headers_add or headers_remove lists
2612 individually rather than the list as a whole. Bug 1452.
2614 Required for reasonable handling of multiple headers_ options when
2615 they may be empty; requires that headers_remove items with embedded
2616 colons must have them doubled (or the list-separator changed).
2618 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2619 view the policy declared in the DMARC record. Currently, $dmarc_status
2620 is a combined value of both the record presence and the result of the
2623 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2625 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2626 dnslookup router and the smtp transport (applying to the forward
2629 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2630 of ldap servers used for a specific lookup. Patch provided by Heiko
2633 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2634 New variable $lookup_dnssec_authenticated for observability.
2636 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2637 Patch submitted by Lars Timman.
2639 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2641 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2642 Requires trusted mode and valid format message id, aborts otherwise.
2643 Patch contributed by Heiko Schlichting.
2645 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2646 certextract with support for various fields. Bug 1358.
2648 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2649 is requested by default, modifiable by smtp transport option
2652 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2653 operate on certificate variables to give certificate fingerprints
2654 Also new ${sha256:cert_variable}.
2656 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2658 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2659 Christian Aistleitner.
2661 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2663 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2664 file. Patch from Wolfgang Breyha.
2666 JH/25 Expand the coverage of the delivery $host and $host_address to
2667 client authenticators run in verify callout. Bug 1476.
2669 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2670 align with daemon_smtp_ports. Bug 72.
2672 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2673 support and error reporting did not work properly.
2675 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2676 and is readable. Patch from Andrew Colin Kissa.
2678 TL/14 Enhance documentation of ${run expansion and how it parses the
2679 commandline after expansion, particularly in the case when an
2680 unquoted variable expansion results in an empty value.
2682 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2684 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2685 by renaming away from it. Addresses GH issue 15, reported by
2688 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2690 TL/15 SECURITY: prevent double expansion in math comparison functions
2691 (can expand unsanitized data). Not remotely exploitable.
2698 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2700 PP/02 Make -n do something, by making it not do something.
2701 When combined with -bP, the name of an option is not output.
2703 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2706 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2707 $sender_host_name and config options to manage this, and basic check
2710 PP/05 DSCP support for outbound connections and control modifier for inbound.
2712 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2713 (Only plugin which currently uses this is kerberos4, which nobody should
2714 be using, but we should make it available and other future plugins might
2715 conceivably use it, even though it would break NAT; stuff *should* be
2716 using channel bindings instead).
2718 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2719 name; added for Sendmail compatibility; requires admin caller.
2720 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2721 just ignore it); requires trusted caller.
2722 Also parse but ignore: -Ac -Am -X<logfile>
2725 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2727 TL/02 Add +smtp_confirmation as a default logging option.
2729 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2730 Patch by Magnus Holmgren from 2007-02-20.
2732 TL/04 Bugzilla 1281 - Spec typo.
2733 Bugzilla 1283 - Spec typo.
2734 Bugzilla 1290 - Spec grammar fixes.
2736 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2738 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2740 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2741 mailing list by by Dmitry Isaikin.
2743 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2745 JH/02 Support "G" suffix to numbers in ${if comparisons.
2747 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2749 NM/01 Bugzilla 1197 - Spec typo
2750 Bugzilla 1196 - Spec examples corrections
2752 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2754 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2755 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2758 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2759 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2761 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2762 "acl {{name}{arg}...}", and optional args on acl condition
2765 JH/05 Permit multiple router/transport headers_add/remove lines.
2767 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2769 JH/07 Avoid using a waiting database for a single-message-only transport.
2770 Performance patch from Paul Fisher. Bugzilla 1262.
2772 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2775 JH/09 Add $headers_added variable, with content from use of ACL modifier
2776 add_header (but not yet added to the message). Bugzilla 199.
2778 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2779 Pulled from Bugzilla 817 by Wolfgang Breyha.
2781 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2783 (nb: this is the same fix as in Exim 4.80.1)
2785 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2788 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2791 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2793 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2794 advertise SMTP AUTH mechanism to us, instead of a generic
2795 protocol violation error. Also, make Exim more robust to bad
2796 data from the Dovecot auth socket.
2798 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2800 When a queue runner is handling a message, Exim first routes the
2801 recipient addresses, during which it prunes them based on the retry
2802 hints database. After that it attempts to deliver the message to
2803 any remaining recipients. It then updates the hints database using
2806 So if a recipient address works intermittently, it can get repeatedly
2807 deferred at routing time. The retry hints record remains fresh so the
2808 address never reaches the final cutoff time.
2810 This is a fairly common occurrence when a user is bumping up against
2811 their storage quota. Exim had some logic in its local delivery code
2812 to deal with this. However it did not apply to per-recipient defers
2813 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2815 This change adds a proper retry rule check during routing so that the
2816 final cutoff time is checked against the message's age. We only do
2817 this check if there is an address retry record and there is not a
2818 domain retry record; this implies that previous attempts to handle
2819 the address had the retry_use_local_parts option turned on. We use
2820 this as an approximation for the destination being like a local
2821 delivery, as in LMTP.
2823 I suspect this new check makes the old local delivery cutoff check
2824 redundant, but I have not verified this so I left the code in place.
2826 TF/02 Correct gecos expansion when From: is a prefix of the username.
2828 Test 0254 submits a message to Exim with the header
2832 When I ran the test suite under the user fanf2, Exim expanded
2833 the header to contain my full name, whereas it should have added
2834 a Resent-Sender: header. It erroneously treats any prefix of the
2835 username as equal to the username.
2837 This change corrects that bug.
2839 GF/01 DCC debug and logging tidyup
2840 Error conditions log to paniclog rather than rejectlog.
2841 Debug lines prefixed by "DCC: " to remove any ambiguity.
2843 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2845 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2846 Bug spotted by Jeremy Harris; was flawed since initial commit.
2847 Would have resulted in OCSP responses post-SNI triggering an Exim
2848 NULL dereference and crash.
2850 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2852 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2853 Bug detection, analysis and fix by Samuel Thibault.
2854 Bugzilla 1331, Debian bug #698092.
2856 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2858 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2859 Server implementation by Todd Lyons, client by JH.
2860 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2861 config variable "prdr_enable" controls whether the server
2862 advertises the facility. If the client requests PRDR a new
2863 acl_data_smtp_prdr ACL is called once for each recipient, after
2864 the body content is received and before the acl_smtp_data ACL.
2865 The client is controlled by both of: a hosts_try_prdr option
2866 on the smtp transport, and the server advertisement.
2867 Default client logging of deliveries and rejections involving
2868 PRDR are flagged with the string "PRDR".
2870 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2871 fclose(). Diagnosis by Todd Lyons.
2873 PP/17 Update configure.default to handle IPv6 localhost better.
2874 Patch by Alain Williams (plus minor tweaks).
2877 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2878 This is now consistent with GnuTLS, and is now documented: the
2879 previous undocumented portable approach to treating the option as
2880 unset was to force an expansion failure. That still works, and
2881 an empty string is now equivalent.
2883 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2884 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2885 not performing validation itself.
2887 PP/20 Added force_command boolean option to pipe transport.
2888 Patch from Nick Koston, of cPanel Inc.
2890 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2893 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2895 PP/21 Fix eximon continuous updating with timestamped log-files.
2896 Broken in a format-string cleanup in 4.80, missed when I repaired the
2897 other false fix of the same issue.
2898 Report and fix from Heiko Schlichting.
2901 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2902 Report from Prashanth Katuri.
2904 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2905 It's SecureTransport, so affects any MacOS clients which use the
2906 system-integrated TLS libraries, including email clients.
2908 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2909 using a MIME ACL for non-SMTP local injection.
2910 Report and assistance in diagnosis by Warren Baker.
2912 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2914 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2916 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2917 last id that failed so it may be referenced in subsequent ACL's.
2919 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2922 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2923 ldap library initialization, allowing self-signed CA's to be
2924 used. Also properly sets require_cert option later in code by
2925 using NULL (global ldap config) instead of ldap handle (per
2926 session). Bug diagnosis and testing by alxgomz.
2928 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2929 the src/util/ subdirectory.
2931 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2932 renamed to Transport Post Delivery Action by Jeremy Harris, as
2935 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2936 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2937 redis_servers = needs to be configured which will be used by the redis
2938 lookup. Patch from Warren Baker, of The Packet Hub.
2940 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2942 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2943 hostname or reverse DNS when processing a host list. Used suggestions
2944 from multiple comments on this bug.
2946 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2948 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2949 Missed a few lines, added it to make the runtest require no keyboard
2952 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2953 contains upper case chars. Make router use caseful_local_part.
2955 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2956 support when GnuTLS has been built with p11-kit.
2962 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2964 This, or similar/improved, will also be change PP/11 of 4.82.
2970 PP/01 Handle short writes when writing local log-files.
2971 In practice, only affects FreeBSD (8 onwards).
2972 Bugzilla 1053, with thanks to Dmitry Isaikin.
2974 NM/01 Bugzilla 949 - Documentation tweak
2976 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2979 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2981 PP/02 Implemented gsasl authenticator.
2983 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2985 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2986 `pkg-config foo` for cflags/libs.
2988 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2989 with rest of GSASL and with heimdal_gssapi.
2991 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2992 `pkg-config foo` for cflags/libs for the TLS implementation.
2994 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2995 properties get this fed in as external SSF. A number of robustness
2996 and debugging improvements to the cyrus_sasl authenticator.
2998 PP/08 cyrus_sasl server now expands the server_realm option.
3000 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3001 Patch by Jeremy Harris.
3003 PP/10 Added dbmjz lookup type.
3005 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3007 PP/12 MAIL args handles TAB as well as SP, for better interop with
3008 non-compliant senders.
3009 Analysis and variant patch by Todd Lyons.
3011 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3012 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3013 Patch from Dirk Mueller <dmueller@suse.com>
3015 PP/13 tls_peerdn now print-escaped for spool files.
3016 Observed some $tls_peerdn in wild which contained \n, which resulted
3017 in spool file corruption.
3019 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3020 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3021 or write after TLS renegotiation, which otherwise led to messages
3024 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3025 as a tracking header (ie: a signed header comes before the signature).
3026 Patch from Wolfgang Breyha.
3028 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3029 comma-sep list; embedded commas doubled.
3031 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3033 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3035 Report and patch from Dmitry Banschikov.
3037 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3038 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3039 failures. We appear to now support TLS1.1+ with Exim.
3041 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3042 lets Exim select keys and certificates based upon TLS SNI from client.
3043 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3044 before an outbound SMTP session. New log_selector, +tls_sni.
3046 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3047 NULL dereference. Report and patch from Alun Jones.
3049 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3050 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3051 Not seeing resolver debug output on NetBSD, but suspect this is a
3052 resolver implementation change.
3054 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3055 Left warnings. Added "eximon gdb" invocation mode.
3057 PP/21 Defaulting "accept_8bitmime" to true, not false.
3059 PP/22 Added -bw for inetd wait mode support.
3061 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3062 locate the relevant includes and libraries. Made this the default.
3064 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3065 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3067 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3068 This may cause build issues on older platforms.
3070 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3071 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3072 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3073 Added SNI support via GnuTLS too.
3074 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3076 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3078 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3079 Applied second patch from Janne, implementing suggestion to default
3080 multiple-strings-in-record handling to match SPF spec.
3082 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3084 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3085 read-only, out of scope).
3086 Patch from Wolfgang Breyha, report from Stuart Northfield.
3088 PP/29 Fix three issues highlighted by clang analyser static analysis.
3089 Only crash-plausible issue would require the Cambridge-specific
3090 iplookup router and a misconfiguration.
3091 Report from Marcin Mirosław.
3093 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3095 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3096 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3097 As part of this, removing so much warning spew let me fix some minor
3098 real issues in debug logging.
3100 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3101 assignment on my part. Fixed.
3103 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3104 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3105 Janne Snabb (who went above and beyond: thank you).
3107 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3108 string otherwise requires a connection and a bunch more work and it's
3109 relatively easy to get wrong. Should also expose TLS library linkage
3112 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3113 64-bit ${eval} (JH/03).
3115 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3116 GNU libc to support some of the 64-bit stuff, should not lead to
3117 conflicts. Defined before os.h is pulled in, so if a given platform
3118 needs to override this, it can.
3120 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3121 protection layer was required, which is not implemented.
3122 Bugzilla 1254, patch from Wolfgang Breyha.
3124 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3125 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3126 tls_dhparam take prime identifiers. Also unbreak combination of
3127 OpenSSL+DH_params+TLSSNI.
3129 PP/39 Disable SSLv2 by default in OpenSSL support.
3135 PP/01 Solaris build fix for Oracle's LDAP libraries.
3136 Bugzilla 1109, patch from Stephen Usher.
3138 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3140 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3143 TF/02 Fix a couple more cases where we did not log the error message
3144 when unlink() failed. See also change 4.74-TF/03.
3146 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3147 lock up or crash if it happened to be inside a call to libc when it
3148 got a SIGUSR1 from exiwhat.
3150 The SIGUSR1 handler appends the current process status to the process
3151 log which is later printed by exiwhat. It used to use the general
3152 purpose logging code to do this, but several functions it calls are
3153 not safe for signals.
3155 The new output code in the SIGUSR1 handler is specific to the process
3156 log, and simple enough that it's easy to inspect for signal safety.
3157 Removing some special cases also simplifies the general logging code.
3158 Removing the spurious timestamps from the process log simplifies
3161 TF/04 Improved ratelimit ACL condition.
3163 The /noupdate option has been deprecated in favour of /readonly which
3164 has clearer semantics. The /leaky, /strict, and /readonly update modes
3165 are mutually exclusive. The update mode is no longer included in the
3166 database key; it just determines when the database is updated. (This
3167 means that when you upgrade Exim will forget old rate measurements.)
3169 Exim now checks that the per_* options are used with an update mode that
3170 makes sense for the current ACL. For example, when Exim is processing a
3171 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3172 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3173 must specify per_mail/readonly. If you omit the update mode it defaults to
3174 /leaky where that makes sense (as before) or /readonly where required.
3176 The /noupdate option is now undocumented but still supported for
3177 backwards compatibility. It is equivalent to /readonly except that in
3178 ACLs where /readonly is required you may specify /leaky/noupdate or
3179 /strict/noupdate which are treated the same as /readonly.
3181 A useful new feature is the /count= option. This is a generalization
3182 of the per_byte option, so that you can measure the throughput of other
3183 aggregate values. For example, the per_byte option is now equivalent
3184 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3186 The per_rcpt option has been generalized using the /count= mechanism
3187 (though it's more complicated than the per_byte equivalence). When it is
3188 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3189 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3190 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3191 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3192 Note that using per_rcpt with a non-readonly update mode in more than
3193 one ACL will cause the recipients to be double-counted. (The per_mail
3194 and per_byte options don't have this problem.)
3196 The handling of very low rates has changed slightly. If the computed rate
3197 is less than the event's count (usually one) then this event is the first
3198 after a long gap. In this case the rate is set to the same as this event's
3199 count, so that the first message of a spam run is counted properly.
3201 The major new feature is a mechanism for counting the rate of unique
3202 events. The new per_addr option counts the number of different
3203 recipients that someone has sent messages to in the last time period. It
3204 behaves like per_rcpt if all the recipient addresses are different, but
3205 duplicate recipient addresses do not increase the measured rate. Like
3206 the /count= option this is a general mechanism, so the per_addr option
3207 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3208 example, measure the rate that a client uses different sender addresses
3209 with the options per_mail/unique=$sender_address. There are further
3210 details in the main documentation.
3212 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3214 TF/06 Removed a few PCRE remnants.
3216 TF/07 Automatically extract Exim's version number from tags in the git
3217 repository when doing development or release builds.
3219 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3220 Bugzilla 879. Patch from Paul Fisher.
3222 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3223 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3226 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3228 PP/05 Variable $av_failed, true if the AV scanner deferred.
3229 Bugzilla 1078. Patch from John Horne.
3231 PP/06 Stop make process more reliably on build failure.
3232 Bugzilla 1087. Patch from Heiko Schlittermann.
3234 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3235 Bugzilla 1089. Patch from Heiko Schlittermann.
3237 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3238 Bugzilla 1131. Patch from Holger Weiß.
3240 PP/09 Handle IPv6 addresses with SPF.
3241 Bugzilla 860. Patch from Wolfgang Breyha.
3243 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3245 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3248 PP/11 match_* no longer expand right-hand-side by default.
3249 New compile-time build option, EXPAND_LISTMATCH_RHS.
3250 New expansion conditions, "inlist", "inlisti".
3252 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3254 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3256 PP/14 fix log_write() format string regression from TF/03.
3257 Bugzilla 1152. Patch from Dmitry Isaikin.
3263 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3265 PP/02 Harmonised TLS library version reporting; only show if debugging.
3266 Layout now matches that introduced for other libraries in 4.74 PP/03.
3268 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3270 PP/04 New "dns_use_edns0" global option.
3272 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3275 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3276 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3278 TK/01 Updated PolarSSL code to 0.14.2.
3279 Bugzilla 1097. Patch from Andreas Metzler.
3281 PP/07 Catch divide-by-zero in ${eval:...}.
3282 Fixes bugzilla 1102.
3284 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3287 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3288 format-string attack -- SECURITY: remote arbitrary code execution.
3290 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3291 time unintentionally subject to list matching rules, letting the header
3292 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3293 arbitrary string expansion). This allowed for information disclosure.
3295 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3296 INT_MIN/-1 -- value coerced to INT_MAX.
3302 NM/01 Workaround for PCRE version dependency in version reporting
3305 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3306 This fixes portability to compilers other than gcc, notably
3307 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3309 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3310 makefiles for portability to HP-UX and POSIX correctness.
3312 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3313 Also via indented variable definition in the Makefile.
3314 (Debugging by Oliver Heesakkers).
3316 PP/02 Restore caching of spamd results with expanded spamd_address.
3317 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3319 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3320 Improves build reliability. Fix from: Frank Elsner
3322 NM/02 Fix wide character breakage in the rfc2047 coding
3323 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3325 NM/03 Allow underscore in dnslist lookups
3326 Fixes bug 1026. Patch from Graeme Fowler
3328 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3329 Code patches from Adam Ciarcinski of NetBSD.
3331 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3334 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3335 is logged, to avoid truncation. Patch from John Horne.
3337 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3338 Patch from Jakob Hirsch.
3340 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3341 SQL string expansion failure details.
3342 Patch from Andrey Oktyabrski.
3344 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3345 Patch from Simon Arlott.
3347 PP/09 New lookups functionality failed to compile on old gcc which rejects
3348 extern declarations in function scope.
3349 Patch from Oliver Fleischmann
3351 PP/10 Use sig_atomic_t for flags set from signal handlers.
3352 Check getgroups() return and improve debugging.
3353 Fixed developed for diagnosis in bug 927 (which turned out to be
3356 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3357 Patch from Mark Zealey.
3359 PP/12 Bugzilla 1056: Improved spamd server selection.
3360 Patch from Mark Zealey.
3362 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3363 Based on patch from Heiko Schlittermann.
3365 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3366 Patch from Uwe Doering, sign-off by Michael Haardt.
3368 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3369 variable declaration deep within a block. Bug and patch from
3372 PP/15 lookups-Makefile IRIX compatibility coercion.
3374 PP/16 Make DISABLE_DKIM build knob functional.
3376 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3377 Patch by Simon Arlott
3379 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3380 variable argument macros. Our copy now differs from upstream.
3386 TF/01 Failure to get a lock on a hints database can have serious
3387 consequences so log it to the panic log.
3389 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3390 controlled using the smtp_confirmation log selector.
3392 TF/03 Include the error message when we fail to unlink a spool file.
3394 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3395 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3396 for maintaining out-of-tree patches for some time.
3398 PP/01 Bugzilla 139: Documentation and portability issues.
3399 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3400 Handle per-OS dynamic-module compilation flags.
3402 PP/02 Let /dev/null have normal permissions.
3403 The 4.73 fixes were a little too stringent and complained about the
3404 permissions on /dev/null. Exempt it from some checks.
3405 Reported by Andreas M. Kirchwitz.
3407 PP/03 Report version information for many libraries, including
3408 Exim version information for dynamically loaded libraries. Created
3409 version.h, now support a version extension string for distributors
3410 who patch heavily. Dynamic module ABI change.
3412 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3413 privilege escalation vulnerability whereby the Exim run-time user
3414 can cause root to append content of the attacker's choosing to
3417 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3420 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3421 If dropping privileges for untrusted macros, we disabled normal logging
3422 on the basis that it would fail; for the Exim run-time user, this is not
3423 the case, and it resulted in successful deliveries going unlogged.
3424 Fixed. Reported by Andreas Metzler.
3430 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3431 only prepend for the Resent-* case. Fixes regression introduced in
3432 Exim 4.70 by NM/22 for Bugzilla 607.
3434 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3435 increasing numbers of administrators be bitten by this.
3437 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3439 PP/03 Bugzilla 994: added openssl_options main configuration option.
3441 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3443 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3445 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3447 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3448 without a peer certificate, leading to a segfault because of an
3449 assumption that peers always have certificates. Be a little more
3450 paranoid. Problem reported by Martin Tscholak.
3452 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3453 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3454 NB: ClamAV planning to remove STREAM in "middle of 2010".
3455 CL also introduces -bmalware, various -d+acl logging additions and
3456 more caution in buffer sizes.
3458 PP/09 Implemented reverse_ip expansion operator.
3460 PP/10 Bugzilla 937: provide a "debug" ACL control.
3462 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3464 PP/12 Bugzilla 973: Implement --version.
3466 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3468 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3470 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3472 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3473 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3474 ignore trailing whitespace.
3476 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3478 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3481 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3482 Notification from Dr Andrew Aitchison.
3484 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3485 ExtendedDetectionInfo response format.
3486 Notification from John Horne.
3488 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3491 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3492 XSL and documented dependency on system catalogs, with examples of how
3495 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3498 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3499 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3500 configuration file which is writeable by the Exim user or group.
3502 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3503 of configuration files to cover files specified with the -C option if
3504 they are going to be used with root privileges, not just the default
3507 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3508 option (effectively making it always true).
3510 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3511 files to be used while preserving root privileges.
3513 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3514 that rogue child processes cannot use them.
3516 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3517 run-time user, instead of root.
3519 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3520 Exim run-time user without dropping privileges.
3522 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3523 result string, instead of calling string_vformat() twice with the same
3526 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3527 for other users. Others should always drop root privileges if they use
3528 -C on the command line, even for a whitelisted configure file.
3530 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3532 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3538 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3539 $data_path, and $header_path variables; fixed documentation bugs and
3542 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3543 exipick to access non-standard spools, including the "frozen" queue
3546 NM/01 Bugzilla 965: Support mysql stored procedures.
3547 Patch from Alain Williams
3549 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3551 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3552 Patch from Andreas Metzler
3554 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3555 Patch from Kirill Miazine
3557 NM/05 Bugzilla 671: Added umask to procmail example.
3559 JJ/03 installed exipick 20100323.0, fixing doc bug
3561 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3562 directory. Notification and patch from Dan Rosenberg.
3564 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3566 TK/02 Improve log output when DKIM signing operation fails.
3568 MH/01 Treat the transport option dkim_domain as a colon separated
3569 list, not as a single string, and sign the message with each element,
3570 omitting multiple occurences of the same signer.
3572 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3573 Bugzilla 985, 986. Patch by Simon Arlott
3575 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3576 Patch by Simon Arlott
3578 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3579 MBX locking. Notification from Dan Rosenberg.
3585 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3587 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3589 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3591 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3593 NM/04 Bugzilla 915: Flag broken perl installation during build.
3599 TK/01 Added patch by Johannes Berg that expands the main option
3600 "spamd_address" if it starts with a dollar sign.
3602 TK/02 Write list of recipients to X-Envelope-Sender header when building
3603 the mbox-format spool file for content scanning (suggested by Jakob
3606 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3607 (http://www.dcc-servers.net/) support via dccifd. Activated by
3608 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3610 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3611 by Mark Daniel Reidel <mr@df.eu>.
3613 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3614 When building exim an external PCRE library is now needed -
3615 PCRE is a system library on the majority of modern systems.
3616 See entry on PCRE_LIBS in EDITME file.
3618 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3619 conversation. Added nologin parameter to request.
3620 Patch contributed by Kirill Miazine.
3622 TF/01 Do not log submission mode rewrites if they do not change the address.
3624 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3626 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3627 log files in place. Contributed by Roberto Lima.
3629 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3631 TF/03 Bugzilla 615: When checking the local_parts router precondition
3632 after a local_part_suffix or local_part_prefix option, Exim now
3633 does not use the address's named list lookup cache, since this
3634 contains cached lookups for the whole local part.
3636 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3637 Robert Millan. Documentation is in experimental-spec.txt.
3639 TF/04 Bugzilla 668: Fix parallel build (make -j).
3641 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3643 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3644 Patch provided by Jan Srzednicki.
3646 TF/05 Leading white space used to be stripped from $spam_report which
3647 wrecked the formatting. Now it is preserved.
3649 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3650 that they are available at delivery time.
3652 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3654 TF/08 TLS error reporting now respects the incoming_interface and
3655 incoming_port log selectors.
3657 TF/09 Produce a more useful error message if an SMTP transport's hosts
3658 setting expands to an empty string.
3660 NM/06 Bugzilla 744: EXPN did not work under TLS.
3661 Patch provided by Phil Pennock.
3663 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3664 Patch provided by Richard Godbee.
3666 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3667 acl_smtp_notquit, added index entry.
3669 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3670 Patch provided by Eugene Bujak.
3672 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3673 accept(). Patch provided by Maxim Dounin.
3675 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3676 Patch provided by Phil Pennock.
3678 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3680 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3681 Patch provided by Brad "anomie" Jorsch.
3683 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3684 Patch provided by Dean Brooks.
3686 TK/05 Add native DKIM support (does not depend on external libraries).
3688 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3689 Patch provided by Graeme Fowler.
3691 NM/16 Bugzilla 851: Documentation example syntax fix.
3693 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3695 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3698 NM/19 Bugzilla 745: TLS version reporting.
3699 Patch provided by Phil Pennock.
3701 NM/20 Bugzilla 167: bool: condition support.
3702 Patch provided by Phil Pennock.
3704 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3705 clients. Patch provided by Phil Pennock.
3707 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3708 Patch provided by Brad "anomie" Jorsch.
3710 NM/23 Bugzilla 687: Fix misparses in eximstats.
3711 Patch provided by Heiko Schlittermann.
3713 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3714 Patch provided by Heiko Schlittermann.
3716 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3717 plus update to original patch.
3719 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3721 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3722 Patch provided by David Brownlee.
3724 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3726 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3728 NM/30 Bugzilla 888: TLS documentation bugfixes.
3730 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3732 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3733 Unlike the original bugzilla I have changed all shell scripts in src tree.
3735 NM/33 Bugzilla 898: Transport filter timeout fix.
3736 Patch by Todd Rinaldo.
3738 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3739 Patch by Serge Demonchaux.
3741 NM/35 Bugzilla 39: Base64 decode bug fixes.
3742 Patch by Jakob Hirsch.
3744 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3746 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3748 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3750 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3756 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3757 ALT-N's libdkim that I have put here:
3758 http://duncanthrax.net/exim-experimental/
3760 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3761 were called 'true' and it seems that C99 defines that as a reserved
3762 keyword to be used with 'bool' variable types. That means you could
3763 not include C99-style headers which use bools without triggering
3764 build errors in sieve.c.
3766 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3767 as mailq or other aliases. Changed the --help handling significantly
3768 to do whats expected. exim_usage() emits usage/help information.
3770 SC/01 Added the -bylocaldomain option to eximstats.
3772 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3774 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3776 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3782 PH/01 Another patch from the Sieve maintainer.
3784 PH/02 When an IPv6 address is converted to a string for single-key lookup
3785 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3786 dots are used instead of colons so that keys in lsearch files need not
3787 contain colons. This was done some time before quoting was made available
3788 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3789 (notated using the quote facility) so as to distinguish them from IPv4
3790 keys. This meant that lookups for IP addresses in host lists did not work
3791 for iplsearch lookups.
3793 This has been fixed by arranging for IPv6 addresses to be expressed with
3794 colons if the lookup type is iplsearch. This is not incompatible, because
3795 previously such lookups could never work.
3797 The situation is now rather anomalous, since one *can* have colons in
3798 ordinary lsearch keys. However, making the change in all cases is
3799 incompatible and would probably break a number of configurations.
3801 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3804 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3805 conversion specification without a maximum field width, thereby enabling
3806 a rogue spamd server to cause a buffer overflow. While nobody in their
3807 right mind would setup Exim to query an untrusted spamd server, an
3808 attacker that gains access to a server running spamd could potentially
3809 exploit this vulnerability to run arbitrary code as the Exim user.
3811 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3812 $primary_hostname instead of what libspf2 thinks the hosts name is.
3814 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3815 a directory entry by the name of the lookup key. Previously, if a
3816 symlink pointed to a non-existing file or a file in a directory that
3817 Exim lacked permissions to read, a lookup for a key matching that
3818 symlink would fail. Now it is enough that a matching directory entry
3819 exists, symlink or not. (Bugzilla 503.)
3821 PH/03 The body_linecount and body_zerocount variables are now exported in the
3824 PH/04 Added the $dnslist_matched variable.
3826 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3827 This means they are set thereafter only if the connection becomes
3830 PH/06 Added the client_condition to authenticators so that some can be skipped
3831 by clients under certain conditions.
3833 PH/07 The error message for a badly-placed control=no_multiline_responses left
3834 "_responses" off the end of the name.
3836 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3838 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3839 (without spaces) instead of just copying the configuration text.
3841 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3843 PH/11 Added $max_received_linelength.
3845 PH/12 Added +ignore_defer and +include_defer to host lists.
3847 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3848 way in which PCRE > 7.0 is built.
3850 PH/14 Implemented queue_only_load_latch.
3852 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3853 MAIL command. The effect was to mangle the value on 64-bit systems.
3855 PH/16 Another patch from the Sieve maintainer.
3857 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3859 PH/18 If a system quota error occurred while trying to create the file for
3860 a maildir delivery, the message "Mailbox is full" was not appended to the
3861 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3862 only to a quota excession during the actual writing of the file.
3864 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3865 characters?) which causes problems in log lines. The DN values are now
3866 passed through string_printing() before being added to log lines.
3868 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3869 and InterBase are left for another time.)
3871 PH/21 Added message_body_newlines option.
3873 PH/22 Guard against possible overflow in moan_check_errorcopy().
3875 PH/23 POSIX allows open() to be a macro; guard against that.
3877 PH/24 If the recipient of an error message contained an @ in the local part
3878 (suitably quoted, of course), incorrect values were put in $domain and
3879 $local_part during the evaluation of errors_copy.
3885 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3886 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3889 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3890 issue a MAIL command.
3892 PH/02 In an ACL statement such as
3894 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3896 if a client was not listed at all, or was listed with a value other than
3897 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3898 the condition was not true (as it should be), so access was not denied.
3899 The bug was that the ! inversion was incorrectly passed on to the second
3900 item. This has been fixed.
3902 PH/03 Added additional dnslists conditions == and =& which are different from
3903 = and & when the dns lookup returns more than one IP address.
3905 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3906 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3908 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3909 FSYNC, which compiles an option called disable_fsync that allows for
3910 bypassing fsync(). The documentation is heavily laced with warnings.
3912 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3914 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3915 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3916 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3917 including adding "make clean"; (3) Added -fPIC when compiling the test
3918 dynamically loaded module, to get rid of a warning.
3920 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3921 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3922 = 0s. The bug is otherwise harmless.
3924 PH/07 There was a bug in the dovecot authenticator such that the value of
3925 $auth1 could be overwritten, and so not correctly preserved, after a
3926 successful authentication. This usually meant that the value preserved by
3927 the server_setid option was incorrect.
3929 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3931 PH/09 Installed PCRE release 7.0.
3933 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3934 run for batched SMTP input. It is now run at the start of every message
3935 in the batch. While fixing this I discovered that the process information
3936 (output by running exiwhat) was not always getting set for -bs and -bS
3937 input. This is fixed, and it now also says "batched" for BSMTP.
3939 PH/11 Added control=no_pipelining.
3941 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3942 patch, slightly modified), and move the expansion of helo_data till after
3943 the connection is made in the smtp transport (so it can use these
3946 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3948 PH/14 Added log_selector = +pid.
3950 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3952 PH/16 Add ${if forany and ${if forall.
3954 PH/17 Added dsn_from option to vary the From: line in DSNs.
3956 PH/18 Flush SMTP output before performing a callout, unless control =
3957 no_callout_flush is set.
3959 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3960 was true (the default) a successful delivery failed to delete the retry
3961 item, thus causing premature timeout of the address. The bug is now
3964 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3966 PH/21 Long custom messages for fakedefer and fakereject are now split up
3967 into multiline responses in the same way that messages for "deny" and
3968 other ACL rejections are.
3970 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3971 with slight modification.
3973 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3974 draft, changing the syntax and factoring some duplicate code".
3976 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3977 for deliveries of the second and subsequent messages over the same SMTP
3980 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3981 ${reduce, with only minor "tidies".
3983 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3985 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3986 expansion side effects.
3988 PH/27 When a message times out after an over-quota error from an Exim-imposed
3989 quota, the bounce message says "mailbox is full". This message was not
3990 being given when it was a system quota that was exceeded. It now should
3993 MH/03 Made $recipients available in local_scan(). local_scan() already has
3994 better access to the recipient list through recipients_list[], but
3995 $recipients can be useful in postmaster-provided expansion strings.
3997 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3998 in the case of a MAIL command with additional options following the
3999 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4000 were accidentally chopped off.
4002 PH/29 SMTP synchronization checks are implemented when a command is read -
4003 there is a check that no more input is waiting when there shouldn't be
4004 any. However, for some commands, a delay in an ACL can mean that it is
4005 some time before the response is written. In this time, more input might
4006 arrive, invalidly. So now there are extra checks after an ACL has run for
4007 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4008 pipelining has not been advertised.
4010 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4012 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4013 returned in the SMTP response when smtp_return_error_details was false.
4014 This has been fixed.
4016 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4017 instead of the C I/O that was originally supplied, because problems were
4018 reported on Solaris.
4020 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4021 Exim which did not show up earlier: it was assuming that a call to
4022 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4023 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4024 was a macro that became an assignment, so it seemed to work. This has
4025 changed to a proper function call with a void return, hence the compile
4026 error. Exim's code has been fixed.
4028 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4031 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4033 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4035 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4036 because it turns out to be too big or there is a timeout) there is no
4037 "Completed" line in the log. When some messages of this type were
4038 selected by exigrep, they were listed as "not completed". Others were
4039 picked up by some special patterns. I have improved the selection
4040 criteria to be more general.
4042 PH/38 The host_find_failed option in the manualroute router can now be set
4043 to "ignore", to completely ignore a host whose IP address cannot be
4044 found. If all hosts are ignored, the behaviour is controlled by the new
4045 host_all_ignored option.
4047 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4048 homing or because of multiple MX records with /mx) generated more than
4049 one IP address, and the following item turned out to be the local host,
4050 all the secondary addresses of the first item were incorrectly removed
4051 from the list, along with the local host and any following hosts (which
4052 is what is supposed to happen).
4054 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4055 whoever called Exim into the -H file. In the case of the daemon it was
4056 behaving confusingly. When first started, it used values for whoever
4057 started the daemon, but after a SIGHUP it used the Exim user (because it
4058 calls itself on a restart). I have changed the code so that it now always
4061 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4062 message are rejected with the same error (e.g. no authentication or bad
4063 sender address), and a DATA command is nevertheless sent (as can happen
4064 with PIPELINING or a stupid MUA), the error message that was given to the
4065 RCPT commands is included in the rejection of the DATA command. This is
4066 intended to be helpful for MUAs that show only the final error to their
4069 PH/42 Another patch from the Sieve maintainer.
4071 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4072 Eximstats - Fixed some broken HTML links and added missing column headers
4074 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4075 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4077 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4079 PH/43 Yet another patch from the Sieve maintainer.
4081 PH/44 I found a way to check for a TCP/IP connection going away before sending
4082 the response to the final '.' that terminates a message, but only in the
4083 case where the client has not sent further data following the '.'
4084 (unfortunately, this is allowed). However, in many cases there won't be
4085 any further data because there won't be any more messages to send. A call
4086 to select() can be used: if it shows that the input is "ready", there is
4087 either input waiting, or the socket has been closed. An attempt to read
4088 the next input character can distinguish the two cases. Previously, Exim
4089 would have sent an OK response which the client would never have see.
4090 This could lead to message repetition. This fix should cure that, at
4091 least in a lot of common cases.
4093 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4094 advertised in response to EHLO.
4100 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4101 fixed by 4.65/MH/01 (is this a record?) are fixed:
4103 (i) An empty string was always treated as zero by the numeric comparison
4104 operators. This behaviour has been restored.
4106 (ii) It is documented that the numeric comparison operators always treat
4107 their arguments as decimal numbers. This was broken in that numbers
4108 starting with 0 were being interpreted as octal.
4110 While fixing these problems I realized that there was another issue that
4111 hadn't been noticed. Values of message_size_limit (both the global option
4112 and the transport option) were treated as octal if they started with 0.
4113 The documentation was vague. These values are now always treated as
4114 decimal, and I will make that clear in the documentation.
4120 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4121 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4124 MH/01 Don't check that the operands of numeric comparison operators are
4125 integers when their expansion is in "skipping" mode (fixes bug
4126 introduced by 4.64-PH/07).
4128 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4129 child addresses, Exim now panics and dies. Previously, because the count
4130 is held in a short int, deliveries were likely to be lost. As such a
4131 large number of recipients for a single message is ridiculous
4132 (performance will be very, very poor), I have chosen to impose a limit
4133 rather than extend the field.
4139 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4140 leftover -K file (the existence of which was triggered by #402).
4141 While we were at it, introduced process PID as part of the -K
4142 filename. This should rule out race conditions when creating
4145 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4146 processing considerably. Previous code took too long for large mails,
4147 triggering a timeout which in turn triggers #401.
4149 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4150 in the DK code in transports.c. sendfile() is not really portable,
4151 hence the _LINUX specificness.
4153 TF/01 In the add_headers option to the mail command in an Exim filter,
4154 there was a bug that Exim would claim a syntax error in any
4155 header after the first one which had an odd number of characters
4158 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4159 callout verification, Exim cached a "reject" for the entire domain. This
4160 is correct for most verifications, but it is not correct for a recipient
4161 verification with use_sender or use_postmaster set, because in that case
4162 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4163 case of MAIL FROM:<> rejection from other early rejections (e.g.
4164 rejection of HELO). When verifying a recipient using a non-null MAIL
4165 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4166 Whatever the result of the callout, the value of the domain cache is
4167 left unchanged (for any other kind of callout, getting as far as trying
4168 RCPT means that the domain itself is ok).
4170 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4173 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4174 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4175 session, the daemon ignores ECONNECT errors and logs others; it now
4176 ignores EPIPE as well.
4178 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4179 (quoted-printable decoding).
4181 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4182 later the small subsequent patch to fix an introduced bug.
4184 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4186 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4188 PH/08 An error is now given if message_size_limit is specified negative.
4190 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4191 to be given (somewhat) arbitrary names.
4193 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4196 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4197 miscellaneous code fixes
4199 PH/10 Added the log_reject_target ACL modifier to specify where to log
4202 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4203 hostname. This is wrong, because it relates to the incoming message (and
4204 probably the interface on which it is arriving) and not to the outgoing
4205 callout (which could be using a different interface). This has been
4206 changed to use the value of the helo_data option from the smtp transport
4207 instead - this is what is used when a message is actually being sent. If
4208 there is no remote transport (possible with a router that sets up host
4209 addresses), $smtp_active_hostname is used.
4211 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4212 tweaks were necessary in order to get it to work (see also 21 below):
4213 (a) The code assumed that strncpy() returns a negative number on buffer
4214 overflow, which isn't the case. Replaced with Exim's string_format()
4216 (b) There were several signed/unsigned issues. I just did the minimum
4217 hacking in of casts. There is scope for a larger refactoring.
4218 (c) The code used strcasecmp() which is not a standard C function.
4219 Replaced with Exim's strcmpic() function.
4220 (d) The code set only $1; it now sets $auth1 as well.
4221 (e) A simple test gave the error "authentication client didn't specify
4222 service in request". It would seem that Dovecot has changed its
4223 interface. Fortunately there's a specification; I followed it and
4224 changed what the client sends and it appears to be working now.
4226 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4229 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4230 address A is aliased to B and C, where B exists and C does not. Without
4231 -v the output is "A verified" because verification stops after a
4232 successful redirection if more than one address is generated. However,
4233 with -v the child addresses are also verified. Exim was outputting "A
4234 failed to verify" and then showing the successful verification for C,
4235 with its parentage. It now outputs "B failed to verify", showing B's
4236 parentage before showing the successful verification of C.
4238 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4239 look up a TXT record in a specific list after matching in a combined
4242 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4243 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4244 they consult the DNS. I had assumed they would set it the way they
4245 wanted; and indeed my experiments on Linux seem to show that in some
4246 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4247 To be on the safe side, however, I have now made the interface to
4248 host_find_byname() similar to host_find_bydns(), with an argument
4249 containing the DNS resolver options. The host_find_byname() function now
4250 sets these options at its start, just as host_find_bydns() does. The smtp
4251 transport options dns_qualify_single and dns_search_parents are passed to
4252 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4253 of host_find_byname() use the default settings of RES_DEFNAMES
4254 (qualify_single) but not RES_DNSRCH (search_parents).
4256 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4257 spool_read_header() do less string testing, by means of a preliminary
4258 switch on the second character of optional "-foo" lines. (This is
4259 overdue, caused by the large number of possibilities that now exist.
4260 Originally there were few.) While I was there, I also converted the
4261 str(n)cmp tests so they don't re-test the leading "-" and the first
4262 character, in the hope this might squeeze out yet more improvement.
4264 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4265 flag allowing group syntax was set by the header_syntax check but not
4266 turned off, possible causing trouble later; (2) The flag was not being
4267 set at all for the header_verify test, causing "group"-style headers to
4268 be rejected. I have now set it in this case, and also caused header_
4269 verify to ignore an empty address taken from a group. While doing this, I
4270 came across some other cases where the code for allowing group syntax
4271 while scanning a header line wasn't quite right (mostly, not resetting
4272 the flag correctly in the right place). These bugs could have caused
4273 trouble for malformed header lines. I hope it is now all correct.
4275 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4276 with the "reply" argument non-NULL. The code, however (which originally
4277 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4278 but it didn't always do it. This confused somebody who was copying the
4279 code for some other use. I have removed all the tests.
4281 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4282 feature that was used to support insecure browsers during the U.S. crypto
4283 embargo. It requires special client support, and Exim is probably the
4284 only MTA that supported it -- and would never use it because real RSA is
4285 always available. This code has been removed, because it had the bad
4286 effect of slowing Exim down by computing (never used) parameters for the
4287 RSA_EXPORT functionality.
4289 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4290 authenticator to fail if there's a tab character in the incoming data
4291 (there should never be unless someone is messing about, as it's supposed
4292 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4293 if the connection is using TLS or if the remote IP is the same as the
4294 local IP, and the "valid-client-cert option" if a client certificate has
4297 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4298 authenticators. This can be used for authorization after authentication
4299 succeeds. (In the case of plaintext, it servers for both authentication
4302 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4303 if any retry times were supplied.
4305 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4306 connection, where there is no client IP address to check. In this
4307 situation, the verify now always succeeds.
4309 PH/25 Applied John Jetmore's -Mset patch.
4311 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4313 PH/27 In a string expansion for a processed (not raw) header when multiple
4314 headers of the same name were present, leading whitespace was being
4315 removed from all of them, but trailing whitespace was being removed only
4316 from the last one. Now trailing whitespace is removed from each header
4317 before concatenation. Completely empty headers in a concatenation (as
4318 before) are ignored.
4320 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4321 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4323 PH/29 [Removed. This was a change that I later backed out, and forgot to
4324 correct the ChangeLog entry (that I had efficiently created) before
4325 committing the later change.]
4327 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4328 address errors (4xx response to RCPT) over the same connection as other
4329 messages routed to the same hosts. Such deliveries are always "forced",
4330 so retry times are not inspected. This resulted in far too many retries
4331 for the affected addresses. The effect occurred only when there were more
4332 hosts than the hosts_max_try setting in the smtp transport when it had
4333 the 4xx errors. Those hosts that it had tried were not added to the list
4334 of hosts for which the message was waiting, so if all were tried, there
4335 was no problem. Two fixes have been applied:
4337 (i) If there are any address or message errors in an SMTP delivery, none
4338 of the hosts (tried or untried) are now added to the list of hosts
4339 for which the message is waiting, so the message should not be a
4340 candidate for sending over the same connection that was used for a
4341 successful delivery of some other message. This seems entirely
4342 reasonable: after all the message is NOT "waiting for some host".
4343 This is so "obvious" that I'm not sure why it wasn't done
4344 previously. Hope I haven't missed anything, but it can't do any
4345 harm, as the worst effect is to miss an optimization.
4347 (ii) If, despite (i), such a delivery is accidentally attempted, the
4348 routing retry time is respected, so at least it doesn't keep
4349 hammering the server.
4351 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4352 in ${readsocket because some servers need this prod.
4354 PH/32 Added some extra debug output when updating a wait-xxx database.
4356 PH/33 The hint "could be header name not terminated by colon", which has been
4357 given for certain expansion errors for a long time, was not being given
4358 for the ${if def:h_colon_omitted{... case.
4360 PH/34 The spec says: "With one important exception, whenever a domain list is
4361 being scanned, $domain contains the subject domain." There was at least
4362 one case where this was not true.
4364 PH/35 The error "getsockname() failed: connection reset by peer" was being
4365 written to the panic log as well as the main log, but it isn't really
4366 panic-worthy as it just means the connection died rather early on. I have
4367 removed the panic log writing for the ECONNRESET error when getsockname()
4370 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4371 runs only) independently of the message's sender address. This meant
4372 that, if the 4xx error was in fact related to the sender, a different
4373 message to the same recipient with a different sender could confuse
4374 things. In particular, this can happen when sending to a greylisting
4375 server, but other circumstances could also provoke similar problems.
4376 I have changed the default so that the retry time for these errors is now
4377 based a combination of the sender and recipient addresses. This change
4378 can be overridden by setting address_retry_include_sender=false in the
4381 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4382 remote server are returned as part of bounce messages. This was not
4383 happening for LMTP over a pipe (the lmtp transport), but now it is the
4384 same for both kinds of LMTP.
4386 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4387 in header lines that were in fact CNAMEs. This is no longer the case.
4389 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4390 and queue runs started by the daemon processed all messages. This has
4391 been fixed so that -R and -S can now usefully be given with -q<time>.
4393 PH/40 Import PCRE release 6.7 (fixes some bugs).
4395 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4397 PH/42 Give an error if -q is specified more than once.
4399 PH/43 Renamed the variables $interface_address and $interface_port as
4400 $received_ip_address and $received_port, to make it clear that these
4401 values apply to message reception, and not to the outgoing interface when
4402 a message is delivered. (The old names remain recognized, of course.)
4404 PH/44 There was no timeout on the connect() call when using a Unix domain
4405 socket in the ${readsocket expansion. There now is.
4407 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4408 be meaningful with "accept".
4410 SC/01 Eximstats V1.43
4411 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4413 SC/02 Eximstats V1.44
4414 Use a glob alias rather than an array ref in the generated
4415 parser. This improves both readability and performance.
4417 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4418 Collect SpamAssassin and rejection statistics.
4419 Don't display local sender or destination tables unless
4420 there is data to show.
4421 Added average volumes into the top table text output.
4423 SC/04 Eximstats V1.46
4424 Collect data on the number of addresses (recipients)
4425 as well as the number of messages.
4427 SC/05 Eximstats V1.47
4428 Added 'Message too big' to the list of mail rejection
4429 reasons (thanks to Marco Gaiarin).
4431 SC/06 Eximstats V1.48
4432 Mainlog lines which have GMT offsets and are too short to
4433 have a flag are now skipped.
4435 SC/07 Eximstats V1.49 (Alain Williams)
4436 Added the -emptyok flag.
4438 SC/08 Eximstats V1.50
4439 Fixes for obtaining the IP address from reject messages.
4441 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4442 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4443 whitespace changes from 4.64-PH/27
4445 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4448 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4449 are found, allow negative numbers in numeric criteria)
4451 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4453 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4454 to match changes made in 4.64-PH/43
4456 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4458 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4459 "domain" is actually an IP address, and force a failure. This locks out
4460 those revolvers/nameservers that support "A-for-A" lookups, in
4461 contravention of the specifications.
4463 PH/48 When a host name was looked up from an IP address, and the subsequent
4464 forward lookup of the name timed out, the host name was left in
4465 $sender_host_name, contrary to the specification.
4467 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4468 restricted to single-key lookups, Exim was not diagnosing an error if
4469 * or *@ was used with a query-style lookup.
4471 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4473 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4474 long ago, but noone interested enough thought of it. Let's just say that
4475 the "1.1" means that there are some new functions that weren't there at
4476 some point in the past.
4478 PH/51 Error processing for expansion failure of helo_data from an smtp
4479 transport during callout processing was broken.
4481 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4482 tested/used via the -bh/-bhc/-bs options.
4484 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4485 bug, fixed in subsequent PCRE releases).
4487 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4488 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4490 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4496 SC/01 Use a glob alias rather than an array ref in eximstats generated
4497 parser. This improves both readability and performance.
4499 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4500 Don't display local sender or destination tables in eximstats unless
4501 there is data to show.
4502 Added average volumes into the eximstats top table text output.
4504 SC/03 Collect data on the number of addresses (recipients) as well
4505 as the number of messages in eximstats.
4507 TF/01 Correct an error in the documentation for the redirect router. Exim
4508 does (usually) call initgroups() when daemonizing.
4510 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4511 with consistent privilege compared to when running as a daemon.
4513 TF/03 Note in the spec that $authenticated_id is not set for local
4514 submissions from trusted users.
4516 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4517 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4519 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4520 by adding some example configuration directives to the default
4521 configuration file. A little bit of work is required to uncomment the
4522 directives and define how usernames and passwords are checked, but
4523 there is now a framework to start from.
4525 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4526 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4527 without this. I don't know how relevant this is to other LDAP libraries.
4529 PH/02 Add the verb name to the "unknown ACL verb" error.
4531 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4533 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4535 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4536 directory not expanded when it should be if an expanded home directory
4537 was set for the address (which is overridden by the transport).
4539 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4542 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4543 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4544 because it is too late at that time, and has no effect.
4546 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4547 security issue with \' (bugzilla #107). I could not use the
4548 PQescapeStringConn() function, because it needs a PGconn value as one of
4551 PH/08 When testing addresses using -bt, indicate those final addresses that
4552 are duplicates that would not cause an additional delivery. At least one
4553 person was confused, thinking that -bt output corresponded to deliveries.
4554 (Suppressing duplicates isn't a good idea as you lose the information
4555 about possibly different redirections that led to the duplicates.)
4557 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4558 systems where poll() doesn't work, in particular OS X.
4560 PH/10 Added more information to debugging output for retry time not reached.
4562 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4563 operations in malware.c.
4565 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4568 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4569 syslog_duplication set false, log lines that would normally be written
4570 both the the main log and to the reject log were not written to syslog at
4573 PH/14 In the default configuration, change the use of "message" in ACL warn
4574 statements to "add_header".
4576 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4577 not followed by a command (e.g. "seen endif").
4579 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4580 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4583 PH/17 Added extra conditions to the default value of delay_warning_condition
4587 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4588 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4589 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4592 The Auto-Submitted: and various List- headers are standardised, whereas I
4593 don't think Precedence: ever was.
4595 PH/18 Refactored debugging code in route_finduser() to show more information,
4596 in particular, the error code if getpwnam() issues one.
4598 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4599 This is apparently needed in addition to the PH/07 change above to avoid
4600 any possible encoding problems.
4602 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4603 but not after initializing Perl.
4605 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4606 output them only if debugging. By default they are written stderr,
4607 apparently, which is not desirable.
4609 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4612 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4615 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4617 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4618 authenticated or an ident call has been made. Suppress the default
4619 values for $authenticated_id and $authenticated_sender (but permit -oMai
4620 and -oMas) when testing with -bh.
4622 PH/24 Re-jigged the order of the tests in the default configuration so that the
4623 tests for valid domains and recipients precede the DNS black list and CSA
4624 tests, on the grounds that those ones are more expensive.
4626 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4627 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4628 This bug exists in every version of Exim that I still have, right back to
4631 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4632 However, an attempt to turn on case-sensitivity in a regex key by
4633 including (?-i) didn't work because the subject string was already
4634 lowercased, and the effects were non-intuitive. It turns out that a
4635 one-line patch can be used to allow (?-i) to work as expected.
4641 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4642 other effects) broke the use of negated acl sub-conditions.
4644 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4647 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4648 "Deny" causes Exim to reject the incoming connection with a 554 error.
4649 Unfortunately, if there is a major crisis, such as a disk failure,
4650 tcp-wrappers gives "deny", whereas what one would like would be some
4651 kind of temporary error. A kludge has been added to help with this.
4652 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4653 554 error is used if errno is still zero or contains ENOENT (which occurs
4654 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4657 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4659 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4660 errors. Otherwise a message that provokes a temporary error (when other
4661 messages do not) can cause a whole host to time out.
4663 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4664 addresses were routed directly to files or pipes from a redirect router.
4665 File deliveries just didn't batch; pipe deliveries might have suffered
4668 PH/06 A failure to get a lock for a hints database would erroneously always say
4669 "Failed to get write lock", even when it was really a read lock.
4671 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4672 of 0600. This has been changed to use the value of the lockfile_mode
4673 option (which defaults to 0600).
4675 PH/08 Applied small patch from the Sieve maintainer.
4677 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4678 folder from quota calculations, a direct delivery into this folder messed
4679 up the contents of the maildirsize file. This was because the regex was
4680 used only to exclude .Trash (or whatever) when the size of the mailbox
4681 was calculated. There was no check that a delivery was happening into an
4682 excluded directory. This bug has been fixed by ignoring all quota
4683 processing for deliveries into excluded directories.
4685 PH/10 Added the maildirfolder_create_regex option to appendfile.
4691 PH/01 The code for finding all the local interface addresses on a FreeBSD
4692 system running IPv6 was broken. This may well have applied to all BSD
4693 systems, as well as to others that have similar system calls. The broken
4694 code found IPv4 interfaces correctly, but gave incorrect values for the
4695 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4696 that it would not match correctly against @[] and not recognize the IPv6
4699 PH/02 The ipliteral router was not recognizing addresses of the form user@
4700 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4702 PH/03 Added disable_ipv6.
4704 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4705 decoded form, because it is most often used to construct To: headers
4706 lines in autoreplies, and the decoded form may well be syntactically
4707 invalid. However, $reply_address has leading white space removed, and all
4708 newlines turned into spaces so that the autoreply transport does not
4711 PH/05 If group was specified without a user on a router, and no group or user
4712 was specified on a transport, the group from the router was ignored.
4714 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4715 for visible compile-time settings that can be used to change these
4716 numbers, for those that want even more. Backwards compatibility with old
4717 spool files has been maintained. However, going back to a previous Exim
4718 release will lost any variables that are in spool files.
4720 PH/07 Two small changes when running in the test harness: increase delay when
4721 passing a TCP/IP connection to a new process, in case the original
4722 process has to generate a bounce, and remove special handling of
4723 127.0.0.2 (sic), which is no longer necessary.
4725 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4726 be the same on different OS.
4728 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4731 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4732 whether --show-vars was specified or not
4734 JJ/02 exipick: Added support for new ACL variable spool format introduced
4737 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4738 syntactically invalid From: or Reply-to: line, and a filter used this to
4739 generate an autoreply, and therefore failed to obtain an address for the
4740 autoreply, Exim could try to deliver to a non-existent relative file
4741 name, causing unrelated and misleading errors. What now happens is that
4742 it logs this as a hard delivery error, but does not attempt to create a
4745 PH/11 The exinext utility has a -C option for testing purposes, but although
4746 the given file was scanned by exinext itself; it wasn't being passed on
4747 when Exim was called.
4749 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4750 an end-of-file indication when reading a command response.
4752 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4753 compiled. In many other places in Exim, IPv6 addresses are always
4754 recognized, so I have changed this. It also means that IPv4 domain
4755 literals of the form [IPV4:n.n.n.n] are now always recognized.
4757 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4758 used if the router is not running as root, for example, when verifying at
4759 ACL time, or when using -bh. The debugging output from this situation was
4760 non-existent - all you got was a failure to exec. I have made two
4763 (a) Failures to set uid/gid, the current directory, or a process leader
4764 in a subprocess such as that created by queryprogram now generate
4765 suitable debugging output when -d is set.
4767 (b) The queryprogram router detects when it is not running as root,
4768 outputs suitable debugging information if -d is set, and then runs
4769 the subprocess without attempting to change uid/gid.
4771 PH/15 Minor change to Makefile for building test_host (undocumented testing
4774 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4775 additional section of a DNS packet that returns MX or SRV records.
4776 Instead, it always explicitly searches for A/AAAA records. This avoids
4777 major problems that occur when a DNS server includes only records of one
4778 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4779 fixed another bug: if SRV records were looked up and the corresponding
4780 address records were *not* found in the additional section, the port
4781 values from the SRV records were lost.
4783 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4784 using the correct key (the original address) when searching the retry
4785 rules in order to find which one to use for generating the retry hint.
4787 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4788 adding the default one. Similarly, if it contains a Reply-To: header, the
4789 errors_reply_to option, if set, is not used.
4791 PH/19 When calculating a retry time, Exim used to measure the "time since
4792 failure" by looking at the "first failed" field in the retry record. Now
4793 it does not use this if it is later than than the arrival time of the
4794 message. Instead it uses the arrival time. This makes for better
4795 behaviour in cases where some deliveries succeed, thus re-setting the
4796 "first failed" field. An example is a quota failure for a huge message
4797 when small messages continue to be delivered. Without this change, the
4798 "time since failure" will always be short, possible causing more frequent
4799 delivery attempts for the huge message than are intended.
4800 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4802 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4803 $1, $2, $3) because the numerical variables can be reset during some
4804 expansion items (e.g. "match"), thereby losing the authentication data.
4806 PH/21 Make -bV show the size of off_t variables so that the test suite can
4807 decide whether to run tests for quotas > 2G.
4809 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4810 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4811 filecount value is greater than 2G or if a quota value is greater than 2G
4812 on a system where the size of off_t is not greater than 4, a panic error
4815 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4816 never match. The debug and -bh output now contains an explicit error
4817 message indicating a malformed IPv4 address or mask.
4819 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4820 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4821 PH/23 above applies.
4823 PH/25 Do not write to syslog when running in the test harness. The only
4824 occasion when this arises is a failure to open the main or panic logs
4825 (for which there is an explicit test).
4827 PH/26 Added the /no_tell option to "control=freeze".
4829 PH/27 If a host name lookup failed very early in a connection, for example, if
4830 the IP address matched host_lookup and the reverse lookup yielded a name
4831 that did not have a forward lookup, an error message of the form "no IP
4832 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4833 could be logged. Now it outputs the IP address instead of "NULL".
4835 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4836 allows the sender and the authenticated sender to be set when
4837 submitting a message from within Exim. Since child_open_exim() is
4838 documented for local_scan(), the new function should be too.
4840 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4841 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4842 results in an empty string is now treated as unset.
4844 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4846 PH/31 Added sender_verify_fail logging option.
4848 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4849 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4850 tidied the source and removed it altogether.
4852 PH/33 When a queue run was abandoned because the load average was too high, a
4853 log line was always written; now it is written only if the queue_run log
4854 selector is set. In addition, the log line for abandonment now contains
4855 information about the queue run such as the pid. This is always present
4856 in "start" and "stop" lines but was omitted from the "abandon" line.
4858 PH/34 Omit spaces between a header name and the colon in the error message that
4859 is given when verify = headers_syntax fails (if there are lots of them,
4860 the message gets confusing).
4862 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4863 names, as there are now some PTR records that contain slashes. This check
4864 is only to protect against broken name servers that fall over on strange
4865 characters, so the fact that it applies to all lookups doesn't matter.
4867 PH/36 Now that the new test suite is complete, we can remove some of the
4868 special code in Exim that was needed for the old test suite. For example,
4869 sorting DNS records because real resolvers return them in an arbitrary
4870 order. The new test suite's fake resolver always returns records in the
4873 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4874 bounces) except when queue_only is set, to avoid logging races between
4875 the different processes.
4877 PH/38 Panic-die if .include specifies a non-absolute path.
4879 PH/39 A tweak to the "H" retry rule from its user.
4881 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4882 a label. They prevented compilation on older perls.
4884 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4885 a warning to be raised on newish perls.
4887 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4888 on queue. Changes to match documented behaviour of showing count of
4889 messages matching specified criteria.
4891 PH/40 Changed the default ident timeout from 30s to 5s.
4893 PH/41 Added support for the use of login_cap features, on those BSD systems
4894 that have them, for controlling the resources used by pipe deliveries.
4896 PH/42 The content-scanning code uses fopen() to create files in which to put
4897 message data. Previously it was not paying any attention to the mode of
4898 the files. Exim runs with umask(0) because the rest of the code creates
4899 files with open(), and sets the required mode explicitly. Thus, these
4900 files were ending up world-writeable. This was not a big issue, because,
4901 being within the spool directory, they were not world-accessible. I have
4902 created a function called modefopen, which takes an additional mode
4903 argument. It sets umask(777), creates the file, chmods it to the required
4904 mode, then resets the umask. All the relevant calls to fopen() in the
4905 content scanning code have been changed to use this function.
4907 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4908 to 24 hours. This avoids potential overflow problems when processing G
4909 and H retry rules. I suspect nobody ever tinkers with this value.
4911 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4913 PH/45 When the plaintext authenticator is running as a client, the server's
4914 challenges are checked to ensure they are valid base64 strings. By
4915 default, the authentication attempt is cancelled if an invalid string is
4916 received. Setting client_ignore_invalid_base64 true ignores these errors.
4917 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4918 they are received. Thus, the responses can be made to depend on the
4919 challenges. If an invalid string is ignored, an empty string is placed in
4922 PH/46 Messages that are created by the autoreply transport now contains a
4923 References: header, in accordance with RFCs 2822 and 3834.
4925 PH/47 Added authenticated_sender_force to the smtp transport.
4927 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4929 PH/49 Installed latest patch from the Sieve maintainer.
4931 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4932 was also set, the appendfile transport was unnecessarily scanning a
4933 directory of message files (e.g. for maildir delivery) to find the count
4934 of files (along with the size), even though it did not need this
4935 information. It now does the scan only if it needs to find either the
4936 size of the count of files.
4938 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4940 PH/52 Two bugs concerned with error handling when the smtp transport is
4943 (i) Exim was not creating retry information for temporary errors given
4944 for individual recipients after the DATA command when the smtp transport
4945 was used in LMTP mode. This meant that they could be retried too
4946 frequently, and not timed out correctly.
4948 (ii) Exim was setting the flag that allows error details to be returned
4949 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4950 recipients that were returned after the DATA command.
4952 PH/53 This is related to PH/52, but is more general: for any failing address,
4953 when detailed error information was permitted to be returned to the
4954 sender, but the error was temporary, then after the final timeout, only
4955 "retry timeout exceeded" was returned. Now it returns the full error as
4956 well as "retry timeout exceeded".
4958 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4959 do this, and (what is worse) MTAs that accept it.
4961 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4962 will now be deprecated.
4964 PH/56 New os.c-cygwin from the Cygwin maintainer.
4966 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4967 formats (previously only available in exim formats via -bpr, -bpru,
4968 and -bpra. Now also available in native and exiqgrep formats)
4970 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4971 with very large, slow to parse queues
4973 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4975 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4977 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4978 responses to authentication challenges, though it was showing the
4979 challenges; (ii) I've removed the CR characters from the debug output for
4982 PH/58 Allow for the insertion of a newline as well as a space when a string
4983 is turned into more than one encoded-word during RFC 2047 encoding. The
4984 Sieve code now uses this.
4986 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4987 data_4xx, lost_connection, tls_required.
4989 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4990 message was being sent as an SMTP response.
4992 PH/61 Add -l and -k options to exicyclog.
4994 PH/62 When verifying, if an address was redirected to one new address, so that
4995 verification continued, and the new address failed or deferred after
4996 having set something in $address_data, the value of $address_data was not
4997 passed back to the ACL. This was different to the case when no
4998 redirection occurred. The value is now passed back in both cases.
5000 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5001 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5002 use login_cap.h, so on its own it isn't the distinguishing feature. The
5003 new name refers directly to the setclassresources() function.
5005 PH/65 Added configuration files for NetBSD3.
5007 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5009 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5010 is preferred over IPv4.
5012 PH/68 The bounce_return_message and bounce_return_body options were not being
5013 honoured for bounces generated during the reception of non-SMTP messages.
5014 In particular, this applied to messages rejected by the ACL. This bug has
5015 been fixed. However, if bounce_return_message is true and bounce_return_
5016 body is false, the headers that are returned for a non-SMTP message
5017 include only those that have been read before the error was detected.
5018 (In the case of an ACL rejection, they have all been read.)
5020 PH/69 The HTML version of the specification is now built in a directory called
5021 spec_html instead of spec.html, because the latter looks like a path with
5022 a MIME-type, and this confuses some software.
5024 PH/70 Catch two compiler warnings in sieve.c.
5026 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5027 function verify_get_ident() calls ip_connect() to connect a socket, but
5028 if the "connect()" function timed out, ip_connect() used to close the
5029 socket. However, verify_get_ident() also closes the socket later, and in
5030 between Exim writes to the log, which may get opened at this point. When
5031 the socket was closed in ip_connect(), the log could get the same file
5032 descriptor number as the socket. This naturally causes chaos. The fix is
5033 not to close the socket in ip_connect(); the socket should be closed by
5034 the function that creates it. There was only one place in the code where
5035 this was missing, in the iplookup router, which I don't think anybody now
5036 uses, but I've fixed it anyway.
5038 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5039 well as to direct DNS lookups. Otherwise the handling of names in host
5040 lists is inconsistent and therefore confusing.
5046 PH/01 Two changes to the default runtime configuration:
5048 (1) Move the checks for relay_from_hosts and authenticated clients from
5049 after to before the (commented out) DNS black list checks.
5051 (2) Add control=submission to the relay_from_hosts and authenticated
5052 clients checks, on the grounds that messages accepted by these
5053 statements are most likely to be submissions.
5055 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5057 (1) Generate an error if the third argument for the ${prvs expansion is
5060 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5063 (3) Reset the variables that are obtained from the first argument of
5064 ${prvscheck and used in the second argument before leaving the code,
5065 because their memory is reclaimed, so using them afterwards may do
5068 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5069 one (it's much easier than Tom thought :-).
5071 (5) Because of (4), we can now allow for the use of $prvscheck_result
5072 inside the third argument.
5074 PH/03 For some reason, the default setting of PATH when running a command from
5075 a pipe transport was just "/usr/bin". I have changed it to
5078 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5079 anything to be listed in the output from -bV.
5081 PH/05 When a filter generated an autoreply, the entire To: header line was
5082 quoted in the delivery log line, like this:
5084 => >A.N.Other <ano@some.domain> <original@ddress> ...
5086 This has been changed so that it extracts the operative address. There
5087 may be more than one such address. If so, they are comma-separated, like
5090 => >ano@some.domain,ona@other.domain <original@ddress> ...
5092 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5093 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5094 not being looked up in the rDNS to get a host name, Exim was showing the
5095 IP address twice in Received: lines, even though the IP addresses were
5096 identical. For example:
5098 Received: from [1.2.3.4] (helo=[1.2.3.4])
5100 However, if the real host name was known, it was omitting the HELO data
5101 if it matched the actual IP address. This has been tidied up so that it
5102 doesn't show the same IP address twice.
5104 PH/07 When both +timestamp and +memory debugging was on, the value given by
5105 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5106 called by the debug printing, thereby overwriting the timestamp buffer.
5107 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5109 PH/08 When the original message was included in an autoreply transport, it
5110 always said "this is a copy of the message, including all the headers",
5111 even if body_only or headers_only was set. It now gives an appropriate
5114 PH/09 Applied a patch from the Sieve maintainer which:
5116 o fixes some comments
5117 o adds the (disabled) notify extension core
5118 o adds some debug output for the result of if/elsif tests
5119 o points to the current vacation draft in the documentation
5120 and documents the missing references header update
5124 o fixes a bug in processing the envelope test (when testing
5125 multiple envelope elements, the last element determined the
5128 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5129 Electronic Mail") by including:
5131 Auto-submitted: auto-generated
5133 in the messages that it generates (bounce messages and others, such as
5134 warnings). In the case of bounce messages for non-SMTP messages, there was
5135 also a typo: it was using "Auto_submitted" (underscore instead of
5136 hyphen). Since every message generated by Exim is necessarily in response
5137 to another message, thes have all been changed to:
5139 Auto-Submitted: auto-replied
5141 in accordance with these statements in the RFC:
5143 The auto-replied keyword:
5145 - SHOULD be used on messages sent in direct response to another
5146 message by an automatic process,
5148 - MUST NOT be used on manually-generated messages,
5150 - MAY be used on Delivery Status Notifications (DSNs) and Message
5151 Disposition Notifications (MDNs),
5153 - MUST NOT be used on messages generated by automatic or periodic
5154 processes, except for messages which are automatic responses to
5157 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5158 to the default Received: header definition.
5160 PH/12 Added log selector acl_warn_skipped (default on).
5162 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5163 variables because (a) they are in the wrong storage pool and (b) even if
5164 they were copied, it wouldn't work properly because of the caching.
5166 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5167 checking when decoding. Apparently there are clients that generate
5168 overlong encoded strings. Why am I not surprised?
5170 PH/15 If the first argument of "${if match_address" was not empty, but did not
5171 contain an "@" character, Exim crashed. Now it writes a panic log message
5172 and treats the condition as false.
5174 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5176 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5177 extension "envelope-auth". The code is finished and in agreement with
5178 other implementations, but there is no documentation so far and in fact,
5179 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5180 not changing the active code.
5182 Print executed "if" and "elsif" statements when debugging is used. This
5183 helps a great deal to understand what a filter does.
5185 Document more things not specified clearly in RFC3028. I had all this
5186 sorted out, when out of a sudden new issues came to my mind. Oops."
5188 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5191 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5192 canonical form (as documented). However, after a host name lookup from
5193 the IP address, check_host() was doing a simple string comparison with
5194 addresses acquired from the DNS when checking that the found name did
5195 have the original IP as one of its addresses. Since any found IPv6
5196 addresses are likely to be in abbreviated form, the comparison could
5197 fail. Luckily, there already exists a function for doing the comparison
5198 by converting both addresses to binary, so now that is used instead of
5199 the text comparison.
5201 PH/20 There was another similar case to PH/19, when a complete host name was
5202 given in a host list; looking up its IP address could give an abbreviated
5203 form, whereas the current host's name might or might not be abbreviated.
5204 The same fix has been applied.
5210 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5211 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5214 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5215 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5217 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5219 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5220 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5221 Also, for "Auto-Submitted", treat anything other than "no" as
5222 non-personal, in accordance with RFC 3834. (Previously it treated
5223 anything starting "auto-" as non-personal.)
5225 TF/01 The control=submission/name=... option had a problem with syntax
5226 errors if the name included a slash character. The /name= option
5227 now slurps the rest of the string, so it can include any characters
5228 but it must come last in the list of options (after /sender_retain
5231 PH/05 Some modifications to the interface to the fake nameserver for the new
5239 TK/01 Added the "success_on_redirect" address verification option. See
5240 NewStuff for rationale and an example.
5242 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5244 PH/02 Patch to exigrep to allow it to work on syslog lines.
5246 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5247 fread() to skip over the body file's header line, because in Cygwin the
5248 header line is locked and is inaccessible.
5250 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5251 co-exist for some time) to make it clear that it is the Exim ID that is
5252 referenced, not the Message-ID: header line.
5254 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5255 string_format() function, because snprintf() does not exist on all
5258 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5259 ${stat: expansion item.
5261 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5262 protocol synchronization error", to keep the pedants happy.
5264 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5265 well as for IRIX systems, when gcc is being used. See the host.c source
5268 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5270 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5273 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5274 transport and to the smtp transport in LMTP mode.
5276 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5278 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5279 run a filter in a subprocess. This could lead to confusion in subsequent
5280 lookups in the parent process. There should also be a search_tidyup() at
5281 the end of the subprocess.
5283 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5284 only if the host matched helo_try_verify_hosts, which caused the
5285 verification to occur when the EHLO/HELO command was issued. The ACL just
5286 tested the remembered result. Now, if a previous verification attempt has
5287 not happened, "verify = helo" does it there and then.
5289 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5291 TK/03 Fix log output including CR from clamd.
5293 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5294 exist provoked a memory error which could cause a segfault.
5296 PH/15 Installed PCRE 6.2
5298 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5300 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5301 of the problem. Specifically, suggested +O2 rather than +O1 for the
5304 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5306 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5307 record, the port was not correctly logged when the outgoing_port log
5308 selector was set (it logged the transort's default port).
5310 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5311 fallback_hosts, and "hosts" in the smtp transport.
5313 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5314 host errors such as "Connection refused".
5316 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5317 authentication with radiusclient 0.4.9:
5319 - Error returned from rc_read_config was caught wrongly
5320 - Username/password not passed on to radius server due to wrong length.
5322 The presumption is that some radiusclient API changes for 4.51/PH/17
5323 were not taken care of correctly. The code is still untested by me (my
5324 Linux distribution still has 0.3.2 of radiusclient), but it was
5325 contributed by a Radius user.
5327 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5328 expanding the "port" option of the smtp transport.
5330 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5331 while reading a MIME header. Thanks to Tom Hughes for a patch.
5333 PH/24 Include config.h inside local_scan.h so that configuration settings are
5336 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5337 that in an ACL for RCPT (for example), you can examine exactly what was
5340 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5341 commands, but it was not correctly comparing the address with the actual
5342 client host address. Thus, it would show the EHLO address in Received:
5343 header lines when this was not necessary.
5345 PH/27 Added the % operator to ${eval:}.
5347 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5348 it should be ignoring failures (because with -C, for example, it has lost
5349 privilege). It wasn't ignoring creation failures other than "already
5352 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5355 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5356 because an input line was too long, either on its own, or by virtue of
5357 too many continuations, the temporary file was not being removed, and the
5358 return code was incorrect.
5360 PH/31 Missing "BOOL" in function definition in filtertest.c.
5362 PH/32 Applied Sieve patches from the maintainer.
5364 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5366 PH/33 Added "verify = not_blind".
5368 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5369 Local/Makefile (with some defaults set). These are used in built scripts
5370 such as exicyclog, but they have never been used in the exim_install
5371 script (though there are many overriding facilities there). I have
5372 arranged that the exim_install script now takes note of these two
5375 PH/35 Installed configuration files for Dragonfly.
5377 PH/36 When a locally submitted message by a trusted user did not contain a
5378 From: header, and the sender address was obtained from -f or from an SMTP
5379 MAIL command, and the trusted user did not use -F to supply a sender
5380 name, $originator_name was incorrectly used when constructing a From:
5381 header. Furthermore, $originator_name was used for submission mode
5382 messages from external hosts without From: headers in a similar way,
5383 which is clearly wrong.
5385 PH/37 Added control=suppress_local_fixups.
5387 PH/38 When log_selector = +received_sender was set, and the addition of the
5388 sender made the log line's construction buffer exactly full, or one byte
5389 less than full, an overflow happened when the terminating "\n" was
5392 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5393 when the result of a list match is failure because a DNS lookup failed.
5395 PH/40 RM_COMMAND is now used in the building process.
5397 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5398 the "build-* directories that it finds.
5400 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5401 address in a domain literal was a prefix of an interface address.
5403 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5404 when verifying a sender address, unless rewrite_headers is false.
5406 PH/44 Wrote a long comment about why errors_to addresses are verified as
5407 recipients, not senders.
5409 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5410 the ratelimit ACL was added.
5412 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5414 PH/46 Added extra information about PostgreSQL errors to the error string.
5416 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5417 avoiding the need to install special zones in a real server. This is
5418 backwards compatible; if it can't find the fake resolver, it drops back.
5419 Thus, both old and new test suites can be run.
5421 TF/02 Added util/ratelimit.pl
5423 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5424 clock is set back in time.
5426 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5427 Candler <B.Candler@pobox.com>.
5429 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5430 OK for addresses that are the result of redirections.
5432 PH/48 A number of further additions for the benefit of the new test suite,
5433 including a fake gethostbyname() that interfaces to the fake DNS resolver
5436 TF/06 The fix for widen_domains has also been applied to qualify_single and
5437 search_parents which are the other dnslookup options that can cause
5440 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5443 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5445 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5446 service" if it encountered an MX record with an empty target hostname.
5447 The message is now "an MX or SRV record indicated no SMTP service".
5449 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5450 if the DNS of the sending site is misconfigured. This is quite a
5451 common situation. This change restores the behaviour of treating a
5452 helo verification defer as a failure.
5454 PH/51 If self=fail was set on a router, the bounce message did not include the
5455 actual error message.
5461 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5463 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5464 command itself ended in error, the underlying message about the transport
5465 filter timeout was being overwritten with the pipe command error. Now the
5466 underlying error message should be appended to the second error message.
5468 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5470 PH/02 Reduce the amount of output that "make" produces by default. Full output
5471 can still be requested.
5473 PH/03 The warning log line about a condition test deferring for a "warn" verb
5474 was being output only once per connection, rather than after each
5475 occurrence (because it was using the same function as for successful
5476 "warn" verbs). This seems wrong, so I have changed it.
5478 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5479 it should not have, which might have caused a crash in the right
5480 circumstances, but probably never did.
5482 PH/04 Installed a modified version of Tony Finch's patch to make submission
5483 mode fix the return path as well as the Sender: header line, and to
5484 add a /name= option so that you can make the user's friendly name appear
5487 TF/03 Added the control = fakedefer ACL modifier.
5489 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5490 Mark Lowes for thorough testing.
5492 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5494 TK/03 Merged latest SRS patch from Miles Wilton.
5496 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5497 with the definition in sysexits.h (which is #included earlier).
5498 Fortunately, Exim does not actually use EX_OK. The code used to try to
5499 preserve the sysexits.h value, by assuming that macro definitions were
5500 scanned for macro replacements. I have been disabused of this notion,
5501 so now the code just undefines EX_OK before #including unistd.h.
5503 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5504 in the smtp transport. When a block could not be written in a single
5505 write() function, the timeout was being re-applied to each part-write.
5506 This seems wrong - if the receiver was accepting one byte at a time it
5507 would take for ever. The timeout is now adjusted when this happens. It
5508 doesn't have to be particularly precise.
5510 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5511 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5513 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5514 without a domain if the check to <postmaster@domain> fails.
5516 SC/01 Eximstats: added -xls and the ability to specify output files
5517 (patch written by Frank Heydlauf).
5519 SC/02 Eximstats: use FileHandles for outputting results.
5521 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5523 SC/04 Eximstats: fixed display of large numbers with -nvr option
5525 SC/05 Eximstats: fixed merging of reports with empty tables.
5527 SC/06 Eximstats: added the -include_original_destination flag
5529 SC/07 Eximstats: removed tabs and trailing whitespace.
5531 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5533 TK/06 MBOX spool code: Add real "From " MBOX separator line
5534 so the .eml file is really in mbox format (even though
5535 most programs do not really care). Patch from Alex Miller.
5537 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5538 The latter is generated from $received_to and is only set if the
5539 message has one envelope recipient. SA can use these headers,
5540 obviously out-of-the-box. Patch from Alex Miller.
5542 PH/08 The ${def test on a variable was returning false if the variable's
5543 value was "0", contrary to what the specification has always said!
5544 The result should be true unless the variable is empty.
5546 PH/09 The syntax error of a character other than { following "${if
5547 def:variable_name" (after optional whitespace) was not being diagnosed.
5548 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5549 accidental colon was present, for example, could give incorrect results.
5551 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5552 result is used (not including appendfile, where other changes are about
5555 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5556 This involved changing a lot of size variables from int to off_t. It
5557 should work with maildirs and everything.
5559 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5560 spamd dying while we are connected to it.
5562 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5565 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5566 function for BDB 4.3.
5568 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5570 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5571 the sources for PCRE have been re-arranged and more files are now
5574 PH/15 The code I had for printing potentially long long variables in PH/11
5575 above was not the best (it lost precision). The length of off_t variables
5576 is now inspected at build time, and an appropriate printing format (%ld
5577 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5578 to be "long long int" or "long int". This is needed for the internal
5579 formatting function string_vformat().
5581 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5582 the configuration file to be ":syslog", then the script "guesses" where
5583 the logs files are, rather than using the compiled in default. In our
5584 case the guess is not the same as the compiled default, so the script
5585 suddenly stopped working when I started to use syslog. The patch checks
5586 to see if log_file_path is "". If so, it attempts to read it from exim
5587 with no configuration file to get the compiled in version, before it
5588 falls back to the previous guessing code."
5590 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5591 implementing BATV in an Exim configuration. See NewStuff for the gory
5594 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5595 Makefile that are specific to HP-UX.
5597 PH/18 If the "use_postmaster" option was set for a recipient callout together
5598 with the "random" option, the postmaster address was used as the MAIL
5599 FROM address for the random test, but not for the subsequent recipient
5600 test. It is now used for both.
5602 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5603 patch removes a few documentation additions to RFC 3028, because the
5604 latest draft now contains them. It adds the new en;ascii-case comparator
5605 and a new error check for 8bit text in MIME parts. Comparator and
5606 require names are now matched exactly. I enabled the subaddress
5607 extension, but it is not well tested yet (read: it works for me)."
5609 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5610 rework some of the code of TK/09 above to avoid the hardwired use of
5611 "%lld" and "long long". Replaced the call to snprintf() with a call to
5614 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5615 records point to non-existent hosts", "retry timeout exceeded", and
5616 "retry time not reached for any host after a long failure period".
5618 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5619 experimental DomainKeys support:
5621 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5622 (2) On an error such as an illegally used "control", the wrong name for
5623 the control was given.
5625 These problems did NOT occur unless DomainKeys support was compiled.
5627 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5629 PH/24 Added ${if match_ip condition.
5631 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5632 databases so that it will be absolutely obvious if a crash occurs in the
5633 DB library. This is a regular occurrence (often caused by mis-matched
5636 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5637 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5638 picked up on a user's system that detects such things. There doesn't seem
5639 to be a gcc warning option for this - only an attribute that has to be
5640 put on the function's prototype. It seems that in Fedora Core 4 they have
5641 set this on a number of new functions. No doubt there will be more in due
5644 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5645 specify a transport. However, if an address that was verified by such a
5646 router was the subject of a callout, Exim crashed because it tried to
5647 read the rcpt_include_affixes from the non-existent transport. Now it
5648 just assumes that the setting of that option is false. This bug was
5649 introduced by 4.51/PH/31.
5651 PH/28 Changed -d+all to exclude +memory, because that information is very
5652 rarely of interest, but it makes the output a lot bigger. People tend to
5653 do -d+all out of habit.
5655 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5656 code in os-type was giving problems when libc.so lives in lib64, like on
5659 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5660 aren't the modern standard, and it seems that some systems' include files
5661 don't always have them. Exim was already checking for some of the newer
5662 ones like T_AAAA, and defining it itself. I've added checks for all the
5663 record types that Exim uses.
5665 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5666 not automatically generating a new one, as it is supposed to. This
5667 prevented TLS from working. If the file did exist, but contained invalid
5668 data, a new version was generated, as expected. It was only the case of a
5669 non-existent file that was broken.
5671 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5672 with a change in libdomainkeys > 0.64.
5674 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5675 from DNS. If the selector record carries the flag, it now has
5676 precedence over the domain-wide flag.
5678 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5680 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5681 the use of a port name that isn't defined in /etc/services) occurred, the
5682 message was deferred as in a normal delivery, and thus remained on the
5683 spool, instead of being failed because of the mua_wrapper setting. This
5684 is now fixed, and I tidied up some of the mua_wrapper messages at the
5687 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5688 the messages in a hash of arrays rather than using individual hashes.
5689 This is a bit cleaner and results in dramatic memory savings, albeit
5690 at a slight CPU cost.
5692 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5693 as requested by Marc Sherman.
5695 SC/10 Eximstats: added histograms for user specified patterns as requested
5698 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5700 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5701 fopen() in the content-scanning modules that did not already have it.
5707 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5708 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5710 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5712 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5714 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5715 to test Sieve filters that use "vacation".
5717 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5718 that changes the way the GnuTLS parameters are stored in the cache file.
5719 The new format can be generated externally. For backward compatibility,
5720 if the data in the cache doesn't make sense, Exim assumes it has read an
5721 old-format file, and it generates new data and writes a new file. This
5722 means that you can't go back to an older release without removing the
5725 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5726 work if there are any delivery delays because "one_time" forces the
5727 parent to be marked "delivered", so its unseen clone is never tried
5728 again. For this reason, Exim now forbids the simultaneous setting of
5731 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5732 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5733 entry if you want to know the details. The fix, however, neglected to
5734 consider the case when local delivery batching is involved. The test for
5735 "previously delivered" was not happening when checking to see if an
5736 address could be batched with a previous (undelivered) one; under
5737 certain circumstances this could lead to multiple deliveries to the same
5740 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5741 in its include files, and this causes problems building Exim.
5743 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5744 header_syntax) but Exim was just ignoring anything given after a slash.
5745 In particular, this caused confusion with an attempt to use "verify =
5746 reverse_host_lookup/defer_ok". An error is now given when options are
5747 supplied for verify items that do not have them. (Maybe reverse_host_
5748 lookup should have a defer_ok option, but that's a different point.)
5750 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5751 defined by RFC 821) to 2048, because there were problems with some AUTH
5752 commands, and RFC 1869 says the size should be increased for extended
5753 SMTP commands that take arguments.
5755 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5758 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5759 "unknown" error; now it says that the functionality isn't in the binary.
5761 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5762 an address' error message when a string expansion fails (syntax or
5763 whatever). Otherwise the password may appear in the log. Following change
5764 PH/42 below, there is no longer a chance of it appearing in a bounce
5767 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5769 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5770 of its addresses was ever tried. (Bugzilla bug #2.)
5772 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5773 the result incorrectly in the debug output. (It correctly added a newline
5774 to what was transported.)
5776 TF/01 Added $received_time.
5778 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5779 commented out examples of how to interface to a virus scanner and to
5780 SpamAssassin. Also added commented examples of av_scanner and
5781 spamd_address settings.
5783 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5784 and controls are allowed in which ACLs. There were a couple of minor
5785 errors. Some of the entries in the conditions table (which is a table of
5786 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5787 negation of where the condition IS allowed.
5789 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5791 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5792 header file does not have a version number, so I've had to invent a new
5793 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5794 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5795 radiusclient), but it was contributed by a Radius user.
5797 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5798 files or directories, for OpenSSL.
5800 PH/19 When an Exim process that is running as root has to create an Exim log
5801 file, it does so in a subprocess that runs as exim:exim so as to get the
5802 ownership right at creation (otherwise, other Exim processes might see
5803 the file with the wrong ownership). There was no test for failure of this
5804 fork() call, which would lead to the process getting stuck as it waited
5805 for a non-existent subprocess. Forks do occasionally fail when resources
5806 run out. I reviewed all the other calls to fork(); they all seem to check
5809 PH/20 When checking for unexpected SMTP input at connect time (before writing
5810 the banner), Exim was not dealing correctly with a non-positive return
5811 from the read() function. If the client had disconnected by this time,
5812 the result was a log entry for a synchronization error with an empty
5813 string after "input=" when read() returned zero. If read() returned -1
5814 (an event I could not check), uninitialized data bytes were printed.
5815 There were reports of junk text (parts of files, etc) appearing after
5818 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5820 PH/22 Added support for macro redefinition, and (re)definition in between
5821 driver and ACL definitions.
5823 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5824 forgetting to use the resulting value; it was using the unexpanded value.
5826 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5827 hadn't been configured. The fix is from Juergen Kreileder, who
5828 understands it better than I do:
5830 "Here's what I see happening with three configured cyrus_sasl
5831 authenticators configured (plain, login, cram-md5):
5833 On startup auth_cyrus_sasl_init() gets called for each of these.
5834 This means three calls to sasl_listmech() without a specified mech_list.
5835 => SASL tests which mechs of all available mechs actually work
5836 => three warnings about OTP not working
5837 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5839 With the patch, sasl_listmech() also gets called three times. But now
5840 SASL's mech_list option is set to the server_mech specified in the the
5841 authenticator. Or in other words, the answer from sasl_listmech()
5842 gets limited to just the mech you're testing for (which is different
5844 => the return list contains just 'plain' or 'login', 'cram-md5' or
5845 nothing depending on the value of ob->server_mech.
5847 I've just tested the patch: Authentication still works fine,
5848 unavailable mechs specified in the exim configuration are still
5849 caught, and the auth.log warnings about OTP are gone."
5851 PH/25 When debugging is enabled, the contents of the command line are added
5852 to the debugging output, even when log_selector=+arguments is not
5855 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5856 answer is "GNU", and only if the return is "GNU/something" is the answer
5859 PH/27 $acl_verify_message is now set immediately after the failure of a
5860 verification in an ACL, and so is available in subsequent modifiers. In
5861 particular, the message can be preserved by coding like this:
5863 warn !verify = sender
5864 set acl_m0 = $acl_verify_message
5866 Previously, $acl_verify_message was set only while expanding "message"
5867 and "log_message" when a very denied access.
5869 PH/28 Modified OS/os.c-Linux with
5871 -#ifndef OS_LOAD_AVERAGE
5872 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5874 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5875 nomenclature these days.)
5877 PH/29 Installed patch from the Sieve maintainer that adds the options
5878 sieve_useraddress and sieve_subaddress to the redirect router.
5880 PH/30 In these circumstances:
5881 . Two addresses routed to the same list of hosts;
5882 . First host does not offer TLS;
5883 . First host accepts first address;
5884 . First host gives temporary error to second address;
5885 . Second host offers TLS and a TLS session is established;
5886 . Second host accepts second address.
5887 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5888 and peerdn, if requested) that were in fact used only for the second
5891 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5892 attention to any local part prefix or suffix that was matched by the
5893 router that accepted the address. It now behaves in the same way as it
5894 does for delivery: the affixes are removed from the local part unless
5895 rcpt_include_affixes is set on the transport.
5897 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5898 timeout during the DATA phase of an incoming message.
5900 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5901 applied a patch sanctioned by the Sieve maintainer.
5903 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5904 the uid or gid is negative. A case of a negative gid caused this to be
5905 noticed. The fix allows for either to be negative.
5907 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5908 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5911 PH/36 The change PH/12 above was broken. Fixed it.
5913 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5914 the grounds that routing the same address twice would always produce the
5915 same answer. This might have been true once, but it is certainly no
5916 longer true now. Routing a child address may depend on the previous
5917 routing that produced that child. Some complicated redirection strategies
5918 went wrong when messages had multiple recipients, and made Exim's
5919 behaviour dependent on the order in which the addresses were given.
5921 I have moved the duplicate checking until after the routing is complete.
5922 Exim scans the addresses that are assigned to local and remote
5923 transports, and removes any duplicates. This means that more work will be
5924 done, as duplicates will always all be routed, but duplicates are
5925 presumably rare, so I don't expect this is of any significance.
5927 For deliveries to pipes, files, and autoreplies, the duplicate checking
5928 still happens during the routing process, since they are not going to be
5931 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5932 It corrects a timeout issue with spamd. This is Ian's comment: "The
5933 background is that sometimes spamd either never reads data from a
5934 connection it has accepted, or it never writes response data. The exiscan
5935 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5936 blindly assumes that writes won't block so it may never time out."
5938 PH/39 Allow G after quota size as well as K and M.
5940 PH/40 The value set for $authenticated_id in an authenticator may not contain
5941 binary zeroes or newlines because the value is written to log lines and
5942 to spool files. There was no check on this. Now the value is run through
5943 the string_printing() function so that such characters are converted to
5944 printable escape sequences.
5946 PH/41 $message_linecount is a new variable that contains the total number of
5947 lines in the message. Compare $body_linecount, which is the count for the
5950 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5951 bounce and delay warning messages, except in certain special cases, which
5954 (a) An SMTP error message from a remote host;
5955 (b) A message specified in a :fail: redirection;
5956 (c) A message specified in a "fail" command in a system filter;
5957 (d) A message specified in a FAIL return from the queryprogram router;
5958 (e) A message specified by the cannot_route_message router option.
5960 In these cases only, Exim does include the error details in bounce and
5961 warning messages. There are also a few cases where bland messages such
5962 as "unrouteable address" or "local delivery error" are given.
5964 PH/43 $value is now also set for the "else" part of a ${run expansion.
5966 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5967 being worked on, but at least Exim now implements the latest version to
5970 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5971 process to complete was treated as a delivery failure, a timeout while
5972 writing the message to the pipe was logged, but erroneously treated as a
5973 successful delivery. Such timeouts include transport filter timeouts. For
5974 consistency with the overall process timeout, these timeouts are now
5975 treated as errors, giving rise to delivery failures by default. However,
5976 there is now a new Boolean option for the pipe transport called
5977 timeout_defer, which, if set TRUE, converts the failures into defers for
5978 both kinds of timeout. A transport filter timeout is now identified in
5981 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5982 systems where "make" and "gmake" are different, calling "gmake" at top
5983 level broke things. I've arranged for the value of $(MAKE) to be passed
5984 from the Makefile to this script so that it can call the same version of
5988 A note about Exim versions 4.44 and 4.50
5989 ----------------------------------------
5991 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5992 changes of various kinds. As a consequence, a big documentation update was
5993 needed. This delayed the release for rather longer than seemed good, especially
5994 in the light of a couple of (minor) security issues. Therefore, the changes
5995 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5996 release. So 4.44 and 4.50 are in effect two different branches that both start
5999 I have left the 4.50 change log unchanged; it contains all the changes since
6000 4.43. The change log for 4.44 is below; many of its items are identical to
6001 those for 4.50. This seems to be the most sensible way to preserve the
6002 historical information.
6008 1. Minor wording change to the doc/README.SIEVE file.
6010 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6011 computation of the current number of files was incorrect.
6013 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6014 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6017 4. Give more explanation in the error message when the command for a transport
6018 filter fails to execute.
6020 5. There are several places where Exim runs a non-Exim command in a
6021 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6022 was being done only for the command run by the queryprogram router. It is
6023 now done for all such subprocesses. The other cases are: ${run, transport
6024 filters, and the commands run by the lmtp and pipe transports.
6026 6. Added CONFIGURE_GROUP build-time option.
6028 7. Some older OS have a limit of 256 on the maximum number of file
6029 descriptors. Exim was using setrlimit() to set 1000 as a large value
6030 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6031 systems. I've change it so that if it can't get 1000, it tries for 256.
6033 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6034 was an oversight, and furthermore, ever since the addition of extra
6035 controls (e.g. 4.43/32), the checks on when to allow different forms of
6036 "control" were broken. There should now be diagnostics for all cases when a
6037 control that does not make sense is encountered.
6039 9. Added the /retain_sender option to "control=submission".
6041 10. $recipients is now available in the predata ACL (oversight).
6043 11. Tidy the search cache before the fork to do a delivery from a message
6044 received from the command line. Otherwise the child will trigger a lookup
6045 failure and thereby defer the delivery if it tries to use (for example) a
6046 cached ldap connection that the parent has called unbind on.
6048 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6049 of $address_data from the recipient verification was clobbered by the
6050 sender verification.
6052 13. The value of address_data from a sender verification is now available in
6053 $sender_address_data in subsequent conditions in the ACL statement.
6055 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6057 15. Added a new option "connect=<time>" to callout options, to set a different
6060 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6061 was its contents. (It was OK if the option was not defined at all.)
6063 17. A "Completed" log line is now written for messages that are removed from
6064 the spool by the -Mrm option.
6066 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6067 information about exactly what failed.
6069 19. Added -dd to debug only the daemon process.
6071 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6072 handles timeouts, both on the server side and network timeouts. Renamed the
6073 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6075 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6076 It is now set to "smtps".
6078 22. $host_address is now set to the target address during the checking of
6079 ignore_target_hosts.
6081 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6082 being passed; this would have caused $sender_host_name to have been used if
6083 matching the list had actually called for a host name (not very likely,
6084 since this list is usually IP addresses). A host name is now passed as
6087 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6088 code that specifies a non-restarting handler (typically sigaction() in
6089 modern systems) in an attempt to fix a rare and obscure crash bug.
6091 25. Narrowed the window for a race in the daemon that could cause it to ignore
6092 SIGCHLD signals. This is not a major problem, because they are used only to
6093 wake it up if nothing else does.
6095 26. A malformed maildirsize file could cause Exim to calculate negative values
6096 for the mailbox size or file count. Odd effects could occur as a result.
6097 The maildirsize information is now recalculated if the size or filecount
6100 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6101 support for a long time. Removed HAVE_SYS_VFS_H.
6103 28. Installed the latest version of exipick from John Jetmore.
6105 29. In an address list, if the pattern was not a regular expression, an empty
6106 subject address (from a bounce message) matched only if the pattern was an
6107 empty string. Non-empty patterns were not even tested. This was the wrong
6108 because it is perfectly reasonable to use an empty address as part of a
6109 database query. An empty address is now tested by patterns that are
6110 lookups. However, all the other forms of pattern expect the subject to
6111 contain a local part and a domain, and therefore, for them, an empty
6112 address still always fails if the pattern is not itself empty.
6114 30. Exim went into a mad DNS loop when attempting to do a callout where the
6115 host was specified on an smtp transport, and looking it up yielded more
6116 than one IP address.
6118 31. Re-factored the code for checking spool and log partition space into a
6119 function that finds that data and another that does the check. The former
6120 is then used to implement four new variables: $spool_space, $log_space,
6121 $spool_inodes, and $log_inodes.
6123 32. The RFC2047 encoding function was originally intended for short strings
6124 such as real names; it was not keeping to the 75-character limit for
6125 encoded words that the RFC imposes. It now respects the limit, and
6126 generates multiple encoded words if necessary. To be on the safe side, I
6127 have increased the buffer size for the ${rfc2047: expansion operator from
6130 33. It is now permitted to omit both strings after an "if" condition; if the
6131 condition is true, the result is "true". As before, when the second string
6132 is omitted, a false condition yields an empty string. This makes it less
6133 cumbersome to write custom ACL and router conditions.
6135 34. Failure to deliver a bounce message always caused it to be frozen, even if
6136 there was an errors_to setting on the router. The errors_to setting is now
6139 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6140 canonical form (fully expanded) before being placed in
6141 $sender_host_address.
6143 36. The table in the code that translates DNS record types into text (T_A to
6144 "A" for instance) was missing entries for NS and CNAME. It is just possible
6145 that this could have caused confusion if both these types were looked up
6146 for the same domain, because the text type is used as part of Exim's
6147 per-process caching. But the chance of anyone hitting this buglet seems
6150 37. The dnsdb lookup has been extended in a number of ways.
6152 (1) There is a new type, "zns", which walks up the domain tree until it
6153 finds some nameserver records. It should be used with care.
6155 (2) There is a new type, "mxh", which is like "mx" except that it returns
6156 just the host names, not the priorities.
6158 (3) It is now possible to give a list of domains (or IP addresses) to be
6159 looked up. The behaviour when one of the lookups defers can be
6160 controlled by a keyword.
6162 (4) It is now possible to specify the separator character for use when
6163 multiple records are returned.
6165 38. The dnslists ACL condition has been extended: it is now possible to supply
6166 a list of IP addresses and/or domains to be looked up in a particular DNS
6169 39. Added log_selector=+queue_time_overall.
6171 40. When running the queue in the test harness, wait just a tad after forking a
6172 delivery process, to get repeatability of debugging output.
6174 41. Include certificate and key file names in error message when GnuTLS fails
6175 to set them up, because the GnuTLS error message doesn't include the name
6176 of the failing file when there is a problem reading it.
6178 42. Allow both -bf and -bF in the same test run.
6180 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6182 44. The "Exiscan patch" is now merged into the mainline Exim source.
6184 45. Sometimes the final signoff response after QUIT could fail to get
6185 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6186 < 0 before doing a fflush(). This bug looks as though it goes back to the
6187 introduction of TLS in release 3.20, but "sometimes" must have been rare
6188 because the tests only now provoked it.
6190 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6191 (this can affect the format of dates).
6193 47. exim_tidydb, when checking for the continued existence of a message for
6194 which it has found a message-specific retry record, was not finding
6195 messages that were in split spool directories. Consequently, it was
6196 deleting retry records that should have stayed in existence.
6198 48. Steve fixed some bugs in eximstats.
6200 49. The SPA authentication driver was not abandoning authentication and moving
6201 on to the next authenticator when an expansion was forced to fail,
6202 contradicting the general specification for all authenticators. Instead it
6203 was generating a temporary error. It now behaves as specified.
6205 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6206 (the order specifies the preference for clients). The order is now AES256,
6207 AES128, 3DES, ARCFOUR128.
6209 51. Small patch to Sieve code - explicitly set From: when generating an
6212 52. Exim crashed if a remote delivery caused a very long error message to be
6213 recorded - for instance if somebody sent an entire SpamAssassin report back
6214 as a large number of 550 error lines. This bug was coincidentally fixed by
6215 increasing the size of one of Exim's internal buffers (big_buffer) that
6216 happened as part of the Exiscan merge. However, to be on the safe side, I
6217 have made the code more robust (and fixed the comments that describe what
6220 53. Now that there can be additional text after "Completed" in log lines (if
6221 the queue_time_overall log selector is set), a one-byte patch to exigrep
6222 was needed to allow it to recognize "Completed" as not the last thing in
6225 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6226 patch that reportedly fixes this has been added. I am not expert enough to
6227 create a test for it. This is what the patch creator wrote:
6229 "I found a little strange behaviour of ldap code when working with
6230 Windows 2003 AD Domain, where users was placed in more than one
6231 Organization Units. When I tried to give exim partial DN, the exit code
6232 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6233 But simultaneously result of request was absolutely normal ldap result,
6234 so I produce this patch..."
6236 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6237 REFERENCE, so I have modified the code to exclude the patch when that macro
6240 55. Some experimental protocols are using DNS PTR records for new purposes. The
6241 keys for these records are domain names, not reversed IP addresses. The
6242 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6243 leaves it alone. Component reversal etc. now happens only for IP addresses.
6246 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6248 57. Double the size of the debug message buffer (to 2048) so that more of very
6249 long debug lines gets shown.
6251 58. The exicyclog utility now does better if the number of log files to keep
6252 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6254 59. Two changes related to the smtp_active_hostname option:
6256 (1) $smtp_active_hostname is now available as a variable.
6257 (2) The default for smtp_banner uses $smtp_active_hostname instead
6258 of $primary_hostname.
6260 60. The host_aton() function is supposed to be passed a string that is known
6261 to be a valid IP address. However, in the case of IPv6 addresses, it was
6262 not checking this. This is a hostage to fortune. Exim now panics and dies
6263 if the condition is not met. A case was found where this could be provoked
6264 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6265 components; fortuitously, this particular loophole had already been fixed
6266 by change 4.50/55 above.
6268 If there are any other similar loopholes, the new check in host_aton()
6269 itself should stop them being exploited. The report I received stated that
6270 data on the command line could provoke the exploit when Exim was running as
6271 exim, but did not say which command line option was involved. All I could
6272 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6273 running as the user.
6276 61. There was a buffer overflow vulnerability in the SPA authentication code
6277 (which came originally from the Samba project). I have added a test to the
6278 spa_base64_to_bits() function which I hope fixes it.
6281 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6282 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6284 63. The daemon start-up calls getloadavg() while still root for those OS that
6285 need the first call to be done as root, but it missed one case: when
6286 deliver_queue_load_max is set with deliver_drop_privilege. This is
6287 necessary for the benefit of the queue runner, because there is no re-exec
6288 when deliver_drop_privilege is set.
6290 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6291 This has been fixed.
6293 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6294 was in use, was not putting the data itself into the right store pool;
6295 consequently, it could be overwritten for a subsequent message in the same
6296 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6299 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6301 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6302 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6303 the function were treating the return as a boolean value, which happened to
6304 work because 0=false and not-0=true, but is not correct code.
6306 68. The host_aton() function was not handling scoped IPv6 addresses (those
6307 with, for example, "%eth0" on the end) correctly.
6309 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6310 negated items (that is, ~something) in unsigned ints. Some compilers
6311 apparently mutter when there is no cast.
6313 70. If an address verification called from an ACL failed, and did not produce a
6314 user-specific message (i.e. there was only a "system" message), nothing was
6315 put in $acl_verify_message. In this situation, it now puts the system
6318 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6319 session; change 4.31/43 added the unwanted input to the log line - except
6320 that it did not do this in the start of session case. It now does.
6322 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6323 This is wrong and can cause the other end to generate a synchronization
6324 error if it is another Exim or anything else that does the synchronization
6325 check. A QUIT command is no longer sent after a timeout.
6327 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6328 during host lookups.
6330 74. The defer_ok option of callout verification was not working if it was used
6331 when verifying addresses in header lines, that is, for this case:
6333 verify = header_sender/callout=defer_ok
6335 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6336 those file descriptors could be used for SMTP connections. If anything
6337 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6338 could be sent to the SMTP client, causing chaos. The daemon now opens
6339 stdin, stdout, and stderr to /dev/null when it puts itself into the
6342 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6343 log by default. The user can override this with suitable Perl magic.
6345 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6346 the log message when discard triggers, was not working for the DATA ACL or
6347 for the non-SMTP ACL.
6349 78. Error message wording change in sieve.c.
6351 79. If smtp_accept_max_per_host was set, the number of connections could be
6352 restricted to fewer than expected, because the daemon was trying to set up
6353 a new connection before checking whether the processes handling previous
6354 connections had finished. The check for completed processes is now done
6355 earlier. On busy systems, this bug wouldn't be noticed because something
6356 else would have woken the daemon, and it would have reaped the completed
6359 80. If a message was submitted locally by a user whose login name contained one
6360 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6361 It caused a spool format error. I have fixed the spool reading code. A
6362 related problem was that the "from" clause in the Received: line became
6363 illegal because of the space(s). It is now covered by ${quote_local_part.
6365 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6368 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6370 83. Patch from Sieve maintainer to fix unterminated string problem in
6371 "vacation" handling.
6373 84. Some minor changes to the Linux configuration files to help with other
6374 OS variants using glibc.
6376 85. One more patch for Sieve to update vacation handling to latest spec.
6379 ----------------------------------------------------
6380 See the note above about the 4.44 and 4.50 releases.
6381 ----------------------------------------------------
6387 1. Change 4.43/35 introduced a bug that caused file counts to be
6388 incorrectly computed when quota_filecount was set in an appendfile
6391 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6392 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6395 3. Give more explanation in the error message when the command for a transport
6396 filter fails to execute.
6398 4. There are several places where Exim runs a non-Exim command in a
6399 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6400 was being done only for the command run by the queryprogram router. It is
6401 now done for all such subprocesses. The other cases are: ${run, transport
6402 filters, and the commands run by the lmtp and pipe transports.
6404 5. Some older OS have a limit of 256 on the maximum number of file
6405 descriptors. Exim was using setrlimit() to set 1000 as a large value
6406 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6407 systems. I've change it so that if it can't get 1000, it tries for 256.
6409 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6410 was an oversight, and furthermore, ever since the addition of extra
6411 controls (e.g. 4.43/32), the checks on when to allow different forms of
6412 "control" were broken. There should now be diagnostics for all cases when a
6413 control that does not make sense is encountered.
6415 7. $recipients is now available in the predata ACL (oversight).
6417 8. Tidy the search cache before the fork to do a delivery from a message
6418 received from the command line. Otherwise the child will trigger a lookup
6419 failure and thereby defer the delivery if it tries to use (for example) a
6420 cached ldap connection that the parent has called unbind on.
6422 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6423 of $address_data from the recipient verification was clobbered by the
6424 sender verification.
6426 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6427 was its contents. (It was OK if the option was not defined at all.)
6429 11. A "Completed" log line is now written for messages that are removed from
6430 the spool by the -Mrm option.
6432 12. $host_address is now set to the target address during the checking of
6433 ignore_target_hosts.
6435 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6436 being passed; this would have caused $sender_host_name to have been used if
6437 matching the list had actually called for a host name (not very likely,
6438 since this list is usually IP addresses). A host name is now passed as
6441 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6442 code that specifies a non-restarting handler (typically sigaction() in
6443 modern systems) in an attempt to fix a rare and obscure crash bug.
6445 15. Narrowed the window for a race in the daemon that could cause it to ignore
6446 SIGCHLD signals. This is not a major problem, because they are used only to
6447 wake it up if nothing else does.
6449 16. A malformed maildirsize file could cause Exim to calculate negative values
6450 for the mailbox size or file count. Odd effects could occur as a result.
6451 The maildirsize information is now recalculated if the size or filecount
6454 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6455 support for a long time. Removed HAVE_SYS_VFS_H.
6457 18. Updated exipick to current release from John Jetmore.
6459 19. Allow an empty sender to be matched against a lookup in an address list.
6460 Previously the only cases considered were a regular expression, or an
6463 20. Exim went into a mad DNS lookup loop when doing a callout where the
6464 host was specified on the transport, if the DNS lookup yielded more than
6467 21. The RFC2047 encoding function was originally intended for short strings
6468 such as real names; it was not keeping to the 75-character limit for
6469 encoded words that the RFC imposes. It now respects the limit, and
6470 generates multiple encoded words if necessary. To be on the safe side, I
6471 have increased the buffer size for the ${rfc2047: expansion operator from
6474 22. Failure to deliver a bounce message always caused it to be frozen, even if
6475 there was an errors_to setting on the router. The errors_to setting is now
6478 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6479 canonical form (fully expanded) before being placed in
6480 $sender_host_address.
6482 24. Updated eximstats to version 1.33
6484 25. Include certificate and key file names in error message when GnuTLS fails
6485 to set them up, because the GnuTLS error message doesn't include the name
6486 of the failing file when there is a problem reading it.
6488 26. Expand error message when OpenSSL has problems setting up cert/key files.
6491 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6492 (this can affect the format of dates).
6494 28. exim_tidydb, when checking for the continued existence of a message for
6495 which it has found a message-specific retry record, was not finding
6496 messages that were in split spool directories. Consequently, it was
6497 deleting retry records that should have stayed in existence.
6499 29. eximstats updated to version 1.35
6500 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6501 1.35 - bugfix such that pie charts by volume are generated correctly
6503 30. The SPA authentication driver was not abandoning authentication and moving
6504 on to the next authenticator when an expansion was forced to fail,
6505 contradicting the general specification for all authenticators. Instead it
6506 was generating a temporary error. It now behaves as specified.
6508 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6509 (the order specifies the preference for clients). The order is now AES256,
6510 AES128, 3DES, ARCFOUR128.
6512 31. Small patch to Sieve code - explicitly set From: when generating an
6515 32. Exim crashed if a remote delivery caused a very long error message to be
6516 recorded - for instance if somebody sent an entire SpamAssassin report back
6517 as a large number of 550 error lines. This bug was coincidentally fixed by
6518 increasing the size of one of Exim's internal buffers (big_buffer) that
6519 happened as part of the Exiscan merge. However, to be on the safe side, I
6520 have made the code more robust (and fixed the comments that describe what
6523 33. Some experimental protocols are using DNS PTR records for new purposes. The
6524 keys for these records are domain names, not reversed IP addresses. The
6525 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6526 leaves it alone. Component reversal etc. now happens only for IP addresses.
6529 34. The host_aton() function is supposed to be passed a string that is known
6530 to be a valid IP address. However, in the case of IPv6 addresses, it was
6531 not checking this. This is a hostage to fortune. Exim now panics and dies
6532 if the condition is not met. A case was found where this could be provoked
6533 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6534 components; fortuitously, this particular loophole had already been fixed
6535 by change 4.50/55 or 4.44/33 above.
6537 If there are any other similar loopholes, the new check in host_aton()
6538 itself should stop them being exploited. The report I received stated that
6539 data on the command line could provoke the exploit when Exim was running as
6540 exim, but did not say which command line option was involved. All I could
6541 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6542 running as the user.
6545 35. There was a buffer overflow vulnerability in the SPA authentication code
6546 (which came originally from the Samba project). I have added a test to the
6547 spa_base64_to_bits() function which I hope fixes it.
6550 36. The daemon start-up calls getloadavg() while still root for those OS that
6551 need the first call to be done as root, but it missed one case: when
6552 deliver_queue_load_max is set with deliver_drop_privilege. This is
6553 necessary for the benefit of the queue runner, because there is no re-exec
6554 when deliver_drop_privilege is set.
6556 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6557 was in use, was not putting the data itself into the right store pool;
6558 consequently, it could be overwritten for a subsequent message in the same
6559 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6562 38. Sometimes the final signoff response after QUIT could fail to get
6563 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6564 < 0 before doing a fflush(). This bug looks as though it goes back to the
6565 introduction of TLS in release 3.20, but "sometimes" must have been rare
6566 because the tests only now provoked it.
6572 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6573 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6574 Now it returns an integer. A number of calls were still expecting a T/F
6575 return. Fortuitously, in all cases, the tests worked in OK situations,
6576 which is the norm. However, things would have gone wrong on any write
6577 failures on the smtp file descriptor. This function is used when sending
6578 messages over SMTP and also when doing verify callouts.
6580 2. When Exim is called to do synchronous delivery of a locally submitted
6581 message (the -odf or -odi options), it no longer closes stderr before doing
6584 3. Implemented the mua_wrapper option.
6586 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6588 5. Implemented the functions header_remove(), header_testname(),
6589 header_add_at_position(), and receive_remove_recipient(), and exported them
6592 6. If an ACL "warn" statement specified the addition of headers, Exim already
6593 inserted X-ACL-Warn: at the start if there was no header name. However, it
6594 was not making this test for the second and subsequent header lines if
6595 there were newlines in the string. This meant that an invalid header could
6596 be inserted if Exim was badly configured.
6598 7. Allow an ACL "warn" statement to add header lines at the start or after all
6599 the Received: headers, as well as at the end.
6601 8. Added the rcpt_4xx retry error code.
6603 9. Added postmaster_mailfrom=xxx to callout verification option.
6605 10. Added mailfrom=xxxx to the callout verification option, for verify=
6608 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6609 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6611 12. Inserted some casts to stop certain compilers warning when using pointer
6612 differences as field lengths or precisions in printf-type calls (mostly
6613 affecting debugging statements).
6615 13. Added optional readline() support for -be (dynamically loaded).
6617 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6618 same clock tick as a message's arrival, so that its received time was the
6619 same as the "first fail" time on the retry record, and that message
6620 remained on the queue past the ultimate address timeout, every queue runner
6621 would try a delivery (because it was past the ultimate address timeout) but
6622 after another failure, the ultimate address timeout, which should have then
6623 bounced the address, did not kick in. This was a "< instead of <=" error;
6624 in most cases the first failure would have been in the next clock tick
6625 after the received time, and all would be well.
6627 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6628 being recognized when the domain list was tested by the match_domain
6629 condition in an expansion string.
6631 16. Added the ${str2b64: operator.
6633 17. Exim was always calling setrlimit() to set a large limit for the number of
6634 processes, without checking whether the existing limit was already
6635 adequate. (It did check for the limit on file descriptors.) Furthermore,
6636 errors from getrlimit() and setrlimit() were being ignored. Now they are
6637 logged to the main and panic logs, but Exim does carry on, to try to do its
6638 job under whatever limits there are.
6640 18. Imported PCRE 5.0.
6642 19. Trivial typo in log message " temporarily refused connection" (the leading
6645 20. If the log selector return_path_on_delivery was set and an address was
6646 redirected to /dev/null, the delivery process crashed because it assumed
6647 that a return path would always be set for a "successful" delivery. In this
6648 case, the whole delivery is bypassed as an optimization, and therefore no
6651 21. Internal re-arrangement: the function for sending a challenge and reading
6652 a response while authentication was assuming a zero-terminated challenge
6653 string. It's now changed to take a pointer and a length, to allow for
6654 binary data in such strings.
6656 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6658 23. Exim was not respecting finduser_retries when seeking the login of the
6659 uid under which it was called; it was always trying 10 times. (The default
6660 setting of finduser_retries is zero.) Also, it was sleeping after the final
6661 failure, which is pointless.
6663 24. Implemented tls_on_connect_ports.
6665 25. Implemented acl_smtp_predata.
6667 26. If the domain in control=submission is set empty, Exim assumes that the
6668 authenticated id is a complete email address when it generates From: or
6669 Sender: header lines.
6671 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6672 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6673 chown and chgrp in /bin and hostname in /usr/bin.
6675 28. Exim was keeping the "process log" file open after each use, just as it
6676 does for the main log. This opens the possibility of it remaining open for
6677 long periods when the USR1 signal hits a daemon. Occasional processlog
6678 errors were reported, that could have been caused by this. Anyway, it seems
6679 much more sensible not to leave this file open at all, so that is what now
6682 29. The long-running daemon process does not normally write to the log once it
6683 has entered its main loop, and it closes the log before doing so. This is
6684 so that log files can straightforwardly be renamed and moved. However,
6685 there are a couple of unusual error situations where the daemon does write
6686 log entries, and I had neglected to close the log afterwards.
6688 30. The text of an SMTP error response that was received during a remote
6689 delivery was being truncated at 512 bytes. This is too short for some of
6690 the long messages that one sometimes sees. I've increased the limit to
6693 31. It is now possible to make retry rules that apply only when a message has a
6694 specific sender, in particular, an empty sender.
6696 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6697 it possible to be selective about when SMTP synchronization is enforced.
6699 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6701 32. Implemented hosts_connection_nolog.
6703 33. Added an ACL for QUIT.
6705 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6708 35. Added mailbox_size and mailbox_filecount to appendfile.
6710 36. Added control = no_multiline_responses to ACLs.
6712 37. There was a bug in the logic of the code that waits for the clock to tick
6713 in the case where the clock went backwards by a substantial amount such
6714 that the microsecond fraction of "now" was more than the microsecond
6715 fraction of "then" (but the whole seconds number was less).
6717 38. Added support for the libradius Radius client library this is found on
6718 FreeBSD (previously only the radiusclient library was supported).
6724 1. When certain lookups returned multiple values in the form name=value, the
6725 quoting of the values was not always being done properly. Specifically:
6726 (a) If the value started with a double quote, but contained no whitespace,
6728 (b) If the value contained whitespace other than a space character (i.e.
6729 tabs or newlines or carriage returns) it was not quoted.
6730 This fix has been applied to the mysql and pgsql lookups by writing a
6731 separate quoting function and calling it from the lookup code. The fix
6732 should probably also be applied to nisplus, ibase and oracle lookups, but
6733 since I cannot test any of those, I have not disturbed their existing code.
6735 2. A hit in the callout cache for a specific address caused a log line with no
6736 reason for rejecting RCPT. Now it says "Previous (cached) callout
6737 verification failure".
6739 3. There was an off-by-one bug in the queryprogram router. An over-long
6740 return line was truncated at 256 instead of 255 characters, thereby
6741 overflowing its buffer with the terminating zero. As well as fixing this, I
6742 have increased the buffer size to 1024 (and made a note to document this).
6744 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6745 when Exim is waiting for an SMTP response from a remote server, Exim
6746 restarts its select() call on the socket, thereby resetting its timeout.
6747 This is not a problem when such interrupts are rare. Somebody set up a cron
6748 job to run exiwhat every 2 minutes, which is less than the normal select()
6749 timeout (5 or 10 minutes). This meant that the select() timeout never
6750 kicked in because it was always reset. I have fixed this by comparing the
6751 time when an interrupt arrives with the time at the start of the first call
6752 to select(). If more time than the timeout has elapsed, the interrupt is
6753 treated as a timeout.
6755 5. Some internal re-factoring in preparation for the addition of Sieve
6756 extensions (by MH). In particular, the "personal" test is moved to a
6757 separate function, and given an option for scanning Cc: and Bcc: (which is
6758 not set for Exim filters).
6760 6. When Exim created an email address using the login of the caller as the
6761 local part (e.g. when creating a From: or Sender: header line), it was not
6762 quoting the local part when it contained special characters such as @.
6764 7. Installed new OpenBSD configuration files.
6766 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6767 try to make them clearer.
6769 9. Callout options, other than the timeout value, were being ignored when
6770 verifying sender addresses in header lines. For example, when using
6772 verify = header_sender/callout=no_cache
6774 the cache was (incorrectly) being used.
6776 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6777 only the Cygwin environment.
6779 11. When return_path_on_delivery was set as a log selector, if different remote
6780 addresses in the same message used different return paths and parallel
6781 remote delivery occurred, the wrong values would sometimes be logged.
6782 (Whenever a remote delivery process finished, the return path value from
6783 the most recently started remote delivery process was logged.)
6785 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6786 lines when AUTH and/or TLS are in use. This is the "received protocol"
6787 field. Exim used to use "asmtp" for authenticated SMTP, without any
6788 indication (in the protocol name) for TLS use. Now it follows the RFC and
6789 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6790 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6791 names appear in log lines as well as in Received: header lines.
6793 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6794 extensions, and comparison tests, and to fix some bugs.
6796 14. Changes to the "personal" filter test:
6798 (1) The test was buggy in that it was just doing the equivalent of
6799 "contains" tests on header lines. For example, if a user's address was
6800 anne@some.where, the "personal" test would incorrectly be true for
6802 To: susanne@some.where
6804 This test is now done by extracting each address from the header in turn,
6805 and checking the entire address. Other tests that are part of "personal"
6806 are now done using regular expressions (for example, to check local parts
6807 of addresses in From: header lines).
6809 (2) The list of non-personal local parts in From: addresses has been
6810 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6811 taken from the Sieve specification recommendations.
6813 (3) If the message contains any header line starting with "List-" it is
6814 treated as non-personal.
6816 (4) The test for "circular" in the Subject: header line has been removed
6817 because it now seems ill-conceived.
6819 15. Minor typos in src/EDITME comments corrected.
6821 16. Installed latest exipick from John Jetmore.
6823 17. If headers_add on a router specified a text string that was too long for
6824 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6825 of string_sprintf() is now avoided.
6827 18. $message_body_size was not set (it was always zero) when running the DATA
6828 ACL and the local_scan() function.
6830 19. For the "mail" command in an Exim filter, no default was being set for
6831 the once_repeat time, causing a random time value to be used if "once" was
6832 specified. (If the value happened to be <= 0, no repeat happened.) The
6833 default is now 0s, meaning "never repeat". The "vacation" command was OK
6834 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6835 (I found it when inspecting the code).
6837 20. There is now an overall timeout for performing a callout verification. It
6838 defaults to 4 times the callout timeout, which applies to individual SMTP
6839 commands during the callout. The overall timeout applies when there is more
6840 than one host that can be tried. The timeout is checked before trying the
6841 next host. This prevents very long delays if there are a large number of
6842 hosts and all are timing out (e.g. when the network connections are timing
6843 out). The value of the overall timeout can be changed by specifying an
6844 additional sub-option for "callout", called "maxwait". For example:
6846 verify = sender/callout=5s,maxwait=20s
6848 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6849 to the end before writing, but this should make it even safer).
6851 22. Exim was forgetting that it had advertised PIPELINING for the second and
6852 subsequent messages on an SMTP connection. It was also not resetting its
6853 memory on STARTTLS and an internal HELO.
6855 23. When Exim logs an SMTP synchronization error within a session, it now
6856 records whether PIPELINING has been advertised or not.
6858 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6859 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6860 rather than long int.
6862 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6864 26. Added the never_mail option to autoreply.
6870 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6871 crash if the getsockname() call failed; this can happen if a connection is
6872 closed very soon after it is established. The problem was simply in the
6873 order in which certain operations were done, causing Exim to try to write
6874 to the SMTP stream before it had set up the file descriptor. The bug has
6875 been fixed by making things happen in the correct order.
6881 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6882 before the connection was closed, thus losing the rejection response.
6884 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6885 some early Solaris releases, but causes trouble in current releases where
6886 socklen_t is defined.
6888 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6891 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6894 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6895 that releasing the top of it at the end releases what was used for sub-
6896 expansions (unless the block got too big). However, discard this block if
6897 the first thing is a variable or header, so that we can use its block when
6898 it is dynamic (useful for very large $message_headers, for example).
6900 6. Lookups now cache *every* query, not just the most recent. A new, separate
6901 store pool is used for this. It can be recovered when all lookup caches are
6902 flushed. Lookups now release memory at the end of their result strings.
6903 This has involved some general refactoring of the lookup sources.
6905 7. Some code has been added to the store_xxx() functions to reduce the amount
6906 of flapping under certain conditions.
6908 8. log_incoming_interface used to affect only the <= reception log lines. Now
6909 it causes the local interface and port to be added to several more SMTP log
6910 lines, for example "SMTP connection from", and rejection lines.
6912 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6914 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6916 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6917 could be overwritten at the end of the current message (or the start of a
6918 new message if it was set in a HELO ACL). The value is now preserved for
6919 the duration of the SMTP connection.
6921 12. If a transport had a headers_rewrite setting, and a matching header line
6922 contained an unqualified address, that address was qualified, even if it
6923 did not match any rewriting rules. The underlying bug was that the values
6924 of the flags that permit the existence of unqualified sender and recipient
6925 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6926 non-local messages, and by -bnq for local messages) were not being
6927 preserved with the message after it was received.
6929 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6930 "next input=" as part of the text comprising the host identity instead of
6931 the correct text. The code was using the same buffer for two different
6932 strings. However, depending on which order the printing function evaluated
6933 its arguments, the bug did not always show up. Under Linux, for example, my
6934 test suite worked just fine.
6936 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6937 doesn't work with some older versions of Perl. It has been changed to "my",
6938 which in any case is probably the better facility to use.
6940 15. A really picky compiler found some instances of statements for creating
6941 error messages that either had too many or two few arguments for the format
6944 16. The size of the buffer for calls to the DNS resolver has been increased
6945 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6946 for addresses that have a lot of PTR records. This alleviates a problem; it
6947 does not fully solve it.
6949 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6950 buffer now truncates the list and logs the incident, which is the same
6951 action as happens when Exim is looking up a host name and its aliases.
6952 Previously in this situation something unpredictable would happen;
6953 sometimes it was "internal error: store_reset failed".
6955 18. If a server dropped the connection unexpectedly when an Exim client was
6956 using GnuTLS and trying to read a response, the client delivery process
6957 crashed while trying to generate an error log message.
6959 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6960 string, for example:
6962 warn message = H1: something\nH2: something
6964 the text was added as a single header line from Exim's point of view
6965 though it ended up OK in the delivered message. However, searching for the
6966 second and subsequent header lines using $h_h2: did not work. This has been
6967 fixed. Similarly, if a system filter added multiple headers in this way,
6968 the routers could not see them.
6970 20. Expanded the error message when iplsearch is called with an invalid key to
6971 suggest using net-iplsearch in a host list.
6973 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6974 ACLs are no longer actually imposed (and a message to that effect is
6977 22. If a "gecos" field in a passwd entry contained escaped characters, in
6978 particular, if it contained a \" sequence, Exim got it wrong when building
6979 a From: or a Sender: header from that name. A second bug also caused
6980 incorrect handling when an unquoted " was present following a character
6981 that needed quoting.
6983 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6984 was not being matched caselessly.
6986 24. Arranged for all hyphens in the exim.8 source to be escaped with
6989 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6990 properly. Recipient callout cache records were still being keyed to include
6991 the sender, even when use_sender was set false. This led to far more
6992 callouts that were necessary. The sender is no longer included in the key
6993 when use_sender is false.
6995 26. Added "control = submission" modifier to ACLs.
6997 27. Added the ${base62d: operator to decode base 62 numbers.
6999 28. dnsdb lookups can now access SRV records.
7001 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7002 the configuration file.
7004 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7005 (-v) mode. This makes the output for a verbose queue run more intelligible.
7007 31. Added a use_postmaster feature to recipient callouts.
7009 32. Added the $body_zerocount variable, containing the number of binary zero
7010 bytes in the message body.
7012 33. The time of last modification of the "new" subdirectory is now used as the
7013 "mailbox time last read" when there is a quota error for a maildir
7016 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7018 35. Added +ignore_unknown as a special item in host lists.
7020 36. Code for decoding IPv6 addresses in host lists is now included, even if
7021 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7022 address was recognized as an IP address, but was then not correctly decoded
7023 into binary, causing unexpected and incorrect effects when compared with
7030 1. Very minor rewording of debugging text in manualroute to say "list of
7031 hosts" instead of "hostlist".
7033 2. If verify=header_syntax was set, and a header line with an unqualified
7034 address (no domain) and a large number of spaces between the end of the
7035 name and the colon was received, the reception process suffered a buffer
7036 overflow, and (when I tested it) crashed. This was caused by some obsolete
7037 code that should have been removed. The fix is to remove it!
7039 3. When running in the test harness, delay a bit after writing a bounce
7040 message to get a bit more predictability in the log output.
7042 4. Added a call to search_tidyup() just before forking a reception process. In
7043 theory, someone could use a lookup in the expansion of smtp_accept_max_
7044 per_host which, without the tidyup, could leave open a database connection.
7046 5. Added the variables $recipient_data and $sender_data which get set from a
7047 lookup success in an ACL "recipients" or "senders" condition, or a router
7048 "senders" option, similar to $domain_data and $local_part_data.
7050 6. Moved the writing of debug_print from before to after the "senders" test
7053 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7054 problems for message scanning, either using a data ACL, or using
7055 local_scan() because the Received: header was not generated till after they
7056 were called (in order to set the time as the time of reception completion).
7057 I have revised the way this works. The header is now generated after the
7058 body is received, but before the ACL or local_scan() are called. After they
7059 are run, the timestamp in the header is updated.
7065 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7066 before starting a queue runner without re-exec. This happened only when
7067 deliver_drop_privilege was set or when the Exim user was set to root. The
7068 effect of the bug was that timeouts during subsequent deliveries caused
7069 crashes instead of being properly handled. The handler is now left at its
7070 default (and expected) setting.
7072 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7073 message, again when deliver_drop_privilege is set or Exim is run as root.
7074 The bug described in (1) was not present in this case, but the tidying up
7075 of the other signals was missing. I have made the two cases consistent.
7077 3. The ignore_target_hosts setting on a manualroute router was being ignored
7078 for hosts that were looked up using the /MX notation.
7080 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7083 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7084 operated on the sender address. After changing the $sender_address to <>
7085 for the sender address verify, Exim was re-instated it as the original
7086 (before rewriting) address, but remembering that it had rewritten it, so it
7087 wasn't rewriting it again. This bug also had the effect of breaking the
7088 sender address verification caching when the sender address was rewritten.
7090 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7091 This has been changed so that if the ip literal address matches
7092 ignore_target_hosts, the router declines.
7094 7. Added expansion conditions match_domain, match_address, and match_local_
7095 part (NOT match_host).
7097 8. The placeholder for the Received: header didn't have a length field set.
7099 9. Added code to Exim itself and to exim_lock to test for a specific race
7100 condition that could lead to file corruption when using MBX delivery. The
7101 issue is with the lockfile that is created in /tmp. If this file is removed
7102 after a process has opened it but before that process has acquired a lock,
7103 there is the potential for a second process to recreate the file and also
7104 acquire a lock. This could lead to two Exim processes writing to the file
7105 at the same time. The added code performs the same test as UW imapd; it
7106 checks after acquiring the lock that its file descriptor still refers to
7107 the same named file.
7109 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7110 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7113 11. Added the smtp_active_hostname option. If used, this will typically be made
7114 to depend on the incoming interface address. Because $interface_address is
7115 not set up until the daemon has forked a reception process, error responses
7116 that can happen earlier (such as "too many connections") no longer contain
7119 12. If an expansion in a condition on a "warn" statement fails because a lookup
7120 defers, the "warn" statement is abandoned, and the next ACL statement is
7121 processed. Previously this caused the whole ACL to be aborted.
7123 13. Added the iplsearch lookup type.
7125 14. Added ident_timeout as a log selector.
7127 15. Added tls_certificate_verified as a log selector.
7129 16. Added a global option tls_require_ciphers (compare the smtp transport
7130 option of the same name). This controls incoming TLS connections.
7132 17. I finally figured out how to make tls_require_ciphers do a similar thing
7133 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7134 before starting the TLS session.
7136 18. Tabs are now shown as \t in -bP output.
7138 19. If the log selector return_path_on_delivery was set, Exim crashed when
7139 bouncing a message because it had too many Received: header lines.
7141 20. If two routers both had headers_remove settings, and the first one included
7142 a superfluous trailing colon, the final name in the first list and the
7143 first name in the second list were incorrectly joined into one item (with a
7144 colon in the middle).
7150 1. Added -C and -D options to the exinext utility, mainly to make it easier
7151 to include in the automated testing, but these could be helpful when
7152 multiple configurations are in use.
7154 2. The exinext utility was not formatting the output nicely when there was
7155 an alternate port involved in the retry record key, nor when there was a
7156 message id as well (for retries that were specific to a specific message
7157 and a specific host). It was also confused by IPv6 addresses, because of
7158 the additional colons they contain. I have fixed the IPv4 problem, and
7159 patched it up to do a reasonable job for IPv6.
7161 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7162 delivery, the log line now contains "pipelined" if PIPELINING was used.
7164 4. An SMTP transport process used to panic and die if the bind() call to set
7165 an explicit outgoing interface failed. This has been changed; it is now
7166 treated in the same way as a connect() failure.
7168 5. A reference to $sender_host_name in the part of a conditional expansion
7169 that was being skipped was still causing a DNS lookup. This no longer
7172 6. The def: expansion condition was not recognizing references to header lines
7173 that used bh_ and bheader_.
7175 7. Added the _cache feature to named lists.
7177 8. The code for checking quota_filecount in the appendfile transport was
7178 allowing one more file than it should have been.
7180 9. For compatibility with Sendmail, the command line option
7188 and sets the incoming protocol and host name (for trusted callers). The
7189 host name and its colon can be omitted when only the protocol is to be set.
7190 Note the Exim already has two private options, -pd and -ps, that refer to
7191 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7192 "s", but I don't think that's a major issue.
7194 10. A number of refactoring changes to the code, none of which should affect
7197 (a) The number of logging options was getting close to filling up the
7198 32-bit word that was used as a bit map. I have split them into two classes:
7199 those that are passed in the argument to log_write(), and those that are
7200 only ever tested independently outside of that function. These are now in
7201 separate 32-bit words, so there is plenty of room for expansion again.
7202 There is no change in the user interface or the logging behaviour.
7204 (b) When building, for example, log lines, the code previously used a
7205 macro that called string_cat() twice, in order to add two strings. This is
7206 not really sufficiently general. Furthermore, there was one instance where
7207 it was actually wrong because one of the argument was used twice, and in
7208 one call a function was used. (As it happened, calling the function twice
7209 did not affect the overall behaviour.) The macro has been replaced by a
7210 function that can join an arbitrary number of extra strings onto a growing
7213 (c) The code for expansion conditions now uses a table and a binary chop
7214 instead of a serial search (which was left over from when there were very
7215 few conditions). Also, it now recognizes conditions like "pam" even when
7216 the relevant support is not compiled in: a suitably worded error message is
7217 given if an attempt is made to use such a condition.
7219 11. Added ${time_interval:xxxxx}.
7221 12. A bug was causing one of the ddress fields not to be passed back correctly
7222 from remote delivery subprocesses. The field in question was not being
7223 subsequently used, so this caused to problems in practice.
7225 13. Added new log selectors queue_time and deliver_time.
7227 14. Might have fixed a bug in maildirsizefile handling that threw up
7228 "unexpected character" debug warnings, and recalculated the data
7229 unnecessarily. In any case, I expanded the warning message to give more
7232 15. Added the message "Restricted characters in address" to the statements in
7233 the default ACL that block characters like @ and % in local parts.
7235 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7236 Three changes have been made:
7238 (a) There was a serious bug; a negative response to MAIL caused the whole
7239 recipient domain to be cached as invalid, thereby blocking all messages
7240 to all local parts at the same domain, from all senders. This bug has
7241 been fixed. The domain is no longer cached after a negative response to
7242 MAIL if the sender used is not empty.
7244 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7247 (c) A new callout option, "use_sender" has been added for people who want
7248 the modified behaviour.
7254 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7257 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7258 indeed breaks things for older releases.
7260 3. Added additional logging to the case where there is a problem reading data
7261 from a filter that is running in a subprocess using a pipe, in order to
7262 try to track down a specific problem.
7264 4. Testing facility fudge: when running in the test harness and attempting
7265 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7266 getting "No route to host". Convert this to a timeout.
7268 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7271 6. Some OS don't have socklen_t but use size_t instead. This affects the
7272 fifth argument of getsockopt() amongst other things. This is now
7273 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7274 can be set for individual OS. I have set it for SunOS5, OSF1, and
7275 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7276 some earlier ones do not.
7278 7. Change 4.30/15 was not doing the test caselessly.
7280 8. The standard form for an IPv6 address literal was being rejected by address
7281 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7282 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7283 this, as well as the form without the "IPv6" on the front (but only when
7284 address literals are enabled, of course).
7286 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7288 10. Exim crashed if a message with an empty sender address specified by -f
7289 encountered a router with an errors_to setting. This could be provoked only
7290 by a command such as
7294 where an empty string was supplied; "<>" did not hit this bug.
7296 11. Installed PCRE release 4.5.
7298 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7299 remained set. It is now erased.
7301 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7302 times from message ids (which are base 36 rather than the normal 62).
7304 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7305 were being counted as actual protocol errors, and logged if the log
7306 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7307 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7308 and DATA following a set of rejected RCPTs do not count as protocol errors.
7309 In other words, Exim assumes they were pipelined, though this may not
7310 actually be the case. Of course, in all cases the client gets an
7311 appropriate error code.
7313 15. If a lookup fails in an ACL condition, a message about the failure may
7314 be available; it is used if testing the ACL cannot continue, because most
7315 such messages specify what the cause of the deferral is. However, some
7316 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7317 that caused an old message to be retained and used if a later statement
7318 caused a defer, replacing the real cause of the deferral.
7320 16. If an IP address had so many PTR records that the DNS lookup buffer
7321 was not large enough to hold them, Exim could crash while trying to process
7322 the truncated data. It now detects and logs this case.
7324 17. Further to 4.21/58, another change has been made: if (and only if) the
7325 first line of a message (the first header line) ends with CRLF, a bare LF
7326 in a subsequent header line has a space inserted after it, so as not to
7327 terminate the header.
7329 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7330 unnecessarily, used atoi() instead of strtol(), and didn't check the
7331 termination when getting file sizes from file names by regex.
7333 19. Completely re-implemented the support for maildirsize files, in the light
7334 of a number of problems with the previous contributed implementation
7335 (4.30/29). In particular:
7337 . If the quota is zero, the maildirsize file is maintained, but no quota is
7340 . If the maildir directory does not exist, it is created before any attempt
7341 to write a maildirsize file.
7343 . The quota value in the file is just a cache; if the quota is changed in
7344 the transport, the new value overrides.
7346 . A regular expression is available for excluding directories from the
7349 20. The autoreply transport checks the characters in options that define the
7350 message's headers; it allows continued headers, but it was checking with
7351 isspace() after an embedded newline instead of explicitly looking for a
7354 21. If all the "regular" hosts to which an address was routed had passed their
7355 expiry times, and had not reached their retry times, the address was
7356 bounced, even if fallback hosts were defined. Now Exim should go on to try
7359 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7360 equivalent code in the SMTP transport. Some hosts send humungous responses
7361 to HELO/EHLO, more than 1024 it seems.
7363 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7364 gives compiler warnings in some environments. I've now done it "properly",
7367 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7368 (because of problems with the built-in one) was declared to return uschar *
7369 instead of char *, causing compiler failure.
7371 25. Fixed a file descriptor leak when processing alias/forward files.
7373 26. Fixed a minor format string issue in dbfn.c.
7375 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7377 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7378 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7381 29. When a sender address is verified, it is cached, to save repeating the test
7382 when there is more than one recipient in a message. However, when the
7383 verification involves a callout, it is possible for different callout
7384 options to be set for different recipients. It is too complicated to keep
7385 track of this in the cache, so now Exim always runs a verification when a
7386 callout is required, relying on the callout cache for the optimization.
7387 The overhead is duplication of the address routing, but this should not be
7390 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7391 to be verified with callout=postmaster, and the main callout worked but the
7392 postmaster check failed, the verification correctly failed. However, if a
7393 subsequent RCPT command asked for sender verification *without* the
7394 postmaster check, incorrect caching caused this verification also to fail,
7397 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7398 it was not caching the DNS options (qualify_single, search_parents) that
7399 were used when the lookup failed. A subsequent lookup with different
7400 options therefore always gave the same answer, though there were cases
7401 where it should not have. (Example: a "domains = !$mx_any" option on a
7402 dnslookup router: the "domains" option is always processed without any
7403 widening, but the router might have qualify_single set.) Now Exim uses the
7404 cached value only when the same options are set.
7406 32. Added John Jetmore's "exipick" utility to the distribution.
7408 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7409 than a timeout (e.g. a certificate is required, and is not provided), an
7410 Exim server now closes the connection immediately. Previously it waited for
7411 the client to close - but if the client is SSL, it seems that they each
7412 wait for each other, leading to a delay before one of them times out.
7414 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7415 maintained 0.8.x compatibility because I don't think many are using it, and
7416 it is clearly obsolete.
7418 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7421 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7422 client certificate was expired. A simple patch fixes this, though I don't
7423 understand the full logic of why the verify callback is called multiple
7426 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7427 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7428 which causes problems with some clients (such as the Certicom SSL Plus
7429 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7430 disables the coutermeasure allowing Eudora to connect."
7432 38. Exim was not checking that a write() to a log file succeeded. This could
7433 lead to Bad Things if a log got too big, in particular if it hit a file
7434 size limit. Exim now panics and dies if it cannot write to a log file, just
7435 as it does if it cannot open a log file.
7437 39. Modified OS/Makefile-Linux so that it now contains
7439 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7441 The two -D definitions ensure that Exim is compiled with large file
7442 support, which makes it possible to handle log files that are bigger than
7445 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7446 instance) a domain was checked against a named list that involved a lookup,
7447 causing $domain_data to be set, then another domain was checked against the
7448 same list, then the first domain was re-checked, the value of $domain_data
7449 after the final check could be wrong. In particular, if the second check
7450 failed, it could be set empty. This bug probably also applied to
7453 41. The strip_trailing_dot option was not being applied to the address given
7454 with the -f command-line option.
7456 42. The code for reading a message's header from the spool was incrementing
7457 $received_count, but never initializing it. This meant that the value was
7458 incorrect (doubled) while delivering a message in the same process in which
7459 it was received. In the most common configuration of Exim, this never
7460 happens - a fresh exec is done - but it can happen when
7461 deliver_drop_privilege is set.
7463 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7464 it now includes up to 150 characters of the unexpected data in the log
7467 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7468 and building data strings. The size of both of these buffers was 10 000
7469 bytes - far larger than anybody would *ever* want, thought I. Needless to
7470 say, somebody hit the limit. I have increased the maximum line length to
7471 20 000 and the maximum data length of concatenated lines to 100 000. I have
7472 also fixed two bugs, because there was no checking on these buffers. Tsk,
7473 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7474 buffer is too small.
7476 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7477 lsearch lookups. Now it does.
7479 46. When parsing a route_list item in a manualroute router, a fixed-length
7480 buffer was used for the list of hosts. I made this 1024 bytes long,
7481 thinking that nobody would ever have a list of hosts that long. Wrong.
7482 Somebody had a whole pile of complicated expansion conditions, and the
7483 string was silently truncated, leading to an expansion error. It turns out
7484 that it is easier to change to an unlimited length (owing to other changes
7485 that have happened since this code was originally written) than to build
7486 structure for giving a limitation error. The length of the item that
7487 expands into the list of hosts is now unlimited.
7489 47. The lsearch lookup could not handle data where the length of text line was
7490 more than 4095 characters. Such lines were truncated, leading to shortened
7491 data being returned. It should now handle lines of any length.
7493 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7494 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7497 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7498 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7499 no longer happens in comment lines. A list of replacements is now placed
7500 at the head of all of the source files, except those whose only change is
7501 to replace PERL_COMMAND in the very first #! line.
7503 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7504 messages on the queue, with a bottom-up merge sort, using code contributed
7505 by Michael Haardt. This should make operations like -bp somewhat faster on
7506 large queues. It won't affect queue runners, except when queue_run_in_order
7509 51. Installed eximstats 1.31 in the distribution.
7511 52. Added support for SRV lookups to the dnslookup router.
7513 53. If an ACL referred to $message_body or $message_body_end, the value was not
7514 reset for any messages that followed in the same SMTP session.
7516 54. The store-handling optimization for building very long strings was not
7517 differentiating between the different store pools. I don't think this
7518 actually made any difference in practice, but I've tidied it.
7520 55. While running the routers to verify a sender address, $sender_address
7521 was still set to the sender address. This is wrong, because when routing to
7522 send a bounce to the sender, it would be empty. Therefore, I have changed
7523 it so that, while verifying a sender address, $sender_address is set to <>.
7524 (There is no change to what happens when verifying a recipient address.)
7526 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7527 target A or AAAA records (if not already returned) without resetting the
7528 qualify_single or search_parents options of the DNS resolver. These are
7529 inappropriate in this case because the targets of MX and SRV records must
7530 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7531 target that, when qualified, matched something in the local domain. These
7532 two options are now turned off when doing these lookups.
7534 57. It seems that at least some releases of Reiserfs (which does not have the
7535 concept of a fixed number of inodes) returns zero and not -1 for the
7536 number of available inodes. This interacted badly with check_spool_inodes,
7537 which assumed that -1 was the "no such thing" setting. What I have done is
7538 to check that the total number of inodes is greater than zero before doing
7539 the test of how many are available.
7541 58. When a "warn" ACL statement has a log_message modifier, the message is
7542 remembered, and not repeated. This is to avoid a lot of repetition when a
7543 message has many recipients that cause the same warning to be written.
7544 However, Exim was preserving the list of already written lines for an
7545 entire SMTP session, which doesn't seem right. The memory is now reset if a
7546 new message is started.
7548 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7549 debugging output unless log_rewrite was also set.
7551 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7552 of (int)(handle) when we know that handle contains (void *)(-1).
7554 61. The Exim daemon panic-logs an error return when it closes the incoming
7555 connection. However "connection reset by peer" seems to be common, and
7556 isn't really an error worthy of noting specially, so that particular error
7559 62. When Exim is trying to find all the local interfaces, it used to panic and
7560 die if the ioctl to get the interface flags failed. However, it seems that
7561 on at least one OS (Solaris 9) it is possible to have an interface that is
7562 included in the list of interfaces, but for which you get a failure error
7563 for this call. This happens when the interface is not "plumbed" into a
7564 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7565 failure of the "get flags" call assumes that the interface is down.
7567 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7568 makes life easier for people who are doing arithmetic on fields extracted
7569 from dates, where you often get leading zeros that should not be
7570 interpreted as octal.
7572 64. Added qualify_domain to the redirect router, to override the global
7575 65. If a pathologically long header line contained very many addresses (the
7576 report of this problem mentioned 10 000) and each of them was rewritten,
7577 Exim could use up a very large amount of memory. (It kept on making new
7578 copies of the header line as it rewrote, and never released the old ones.)
7579 At the expense of a bit more processing, the header rewriting function has
7580 been changed so that it no longer eats memory in this way.
7582 66. The generation of the Received: header has been moved from the time that a
7583 message starts to be received, to the time that it finishes. The timestamp
7584 in the Received: header should now be very close to that of the <= log
7585 line. There are two side-effects of this change:
7587 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7588 logged header lines no longer include the local Received: line, because
7589 it has not yet been created. The same applies to a copy of the message
7590 that is returned to a non-SMTP sender when a message is rejected.
7592 (b) When a filter file is tested using -bf, no additional Received: header
7593 is added to the test message. After some thought, I decided that this
7596 This change does not affect the value of $received_for. It is still set
7597 after address rewriting, but before local_scan() is called.
7599 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7601 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7602 gave an unhelpful panic error message, and a defer error. I have managed to
7603 change this behaviour so that it now rejects any supplied certificate,
7604 which seems right, as the list of acceptable certificates is empty.
7606 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7607 gave an unhelpful defer error. I have not managed to make this reject any
7608 supplied certificates, but the error message it gives is "no certificate
7609 supplied", which is not helpful.
7611 70. exigrep's output now also includes lines that are not associated with any
7612 message, but which match the given pattern. Implemented by a patch from
7613 Martin Sluka, which also tidied up the Perl a bit.
7615 71. Recipient callout verification, like sender verification, was using <> in
7616 the MAIL FROM command. This isn't really the right thing, since the actual
7617 sender may affect whether the remote host accepts the recipient or not. I
7618 have changed it to use the actual sender in the callout; this means that
7619 the cache record is now keyed on a recipient/sender pair, not just the
7620 recipient address. There doesn't seem to be a real danger of callout loops,
7621 since a callout by the remote host to check the sender would use <>.
7622 [SEE ABOVE: changed after hitting problems.]
7624 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7625 temporary errors. However, in the case of such a code being given after
7626 the end of a data transmission (i.e. after ".") Exim was failing to write
7627 a retry record for the message. (Yes, there was some broken host that was
7628 actually sending 8xx at this point.)
7630 73. An unknown lookup type in a host list could cause Exim to panic-die when
7631 the list was checked. (An example that provoked this was putting <; in the
7632 middle of a list instead of at the start.) If this happened during a DATA
7633 ACL check, a -D file could be left lying around. This kind of configuration
7634 error no longer causes Exim to die; instead it causes a defer error. The
7635 incident is still logged to the main and panic logs.
7637 74. Buglet left over from Exim 3 conversion. The message "too many messages
7638 in one connection" was written to the rejectlog but not the mainlog, except
7639 when address rewriting (yes!) was being logged.
7641 75. Added write_rejectlog option.
7643 76. When a system filter was run not as root (that is, when system_filter_user
7644 was set), the values of the $n variables were not being returned to the
7645 main process; thus, they were not subsequently available in the $sn
7648 77. Added +return_path_on_delivery log selector.
7650 78. A connection timeout was being treated differently from recipients deferred
7651 when testing hosts_max_try with a message that was older than the host's
7652 retry timeout. (The host should not be counted, thus allowing all hosts to
7653 be tried at least once before bouncing.) This may have been the cause of an
7654 occasionally reported bug whereby a message would remain on the queue
7655 longer than the retry timeout, but would be bounced if a delivery was
7656 forced. I say "may" because I never totally pinned down the problem;
7657 setting up timeout/retry tests is difficult. See also the next item.
7659 79. The ultimate address timeout was not being applied to errors that involved
7660 a combination of host plus message (for example, a timeout on a MAIL
7661 command). When an address resolved to a number of possible hosts, and they
7662 were not all tried for each delivery (e.g. because of hosts_max_try), a
7663 message could remain on the queue longer than the retry timeout.
7665 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7668 81. Fixed an obscure SMTP outgoing bug which required at least the following
7669 conditions: (a) there was another message waiting for the same server;
7670 (b) the server returned 5xx to all RCPT commands in the first message so
7671 that the message was not completed; (c) the server dropped the connection
7672 or gave a negative response to the RSET that Exim sends to abort the
7673 transaction. The observed case was a dropped connection after DATA that had
7674 been sent in pipelining mode. That is, the server had advertised PIPELINING
7675 but was not implementing it correctly. The effect of the bug was incorrect
7676 behaviour, such as trying another host, and this could lead to a crash.
7682 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7683 and daemon.c were passed as pointers to ints; they should have been
7684 pointers to socklen_t variables (which are typically unsigned ints).
7686 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7689 3. Fixed a really odd bug that affected only the testing scheme; patching a
7690 certain fixed string in the binary changed the value of another string that
7691 happened to be identical to the end of the original first string.
7693 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7694 name", it returns that address as the IP address. On some operating
7695 systems (e.g. Solaris), it also passes back the IP address string as the
7696 "host name". However, on others (e.g. Linux), it passes back an empty
7697 string. Exim wasn't checking for this, and was changing the host name to an
7698 empty string, assuming it had been canonicalized.
7700 5. Although rare, it is permitted to have more than one PTR record for a given
7701 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7702 all the names associated with an address, because they do in Solaris.
7703 However, it seems that they do not in Linux for data that comes from the
7704 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7705 I found this out when I moved to a new Linux workstation and tried to run
7706 the Exim test suite.
7708 To get round this problem I have changed the code so that it now does its
7709 own call to the DNS to look up PTR records when searching for a host name.
7710 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7711 addresses that are only in /etc/hosts are still found.
7713 This behaviour is, however, controlled by an option called host_lookup_
7714 order, which defaults to "bydns:byaddr". If people want to use the other
7715 order, or indeed, just use one or the other means of lookup, they can
7716 specify it in this variable.
7718 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7719 some operating systems, this comes back from gethostbyaddr() as an empty
7720 string, and this is what Exim used to test for. However, it seems that in
7721 other systems, "." is yielded. Exim now tests for this case too.
7723 7. The values of check_spool_space and check_log_space are now held internally
7724 as a number of kilobytes instead of an absolute number of bytes. If a
7725 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7726 kilobyte. This means that much larger values can be stored.
7728 8. Exim monitor: an attempt to get the action menu when not actually pointing
7729 at a message produces an empty menu entitled "No message selected". This
7730 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7731 no entries in it ("Shell widget menu has zero width and/or height"). So I
7732 have added a single, blank menu entry in this case.
7734 9. Added ${quote_local_part.
7736 10. MIME decoding is now applied to the contents of Subject: header lines when
7739 11. Now that a reference to $sender_host_address automatically causes a reverse
7740 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7741 host lookup before query-style lookups in lists that might use this
7742 variable. This has therefore been abolished, and the "net-" prefix is no
7743 longer necessary for query-style lookups.
7745 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7746 have been a typo for LFLAGS, so it has been changed.
7748 13. The install script calls Exim with "-C /dev/null" in order to find the
7749 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7750 to be output. However, since Exim outputs its version number before the
7751 error, it didn't break the script. It just looked ugly. I fixed this by
7752 always allowing "-C /dev/null" if the caller is root.
7754 14. Ignore overlarge ACL variable number when reading spool file - insurance
7755 against a later release with more variables having written the file.
7757 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7758 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7759 this, as well as the form without the "IPv6" on the front.
7761 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7762 OS/Makefile-Darwin file.
7764 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7765 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7767 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7768 variable. However, some OS use ints for inodes. I've added cast to long int
7769 to get rid of the compiler warning.
7771 19. I had forgotten to lock out "/../" in configuration file names when
7772 ALT_CONFIG_PREFIX was set.
7774 20. Routers used for verification do not need to specify transports. However,
7775 if such a router generated a host list, and callout was configured, Exim
7776 crashed, because it could not find a port number from the (non-existent)
7777 transport. It now assumes port 25 in this circumstance.
7779 21. Added the -t option to exigrep.
7781 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7782 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7783 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7784 them is set, LOOKUP_LSEARCH is forced.
7786 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7788 24. Added sender and host information to the "rejected by local_scan()" log
7789 line; previously there was no indication of these.
7791 25. Added .include_if_exists.
7793 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7794 Linux. It turns out that not all file systems support this. Apparently some
7795 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7796 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7797 not supported on the file descriptor, is now ignored when Exim is trying to
7798 sync a directory. This applies only to Linux.
7800 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7802 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7803 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7806 29. Added support for maildirsize files from supplied patch (modified a bit).
7808 30. The use of :fail: followed by an empty string could lead Exim to respond to
7809 sender verification failures with (e.g.):
7811 550 Verification failed for <xxx>
7812 550 Sender verify failed
7814 where the first response line was missing the '-' that indicates it is not
7815 the final line of the response.
7817 31. The loop for finding the name of the user that called Exim had a hardwired
7818 limit of 10; it now uses the value of finduser_retries, which is used for
7819 all other user lookups.
7821 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7824 33. Exim was neglecting to zero errno before one call of strtol() when
7825 expanding a string and expecting an integer value. On some systems this
7826 resulted in spurious "integer overflow" errors. Also, it was casting the
7827 result into an int without checking.
7829 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7830 did not work. The code looks as if it has *never* worked, though it appears
7831 to have been documented since at least release 1.62. I have made it work.
7833 35. The "timeout_DNS" error in retry rules, also documented since at least
7834 1.62, also never worked. As it isn't clear exactly what this means, and
7835 clearly it isn't a major issue, I have abolished the feature by treating it
7836 as "timeout", and writing a warning to the main and panic logs.
7838 36. The display of retry rules for -brt wasn't always showing the error code
7841 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7842 timeout_connect_A, timeout_connect_MX.
7844 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7845 to the empty sender.
7847 39. The daemon was not analysing the content of -oX till after it had closed
7848 stderr and disconnected from the controlling terminal. This meant that any
7849 syntax errors were only noted on the panic log, and the return code from
7850 the command was 0. By re-arranging the code a little, I've made the
7851 decoding happen first, so such errors now appear on stderr, and the return
7852 code is 1. However, the actual setting up of the sockets still happens in
7853 the disconnected process, so errors there are still only recorded on the
7856 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7857 connections (as happens on some IP stacks) was logged at start up time as
7858 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7859 it from "IPv6 and IPv4", which means that two separate sockets are being
7862 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7863 says whether AF_INET or AF_INET6 was passed as an argument.
7865 42. Exiwhat output was messed up when time zones were included in log
7868 43. Exiwhat now gives more information about the daemon's listening ports,
7869 and whether -tls-on-connect was used.
7871 44. The "port" option of the smtp transport is now expanded.
7873 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7874 silently ignored. Now an error message is written to the main and panic
7877 46. There's a new ACL modifier called "logwrite" which writes to a log file
7878 as soon as it is encountered.
7880 47. Added $local_user_uid and $local_user_gid at routing time.
7882 48. Exim crashed when trying to verify a sender address that was being
7885 49. Exim was recognizing only a space character after ".include". It now also
7886 recognizes a tab character.
7888 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7889 extracting the relevant information from the specification. The man page no
7890 longer contains scrambled data for the -d option, and I've added a section
7891 at the front about calling Exim under different names.
7893 51. Added "extra_headers" argument to the "mail" command in filter files.
7895 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7898 53. Installed eximstats 1.29.
7900 54. Added transport_filter_timeout as a generic transport option.
7902 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7903 Cc: header lines. This was required by RFC 822, but it not required by RFC
7906 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7907 incoming messages that did not have them. Now it does so only if the
7908 message originates locally, that is, if there is no associated remote host
7909 address. When Resent- header lines are present, this applies to the Resent-
7910 lines rather than the non-Resent- lines.
7912 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7913 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7915 58. Messages for configuration errors now include the name of the main
7916 configuration file - useful now that there may be more than one file in a
7917 list (.included file names were always shown).
7919 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7920 for those rare installations that do not start the daemon as root or run it
7921 setuid root. I've cut out the call to initgroups() if the daemon is not
7924 60. The Exim user and group can now be bound into the binary as text strings
7925 that are looked up at the start of Exim's processing.
7927 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7929 62. Added $mailstore_basename variable.
7931 63. Installed patch to sieve.c from Michael Haardt.
7933 64. When Exim failed to open the panic log after failing to open the main log,
7934 the original message it was trying to log was written to stderr and debug
7935 output, but if they were not available (the usual case in production), it
7936 was lost. Now it is written to syslog before the two lines that record the
7937 failures to open the logs.
7939 65. Users' Exim filters run in subprocesses under the user's uid. It is
7940 possible for a "deliver" command or an alias in a "personal" command to
7941 provoke an address rewrite. If logging of address rewriting is configured,
7942 this fails because the process is not running as root or exim. There may be
7943 a better way of dealing with this, but for the moment (because 4.30 needs
7944 to be released), I have disabled address rewrite logging when running a
7945 filter in a non-root, non-exim process.
7951 1. The buildconfig auxiliary program wasn't quoting the value set for
7952 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7953 not defined. This bug was masked in 4.22 by the effect that was fixed in
7956 2. Some messages that were rejected after a message id was allocated were
7957 shown as "incomplete" by exigrep. It no longer does this for messages that
7958 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7960 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7961 have allow_domain_literals set, the ID did not get logged in the <= line.
7962 Domain literals are now always recognized in Message-ID: header lines.
7964 4. The first argument for a ${extract expansion item is the key name or field
7965 number. Leading and trailing spaces in this item were not being ignored,
7966 causing some misleading effects.
7968 5. When deliver_drop_privilege was set, single queue runner processes started
7969 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7970 same command in the process it spins off) were not dropping privilege.
7972 6. When the daemon running as "exim" started a queue runner, it always
7973 re-executed Exim in the spun-off process. This is a waste of effort when
7974 deliver_drop_privilege is set. The new process now just calls the
7975 queue-runner function directly.
7981 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7984 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7985 ignored. Though the use of -oP was forcing the writing of a pid file, it
7986 was always written to the default place.
7988 3. If the message "no IP address found for host xxxx" is generated during
7989 incoming verification, it is now followed by identification of the incoming
7990 connection (so you can more easily find what provoked it).
7992 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7994 5. Added some features to "harden" Exim a bit more against certain attacks:
7996 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7997 be put in Local/Makefile. This is like the never_users runtime option,
7998 but it cannot be overridden. The default setting is "root".
8000 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8001 prefix string with which any file named in a -C command line option
8004 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8005 is retained for -C and -D only if the caller of Exim is root. Without
8006 it, the exim user may also use -C and -D and retain privilege.
8008 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8009 command line option is disabled.
8011 6. Macro names set by the -D option must start with an upper case letter, just
8012 like macro names defined in the configuration file.
8014 7. Added "dereference=" facility to LDAP.
8016 8. Two instances of the typo "uknown" in the source files are fixed.
8018 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8019 the Configure-Makefile script screwed up while processing it.
8021 10. Incorporated PCRE 4.4.
8023 11. The SMTP synchronization check was not operating right at the start of an
8024 SMTP session. For example, it could not catch a HELO sent before the client
8025 waited for the greeting. There is now a check for outstanding input at the
8026 point when the greeting is written. Because of the duplex, asynchronous
8027 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8028 way, but not yet received, when the check is performed.
8030 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8031 on TCP/IP sockets, because this apparently causes some broken clients to
8034 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8035 unchanged) from the Cygwin maintainer.
8037 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8038 was supported instead of testing for mbx. Effectively a typo.
8040 15. The spa authenticator server code was not checking that the input it
8041 received was valid base64.
8043 16. The debug output line for the "set" modifier in ACLs was not showing the
8044 name of the variable that was being set.
8046 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8048 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8049 lookup on its own. Something else was needed to trigger the lookup. For
8050 example, a match in host_lookup or the need for a host name in a host list.
8051 Now, if $sender_host_name is referenced and the host name has not yet been
8052 looked up, a lookup is performed. If the lookup fails, the variable remains
8053 empty, and $host_lookup_failed is set to "1".
8055 19. Added "eqi" as a case-independent comparison operator.
8057 20. The saslauthd authentication condition could segfault if neither service
8058 nor realm was specified.
8060 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8061 error message that was logged was misleading, and incoming SMTP
8062 connections were dropped. The message is now more accurate, and temporary
8063 errors are given to SMTP connections.
8065 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8066 (except RSET) with a 421 error, until QUIT is received. However, it was
8067 failing to send a response to QUIT.
8069 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8070 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8071 ACL is now run after the helo_try_verify_hosts code.
8073 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8074 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8075 case-independent, so other case variants are also recognized). Apparently
8076 some systems use these upper case variants.
8078 25. If more than two messages were waiting for the same host, and a transport
8079 filter was specified for the transport, Exim sent two messages over the
8080 same TCP/IP connection, and then failed with "socket operation on non-
8081 socket" when it tried to send the third.
8083 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8085 27. The extern definition of crypt16() in expand.c was not being excluded when
8086 the OS had its own crypt16() function.
8088 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8089 as a preferred synonym for return_size_limit, both as an option and as an
8092 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8094 30. Changed the default configuration ACL to relax the local part checking rule
8095 for addresses that are not in any local domains. For these addresses,
8096 slashes and pipe symbols are allowed within local parts, but the sequence
8097 /../ is explicitly forbidden.
8099 31. SPA server authentication was not clearing the challenge buffer before
8102 32. log_message in a "warn" ACL statement was writing to the reject log as
8103 well as to the main log, which contradicts the documentation and doesn't
8104 seem right (because no rejection is happening). So I have stopped it.
8106 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8107 However, I am unable to do any testing of this.
8109 34. Fixed an infelicity in the appendfile transport. When checking directories
8110 for a mailbox, to see if any needed to be created, it was accidentally
8111 using path names with one or more superfluous leading slashes; tracing
8112 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8114 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8115 contents are added to the log line that is written for every discarded
8116 recipient. (Previously a log_message setting was ignored.)
8118 36. The ${quote: operator now quotes the string if it is empty.
8120 37. The install script runs exim in order to find its version number. If for
8121 some reason other than non-existence or emptiness, which it checks, it
8122 could not run './exim', it was installing it with an empty version number,
8123 i.e. as "exim-". This error state is now caught, and the installation is
8126 38. An argument was missing from the function that creates an error message
8127 when Exim fails to connect to the socket for saslauthd authentication.
8128 This could cause Exim to crash, or give a corrupted message.
8130 39. Added isip, isip4, and isip6 to ${if conditions.
8132 40. The ACL variables $acl_xx are now saved with the message, and can be
8133 accessed later in routers, transports, and filters.
8135 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8136 strings in the file are not string-expanded.
8138 42. If a MAIL command specified a SIZE value that was too large to fit into an
8139 int variable, the check against message_size_limit failed. Such values are
8140 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8141 day this will have to be increased, but I don't think I want to be around
8142 when emails are that large.
8149 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8150 iconv() is not standard in FreeBSD.
8152 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8153 IPv6 enabled. The observed symptom was a segmentation fault on return from
8154 the function os_common_find_running_interfaces() in src/os.c.
8156 3. In the check_special_case() function in daemon.c I had used "errno" as an
8157 argument name, which causes warnings on some systems. This was basically a
8158 typo, since it was named "eno" in the comments!
8160 4. The code that waits for the clock to tick (at a resolution of some fraction
8161 of a second) so as to ensure message-id uniqueness was always waiting for
8162 at least one whole tick, when it could have waited for less. [This is
8163 almost certainly not relevant at current processor speeds, where it is
8164 unlikely to ever wait at all. But we try to future-proof.]
8166 5. The function that sleeps for a time interval that includes fractions of a
8167 second contained a race. It did not block SIGALRM between setting the
8168 timer, and suspending (a couple of lines later). If the interval was short
8169 and the sigsuspend() was delayed until after it had expired, the suspension
8170 never ended. On busy systems this could lead to processes getting stuck for
8173 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8174 process, before it forks any delivery processes. The open lookup caching
8175 mechanism meant that the open file or database connection was passed into
8176 the delivery process. The problem was that delivery processes always tidy
8177 up cached lookup data. This could cause a problem for the next delivery
8178 process started by the queue runner, because the external queue runner
8179 process does not know about the closure. So the next delivery process
8180 still has data in the lookup cache. In the case of a file lookup, there was
8181 no problem because closing a file descriptor in a subprocess doesn't affect
8182 the parent. However, if the lookup was caching a connection to a database,
8183 the connection was closed, and the second delivery process was likely to
8184 see errors such as "PGSQL: query failed: server closed the connection
8185 unexpectedly". The problem has been fixed by closing all cached lookups
8186 in a queue runner before running a delivery process.
8188 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8189 seem to have the "const" qualifier which it has on other OS. I've
8192 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8193 *of the same type* (client or server) with the same public name that an
8194 error should be diagnosed.
8196 9. When Exim looked up a host name for an IP address, but failed to find the
8197 original IP address when looking up the host name (a safety check), it
8198 output the message "<ip address> does not match any IP for NULL", which was
8199 confusing, to say the least. The bug was that the host name should have
8200 appeared instead of "NULL".
8202 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8203 user that is built into the binary, and the -C or -D options is used, root
8204 privilege is dropped before the configuration file is read. In addition,
8205 logging is switched to stderr instead of the normal log files. If the
8206 configuration then re-defines the Exim user, the unprivileged environment
8207 is probably not what is expected, so Exim logs a panic warning message (but
8210 However, if deliver_drop_privilege is set, the unprivileged state may well
8211 be exactly what is intended, so the warning has been cut out in that case,
8212 and Exim is allowed to try to write to its normal log files.
8218 1. smtp_return_error_details was not giving details for temporary sender
8219 or receiver verification errors.
8221 2. Diagnose a configuration error if two authenticators have the same public
8224 3. Exim used not to create the message log file for a message until the first
8225 delivery attempt. This could be confusing when incoming messages were held
8226 for policy or load reasons. The message log file is now created at the time
8227 the message is received, and an initial "Received" line is written to it.
8229 4. The automatically generated man page for command line options had a minor
8230 bug that caused no ill effects; however, a more serious problem was that
8231 the procedure for building the man page automatically didn't always
8232 operate. Consequently, release 4.20 contains an out-of-date version. This
8233 shouldn't happen again.
8235 5. When building Exim with embedded Perl support, the script that builds the
8236 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8237 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8239 6. The freeze_tell option was not being used for messages that were frozen on
8240 arrival, either by an ACL or by local_scan().
8242 7. Added the smtp_incomplete_transaction log selector.
8244 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8245 was accepting AUTH without a new EHLO.
8247 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8248 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8251 10. Logging of TCP/IP connections (when configured) now happens in the main
8252 daemon process instead of the child process, so that the TCP/IP connection
8253 count is more accurate (but it can never be perfect).
8255 11. The use of "drop" in a nested ACL was not being handled correctly in the
8256 outer ACL. Now, if condition failure induced by the nested "drop" causes
8257 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8258 or "require"), the connection is dropped.
8260 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8261 that yield "discard" can now be used with an "accept" or a "discard" verb,
8262 but an error is generated for any others (because I can't see a useful way
8263 to define what should happen).
8265 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8266 are now processed in the same way as lines in the Exim configuration file.
8267 In particular, continuation lines are supported.
8269 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8271 15. Added -ti meaning -t -i.
8273 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8274 domains, and warn by logging if others are found.
8276 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8277 returned from GIFCONF when Exim is trying to find the list of interfaces on
8278 a host. The code in os.c has been modified to copy each ifreq to an aligned
8279 structure in all cases.
8281 Also, in some cases, the returned ifreq's were being copied to a 'struct
8282 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8283 means the last couple of bytes of an IPv6 address could be chopped if the
8284 ifreq contained only a normal sockaddr (14 bytes storage).
8286 18. Named domain lists were not supported in the hosts_treat_as_local option.
8287 An entry such as +xxxx was not recognized, and was treated as a literal
8290 19. Ensure that header lines added by a DATA ACL are included in the reject log
8291 if the ACL subsequently rejects the message.
8293 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8294 MD5 (which is deprecated).
8296 21. When testing a filter file using -bf, Exim was writing a message when it
8297 took the sender from a "From " line in the message, but it was not doing so
8298 when it took $return_path from a Return-Path: header line. It now does.
8300 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8301 with a valid header line field name (a series of printing characters
8302 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8304 23. Changed "disc" in the source to "disk" to conform to the documentation and
8305 the book and for uniformity.
8307 24. Ignore Sendmail's -Ooption=value command line item.
8309 25. When execve() failed while trying to run a command in a pipe transport,
8310 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8311 could be confused with a return value of 69 from the command itself. This
8312 has been changed to 127, the value the shell returns if it is asked to run
8313 a non-existent command. The wording for the related log line suggests a
8314 non-existent command as the problem.
8316 26. If received_header_text expands to an empty string, do not add a Received:
8317 header line to the message. (Well, it adds a token one on the spool, but
8318 marks it "old" so that it doesn't get used or transmitted.)
8320 27. Installed eximstats 1.28 (addition of -nt option).
8322 28. There was no check for failure on the call to getsockname() in the daemon
8323 code. This can fail if there is a shortage of resources on the system, with
8324 ENOMEM, for example. A temporary error is now given on failure.
8326 29. Contrary to the C standard, it seems that in some environments, the
8327 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8328 program. Exim now does this explicitly; it affects the formatting of
8329 timestamps using strftime().
8331 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8332 complaints. I've now initialized all the variables, to avoid this.
8334 32. Header lines added by a system filter were not being "seen" during
8335 transport-time rewrites.
8337 33. The info_callback() function passed to OpenSSL is set up with type void
8338 (*)(SSL *, int, int), as described somewhere. However, when calling the
8339 function (actually a macro) that sets it up, the type void(*)() is
8340 expected. I've put in a cast to prevent warnings from picky compilers.
8342 34. If a DNS black list lookup found a CNAME record, but there were no A
8343 records associated with the domain it pointed at, Exim crashed.
8345 35. If a DNS black list lookup returned more than one A record, Exim ignored
8346 all but the first. It now scans all returned addresses if a particular IP
8347 value is being sought. In this situation, the contents of the
8348 $dnslist_value variable are a list of all the addresses, separated by a
8351 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8352 to accept a host name and all its aliases if the forward lookup for any of
8353 them yielded the IP address of the incoming connection. Now it accepts only
8354 those names whose forward lookup yields the correct IP address. Any other
8355 names are discarded. This closes a loophole whereby a rogue DNS
8356 administrator could create reverse DNS records to break through a
8357 wildcarded host restriction in an ACL.
8359 37. If a user filter or a system filter that ran in a subprocess used any of
8360 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8361 the wrong values were passed to the pipe command ($thisaddress had the
8362 value of $0, $0 had the value of $1, etc). This bug was introduced by
8363 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8365 38. Improved the line breaking for long SMTP error messages from ACLs.
8366 Previously, if there was no break point between 40 and 75 characters, Exim
8367 left the rest of the message alone. Two changes have been made: (a) I've
8368 reduced the minimum length to 35 characters; (b) if it can't find a break
8369 point between 35 and 75 characters, it looks ahead and uses the first one
8370 that it finds. This may give the occasional overlong line, but at least the
8371 remaining text gets split now.
8373 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8374 file descriptors might be low, and that setting 1000 would always raise it.
8375 It turns out that in some environments, the limit is already over 1000 and
8376 that lowering it causes trouble. So now Exim takes care not to decrease it.
8378 40. When delivering a message, the value of $return_path is set to $sender_
8379 address at the start of routing (routers may change the value). By an
8380 oversight, this default was not being set up when an address was tested by
8381 -bt or -bv, which affected the outcome if any router or filter referred to
8384 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8385 log only when "message" or "log_message" are set. However, if one of the
8386 conditions was an address verification, or a call to a nested ACL, the
8387 messages generated by the underlying test were being passed through. This
8388 no longer happens. The underlying message is available in $acl_verify_
8389 message for both "message" and "log_message" expansions, so it can be
8390 passed through if needed.
8392 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8393 new expansion $bh_ to give the encoded byte string without charset
8394 translation. Translation happens only if iconv() is available; HAVE_ICONV
8395 indicates this at build time. HEADERS_CHARSET gives the charset to
8396 translate to; headers_charset can change it in the configuration, and
8397 "headers charset" can change it in an individual filter file.
8399 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8400 that creates RFC 2047 encoded "words" labels them as that charset instead
8401 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8402 expansion operator; (ii) when Exim creates a From: line for a local
8403 message; (iii) when a header line is rewritten to include a "phrase" part.
8405 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8406 buggy, causing it to skip the first lines of messages whose message ID
8407 ended in 'D'. This would not have bitten before Exim release 4.14, because
8408 message IDs were unlikely to end in 'D' before then. The effect was to have
8409 incorrect size information for certain domains.
8411 45. #include "config.h" was missing at the start of the crypt16.c module. This
8412 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8415 46. If there was a timeout during a "random" callout check, Exim treated it as
8416 a failure of the random address, and carried on sending RSET and the real
8417 address. If the delay was just some slowness somewhere, the response to the
8418 original RCPT would be taken as a response to RSET and so on, causing
8419 mayhem of various kinds.
8421 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8422 when I implemented it. It didn't allow for the fact that some option values
8423 may legitimately be negative (e.g. size_addition), and it didn't even do
8424 the right test for positive values.
8426 48. Domain names in DNS records are case-independent. Exim always looks them up
8427 in lower case. Some resolvers return domain names in exactly the case they
8428 appear in the zone file, that is, they may contain uppercase letters. Not
8429 all resolvers do this - some return always lower case. Exim was treating a
8430 change of case by a resolver as a change of domain, similar to a widening
8431 of a domain abbreviation. This triggered its re-routing code and so it was
8432 trying to route what was effectively the same domain again. This normally
8433 caused routing to fail (because the router wouldn't handle the domain
8434 twice). Now Exim checks for this case specially, and just changes the
8435 casing of the domain that it ultimately uses when it transmits the message
8438 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8441 50. If a filter generated a file delivery with a non-absolute name (possible if
8442 no home directory exists for the router), the forbid_file option was not
8445 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8446 the existing equality matching.
8448 52. Exim was using ints instead of ino_t variables in some places where it was
8449 dealing with inode numbers.
8451 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8452 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8453 called TMPDIR, and if it finds it is different, it changes its value.
8455 54. The smtp_printf() function is now made available to local_scan() so
8456 additional output lines can be written before returning. There is also an
8457 smtp_fflush() function to enable the detection of a dropped connection.
8458 The variables smtp_input and smtp_batched_input are exported to
8461 55. Changed the default runtime configuration: the message "Unknown user"
8462 has been removed from the ACL, and instead placed on the localuser router,
8463 using the cannot_route_message feature. This means that any verification
8464 failures that generate their own messages won't get overridden. Similarly,
8465 the "Unrouteable address" message that was in the ACL for unverifiable
8466 relay addresses has also been removed.
8468 56. Added hosts_avoid_esmtp to the smtp transport.
8470 57. The exicyclog script was not checking for the esoteric option
8471 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8472 will work only if exicyclog is run under the appropriate euid.
8474 58. Following a discussion on the list, the rules by which Exim recognises line
8475 endings on incoming messages have been changed. The -dropcr and drop_cr
8476 options are now no-ops, retained only for backwards compatibility. The
8477 following line terminators are recognized: LF CRLF CR. However, special
8478 processing applies to CR:
8480 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8481 nor a local message in the state where . is a terminator.
8483 (ii) If a bare CR is encountered in a header line, an extra space is added
8484 after the line terminator so as not to end the header. The reasoning
8485 behind this is that bare CRs in header lines are most likely either
8486 to be mistakes, or people trying to play silly games.
8488 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8489 was being incorrectly given as 18 bytes larger than it should have been.
8490 This is a VOB (very old bug).
8492 60. This may never have affected anything current, but just in case it has:
8493 When the local host is found other than at the start of a list of hosts,
8494 the local host, those with the same MX, and any that follow, are discarded.
8495 When the list in question was part of a longer list of hosts, the following
8496 hosts (not currently being processed) were also being discarded. This no
8497 longer happens. I'm not sure if this situation could ever has previously
8500 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8503 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8504 header. This is something that is recommended in a new Internet Draft, and
8505 is something that is documented as being done by Sendmail. There are two
8506 possible values. For messages generated by the autoreply transport, Exim
8509 Auto-Submitted: auto-replied
8511 whereas for all other generated messages (e.g. bounces) it adds
8513 Auto-Submitted: auto-generated
8515 63. The "personal" condition in filters now includes a test for the
8516 Auto-Submitted: header. If it contains the string "auto-" the message it
8517 not considered personal.
8519 64. Added rcpt_include_affixes as a generic transport option.
8521 65. Added queue_only_override (default true).
8523 66. Added the syslog_duplication option.
8525 67. If what should have been the first header line of a message consisted of
8526 a space followed by a colon, Exim was mis-interpreting it as a header line.
8527 It isn't of course - it is syntactically invalid and should therefore be
8528 treated as the start of the message body. The misbehaviour could have
8529 caused a number of strange effects, including loss of data in subsequent
8530 header lines, and spool format errors.
8532 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8533 client host had authenticated. This control can now be exercised by an ACL
8534 for more flexibility.
8536 69. By default, callouts do not happen when testing with -bh. There is now a
8537 variant, -bhc, which does actually run the callout code, including
8538 consulting and updating the callout cache.
8540 70. Added support for saslauthd authentication, courtesy of Alexander
8543 71. If statvfs() failed on the spool or log directories while checking their
8544 size for availability, Exim confusingly gave the error "space shortage".
8545 Furthermore, in debugging mode it crashed with a floating point exception.
8546 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8547 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8548 serious problem, Exim now writes to the main and panic logs when this
8549 happens, with details of the failure. It then refuses to accept the
8550 incoming message, giving the message "spool directory problem" or "log
8551 directory problem" with a 421 code for SMTP messages.
8553 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8554 0, 1, and 2 exist, because some OS complain for execs without them (see
8555 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8556 descriptors. However, the code omitted to check that the open succeeded,
8557 causing mysterious errors if for some reason the permissions on /dev/null
8558 got screwed. Now Exim writes a message to the main and panic logs, and
8559 bombs out if it can't open /dev/null.
8561 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8562 interact so that it is all more flexible. It is supposed to remain
8563 backwards compatible. Also added extra_local_interfaces.
8565 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8566 to bomb out with an assertion failure - to the client this appears as a
8567 connection drop. This problem occurs in the part of the code that was taken
8568 from the Samba project. Fortunately, the assertion is in a very simple
8569 function, so I have fixed this by reproducing the function inline in the
8570 one place where it is called, and arranging for authentication to fail
8571 instead of killing the process with assert().
8573 75. The SPA client code was not working when the server requested OEM rather
8574 than Unicode encoding.
8576 76. Added code to make require_files with a specific uid setting more usable in
8577 the case where statting the file as root fails - usually a non-root-mounted
8578 NFS file system. When this happens and the failure is EACCES, Exim now
8579 forks a subprocess and does the per-uid checking as the relevant uid.
8581 77. Added process_log_path.
8583 78. If log_file_path was not explicitly set, a setting of check_log_space or
8584 check_log_inodes was ignored.
8586 79. If a space check for the spool or log partitions fails, the incident is now
8587 logged. Of course, in the latter case the data may get lost...
8589 80. Added the %p formatting code to string_format() so that it can be used to
8590 print addresses in debug_print(). Adjusted all the address printing in the
8591 debugging in store.c to use %p rather than %d.
8593 81. There was a concern that a line of code in smtp_in.c could overflow a
8594 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8595 initially expressed, the concern was not well-founded, because trailing
8596 spaces are removed early. However, if the trailing spaces were followed by
8597 a NULL, they did not get removed, so the overflow was possible. Two fixes
8600 (a) I re-wrote the offending code in a cleaner fashion.
8601 (b) If an incoming SMTP command contains a NULL character, it is rejected
8604 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8605 runs initgroups(), so that if the Exim user is in any additional groups,
8606 they will be used during message reception.
8612 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob