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 trtying 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
86 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
87 SMTP connection" log lines.
89 JH/02 Option default value updates:
90 - queue_fast_ramp (main) true (was false)
91 - remote_max_parallel (main) 4 (was 2)
93 JH/03 Cache static regex pattern compilations, for use by ACLs.
95 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
96 Make the rewrite never match and keep the logging. Trust the
97 admin to be using verify=header-syntax (to actually reject the message).
99 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
100 (under Linux) we watch the dir containing the final file; previously
101 it would be the dir with the first symlink. We still do not monitor
104 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
105 dn_expand() is happy to pass them through.
107 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
108 the file with the proof had an unchanged name, the new proof(s) were
109 loaded on top of the old ones (and nover used; the old ones were stapled).
111 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
112 more than one message arrived in a single connection a reference from
113 the earlier message could be re-used. Often a sigsegv resulted.
114 These variables were introduced in Exim 4.87.
115 Debug help from Graeme Fowler.
117 JH/09 Fix ${filter } for conditions that modify $value. Previously the
118 modified version would be used in construction the result, and a memory
121 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
122 Find and fix by Jasen Betts.
124 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
125 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
126 the systemwide configuration to override the Exim config.
128 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
129 API changes in libopendmarc.
131 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
132 pid 1, in the normal "background daemon" mode, having to drop process-
133 group leadership also lost track of needing to create listener sockets.
135 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
136 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
138 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
139 a capture group which obtained no text (eg. "(abc)*" matching zero
140 occurrences) could cause a segfault if the corresponding $<n> was
143 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
144 included a close-brace character (eg. it itself used an expansion) an
147 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
148 starting TLS. Previously it was after, meaning that attackers on such
149 ports had to be screened using the host_reject_connection main config
150 option. The new sequence aligns better with the STARTTLS behaviour, and
151 permits defences against crypto-processing load attacks, even though it
152 is strictly an incompatible change.
153 Also, avoid sending any SMTP fail response for either the connect ACL
154 or host_reject_connection, for TLS-on-connect ports.
156 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
157 Previously this was not permitted, but it makes reasonable sense.
158 While there, restore a restriction on using it from a connect ACL; given
159 the change JH/16 it could only return false (and before 4.91 was not
162 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
163 was exactly sized compared to the log buffer, a crash occurred with the
164 misleading message "bad memory reference; pool not found".
165 Found and traced by Jasen Betts.
167 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
168 dns_again_means_nonexist included an element causing a DNS lookup which
169 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
170 included (though probably not limited to) a process crash from stack
171 memory limit, or from excessive open files. Replace this with a paniclog
172 whine (as this is likely a configuration error), and returning
175 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
176 this always failed, probably leading to the usual downgrade to in-clear
179 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
180 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
181 to not checking that list for these lookups.
183 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
186 JH/23 Fix crash in string expansions. Previously, if an empty variable was
187 immediately followed by an expansion operator, a null-indirection read
188 was done, killing the process.
190 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
191 include an SMTP response string which is longer than that supported
192 by the delivering transport. Alleviate by wrapping such lines before
195 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
196 chars (RFC limit). Previously a limit of 12 items was made, which with
197 a not-impossible References: in the message being bounced could still
198 be over-large and get stopped in the transport.
200 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
201 close. Previously a bare socket close was done.
203 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
206 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
207 message being considered for delivery was already being handled by
208 another process, and having an SMTP connection already open, the function
209 to close it tried to use an uninitialized variable. This would afftect
210 high-volume sites more, especially when running mailing-list-style loads.
211 Pollution of logs was the major effect, as the other process delivered
212 the message. Found and partly investigated by Graeme Fowler.
214 JH/29 Change format of the internal ID used for message identification. The old
215 version only supported 31 bits for a PID element; the new 64 (on systems
216 which can use Base-62 encoding, which is all currently supported ones
217 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
218 and must use Base-36). The new ID is 23 characters rather than 16, and is
219 visible in various places - notably logs, message headers, and spool file
220 names. Various of the ancillary utilities also have to know the format.
221 As well as the expanded PID portion, the sub-second part of the time
222 recorded in the ID is expanded to support finer precision. Theoretically
223 this permits a receive rate from a single comms channel of better than the
225 The major timestamp part of the ID is not changed; at 6 characters it is
226 usable until about year 3700.
227 Updating from previously releases is fully supported: old-format spool
228 files are still usable, and the utilities support both formats. New
229 message will use the new format. The one hints-DB file type which uses
230 message-IDs (the transport wait- DB) will be discarded if an old-format ID
231 is seen; new ones will be built with only new-format IDs.
232 Optionally, a utility can be used to convert spool files from old to new,
233 but this is only an efficiency measure not a requirement for operation
234 Downgrading from new to old requires running a provided utility, having
235 first stopped all operations. This will convert any spool files from new
236 back to old (losing time-precision and PID information) and remove any
237 wait- hints databases.
239 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
240 we treated them as item separators when parsing for a list item, but they
241 need to be protected by the doublequotes. While there, add handling for
244 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
245 Found and fixed by Jasen Betts. No testcase for this as my usual text
246 editor insists on emitting only valid UTF-8.
248 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
249 a null-indirection SIGSEGV for the receive process.
251 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
252 -bh use. Internal checking would spot this and take a panic.
254 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
255 In 4.96 this would expand to empty.
257 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
258 certificate. Find and fix by Andreas Metzler.
260 JH/36 Add ARC info to DMARC hostory records.
262 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
263 or fakedefer. Previously the sender could discover that the message
264 had in fact been accepted.
266 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
267 sequences. Previously the input was not noted as being tainted; notably
268 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
269 bad coding of authenticators.
271 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
272 and ${tr...}. Found and diagnosed by Heiko Schlichting.
274 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
275 could be triggered by externally-supplied input. Found by Trend Micro.
278 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
279 be triggered by externally-controlled input. Found by Trend Micro.
282 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
283 be triggered by externally-controlled input. Found by Trend Micro.
286 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
287 Make the rewrite never match and keep the logging. Trust the
288 admin to be using verify=header-syntax (to actually reject the message).
290 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
293 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
299 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
300 after reception to before a subsequent reception. This should
301 mean slightly faster delivery, and also confirmation of reception
304 JH/02 Move from using the pcre library to pcre2. The former is no longer
305 being developed or supported (by the original developer).
307 JH/03 Constification work in the filters module required a major version
308 bump for the local-scan API. Specifically, the "headers_charset"
309 global which is visible via the API is now const and may therefore
310 not be modified by local-scan code.
312 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
313 sendfile() didi not account for the way the ClamAV driver code called it.
315 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
316 time check was being done for every character; replace that with one
319 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
320 sent was prefixed with a length byte.
322 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
323 RFC 5321. Previously Dovecot (at least) would log errors during
326 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
327 in 4.61 without a reason logged. Binaries will be bigger, which might
328 matter on diskspace-constrained systems, but debug is easier.
330 JH/09 Fix macro-definition during "-be" expansion testing. The move to
331 write-protected store for macros had not accounted for these runtime
332 additions; fix by removing this protection for "-be" mode.
334 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
335 handing out large-numbered file descriptors, violating the usual Unix
336 assumption (and required by Posix) that the lowest possible number will be
337 allocated by the kernel when a new one is needed. In the daemon, and any
338 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
339 are not useable for FD_SET() [and hence select()] and overwrite the stack.
340 Assorted crashes happen.
342 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
343 main-section options or in a connect ACL, the value from the first ever
344 connection was never replaced for subsequent connections. Found by
347 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
348 though only once PCRE2 was introduced: the memory accounting used under
349 debug offset allocations by an int, giving a hard trap in early startup.
350 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
352 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
353 with underbars is given. The write-protection of configuration introduced
354 in 4.95 trapped when normalisation was applied to an option not needing
357 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
359 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
360 failing to close the certificates directory, every hour or any time it
363 JH/16 Debugging initiated by an ACL control now continues through into routing
364 and transport processes. Previously debugging stopped any time Exim
365 re-execs, or for processing a queued message.
367 JH/17 The "expand" debug selector now gives more detail, specifically on the
368 result of expansion operators and items.
370 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
371 bad comparison between the option value and the name of the file to
372 be included was done, and a mismatch was wrongly identified.
373 4.88 to 4.95 are affected.
375 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
377 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
378 supplied as the db file-pair basename. Previously, if a directory
379 path was given, for example via the autoreply "once" option, the DB
380 file.pag and file.dir files would be created in that directory's
383 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
384 "taint" log_selector. These were previously deprecated.
386 JH/22 Fix static address-list lookups to properly return the matched item.
387 Previously only the domain part was returned.
389 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
390 the call into OpenSSL to send a TLS Close was being repeated; this
391 resulted in the library waiting for the peer's Close. If that was never
392 sent we waited forever. Fix by tracking send calls.
394 JH/24 The ${run} expansion item now expands its command string elements after
395 splitting. Previously it was before; the new ordering makes handling
396 zero-length arguments simpler. The old ordering can be obtained by
397 appending a new option "preexpand", after a comma, to the "run".
399 JH/25 Taint-check exec arguments for transport-initiated external processes.
400 Previously, tainted values could be used. This affects "pipe", "lmtp" and
401 "queryprogram" transport, transport-filter, and ETRN commands.
402 The ${run} expansion is also affected: in "preexpand" mode no part of
403 the command line may be tainted, in default mode the executable name
406 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
407 the the facility was not passed across execs, and only the first message
408 passed over a connection could use BDAT; any further ones using DATA.
410 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
411 uses $sending_ip_address and an interface is specified.
412 Previously any use of the local address in the EHLO name disabled
413 PIPECONNECT, the common case being to use the rDNS of it.
415 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
416 resumption. Previously verify failed because no certificate status is
417 passed on the wire for the restarted session. Fix by using the recorded
418 ocsp status of the stored session for the new connection.
420 JH/29 TLS resumption: the key for session lookup in the client now includes
421 more info that a server could potentially use in configuring a TLS
422 session, avoiding oferring mismatching sessions to such a server.
423 Previously only the server IP was used.
425 JH/30 Fix string_copyn() for limit greater than actual string length.
426 Previously the copied amount was the limit, which could result in a
427 overlapping memcpy for newly allocated destination soon after a
428 source string shorter than the limit. Found/investigated by KM.
430 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
431 close; it may be needed for a subsequent connection. This caused a
432 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
434 JH/32 Fix CHUNKING for a second message on a connection when the first was
435 rejected. Previously we did not reset the chunking-offered state, and
436 erroneously rejected the BDAT command. Investigation help from
439 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
440 an empty address. Previously the expansion returned an error.
442 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
443 proxy. Previously these were misparsed, leading to paniclog entries.
449 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
450 related applications. Previously an "H" was used where available info
451 says that "M" should be, so change to match.
453 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
454 as arguments, so an implementation trying to copy these into a local
455 buffer was taking a taint-enforcement trap. Fix by using dynamically
456 created buffers. Similar fix for radius expansion condition.
458 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
459 reasonable, eg. to count headers. Fix by using dynamically created
460 buffers rather than a local. Do similar fixes for ACL actions "dcc",
461 "log_reject_target", "malware" and "spam"; the arguments are expanded
462 so could be handling tainted values.
464 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
465 broken the (no-op) support for this sendmail command. Restore it
466 to doing nothing, silently, and returning good status.
468 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
469 record path was given (or the default used) without a leading directory
470 path, an error occurred on trying to open it. Use the transport's working
473 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
474 transport. Previously it was the name on the DNS A-record; use instead
475 the head of the CNAME chain leading there (if there is one). This seems
476 to align better with RFC 6125.
478 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
479 smtp_accept_max_per_host allocated resources which were not released
480 when the limit was exceeded. This eventually crashed the daemon. Fix
481 by adding a release action in that path.
483 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
484 expanded; previously using tainted values was rejected. Fix by using
485 dynamically-created buffers.
487 JH/09 Relax restrictions on ACL verify condition needing access to message
488 headers. Previously they were only permitted in data and non-smtp ACLs;
489 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
490 not_blind, header_sender and header_names_ascii verification.
492 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
493 Previously a macro used one argument twice; when called with the
494 argument as an expression having side-effects, incorrect operation
495 resulted. Use an inlineable function.
497 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
498 held open for a verify callout. Previously this wan not accounted for
499 and a corrupt onward SMTP conversation resulted.
501 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
502 Previously they were embedded naively in the constructed address; when
503 needed, strip the quoting and quote the entire local_part.
504 Also make the inbound_srs expansion condition handle quoting.
506 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
507 excluded, not matching the documentation.
509 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
510 was given for the sqlite_dbfile a trap resulted.
512 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
513 "name" argument resulted in a trap. There is no reason to disallow such;
514 this was a coding error.
516 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
517 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
518 spent suspended, ignoring the POSIX definition. Previously we assumed
519 it did and a constant offset from real time could be used as a correction.
520 Change to using the same clock source for the start-of-message and the
521 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
522 exists, just to get a clock slightly more aligned to reality.
524 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
525 RFC says it is optional some validators care. The missing char was not
526 intended but triggered by a line-wrap alignment. Discovery and fix by
527 Guillaume Outters, hacked on by JH.
529 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
530 name being quoted was tainted a trap would be taken. Fix by using
531 dynamically created buffers. The routine could have been called by a
532 rewrite with the "h" flag, by using the "-F" command-line option, or
533 by using a "name=" option on a control=submission ACL modifier.
535 JH/19 SPF: change the Authentication-Results expansion component to give
536 smtp.helo when the sender domain is empty. Previously it gave
539 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
540 not in 127.0.0.0/8 to help in spotting list domains taken over by a
541 domain-parking registrar.
543 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
544 Previously when a whitespace character was specified it was not inserted
545 after removing the newline.
547 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
548 the domain part of the recipient address. This overrides any tls_sni
549 option set, which was previously used.
551 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
554 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
555 is_tainted() had an off-by-one error in the overenthusiastic direction.
556 Find and fix by Gavan. Although NetBSD is not a supported platform for
557 4.94 this bug could affect other platforms.
559 PP/01 Fix default prime selection to be consistent.
560 One path used ike23 still, instead of exim.dev.20160529.3; now both
561 execution flows will use the same DH primes (currently
562 exim.dev.20160529.3).
564 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
565 option in smtp transport, to match the documentation. Previously
566 verification was not being done in some cases where it should have been.
568 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
569 than one server was defined and depending on the platform memory layout
570 details, an internal consistency trap could be hit while walking the list
573 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
574 files. The value is used by the authresults expansion item. Previously
575 if this was used in a router or transport, a crash could result.
577 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
578 have a sufficient incidence of EINTR returns from select that an
579 interaction with other operations done by the main daemon loop exposed
580 a bug in the error-handling. This was benign apart from the log
583 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
584 for consistency with delivered "=>" and failed "**" lines. While we're
585 there, handle PRX and TFO.
587 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
588 applied. This resulted, if any header-line rewrite rules were configured,
589 in a panic-log triggerable by sending a message with a long address in
590 a header. Fix by increasing the arbitrary limit to larger than a single
591 (dewrapped) 5322 header line maximum size.
593 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
594 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
595 RFC 6648 which deprecates X- options in protocols as a general practice.
596 Changeover between the implementations is handled by the mechanisms
599 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
600 a remote which uses callout/hold. Previously the local was queued.
602 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
605 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
606 log line. The asterisk is now consistently appended to the remote IP
607 (and port, if given), and will also be provided on defer and fail log
608 lines. Previously it could be placed on the local IP if that was being
609 logged, and was only provided on delivery lines.
611 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
613 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
614 authenticator client_send option. Previously the next char, after a pair
615 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
616 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
617 documentation. There is still no way to get a leading ^ immediately
618 after a NUL (ie. for the password of a PLAIN method authenticator.
620 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
621 files. For bad sizes read, delete the record and whine to paniclog.
623 JH/38 When logging an AUTH failure, as server, do not include sensitive
624 information. Previously, the credentials would be included if given
625 as part of the AUTH command line and an ACL denied authentication.
627 JH/39 Bug 2691: fix $local_part_data. When the matching list element
628 referred to a file, bad data was returned. This likely also affected
631 JH/40 The gsasl authenticator now supports caching of the salted password
632 generated by the client-side implementation. This required the addition
633 of a new variable: $auth4.
635 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
636 left undeleted; the attempt to re-create it then failed - resulting in
637 the usual "SIGHUP tp have daemon reload configuration" to not work.
638 This affected any platform not supporting "abstract" Unix-domain
639 sockets (i.e. not Linux).
641 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
642 recipients" response to RCPT in a later response, with a 250. The
643 previous coding assumed this would not happen, and under PIPELINING
644 would result in both lost and duplicate recipients for a message.
646 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
647 Previously the weighting was incorrectly applied. Similar fix for socks
648 proxies. Found and fixed by Heiko Schlichting.
650 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
651 not handle sub-lists included using the +namedlist syntax. While
652 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
653 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
656 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
657 Previously the permanent pool was used, so the sore could not be freed.
658 This meant a connection with many messages would use continually-growing
661 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
662 per-pool so as not to waste too much space. Previously a constant size
663 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
664 DOS attacks harder. The cost is wasted memory use in the larger blocks.
666 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
667 same space repeatedly, and a smaller process footprint.
669 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
670 for GnuTLS. Previously the name started "%s" which, while not a bug,
671 looked as if if might be one.
673 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
674 "continued connection") make the $proxy_* variables available. Previously
675 the information was not passed across the exec() call for subsequent
676 transport executions. This also mean that the log lines for the
677 messages can show the proxy information.
679 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
680 receive time. With modern systems the difference is significant.
681 The historical behaviour can be restored by disabling (a new) log_selector
682 "queue_time_exclusive".
684 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
685 content) were specifically tested for. Now, also cover expansions
686 resulting in ACL names and inline ACL content.
688 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
689 making it unusable in complex expressions.
691 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
692 fail with a taint-check on the spoolfile name, and leave the message
695 HS/01 Enforce absolute PID file path name.
697 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
699 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
701 PP/02 Bug 2643: Correct TLS DH constants.
702 A missing NUL termination in our code-generation tool had led to some
703 incorrect Diffie-Hellman constants in the Exim source.
704 Reported by kylon94, code-gen tool fix by Simon Arlott.
706 PP/03 Impose security length checks on various command-line options.
707 Fixes CVE-2020-SPRSS reported by Qualys.
709 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
710 better. Reported by Qualys.
712 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
713 providing a particularly obnoxious sender full name.
716 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
718 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
721 PP/08 Change default for recipients_max from unlimited to 50,000.
723 PP/09 Fix security issue with too many recipients on a message (to remove a
724 known security problem if someone does set recipients_max to unlimited,
725 or if local additions add to the recipient list).
726 Fixes CVE-2020-RCPTL reported by Qualys.
728 PP/10 Fix security issue in SMTP verb option parsing
729 Fixes CVE-2020-EXOPT reported by Qualys.
731 PP/11 Fix security issue in BDAT state confusion.
732 Ensure we reset known-good where we know we need to not be reading BDAT
733 data, as a general case fix, and move the places where we switch to BDAT
734 mode until after various protocol state checks.
735 Fixes CVE-2020-BDATA reported by Qualys.
737 HS/03 Die on "/../" in msglog file names
739 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
740 the Exim runtime user are allowed to create files.
742 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
745 QS/03 When reading the output from interpreted forward files we do not
746 pass the pipe between the parent and the interpreting process to
747 executed child processes (if any).
749 QS/04 Always die if requested from internal logging, even is logging is
752 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
753 the API; compilation noo longer completes with DMARC support included.
754 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
755 been reported on other platforms.
757 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
759 JH/56 Make the majority of info read from config files readonly, for defence-in-
760 depth against exploits. Suggestion by Qualys.
761 Not supported on Solaris 10.
763 JH/57 Fix control=fakreject for a custom message containing tainted data.
764 Previously this resulted in a log complaint, due to a re-expansion present
765 since fakereject was originally introduced.
767 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
768 modifier was given, a loop resulted.
770 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
771 pipelined SMTP command followed the BDAT LAST then it would be
772 incorrectly treated as part of the message body, causing a verification
775 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
776 require looser validation rules than those for 821-level addresses,
777 which only permit IP addresses.
783 JH/01 Avoid costly startup code when not strictly needed. This reduces time
784 for some exim process initialisations. It does mean that the logging
785 of TLS configuration problems is only done for the daemon startup.
787 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
789 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
790 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
793 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
794 DKIM signing is being done. Previously a transport_filter would always
795 disable CHUNKING, falling back to traditional DATA.
797 JH/05 Regard command-line recipients as tainted.
799 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
801 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
802 PAM library frees one of the arguments given to it, despite the
803 documentation. Therefore a plain malloc must be used.
805 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
806 on-stack buffers were used, resulting in a taint trap when DSN information
807 copied from a received message was written into the buffer.
809 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
810 the ordering of its ARC headers. This caused a crash.
812 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
813 a new record was being constructed with information from the peer, a trap
816 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
817 installation would get error messages from DMARC verify, when it hit the
818 nonexistent file indicated by the default. Distros wanting DMARC enabled
819 should both provide the file and set the option.
820 Also enforce no DMARC verification for command-line sourced messages.
822 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
823 could, depending on the platform, hang at the STARTTLS response.
825 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
826 message on a connection. Previously if one message had ARC headers and
827 the following one did not, a crash could result when adding an
828 Authentication-Results: header.
830 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
831 Exim main code and Exim-related utities. The introduction of taint
832 tracking also did many adjustments to string handling. Since then, eximon
833 frequently terminated with an assert failure.
835 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
836 check for 452 responses. This slightly helps the inefficieny of doing
837 a large alias-expansion into a recipient-limited target. The max_rcpt
838 transport option still applies (and at the current default, will override
839 the new feature). The check is done for either cause of synch, and forces
840 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
841 connection. The new facility is not tunable at this time.
843 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
844 library live data was being used, so the results became garbage. Make
845 copies while it is still usable.
847 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
848 on delivery deferred (==) and failed (**) lines (if a delivery was
849 attemtped). Previously it was only on completion (=>) lines.
851 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
852 for the expansion of the server_scram_iter and server_scram_salt options.
854 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
855 are now specifically given a NO_DATA response without hitting the system
856 resolver. The library goes on to do the now-standard TXT lookup.
857 Use of dnsdb lookups is not affected.
859 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
860 only retrieve the errormessage once. Previously two calls to dlerror()
861 were used, and the second one (for mainlog/paniclog) retrieved null
864 JH/20 Taint checking: disallow use of tainted data for
865 - the appendfile transport file and directory options
866 - the pipe transport command
867 - the autoreply transport file, log and once options
868 - file names used by the redirect router (including filter files)
870 - paths used by single-key lookups
871 Previously this was permitted.
873 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
874 adjusted the size of a major service buffer; this failed because the
875 buffer was in use at the time. Change to a compile-time increase in the
876 buffer size, when this authenticator is compiled into exim.
878 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
879 previous fast-mode was untenable in the face of glibs using mmap to
880 support larger malloc requests.
882 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
883 New values supported, if defined on system where compiled:
884 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
885 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
887 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
888 running a limited number of proceses in parallel, a benefit is gained. The
889 amount varies with the platform hardware and load. The use of the option
890 queue_run_in_order means we cannot do this, as ordering becomes
893 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
894 had introduced a string-copy (for ensuring NUL-termination) which was not
895 appropriate for that case, which can include embedded NUL bytes in the
896 block of data. Investigation showed the copy to actually be needless, the
897 data being length-specified.
899 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
900 done during a receiving connection, and both used TLS, global info was
901 used rather than per-connection info for tracking the state of data
902 queued for transmission. This could result in a connection hang.
904 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
905 Previously, when delivering serveral messages down a single connection
906 only the first would provide a SIZE. This was due to the size information
907 not being properly tracked.
909 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
910 TAI (at 37 seconds currently), pretend to be in UTC for time-related
911 expansion and logging. Previously, spurious values such as a future
912 minute could be seen.
914 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
915 it could crash from a null-deref. This could also affect the
916 ${addresses: } operator and ${readsock } item.
918 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
919 message following a mime one, the variable was not reset.
921 JH/30 When an pipelined-connect fails at the first response, assume incorrect
922 cached capability (perhaps the peer reneged?) and immediately retry in
925 JH/31 Fix spurious detection of timeout while writing to transport filter.
927 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
928 an attempt to copy the string was made before checking it.
930 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
931 taint of the lookup key was maintained; we now regard the presence in the
932 filesystem as sufficient validation.
934 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
935 argument is supplied.
937 JH/35 The dsearch lookup now requires that the directory is an absolute path.
938 Previously this was not checked, and nonempty relative paths made an
939 access under Exim's current working directory.
941 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
942 Previously no event was raised.
944 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
945 parameter supplied by the sender MAIL FROM command. Previously it was
946 ignored, and only the check_spool_space option value for the required
949 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
950 the size of the signing public-key. Previously it was instead giving
951 the size of the signature hash.
953 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
954 the default. See the (new) dkim_verify_min_keysizes option.
956 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
957 and an ACL use a lookup for checking either the local_part or domain,
958 stale data could be accessed. Ensure that variable references are
959 dropped between messages.
961 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
962 by the client was not checked as pointing within response data before
963 being used. A malicious client could thus cause an out-of-bounds read and
964 possibly gain authentication. Fix by adding the check.
966 JH/42 Internationalisation: change the default for downconversion in the smtp
967 transport to be "if needed". Previously it was "as previously set" for
968 the message, which usually meant "if needed" for message-submission but
969 "no" for everything else. However, MTAs have been seen using SMTPUTF8
970 even when the envelope addresses did not need it, resulting in forwarding
971 failures to non-supporting MTAs. A downconvert in such cases will be
972 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
973 transport. The change does mean that addresses needing conversion will
974 be converted when previously a delivery failure would occur.
976 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
977 response was received it would be used unchecked in a fail-DSN, violating
978 standards on line-length limits. Truncate if needed.
980 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
981 doesn't work. (Additionally add a new main config option to configure the
988 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
989 side, to decode a TLS 1.3 packet capture.
991 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
992 Previously the default library behaviour applied, sending two, each in
995 JH/03 Debug output for ACL now gives the config file name and line number for
998 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1000 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1002 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1003 buffer overrun for (non-chunking) other transports.
1005 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1006 TLS1.3, means that a server rejecting a client certificate is not visible
1007 to the client until the first read of encrypted data (typically the
1008 response to EHLO). Add detection for that case and treat it as a failed
1009 TLS connection attempt, so that the normal retry-in-clear can work (if
1010 suitably configured).
1012 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1013 and/or domain. Found and fixed by Jason Betts.
1015 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1016 configuration). If a CNAME target was not a wellformed name pattern, a
1019 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1020 the OS reports them interleaved with other addresses.
1022 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1023 used both for input and for a verify callout, both encrypted, SMTP
1024 responses being sent by the server could be lost. This resulted in
1025 dropped connections and sometimes bounces generated by a peer sending
1028 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1029 string. Previously it was possible to cause undefined behaviour in a
1030 library routine (usually a crash). Found by "zerons".
1032 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1035 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1036 API was removed, so update to use the newer ones.
1038 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1039 any timeout set, is taking a long time. Previously we would hang on to a
1040 rotated logfile "forever" if the input was arriving with long gaps
1041 (a previous attempt to fix addressed lack, for a long time, of initial
1044 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1045 shared (NFS) environment. The length of the tempfile name is now
1046 4 + 16 ("hdr.$message_exim_id") which might break on file
1047 systems which restrict the file name length to lower values.
1048 (It was "hdr.$pid".)
1050 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1051 shared (NFS) environment.
1053 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1054 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1057 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1058 on some platforms for bit 31.
1060 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1061 to changes apparently associated with TLS1.3 handling some of the APIs
1062 previously used were either nonfunctional or inappropriate. Strings
1063 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1064 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1065 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1066 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1067 and the use of specific cipher names in the encrypted= ACL condition.
1069 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1071 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1072 verification result was not updated unless hosts_require_ocsp applied.
1074 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1075 queue_list_requires_admin set to false, non-admin users were denied the
1078 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1079 directory-of-certs mode. Previously they were advertised despite the
1082 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1083 A single TCP connection by a client will now hold a TLS connection open
1084 for multiple message deliveries, by default. Previously the default was to
1087 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1088 default. If built with the facility, DANE will be used. The facility
1089 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1091 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1092 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1093 must be defined and you must still, unless you define DISABLE_TLS, manage
1094 the the include-dir and library-file requirements that go with that
1095 choice. Non-TLS builds are still supported.
1097 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1100 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1101 default. If the platform supports and has the facility enabled, it will
1102 be requested on all coneections.
1104 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1105 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1107 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1109 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1110 success-DSN messages. Previously the From: header was always the default
1111 one for these; the option was ignored.
1113 JH/28 Fix the timeout on smtp response to apply to the whole response.
1114 Previously it was reset for every read, so a teergrubing peer sending
1115 single bytes within the time limit could extend the connection for a
1116 long time. Credit to Qualsys Security Advisory Team for the discovery.
1118 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1119 delivery address, which leaked information of the results of local
1120 forwarding. Change to the original envelope recipient address, per
1123 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1124 requested. Previously not bounce was generated and a log entry of
1125 error ignored was made.
1127 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1129 JH/32 Introduce a general tainting mechanism for values read from the input
1130 channel, and values derived from them. Refuse to expand any tainted
1131 values, to catch one form of exploit.
1133 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1134 was unused and the unexpanded text used for the test. Found and
1135 fixed by Ruben Jenster.
1137 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1138 an attempt to use a TLS library read routine dereffed a nul pointer,
1141 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1142 them in our smtp response.
1144 JH/36 Have the general router option retry_use_local_part default to true when
1145 any of the restrictive preconditions are set (to anything). Previously it
1146 was only for check_local user. The change removes one item of manual
1147 configuration which is required for proper retries when a remote router
1148 handles a subset of addresses for a domain.
1150 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1151 link count into consideration.
1153 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1154 caused the extension of big_buffer, the following lines were ignored.
1156 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1157 accordance with RFC 2308. Previously there was no expiry, so a longlived
1158 receive process (eg. due to ACL delays) versus a short SOA value could
1161 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1163 JH/39 Promote DMARC support to mainline.
1165 JH/40 Bug 2452: Add a References: header to DSNs.
1167 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1168 parameters. The relevant library call is documented as "Deprecated: This
1169 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1170 3.6.0, DH parameters are negotiated following RFC7919."
1172 HS/06 Change the default of dnssec_request_domains to "*"
1174 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1175 carried on and emitted a BDAT command, even when PIPELINING was not
1178 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1179 buffer was used for the filename, resulting in a trap when tainted
1180 arguments (eg. $domain) were used.
1182 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1183 recommended to avoid a possible server-load attack. The feature can be
1184 re-enabled via the openssl_options main cofiguration option.
1186 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1187 for version 4.90 - adding a "more data" boolean to the arguments.
1188 Bumped the ABI version number also, this having been missed previously;
1189 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1190 respect of usage of smtp_printf() by either local_scan code or libraries
1191 accessed via the ${dlfunc } expansion item. Both will need coding
1192 adjustment for any calls to smtp_printf() to match the new function
1193 signature; a FALSE value for the new argument is always safe.
1195 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1196 the file-offset (which the Linux syscall does, and exim expects); this
1197 resulted in an indefinite loop.
1199 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1200 to do ARC verification. The Authentication-Results: header line added
1201 by the configuration then had no ARC item.
1203 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1204 an incorrect mode was used for reading the result, resulting in it being
1211 JH/01 Remove code calling the customisable local_scan function, unless a new
1212 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1214 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1215 non-signal-safe functions being used.
1217 JH/03 Bug 2269: When presented with a received message having a stupidly large
1218 number of DKIM-Signature headers, disable DKIM verification to avoid
1219 a resource-consumption attack. The limit is set at twenty.
1221 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1222 report of oldest_pass in ${authres } in consequence, and separate out
1223 some descriptions of reasons for verification fail.
1225 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1226 files in the spool were present and unlocked. A queue-runner could spot
1227 them, resulting in a duplicate delivery. Fix that by doing the unlock
1228 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1229 add more error-checking on spoolfile handling while that code is being
1232 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1233 No known attacks, no CVE, this is defensive hardening.
1235 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1236 a queue-runner could start a delivery while other operations were ongoing.
1237 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1238 Found and investigated by Tim Stewart. Fix by using the open message data
1239 file handle rather than opening another, and not locally closing it (which
1240 releases a lock) for that case, while creating the temporary .eml format
1241 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1243 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1244 $sender_verify_failure/$recipient_verify_failure to "random".
1246 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1249 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1250 Previously this would segfault.
1252 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1255 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1256 like zero, since the resolver should be doing this for us, But we need one
1257 as a CNAME but no MX presence gets the CNAME returned; we need to check
1258 that doesn't point to an MX to declare it "no MX returned" rather than
1259 "error, loop". A new main option is added so the older capability of
1260 following some limited number of chain links is maintained.
1262 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1264 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1265 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1266 now-deprecated earlier definitions used only specified the range up to TLS
1267 1.2 (in the older-version library docs).
1269 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1271 JH/15 Rework TLS client-side context management. Stop using a global, and
1272 explicitly pass a context around. This enables future use of TLS for
1273 connections to service-daemons (eg. malware scanning) while a client smtp
1274 connection is using TLS; with cutthrough connections this is quite likely.
1276 JH/16 Fix ARC verification to do AS checks in reverse order.
1278 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1280 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1281 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1282 Previously the "utf8" would be re-prepended for every additional message.
1284 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1285 Previously thery were accepted, resulting in issues when attempting to
1286 forward messages to a non-supporting MTA.
1288 PP/02 Let -n work with printing macros too, not just options.
1290 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1291 one parent address was copied, and bogus data was used at delivery-logging
1292 time. Either a crash (after delivery) or bogus log data could result.
1293 Discovery and analysis by Tim Stewart.
1295 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1296 Previously if the string ended mid-character, we did not insert the
1297 promised '?' replacement.
1299 PP/04 Documentation: current string operators work on bytes, not codepoints.
1301 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1302 should pack well giving a smaller memory footprint so better caching and
1303 therefore performance. Group the declarations where this can't be done so
1304 that the byte-sized flag variables are not interspersed among pointer
1305 variables, giving a better chance of good packing by the compiler.
1307 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1308 non-null, to avoid issues with sites running BATV. Previously reports were
1309 sent with an empty envelope sender so looked like bounces.
1311 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1312 The ignore_error flag wasn't being returned from the filter subprocess so
1313 was not set for later routers. Investigation and fix by Matthias Kurz.
1315 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1316 and a msg:complete for the whole, when a message is manually removed using
1317 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1319 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1320 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1321 hope that is portable enough.
1323 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1324 requiring a known-CA anchor certificate; make it now rely entirely on the
1325 TLSA as an anchor. Checking the name on the leaf cert against the name
1326 on the A-record for the host is still done for TA (but not for EE mode).
1328 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1329 included in delivery lines for non-proxied connections, when compiled with
1330 SUPPORT_SOCKS and running with proxy logging enabled.
1332 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1333 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1334 move the existing event to fire before the normal logging of message
1335 failure so that custom logging is bracketed by normal logging.
1337 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1338 msg:fail:internal event. Developement by Matthias Kurz.
1340 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1341 far too small for todays use of crypto signatures stored there. Go all
1342 the way to the max DNS message size of 64kB, even though this might be
1343 overmuch for IOT constrained device use.
1345 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1346 copy a string over itself. The library routine is documented as not
1347 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1349 JH/32 For main options check_spool_space and check_inode_space, where the
1350 platform supports 64b integers, support more than the previous 2^31 kB
1351 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1352 the previous G, M, k.
1354 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1355 $authenticated_fail_id variable on authentication failure. Previously
1358 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1359 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1360 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1361 GNUTLS_SEC_PARAM_MEDIUM.
1363 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1364 side. Previously we would continue as if no SNI had been received.
1366 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1367 "<" character, which should be a list-separator specification, we walked
1368 off past the nul-terimation.
1370 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1371 causes) even when the retry time is not yet met. Previously they were
1372 not, meaning that when (say) an account was over-quota and temp-rejecting,
1373 and multiple senders' messages were queued, only one sender would get
1374 notified on each configured delay_warning cycle.
1376 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1378 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1379 cipher-suites, an error can be left on the stack even for a succeeding
1380 accept; this results in impossible error messages when a later operation
1383 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1384 return error codes indicating retry. Under TLS1.3 this becomes required.
1386 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1387 it only wrote the new authenticators, resulting in a lack of tracking of
1388 peer changes of ESMTP extensions until the next cache flush.
1390 JH/41 Fix the loop reading a message header line to check for integer overflow,
1391 and more-often against header_maxsize. Previously a crafted message could
1392 induce a crash of the recive process; now the message is cleanly rejected.
1394 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1395 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1401 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1402 When redis_servers is set to a list of > 1 element, and the Redis servers
1403 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1404 case of MOVED into a DEFER case instead, thus moving the query onto the
1405 next server in the list. For a cluster of N elements, all N servers must
1406 be defined in redis_servers.
1408 GF/02 Catch and remove uninitialized value warning in exiqsumm
1409 Check for existence of @ARGV before looking at $ARGV[0]
1411 JH/01 Replace the store_release() internal interface with store_newblock(),
1412 which internalises the check required to safely use the old one, plus
1413 the allocate and data copy operations duplicated in both (!) of the
1414 extant use locations.
1416 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1417 modifier. This matches the restriction on the commandline.
1419 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1420 Previously only the last row was returned.
1422 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1423 we assumed that tags in the header were well-formed, and parsed the
1424 element content after inspecting only the first char of the tag.
1425 Assumptions at that stage could crash the receive process on malformed
1428 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1429 While running the DKIM ACL we operate on the Permanent memory pool so that
1430 variables created with "set" persist to the DATA ACL. Also (at any time)
1431 DNS lookups that fail create cache records using the Permanent pool. But
1432 expansions release any allocations made on the current pool - so a dnsdb
1433 lookup expansion done in the DKIM ACL releases the memory used for the
1434 DNS negative-cache, and bad things result. Solution is to switch to the
1435 Main pool for expansions.
1436 While we're in that code, add checks on the DNS cache during store_reset,
1437 active in the testsuite.
1438 Problem spotted, and debugging aided, by Wolfgang Breyha.
1440 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1441 When none of the hosts presented to a transport match an already-open
1442 connection, close it and proceed with the list. Previously we would
1443 queue the message. Spotted by Lena with Yahoo, probably involving
1446 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1447 Previously a spurious "250 OK id=" response was appended to the proper
1450 JH/08 The "support for" informational output now, which built with Content
1451 Scanning support, has a line for the malware scanner interfaces compiled
1452 in. Interface can be individually included or not at build time.
1454 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1455 by the template makefile "src/EDITME". The "STREAM" support for an older
1456 ClamAV interface method is removed.
1458 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1459 rows affected is given instead).
1461 JH/11 The runtime Berkeley DB library version is now additionally output by
1462 "exim -d -bV". Previously only the compile-time version was shown.
1464 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1465 SMTP connection. Previously, when one had more recipients than the
1466 first, an abortive onward connection was made. Move to full support for
1467 multiple onward connections in sequence, handling cutthrough connection
1468 for all multi-message initiating connections.
1470 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1471 routers. Previously, a multi-recipient message would fail to match the
1472 onward-connection opened for the first recipient, and cause its closure.
1474 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1475 a timeout on read on a GnuTLS initiating connection, resulting in the
1476 initiating connection being dropped. This mattered most when the callout
1477 was marked defer_ok. Fix to keep the two timeout-detection methods
1480 JH/15 Relax results from ACL control request to enable cutthrough, in
1481 unsupported situations, from error to silently (except under debug)
1482 ignoring. This covers use with PRDR, frozen messages, queue-only and
1485 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1487 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1488 metadata, resulting in a crash in free().
1490 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1491 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1492 Broken also in d185889f4, with init system revamp.
1494 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1495 Previously we abruptly closed the connection after reading a malware-
1496 found indication; now we go on to read the "scan ok" response line,
1499 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1500 ACL. Previously, a crash would result.
1502 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1503 macro text after a replacement (previously it was only once per line) and
1504 by skipping builtin macros when searching for an uppercase lead character.
1506 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1507 for the build is renamed.
1509 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1510 was allocated for every new TLS startup, meaning one per message. Fix
1511 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1513 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1514 reported the original. Fix to report (as far as possible) the ACL
1515 result replacing the original.
1517 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1518 OpenSSL. Certificate information is loaded for every new TLS startup,
1519 and the resources needed to be freed.
1521 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1523 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1526 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1527 DATA response info to the (existing) per-recipient response info for
1528 the "C=" log element. It can have useful tracking info from the
1529 destination system. Patch from Simon Arlott.
1531 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1532 length value. Previously this would segfault.
1534 HS/02 Support Avast multiline protoocol, this allows passing flags to
1535 newer versions of the scanner.
1537 JH/28 Ensure that variables possibly set during message acceptance are marked
1538 dead before release of memory in the daemon loop. This stops complaints
1539 about them when the debug_store option is enabled. Discovered specifically
1540 for sender_rate_period, but applies to a whole set of variables.
1541 Do the same for the queue-runner and queue-list loops, for variables set
1542 from spool message files. Do the same for the SMTP per-message loop, for
1543 certain variables indirectly set in ACL operations.
1545 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1546 as a multi-recipient message from a mailinglist manager). The coding had
1547 an arbitrary cutoff number of characters while checking for more input;
1548 enforced by writing a NUL into the buffer. This corrupted long / fast
1549 input. The problem was exposed more widely when more pipelineing of SMTP
1550 responses was introduced, and one Exim system was feeding another.
1551 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1552 receiving system, and refused recipients seen by the sending system
1553 (propating to people being dropped from mailing lists).
1554 Discovered and pinpointed by David Carter.
1556 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1557 replaced by the ${authresults } expansion.
1559 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1561 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1562 allows proper process termination in container environments.
1564 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1565 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1567 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1568 and "err_perm", deprecated since 4.83 when the RFC-defined words
1569 "temperror" and "permerror" were introduced.
1571 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1572 transport-filters or DKIM-signing. The restriction was lost in the
1573 consolidation of verify-callout and delivery SMTP handling.
1574 Extend the restriction to also cover ARC-signing.
1576 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1577 in defer=pass mode supply a 450 to the initiator. Previously the message
1580 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1581 tls_require_ciphers is used as before.
1583 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1584 "pass_unscanned". Only tmpfails from the scanner are written to
1585 the paniclog, as they may require admin intervention (permission
1586 denied, license issues). Other scanner errors (like decompression
1587 bombs) do not cause a paniclog entry.
1589 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1590 Previously it was possible to log spurious information in receive log
1593 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1594 triggered odd behaviour from Outlook Express clients.
1596 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1599 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1600 since the IETF WG has not yet settled on that versus the original
1601 "bare" representation.
1603 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1604 Previously the millisecond value corrupted the output.
1605 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1606 corrupted the output.
1612 JH/01 Rework error string handling in TLS interface so that the caller in
1613 more cases is responsible for logging. This permits library-sourced
1614 string to be attached to addresses during delivery, and collapses
1615 pairs of long lines into single ones.
1617 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1618 during configuration. Wildcards are allowed and expanded.
1620 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1621 This permits better logging.
1623 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1624 do not close it down and have the child transport start it up again on
1625 the passed-on TCP connection. Instead, proxy the child (and any
1626 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1627 affected: the continued delivery log lines do not have any DNSSEC, TLS
1628 Certificate or OCSP information. TLS cipher information is still logged.
1630 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1631 identical IP addresses on different listening ports. Will also affect
1634 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1635 add noisy ifdef guards to special-case this sillyness.
1636 Patch from Bernd Kuhls.
1638 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1639 than 255 are no longer allowed.
1641 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1642 Disable the session-cache too, which might reduce our load. Since we
1643 currrectly use a new context for every connection, both as server and
1644 client, there is no benefit for these.
1645 GnuTLS appears to not support tickets server-side by default (we don't
1646 call gnutls_session_ticket_enable_server()) but client side is enabled
1647 by default on recent versions (3.1.3 +) unless the PFS priority string
1650 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1651 <https://reproducible-builds.org/specs/source-date-epoch/>.
1653 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1654 the check for any unsuccessful recipients did not notice the limit, and
1655 erroneously found still-pending ones.
1657 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1658 MSG_MORE. Only in-clear (not on TLS connections).
1660 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1661 transport-filter is involved we can buffer the headers in memory for
1662 creating the signature, and read the spool data file once for the
1663 signature and again for transmission.
1665 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1666 4.77 as the kernel support then wasn't solid, having issues in 64bit
1667 mode. Now, it's been long enough. Add support for FreeBSD also.
1669 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1670 case where the routing stage had gathered several addresses to send to
1671 a host before calling the transport for the first, we previously failed
1672 to close down TLS in the old transport process before passing the TCP
1673 connection to the new process. The new one sent a STARTTLS command
1674 which naturally failed, giving a failed delivery and bloating the retry
1675 database. Investigation and fix prototype from Wolfgang Breyha.
1677 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1678 false-negatives in the check that the sender had not preempted a response
1679 or prompt from Exim (running as a server), due to that code's lack of
1680 awareness of the SMTP input buffering.
1682 PP/04 Add commandline_checks_require_admin option.
1683 Exim drops privileges sanely, various checks such as -be aren't a
1684 security problem, as long as you trust local users with access to their
1685 own account. When invoked by services which pass untrusted data to
1686 Exim, this might be an issue. Set this option in main configuration
1687 AND make fixes to the calling application, such as using `--` to stop
1690 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1691 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1692 responses to those, into a single TLS record each way (this usually means
1693 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1696 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1697 affects you only if you're dancing at the edge of the param size limits.
1698 If you are, and this message makes sense to you, then: raise the
1699 configured limit or use OpenSSL 1.1. Nothing we can do for older
1702 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1703 cmdline element to get the documented default one. Previously it was
1706 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1707 are unsuable for non-message-specific reasons - eg. connection timeout,
1708 banner-time rejection.
1710 JH/16 Fix logging of delivery remote port, when specified by router, under
1713 PP/06 Repair manualroute's ability to take options in any order, even if one
1714 is the name of a transport.
1717 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1719 JH/17 Change the list-building routines interface to use the expanding-string
1720 triplet model, for better allocation and copying behaviour.
1722 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1723 Previously it was constructed the first time a possibly-matching string
1724 was met in the configuration file input during startup; now it is done
1727 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1728 compatible one, to avoid the (poorly documented) possibility of a config
1729 file in the working directory redirecting the DB files, possibly correpting
1730 some existing file. CVE-2017-10140 assigned for BDB.
1732 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1733 cache-hot. Previously, although the result was properly cached, the
1734 initial verify call returned a defer.
1736 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1737 the main verify for receipient in uncached-mode.
1739 JH/22 Retire historical build files to an "unsupported" subdir. These are
1740 defined as "ones for which we have no current evidence of testing".
1742 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1743 if present. Previously it was ignored.
1745 JH/24 Start using specified-initialisers in C structure init coding. This is
1746 a C99 feature (it's 2017, so now considered safe).
1748 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1749 if was a fixed-sized field and bitmask ops via macros; it is now more
1752 PP/07 GitHub PR 56: Apply MariaDB build fix.
1753 Patch provided by Jaroslav Škarvada.
1755 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1756 during Coverity cleanups [4.87 JH/47]
1757 Diagnosis and fix provided by Michael Fischer v. Mollard.
1759 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1760 the right size to place the terminating semicolon on its own folded
1761 line, the header hash was calculated to an incorrect value thanks to
1762 the (relaxed) space the fold became.
1764 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1765 and confused the parent.
1767 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1768 which could crash as a result. This could lead to undeliverable messages.
1770 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1773 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1774 matters on fast-turnover and PID-randomising systems, which were getting
1775 out-of-order delivery.
1777 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1778 a possibly-overlapping copy. The symptom was that "Remote host closed
1779 connection in response to HELO" was logged instead of the actual 4xx
1782 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1783 Previously only that bufferd was discarded, resulting in SYMTP command
1786 JH/32 DKIM: when a message has multiple signatures matching an identity given
1787 in dkim_verify_signers, run the dkim acl once for each. Previously only
1788 one run was done. Bug 2189.
1790 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1791 "panic the current process" to "deliberately defer". The panic log is
1792 still written with the problem list name; the mail and reject logs now
1793 get a temp-reject line for the message that was being handled, saying
1794 something like "domains check lookup or other defer". The SMTP 451
1795 message is still "Temporary local problem".
1797 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1798 A crafted sequence of BDAT commands could result in in-use memory beeing
1799 freed. CVE-2017-16943.
1801 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1802 from SMTP input. Previously it was always done; now only done for DATA
1803 and not BDAT commands. CVE-2017-16944.
1805 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1806 to the message (such as an overlong header line). Previously this was
1807 not done and we did not exit BDAT mode. Followon from the previous item
1808 though a different problem.
1814 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1815 than -2003 did; needs libidn2 in addition to libidn.
1817 JH/02 The path option on a pipe transport is now expanded before use.
1819 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1820 Patch provided by "Björn", documentation fix added too.
1822 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1823 missing a wire-to-host endian conversion.
1825 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1826 close after a BDAT command line could be taken as a following command,
1827 giving a synch failure. Fix by only checking for synch immediately
1828 before acknowledging the chunk.
1830 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1831 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1832 macro. Patches provided by Josh Soref.
1834 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1835 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1836 by IANA as a service extension. However, John Klensin suggests that we
1839 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1840 the dkim code may be unix-mode line endings rather than smtp wire-format
1841 CRLF, so prepend a CR to any bare LF.
1843 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1844 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1846 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1847 the first were themselves being wrongly included in the feed into dkim
1848 processing; with most chunk sizes in use this resulted in an incorrect
1849 body hash calculated value.
1851 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1852 DKIM signature block, for verification. Although advised against by
1853 standards it is specifically not ruled illegal.
1855 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1857 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1858 missing a body hash (the bh= tag).
1860 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1861 It seems that HAProxy sends the Proxy Protocol information in clear and
1862 only then does a TLS startup, so do the same.
1864 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1865 TCP connections (such as for Spamd) unless the daemon successfully set
1866 Fast Open mode on its listening sockets. This fixes breakage seen on
1867 too-old kernels or those not configured for Fast Open, at the cost of
1868 requiring both directions being enabled for TFO, and TFO never being used
1869 by non-daemon-related Exim processes.
1871 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1872 endings, at least on the first header line. Try to canonify any that get
1873 past that check, despite the cost.
1875 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1876 now limited to an arbitrary five deep, while parsing addresses with the
1877 strip_excess_angle_brackets option enabled.
1879 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1880 instead leave the unprompted TLS handshake in socket buffer for the
1881 TLS library to consume.
1883 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1885 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1887 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1888 region in one operation, for speed, and this leaves assigned pointers
1889 dangling. Add checks run only under the testsuite which checks all
1890 variables at a store-reset and panics on a dangling pointer; add code
1891 explicitly nulling out all the variables discovered. Fixes one known
1892 bug: a transport crash, where a dangling pointer for $sending_ip_address
1893 originally assigned in a verify callout, is re-used.
1895 PP/06 Drop '.' from @INC in various Perl scripts.
1897 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1899 PP/08 Reduce a number of compilation warnings under clang; building with
1900 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1901 should be warning-free.
1903 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1905 HS/01 Fix portability problems introduced by PP/08 for platforms where
1906 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1908 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1909 chunk. This allows us to accept broken chunked messages. We need a more
1910 general solution here.
1912 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1913 already-broken messages in the queue.
1915 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1917 JH/19 Fix reference counting bug in routing-generated-address tracking.
1923 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1924 supports it and a size is available (ie. the sending peer gave us one).
1926 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1927 years of being deprecated). The replacements are the ACLs
1928 acl_smtp_mime and acl_not_smtp_mime.
1930 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1931 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1932 or even an in-clear connection were permitted. Now, if the host lookup
1933 was dnssec and dane was requested then the host is only used if the
1934 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1935 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1936 if one fails this test.
1937 This means that a poorly-configured remote DNS will make it incommunicado;
1938 but it protects against a DNS-interception attack on it.
1940 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1941 non-noisy when a race steals the message being considered.
1943 JH/05 If main configuration option tls_certificate is unset, generate a
1944 self-signed certificate for inbound TLS connections.
1946 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1947 in rewrites and routers.
1949 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1950 and logged a warning sing 4.83; now they are a configuration file error.
1952 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1953 (lacking @domain). Apply the same qualification processing as RCPT.
1955 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1957 JH/10 Support ${sha256:} applied to a string (as well as the previous
1960 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1961 a cutthrough deliver is pending, as we always want to make a connection.
1962 This also avoids re-routing the message when later placing the cutthrough
1963 connection after a verify cache hit.
1964 Do not update it with the verify result either.
1966 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1967 when routing results in more than one destination address.
1969 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1970 signing (which inhibits the cutthrough capability). Previously only
1971 the presence of an option was tested; now an expansion evaluating as
1972 empty is permissible (obviously it should depend only on data available
1973 when the cutthrough connection is made).
1975 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1976 the relevant preceding SMTP command did not note the pipelining mode.
1978 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1979 Previously they were not counted.
1981 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1982 as one having no matching records. Previously we deferred the message
1983 that needed the lookup.
1985 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1986 distinguished as "(=".
1988 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1989 for missing MX records. Previously it only worked for missing A records.
1991 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1993 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1994 after the data-go-ahead and data-ack. Patch from Jason Betts.
1996 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1997 even for a "none" policy. Patch from Tony Meyer.
1999 JH/22 Fix continued use of a connection for further deliveries. If a port was
2000 specified by a router, it must also match for the delivery to be
2003 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2004 When one of the recipients of a message was unsuitable for the connection
2005 (has no matching addresses), we lost track of needing to mark it
2006 deferred. As a result mail would be lost.
2008 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2010 JH/25 Decoding ACL controls is now done using a binary search; the source code
2011 takes up less space and should be simpler to maintain. Merge the ACL
2012 condition decode tables also, with similar effect.
2014 JH/26 Fix problem with one_time used on a redirect router which returned the
2015 parent address unchanged. A retry would see the parent address marked as
2016 delivered, so not attempt the (identical) child. As a result mail would
2019 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2020 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2021 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2024 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2025 Main config options check_{log,spool}_{inodes,space} are now
2026 100 inodes, 10MB unless set otherwise in the configuration.
2028 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2029 Previously it only applied to the main-section connection policy
2032 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2034 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2035 by me. Added RFC7919 DH primes as an alternative.
2037 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2038 are not in the system include path.
2040 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2041 GnuTLS, when a session startup failed (eg because the client disconnected)
2042 Exim did stdio operations after fclose. This was exposed by a recent
2043 change which nulled out the file handle after the fclose.
2045 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2046 signed directly by the cert-signing cert, rather than an intermediate
2047 OCSP-signing cert. This is the model used by LetsEncrypt.
2049 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2051 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2052 an incoming connection.
2054 HS/02 Bug 1802: Do not half-close the connection after sending a request
2057 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2058 fallback to "prime256v1".
2060 JH/34 SECURITY: Use proper copy of DATA command in error message.
2061 Could leak key material. Remotely exploitable. CVE-2016-9963.
2067 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2068 and 3.4.4 - once the server is enabled to respond to an OCSP request
2069 it does even when not requested, resulting in a stapling non-aware
2070 client dropping the TLS connection.
2072 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2073 support variable-length bit vectors. No functional change.
2075 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2076 The I= interface field on outgoing lines is now after the H= remote
2077 host field, same as incoming lines. There is a separate
2078 outgoing_interface log selector which allows you to disable the
2081 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2082 If not running log_selector +smtp_connection the mainlog would be held
2083 open indefinitely after a "too many connections" event, including to a
2084 deleted file after a log rotate. Leave the per net connection logging
2085 leaving it open for efficiency as that will be quickly detected by the
2086 check on the next write.
2088 HS/01 Bug 1671: Fix post transport crash.
2089 Processing the wait-<transport> messages could crash the delivery
2090 process if the message IDs didn't exist for some reason. When
2091 using 'split_spool_directory=yes' the construction of the spool
2092 file name failed already, exposing the same netto behaviour.
2094 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2095 mime_regex ACL conditions.
2097 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2098 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2099 to HELO, local diagnostic string.
2101 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2102 log line to debug. Even when configured with a tls authenticator many
2103 client connections are expected to not authenticate in this way, so
2104 an authenticate fail is not an error.
2106 HS/02 Add the Exim version string to the process info. This way exiwhat
2107 gives some more detail about the running daemon.
2109 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2110 matter for fast-change records such as DNSBLs.
2112 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2113 retry record, even if constant. There may be multiple transports with
2114 different interface settings and the retry behaviour needs to be kept
2117 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2119 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2121 JH/10 Bug 840: fix log_defer_output option of pipe transport
2123 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2124 hosts_randomize. This matters a lot when combined with mua_wrapper.
2126 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2127 ${quote_pgsql:<string>} operator.
2129 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2131 JH/14 Change header file name for freeradius-client. Relevant if compiling
2132 with Radius support; from the Gentoo tree and checked under Fedora.
2134 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2136 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2137 option transport_filter, meaning no filtering.
2139 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2141 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2142 defaults to "*" (all hosts). The variable is now available when not built
2143 with TLS, default unset, mainly to enable keeping the testsuite sane.
2144 If a server certificate is not supplied (via tls_certificate) an error is
2145 logged, and clients will find TLS connections fail on startup. Presumably
2146 they will retry in-clear.
2147 Packagers of Exim are strongly encouraged to create a server certificate
2148 at installation time.
2150 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2151 with the $config_file variable.
2153 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2154 in transport context, after the attempt, and per-recipient. The latter type
2155 is per host attempted. The event data is the error message, and the errno
2156 information encodes the lookup type (A vs. MX) used for the (first) host,
2157 and the trailing two digits of the smtp 4xx response.
2159 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2160 to write to mainlog (or rejectlog, paniclog) in the window between file
2161 creation and permissions/ownership being changed. Particularly affects
2162 installations where exicyclog is run as root, rather than exim user;
2163 result is that the running daemon panics and dies.
2165 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2167 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2168 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2169 "pri" and "weight". Note that the previous implicit priority given by the
2170 list order is no longer honoured.
2172 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2173 for DKIM processing.
2175 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2176 by defining SUPPORT_SOCKS.
2178 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2179 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2180 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2181 variables are renamed to proxy_{local,external}_{address,port}.
2183 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2184 for a build by defining SUPPORT_I18N
2186 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2187 of the query string, and make ${quote_redis:} do that quoting.
2189 JH/29 Move Events support from Experimental to mainline, enabled by default
2190 and removable for a build by defining DISABLE_EVENT.
2192 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2194 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2195 cached by the daemon.
2197 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2198 by defining LOOKUP_REDIS. The libhiredis library is required.
2200 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2201 keys are given for lookup.
2203 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2204 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2205 only supported when built with TLS support. The PolarSSL SHA routines
2206 are still used when the TLS library is too old for convenient support.
2208 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2209 openssl_options), for security. OpenSSL forces this from version 1.1.0
2210 server-side so match that on older versions.
2212 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2213 allocation for $value could be released as the expansion processing
2214 concluded, but leaving the global pointer active for it.
2216 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2217 and to use the domains and local_parts ACL conditions.
2219 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2220 incorrectly not doubled on cutthrough transmission, hence seen as a
2221 body-termination at the receiving system - resulting in truncated mails.
2222 Commonly the sender saw a TCP-level error, and retransmitted the message
2223 via the normal store-and-forward channel. This could result in duplicates
2224 received - but deduplicating mailstores were liable to retain only the
2225 initial truncated version.
2227 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2229 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2231 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2232 we're in there, support oversigning also; bug 1309.
2234 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2236 HS/04 Add support for keep_environment and add_environment options.
2238 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2239 either intentional arithmetic overflow during PRNG, or testing config-
2242 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2243 delivery resulted in actual delivery. Cancel cutthrough before DATA
2246 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2247 rejects a recipient: pass the reject to the originator.
2249 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2250 Many were false-positives and ignorable, but it's worth fixing the
2253 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2254 for the new environment-manipulation done at startup. Move the routines
2255 from being local to tls.c to being global via the os.c file.
2257 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2258 an extract embedded as result-arg for a map, the first arg for extract
2259 is unavailable so we cannot tell if this is a numbered or keyed
2260 extraction. Accept either.
2266 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2269 JH/02 The smtp transport option "multi_domain" is now expanded.
2271 JH/03 The smtp transport now requests PRDR by default, if the server offers
2274 JH/04 Certificate name checking on server certificates, when exim is a client,
2275 is now done by default. The transport option tls_verify_cert_hostnames
2276 can be used to disable this per-host. The build option
2277 EXPERIMENTAL_CERTNAMES is withdrawn.
2279 JH/05 The value of the tls_verify_certificates smtp transport and main options
2280 default to the word "system" to access the system default CA bundle.
2281 For GnuTLS, only version 3.0.20 or later.
2283 JH/06 Verification of the server certificate for a TLS connection is now tried
2284 (but not required) by default. The verification status is now logged by
2285 default, for both outbound TLS and client-certificate supplying inbound
2288 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2291 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2292 Status Notification (bounce) messages are now MIME format per RFC 3464.
2293 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2294 under the control of the dsn_advertise_hosts option, and routers may
2295 have a dsn_lasthop option.
2297 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2298 default, modifiable by a malware= option. The list separator for
2299 the options can now be changed in the usual way. Bug 68.
2301 JH/10 The smtp_receive_timeout main option is now expanded before use.
2303 JH/11 The incoming_interface log option now also enables logging of the
2304 local interface on delivery outgoing connections.
2306 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2307 if the interface and destination host and port all match.
2309 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2312 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2313 Patch from Andrew Lewis.
2315 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2316 now supports optional time-restrictions, weighting, and priority
2317 modifiers per server. Patch originally by <rommer@active.by>.
2319 JH/16 The spamd_address main option now supports a mixed list of local
2320 and remote servers. Remote servers can be IPv6 addresses, and
2321 specify a port-range.
2323 JH/17 Bug 68: The spamd_address main option now supports an optional
2324 timeout value per server.
2326 JH/18 Bug 1581: Router and transport options headers_add/remove can
2327 now have the list separator specified.
2329 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2332 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2335 JH/21 Support for the A6 type of dns record is withdrawn.
2337 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2338 rather than the verbs used.
2340 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2341 from 255 to 1024 chars.
2343 JH/24 Verification callouts now attempt to use TLS by default.
2345 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2346 are generic router options now. The defaults didn't change.
2348 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2349 Original patch from Alexander Shikoff, worked over by JH.
2351 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2352 files if ZCAT_COMMAND is not executable.
2354 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2356 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2358 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2359 Normally benign, it bites when the pair was led to by a CNAME;
2360 modern usage is to not canonicalize the domain to a CNAME target
2361 (and we were inconsistent anyway for A-only vs AAAA+A).
2363 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2365 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2366 when evaluating $sender_host_dnssec.
2368 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2369 $sender_helo_dnssec variable.
2371 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2373 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2375 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2377 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2378 documented as working, but never had. Support all but $spam_report.
2380 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2381 added for tls authenticator.
2383 HS/03 Add perl_taintmode main config option
2389 TL/01 When running the test suite, the README says that variables such as
2390 no_msglog_check are global and can be placed anywhere in a specific
2391 test's script, however it was observed that placement needed to be near
2392 the beginning for it to behave that way. Changed the runtest perl
2393 script to read through the entire script once to detect and set these
2394 variables, reset to the beginning of the script, and then run through
2395 the script parsing/test process like normal.
2397 TL/02 The BSD's have an arc4random API. One of the functions to induce
2398 adding randomness was arc4random_stir(), but it has been removed in
2399 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2400 function when detected.
2402 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2403 cause callback expansion.
2405 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2406 syntax errors in an expansion can be treated as a string instead of
2407 logging or causing an error, due to the internal use of bool_lax
2408 instead of bool when processing it.
2410 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2411 server certificates when making smtp deliveries.
2413 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2415 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2417 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2419 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2420 Merged patch from Sebastian Wiedenroth.
2422 JH/05 Fix results-pipe from transport process. Several recipients, combined
2423 with certificate use, exposed issues where response data items split
2424 over buffer boundaries were not parsed properly. This eventually
2425 resulted in duplicates being sent. This issue only became common enough
2426 to notice due to the introduction of connection certificate information,
2427 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2429 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2430 size buffer was used, resulting in syntax errors when an expansion
2433 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2434 version 3.3.6 or later.
2436 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2437 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2438 both become "event_action", the variables become $event_name, $event_data
2439 and $event_defer_errno. There is a new variable $verify_mode, usable in
2440 routers, transports and related events. The tls:cert event is now also
2441 raised for inbound connections, if the main configuration event_action
2444 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2445 early OCSP support, but no stapling (appears to be less than 1.0.0).
2447 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2448 server certificate names available under the smtp transport option
2449 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2452 JH/10 Time-related extraction expansions from certificates now use the main
2453 option "timezone" setting for output formatting, and are consistent
2454 between OpenSSL and GnuTLS compilations. Bug 1541.
2456 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2457 encoded parameter in the incoming message. Bug 1558.
2459 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2460 include certificate info, eximon was claiming there were spoolfile
2463 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2465 JH/14 Log delivery-related information more consistently, using the sequence
2466 "H=<name> [<ip>]" wherever possible.
2468 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2469 are problematic for Debian distribution, omit them from the release
2472 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2474 JH/16 Fix string representation of time values on 64bit time_t architectures.
2477 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2478 output list separator was used.
2483 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2484 checkers that were complaining about end of non-void function with no
2487 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2488 This was a regression introduced in 4.83 by another bugfix.
2490 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2492 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2493 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2499 TF/01 Correctly close the server side of TLS when forking for delivery.
2501 When a message was received over SMTP with TLS, Exim failed to clear up
2502 the incoming connection properly after forking off the child process to
2503 deliver the message. In some situations the subsequent outgoing
2504 delivery connection happened to have the same fd number as the incoming
2505 connection previously had. Exim would try to use TLS and fail, logging
2506 a "Bad file descriptor" error.
2508 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2509 utilities have not been installed.
2511 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2512 temporary space as the ACL may create new global variables.
2514 TL/01 LDAP support uses per connection or global context settings, depending
2515 upon the detected version of the libraries at build time.
2517 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2518 to extract and use the src ip:port in logging and expansions as if it
2519 were a direct connection from the outside internet. PPv2 support was
2520 updated based on HAProxy spec change in May 2014.
2522 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2524 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2525 Properly escape header and check for NULL return.
2527 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2530 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2532 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2533 characters in header names, implemented as a verify condition.
2534 Contributed by Michael Fischer v. Mollard.
2536 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2537 results permerror and temperror. Previous values are deprecated but
2538 still accepted. In a future release, err_perm and err_temp will be
2539 completely removed, which will be a backward incompatibility if the
2540 ACL tests for either of these two old results. Patch contributed by
2541 user bes-internal on the mailing list.
2543 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2545 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2546 selectors, in both main and reject logs.
2548 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2551 JH/07 Add malware type "sock" for talking to simple daemon.
2553 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2555 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2556 routers/transports under cutthrough routing.
2558 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2559 numbers. Touch up "bool" conditional to keep the same definition.
2561 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2563 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2565 JH/12 Expand items in router/transport headers_add or headers_remove lists
2566 individually rather than the list as a whole. Bug 1452.
2568 Required for reasonable handling of multiple headers_ options when
2569 they may be empty; requires that headers_remove items with embedded
2570 colons must have them doubled (or the list-separator changed).
2572 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2573 view the policy declared in the DMARC record. Currently, $dmarc_status
2574 is a combined value of both the record presence and the result of the
2577 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2579 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2580 dnslookup router and the smtp transport (applying to the forward
2583 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2584 of ldap servers used for a specific lookup. Patch provided by Heiko
2587 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2588 New variable $lookup_dnssec_authenticated for observability.
2590 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2591 Patch submitted by Lars Timman.
2593 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2595 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2596 Requires trusted mode and valid format message id, aborts otherwise.
2597 Patch contributed by Heiko Schlichting.
2599 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2600 certextract with support for various fields. Bug 1358.
2602 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2603 is requested by default, modifiable by smtp transport option
2606 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2607 operate on certificate variables to give certificate fingerprints
2608 Also new ${sha256:cert_variable}.
2610 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2612 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2613 Christian Aistleitner.
2615 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2617 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2618 file. Patch from Wolfgang Breyha.
2620 JH/25 Expand the coverage of the delivery $host and $host_address to
2621 client authenticators run in verify callout. Bug 1476.
2623 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2624 align with daemon_smtp_ports. Bug 72.
2626 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2627 support and error reporting did not work properly.
2629 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2630 and is readable. Patch from Andrew Colin Kissa.
2632 TL/14 Enhance documentation of ${run expansion and how it parses the
2633 commandline after expansion, particularly in the case when an
2634 unquoted variable expansion results in an empty value.
2636 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2638 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2639 by renaming away from it. Addresses GH issue 15, reported by
2642 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2644 TL/15 SECURITY: prevent double expansion in math comparison functions
2645 (can expand unsanitized data). Not remotely exploitable.
2652 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2654 PP/02 Make -n do something, by making it not do something.
2655 When combined with -bP, the name of an option is not output.
2657 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2660 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2661 $sender_host_name and config options to manage this, and basic check
2664 PP/05 DSCP support for outbound connections and control modifier for inbound.
2666 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2667 (Only plugin which currently uses this is kerberos4, which nobody should
2668 be using, but we should make it available and other future plugins might
2669 conceivably use it, even though it would break NAT; stuff *should* be
2670 using channel bindings instead).
2672 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2673 name; added for Sendmail compatibility; requires admin caller.
2674 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2675 just ignore it); requires trusted caller.
2676 Also parse but ignore: -Ac -Am -X<logfile>
2679 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2681 TL/02 Add +smtp_confirmation as a default logging option.
2683 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2684 Patch by Magnus Holmgren from 2007-02-20.
2686 TL/04 Bugzilla 1281 - Spec typo.
2687 Bugzilla 1283 - Spec typo.
2688 Bugzilla 1290 - Spec grammar fixes.
2690 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2692 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2694 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2695 mailing list by by Dmitry Isaikin.
2697 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2699 JH/02 Support "G" suffix to numbers in ${if comparisons.
2701 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2703 NM/01 Bugzilla 1197 - Spec typo
2704 Bugzilla 1196 - Spec examples corrections
2706 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2708 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2709 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2712 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2713 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2715 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2716 "acl {{name}{arg}...}", and optional args on acl condition
2719 JH/05 Permit multiple router/transport headers_add/remove lines.
2721 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2723 JH/07 Avoid using a waiting database for a single-message-only transport.
2724 Performance patch from Paul Fisher. Bugzilla 1262.
2726 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2729 JH/09 Add $headers_added variable, with content from use of ACL modifier
2730 add_header (but not yet added to the message). Bugzilla 199.
2732 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2733 Pulled from Bugzilla 817 by Wolfgang Breyha.
2735 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2737 (nb: this is the same fix as in Exim 4.80.1)
2739 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2742 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2745 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2747 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2748 advertise SMTP AUTH mechanism to us, instead of a generic
2749 protocol violation error. Also, make Exim more robust to bad
2750 data from the Dovecot auth socket.
2752 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2754 When a queue runner is handling a message, Exim first routes the
2755 recipient addresses, during which it prunes them based on the retry
2756 hints database. After that it attempts to deliver the message to
2757 any remaining recipients. It then updates the hints database using
2760 So if a recipient address works intermittently, it can get repeatedly
2761 deferred at routing time. The retry hints record remains fresh so the
2762 address never reaches the final cutoff time.
2764 This is a fairly common occurrence when a user is bumping up against
2765 their storage quota. Exim had some logic in its local delivery code
2766 to deal with this. However it did not apply to per-recipient defers
2767 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2769 This change adds a proper retry rule check during routing so that the
2770 final cutoff time is checked against the message's age. We only do
2771 this check if there is an address retry record and there is not a
2772 domain retry record; this implies that previous attempts to handle
2773 the address had the retry_use_local_parts option turned on. We use
2774 this as an approximation for the destination being like a local
2775 delivery, as in LMTP.
2777 I suspect this new check makes the old local delivery cutoff check
2778 redundant, but I have not verified this so I left the code in place.
2780 TF/02 Correct gecos expansion when From: is a prefix of the username.
2782 Test 0254 submits a message to Exim with the header
2786 When I ran the test suite under the user fanf2, Exim expanded
2787 the header to contain my full name, whereas it should have added
2788 a Resent-Sender: header. It erroneously treats any prefix of the
2789 username as equal to the username.
2791 This change corrects that bug.
2793 GF/01 DCC debug and logging tidyup
2794 Error conditions log to paniclog rather than rejectlog.
2795 Debug lines prefixed by "DCC: " to remove any ambiguity.
2797 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2799 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2800 Bug spotted by Jeremy Harris; was flawed since initial commit.
2801 Would have resulted in OCSP responses post-SNI triggering an Exim
2802 NULL dereference and crash.
2804 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2806 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2807 Bug detection, analysis and fix by Samuel Thibault.
2808 Bugzilla 1331, Debian bug #698092.
2810 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2812 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2813 Server implementation by Todd Lyons, client by JH.
2814 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2815 config variable "prdr_enable" controls whether the server
2816 advertises the facility. If the client requests PRDR a new
2817 acl_data_smtp_prdr ACL is called once for each recipient, after
2818 the body content is received and before the acl_smtp_data ACL.
2819 The client is controlled by both of: a hosts_try_prdr option
2820 on the smtp transport, and the server advertisement.
2821 Default client logging of deliveries and rejections involving
2822 PRDR are flagged with the string "PRDR".
2824 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2825 fclose(). Diagnosis by Todd Lyons.
2827 PP/17 Update configure.default to handle IPv6 localhost better.
2828 Patch by Alain Williams (plus minor tweaks).
2831 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2832 This is now consistent with GnuTLS, and is now documented: the
2833 previous undocumented portable approach to treating the option as
2834 unset was to force an expansion failure. That still works, and
2835 an empty string is now equivalent.
2837 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2838 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2839 not performing validation itself.
2841 PP/20 Added force_command boolean option to pipe transport.
2842 Patch from Nick Koston, of cPanel Inc.
2844 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2847 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2849 PP/21 Fix eximon continuous updating with timestamped log-files.
2850 Broken in a format-string cleanup in 4.80, missed when I repaired the
2851 other false fix of the same issue.
2852 Report and fix from Heiko Schlichting.
2855 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2856 Report from Prashanth Katuri.
2858 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2859 It's SecureTransport, so affects any MacOS clients which use the
2860 system-integrated TLS libraries, including email clients.
2862 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2863 using a MIME ACL for non-SMTP local injection.
2864 Report and assistance in diagnosis by Warren Baker.
2866 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2868 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2870 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2871 last id that failed so it may be referenced in subsequent ACL's.
2873 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2876 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2877 ldap library initialization, allowing self-signed CA's to be
2878 used. Also properly sets require_cert option later in code by
2879 using NULL (global ldap config) instead of ldap handle (per
2880 session). Bug diagnosis and testing by alxgomz.
2882 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2883 the src/util/ subdirectory.
2885 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2886 renamed to Transport Post Delivery Action by Jeremy Harris, as
2889 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2890 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2891 redis_servers = needs to be configured which will be used by the redis
2892 lookup. Patch from Warren Baker, of The Packet Hub.
2894 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2896 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2897 hostname or reverse DNS when processing a host list. Used suggestions
2898 from multiple comments on this bug.
2900 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2902 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2903 Missed a few lines, added it to make the runtest require no keyboard
2906 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2907 contains upper case chars. Make router use caseful_local_part.
2909 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2910 support when GnuTLS has been built with p11-kit.
2916 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2918 This, or similar/improved, will also be change PP/11 of 4.82.
2924 PP/01 Handle short writes when writing local log-files.
2925 In practice, only affects FreeBSD (8 onwards).
2926 Bugzilla 1053, with thanks to Dmitry Isaikin.
2928 NM/01 Bugzilla 949 - Documentation tweak
2930 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2933 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2935 PP/02 Implemented gsasl authenticator.
2937 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2939 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2940 `pkg-config foo` for cflags/libs.
2942 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2943 with rest of GSASL and with heimdal_gssapi.
2945 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2946 `pkg-config foo` for cflags/libs for the TLS implementation.
2948 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2949 properties get this fed in as external SSF. A number of robustness
2950 and debugging improvements to the cyrus_sasl authenticator.
2952 PP/08 cyrus_sasl server now expands the server_realm option.
2954 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2955 Patch by Jeremy Harris.
2957 PP/10 Added dbmjz lookup type.
2959 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2961 PP/12 MAIL args handles TAB as well as SP, for better interop with
2962 non-compliant senders.
2963 Analysis and variant patch by Todd Lyons.
2965 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2966 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2967 Patch from Dirk Mueller <dmueller@suse.com>
2969 PP/13 tls_peerdn now print-escaped for spool files.
2970 Observed some $tls_peerdn in wild which contained \n, which resulted
2971 in spool file corruption.
2973 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2974 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2975 or write after TLS renegotiation, which otherwise led to messages
2978 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2979 as a tracking header (ie: a signed header comes before the signature).
2980 Patch from Wolfgang Breyha.
2982 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2983 comma-sep list; embedded commas doubled.
2985 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2987 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2989 Report and patch from Dmitry Banschikov.
2991 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2992 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2993 failures. We appear to now support TLS1.1+ with Exim.
2995 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2996 lets Exim select keys and certificates based upon TLS SNI from client.
2997 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2998 before an outbound SMTP session. New log_selector, +tls_sni.
3000 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3001 NULL dereference. Report and patch from Alun Jones.
3003 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3004 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3005 Not seeing resolver debug output on NetBSD, but suspect this is a
3006 resolver implementation change.
3008 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3009 Left warnings. Added "eximon gdb" invocation mode.
3011 PP/21 Defaulting "accept_8bitmime" to true, not false.
3013 PP/22 Added -bw for inetd wait mode support.
3015 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3016 locate the relevant includes and libraries. Made this the default.
3018 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3019 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3021 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3022 This may cause build issues on older platforms.
3024 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3025 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3026 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3027 Added SNI support via GnuTLS too.
3028 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3030 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3032 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3033 Applied second patch from Janne, implementing suggestion to default
3034 multiple-strings-in-record handling to match SPF spec.
3036 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3038 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3039 read-only, out of scope).
3040 Patch from Wolfgang Breyha, report from Stuart Northfield.
3042 PP/29 Fix three issues highlighted by clang analyser static analysis.
3043 Only crash-plausible issue would require the Cambridge-specific
3044 iplookup router and a misconfiguration.
3045 Report from Marcin Mirosław.
3047 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3049 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3050 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3051 As part of this, removing so much warning spew let me fix some minor
3052 real issues in debug logging.
3054 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3055 assignment on my part. Fixed.
3057 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3058 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3059 Janne Snabb (who went above and beyond: thank you).
3061 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3062 string otherwise requires a connection and a bunch more work and it's
3063 relatively easy to get wrong. Should also expose TLS library linkage
3066 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3067 64-bit ${eval} (JH/03).
3069 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3070 GNU libc to support some of the 64-bit stuff, should not lead to
3071 conflicts. Defined before os.h is pulled in, so if a given platform
3072 needs to override this, it can.
3074 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3075 protection layer was required, which is not implemented.
3076 Bugzilla 1254, patch from Wolfgang Breyha.
3078 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3079 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3080 tls_dhparam take prime identifiers. Also unbreak combination of
3081 OpenSSL+DH_params+TLSSNI.
3083 PP/39 Disable SSLv2 by default in OpenSSL support.
3089 PP/01 Solaris build fix for Oracle's LDAP libraries.
3090 Bugzilla 1109, patch from Stephen Usher.
3092 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3094 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3097 TF/02 Fix a couple more cases where we did not log the error message
3098 when unlink() failed. See also change 4.74-TF/03.
3100 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3101 lock up or crash if it happened to be inside a call to libc when it
3102 got a SIGUSR1 from exiwhat.
3104 The SIGUSR1 handler appends the current process status to the process
3105 log which is later printed by exiwhat. It used to use the general
3106 purpose logging code to do this, but several functions it calls are
3107 not safe for signals.
3109 The new output code in the SIGUSR1 handler is specific to the process
3110 log, and simple enough that it's easy to inspect for signal safety.
3111 Removing some special cases also simplifies the general logging code.
3112 Removing the spurious timestamps from the process log simplifies
3115 TF/04 Improved ratelimit ACL condition.
3117 The /noupdate option has been deprecated in favour of /readonly which
3118 has clearer semantics. The /leaky, /strict, and /readonly update modes
3119 are mutually exclusive. The update mode is no longer included in the
3120 database key; it just determines when the database is updated. (This
3121 means that when you upgrade Exim will forget old rate measurements.)
3123 Exim now checks that the per_* options are used with an update mode that
3124 makes sense for the current ACL. For example, when Exim is processing a
3125 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3126 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3127 must specify per_mail/readonly. If you omit the update mode it defaults to
3128 /leaky where that makes sense (as before) or /readonly where required.
3130 The /noupdate option is now undocumented but still supported for
3131 backwards compatibility. It is equivalent to /readonly except that in
3132 ACLs where /readonly is required you may specify /leaky/noupdate or
3133 /strict/noupdate which are treated the same as /readonly.
3135 A useful new feature is the /count= option. This is a generalization
3136 of the per_byte option, so that you can measure the throughput of other
3137 aggregate values. For example, the per_byte option is now equivalent
3138 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3140 The per_rcpt option has been generalized using the /count= mechanism
3141 (though it's more complicated than the per_byte equivalence). When it is
3142 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3143 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3144 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3145 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3146 Note that using per_rcpt with a non-readonly update mode in more than
3147 one ACL will cause the recipients to be double-counted. (The per_mail
3148 and per_byte options don't have this problem.)
3150 The handling of very low rates has changed slightly. If the computed rate
3151 is less than the event's count (usually one) then this event is the first
3152 after a long gap. In this case the rate is set to the same as this event's
3153 count, so that the first message of a spam run is counted properly.
3155 The major new feature is a mechanism for counting the rate of unique
3156 events. The new per_addr option counts the number of different
3157 recipients that someone has sent messages to in the last time period. It
3158 behaves like per_rcpt if all the recipient addresses are different, but
3159 duplicate recipient addresses do not increase the measured rate. Like
3160 the /count= option this is a general mechanism, so the per_addr option
3161 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3162 example, measure the rate that a client uses different sender addresses
3163 with the options per_mail/unique=$sender_address. There are further
3164 details in the main documentation.
3166 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3168 TF/06 Removed a few PCRE remnants.
3170 TF/07 Automatically extract Exim's version number from tags in the git
3171 repository when doing development or release builds.
3173 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3174 Bugzilla 879. Patch from Paul Fisher.
3176 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3177 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3180 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3182 PP/05 Variable $av_failed, true if the AV scanner deferred.
3183 Bugzilla 1078. Patch from John Horne.
3185 PP/06 Stop make process more reliably on build failure.
3186 Bugzilla 1087. Patch from Heiko Schlittermann.
3188 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3189 Bugzilla 1089. Patch from Heiko Schlittermann.
3191 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3192 Bugzilla 1131. Patch from Holger Weiß.
3194 PP/09 Handle IPv6 addresses with SPF.
3195 Bugzilla 860. Patch from Wolfgang Breyha.
3197 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3199 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3202 PP/11 match_* no longer expand right-hand-side by default.
3203 New compile-time build option, EXPAND_LISTMATCH_RHS.
3204 New expansion conditions, "inlist", "inlisti".
3206 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3208 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3210 PP/14 fix log_write() format string regression from TF/03.
3211 Bugzilla 1152. Patch from Dmitry Isaikin.
3217 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3219 PP/02 Harmonised TLS library version reporting; only show if debugging.
3220 Layout now matches that introduced for other libraries in 4.74 PP/03.
3222 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3224 PP/04 New "dns_use_edns0" global option.
3226 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3229 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3230 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3232 TK/01 Updated PolarSSL code to 0.14.2.
3233 Bugzilla 1097. Patch from Andreas Metzler.
3235 PP/07 Catch divide-by-zero in ${eval:...}.
3236 Fixes bugzilla 1102.
3238 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3241 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3242 format-string attack -- SECURITY: remote arbitrary code execution.
3244 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3245 time unintentionally subject to list matching rules, letting the header
3246 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3247 arbitrary string expansion). This allowed for information disclosure.
3249 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3250 INT_MIN/-1 -- value coerced to INT_MAX.
3256 NM/01 Workaround for PCRE version dependency in version reporting
3259 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3260 This fixes portability to compilers other than gcc, notably
3261 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3263 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3264 makefiles for portability to HP-UX and POSIX correctness.
3266 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3267 Also via indented variable definition in the Makefile.
3268 (Debugging by Oliver Heesakkers).
3270 PP/02 Restore caching of spamd results with expanded spamd_address.
3271 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3273 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3274 Improves build reliability. Fix from: Frank Elsner
3276 NM/02 Fix wide character breakage in the rfc2047 coding
3277 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3279 NM/03 Allow underscore in dnslist lookups
3280 Fixes bug 1026. Patch from Graeme Fowler
3282 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3283 Code patches from Adam Ciarcinski of NetBSD.
3285 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3288 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3289 is logged, to avoid truncation. Patch from John Horne.
3291 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3292 Patch from Jakob Hirsch.
3294 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3295 SQL string expansion failure details.
3296 Patch from Andrey Oktyabrski.
3298 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3299 Patch from Simon Arlott.
3301 PP/09 New lookups functionality failed to compile on old gcc which rejects
3302 extern declarations in function scope.
3303 Patch from Oliver Fleischmann
3305 PP/10 Use sig_atomic_t for flags set from signal handlers.
3306 Check getgroups() return and improve debugging.
3307 Fixed developed for diagnosis in bug 927 (which turned out to be
3310 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3311 Patch from Mark Zealey.
3313 PP/12 Bugzilla 1056: Improved spamd server selection.
3314 Patch from Mark Zealey.
3316 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3317 Based on patch from Heiko Schlittermann.
3319 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3320 Patch from Uwe Doering, sign-off by Michael Haardt.
3322 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3323 variable declaration deep within a block. Bug and patch from
3326 PP/15 lookups-Makefile IRIX compatibility coercion.
3328 PP/16 Make DISABLE_DKIM build knob functional.
3330 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3331 Patch by Simon Arlott
3333 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3334 variable argument macros. Our copy now differs from upstream.
3340 TF/01 Failure to get a lock on a hints database can have serious
3341 consequences so log it to the panic log.
3343 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3344 controlled using the smtp_confirmation log selector.
3346 TF/03 Include the error message when we fail to unlink a spool file.
3348 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3349 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3350 for maintaining out-of-tree patches for some time.
3352 PP/01 Bugzilla 139: Documentation and portability issues.
3353 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3354 Handle per-OS dynamic-module compilation flags.
3356 PP/02 Let /dev/null have normal permissions.
3357 The 4.73 fixes were a little too stringent and complained about the
3358 permissions on /dev/null. Exempt it from some checks.
3359 Reported by Andreas M. Kirchwitz.
3361 PP/03 Report version information for many libraries, including
3362 Exim version information for dynamically loaded libraries. Created
3363 version.h, now support a version extension string for distributors
3364 who patch heavily. Dynamic module ABI change.
3366 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3367 privilege escalation vulnerability whereby the Exim run-time user
3368 can cause root to append content of the attacker's choosing to
3371 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3374 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3375 If dropping privileges for untrusted macros, we disabled normal logging
3376 on the basis that it would fail; for the Exim run-time user, this is not
3377 the case, and it resulted in successful deliveries going unlogged.
3378 Fixed. Reported by Andreas Metzler.
3384 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3385 only prepend for the Resent-* case. Fixes regression introduced in
3386 Exim 4.70 by NM/22 for Bugzilla 607.
3388 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3389 increasing numbers of administrators be bitten by this.
3391 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3393 PP/03 Bugzilla 994: added openssl_options main configuration option.
3395 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3397 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3399 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3401 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3402 without a peer certificate, leading to a segfault because of an
3403 assumption that peers always have certificates. Be a little more
3404 paranoid. Problem reported by Martin Tscholak.
3406 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3407 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3408 NB: ClamAV planning to remove STREAM in "middle of 2010".
3409 CL also introduces -bmalware, various -d+acl logging additions and
3410 more caution in buffer sizes.
3412 PP/09 Implemented reverse_ip expansion operator.
3414 PP/10 Bugzilla 937: provide a "debug" ACL control.
3416 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3418 PP/12 Bugzilla 973: Implement --version.
3420 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3422 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3424 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3426 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3427 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3428 ignore trailing whitespace.
3430 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3432 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3435 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3436 Notification from Dr Andrew Aitchison.
3438 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3439 ExtendedDetectionInfo response format.
3440 Notification from John Horne.
3442 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3445 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3446 XSL and documented dependency on system catalogs, with examples of how
3449 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3452 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3453 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3454 configuration file which is writeable by the Exim user or group.
3456 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3457 of configuration files to cover files specified with the -C option if
3458 they are going to be used with root privileges, not just the default
3461 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3462 option (effectively making it always true).
3464 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3465 files to be used while preserving root privileges.
3467 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3468 that rogue child processes cannot use them.
3470 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3471 run-time user, instead of root.
3473 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3474 Exim run-time user without dropping privileges.
3476 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3477 result string, instead of calling string_vformat() twice with the same
3480 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3481 for other users. Others should always drop root privileges if they use
3482 -C on the command line, even for a whitelisted configure file.
3484 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3486 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3492 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3493 $data_path, and $header_path variables; fixed documentation bugs and
3496 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3497 exipick to access non-standard spools, including the "frozen" queue
3500 NM/01 Bugzilla 965: Support mysql stored procedures.
3501 Patch from Alain Williams
3503 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3505 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3506 Patch from Andreas Metzler
3508 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3509 Patch from Kirill Miazine
3511 NM/05 Bugzilla 671: Added umask to procmail example.
3513 JJ/03 installed exipick 20100323.0, fixing doc bug
3515 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3516 directory. Notification and patch from Dan Rosenberg.
3518 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3520 TK/02 Improve log output when DKIM signing operation fails.
3522 MH/01 Treat the transport option dkim_domain as a colon separated
3523 list, not as a single string, and sign the message with each element,
3524 omitting multiple occurences of the same signer.
3526 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3527 Bugzilla 985, 986. Patch by Simon Arlott
3529 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3530 Patch by Simon Arlott
3532 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3533 MBX locking. Notification from Dan Rosenberg.
3539 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3541 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3543 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3545 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3547 NM/04 Bugzilla 915: Flag broken perl installation during build.
3553 TK/01 Added patch by Johannes Berg that expands the main option
3554 "spamd_address" if it starts with a dollar sign.
3556 TK/02 Write list of recipients to X-Envelope-Sender header when building
3557 the mbox-format spool file for content scanning (suggested by Jakob
3560 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3561 (http://www.dcc-servers.net/) support via dccifd. Activated by
3562 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3564 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3565 by Mark Daniel Reidel <mr@df.eu>.
3567 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3568 When building exim an external PCRE library is now needed -
3569 PCRE is a system library on the majority of modern systems.
3570 See entry on PCRE_LIBS in EDITME file.
3572 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3573 conversation. Added nologin parameter to request.
3574 Patch contributed by Kirill Miazine.
3576 TF/01 Do not log submission mode rewrites if they do not change the address.
3578 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3580 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3581 log files in place. Contributed by Roberto Lima.
3583 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3585 TF/03 Bugzilla 615: When checking the local_parts router precondition
3586 after a local_part_suffix or local_part_prefix option, Exim now
3587 does not use the address's named list lookup cache, since this
3588 contains cached lookups for the whole local part.
3590 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3591 Robert Millan. Documentation is in experimental-spec.txt.
3593 TF/04 Bugzilla 668: Fix parallel build (make -j).
3595 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3597 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3598 Patch provided by Jan Srzednicki.
3600 TF/05 Leading white space used to be stripped from $spam_report which
3601 wrecked the formatting. Now it is preserved.
3603 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3604 that they are available at delivery time.
3606 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3608 TF/08 TLS error reporting now respects the incoming_interface and
3609 incoming_port log selectors.
3611 TF/09 Produce a more useful error message if an SMTP transport's hosts
3612 setting expands to an empty string.
3614 NM/06 Bugzilla 744: EXPN did not work under TLS.
3615 Patch provided by Phil Pennock.
3617 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3618 Patch provided by Richard Godbee.
3620 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3621 acl_smtp_notquit, added index entry.
3623 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3624 Patch provided by Eugene Bujak.
3626 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3627 accept(). Patch provided by Maxim Dounin.
3629 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3630 Patch provided by Phil Pennock.
3632 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3634 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3635 Patch provided by Brad "anomie" Jorsch.
3637 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3638 Patch provided by Dean Brooks.
3640 TK/05 Add native DKIM support (does not depend on external libraries).
3642 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3643 Patch provided by Graeme Fowler.
3645 NM/16 Bugzilla 851: Documentation example syntax fix.
3647 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3649 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3652 NM/19 Bugzilla 745: TLS version reporting.
3653 Patch provided by Phil Pennock.
3655 NM/20 Bugzilla 167: bool: condition support.
3656 Patch provided by Phil Pennock.
3658 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3659 clients. Patch provided by Phil Pennock.
3661 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3662 Patch provided by Brad "anomie" Jorsch.
3664 NM/23 Bugzilla 687: Fix misparses in eximstats.
3665 Patch provided by Heiko Schlittermann.
3667 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3668 Patch provided by Heiko Schlittermann.
3670 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3671 plus update to original patch.
3673 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3675 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3676 Patch provided by David Brownlee.
3678 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3680 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3682 NM/30 Bugzilla 888: TLS documentation bugfixes.
3684 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3686 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3687 Unlike the original bugzilla I have changed all shell scripts in src tree.
3689 NM/33 Bugzilla 898: Transport filter timeout fix.
3690 Patch by Todd Rinaldo.
3692 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3693 Patch by Serge Demonchaux.
3695 NM/35 Bugzilla 39: Base64 decode bug fixes.
3696 Patch by Jakob Hirsch.
3698 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3700 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3702 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3704 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3710 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3711 ALT-N's libdkim that I have put here:
3712 http://duncanthrax.net/exim-experimental/
3714 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3715 were called 'true' and it seems that C99 defines that as a reserved
3716 keyword to be used with 'bool' variable types. That means you could
3717 not include C99-style headers which use bools without triggering
3718 build errors in sieve.c.
3720 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3721 as mailq or other aliases. Changed the --help handling significantly
3722 to do whats expected. exim_usage() emits usage/help information.
3724 SC/01 Added the -bylocaldomain option to eximstats.
3726 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3728 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3730 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3736 PH/01 Another patch from the Sieve maintainer.
3738 PH/02 When an IPv6 address is converted to a string for single-key lookup
3739 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3740 dots are used instead of colons so that keys in lsearch files need not
3741 contain colons. This was done some time before quoting was made available
3742 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3743 (notated using the quote facility) so as to distinguish them from IPv4
3744 keys. This meant that lookups for IP addresses in host lists did not work
3745 for iplsearch lookups.
3747 This has been fixed by arranging for IPv6 addresses to be expressed with
3748 colons if the lookup type is iplsearch. This is not incompatible, because
3749 previously such lookups could never work.
3751 The situation is now rather anomalous, since one *can* have colons in
3752 ordinary lsearch keys. However, making the change in all cases is
3753 incompatible and would probably break a number of configurations.
3755 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3758 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3759 conversion specification without a maximum field width, thereby enabling
3760 a rogue spamd server to cause a buffer overflow. While nobody in their
3761 right mind would setup Exim to query an untrusted spamd server, an
3762 attacker that gains access to a server running spamd could potentially
3763 exploit this vulnerability to run arbitrary code as the Exim user.
3765 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3766 $primary_hostname instead of what libspf2 thinks the hosts name is.
3768 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3769 a directory entry by the name of the lookup key. Previously, if a
3770 symlink pointed to a non-existing file or a file in a directory that
3771 Exim lacked permissions to read, a lookup for a key matching that
3772 symlink would fail. Now it is enough that a matching directory entry
3773 exists, symlink or not. (Bugzilla 503.)
3775 PH/03 The body_linecount and body_zerocount variables are now exported in the
3778 PH/04 Added the $dnslist_matched variable.
3780 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3781 This means they are set thereafter only if the connection becomes
3784 PH/06 Added the client_condition to authenticators so that some can be skipped
3785 by clients under certain conditions.
3787 PH/07 The error message for a badly-placed control=no_multiline_responses left
3788 "_responses" off the end of the name.
3790 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3792 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3793 (without spaces) instead of just copying the configuration text.
3795 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3797 PH/11 Added $max_received_linelength.
3799 PH/12 Added +ignore_defer and +include_defer to host lists.
3801 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3802 way in which PCRE > 7.0 is built.
3804 PH/14 Implemented queue_only_load_latch.
3806 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3807 MAIL command. The effect was to mangle the value on 64-bit systems.
3809 PH/16 Another patch from the Sieve maintainer.
3811 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3813 PH/18 If a system quota error occurred while trying to create the file for
3814 a maildir delivery, the message "Mailbox is full" was not appended to the
3815 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3816 only to a quota excession during the actual writing of the file.
3818 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3819 characters?) which causes problems in log lines. The DN values are now
3820 passed through string_printing() before being added to log lines.
3822 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3823 and InterBase are left for another time.)
3825 PH/21 Added message_body_newlines option.
3827 PH/22 Guard against possible overflow in moan_check_errorcopy().
3829 PH/23 POSIX allows open() to be a macro; guard against that.
3831 PH/24 If the recipient of an error message contained an @ in the local part
3832 (suitably quoted, of course), incorrect values were put in $domain and
3833 $local_part during the evaluation of errors_copy.
3839 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3840 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3843 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3844 issue a MAIL command.
3846 PH/02 In an ACL statement such as
3848 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3850 if a client was not listed at all, or was listed with a value other than
3851 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3852 the condition was not true (as it should be), so access was not denied.
3853 The bug was that the ! inversion was incorrectly passed on to the second
3854 item. This has been fixed.
3856 PH/03 Added additional dnslists conditions == and =& which are different from
3857 = and & when the dns lookup returns more than one IP address.
3859 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3860 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3862 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3863 FSYNC, which compiles an option called disable_fsync that allows for
3864 bypassing fsync(). The documentation is heavily laced with warnings.
3866 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3868 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3869 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3870 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3871 including adding "make clean"; (3) Added -fPIC when compiling the test
3872 dynamically loaded module, to get rid of a warning.
3874 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3875 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3876 = 0s. The bug is otherwise harmless.
3878 PH/07 There was a bug in the dovecot authenticator such that the value of
3879 $auth1 could be overwritten, and so not correctly preserved, after a
3880 successful authentication. This usually meant that the value preserved by
3881 the server_setid option was incorrect.
3883 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3885 PH/09 Installed PCRE release 7.0.
3887 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3888 run for batched SMTP input. It is now run at the start of every message
3889 in the batch. While fixing this I discovered that the process information
3890 (output by running exiwhat) was not always getting set for -bs and -bS
3891 input. This is fixed, and it now also says "batched" for BSMTP.
3893 PH/11 Added control=no_pipelining.
3895 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3896 patch, slightly modified), and move the expansion of helo_data till after
3897 the connection is made in the smtp transport (so it can use these
3900 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3902 PH/14 Added log_selector = +pid.
3904 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3906 PH/16 Add ${if forany and ${if forall.
3908 PH/17 Added dsn_from option to vary the From: line in DSNs.
3910 PH/18 Flush SMTP output before performing a callout, unless control =
3911 no_callout_flush is set.
3913 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3914 was true (the default) a successful delivery failed to delete the retry
3915 item, thus causing premature timeout of the address. The bug is now
3918 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3920 PH/21 Long custom messages for fakedefer and fakereject are now split up
3921 into multiline responses in the same way that messages for "deny" and
3922 other ACL rejections are.
3924 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3925 with slight modification.
3927 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3928 draft, changing the syntax and factoring some duplicate code".
3930 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3931 for deliveries of the second and subsequent messages over the same SMTP
3934 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3935 ${reduce, with only minor "tidies".
3937 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3939 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3940 expansion side effects.
3942 PH/27 When a message times out after an over-quota error from an Exim-imposed
3943 quota, the bounce message says "mailbox is full". This message was not
3944 being given when it was a system quota that was exceeded. It now should
3947 MH/03 Made $recipients available in local_scan(). local_scan() already has
3948 better access to the recipient list through recipients_list[], but
3949 $recipients can be useful in postmaster-provided expansion strings.
3951 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3952 in the case of a MAIL command with additional options following the
3953 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3954 were accidentally chopped off.
3956 PH/29 SMTP synchronization checks are implemented when a command is read -
3957 there is a check that no more input is waiting when there shouldn't be
3958 any. However, for some commands, a delay in an ACL can mean that it is
3959 some time before the response is written. In this time, more input might
3960 arrive, invalidly. So now there are extra checks after an ACL has run for
3961 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3962 pipelining has not been advertised.
3964 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3966 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3967 returned in the SMTP response when smtp_return_error_details was false.
3968 This has been fixed.
3970 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3971 instead of the C I/O that was originally supplied, because problems were
3972 reported on Solaris.
3974 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3975 Exim which did not show up earlier: it was assuming that a call to
3976 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3977 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3978 was a macro that became an assignment, so it seemed to work. This has
3979 changed to a proper function call with a void return, hence the compile
3980 error. Exim's code has been fixed.
3982 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3985 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3987 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3989 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3990 because it turns out to be too big or there is a timeout) there is no
3991 "Completed" line in the log. When some messages of this type were
3992 selected by exigrep, they were listed as "not completed". Others were
3993 picked up by some special patterns. I have improved the selection
3994 criteria to be more general.
3996 PH/38 The host_find_failed option in the manualroute router can now be set
3997 to "ignore", to completely ignore a host whose IP address cannot be
3998 found. If all hosts are ignored, the behaviour is controlled by the new
3999 host_all_ignored option.
4001 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4002 homing or because of multiple MX records with /mx) generated more than
4003 one IP address, and the following item turned out to be the local host,
4004 all the secondary addresses of the first item were incorrectly removed
4005 from the list, along with the local host and any following hosts (which
4006 is what is supposed to happen).
4008 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4009 whoever called Exim into the -H file. In the case of the daemon it was
4010 behaving confusingly. When first started, it used values for whoever
4011 started the daemon, but after a SIGHUP it used the Exim user (because it
4012 calls itself on a restart). I have changed the code so that it now always
4015 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4016 message are rejected with the same error (e.g. no authentication or bad
4017 sender address), and a DATA command is nevertheless sent (as can happen
4018 with PIPELINING or a stupid MUA), the error message that was given to the
4019 RCPT commands is included in the rejection of the DATA command. This is
4020 intended to be helpful for MUAs that show only the final error to their
4023 PH/42 Another patch from the Sieve maintainer.
4025 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4026 Eximstats - Fixed some broken HTML links and added missing column headers
4028 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4029 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4031 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4033 PH/43 Yet another patch from the Sieve maintainer.
4035 PH/44 I found a way to check for a TCP/IP connection going away before sending
4036 the response to the final '.' that terminates a message, but only in the
4037 case where the client has not sent further data following the '.'
4038 (unfortunately, this is allowed). However, in many cases there won't be
4039 any further data because there won't be any more messages to send. A call
4040 to select() can be used: if it shows that the input is "ready", there is
4041 either input waiting, or the socket has been closed. An attempt to read
4042 the next input character can distinguish the two cases. Previously, Exim
4043 would have sent an OK response which the client would never have see.
4044 This could lead to message repetition. This fix should cure that, at
4045 least in a lot of common cases.
4047 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4048 advertised in response to EHLO.
4054 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4055 fixed by 4.65/MH/01 (is this a record?) are fixed:
4057 (i) An empty string was always treated as zero by the numeric comparison
4058 operators. This behaviour has been restored.
4060 (ii) It is documented that the numeric comparison operators always treat
4061 their arguments as decimal numbers. This was broken in that numbers
4062 starting with 0 were being interpreted as octal.
4064 While fixing these problems I realized that there was another issue that
4065 hadn't been noticed. Values of message_size_limit (both the global option
4066 and the transport option) were treated as octal if they started with 0.
4067 The documentation was vague. These values are now always treated as
4068 decimal, and I will make that clear in the documentation.
4074 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4075 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4078 MH/01 Don't check that the operands of numeric comparison operators are
4079 integers when their expansion is in "skipping" mode (fixes bug
4080 introduced by 4.64-PH/07).
4082 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4083 child addresses, Exim now panics and dies. Previously, because the count
4084 is held in a short int, deliveries were likely to be lost. As such a
4085 large number of recipients for a single message is ridiculous
4086 (performance will be very, very poor), I have chosen to impose a limit
4087 rather than extend the field.
4093 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4094 leftover -K file (the existence of which was triggered by #402).
4095 While we were at it, introduced process PID as part of the -K
4096 filename. This should rule out race conditions when creating
4099 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4100 processing considerably. Previous code took too long for large mails,
4101 triggering a timeout which in turn triggers #401.
4103 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4104 in the DK code in transports.c. sendfile() is not really portable,
4105 hence the _LINUX specificness.
4107 TF/01 In the add_headers option to the mail command in an Exim filter,
4108 there was a bug that Exim would claim a syntax error in any
4109 header after the first one which had an odd number of characters
4112 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4113 callout verification, Exim cached a "reject" for the entire domain. This
4114 is correct for most verifications, but it is not correct for a recipient
4115 verification with use_sender or use_postmaster set, because in that case
4116 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4117 case of MAIL FROM:<> rejection from other early rejections (e.g.
4118 rejection of HELO). When verifying a recipient using a non-null MAIL
4119 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4120 Whatever the result of the callout, the value of the domain cache is
4121 left unchanged (for any other kind of callout, getting as far as trying
4122 RCPT means that the domain itself is ok).
4124 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4127 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4128 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4129 session, the daemon ignores ECONNECT errors and logs others; it now
4130 ignores EPIPE as well.
4132 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4133 (quoted-printable decoding).
4135 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4136 later the small subsequent patch to fix an introduced bug.
4138 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4140 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4142 PH/08 An error is now given if message_size_limit is specified negative.
4144 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4145 to be given (somewhat) arbitrary names.
4147 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4150 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4151 miscellaneous code fixes
4153 PH/10 Added the log_reject_target ACL modifier to specify where to log
4156 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4157 hostname. This is wrong, because it relates to the incoming message (and
4158 probably the interface on which it is arriving) and not to the outgoing
4159 callout (which could be using a different interface). This has been
4160 changed to use the value of the helo_data option from the smtp transport
4161 instead - this is what is used when a message is actually being sent. If
4162 there is no remote transport (possible with a router that sets up host
4163 addresses), $smtp_active_hostname is used.
4165 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4166 tweaks were necessary in order to get it to work (see also 21 below):
4167 (a) The code assumed that strncpy() returns a negative number on buffer
4168 overflow, which isn't the case. Replaced with Exim's string_format()
4170 (b) There were several signed/unsigned issues. I just did the minimum
4171 hacking in of casts. There is scope for a larger refactoring.
4172 (c) The code used strcasecmp() which is not a standard C function.
4173 Replaced with Exim's strcmpic() function.
4174 (d) The code set only $1; it now sets $auth1 as well.
4175 (e) A simple test gave the error "authentication client didn't specify
4176 service in request". It would seem that Dovecot has changed its
4177 interface. Fortunately there's a specification; I followed it and
4178 changed what the client sends and it appears to be working now.
4180 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4183 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4184 address A is aliased to B and C, where B exists and C does not. Without
4185 -v the output is "A verified" because verification stops after a
4186 successful redirection if more than one address is generated. However,
4187 with -v the child addresses are also verified. Exim was outputting "A
4188 failed to verify" and then showing the successful verification for C,
4189 with its parentage. It now outputs "B failed to verify", showing B's
4190 parentage before showing the successful verification of C.
4192 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4193 look up a TXT record in a specific list after matching in a combined
4196 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4197 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4198 they consult the DNS. I had assumed they would set it the way they
4199 wanted; and indeed my experiments on Linux seem to show that in some
4200 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4201 To be on the safe side, however, I have now made the interface to
4202 host_find_byname() similar to host_find_bydns(), with an argument
4203 containing the DNS resolver options. The host_find_byname() function now
4204 sets these options at its start, just as host_find_bydns() does. The smtp
4205 transport options dns_qualify_single and dns_search_parents are passed to
4206 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4207 of host_find_byname() use the default settings of RES_DEFNAMES
4208 (qualify_single) but not RES_DNSRCH (search_parents).
4210 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4211 spool_read_header() do less string testing, by means of a preliminary
4212 switch on the second character of optional "-foo" lines. (This is
4213 overdue, caused by the large number of possibilities that now exist.
4214 Originally there were few.) While I was there, I also converted the
4215 str(n)cmp tests so they don't re-test the leading "-" and the first
4216 character, in the hope this might squeeze out yet more improvement.
4218 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4219 flag allowing group syntax was set by the header_syntax check but not
4220 turned off, possible causing trouble later; (2) The flag was not being
4221 set at all for the header_verify test, causing "group"-style headers to
4222 be rejected. I have now set it in this case, and also caused header_
4223 verify to ignore an empty address taken from a group. While doing this, I
4224 came across some other cases where the code for allowing group syntax
4225 while scanning a header line wasn't quite right (mostly, not resetting
4226 the flag correctly in the right place). These bugs could have caused
4227 trouble for malformed header lines. I hope it is now all correct.
4229 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4230 with the "reply" argument non-NULL. The code, however (which originally
4231 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4232 but it didn't always do it. This confused somebody who was copying the
4233 code for some other use. I have removed all the tests.
4235 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4236 feature that was used to support insecure browsers during the U.S. crypto
4237 embargo. It requires special client support, and Exim is probably the
4238 only MTA that supported it -- and would never use it because real RSA is
4239 always available. This code has been removed, because it had the bad
4240 effect of slowing Exim down by computing (never used) parameters for the
4241 RSA_EXPORT functionality.
4243 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4244 authenticator to fail if there's a tab character in the incoming data
4245 (there should never be unless someone is messing about, as it's supposed
4246 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4247 if the connection is using TLS or if the remote IP is the same as the
4248 local IP, and the "valid-client-cert option" if a client certificate has
4251 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4252 authenticators. This can be used for authorization after authentication
4253 succeeds. (In the case of plaintext, it servers for both authentication
4256 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4257 if any retry times were supplied.
4259 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4260 connection, where there is no client IP address to check. In this
4261 situation, the verify now always succeeds.
4263 PH/25 Applied John Jetmore's -Mset patch.
4265 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4267 PH/27 In a string expansion for a processed (not raw) header when multiple
4268 headers of the same name were present, leading whitespace was being
4269 removed from all of them, but trailing whitespace was being removed only
4270 from the last one. Now trailing whitespace is removed from each header
4271 before concatenation. Completely empty headers in a concatenation (as
4272 before) are ignored.
4274 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4275 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4277 PH/29 [Removed. This was a change that I later backed out, and forgot to
4278 correct the ChangeLog entry (that I had efficiently created) before
4279 committing the later change.]
4281 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4282 address errors (4xx response to RCPT) over the same connection as other
4283 messages routed to the same hosts. Such deliveries are always "forced",
4284 so retry times are not inspected. This resulted in far too many retries
4285 for the affected addresses. The effect occurred only when there were more
4286 hosts than the hosts_max_try setting in the smtp transport when it had
4287 the 4xx errors. Those hosts that it had tried were not added to the list
4288 of hosts for which the message was waiting, so if all were tried, there
4289 was no problem. Two fixes have been applied:
4291 (i) If there are any address or message errors in an SMTP delivery, none
4292 of the hosts (tried or untried) are now added to the list of hosts
4293 for which the message is waiting, so the message should not be a
4294 candidate for sending over the same connection that was used for a
4295 successful delivery of some other message. This seems entirely
4296 reasonable: after all the message is NOT "waiting for some host".
4297 This is so "obvious" that I'm not sure why it wasn't done
4298 previously. Hope I haven't missed anything, but it can't do any
4299 harm, as the worst effect is to miss an optimization.
4301 (ii) If, despite (i), such a delivery is accidentally attempted, the
4302 routing retry time is respected, so at least it doesn't keep
4303 hammering the server.
4305 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4306 in ${readsocket because some servers need this prod.
4308 PH/32 Added some extra debug output when updating a wait-xxx database.
4310 PH/33 The hint "could be header name not terminated by colon", which has been
4311 given for certain expansion errors for a long time, was not being given
4312 for the ${if def:h_colon_omitted{... case.
4314 PH/34 The spec says: "With one important exception, whenever a domain list is
4315 being scanned, $domain contains the subject domain." There was at least
4316 one case where this was not true.
4318 PH/35 The error "getsockname() failed: connection reset by peer" was being
4319 written to the panic log as well as the main log, but it isn't really
4320 panic-worthy as it just means the connection died rather early on. I have
4321 removed the panic log writing for the ECONNRESET error when getsockname()
4324 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4325 runs only) independently of the message's sender address. This meant
4326 that, if the 4xx error was in fact related to the sender, a different
4327 message to the same recipient with a different sender could confuse
4328 things. In particular, this can happen when sending to a greylisting
4329 server, but other circumstances could also provoke similar problems.
4330 I have changed the default so that the retry time for these errors is now
4331 based a combination of the sender and recipient addresses. This change
4332 can be overridden by setting address_retry_include_sender=false in the
4335 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4336 remote server are returned as part of bounce messages. This was not
4337 happening for LMTP over a pipe (the lmtp transport), but now it is the
4338 same for both kinds of LMTP.
4340 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4341 in header lines that were in fact CNAMEs. This is no longer the case.
4343 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4344 and queue runs started by the daemon processed all messages. This has
4345 been fixed so that -R and -S can now usefully be given with -q<time>.
4347 PH/40 Import PCRE release 6.7 (fixes some bugs).
4349 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4351 PH/42 Give an error if -q is specified more than once.
4353 PH/43 Renamed the variables $interface_address and $interface_port as
4354 $received_ip_address and $received_port, to make it clear that these
4355 values apply to message reception, and not to the outgoing interface when
4356 a message is delivered. (The old names remain recognized, of course.)
4358 PH/44 There was no timeout on the connect() call when using a Unix domain
4359 socket in the ${readsocket expansion. There now is.
4361 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4362 be meaningful with "accept".
4364 SC/01 Eximstats V1.43
4365 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4367 SC/02 Eximstats V1.44
4368 Use a glob alias rather than an array ref in the generated
4369 parser. This improves both readability and performance.
4371 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4372 Collect SpamAssassin and rejection statistics.
4373 Don't display local sender or destination tables unless
4374 there is data to show.
4375 Added average volumes into the top table text output.
4377 SC/04 Eximstats V1.46
4378 Collect data on the number of addresses (recipients)
4379 as well as the number of messages.
4381 SC/05 Eximstats V1.47
4382 Added 'Message too big' to the list of mail rejection
4383 reasons (thanks to Marco Gaiarin).
4385 SC/06 Eximstats V1.48
4386 Mainlog lines which have GMT offsets and are too short to
4387 have a flag are now skipped.
4389 SC/07 Eximstats V1.49 (Alain Williams)
4390 Added the -emptyok flag.
4392 SC/08 Eximstats V1.50
4393 Fixes for obtaining the IP address from reject messages.
4395 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4396 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4397 whitespace changes from 4.64-PH/27
4399 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4402 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4403 are found, allow negative numbers in numeric criteria)
4405 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4407 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4408 to match changes made in 4.64-PH/43
4410 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4412 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4413 "domain" is actually an IP address, and force a failure. This locks out
4414 those revolvers/nameservers that support "A-for-A" lookups, in
4415 contravention of the specifications.
4417 PH/48 When a host name was looked up from an IP address, and the subsequent
4418 forward lookup of the name timed out, the host name was left in
4419 $sender_host_name, contrary to the specification.
4421 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4422 restricted to single-key lookups, Exim was not diagnosing an error if
4423 * or *@ was used with a query-style lookup.
4425 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4427 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4428 long ago, but noone interested enough thought of it. Let's just say that
4429 the "1.1" means that there are some new functions that weren't there at
4430 some point in the past.
4432 PH/51 Error processing for expansion failure of helo_data from an smtp
4433 transport during callout processing was broken.
4435 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4436 tested/used via the -bh/-bhc/-bs options.
4438 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4439 bug, fixed in subsequent PCRE releases).
4441 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4442 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4444 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4450 SC/01 Use a glob alias rather than an array ref in eximstats generated
4451 parser. This improves both readability and performance.
4453 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4454 Don't display local sender or destination tables in eximstats unless
4455 there is data to show.
4456 Added average volumes into the eximstats top table text output.
4458 SC/03 Collect data on the number of addresses (recipients) as well
4459 as the number of messages in eximstats.
4461 TF/01 Correct an error in the documentation for the redirect router. Exim
4462 does (usually) call initgroups() when daemonizing.
4464 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4465 with consistent privilege compared to when running as a daemon.
4467 TF/03 Note in the spec that $authenticated_id is not set for local
4468 submissions from trusted users.
4470 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4471 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4473 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4474 by adding some example configuration directives to the default
4475 configuration file. A little bit of work is required to uncomment the
4476 directives and define how usernames and passwords are checked, but
4477 there is now a framework to start from.
4479 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4480 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4481 without this. I don't know how relevant this is to other LDAP libraries.
4483 PH/02 Add the verb name to the "unknown ACL verb" error.
4485 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4487 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4489 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4490 directory not expanded when it should be if an expanded home directory
4491 was set for the address (which is overridden by the transport).
4493 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4496 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4497 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4498 because it is too late at that time, and has no effect.
4500 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4501 security issue with \' (bugzilla #107). I could not use the
4502 PQescapeStringConn() function, because it needs a PGconn value as one of
4505 PH/08 When testing addresses using -bt, indicate those final addresses that
4506 are duplicates that would not cause an additional delivery. At least one
4507 person was confused, thinking that -bt output corresponded to deliveries.
4508 (Suppressing duplicates isn't a good idea as you lose the information
4509 about possibly different redirections that led to the duplicates.)
4511 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4512 systems where poll() doesn't work, in particular OS X.
4514 PH/10 Added more information to debugging output for retry time not reached.
4516 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4517 operations in malware.c.
4519 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4522 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4523 syslog_duplication set false, log lines that would normally be written
4524 both the the main log and to the reject log were not written to syslog at
4527 PH/14 In the default configuration, change the use of "message" in ACL warn
4528 statements to "add_header".
4530 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4531 not followed by a command (e.g. "seen endif").
4533 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4534 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4537 PH/17 Added extra conditions to the default value of delay_warning_condition
4541 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4542 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4543 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4546 The Auto-Submitted: and various List- headers are standardised, whereas I
4547 don't think Precedence: ever was.
4549 PH/18 Refactored debugging code in route_finduser() to show more information,
4550 in particular, the error code if getpwnam() issues one.
4552 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4553 This is apparently needed in addition to the PH/07 change above to avoid
4554 any possible encoding problems.
4556 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4557 but not after initializing Perl.
4559 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4560 output them only if debugging. By default they are written stderr,
4561 apparently, which is not desirable.
4563 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4566 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4569 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4571 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4572 authenticated or an ident call has been made. Suppress the default
4573 values for $authenticated_id and $authenticated_sender (but permit -oMai
4574 and -oMas) when testing with -bh.
4576 PH/24 Re-jigged the order of the tests in the default configuration so that the
4577 tests for valid domains and recipients precede the DNS black list and CSA
4578 tests, on the grounds that those ones are more expensive.
4580 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4581 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4582 This bug exists in every version of Exim that I still have, right back to
4585 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4586 However, an attempt to turn on case-sensitivity in a regex key by
4587 including (?-i) didn't work because the subject string was already
4588 lowercased, and the effects were non-intuitive. It turns out that a
4589 one-line patch can be used to allow (?-i) to work as expected.
4595 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4596 other effects) broke the use of negated acl sub-conditions.
4598 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4601 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4602 "Deny" causes Exim to reject the incoming connection with a 554 error.
4603 Unfortunately, if there is a major crisis, such as a disk failure,
4604 tcp-wrappers gives "deny", whereas what one would like would be some
4605 kind of temporary error. A kludge has been added to help with this.
4606 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4607 554 error is used if errno is still zero or contains ENOENT (which occurs
4608 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4611 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4613 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4614 errors. Otherwise a message that provokes a temporary error (when other
4615 messages do not) can cause a whole host to time out.
4617 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4618 addresses were routed directly to files or pipes from a redirect router.
4619 File deliveries just didn't batch; pipe deliveries might have suffered
4622 PH/06 A failure to get a lock for a hints database would erroneously always say
4623 "Failed to get write lock", even when it was really a read lock.
4625 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4626 of 0600. This has been changed to use the value of the lockfile_mode
4627 option (which defaults to 0600).
4629 PH/08 Applied small patch from the Sieve maintainer.
4631 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4632 folder from quota calculations, a direct delivery into this folder messed
4633 up the contents of the maildirsize file. This was because the regex was
4634 used only to exclude .Trash (or whatever) when the size of the mailbox
4635 was calculated. There was no check that a delivery was happening into an
4636 excluded directory. This bug has been fixed by ignoring all quota
4637 processing for deliveries into excluded directories.
4639 PH/10 Added the maildirfolder_create_regex option to appendfile.
4645 PH/01 The code for finding all the local interface addresses on a FreeBSD
4646 system running IPv6 was broken. This may well have applied to all BSD
4647 systems, as well as to others that have similar system calls. The broken
4648 code found IPv4 interfaces correctly, but gave incorrect values for the
4649 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4650 that it would not match correctly against @[] and not recognize the IPv6
4653 PH/02 The ipliteral router was not recognizing addresses of the form user@
4654 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4656 PH/03 Added disable_ipv6.
4658 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4659 decoded form, because it is most often used to construct To: headers
4660 lines in autoreplies, and the decoded form may well be syntactically
4661 invalid. However, $reply_address has leading white space removed, and all
4662 newlines turned into spaces so that the autoreply transport does not
4665 PH/05 If group was specified without a user on a router, and no group or user
4666 was specified on a transport, the group from the router was ignored.
4668 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4669 for visible compile-time settings that can be used to change these
4670 numbers, for those that want even more. Backwards compatibility with old
4671 spool files has been maintained. However, going back to a previous Exim
4672 release will lost any variables that are in spool files.
4674 PH/07 Two small changes when running in the test harness: increase delay when
4675 passing a TCP/IP connection to a new process, in case the original
4676 process has to generate a bounce, and remove special handling of
4677 127.0.0.2 (sic), which is no longer necessary.
4679 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4680 be the same on different OS.
4682 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4685 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4686 whether --show-vars was specified or not
4688 JJ/02 exipick: Added support for new ACL variable spool format introduced
4691 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4692 syntactically invalid From: or Reply-to: line, and a filter used this to
4693 generate an autoreply, and therefore failed to obtain an address for the
4694 autoreply, Exim could try to deliver to a non-existent relative file
4695 name, causing unrelated and misleading errors. What now happens is that
4696 it logs this as a hard delivery error, but does not attempt to create a
4699 PH/11 The exinext utility has a -C option for testing purposes, but although
4700 the given file was scanned by exinext itself; it wasn't being passed on
4701 when Exim was called.
4703 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4704 an end-of-file indication when reading a command response.
4706 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4707 compiled. In many other places in Exim, IPv6 addresses are always
4708 recognized, so I have changed this. It also means that IPv4 domain
4709 literals of the form [IPV4:n.n.n.n] are now always recognized.
4711 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4712 used if the router is not running as root, for example, when verifying at
4713 ACL time, or when using -bh. The debugging output from this situation was
4714 non-existent - all you got was a failure to exec. I have made two
4717 (a) Failures to set uid/gid, the current directory, or a process leader
4718 in a subprocess such as that created by queryprogram now generate
4719 suitable debugging output when -d is set.
4721 (b) The queryprogram router detects when it is not running as root,
4722 outputs suitable debugging information if -d is set, and then runs
4723 the subprocess without attempting to change uid/gid.
4725 PH/15 Minor change to Makefile for building test_host (undocumented testing
4728 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4729 additional section of a DNS packet that returns MX or SRV records.
4730 Instead, it always explicitly searches for A/AAAA records. This avoids
4731 major problems that occur when a DNS server includes only records of one
4732 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4733 fixed another bug: if SRV records were looked up and the corresponding
4734 address records were *not* found in the additional section, the port
4735 values from the SRV records were lost.
4737 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4738 using the correct key (the original address) when searching the retry
4739 rules in order to find which one to use for generating the retry hint.
4741 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4742 adding the default one. Similarly, if it contains a Reply-To: header, the
4743 errors_reply_to option, if set, is not used.
4745 PH/19 When calculating a retry time, Exim used to measure the "time since
4746 failure" by looking at the "first failed" field in the retry record. Now
4747 it does not use this if it is later than than the arrival time of the
4748 message. Instead it uses the arrival time. This makes for better
4749 behaviour in cases where some deliveries succeed, thus re-setting the
4750 "first failed" field. An example is a quota failure for a huge message
4751 when small messages continue to be delivered. Without this change, the
4752 "time since failure" will always be short, possible causing more frequent
4753 delivery attempts for the huge message than are intended.
4754 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4756 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4757 $1, $2, $3) because the numerical variables can be reset during some
4758 expansion items (e.g. "match"), thereby losing the authentication data.
4760 PH/21 Make -bV show the size of off_t variables so that the test suite can
4761 decide whether to run tests for quotas > 2G.
4763 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4764 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4765 filecount value is greater than 2G or if a quota value is greater than 2G
4766 on a system where the size of off_t is not greater than 4, a panic error
4769 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4770 never match. The debug and -bh output now contains an explicit error
4771 message indicating a malformed IPv4 address or mask.
4773 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4774 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4775 PH/23 above applies.
4777 PH/25 Do not write to syslog when running in the test harness. The only
4778 occasion when this arises is a failure to open the main or panic logs
4779 (for which there is an explicit test).
4781 PH/26 Added the /no_tell option to "control=freeze".
4783 PH/27 If a host name lookup failed very early in a connection, for example, if
4784 the IP address matched host_lookup and the reverse lookup yielded a name
4785 that did not have a forward lookup, an error message of the form "no IP
4786 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4787 could be logged. Now it outputs the IP address instead of "NULL".
4789 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4790 allows the sender and the authenticated sender to be set when
4791 submitting a message from within Exim. Since child_open_exim() is
4792 documented for local_scan(), the new function should be too.
4794 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4795 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4796 results in an empty string is now treated as unset.
4798 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4800 PH/31 Added sender_verify_fail logging option.
4802 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4803 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4804 tidied the source and removed it altogether.
4806 PH/33 When a queue run was abandoned because the load average was too high, a
4807 log line was always written; now it is written only if the queue_run log
4808 selector is set. In addition, the log line for abandonment now contains
4809 information about the queue run such as the pid. This is always present
4810 in "start" and "stop" lines but was omitted from the "abandon" line.
4812 PH/34 Omit spaces between a header name and the colon in the error message that
4813 is given when verify = headers_syntax fails (if there are lots of them,
4814 the message gets confusing).
4816 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4817 names, as there are now some PTR records that contain slashes. This check
4818 is only to protect against broken name servers that fall over on strange
4819 characters, so the fact that it applies to all lookups doesn't matter.
4821 PH/36 Now that the new test suite is complete, we can remove some of the
4822 special code in Exim that was needed for the old test suite. For example,
4823 sorting DNS records because real resolvers return them in an arbitrary
4824 order. The new test suite's fake resolver always returns records in the
4827 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4828 bounces) except when queue_only is set, to avoid logging races between
4829 the different processes.
4831 PH/38 Panic-die if .include specifies a non-absolute path.
4833 PH/39 A tweak to the "H" retry rule from its user.
4835 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4836 a label. They prevented compilation on older perls.
4838 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4839 a warning to be raised on newish perls.
4841 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4842 on queue. Changes to match documented behaviour of showing count of
4843 messages matching specified criteria.
4845 PH/40 Changed the default ident timeout from 30s to 5s.
4847 PH/41 Added support for the use of login_cap features, on those BSD systems
4848 that have them, for controlling the resources used by pipe deliveries.
4850 PH/42 The content-scanning code uses fopen() to create files in which to put
4851 message data. Previously it was not paying any attention to the mode of
4852 the files. Exim runs with umask(0) because the rest of the code creates
4853 files with open(), and sets the required mode explicitly. Thus, these
4854 files were ending up world-writeable. This was not a big issue, because,
4855 being within the spool directory, they were not world-accessible. I have
4856 created a function called modefopen, which takes an additional mode
4857 argument. It sets umask(777), creates the file, chmods it to the required
4858 mode, then resets the umask. All the relevant calls to fopen() in the
4859 content scanning code have been changed to use this function.
4861 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4862 to 24 hours. This avoids potential overflow problems when processing G
4863 and H retry rules. I suspect nobody ever tinkers with this value.
4865 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4867 PH/45 When the plaintext authenticator is running as a client, the server's
4868 challenges are checked to ensure they are valid base64 strings. By
4869 default, the authentication attempt is cancelled if an invalid string is
4870 received. Setting client_ignore_invalid_base64 true ignores these errors.
4871 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4872 they are received. Thus, the responses can be made to depend on the
4873 challenges. If an invalid string is ignored, an empty string is placed in
4876 PH/46 Messages that are created by the autoreply transport now contains a
4877 References: header, in accordance with RFCs 2822 and 3834.
4879 PH/47 Added authenticated_sender_force to the smtp transport.
4881 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4883 PH/49 Installed latest patch from the Sieve maintainer.
4885 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4886 was also set, the appendfile transport was unnecessarily scanning a
4887 directory of message files (e.g. for maildir delivery) to find the count
4888 of files (along with the size), even though it did not need this
4889 information. It now does the scan only if it needs to find either the
4890 size of the count of files.
4892 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4894 PH/52 Two bugs concerned with error handling when the smtp transport is
4897 (i) Exim was not creating retry information for temporary errors given
4898 for individual recipients after the DATA command when the smtp transport
4899 was used in LMTP mode. This meant that they could be retried too
4900 frequently, and not timed out correctly.
4902 (ii) Exim was setting the flag that allows error details to be returned
4903 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4904 recipients that were returned after the DATA command.
4906 PH/53 This is related to PH/52, but is more general: for any failing address,
4907 when detailed error information was permitted to be returned to the
4908 sender, but the error was temporary, then after the final timeout, only
4909 "retry timeout exceeded" was returned. Now it returns the full error as
4910 well as "retry timeout exceeded".
4912 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4913 do this, and (what is worse) MTAs that accept it.
4915 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4916 will now be deprecated.
4918 PH/56 New os.c-cygwin from the Cygwin maintainer.
4920 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4921 formats (previously only available in exim formats via -bpr, -bpru,
4922 and -bpra. Now also available in native and exiqgrep formats)
4924 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4925 with very large, slow to parse queues
4927 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4929 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4931 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4932 responses to authentication challenges, though it was showing the
4933 challenges; (ii) I've removed the CR characters from the debug output for
4936 PH/58 Allow for the insertion of a newline as well as a space when a string
4937 is turned into more than one encoded-word during RFC 2047 encoding. The
4938 Sieve code now uses this.
4940 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4941 data_4xx, lost_connection, tls_required.
4943 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4944 message was being sent as an SMTP response.
4946 PH/61 Add -l and -k options to exicyclog.
4948 PH/62 When verifying, if an address was redirected to one new address, so that
4949 verification continued, and the new address failed or deferred after
4950 having set something in $address_data, the value of $address_data was not
4951 passed back to the ACL. This was different to the case when no
4952 redirection occurred. The value is now passed back in both cases.
4954 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4955 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4956 use login_cap.h, so on its own it isn't the distinguishing feature. The
4957 new name refers directly to the setclassresources() function.
4959 PH/65 Added configuration files for NetBSD3.
4961 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4963 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4964 is preferred over IPv4.
4966 PH/68 The bounce_return_message and bounce_return_body options were not being
4967 honoured for bounces generated during the reception of non-SMTP messages.
4968 In particular, this applied to messages rejected by the ACL. This bug has
4969 been fixed. However, if bounce_return_message is true and bounce_return_
4970 body is false, the headers that are returned for a non-SMTP message
4971 include only those that have been read before the error was detected.
4972 (In the case of an ACL rejection, they have all been read.)
4974 PH/69 The HTML version of the specification is now built in a directory called
4975 spec_html instead of spec.html, because the latter looks like a path with
4976 a MIME-type, and this confuses some software.
4978 PH/70 Catch two compiler warnings in sieve.c.
4980 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4981 function verify_get_ident() calls ip_connect() to connect a socket, but
4982 if the "connect()" function timed out, ip_connect() used to close the
4983 socket. However, verify_get_ident() also closes the socket later, and in
4984 between Exim writes to the log, which may get opened at this point. When
4985 the socket was closed in ip_connect(), the log could get the same file
4986 descriptor number as the socket. This naturally causes chaos. The fix is
4987 not to close the socket in ip_connect(); the socket should be closed by
4988 the function that creates it. There was only one place in the code where
4989 this was missing, in the iplookup router, which I don't think anybody now
4990 uses, but I've fixed it anyway.
4992 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4993 well as to direct DNS lookups. Otherwise the handling of names in host
4994 lists is inconsistent and therefore confusing.
5000 PH/01 Two changes to the default runtime configuration:
5002 (1) Move the checks for relay_from_hosts and authenticated clients from
5003 after to before the (commented out) DNS black list checks.
5005 (2) Add control=submission to the relay_from_hosts and authenticated
5006 clients checks, on the grounds that messages accepted by these
5007 statements are most likely to be submissions.
5009 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5011 (1) Generate an error if the third argument for the ${prvs expansion is
5014 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5017 (3) Reset the variables that are obtained from the first argument of
5018 ${prvscheck and used in the second argument before leaving the code,
5019 because their memory is reclaimed, so using them afterwards may do
5022 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5023 one (it's much easier than Tom thought :-).
5025 (5) Because of (4), we can now allow for the use of $prvscheck_result
5026 inside the third argument.
5028 PH/03 For some reason, the default setting of PATH when running a command from
5029 a pipe transport was just "/usr/bin". I have changed it to
5032 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5033 anything to be listed in the output from -bV.
5035 PH/05 When a filter generated an autoreply, the entire To: header line was
5036 quoted in the delivery log line, like this:
5038 => >A.N.Other <ano@some.domain> <original@ddress> ...
5040 This has been changed so that it extracts the operative address. There
5041 may be more than one such address. If so, they are comma-separated, like
5044 => >ano@some.domain,ona@other.domain <original@ddress> ...
5046 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5047 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5048 not being looked up in the rDNS to get a host name, Exim was showing the
5049 IP address twice in Received: lines, even though the IP addresses were
5050 identical. For example:
5052 Received: from [1.2.3.4] (helo=[1.2.3.4])
5054 However, if the real host name was known, it was omitting the HELO data
5055 if it matched the actual IP address. This has been tidied up so that it
5056 doesn't show the same IP address twice.
5058 PH/07 When both +timestamp and +memory debugging was on, the value given by
5059 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5060 called by the debug printing, thereby overwriting the timestamp buffer.
5061 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5063 PH/08 When the original message was included in an autoreply transport, it
5064 always said "this is a copy of the message, including all the headers",
5065 even if body_only or headers_only was set. It now gives an appropriate
5068 PH/09 Applied a patch from the Sieve maintainer which:
5070 o fixes some comments
5071 o adds the (disabled) notify extension core
5072 o adds some debug output for the result of if/elsif tests
5073 o points to the current vacation draft in the documentation
5074 and documents the missing references header update
5078 o fixes a bug in processing the envelope test (when testing
5079 multiple envelope elements, the last element determined the
5082 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5083 Electronic Mail") by including:
5085 Auto-submitted: auto-generated
5087 in the messages that it generates (bounce messages and others, such as
5088 warnings). In the case of bounce messages for non-SMTP messages, there was
5089 also a typo: it was using "Auto_submitted" (underscore instead of
5090 hyphen). Since every message generated by Exim is necessarily in response
5091 to another message, thes have all been changed to:
5093 Auto-Submitted: auto-replied
5095 in accordance with these statements in the RFC:
5097 The auto-replied keyword:
5099 - SHOULD be used on messages sent in direct response to another
5100 message by an automatic process,
5102 - MUST NOT be used on manually-generated messages,
5104 - MAY be used on Delivery Status Notifications (DSNs) and Message
5105 Disposition Notifications (MDNs),
5107 - MUST NOT be used on messages generated by automatic or periodic
5108 processes, except for messages which are automatic responses to
5111 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5112 to the default Received: header definition.
5114 PH/12 Added log selector acl_warn_skipped (default on).
5116 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5117 variables because (a) they are in the wrong storage pool and (b) even if
5118 they were copied, it wouldn't work properly because of the caching.
5120 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5121 checking when decoding. Apparently there are clients that generate
5122 overlong encoded strings. Why am I not surprised?
5124 PH/15 If the first argument of "${if match_address" was not empty, but did not
5125 contain an "@" character, Exim crashed. Now it writes a panic log message
5126 and treats the condition as false.
5128 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5130 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5131 extension "envelope-auth". The code is finished and in agreement with
5132 other implementations, but there is no documentation so far and in fact,
5133 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5134 not changing the active code.
5136 Print executed "if" and "elsif" statements when debugging is used. This
5137 helps a great deal to understand what a filter does.
5139 Document more things not specified clearly in RFC3028. I had all this
5140 sorted out, when out of a sudden new issues came to my mind. Oops."
5142 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5145 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5146 canonical form (as documented). However, after a host name lookup from
5147 the IP address, check_host() was doing a simple string comparison with
5148 addresses acquired from the DNS when checking that the found name did
5149 have the original IP as one of its addresses. Since any found IPv6
5150 addresses are likely to be in abbreviated form, the comparison could
5151 fail. Luckily, there already exists a function for doing the comparison
5152 by converting both addresses to binary, so now that is used instead of
5153 the text comparison.
5155 PH/20 There was another similar case to PH/19, when a complete host name was
5156 given in a host list; looking up its IP address could give an abbreviated
5157 form, whereas the current host's name might or might not be abbreviated.
5158 The same fix has been applied.
5164 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5165 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5168 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5169 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5171 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5173 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5174 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5175 Also, for "Auto-Submitted", treat anything other than "no" as
5176 non-personal, in accordance with RFC 3834. (Previously it treated
5177 anything starting "auto-" as non-personal.)
5179 TF/01 The control=submission/name=... option had a problem with syntax
5180 errors if the name included a slash character. The /name= option
5181 now slurps the rest of the string, so it can include any characters
5182 but it must come last in the list of options (after /sender_retain
5185 PH/05 Some modifications to the interface to the fake nameserver for the new
5193 TK/01 Added the "success_on_redirect" address verification option. See
5194 NewStuff for rationale and an example.
5196 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5198 PH/02 Patch to exigrep to allow it to work on syslog lines.
5200 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5201 fread() to skip over the body file's header line, because in Cygwin the
5202 header line is locked and is inaccessible.
5204 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5205 co-exist for some time) to make it clear that it is the Exim ID that is
5206 referenced, not the Message-ID: header line.
5208 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5209 string_format() function, because snprintf() does not exist on all
5212 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5213 ${stat: expansion item.
5215 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5216 protocol synchronization error", to keep the pedants happy.
5218 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5219 well as for IRIX systems, when gcc is being used. See the host.c source
5222 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5224 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5227 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5228 transport and to the smtp transport in LMTP mode.
5230 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5232 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5233 run a filter in a subprocess. This could lead to confusion in subsequent
5234 lookups in the parent process. There should also be a search_tidyup() at
5235 the end of the subprocess.
5237 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5238 only if the host matched helo_try_verify_hosts, which caused the
5239 verification to occur when the EHLO/HELO command was issued. The ACL just
5240 tested the remembered result. Now, if a previous verification attempt has
5241 not happened, "verify = helo" does it there and then.
5243 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5245 TK/03 Fix log output including CR from clamd.
5247 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5248 exist provoked a memory error which could cause a segfault.
5250 PH/15 Installed PCRE 6.2
5252 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5254 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5255 of the problem. Specifically, suggested +O2 rather than +O1 for the
5258 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5260 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5261 record, the port was not correctly logged when the outgoing_port log
5262 selector was set (it logged the transort's default port).
5264 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5265 fallback_hosts, and "hosts" in the smtp transport.
5267 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5268 host errors such as "Connection refused".
5270 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5271 authentication with radiusclient 0.4.9:
5273 - Error returned from rc_read_config was caught wrongly
5274 - Username/password not passed on to radius server due to wrong length.
5276 The presumption is that some radiusclient API changes for 4.51/PH/17
5277 were not taken care of correctly. The code is still untested by me (my
5278 Linux distribution still has 0.3.2 of radiusclient), but it was
5279 contributed by a Radius user.
5281 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5282 expanding the "port" option of the smtp transport.
5284 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5285 while reading a MIME header. Thanks to Tom Hughes for a patch.
5287 PH/24 Include config.h inside local_scan.h so that configuration settings are
5290 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5291 that in an ACL for RCPT (for example), you can examine exactly what was
5294 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5295 commands, but it was not correctly comparing the address with the actual
5296 client host address. Thus, it would show the EHLO address in Received:
5297 header lines when this was not necessary.
5299 PH/27 Added the % operator to ${eval:}.
5301 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5302 it should be ignoring failures (because with -C, for example, it has lost
5303 privilege). It wasn't ignoring creation failures other than "already
5306 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5309 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5310 because an input line was too long, either on its own, or by virtue of
5311 too many continuations, the temporary file was not being removed, and the
5312 return code was incorrect.
5314 PH/31 Missing "BOOL" in function definition in filtertest.c.
5316 PH/32 Applied Sieve patches from the maintainer.
5318 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5320 PH/33 Added "verify = not_blind".
5322 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5323 Local/Makefile (with some defaults set). These are used in built scripts
5324 such as exicyclog, but they have never been used in the exim_install
5325 script (though there are many overriding facilities there). I have
5326 arranged that the exim_install script now takes note of these two
5329 PH/35 Installed configuration files for Dragonfly.
5331 PH/36 When a locally submitted message by a trusted user did not contain a
5332 From: header, and the sender address was obtained from -f or from an SMTP
5333 MAIL command, and the trusted user did not use -F to supply a sender
5334 name, $originator_name was incorrectly used when constructing a From:
5335 header. Furthermore, $originator_name was used for submission mode
5336 messages from external hosts without From: headers in a similar way,
5337 which is clearly wrong.
5339 PH/37 Added control=suppress_local_fixups.
5341 PH/38 When log_selector = +received_sender was set, and the addition of the
5342 sender made the log line's construction buffer exactly full, or one byte
5343 less than full, an overflow happened when the terminating "\n" was
5346 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5347 when the result of a list match is failure because a DNS lookup failed.
5349 PH/40 RM_COMMAND is now used in the building process.
5351 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5352 the "build-* directories that it finds.
5354 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5355 address in a domain literal was a prefix of an interface address.
5357 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5358 when verifying a sender address, unless rewrite_headers is false.
5360 PH/44 Wrote a long comment about why errors_to addresses are verified as
5361 recipients, not senders.
5363 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5364 the ratelimit ACL was added.
5366 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5368 PH/46 Added extra information about PostgreSQL errors to the error string.
5370 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5371 avoiding the need to install special zones in a real server. This is
5372 backwards compatible; if it can't find the fake resolver, it drops back.
5373 Thus, both old and new test suites can be run.
5375 TF/02 Added util/ratelimit.pl
5377 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5378 clock is set back in time.
5380 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5381 Candler <B.Candler@pobox.com>.
5383 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5384 OK for addresses that are the result of redirections.
5386 PH/48 A number of further additions for the benefit of the new test suite,
5387 including a fake gethostbyname() that interfaces to the fake DNS resolver
5390 TF/06 The fix for widen_domains has also been applied to qualify_single and
5391 search_parents which are the other dnslookup options that can cause
5394 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5397 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5399 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5400 service" if it encountered an MX record with an empty target hostname.
5401 The message is now "an MX or SRV record indicated no SMTP service".
5403 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5404 if the DNS of the sending site is misconfigured. This is quite a
5405 common situation. This change restores the behaviour of treating a
5406 helo verification defer as a failure.
5408 PH/51 If self=fail was set on a router, the bounce message did not include the
5409 actual error message.
5415 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5417 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5418 command itself ended in error, the underlying message about the transport
5419 filter timeout was being overwritten with the pipe command error. Now the
5420 underlying error message should be appended to the second error message.
5422 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5424 PH/02 Reduce the amount of output that "make" produces by default. Full output
5425 can still be requested.
5427 PH/03 The warning log line about a condition test deferring for a "warn" verb
5428 was being output only once per connection, rather than after each
5429 occurrence (because it was using the same function as for successful
5430 "warn" verbs). This seems wrong, so I have changed it.
5432 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5433 it should not have, which might have caused a crash in the right
5434 circumstances, but probably never did.
5436 PH/04 Installed a modified version of Tony Finch's patch to make submission
5437 mode fix the return path as well as the Sender: header line, and to
5438 add a /name= option so that you can make the user's friendly name appear
5441 TF/03 Added the control = fakedefer ACL modifier.
5443 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5444 Mark Lowes for thorough testing.
5446 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5448 TK/03 Merged latest SRS patch from Miles Wilton.
5450 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5451 with the definition in sysexits.h (which is #included earlier).
5452 Fortunately, Exim does not actually use EX_OK. The code used to try to
5453 preserve the sysexits.h value, by assuming that macro definitions were
5454 scanned for macro replacements. I have been disabused of this notion,
5455 so now the code just undefines EX_OK before #including unistd.h.
5457 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5458 in the smtp transport. When a block could not be written in a single
5459 write() function, the timeout was being re-applied to each part-write.
5460 This seems wrong - if the receiver was accepting one byte at a time it
5461 would take for ever. The timeout is now adjusted when this happens. It
5462 doesn't have to be particularly precise.
5464 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5465 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5467 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5468 without a domain if the check to <postmaster@domain> fails.
5470 SC/01 Eximstats: added -xls and the ability to specify output files
5471 (patch written by Frank Heydlauf).
5473 SC/02 Eximstats: use FileHandles for outputting results.
5475 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5477 SC/04 Eximstats: fixed display of large numbers with -nvr option
5479 SC/05 Eximstats: fixed merging of reports with empty tables.
5481 SC/06 Eximstats: added the -include_original_destination flag
5483 SC/07 Eximstats: removed tabs and trailing whitespace.
5485 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5487 TK/06 MBOX spool code: Add real "From " MBOX separator line
5488 so the .eml file is really in mbox format (even though
5489 most programs do not really care). Patch from Alex Miller.
5491 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5492 The latter is generated from $received_to and is only set if the
5493 message has one envelope recipient. SA can use these headers,
5494 obviously out-of-the-box. Patch from Alex Miller.
5496 PH/08 The ${def test on a variable was returning false if the variable's
5497 value was "0", contrary to what the specification has always said!
5498 The result should be true unless the variable is empty.
5500 PH/09 The syntax error of a character other than { following "${if
5501 def:variable_name" (after optional whitespace) was not being diagnosed.
5502 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5503 accidental colon was present, for example, could give incorrect results.
5505 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5506 result is used (not including appendfile, where other changes are about
5509 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5510 This involved changing a lot of size variables from int to off_t. It
5511 should work with maildirs and everything.
5513 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5514 spamd dying while we are connected to it.
5516 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5519 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5520 function for BDB 4.3.
5522 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5524 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5525 the sources for PCRE have been re-arranged and more files are now
5528 PH/15 The code I had for printing potentially long long variables in PH/11
5529 above was not the best (it lost precision). The length of off_t variables
5530 is now inspected at build time, and an appropriate printing format (%ld
5531 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5532 to be "long long int" or "long int". This is needed for the internal
5533 formatting function string_vformat().
5535 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5536 the configuration file to be ":syslog", then the script "guesses" where
5537 the logs files are, rather than using the compiled in default. In our
5538 case the guess is not the same as the compiled default, so the script
5539 suddenly stopped working when I started to use syslog. The patch checks
5540 to see if log_file_path is "". If so, it attempts to read it from exim
5541 with no configuration file to get the compiled in version, before it
5542 falls back to the previous guessing code."
5544 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5545 implementing BATV in an Exim configuration. See NewStuff for the gory
5548 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5549 Makefile that are specific to HP-UX.
5551 PH/18 If the "use_postmaster" option was set for a recipient callout together
5552 with the "random" option, the postmaster address was used as the MAIL
5553 FROM address for the random test, but not for the subsequent recipient
5554 test. It is now used for both.
5556 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5557 patch removes a few documentation additions to RFC 3028, because the
5558 latest draft now contains them. It adds the new en;ascii-case comparator
5559 and a new error check for 8bit text in MIME parts. Comparator and
5560 require names are now matched exactly. I enabled the subaddress
5561 extension, but it is not well tested yet (read: it works for me)."
5563 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5564 rework some of the code of TK/09 above to avoid the hardwired use of
5565 "%lld" and "long long". Replaced the call to snprintf() with a call to
5568 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5569 records point to non-existent hosts", "retry timeout exceeded", and
5570 "retry time not reached for any host after a long failure period".
5572 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5573 experimental DomainKeys support:
5575 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5576 (2) On an error such as an illegally used "control", the wrong name for
5577 the control was given.
5579 These problems did NOT occur unless DomainKeys support was compiled.
5581 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5583 PH/24 Added ${if match_ip condition.
5585 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5586 databases so that it will be absolutely obvious if a crash occurs in the
5587 DB library. This is a regular occurrence (often caused by mis-matched
5590 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5591 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5592 picked up on a user's system that detects such things. There doesn't seem
5593 to be a gcc warning option for this - only an attribute that has to be
5594 put on the function's prototype. It seems that in Fedora Core 4 they have
5595 set this on a number of new functions. No doubt there will be more in due
5598 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5599 specify a transport. However, if an address that was verified by such a
5600 router was the subject of a callout, Exim crashed because it tried to
5601 read the rcpt_include_affixes from the non-existent transport. Now it
5602 just assumes that the setting of that option is false. This bug was
5603 introduced by 4.51/PH/31.
5605 PH/28 Changed -d+all to exclude +memory, because that information is very
5606 rarely of interest, but it makes the output a lot bigger. People tend to
5607 do -d+all out of habit.
5609 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5610 code in os-type was giving problems when libc.so lives in lib64, like on
5613 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5614 aren't the modern standard, and it seems that some systems' include files
5615 don't always have them. Exim was already checking for some of the newer
5616 ones like T_AAAA, and defining it itself. I've added checks for all the
5617 record types that Exim uses.
5619 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5620 not automatically generating a new one, as it is supposed to. This
5621 prevented TLS from working. If the file did exist, but contained invalid
5622 data, a new version was generated, as expected. It was only the case of a
5623 non-existent file that was broken.
5625 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5626 with a change in libdomainkeys > 0.64.
5628 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5629 from DNS. If the selector record carries the flag, it now has
5630 precedence over the domain-wide flag.
5632 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5634 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5635 the use of a port name that isn't defined in /etc/services) occurred, the
5636 message was deferred as in a normal delivery, and thus remained on the
5637 spool, instead of being failed because of the mua_wrapper setting. This
5638 is now fixed, and I tidied up some of the mua_wrapper messages at the
5641 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5642 the messages in a hash of arrays rather than using individual hashes.
5643 This is a bit cleaner and results in dramatic memory savings, albeit
5644 at a slight CPU cost.
5646 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5647 as requested by Marc Sherman.
5649 SC/10 Eximstats: added histograms for user specified patterns as requested
5652 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5654 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5655 fopen() in the content-scanning modules that did not already have it.
5661 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5662 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5664 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5666 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5668 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5669 to test Sieve filters that use "vacation".
5671 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5672 that changes the way the GnuTLS parameters are stored in the cache file.
5673 The new format can be generated externally. For backward compatibility,
5674 if the data in the cache doesn't make sense, Exim assumes it has read an
5675 old-format file, and it generates new data and writes a new file. This
5676 means that you can't go back to an older release without removing the
5679 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5680 work if there are any delivery delays because "one_time" forces the
5681 parent to be marked "delivered", so its unseen clone is never tried
5682 again. For this reason, Exim now forbids the simultaneous setting of
5685 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5686 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5687 entry if you want to know the details. The fix, however, neglected to
5688 consider the case when local delivery batching is involved. The test for
5689 "previously delivered" was not happening when checking to see if an
5690 address could be batched with a previous (undelivered) one; under
5691 certain circumstances this could lead to multiple deliveries to the same
5694 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5695 in its include files, and this causes problems building Exim.
5697 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5698 header_syntax) but Exim was just ignoring anything given after a slash.
5699 In particular, this caused confusion with an attempt to use "verify =
5700 reverse_host_lookup/defer_ok". An error is now given when options are
5701 supplied for verify items that do not have them. (Maybe reverse_host_
5702 lookup should have a defer_ok option, but that's a different point.)
5704 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5705 defined by RFC 821) to 2048, because there were problems with some AUTH
5706 commands, and RFC 1869 says the size should be increased for extended
5707 SMTP commands that take arguments.
5709 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5712 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5713 "unknown" error; now it says that the functionality isn't in the binary.
5715 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5716 an address' error message when a string expansion fails (syntax or
5717 whatever). Otherwise the password may appear in the log. Following change
5718 PH/42 below, there is no longer a chance of it appearing in a bounce
5721 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5723 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5724 of its addresses was ever tried. (Bugzilla bug #2.)
5726 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5727 the result incorrectly in the debug output. (It correctly added a newline
5728 to what was transported.)
5730 TF/01 Added $received_time.
5732 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5733 commented out examples of how to interface to a virus scanner and to
5734 SpamAssassin. Also added commented examples of av_scanner and
5735 spamd_address settings.
5737 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5738 and controls are allowed in which ACLs. There were a couple of minor
5739 errors. Some of the entries in the conditions table (which is a table of
5740 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5741 negation of where the condition IS allowed.
5743 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5745 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5746 header file does not have a version number, so I've had to invent a new
5747 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5748 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5749 radiusclient), but it was contributed by a Radius user.
5751 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5752 files or directories, for OpenSSL.
5754 PH/19 When an Exim process that is running as root has to create an Exim log
5755 file, it does so in a subprocess that runs as exim:exim so as to get the
5756 ownership right at creation (otherwise, other Exim processes might see
5757 the file with the wrong ownership). There was no test for failure of this
5758 fork() call, which would lead to the process getting stuck as it waited
5759 for a non-existent subprocess. Forks do occasionally fail when resources
5760 run out. I reviewed all the other calls to fork(); they all seem to check
5763 PH/20 When checking for unexpected SMTP input at connect time (before writing
5764 the banner), Exim was not dealing correctly with a non-positive return
5765 from the read() function. If the client had disconnected by this time,
5766 the result was a log entry for a synchronization error with an empty
5767 string after "input=" when read() returned zero. If read() returned -1
5768 (an event I could not check), uninitialized data bytes were printed.
5769 There were reports of junk text (parts of files, etc) appearing after
5772 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5774 PH/22 Added support for macro redefinition, and (re)definition in between
5775 driver and ACL definitions.
5777 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5778 forgetting to use the resulting value; it was using the unexpanded value.
5780 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5781 hadn't been configured. The fix is from Juergen Kreileder, who
5782 understands it better than I do:
5784 "Here's what I see happening with three configured cyrus_sasl
5785 authenticators configured (plain, login, cram-md5):
5787 On startup auth_cyrus_sasl_init() gets called for each of these.
5788 This means three calls to sasl_listmech() without a specified mech_list.
5789 => SASL tests which mechs of all available mechs actually work
5790 => three warnings about OTP not working
5791 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5793 With the patch, sasl_listmech() also gets called three times. But now
5794 SASL's mech_list option is set to the server_mech specified in the the
5795 authenticator. Or in other words, the answer from sasl_listmech()
5796 gets limited to just the mech you're testing for (which is different
5798 => the return list contains just 'plain' or 'login', 'cram-md5' or
5799 nothing depending on the value of ob->server_mech.
5801 I've just tested the patch: Authentication still works fine,
5802 unavailable mechs specified in the exim configuration are still
5803 caught, and the auth.log warnings about OTP are gone."
5805 PH/25 When debugging is enabled, the contents of the command line are added
5806 to the debugging output, even when log_selector=+arguments is not
5809 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5810 answer is "GNU", and only if the return is "GNU/something" is the answer
5813 PH/27 $acl_verify_message is now set immediately after the failure of a
5814 verification in an ACL, and so is available in subsequent modifiers. In
5815 particular, the message can be preserved by coding like this:
5817 warn !verify = sender
5818 set acl_m0 = $acl_verify_message
5820 Previously, $acl_verify_message was set only while expanding "message"
5821 and "log_message" when a very denied access.
5823 PH/28 Modified OS/os.c-Linux with
5825 -#ifndef OS_LOAD_AVERAGE
5826 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5828 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5829 nomenclature these days.)
5831 PH/29 Installed patch from the Sieve maintainer that adds the options
5832 sieve_useraddress and sieve_subaddress to the redirect router.
5834 PH/30 In these circumstances:
5835 . Two addresses routed to the same list of hosts;
5836 . First host does not offer TLS;
5837 . First host accepts first address;
5838 . First host gives temporary error to second address;
5839 . Second host offers TLS and a TLS session is established;
5840 . Second host accepts second address.
5841 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5842 and peerdn, if requested) that were in fact used only for the second
5845 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5846 attention to any local part prefix or suffix that was matched by the
5847 router that accepted the address. It now behaves in the same way as it
5848 does for delivery: the affixes are removed from the local part unless
5849 rcpt_include_affixes is set on the transport.
5851 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5852 timeout during the DATA phase of an incoming message.
5854 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5855 applied a patch sanctioned by the Sieve maintainer.
5857 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5858 the uid or gid is negative. A case of a negative gid caused this to be
5859 noticed. The fix allows for either to be negative.
5861 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5862 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5865 PH/36 The change PH/12 above was broken. Fixed it.
5867 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5868 the grounds that routing the same address twice would always produce the
5869 same answer. This might have been true once, but it is certainly no
5870 longer true now. Routing a child address may depend on the previous
5871 routing that produced that child. Some complicated redirection strategies
5872 went wrong when messages had multiple recipients, and made Exim's
5873 behaviour dependent on the order in which the addresses were given.
5875 I have moved the duplicate checking until after the routing is complete.
5876 Exim scans the addresses that are assigned to local and remote
5877 transports, and removes any duplicates. This means that more work will be
5878 done, as duplicates will always all be routed, but duplicates are
5879 presumably rare, so I don't expect this is of any significance.
5881 For deliveries to pipes, files, and autoreplies, the duplicate checking
5882 still happens during the routing process, since they are not going to be
5885 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5886 It corrects a timeout issue with spamd. This is Ian's comment: "The
5887 background is that sometimes spamd either never reads data from a
5888 connection it has accepted, or it never writes response data. The exiscan
5889 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5890 blindly assumes that writes won't block so it may never time out."
5892 PH/39 Allow G after quota size as well as K and M.
5894 PH/40 The value set for $authenticated_id in an authenticator may not contain
5895 binary zeroes or newlines because the value is written to log lines and
5896 to spool files. There was no check on this. Now the value is run through
5897 the string_printing() function so that such characters are converted to
5898 printable escape sequences.
5900 PH/41 $message_linecount is a new variable that contains the total number of
5901 lines in the message. Compare $body_linecount, which is the count for the
5904 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5905 bounce and delay warning messages, except in certain special cases, which
5908 (a) An SMTP error message from a remote host;
5909 (b) A message specified in a :fail: redirection;
5910 (c) A message specified in a "fail" command in a system filter;
5911 (d) A message specified in a FAIL return from the queryprogram router;
5912 (e) A message specified by the cannot_route_message router option.
5914 In these cases only, Exim does include the error details in bounce and
5915 warning messages. There are also a few cases where bland messages such
5916 as "unrouteable address" or "local delivery error" are given.
5918 PH/43 $value is now also set for the "else" part of a ${run expansion.
5920 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5921 being worked on, but at least Exim now implements the latest version to
5924 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5925 process to complete was treated as a delivery failure, a timeout while
5926 writing the message to the pipe was logged, but erroneously treated as a
5927 successful delivery. Such timeouts include transport filter timeouts. For
5928 consistency with the overall process timeout, these timeouts are now
5929 treated as errors, giving rise to delivery failures by default. However,
5930 there is now a new Boolean option for the pipe transport called
5931 timeout_defer, which, if set TRUE, converts the failures into defers for
5932 both kinds of timeout. A transport filter timeout is now identified in
5935 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5936 systems where "make" and "gmake" are different, calling "gmake" at top
5937 level broke things. I've arranged for the value of $(MAKE) to be passed
5938 from the Makefile to this script so that it can call the same version of
5942 A note about Exim versions 4.44 and 4.50
5943 ----------------------------------------
5945 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5946 changes of various kinds. As a consequence, a big documentation update was
5947 needed. This delayed the release for rather longer than seemed good, especially
5948 in the light of a couple of (minor) security issues. Therefore, the changes
5949 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5950 release. So 4.44 and 4.50 are in effect two different branches that both start
5953 I have left the 4.50 change log unchanged; it contains all the changes since
5954 4.43. The change log for 4.44 is below; many of its items are identical to
5955 those for 4.50. This seems to be the most sensible way to preserve the
5956 historical information.
5962 1. Minor wording change to the doc/README.SIEVE file.
5964 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5965 computation of the current number of files was incorrect.
5967 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5968 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5971 4. Give more explanation in the error message when the command for a transport
5972 filter fails to execute.
5974 5. There are several places where Exim runs a non-Exim command in a
5975 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5976 was being done only for the command run by the queryprogram router. It is
5977 now done for all such subprocesses. The other cases are: ${run, transport
5978 filters, and the commands run by the lmtp and pipe transports.
5980 6. Added CONFIGURE_GROUP build-time option.
5982 7. Some older OS have a limit of 256 on the maximum number of file
5983 descriptors. Exim was using setrlimit() to set 1000 as a large value
5984 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5985 systems. I've change it so that if it can't get 1000, it tries for 256.
5987 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5988 was an oversight, and furthermore, ever since the addition of extra
5989 controls (e.g. 4.43/32), the checks on when to allow different forms of
5990 "control" were broken. There should now be diagnostics for all cases when a
5991 control that does not make sense is encountered.
5993 9. Added the /retain_sender option to "control=submission".
5995 10. $recipients is now available in the predata ACL (oversight).
5997 11. Tidy the search cache before the fork to do a delivery from a message
5998 received from the command line. Otherwise the child will trigger a lookup
5999 failure and thereby defer the delivery if it tries to use (for example) a
6000 cached ldap connection that the parent has called unbind on.
6002 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6003 of $address_data from the recipient verification was clobbered by the
6004 sender verification.
6006 13. The value of address_data from a sender verification is now available in
6007 $sender_address_data in subsequent conditions in the ACL statement.
6009 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6011 15. Added a new option "connect=<time>" to callout options, to set a different
6014 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6015 was its contents. (It was OK if the option was not defined at all.)
6017 17. A "Completed" log line is now written for messages that are removed from
6018 the spool by the -Mrm option.
6020 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6021 information about exactly what failed.
6023 19. Added -dd to debug only the daemon process.
6025 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6026 handles timeouts, both on the server side and network timeouts. Renamed the
6027 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6029 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6030 It is now set to "smtps".
6032 22. $host_address is now set to the target address during the checking of
6033 ignore_target_hosts.
6035 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6036 being passed; this would have caused $sender_host_name to have been used if
6037 matching the list had actually called for a host name (not very likely,
6038 since this list is usually IP addresses). A host name is now passed as
6041 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6042 code that specifies a non-restarting handler (typically sigaction() in
6043 modern systems) in an attempt to fix a rare and obscure crash bug.
6045 25. Narrowed the window for a race in the daemon that could cause it to ignore
6046 SIGCHLD signals. This is not a major problem, because they are used only to
6047 wake it up if nothing else does.
6049 26. A malformed maildirsize file could cause Exim to calculate negative values
6050 for the mailbox size or file count. Odd effects could occur as a result.
6051 The maildirsize information is now recalculated if the size or filecount
6054 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6055 support for a long time. Removed HAVE_SYS_VFS_H.
6057 28. Installed the latest version of exipick from John Jetmore.
6059 29. In an address list, if the pattern was not a regular expression, an empty
6060 subject address (from a bounce message) matched only if the pattern was an
6061 empty string. Non-empty patterns were not even tested. This was the wrong
6062 because it is perfectly reasonable to use an empty address as part of a
6063 database query. An empty address is now tested by patterns that are
6064 lookups. However, all the other forms of pattern expect the subject to
6065 contain a local part and a domain, and therefore, for them, an empty
6066 address still always fails if the pattern is not itself empty.
6068 30. Exim went into a mad DNS loop when attempting to do a callout where the
6069 host was specified on an smtp transport, and looking it up yielded more
6070 than one IP address.
6072 31. Re-factored the code for checking spool and log partition space into a
6073 function that finds that data and another that does the check. The former
6074 is then used to implement four new variables: $spool_space, $log_space,
6075 $spool_inodes, and $log_inodes.
6077 32. The RFC2047 encoding function was originally intended for short strings
6078 such as real names; it was not keeping to the 75-character limit for
6079 encoded words that the RFC imposes. It now respects the limit, and
6080 generates multiple encoded words if necessary. To be on the safe side, I
6081 have increased the buffer size for the ${rfc2047: expansion operator from
6084 33. It is now permitted to omit both strings after an "if" condition; if the
6085 condition is true, the result is "true". As before, when the second string
6086 is omitted, a false condition yields an empty string. This makes it less
6087 cumbersome to write custom ACL and router conditions.
6089 34. Failure to deliver a bounce message always caused it to be frozen, even if
6090 there was an errors_to setting on the router. The errors_to setting is now
6093 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6094 canonical form (fully expanded) before being placed in
6095 $sender_host_address.
6097 36. The table in the code that translates DNS record types into text (T_A to
6098 "A" for instance) was missing entries for NS and CNAME. It is just possible
6099 that this could have caused confusion if both these types were looked up
6100 for the same domain, because the text type is used as part of Exim's
6101 per-process caching. But the chance of anyone hitting this buglet seems
6104 37. The dnsdb lookup has been extended in a number of ways.
6106 (1) There is a new type, "zns", which walks up the domain tree until it
6107 finds some nameserver records. It should be used with care.
6109 (2) There is a new type, "mxh", which is like "mx" except that it returns
6110 just the host names, not the priorities.
6112 (3) It is now possible to give a list of domains (or IP addresses) to be
6113 looked up. The behaviour when one of the lookups defers can be
6114 controlled by a keyword.
6116 (4) It is now possible to specify the separator character for use when
6117 multiple records are returned.
6119 38. The dnslists ACL condition has been extended: it is now possible to supply
6120 a list of IP addresses and/or domains to be looked up in a particular DNS
6123 39. Added log_selector=+queue_time_overall.
6125 40. When running the queue in the test harness, wait just a tad after forking a
6126 delivery process, to get repeatability of debugging output.
6128 41. Include certificate and key file names in error message when GnuTLS fails
6129 to set them up, because the GnuTLS error message doesn't include the name
6130 of the failing file when there is a problem reading it.
6132 42. Allow both -bf and -bF in the same test run.
6134 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6136 44. The "Exiscan patch" is now merged into the mainline Exim source.
6138 45. Sometimes the final signoff response after QUIT could fail to get
6139 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6140 < 0 before doing a fflush(). This bug looks as though it goes back to the
6141 introduction of TLS in release 3.20, but "sometimes" must have been rare
6142 because the tests only now provoked it.
6144 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6145 (this can affect the format of dates).
6147 47. exim_tidydb, when checking for the continued existence of a message for
6148 which it has found a message-specific retry record, was not finding
6149 messages that were in split spool directories. Consequently, it was
6150 deleting retry records that should have stayed in existence.
6152 48. Steve fixed some bugs in eximstats.
6154 49. The SPA authentication driver was not abandoning authentication and moving
6155 on to the next authenticator when an expansion was forced to fail,
6156 contradicting the general specification for all authenticators. Instead it
6157 was generating a temporary error. It now behaves as specified.
6159 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6160 (the order specifies the preference for clients). The order is now AES256,
6161 AES128, 3DES, ARCFOUR128.
6163 51. Small patch to Sieve code - explicitly set From: when generating an
6166 52. Exim crashed if a remote delivery caused a very long error message to be
6167 recorded - for instance if somebody sent an entire SpamAssassin report back
6168 as a large number of 550 error lines. This bug was coincidentally fixed by
6169 increasing the size of one of Exim's internal buffers (big_buffer) that
6170 happened as part of the Exiscan merge. However, to be on the safe side, I
6171 have made the code more robust (and fixed the comments that describe what
6174 53. Now that there can be additional text after "Completed" in log lines (if
6175 the queue_time_overall log selector is set), a one-byte patch to exigrep
6176 was needed to allow it to recognize "Completed" as not the last thing in
6179 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6180 patch that reportedly fixes this has been added. I am not expert enough to
6181 create a test for it. This is what the patch creator wrote:
6183 "I found a little strange behaviour of ldap code when working with
6184 Windows 2003 AD Domain, where users was placed in more than one
6185 Organization Units. When I tried to give exim partial DN, the exit code
6186 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6187 But simultaneously result of request was absolutely normal ldap result,
6188 so I produce this patch..."
6190 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6191 REFERENCE, so I have modified the code to exclude the patch when that macro
6194 55. Some experimental protocols are using DNS PTR records for new purposes. The
6195 keys for these records are domain names, not reversed IP addresses. The
6196 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6197 leaves it alone. Component reversal etc. now happens only for IP addresses.
6200 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6202 57. Double the size of the debug message buffer (to 2048) so that more of very
6203 long debug lines gets shown.
6205 58. The exicyclog utility now does better if the number of log files to keep
6206 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6208 59. Two changes related to the smtp_active_hostname option:
6210 (1) $smtp_active_hostname is now available as a variable.
6211 (2) The default for smtp_banner uses $smtp_active_hostname instead
6212 of $primary_hostname.
6214 60. The host_aton() function is supposed to be passed a string that is known
6215 to be a valid IP address. However, in the case of IPv6 addresses, it was
6216 not checking this. This is a hostage to fortune. Exim now panics and dies
6217 if the condition is not met. A case was found where this could be provoked
6218 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6219 components; fortuitously, this particular loophole had already been fixed
6220 by change 4.50/55 above.
6222 If there are any other similar loopholes, the new check in host_aton()
6223 itself should stop them being exploited. The report I received stated that
6224 data on the command line could provoke the exploit when Exim was running as
6225 exim, but did not say which command line option was involved. All I could
6226 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6227 running as the user.
6230 61. There was a buffer overflow vulnerability in the SPA authentication code
6231 (which came originally from the Samba project). I have added a test to the
6232 spa_base64_to_bits() function which I hope fixes it.
6235 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6236 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6238 63. The daemon start-up calls getloadavg() while still root for those OS that
6239 need the first call to be done as root, but it missed one case: when
6240 deliver_queue_load_max is set with deliver_drop_privilege. This is
6241 necessary for the benefit of the queue runner, because there is no re-exec
6242 when deliver_drop_privilege is set.
6244 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6245 This has been fixed.
6247 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6248 was in use, was not putting the data itself into the right store pool;
6249 consequently, it could be overwritten for a subsequent message in the same
6250 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6253 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6255 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6256 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6257 the function were treating the return as a boolean value, which happened to
6258 work because 0=false and not-0=true, but is not correct code.
6260 68. The host_aton() function was not handling scoped IPv6 addresses (those
6261 with, for example, "%eth0" on the end) correctly.
6263 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6264 negated items (that is, ~something) in unsigned ints. Some compilers
6265 apparently mutter when there is no cast.
6267 70. If an address verification called from an ACL failed, and did not produce a
6268 user-specific message (i.e. there was only a "system" message), nothing was
6269 put in $acl_verify_message. In this situation, it now puts the system
6272 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6273 session; change 4.31/43 added the unwanted input to the log line - except
6274 that it did not do this in the start of session case. It now does.
6276 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6277 This is wrong and can cause the other end to generate a synchronization
6278 error if it is another Exim or anything else that does the synchronization
6279 check. A QUIT command is no longer sent after a timeout.
6281 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6282 during host lookups.
6284 74. The defer_ok option of callout verification was not working if it was used
6285 when verifying addresses in header lines, that is, for this case:
6287 verify = header_sender/callout=defer_ok
6289 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6290 those file descriptors could be used for SMTP connections. If anything
6291 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6292 could be sent to the SMTP client, causing chaos. The daemon now opens
6293 stdin, stdout, and stderr to /dev/null when it puts itself into the
6296 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6297 log by default. The user can override this with suitable Perl magic.
6299 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6300 the log message when discard triggers, was not working for the DATA ACL or
6301 for the non-SMTP ACL.
6303 78. Error message wording change in sieve.c.
6305 79. If smtp_accept_max_per_host was set, the number of connections could be
6306 restricted to fewer than expected, because the daemon was trying to set up
6307 a new connection before checking whether the processes handling previous
6308 connections had finished. The check for completed processes is now done
6309 earlier. On busy systems, this bug wouldn't be noticed because something
6310 else would have woken the daemon, and it would have reaped the completed
6313 80. If a message was submitted locally by a user whose login name contained one
6314 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6315 It caused a spool format error. I have fixed the spool reading code. A
6316 related problem was that the "from" clause in the Received: line became
6317 illegal because of the space(s). It is now covered by ${quote_local_part.
6319 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6322 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6324 83. Patch from Sieve maintainer to fix unterminated string problem in
6325 "vacation" handling.
6327 84. Some minor changes to the Linux configuration files to help with other
6328 OS variants using glibc.
6330 85. One more patch for Sieve to update vacation handling to latest spec.
6333 ----------------------------------------------------
6334 See the note above about the 4.44 and 4.50 releases.
6335 ----------------------------------------------------
6341 1. Change 4.43/35 introduced a bug that caused file counts to be
6342 incorrectly computed when quota_filecount was set in an appendfile
6345 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6346 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6349 3. Give more explanation in the error message when the command for a transport
6350 filter fails to execute.
6352 4. There are several places where Exim runs a non-Exim command in a
6353 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6354 was being done only for the command run by the queryprogram router. It is
6355 now done for all such subprocesses. The other cases are: ${run, transport
6356 filters, and the commands run by the lmtp and pipe transports.
6358 5. Some older OS have a limit of 256 on the maximum number of file
6359 descriptors. Exim was using setrlimit() to set 1000 as a large value
6360 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6361 systems. I've change it so that if it can't get 1000, it tries for 256.
6363 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6364 was an oversight, and furthermore, ever since the addition of extra
6365 controls (e.g. 4.43/32), the checks on when to allow different forms of
6366 "control" were broken. There should now be diagnostics for all cases when a
6367 control that does not make sense is encountered.
6369 7. $recipients is now available in the predata ACL (oversight).
6371 8. Tidy the search cache before the fork to do a delivery from a message
6372 received from the command line. Otherwise the child will trigger a lookup
6373 failure and thereby defer the delivery if it tries to use (for example) a
6374 cached ldap connection that the parent has called unbind on.
6376 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6377 of $address_data from the recipient verification was clobbered by the
6378 sender verification.
6380 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6381 was its contents. (It was OK if the option was not defined at all.)
6383 11. A "Completed" log line is now written for messages that are removed from
6384 the spool by the -Mrm option.
6386 12. $host_address is now set to the target address during the checking of
6387 ignore_target_hosts.
6389 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6390 being passed; this would have caused $sender_host_name to have been used if
6391 matching the list had actually called for a host name (not very likely,
6392 since this list is usually IP addresses). A host name is now passed as
6395 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6396 code that specifies a non-restarting handler (typically sigaction() in
6397 modern systems) in an attempt to fix a rare and obscure crash bug.
6399 15. Narrowed the window for a race in the daemon that could cause it to ignore
6400 SIGCHLD signals. This is not a major problem, because they are used only to
6401 wake it up if nothing else does.
6403 16. A malformed maildirsize file could cause Exim to calculate negative values
6404 for the mailbox size or file count. Odd effects could occur as a result.
6405 The maildirsize information is now recalculated if the size or filecount
6408 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6409 support for a long time. Removed HAVE_SYS_VFS_H.
6411 18. Updated exipick to current release from John Jetmore.
6413 19. Allow an empty sender to be matched against a lookup in an address list.
6414 Previously the only cases considered were a regular expression, or an
6417 20. Exim went into a mad DNS lookup loop when doing a callout where the
6418 host was specified on the transport, if the DNS lookup yielded more than
6421 21. The RFC2047 encoding function was originally intended for short strings
6422 such as real names; it was not keeping to the 75-character limit for
6423 encoded words that the RFC imposes. It now respects the limit, and
6424 generates multiple encoded words if necessary. To be on the safe side, I
6425 have increased the buffer size for the ${rfc2047: expansion operator from
6428 22. Failure to deliver a bounce message always caused it to be frozen, even if
6429 there was an errors_to setting on the router. The errors_to setting is now
6432 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6433 canonical form (fully expanded) before being placed in
6434 $sender_host_address.
6436 24. Updated eximstats to version 1.33
6438 25. Include certificate and key file names in error message when GnuTLS fails
6439 to set them up, because the GnuTLS error message doesn't include the name
6440 of the failing file when there is a problem reading it.
6442 26. Expand error message when OpenSSL has problems setting up cert/key files.
6445 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6446 (this can affect the format of dates).
6448 28. exim_tidydb, when checking for the continued existence of a message for
6449 which it has found a message-specific retry record, was not finding
6450 messages that were in split spool directories. Consequently, it was
6451 deleting retry records that should have stayed in existence.
6453 29. eximstats updated to version 1.35
6454 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6455 1.35 - bugfix such that pie charts by volume are generated correctly
6457 30. The SPA authentication driver was not abandoning authentication and moving
6458 on to the next authenticator when an expansion was forced to fail,
6459 contradicting the general specification for all authenticators. Instead it
6460 was generating a temporary error. It now behaves as specified.
6462 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6463 (the order specifies the preference for clients). The order is now AES256,
6464 AES128, 3DES, ARCFOUR128.
6466 31. Small patch to Sieve code - explicitly set From: when generating an
6469 32. Exim crashed if a remote delivery caused a very long error message to be
6470 recorded - for instance if somebody sent an entire SpamAssassin report back
6471 as a large number of 550 error lines. This bug was coincidentally fixed by
6472 increasing the size of one of Exim's internal buffers (big_buffer) that
6473 happened as part of the Exiscan merge. However, to be on the safe side, I
6474 have made the code more robust (and fixed the comments that describe what
6477 33. Some experimental protocols are using DNS PTR records for new purposes. The
6478 keys for these records are domain names, not reversed IP addresses. The
6479 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6480 leaves it alone. Component reversal etc. now happens only for IP addresses.
6483 34. The host_aton() function is supposed to be passed a string that is known
6484 to be a valid IP address. However, in the case of IPv6 addresses, it was
6485 not checking this. This is a hostage to fortune. Exim now panics and dies
6486 if the condition is not met. A case was found where this could be provoked
6487 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6488 components; fortuitously, this particular loophole had already been fixed
6489 by change 4.50/55 or 4.44/33 above.
6491 If there are any other similar loopholes, the new check in host_aton()
6492 itself should stop them being exploited. The report I received stated that
6493 data on the command line could provoke the exploit when Exim was running as
6494 exim, but did not say which command line option was involved. All I could
6495 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6496 running as the user.
6499 35. There was a buffer overflow vulnerability in the SPA authentication code
6500 (which came originally from the Samba project). I have added a test to the
6501 spa_base64_to_bits() function which I hope fixes it.
6504 36. The daemon start-up calls getloadavg() while still root for those OS that
6505 need the first call to be done as root, but it missed one case: when
6506 deliver_queue_load_max is set with deliver_drop_privilege. This is
6507 necessary for the benefit of the queue runner, because there is no re-exec
6508 when deliver_drop_privilege is set.
6510 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6511 was in use, was not putting the data itself into the right store pool;
6512 consequently, it could be overwritten for a subsequent message in the same
6513 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6516 38. Sometimes the final signoff response after QUIT could fail to get
6517 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6518 < 0 before doing a fflush(). This bug looks as though it goes back to the
6519 introduction of TLS in release 3.20, but "sometimes" must have been rare
6520 because the tests only now provoked it.
6526 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6527 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6528 Now it returns an integer. A number of calls were still expecting a T/F
6529 return. Fortuitously, in all cases, the tests worked in OK situations,
6530 which is the norm. However, things would have gone wrong on any write
6531 failures on the smtp file descriptor. This function is used when sending
6532 messages over SMTP and also when doing verify callouts.
6534 2. When Exim is called to do synchronous delivery of a locally submitted
6535 message (the -odf or -odi options), it no longer closes stderr before doing
6538 3. Implemented the mua_wrapper option.
6540 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6542 5. Implemented the functions header_remove(), header_testname(),
6543 header_add_at_position(), and receive_remove_recipient(), and exported them
6546 6. If an ACL "warn" statement specified the addition of headers, Exim already
6547 inserted X-ACL-Warn: at the start if there was no header name. However, it
6548 was not making this test for the second and subsequent header lines if
6549 there were newlines in the string. This meant that an invalid header could
6550 be inserted if Exim was badly configured.
6552 7. Allow an ACL "warn" statement to add header lines at the start or after all
6553 the Received: headers, as well as at the end.
6555 8. Added the rcpt_4xx retry error code.
6557 9. Added postmaster_mailfrom=xxx to callout verification option.
6559 10. Added mailfrom=xxxx to the callout verification option, for verify=
6562 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6563 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6565 12. Inserted some casts to stop certain compilers warning when using pointer
6566 differences as field lengths or precisions in printf-type calls (mostly
6567 affecting debugging statements).
6569 13. Added optional readline() support for -be (dynamically loaded).
6571 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6572 same clock tick as a message's arrival, so that its received time was the
6573 same as the "first fail" time on the retry record, and that message
6574 remained on the queue past the ultimate address timeout, every queue runner
6575 would try a delivery (because it was past the ultimate address timeout) but
6576 after another failure, the ultimate address timeout, which should have then
6577 bounced the address, did not kick in. This was a "< instead of <=" error;
6578 in most cases the first failure would have been in the next clock tick
6579 after the received time, and all would be well.
6581 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6582 being recognized when the domain list was tested by the match_domain
6583 condition in an expansion string.
6585 16. Added the ${str2b64: operator.
6587 17. Exim was always calling setrlimit() to set a large limit for the number of
6588 processes, without checking whether the existing limit was already
6589 adequate. (It did check for the limit on file descriptors.) Furthermore,
6590 errors from getrlimit() and setrlimit() were being ignored. Now they are
6591 logged to the main and panic logs, but Exim does carry on, to try to do its
6592 job under whatever limits there are.
6594 18. Imported PCRE 5.0.
6596 19. Trivial typo in log message " temporarily refused connection" (the leading
6599 20. If the log selector return_path_on_delivery was set and an address was
6600 redirected to /dev/null, the delivery process crashed because it assumed
6601 that a return path would always be set for a "successful" delivery. In this
6602 case, the whole delivery is bypassed as an optimization, and therefore no
6605 21. Internal re-arrangement: the function for sending a challenge and reading
6606 a response while authentication was assuming a zero-terminated challenge
6607 string. It's now changed to take a pointer and a length, to allow for
6608 binary data in such strings.
6610 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6612 23. Exim was not respecting finduser_retries when seeking the login of the
6613 uid under which it was called; it was always trying 10 times. (The default
6614 setting of finduser_retries is zero.) Also, it was sleeping after the final
6615 failure, which is pointless.
6617 24. Implemented tls_on_connect_ports.
6619 25. Implemented acl_smtp_predata.
6621 26. If the domain in control=submission is set empty, Exim assumes that the
6622 authenticated id is a complete email address when it generates From: or
6623 Sender: header lines.
6625 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6626 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6627 chown and chgrp in /bin and hostname in /usr/bin.
6629 28. Exim was keeping the "process log" file open after each use, just as it
6630 does for the main log. This opens the possibility of it remaining open for
6631 long periods when the USR1 signal hits a daemon. Occasional processlog
6632 errors were reported, that could have been caused by this. Anyway, it seems
6633 much more sensible not to leave this file open at all, so that is what now
6636 29. The long-running daemon process does not normally write to the log once it
6637 has entered its main loop, and it closes the log before doing so. This is
6638 so that log files can straightforwardly be renamed and moved. However,
6639 there are a couple of unusual error situations where the daemon does write
6640 log entries, and I had neglected to close the log afterwards.
6642 30. The text of an SMTP error response that was received during a remote
6643 delivery was being truncated at 512 bytes. This is too short for some of
6644 the long messages that one sometimes sees. I've increased the limit to
6647 31. It is now possible to make retry rules that apply only when a message has a
6648 specific sender, in particular, an empty sender.
6650 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6651 it possible to be selective about when SMTP synchronization is enforced.
6653 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6655 32. Implemented hosts_connection_nolog.
6657 33. Added an ACL for QUIT.
6659 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6662 35. Added mailbox_size and mailbox_filecount to appendfile.
6664 36. Added control = no_multiline_responses to ACLs.
6666 37. There was a bug in the logic of the code that waits for the clock to tick
6667 in the case where the clock went backwards by a substantial amount such
6668 that the microsecond fraction of "now" was more than the microsecond
6669 fraction of "then" (but the whole seconds number was less).
6671 38. Added support for the libradius Radius client library this is found on
6672 FreeBSD (previously only the radiusclient library was supported).
6678 1. When certain lookups returned multiple values in the form name=value, the
6679 quoting of the values was not always being done properly. Specifically:
6680 (a) If the value started with a double quote, but contained no whitespace,
6682 (b) If the value contained whitespace other than a space character (i.e.
6683 tabs or newlines or carriage returns) it was not quoted.
6684 This fix has been applied to the mysql and pgsql lookups by writing a
6685 separate quoting function and calling it from the lookup code. The fix
6686 should probably also be applied to nisplus, ibase and oracle lookups, but
6687 since I cannot test any of those, I have not disturbed their existing code.
6689 2. A hit in the callout cache for a specific address caused a log line with no
6690 reason for rejecting RCPT. Now it says "Previous (cached) callout
6691 verification failure".
6693 3. There was an off-by-one bug in the queryprogram router. An over-long
6694 return line was truncated at 256 instead of 255 characters, thereby
6695 overflowing its buffer with the terminating zero. As well as fixing this, I
6696 have increased the buffer size to 1024 (and made a note to document this).
6698 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6699 when Exim is waiting for an SMTP response from a remote server, Exim
6700 restarts its select() call on the socket, thereby resetting its timeout.
6701 This is not a problem when such interrupts are rare. Somebody set up a cron
6702 job to run exiwhat every 2 minutes, which is less than the normal select()
6703 timeout (5 or 10 minutes). This meant that the select() timeout never
6704 kicked in because it was always reset. I have fixed this by comparing the
6705 time when an interrupt arrives with the time at the start of the first call
6706 to select(). If more time than the timeout has elapsed, the interrupt is
6707 treated as a timeout.
6709 5. Some internal re-factoring in preparation for the addition of Sieve
6710 extensions (by MH). In particular, the "personal" test is moved to a
6711 separate function, and given an option for scanning Cc: and Bcc: (which is
6712 not set for Exim filters).
6714 6. When Exim created an email address using the login of the caller as the
6715 local part (e.g. when creating a From: or Sender: header line), it was not
6716 quoting the local part when it contained special characters such as @.
6718 7. Installed new OpenBSD configuration files.
6720 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6721 try to make them clearer.
6723 9. Callout options, other than the timeout value, were being ignored when
6724 verifying sender addresses in header lines. For example, when using
6726 verify = header_sender/callout=no_cache
6728 the cache was (incorrectly) being used.
6730 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6731 only the Cygwin environment.
6733 11. When return_path_on_delivery was set as a log selector, if different remote
6734 addresses in the same message used different return paths and parallel
6735 remote delivery occurred, the wrong values would sometimes be logged.
6736 (Whenever a remote delivery process finished, the return path value from
6737 the most recently started remote delivery process was logged.)
6739 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6740 lines when AUTH and/or TLS are in use. This is the "received protocol"
6741 field. Exim used to use "asmtp" for authenticated SMTP, without any
6742 indication (in the protocol name) for TLS use. Now it follows the RFC and
6743 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6744 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6745 names appear in log lines as well as in Received: header lines.
6747 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6748 extensions, and comparison tests, and to fix some bugs.
6750 14. Changes to the "personal" filter test:
6752 (1) The test was buggy in that it was just doing the equivalent of
6753 "contains" tests on header lines. For example, if a user's address was
6754 anne@some.where, the "personal" test would incorrectly be true for
6756 To: susanne@some.where
6758 This test is now done by extracting each address from the header in turn,
6759 and checking the entire address. Other tests that are part of "personal"
6760 are now done using regular expressions (for example, to check local parts
6761 of addresses in From: header lines).
6763 (2) The list of non-personal local parts in From: addresses has been
6764 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6765 taken from the Sieve specification recommendations.
6767 (3) If the message contains any header line starting with "List-" it is
6768 treated as non-personal.
6770 (4) The test for "circular" in the Subject: header line has been removed
6771 because it now seems ill-conceived.
6773 15. Minor typos in src/EDITME comments corrected.
6775 16. Installed latest exipick from John Jetmore.
6777 17. If headers_add on a router specified a text string that was too long for
6778 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6779 of string_sprintf() is now avoided.
6781 18. $message_body_size was not set (it was always zero) when running the DATA
6782 ACL and the local_scan() function.
6784 19. For the "mail" command in an Exim filter, no default was being set for
6785 the once_repeat time, causing a random time value to be used if "once" was
6786 specified. (If the value happened to be <= 0, no repeat happened.) The
6787 default is now 0s, meaning "never repeat". The "vacation" command was OK
6788 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6789 (I found it when inspecting the code).
6791 20. There is now an overall timeout for performing a callout verification. It
6792 defaults to 4 times the callout timeout, which applies to individual SMTP
6793 commands during the callout. The overall timeout applies when there is more
6794 than one host that can be tried. The timeout is checked before trying the
6795 next host. This prevents very long delays if there are a large number of
6796 hosts and all are timing out (e.g. when the network connections are timing
6797 out). The value of the overall timeout can be changed by specifying an
6798 additional sub-option for "callout", called "maxwait". For example:
6800 verify = sender/callout=5s,maxwait=20s
6802 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6803 to the end before writing, but this should make it even safer).
6805 22. Exim was forgetting that it had advertised PIPELINING for the second and
6806 subsequent messages on an SMTP connection. It was also not resetting its
6807 memory on STARTTLS and an internal HELO.
6809 23. When Exim logs an SMTP synchronization error within a session, it now
6810 records whether PIPELINING has been advertised or not.
6812 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6813 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6814 rather than long int.
6816 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6818 26. Added the never_mail option to autoreply.
6824 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6825 crash if the getsockname() call failed; this can happen if a connection is
6826 closed very soon after it is established. The problem was simply in the
6827 order in which certain operations were done, causing Exim to try to write
6828 to the SMTP stream before it had set up the file descriptor. The bug has
6829 been fixed by making things happen in the correct order.
6835 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6836 before the connection was closed, thus losing the rejection response.
6838 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6839 some early Solaris releases, but causes trouble in current releases where
6840 socklen_t is defined.
6842 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6845 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6848 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6849 that releasing the top of it at the end releases what was used for sub-
6850 expansions (unless the block got too big). However, discard this block if
6851 the first thing is a variable or header, so that we can use its block when
6852 it is dynamic (useful for very large $message_headers, for example).
6854 6. Lookups now cache *every* query, not just the most recent. A new, separate
6855 store pool is used for this. It can be recovered when all lookup caches are
6856 flushed. Lookups now release memory at the end of their result strings.
6857 This has involved some general refactoring of the lookup sources.
6859 7. Some code has been added to the store_xxx() functions to reduce the amount
6860 of flapping under certain conditions.
6862 8. log_incoming_interface used to affect only the <= reception log lines. Now
6863 it causes the local interface and port to be added to several more SMTP log
6864 lines, for example "SMTP connection from", and rejection lines.
6866 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6868 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6870 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6871 could be overwritten at the end of the current message (or the start of a
6872 new message if it was set in a HELO ACL). The value is now preserved for
6873 the duration of the SMTP connection.
6875 12. If a transport had a headers_rewrite setting, and a matching header line
6876 contained an unqualified address, that address was qualified, even if it
6877 did not match any rewriting rules. The underlying bug was that the values
6878 of the flags that permit the existence of unqualified sender and recipient
6879 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6880 non-local messages, and by -bnq for local messages) were not being
6881 preserved with the message after it was received.
6883 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6884 "next input=" as part of the text comprising the host identity instead of
6885 the correct text. The code was using the same buffer for two different
6886 strings. However, depending on which order the printing function evaluated
6887 its arguments, the bug did not always show up. Under Linux, for example, my
6888 test suite worked just fine.
6890 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6891 doesn't work with some older versions of Perl. It has been changed to "my",
6892 which in any case is probably the better facility to use.
6894 15. A really picky compiler found some instances of statements for creating
6895 error messages that either had too many or two few arguments for the format
6898 16. The size of the buffer for calls to the DNS resolver has been increased
6899 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6900 for addresses that have a lot of PTR records. This alleviates a problem; it
6901 does not fully solve it.
6903 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6904 buffer now truncates the list and logs the incident, which is the same
6905 action as happens when Exim is looking up a host name and its aliases.
6906 Previously in this situation something unpredictable would happen;
6907 sometimes it was "internal error: store_reset failed".
6909 18. If a server dropped the connection unexpectedly when an Exim client was
6910 using GnuTLS and trying to read a response, the client delivery process
6911 crashed while trying to generate an error log message.
6913 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6914 string, for example:
6916 warn message = H1: something\nH2: something
6918 the text was added as a single header line from Exim's point of view
6919 though it ended up OK in the delivered message. However, searching for the
6920 second and subsequent header lines using $h_h2: did not work. This has been
6921 fixed. Similarly, if a system filter added multiple headers in this way,
6922 the routers could not see them.
6924 20. Expanded the error message when iplsearch is called with an invalid key to
6925 suggest using net-iplsearch in a host list.
6927 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6928 ACLs are no longer actually imposed (and a message to that effect is
6931 22. If a "gecos" field in a passwd entry contained escaped characters, in
6932 particular, if it contained a \" sequence, Exim got it wrong when building
6933 a From: or a Sender: header from that name. A second bug also caused
6934 incorrect handling when an unquoted " was present following a character
6935 that needed quoting.
6937 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6938 was not being matched caselessly.
6940 24. Arranged for all hyphens in the exim.8 source to be escaped with
6943 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6944 properly. Recipient callout cache records were still being keyed to include
6945 the sender, even when use_sender was set false. This led to far more
6946 callouts that were necessary. The sender is no longer included in the key
6947 when use_sender is false.
6949 26. Added "control = submission" modifier to ACLs.
6951 27. Added the ${base62d: operator to decode base 62 numbers.
6953 28. dnsdb lookups can now access SRV records.
6955 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6956 the configuration file.
6958 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6959 (-v) mode. This makes the output for a verbose queue run more intelligible.
6961 31. Added a use_postmaster feature to recipient callouts.
6963 32. Added the $body_zerocount variable, containing the number of binary zero
6964 bytes in the message body.
6966 33. The time of last modification of the "new" subdirectory is now used as the
6967 "mailbox time last read" when there is a quota error for a maildir
6970 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6972 35. Added +ignore_unknown as a special item in host lists.
6974 36. Code for decoding IPv6 addresses in host lists is now included, even if
6975 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6976 address was recognized as an IP address, but was then not correctly decoded
6977 into binary, causing unexpected and incorrect effects when compared with
6984 1. Very minor rewording of debugging text in manualroute to say "list of
6985 hosts" instead of "hostlist".
6987 2. If verify=header_syntax was set, and a header line with an unqualified
6988 address (no domain) and a large number of spaces between the end of the
6989 name and the colon was received, the reception process suffered a buffer
6990 overflow, and (when I tested it) crashed. This was caused by some obsolete
6991 code that should have been removed. The fix is to remove it!
6993 3. When running in the test harness, delay a bit after writing a bounce
6994 message to get a bit more predictability in the log output.
6996 4. Added a call to search_tidyup() just before forking a reception process. In
6997 theory, someone could use a lookup in the expansion of smtp_accept_max_
6998 per_host which, without the tidyup, could leave open a database connection.
7000 5. Added the variables $recipient_data and $sender_data which get set from a
7001 lookup success in an ACL "recipients" or "senders" condition, or a router
7002 "senders" option, similar to $domain_data and $local_part_data.
7004 6. Moved the writing of debug_print from before to after the "senders" test
7007 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7008 problems for message scanning, either using a data ACL, or using
7009 local_scan() because the Received: header was not generated till after they
7010 were called (in order to set the time as the time of reception completion).
7011 I have revised the way this works. The header is now generated after the
7012 body is received, but before the ACL or local_scan() are called. After they
7013 are run, the timestamp in the header is updated.
7019 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7020 before starting a queue runner without re-exec. This happened only when
7021 deliver_drop_privilege was set or when the Exim user was set to root. The
7022 effect of the bug was that timeouts during subsequent deliveries caused
7023 crashes instead of being properly handled. The handler is now left at its
7024 default (and expected) setting.
7026 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7027 message, again when deliver_drop_privilege is set or Exim is run as root.
7028 The bug described in (1) was not present in this case, but the tidying up
7029 of the other signals was missing. I have made the two cases consistent.
7031 3. The ignore_target_hosts setting on a manualroute router was being ignored
7032 for hosts that were looked up using the /MX notation.
7034 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7037 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7038 operated on the sender address. After changing the $sender_address to <>
7039 for the sender address verify, Exim was re-instated it as the original
7040 (before rewriting) address, but remembering that it had rewritten it, so it
7041 wasn't rewriting it again. This bug also had the effect of breaking the
7042 sender address verification caching when the sender address was rewritten.
7044 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7045 This has been changed so that if the ip literal address matches
7046 ignore_target_hosts, the router declines.
7048 7. Added expansion conditions match_domain, match_address, and match_local_
7049 part (NOT match_host).
7051 8. The placeholder for the Received: header didn't have a length field set.
7053 9. Added code to Exim itself and to exim_lock to test for a specific race
7054 condition that could lead to file corruption when using MBX delivery. The
7055 issue is with the lockfile that is created in /tmp. If this file is removed
7056 after a process has opened it but before that process has acquired a lock,
7057 there is the potential for a second process to recreate the file and also
7058 acquire a lock. This could lead to two Exim processes writing to the file
7059 at the same time. The added code performs the same test as UW imapd; it
7060 checks after acquiring the lock that its file descriptor still refers to
7061 the same named file.
7063 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7064 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7067 11. Added the smtp_active_hostname option. If used, this will typically be made
7068 to depend on the incoming interface address. Because $interface_address is
7069 not set up until the daemon has forked a reception process, error responses
7070 that can happen earlier (such as "too many connections") no longer contain
7073 12. If an expansion in a condition on a "warn" statement fails because a lookup
7074 defers, the "warn" statement is abandoned, and the next ACL statement is
7075 processed. Previously this caused the whole ACL to be aborted.
7077 13. Added the iplsearch lookup type.
7079 14. Added ident_timeout as a log selector.
7081 15. Added tls_certificate_verified as a log selector.
7083 16. Added a global option tls_require_ciphers (compare the smtp transport
7084 option of the same name). This controls incoming TLS connections.
7086 17. I finally figured out how to make tls_require_ciphers do a similar thing
7087 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7088 before starting the TLS session.
7090 18. Tabs are now shown as \t in -bP output.
7092 19. If the log selector return_path_on_delivery was set, Exim crashed when
7093 bouncing a message because it had too many Received: header lines.
7095 20. If two routers both had headers_remove settings, and the first one included
7096 a superfluous trailing colon, the final name in the first list and the
7097 first name in the second list were incorrectly joined into one item (with a
7098 colon in the middle).
7104 1. Added -C and -D options to the exinext utility, mainly to make it easier
7105 to include in the automated testing, but these could be helpful when
7106 multiple configurations are in use.
7108 2. The exinext utility was not formatting the output nicely when there was
7109 an alternate port involved in the retry record key, nor when there was a
7110 message id as well (for retries that were specific to a specific message
7111 and a specific host). It was also confused by IPv6 addresses, because of
7112 the additional colons they contain. I have fixed the IPv4 problem, and
7113 patched it up to do a reasonable job for IPv6.
7115 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7116 delivery, the log line now contains "pipelined" if PIPELINING was used.
7118 4. An SMTP transport process used to panic and die if the bind() call to set
7119 an explicit outgoing interface failed. This has been changed; it is now
7120 treated in the same way as a connect() failure.
7122 5. A reference to $sender_host_name in the part of a conditional expansion
7123 that was being skipped was still causing a DNS lookup. This no longer
7126 6. The def: expansion condition was not recognizing references to header lines
7127 that used bh_ and bheader_.
7129 7. Added the _cache feature to named lists.
7131 8. The code for checking quota_filecount in the appendfile transport was
7132 allowing one more file than it should have been.
7134 9. For compatibility with Sendmail, the command line option
7142 and sets the incoming protocol and host name (for trusted callers). The
7143 host name and its colon can be omitted when only the protocol is to be set.
7144 Note the Exim already has two private options, -pd and -ps, that refer to
7145 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7146 "s", but I don't think that's a major issue.
7148 10. A number of refactoring changes to the code, none of which should affect
7151 (a) The number of logging options was getting close to filling up the
7152 32-bit word that was used as a bit map. I have split them into two classes:
7153 those that are passed in the argument to log_write(), and those that are
7154 only ever tested independently outside of that function. These are now in
7155 separate 32-bit words, so there is plenty of room for expansion again.
7156 There is no change in the user interface or the logging behaviour.
7158 (b) When building, for example, log lines, the code previously used a
7159 macro that called string_cat() twice, in order to add two strings. This is
7160 not really sufficiently general. Furthermore, there was one instance where
7161 it was actually wrong because one of the argument was used twice, and in
7162 one call a function was used. (As it happened, calling the function twice
7163 did not affect the overall behaviour.) The macro has been replaced by a
7164 function that can join an arbitrary number of extra strings onto a growing
7167 (c) The code for expansion conditions now uses a table and a binary chop
7168 instead of a serial search (which was left over from when there were very
7169 few conditions). Also, it now recognizes conditions like "pam" even when
7170 the relevant support is not compiled in: a suitably worded error message is
7171 given if an attempt is made to use such a condition.
7173 11. Added ${time_interval:xxxxx}.
7175 12. A bug was causing one of the ddress fields not to be passed back correctly
7176 from remote delivery subprocesses. The field in question was not being
7177 subsequently used, so this caused to problems in practice.
7179 13. Added new log selectors queue_time and deliver_time.
7181 14. Might have fixed a bug in maildirsizefile handling that threw up
7182 "unexpected character" debug warnings, and recalculated the data
7183 unnecessarily. In any case, I expanded the warning message to give more
7186 15. Added the message "Restricted characters in address" to the statements in
7187 the default ACL that block characters like @ and % in local parts.
7189 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7190 Three changes have been made:
7192 (a) There was a serious bug; a negative response to MAIL caused the whole
7193 recipient domain to be cached as invalid, thereby blocking all messages
7194 to all local parts at the same domain, from all senders. This bug has
7195 been fixed. The domain is no longer cached after a negative response to
7196 MAIL if the sender used is not empty.
7198 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7201 (c) A new callout option, "use_sender" has been added for people who want
7202 the modified behaviour.
7208 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7211 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7212 indeed breaks things for older releases.
7214 3. Added additional logging to the case where there is a problem reading data
7215 from a filter that is running in a subprocess using a pipe, in order to
7216 try to track down a specific problem.
7218 4. Testing facility fudge: when running in the test harness and attempting
7219 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7220 getting "No route to host". Convert this to a timeout.
7222 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7225 6. Some OS don't have socklen_t but use size_t instead. This affects the
7226 fifth argument of getsockopt() amongst other things. This is now
7227 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7228 can be set for individual OS. I have set it for SunOS5, OSF1, and
7229 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7230 some earlier ones do not.
7232 7. Change 4.30/15 was not doing the test caselessly.
7234 8. The standard form for an IPv6 address literal was being rejected by address
7235 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7236 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7237 this, as well as the form without the "IPv6" on the front (but only when
7238 address literals are enabled, of course).
7240 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7242 10. Exim crashed if a message with an empty sender address specified by -f
7243 encountered a router with an errors_to setting. This could be provoked only
7244 by a command such as
7248 where an empty string was supplied; "<>" did not hit this bug.
7250 11. Installed PCRE release 4.5.
7252 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7253 remained set. It is now erased.
7255 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7256 times from message ids (which are base 36 rather than the normal 62).
7258 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7259 were being counted as actual protocol errors, and logged if the log
7260 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7261 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7262 and DATA following a set of rejected RCPTs do not count as protocol errors.
7263 In other words, Exim assumes they were pipelined, though this may not
7264 actually be the case. Of course, in all cases the client gets an
7265 appropriate error code.
7267 15. If a lookup fails in an ACL condition, a message about the failure may
7268 be available; it is used if testing the ACL cannot continue, because most
7269 such messages specify what the cause of the deferral is. However, some
7270 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7271 that caused an old message to be retained and used if a later statement
7272 caused a defer, replacing the real cause of the deferral.
7274 16. If an IP address had so many PTR records that the DNS lookup buffer
7275 was not large enough to hold them, Exim could crash while trying to process
7276 the truncated data. It now detects and logs this case.
7278 17. Further to 4.21/58, another change has been made: if (and only if) the
7279 first line of a message (the first header line) ends with CRLF, a bare LF
7280 in a subsequent header line has a space inserted after it, so as not to
7281 terminate the header.
7283 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7284 unnecessarily, used atoi() instead of strtol(), and didn't check the
7285 termination when getting file sizes from file names by regex.
7287 19. Completely re-implemented the support for maildirsize files, in the light
7288 of a number of problems with the previous contributed implementation
7289 (4.30/29). In particular:
7291 . If the quota is zero, the maildirsize file is maintained, but no quota is
7294 . If the maildir directory does not exist, it is created before any attempt
7295 to write a maildirsize file.
7297 . The quota value in the file is just a cache; if the quota is changed in
7298 the transport, the new value overrides.
7300 . A regular expression is available for excluding directories from the
7303 20. The autoreply transport checks the characters in options that define the
7304 message's headers; it allows continued headers, but it was checking with
7305 isspace() after an embedded newline instead of explicitly looking for a
7308 21. If all the "regular" hosts to which an address was routed had passed their
7309 expiry times, and had not reached their retry times, the address was
7310 bounced, even if fallback hosts were defined. Now Exim should go on to try
7313 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7314 equivalent code in the SMTP transport. Some hosts send humungous responses
7315 to HELO/EHLO, more than 1024 it seems.
7317 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7318 gives compiler warnings in some environments. I've now done it "properly",
7321 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7322 (because of problems with the built-in one) was declared to return uschar *
7323 instead of char *, causing compiler failure.
7325 25. Fixed a file descriptor leak when processing alias/forward files.
7327 26. Fixed a minor format string issue in dbfn.c.
7329 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7331 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7332 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7335 29. When a sender address is verified, it is cached, to save repeating the test
7336 when there is more than one recipient in a message. However, when the
7337 verification involves a callout, it is possible for different callout
7338 options to be set for different recipients. It is too complicated to keep
7339 track of this in the cache, so now Exim always runs a verification when a
7340 callout is required, relying on the callout cache for the optimization.
7341 The overhead is duplication of the address routing, but this should not be
7344 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7345 to be verified with callout=postmaster, and the main callout worked but the
7346 postmaster check failed, the verification correctly failed. However, if a
7347 subsequent RCPT command asked for sender verification *without* the
7348 postmaster check, incorrect caching caused this verification also to fail,
7351 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7352 it was not caching the DNS options (qualify_single, search_parents) that
7353 were used when the lookup failed. A subsequent lookup with different
7354 options therefore always gave the same answer, though there were cases
7355 where it should not have. (Example: a "domains = !$mx_any" option on a
7356 dnslookup router: the "domains" option is always processed without any
7357 widening, but the router might have qualify_single set.) Now Exim uses the
7358 cached value only when the same options are set.
7360 32. Added John Jetmore's "exipick" utility to the distribution.
7362 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7363 than a timeout (e.g. a certificate is required, and is not provided), an
7364 Exim server now closes the connection immediately. Previously it waited for
7365 the client to close - but if the client is SSL, it seems that they each
7366 wait for each other, leading to a delay before one of them times out.
7368 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7369 maintained 0.8.x compatibility because I don't think many are using it, and
7370 it is clearly obsolete.
7372 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7375 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7376 client certificate was expired. A simple patch fixes this, though I don't
7377 understand the full logic of why the verify callback is called multiple
7380 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7381 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7382 which causes problems with some clients (such as the Certicom SSL Plus
7383 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7384 disables the coutermeasure allowing Eudora to connect."
7386 38. Exim was not checking that a write() to a log file succeeded. This could
7387 lead to Bad Things if a log got too big, in particular if it hit a file
7388 size limit. Exim now panics and dies if it cannot write to a log file, just
7389 as it does if it cannot open a log file.
7391 39. Modified OS/Makefile-Linux so that it now contains
7393 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7395 The two -D definitions ensure that Exim is compiled with large file
7396 support, which makes it possible to handle log files that are bigger than
7399 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7400 instance) a domain was checked against a named list that involved a lookup,
7401 causing $domain_data to be set, then another domain was checked against the
7402 same list, then the first domain was re-checked, the value of $domain_data
7403 after the final check could be wrong. In particular, if the second check
7404 failed, it could be set empty. This bug probably also applied to
7407 41. The strip_trailing_dot option was not being applied to the address given
7408 with the -f command-line option.
7410 42. The code for reading a message's header from the spool was incrementing
7411 $received_count, but never initializing it. This meant that the value was
7412 incorrect (doubled) while delivering a message in the same process in which
7413 it was received. In the most common configuration of Exim, this never
7414 happens - a fresh exec is done - but it can happen when
7415 deliver_drop_privilege is set.
7417 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7418 it now includes up to 150 characters of the unexpected data in the log
7421 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7422 and building data strings. The size of both of these buffers was 10 000
7423 bytes - far larger than anybody would *ever* want, thought I. Needless to
7424 say, somebody hit the limit. I have increased the maximum line length to
7425 20 000 and the maximum data length of concatenated lines to 100 000. I have
7426 also fixed two bugs, because there was no checking on these buffers. Tsk,
7427 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7428 buffer is too small.
7430 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7431 lsearch lookups. Now it does.
7433 46. When parsing a route_list item in a manualroute router, a fixed-length
7434 buffer was used for the list of hosts. I made this 1024 bytes long,
7435 thinking that nobody would ever have a list of hosts that long. Wrong.
7436 Somebody had a whole pile of complicated expansion conditions, and the
7437 string was silently truncated, leading to an expansion error. It turns out
7438 that it is easier to change to an unlimited length (owing to other changes
7439 that have happened since this code was originally written) than to build
7440 structure for giving a limitation error. The length of the item that
7441 expands into the list of hosts is now unlimited.
7443 47. The lsearch lookup could not handle data where the length of text line was
7444 more than 4095 characters. Such lines were truncated, leading to shortened
7445 data being returned. It should now handle lines of any length.
7447 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7448 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7451 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7452 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7453 no longer happens in comment lines. A list of replacements is now placed
7454 at the head of all of the source files, except those whose only change is
7455 to replace PERL_COMMAND in the very first #! line.
7457 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7458 messages on the queue, with a bottom-up merge sort, using code contributed
7459 by Michael Haardt. This should make operations like -bp somewhat faster on
7460 large queues. It won't affect queue runners, except when queue_run_in_order
7463 51. Installed eximstats 1.31 in the distribution.
7465 52. Added support for SRV lookups to the dnslookup router.
7467 53. If an ACL referred to $message_body or $message_body_end, the value was not
7468 reset for any messages that followed in the same SMTP session.
7470 54. The store-handling optimization for building very long strings was not
7471 differentiating between the different store pools. I don't think this
7472 actually made any difference in practice, but I've tidied it.
7474 55. While running the routers to verify a sender address, $sender_address
7475 was still set to the sender address. This is wrong, because when routing to
7476 send a bounce to the sender, it would be empty. Therefore, I have changed
7477 it so that, while verifying a sender address, $sender_address is set to <>.
7478 (There is no change to what happens when verifying a recipient address.)
7480 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7481 target A or AAAA records (if not already returned) without resetting the
7482 qualify_single or search_parents options of the DNS resolver. These are
7483 inappropriate in this case because the targets of MX and SRV records must
7484 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7485 target that, when qualified, matched something in the local domain. These
7486 two options are now turned off when doing these lookups.
7488 57. It seems that at least some releases of Reiserfs (which does not have the
7489 concept of a fixed number of inodes) returns zero and not -1 for the
7490 number of available inodes. This interacted badly with check_spool_inodes,
7491 which assumed that -1 was the "no such thing" setting. What I have done is
7492 to check that the total number of inodes is greater than zero before doing
7493 the test of how many are available.
7495 58. When a "warn" ACL statement has a log_message modifier, the message is
7496 remembered, and not repeated. This is to avoid a lot of repetition when a
7497 message has many recipients that cause the same warning to be written.
7498 However, Exim was preserving the list of already written lines for an
7499 entire SMTP session, which doesn't seem right. The memory is now reset if a
7500 new message is started.
7502 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7503 debugging output unless log_rewrite was also set.
7505 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7506 of (int)(handle) when we know that handle contains (void *)(-1).
7508 61. The Exim daemon panic-logs an error return when it closes the incoming
7509 connection. However "connection reset by peer" seems to be common, and
7510 isn't really an error worthy of noting specially, so that particular error
7513 62. When Exim is trying to find all the local interfaces, it used to panic and
7514 die if the ioctl to get the interface flags failed. However, it seems that
7515 on at least one OS (Solaris 9) it is possible to have an interface that is
7516 included in the list of interfaces, but for which you get a failure error
7517 for this call. This happens when the interface is not "plumbed" into a
7518 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7519 failure of the "get flags" call assumes that the interface is down.
7521 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7522 makes life easier for people who are doing arithmetic on fields extracted
7523 from dates, where you often get leading zeros that should not be
7524 interpreted as octal.
7526 64. Added qualify_domain to the redirect router, to override the global
7529 65. If a pathologically long header line contained very many addresses (the
7530 report of this problem mentioned 10 000) and each of them was rewritten,
7531 Exim could use up a very large amount of memory. (It kept on making new
7532 copies of the header line as it rewrote, and never released the old ones.)
7533 At the expense of a bit more processing, the header rewriting function has
7534 been changed so that it no longer eats memory in this way.
7536 66. The generation of the Received: header has been moved from the time that a
7537 message starts to be received, to the time that it finishes. The timestamp
7538 in the Received: header should now be very close to that of the <= log
7539 line. There are two side-effects of this change:
7541 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7542 logged header lines no longer include the local Received: line, because
7543 it has not yet been created. The same applies to a copy of the message
7544 that is returned to a non-SMTP sender when a message is rejected.
7546 (b) When a filter file is tested using -bf, no additional Received: header
7547 is added to the test message. After some thought, I decided that this
7550 This change does not affect the value of $received_for. It is still set
7551 after address rewriting, but before local_scan() is called.
7553 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7555 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7556 gave an unhelpful panic error message, and a defer error. I have managed to
7557 change this behaviour so that it now rejects any supplied certificate,
7558 which seems right, as the list of acceptable certificates is empty.
7560 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7561 gave an unhelpful defer error. I have not managed to make this reject any
7562 supplied certificates, but the error message it gives is "no certificate
7563 supplied", which is not helpful.
7565 70. exigrep's output now also includes lines that are not associated with any
7566 message, but which match the given pattern. Implemented by a patch from
7567 Martin Sluka, which also tidied up the Perl a bit.
7569 71. Recipient callout verification, like sender verification, was using <> in
7570 the MAIL FROM command. This isn't really the right thing, since the actual
7571 sender may affect whether the remote host accepts the recipient or not. I
7572 have changed it to use the actual sender in the callout; this means that
7573 the cache record is now keyed on a recipient/sender pair, not just the
7574 recipient address. There doesn't seem to be a real danger of callout loops,
7575 since a callout by the remote host to check the sender would use <>.
7576 [SEE ABOVE: changed after hitting problems.]
7578 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7579 temporary errors. However, in the case of such a code being given after
7580 the end of a data transmission (i.e. after ".") Exim was failing to write
7581 a retry record for the message. (Yes, there was some broken host that was
7582 actually sending 8xx at this point.)
7584 73. An unknown lookup type in a host list could cause Exim to panic-die when
7585 the list was checked. (An example that provoked this was putting <; in the
7586 middle of a list instead of at the start.) If this happened during a DATA
7587 ACL check, a -D file could be left lying around. This kind of configuration
7588 error no longer causes Exim to die; instead it causes a defer error. The
7589 incident is still logged to the main and panic logs.
7591 74. Buglet left over from Exim 3 conversion. The message "too many messages
7592 in one connection" was written to the rejectlog but not the mainlog, except
7593 when address rewriting (yes!) was being logged.
7595 75. Added write_rejectlog option.
7597 76. When a system filter was run not as root (that is, when system_filter_user
7598 was set), the values of the $n variables were not being returned to the
7599 main process; thus, they were not subsequently available in the $sn
7602 77. Added +return_path_on_delivery log selector.
7604 78. A connection timeout was being treated differently from recipients deferred
7605 when testing hosts_max_try with a message that was older than the host's
7606 retry timeout. (The host should not be counted, thus allowing all hosts to
7607 be tried at least once before bouncing.) This may have been the cause of an
7608 occasionally reported bug whereby a message would remain on the queue
7609 longer than the retry timeout, but would be bounced if a delivery was
7610 forced. I say "may" because I never totally pinned down the problem;
7611 setting up timeout/retry tests is difficult. See also the next item.
7613 79. The ultimate address timeout was not being applied to errors that involved
7614 a combination of host plus message (for example, a timeout on a MAIL
7615 command). When an address resolved to a number of possible hosts, and they
7616 were not all tried for each delivery (e.g. because of hosts_max_try), a
7617 message could remain on the queue longer than the retry timeout.
7619 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7622 81. Fixed an obscure SMTP outgoing bug which required at least the following
7623 conditions: (a) there was another message waiting for the same server;
7624 (b) the server returned 5xx to all RCPT commands in the first message so
7625 that the message was not completed; (c) the server dropped the connection
7626 or gave a negative response to the RSET that Exim sends to abort the
7627 transaction. The observed case was a dropped connection after DATA that had
7628 been sent in pipelining mode. That is, the server had advertised PIPELINING
7629 but was not implementing it correctly. The effect of the bug was incorrect
7630 behaviour, such as trying another host, and this could lead to a crash.
7636 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7637 and daemon.c were passed as pointers to ints; they should have been
7638 pointers to socklen_t variables (which are typically unsigned ints).
7640 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7643 3. Fixed a really odd bug that affected only the testing scheme; patching a
7644 certain fixed string in the binary changed the value of another string that
7645 happened to be identical to the end of the original first string.
7647 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7648 name", it returns that address as the IP address. On some operating
7649 systems (e.g. Solaris), it also passes back the IP address string as the
7650 "host name". However, on others (e.g. Linux), it passes back an empty
7651 string. Exim wasn't checking for this, and was changing the host name to an
7652 empty string, assuming it had been canonicalized.
7654 5. Although rare, it is permitted to have more than one PTR record for a given
7655 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7656 all the names associated with an address, because they do in Solaris.
7657 However, it seems that they do not in Linux for data that comes from the
7658 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7659 I found this out when I moved to a new Linux workstation and tried to run
7660 the Exim test suite.
7662 To get round this problem I have changed the code so that it now does its
7663 own call to the DNS to look up PTR records when searching for a host name.
7664 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7665 addresses that are only in /etc/hosts are still found.
7667 This behaviour is, however, controlled by an option called host_lookup_
7668 order, which defaults to "bydns:byaddr". If people want to use the other
7669 order, or indeed, just use one or the other means of lookup, they can
7670 specify it in this variable.
7672 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7673 some operating systems, this comes back from gethostbyaddr() as an empty
7674 string, and this is what Exim used to test for. However, it seems that in
7675 other systems, "." is yielded. Exim now tests for this case too.
7677 7. The values of check_spool_space and check_log_space are now held internally
7678 as a number of kilobytes instead of an absolute number of bytes. If a
7679 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7680 kilobyte. This means that much larger values can be stored.
7682 8. Exim monitor: an attempt to get the action menu when not actually pointing
7683 at a message produces an empty menu entitled "No message selected". This
7684 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7685 no entries in it ("Shell widget menu has zero width and/or height"). So I
7686 have added a single, blank menu entry in this case.
7688 9. Added ${quote_local_part.
7690 10. MIME decoding is now applied to the contents of Subject: header lines when
7693 11. Now that a reference to $sender_host_address automatically causes a reverse
7694 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7695 host lookup before query-style lookups in lists that might use this
7696 variable. This has therefore been abolished, and the "net-" prefix is no
7697 longer necessary for query-style lookups.
7699 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7700 have been a typo for LFLAGS, so it has been changed.
7702 13. The install script calls Exim with "-C /dev/null" in order to find the
7703 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7704 to be output. However, since Exim outputs its version number before the
7705 error, it didn't break the script. It just looked ugly. I fixed this by
7706 always allowing "-C /dev/null" if the caller is root.
7708 14. Ignore overlarge ACL variable number when reading spool file - insurance
7709 against a later release with more variables having written the file.
7711 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7712 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7713 this, as well as the form without the "IPv6" on the front.
7715 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7716 OS/Makefile-Darwin file.
7718 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7719 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7721 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7722 variable. However, some OS use ints for inodes. I've added cast to long int
7723 to get rid of the compiler warning.
7725 19. I had forgotten to lock out "/../" in configuration file names when
7726 ALT_CONFIG_PREFIX was set.
7728 20. Routers used for verification do not need to specify transports. However,
7729 if such a router generated a host list, and callout was configured, Exim
7730 crashed, because it could not find a port number from the (non-existent)
7731 transport. It now assumes port 25 in this circumstance.
7733 21. Added the -t option to exigrep.
7735 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7736 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7737 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7738 them is set, LOOKUP_LSEARCH is forced.
7740 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7742 24. Added sender and host information to the "rejected by local_scan()" log
7743 line; previously there was no indication of these.
7745 25. Added .include_if_exists.
7747 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7748 Linux. It turns out that not all file systems support this. Apparently some
7749 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7750 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7751 not supported on the file descriptor, is now ignored when Exim is trying to
7752 sync a directory. This applies only to Linux.
7754 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7756 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7757 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7760 29. Added support for maildirsize files from supplied patch (modified a bit).
7762 30. The use of :fail: followed by an empty string could lead Exim to respond to
7763 sender verification failures with (e.g.):
7765 550 Verification failed for <xxx>
7766 550 Sender verify failed
7768 where the first response line was missing the '-' that indicates it is not
7769 the final line of the response.
7771 31. The loop for finding the name of the user that called Exim had a hardwired
7772 limit of 10; it now uses the value of finduser_retries, which is used for
7773 all other user lookups.
7775 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7778 33. Exim was neglecting to zero errno before one call of strtol() when
7779 expanding a string and expecting an integer value. On some systems this
7780 resulted in spurious "integer overflow" errors. Also, it was casting the
7781 result into an int without checking.
7783 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7784 did not work. The code looks as if it has *never* worked, though it appears
7785 to have been documented since at least release 1.62. I have made it work.
7787 35. The "timeout_DNS" error in retry rules, also documented since at least
7788 1.62, also never worked. As it isn't clear exactly what this means, and
7789 clearly it isn't a major issue, I have abolished the feature by treating it
7790 as "timeout", and writing a warning to the main and panic logs.
7792 36. The display of retry rules for -brt wasn't always showing the error code
7795 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7796 timeout_connect_A, timeout_connect_MX.
7798 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7799 to the empty sender.
7801 39. The daemon was not analysing the content of -oX till after it had closed
7802 stderr and disconnected from the controlling terminal. This meant that any
7803 syntax errors were only noted on the panic log, and the return code from
7804 the command was 0. By re-arranging the code a little, I've made the
7805 decoding happen first, so such errors now appear on stderr, and the return
7806 code is 1. However, the actual setting up of the sockets still happens in
7807 the disconnected process, so errors there are still only recorded on the
7810 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7811 connections (as happens on some IP stacks) was logged at start up time as
7812 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7813 it from "IPv6 and IPv4", which means that two separate sockets are being
7816 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7817 says whether AF_INET or AF_INET6 was passed as an argument.
7819 42. Exiwhat output was messed up when time zones were included in log
7822 43. Exiwhat now gives more information about the daemon's listening ports,
7823 and whether -tls-on-connect was used.
7825 44. The "port" option of the smtp transport is now expanded.
7827 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7828 silently ignored. Now an error message is written to the main and panic
7831 46. There's a new ACL modifier called "logwrite" which writes to a log file
7832 as soon as it is encountered.
7834 47. Added $local_user_uid and $local_user_gid at routing time.
7836 48. Exim crashed when trying to verify a sender address that was being
7839 49. Exim was recognizing only a space character after ".include". It now also
7840 recognizes a tab character.
7842 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7843 extracting the relevant information from the specification. The man page no
7844 longer contains scrambled data for the -d option, and I've added a section
7845 at the front about calling Exim under different names.
7847 51. Added "extra_headers" argument to the "mail" command in filter files.
7849 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7852 53. Installed eximstats 1.29.
7854 54. Added transport_filter_timeout as a generic transport option.
7856 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7857 Cc: header lines. This was required by RFC 822, but it not required by RFC
7860 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7861 incoming messages that did not have them. Now it does so only if the
7862 message originates locally, that is, if there is no associated remote host
7863 address. When Resent- header lines are present, this applies to the Resent-
7864 lines rather than the non-Resent- lines.
7866 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7867 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7869 58. Messages for configuration errors now include the name of the main
7870 configuration file - useful now that there may be more than one file in a
7871 list (.included file names were always shown).
7873 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7874 for those rare installations that do not start the daemon as root or run it
7875 setuid root. I've cut out the call to initgroups() if the daemon is not
7878 60. The Exim user and group can now be bound into the binary as text strings
7879 that are looked up at the start of Exim's processing.
7881 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7883 62. Added $mailstore_basename variable.
7885 63. Installed patch to sieve.c from Michael Haardt.
7887 64. When Exim failed to open the panic log after failing to open the main log,
7888 the original message it was trying to log was written to stderr and debug
7889 output, but if they were not available (the usual case in production), it
7890 was lost. Now it is written to syslog before the two lines that record the
7891 failures to open the logs.
7893 65. Users' Exim filters run in subprocesses under the user's uid. It is
7894 possible for a "deliver" command or an alias in a "personal" command to
7895 provoke an address rewrite. If logging of address rewriting is configured,
7896 this fails because the process is not running as root or exim. There may be
7897 a better way of dealing with this, but for the moment (because 4.30 needs
7898 to be released), I have disabled address rewrite logging when running a
7899 filter in a non-root, non-exim process.
7905 1. The buildconfig auxiliary program wasn't quoting the value set for
7906 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7907 not defined. This bug was masked in 4.22 by the effect that was fixed in
7910 2. Some messages that were rejected after a message id was allocated were
7911 shown as "incomplete" by exigrep. It no longer does this for messages that
7912 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7914 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7915 have allow_domain_literals set, the ID did not get logged in the <= line.
7916 Domain literals are now always recognized in Message-ID: header lines.
7918 4. The first argument for a ${extract expansion item is the key name or field
7919 number. Leading and trailing spaces in this item were not being ignored,
7920 causing some misleading effects.
7922 5. When deliver_drop_privilege was set, single queue runner processes started
7923 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7924 same command in the process it spins off) were not dropping privilege.
7926 6. When the daemon running as "exim" started a queue runner, it always
7927 re-executed Exim in the spun-off process. This is a waste of effort when
7928 deliver_drop_privilege is set. The new process now just calls the
7929 queue-runner function directly.
7935 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7938 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7939 ignored. Though the use of -oP was forcing the writing of a pid file, it
7940 was always written to the default place.
7942 3. If the message "no IP address found for host xxxx" is generated during
7943 incoming verification, it is now followed by identification of the incoming
7944 connection (so you can more easily find what provoked it).
7946 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7948 5. Added some features to "harden" Exim a bit more against certain attacks:
7950 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7951 be put in Local/Makefile. This is like the never_users runtime option,
7952 but it cannot be overridden. The default setting is "root".
7954 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7955 prefix string with which any file named in a -C command line option
7958 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7959 is retained for -C and -D only if the caller of Exim is root. Without
7960 it, the exim user may also use -C and -D and retain privilege.
7962 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7963 command line option is disabled.
7965 6. Macro names set by the -D option must start with an upper case letter, just
7966 like macro names defined in the configuration file.
7968 7. Added "dereference=" facility to LDAP.
7970 8. Two instances of the typo "uknown" in the source files are fixed.
7972 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7973 the Configure-Makefile script screwed up while processing it.
7975 10. Incorporated PCRE 4.4.
7977 11. The SMTP synchronization check was not operating right at the start of an
7978 SMTP session. For example, it could not catch a HELO sent before the client
7979 waited for the greeting. There is now a check for outstanding input at the
7980 point when the greeting is written. Because of the duplex, asynchronous
7981 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7982 way, but not yet received, when the check is performed.
7984 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7985 on TCP/IP sockets, because this apparently causes some broken clients to
7988 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7989 unchanged) from the Cygwin maintainer.
7991 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7992 was supported instead of testing for mbx. Effectively a typo.
7994 15. The spa authenticator server code was not checking that the input it
7995 received was valid base64.
7997 16. The debug output line for the "set" modifier in ACLs was not showing the
7998 name of the variable that was being set.
8000 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8002 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8003 lookup on its own. Something else was needed to trigger the lookup. For
8004 example, a match in host_lookup or the need for a host name in a host list.
8005 Now, if $sender_host_name is referenced and the host name has not yet been
8006 looked up, a lookup is performed. If the lookup fails, the variable remains
8007 empty, and $host_lookup_failed is set to "1".
8009 19. Added "eqi" as a case-independent comparison operator.
8011 20. The saslauthd authentication condition could segfault if neither service
8012 nor realm was specified.
8014 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8015 error message that was logged was misleading, and incoming SMTP
8016 connections were dropped. The message is now more accurate, and temporary
8017 errors are given to SMTP connections.
8019 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8020 (except RSET) with a 421 error, until QUIT is received. However, it was
8021 failing to send a response to QUIT.
8023 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8024 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8025 ACL is now run after the helo_try_verify_hosts code.
8027 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8028 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8029 case-independent, so other case variants are also recognized). Apparently
8030 some systems use these upper case variants.
8032 25. If more than two messages were waiting for the same host, and a transport
8033 filter was specified for the transport, Exim sent two messages over the
8034 same TCP/IP connection, and then failed with "socket operation on non-
8035 socket" when it tried to send the third.
8037 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8039 27. The extern definition of crypt16() in expand.c was not being excluded when
8040 the OS had its own crypt16() function.
8042 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8043 as a preferred synonym for return_size_limit, both as an option and as an
8046 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8048 30. Changed the default configuration ACL to relax the local part checking rule
8049 for addresses that are not in any local domains. For these addresses,
8050 slashes and pipe symbols are allowed within local parts, but the sequence
8051 /../ is explicitly forbidden.
8053 31. SPA server authentication was not clearing the challenge buffer before
8056 32. log_message in a "warn" ACL statement was writing to the reject log as
8057 well as to the main log, which contradicts the documentation and doesn't
8058 seem right (because no rejection is happening). So I have stopped it.
8060 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8061 However, I am unable to do any testing of this.
8063 34. Fixed an infelicity in the appendfile transport. When checking directories
8064 for a mailbox, to see if any needed to be created, it was accidentally
8065 using path names with one or more superfluous leading slashes; tracing
8066 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8068 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8069 contents are added to the log line that is written for every discarded
8070 recipient. (Previously a log_message setting was ignored.)
8072 36. The ${quote: operator now quotes the string if it is empty.
8074 37. The install script runs exim in order to find its version number. If for
8075 some reason other than non-existence or emptiness, which it checks, it
8076 could not run './exim', it was installing it with an empty version number,
8077 i.e. as "exim-". This error state is now caught, and the installation is
8080 38. An argument was missing from the function that creates an error message
8081 when Exim fails to connect to the socket for saslauthd authentication.
8082 This could cause Exim to crash, or give a corrupted message.
8084 39. Added isip, isip4, and isip6 to ${if conditions.
8086 40. The ACL variables $acl_xx are now saved with the message, and can be
8087 accessed later in routers, transports, and filters.
8089 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8090 strings in the file are not string-expanded.
8092 42. If a MAIL command specified a SIZE value that was too large to fit into an
8093 int variable, the check against message_size_limit failed. Such values are
8094 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8095 day this will have to be increased, but I don't think I want to be around
8096 when emails are that large.
8103 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8104 iconv() is not standard in FreeBSD.
8106 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8107 IPv6 enabled. The observed symptom was a segmentation fault on return from
8108 the function os_common_find_running_interfaces() in src/os.c.
8110 3. In the check_special_case() function in daemon.c I had used "errno" as an
8111 argument name, which causes warnings on some systems. This was basically a
8112 typo, since it was named "eno" in the comments!
8114 4. The code that waits for the clock to tick (at a resolution of some fraction
8115 of a second) so as to ensure message-id uniqueness was always waiting for
8116 at least one whole tick, when it could have waited for less. [This is
8117 almost certainly not relevant at current processor speeds, where it is
8118 unlikely to ever wait at all. But we try to future-proof.]
8120 5. The function that sleeps for a time interval that includes fractions of a
8121 second contained a race. It did not block SIGALRM between setting the
8122 timer, and suspending (a couple of lines later). If the interval was short
8123 and the sigsuspend() was delayed until after it had expired, the suspension
8124 never ended. On busy systems this could lead to processes getting stuck for
8127 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8128 process, before it forks any delivery processes. The open lookup caching
8129 mechanism meant that the open file or database connection was passed into
8130 the delivery process. The problem was that delivery processes always tidy
8131 up cached lookup data. This could cause a problem for the next delivery
8132 process started by the queue runner, because the external queue runner
8133 process does not know about the closure. So the next delivery process
8134 still has data in the lookup cache. In the case of a file lookup, there was
8135 no problem because closing a file descriptor in a subprocess doesn't affect
8136 the parent. However, if the lookup was caching a connection to a database,
8137 the connection was closed, and the second delivery process was likely to
8138 see errors such as "PGSQL: query failed: server closed the connection
8139 unexpectedly". The problem has been fixed by closing all cached lookups
8140 in a queue runner before running a delivery process.
8142 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8143 seem to have the "const" qualifier which it has on other OS. I've
8146 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8147 *of the same type* (client or server) with the same public name that an
8148 error should be diagnosed.
8150 9. When Exim looked up a host name for an IP address, but failed to find the
8151 original IP address when looking up the host name (a safety check), it
8152 output the message "<ip address> does not match any IP for NULL", which was
8153 confusing, to say the least. The bug was that the host name should have
8154 appeared instead of "NULL".
8156 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8157 user that is built into the binary, and the -C or -D options is used, root
8158 privilege is dropped before the configuration file is read. In addition,
8159 logging is switched to stderr instead of the normal log files. If the
8160 configuration then re-defines the Exim user, the unprivileged environment
8161 is probably not what is expected, so Exim logs a panic warning message (but
8164 However, if deliver_drop_privilege is set, the unprivileged state may well
8165 be exactly what is intended, so the warning has been cut out in that case,
8166 and Exim is allowed to try to write to its normal log files.
8172 1. smtp_return_error_details was not giving details for temporary sender
8173 or receiver verification errors.
8175 2. Diagnose a configuration error if two authenticators have the same public
8178 3. Exim used not to create the message log file for a message until the first
8179 delivery attempt. This could be confusing when incoming messages were held
8180 for policy or load reasons. The message log file is now created at the time
8181 the message is received, and an initial "Received" line is written to it.
8183 4. The automatically generated man page for command line options had a minor
8184 bug that caused no ill effects; however, a more serious problem was that
8185 the procedure for building the man page automatically didn't always
8186 operate. Consequently, release 4.20 contains an out-of-date version. This
8187 shouldn't happen again.
8189 5. When building Exim with embedded Perl support, the script that builds the
8190 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8191 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8193 6. The freeze_tell option was not being used for messages that were frozen on
8194 arrival, either by an ACL or by local_scan().
8196 7. Added the smtp_incomplete_transaction log selector.
8198 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8199 was accepting AUTH without a new EHLO.
8201 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8202 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8205 10. Logging of TCP/IP connections (when configured) now happens in the main
8206 daemon process instead of the child process, so that the TCP/IP connection
8207 count is more accurate (but it can never be perfect).
8209 11. The use of "drop" in a nested ACL was not being handled correctly in the
8210 outer ACL. Now, if condition failure induced by the nested "drop" causes
8211 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8212 or "require"), the connection is dropped.
8214 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8215 that yield "discard" can now be used with an "accept" or a "discard" verb,
8216 but an error is generated for any others (because I can't see a useful way
8217 to define what should happen).
8219 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8220 are now processed in the same way as lines in the Exim configuration file.
8221 In particular, continuation lines are supported.
8223 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8225 15. Added -ti meaning -t -i.
8227 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8228 domains, and warn by logging if others are found.
8230 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8231 returned from GIFCONF when Exim is trying to find the list of interfaces on
8232 a host. The code in os.c has been modified to copy each ifreq to an aligned
8233 structure in all cases.
8235 Also, in some cases, the returned ifreq's were being copied to a 'struct
8236 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8237 means the last couple of bytes of an IPv6 address could be chopped if the
8238 ifreq contained only a normal sockaddr (14 bytes storage).
8240 18. Named domain lists were not supported in the hosts_treat_as_local option.
8241 An entry such as +xxxx was not recognized, and was treated as a literal
8244 19. Ensure that header lines added by a DATA ACL are included in the reject log
8245 if the ACL subsequently rejects the message.
8247 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8248 MD5 (which is deprecated).
8250 21. When testing a filter file using -bf, Exim was writing a message when it
8251 took the sender from a "From " line in the message, but it was not doing so
8252 when it took $return_path from a Return-Path: header line. It now does.
8254 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8255 with a valid header line field name (a series of printing characters
8256 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8258 23. Changed "disc" in the source to "disk" to conform to the documentation and
8259 the book and for uniformity.
8261 24. Ignore Sendmail's -Ooption=value command line item.
8263 25. When execve() failed while trying to run a command in a pipe transport,
8264 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8265 could be confused with a return value of 69 from the command itself. This
8266 has been changed to 127, the value the shell returns if it is asked to run
8267 a non-existent command. The wording for the related log line suggests a
8268 non-existent command as the problem.
8270 26. If received_header_text expands to an empty string, do not add a Received:
8271 header line to the message. (Well, it adds a token one on the spool, but
8272 marks it "old" so that it doesn't get used or transmitted.)
8274 27. Installed eximstats 1.28 (addition of -nt option).
8276 28. There was no check for failure on the call to getsockname() in the daemon
8277 code. This can fail if there is a shortage of resources on the system, with
8278 ENOMEM, for example. A temporary error is now given on failure.
8280 29. Contrary to the C standard, it seems that in some environments, the
8281 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8282 program. Exim now does this explicitly; it affects the formatting of
8283 timestamps using strftime().
8285 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8286 complaints. I've now initialized all the variables, to avoid this.
8288 32. Header lines added by a system filter were not being "seen" during
8289 transport-time rewrites.
8291 33. The info_callback() function passed to OpenSSL is set up with type void
8292 (*)(SSL *, int, int), as described somewhere. However, when calling the
8293 function (actually a macro) that sets it up, the type void(*)() is
8294 expected. I've put in a cast to prevent warnings from picky compilers.
8296 34. If a DNS black list lookup found a CNAME record, but there were no A
8297 records associated with the domain it pointed at, Exim crashed.
8299 35. If a DNS black list lookup returned more than one A record, Exim ignored
8300 all but the first. It now scans all returned addresses if a particular IP
8301 value is being sought. In this situation, the contents of the
8302 $dnslist_value variable are a list of all the addresses, separated by a
8305 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8306 to accept a host name and all its aliases if the forward lookup for any of
8307 them yielded the IP address of the incoming connection. Now it accepts only
8308 those names whose forward lookup yields the correct IP address. Any other
8309 names are discarded. This closes a loophole whereby a rogue DNS
8310 administrator could create reverse DNS records to break through a
8311 wildcarded host restriction in an ACL.
8313 37. If a user filter or a system filter that ran in a subprocess used any of
8314 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8315 the wrong values were passed to the pipe command ($thisaddress had the
8316 value of $0, $0 had the value of $1, etc). This bug was introduced by
8317 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8319 38. Improved the line breaking for long SMTP error messages from ACLs.
8320 Previously, if there was no break point between 40 and 75 characters, Exim
8321 left the rest of the message alone. Two changes have been made: (a) I've
8322 reduced the minimum length to 35 characters; (b) if it can't find a break
8323 point between 35 and 75 characters, it looks ahead and uses the first one
8324 that it finds. This may give the occasional overlong line, but at least the
8325 remaining text gets split now.
8327 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8328 file descriptors might be low, and that setting 1000 would always raise it.
8329 It turns out that in some environments, the limit is already over 1000 and
8330 that lowering it causes trouble. So now Exim takes care not to decrease it.
8332 40. When delivering a message, the value of $return_path is set to $sender_
8333 address at the start of routing (routers may change the value). By an
8334 oversight, this default was not being set up when an address was tested by
8335 -bt or -bv, which affected the outcome if any router or filter referred to
8338 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8339 log only when "message" or "log_message" are set. However, if one of the
8340 conditions was an address verification, or a call to a nested ACL, the
8341 messages generated by the underlying test were being passed through. This
8342 no longer happens. The underlying message is available in $acl_verify_
8343 message for both "message" and "log_message" expansions, so it can be
8344 passed through if needed.
8346 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8347 new expansion $bh_ to give the encoded byte string without charset
8348 translation. Translation happens only if iconv() is available; HAVE_ICONV
8349 indicates this at build time. HEADERS_CHARSET gives the charset to
8350 translate to; headers_charset can change it in the configuration, and
8351 "headers charset" can change it in an individual filter file.
8353 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8354 that creates RFC 2047 encoded "words" labels them as that charset instead
8355 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8356 expansion operator; (ii) when Exim creates a From: line for a local
8357 message; (iii) when a header line is rewritten to include a "phrase" part.
8359 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8360 buggy, causing it to skip the first lines of messages whose message ID
8361 ended in 'D'. This would not have bitten before Exim release 4.14, because
8362 message IDs were unlikely to end in 'D' before then. The effect was to have
8363 incorrect size information for certain domains.
8365 45. #include "config.h" was missing at the start of the crypt16.c module. This
8366 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8369 46. If there was a timeout during a "random" callout check, Exim treated it as
8370 a failure of the random address, and carried on sending RSET and the real
8371 address. If the delay was just some slowness somewhere, the response to the
8372 original RCPT would be taken as a response to RSET and so on, causing
8373 mayhem of various kinds.
8375 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8376 when I implemented it. It didn't allow for the fact that some option values
8377 may legitimately be negative (e.g. size_addition), and it didn't even do
8378 the right test for positive values.
8380 48. Domain names in DNS records are case-independent. Exim always looks them up
8381 in lower case. Some resolvers return domain names in exactly the case they
8382 appear in the zone file, that is, they may contain uppercase letters. Not
8383 all resolvers do this - some return always lower case. Exim was treating a
8384 change of case by a resolver as a change of domain, similar to a widening
8385 of a domain abbreviation. This triggered its re-routing code and so it was
8386 trying to route what was effectively the same domain again. This normally
8387 caused routing to fail (because the router wouldn't handle the domain
8388 twice). Now Exim checks for this case specially, and just changes the
8389 casing of the domain that it ultimately uses when it transmits the message
8392 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8395 50. If a filter generated a file delivery with a non-absolute name (possible if
8396 no home directory exists for the router), the forbid_file option was not
8399 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8400 the existing equality matching.
8402 52. Exim was using ints instead of ino_t variables in some places where it was
8403 dealing with inode numbers.
8405 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8406 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8407 called TMPDIR, and if it finds it is different, it changes its value.
8409 54. The smtp_printf() function is now made available to local_scan() so
8410 additional output lines can be written before returning. There is also an
8411 smtp_fflush() function to enable the detection of a dropped connection.
8412 The variables smtp_input and smtp_batched_input are exported to
8415 55. Changed the default runtime configuration: the message "Unknown user"
8416 has been removed from the ACL, and instead placed on the localuser router,
8417 using the cannot_route_message feature. This means that any verification
8418 failures that generate their own messages won't get overridden. Similarly,
8419 the "Unrouteable address" message that was in the ACL for unverifiable
8420 relay addresses has also been removed.
8422 56. Added hosts_avoid_esmtp to the smtp transport.
8424 57. The exicyclog script was not checking for the esoteric option
8425 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8426 will work only if exicyclog is run under the appropriate euid.
8428 58. Following a discussion on the list, the rules by which Exim recognises line
8429 endings on incoming messages have been changed. The -dropcr and drop_cr
8430 options are now no-ops, retained only for backwards compatibility. The
8431 following line terminators are recognized: LF CRLF CR. However, special
8432 processing applies to CR:
8434 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8435 nor a local message in the state where . is a terminator.
8437 (ii) If a bare CR is encountered in a header line, an extra space is added
8438 after the line terminator so as not to end the header. The reasoning
8439 behind this is that bare CRs in header lines are most likely either
8440 to be mistakes, or people trying to play silly games.
8442 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8443 was being incorrectly given as 18 bytes larger than it should have been.
8444 This is a VOB (very old bug).
8446 60. This may never have affected anything current, but just in case it has:
8447 When the local host is found other than at the start of a list of hosts,
8448 the local host, those with the same MX, and any that follow, are discarded.
8449 When the list in question was part of a longer list of hosts, the following
8450 hosts (not currently being processed) were also being discarded. This no
8451 longer happens. I'm not sure if this situation could ever has previously
8454 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8457 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8458 header. This is something that is recommended in a new Internet Draft, and
8459 is something that is documented as being done by Sendmail. There are two
8460 possible values. For messages generated by the autoreply transport, Exim
8463 Auto-Submitted: auto-replied
8465 whereas for all other generated messages (e.g. bounces) it adds
8467 Auto-Submitted: auto-generated
8469 63. The "personal" condition in filters now includes a test for the
8470 Auto-Submitted: header. If it contains the string "auto-" the message it
8471 not considered personal.
8473 64. Added rcpt_include_affixes as a generic transport option.
8475 65. Added queue_only_override (default true).
8477 66. Added the syslog_duplication option.
8479 67. If what should have been the first header line of a message consisted of
8480 a space followed by a colon, Exim was mis-interpreting it as a header line.
8481 It isn't of course - it is syntactically invalid and should therefore be
8482 treated as the start of the message body. The misbehaviour could have
8483 caused a number of strange effects, including loss of data in subsequent
8484 header lines, and spool format errors.
8486 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8487 client host had authenticated. This control can now be exercised by an ACL
8488 for more flexibility.
8490 69. By default, callouts do not happen when testing with -bh. There is now a
8491 variant, -bhc, which does actually run the callout code, including
8492 consulting and updating the callout cache.
8494 70. Added support for saslauthd authentication, courtesy of Alexander
8497 71. If statvfs() failed on the spool or log directories while checking their
8498 size for availability, Exim confusingly gave the error "space shortage".
8499 Furthermore, in debugging mode it crashed with a floating point exception.
8500 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8501 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8502 serious problem, Exim now writes to the main and panic logs when this
8503 happens, with details of the failure. It then refuses to accept the
8504 incoming message, giving the message "spool directory problem" or "log
8505 directory problem" with a 421 code for SMTP messages.
8507 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8508 0, 1, and 2 exist, because some OS complain for execs without them (see
8509 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8510 descriptors. However, the code omitted to check that the open succeeded,
8511 causing mysterious errors if for some reason the permissions on /dev/null
8512 got screwed. Now Exim writes a message to the main and panic logs, and
8513 bombs out if it can't open /dev/null.
8515 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8516 interact so that it is all more flexible. It is supposed to remain
8517 backwards compatible. Also added extra_local_interfaces.
8519 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8520 to bomb out with an assertion failure - to the client this appears as a
8521 connection drop. This problem occurs in the part of the code that was taken
8522 from the Samba project. Fortunately, the assertion is in a very simple
8523 function, so I have fixed this by reproducing the function inline in the
8524 one place where it is called, and arranging for authentication to fail
8525 instead of killing the process with assert().
8527 75. The SPA client code was not working when the server requested OEM rather
8528 than Unicode encoding.
8530 76. Added code to make require_files with a specific uid setting more usable in
8531 the case where statting the file as root fails - usually a non-root-mounted
8532 NFS file system. When this happens and the failure is EACCES, Exim now
8533 forks a subprocess and does the per-uid checking as the relevant uid.
8535 77. Added process_log_path.
8537 78. If log_file_path was not explicitly set, a setting of check_log_space or
8538 check_log_inodes was ignored.
8540 79. If a space check for the spool or log partitions fails, the incident is now
8541 logged. Of course, in the latter case the data may get lost...
8543 80. Added the %p formatting code to string_format() so that it can be used to
8544 print addresses in debug_print(). Adjusted all the address printing in the
8545 debugging in store.c to use %p rather than %d.
8547 81. There was a concern that a line of code in smtp_in.c could overflow a
8548 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8549 initially expressed, the concern was not well-founded, because trailing
8550 spaces are removed early. However, if the trailing spaces were followed by
8551 a NULL, they did not get removed, so the overflow was possible. Two fixes
8554 (a) I re-wrote the offending code in a cleaner fashion.
8555 (b) If an incoming SMTP command contains a NULL character, it is rejected
8558 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8559 runs initgroups(), so that if the Exim user is in any additional groups,
8560 they will be used during message reception.
8566 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob