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.
83 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
84 SMTP connection" log lines.
86 JH/02 Option default value updates:
87 - queue_fast_ramp (main) true (was false)
88 - remote_max_parallel (main) 4 (was 2)
90 JH/03 Cache static regex pattern compilations, for use by ACLs.
92 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
93 Make the rewrite never match and keep the logging. Trust the
94 admin to be using verify=header-syntax (to actually reject the message).
96 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
97 (under Linux) we watch the dir containing the final file; previously
98 it would be the dir with the first symlink. We still do not monitor
101 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
102 dn_expand() is happy to pass them through.
104 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
105 the file with the proof had an unchanged name, the new proof(s) were
106 loaded on top of the old ones (and nover used; the old ones were stapled).
108 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
109 more than one message arrived in a single connection a reference from
110 the earlier message could be re-used. Often a sigsegv resulted.
111 These variables were introduced in Exim 4.87.
112 Debug help from Graeme Fowler.
114 JH/09 Fix ${filter } for conditions that modify $value. Previously the
115 modified version would be used in construction the result, and a memory
118 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
119 Find and fix by Jasen Betts.
121 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
122 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
123 the systemwide configuration to override the Exim config.
125 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
126 API changes in libopendmarc.
128 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
129 pid 1, in the normal "background daemon" mode, having to drop process-
130 group leadership also lost track of needing to create listener sockets.
132 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
133 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
135 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
136 a capture group which obtained no text (eg. "(abc)*" matching zero
137 occurrences) could cause a segfault if the corresponding $<n> was
140 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
141 included a close-brace character (eg. it itself used an expansion) an
144 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
145 starting TLS. Previously it was after, meaning that attackers on such
146 ports had to be screened using the host_reject_connection main config
147 option. The new sequence aligns better with the STARTTLS behaviour, and
148 permits defences against crypto-processing load attacks, even though it
149 is strictly an incompatible change.
150 Also, avoid sending any SMTP fail response for either the connect ACL
151 or host_reject_connection, for TLS-on-connect ports.
153 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
154 Previously this was not permitted, but it makes reasonable sense.
155 While there, restore a restriction on using it from a connect ACL; given
156 the change JH/16 it could only return false (and before 4.91 was not
159 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
160 was exactly sized compared to the log buffer, a crash occurred with the
161 misleading message "bad memory reference; pool not found".
162 Found and traced by Jasen Betts.
164 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
165 dns_again_means_nonexist included an element causing a DNS lookup which
166 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
167 included (though probably not limited to) a process crash from stack
168 memory limit, or from excessive open files. Replace this with a paniclog
169 whine (as this is likely a configuration error), and returning
172 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
173 this always failed, probably leading to the usual downgrade to in-clear
176 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
177 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
178 to not checking that list for these lookups.
180 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
183 JH/23 Fix crash in string expansions. Previously, if an empty variable was
184 immediately followed by an expansion operator, a null-indirection read
185 was done, killing the process.
187 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
188 include an SMTP response string which is longer than that supported
189 by the delivering transport. Alleviate by wrapping such lines before
192 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
193 chars (RFC limit). Previously a limit of 12 items was made, which with
194 a not-impossible References: in the message being bounced could still
195 be over-large and get stopped in the transport.
197 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
198 close. Previously a bare socket close was done.
200 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
203 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
204 message being considered for delivery was already being handled by
205 another process, and having an SMTP connection already open, the function
206 to close it tried to use an uninitialized variable. This would afftect
207 high-volume sites more, especially when running mailing-list-style loads.
208 Pollution of logs was the major effect, as the other process delivered
209 the message. Found and partly investigated by Graeme Fowler.
211 JH/29 Change format of the internal ID used for message identification. The old
212 version only supported 31 bits for a PID element; the new 64 (on systems
213 which can use Base-62 encoding, which is all currently supported ones
214 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
215 and must use Base-36). The new ID is 23 characters rather than 16, and is
216 visible in various places - notably logs, message headers, and spool file
217 names. Various of the ancillary utilities also have to know the format.
218 As well as the expanded PID portion, the sub-second part of the time
219 recorded in the ID is expanded to support finer precision. Theoretically
220 this permits a receive rate from a single comms channel of better than the
222 The major timestamp part of the ID is not changed; at 6 characters it is
223 usable until about year 3700.
224 Updating from previously releases is fully supported: old-format spool
225 files are still usable, and the utilities support both formats. New
226 message will use the new format. The one hints-DB file type which uses
227 message-IDs (the transport wait- DB) will be discarded if an old-format ID
228 is seen; new ones will be built with only new-format IDs.
229 Optionally, a utility can be used to convert spool files from old to new,
230 but this is only an efficiency measure not a requirement for operation
231 Downgrading from new to old requires running a provided utility, having
232 first stopped all operations. This will convert any spool files from new
233 back to old (losing time-precision and PID information) and remove any
234 wait- hints databases.
236 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
237 we treated them as item separators when parsing for a list item, but they
238 need to be protected by the doublequotes. While there, add handling for
241 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
242 Found and fixed by Jasen Betts. No testcase for this as my usual text
243 editor insists on emitting only valid UTF-8.
245 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
246 a null-indirection SIGSEGV for the receive process.
248 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
249 -bh use. Internal checking would spot this and take a panic.
251 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
252 In 4.96 this would expand to empty.
254 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
255 certificate. Find and fix by Andreas Metzler.
257 JH/36 Add ARC info to DMARC hostory records.
259 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
260 or fakedefer. Previously the sender could discover that the message
261 had in fact been accepted.
263 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
264 sequences. Previously the input was not noted as being tainted; notably
265 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
266 bad coding of authenticators.
268 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
269 and ${tr...}. Found and diagnosed by Heiko Schlichting.
271 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
272 could be triggered by externally-supplied input. Found by Trend Micro.
275 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
276 be triggered by externally-controlled input. Found by Trend Micro.
279 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
280 be triggered by externally-controlled input. Found by Trend Micro.
283 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
284 Make the rewrite never match and keep the logging. Trust the
285 admin to be using verify=header-syntax (to actually reject the message).
287 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
290 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
296 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
297 after reception to before a subsequent reception. This should
298 mean slightly faster delivery, and also confirmation of reception
301 JH/02 Move from using the pcre library to pcre2. The former is no longer
302 being developed or supported (by the original developer).
304 JH/03 Constification work in the filters module required a major version
305 bump for the local-scan API. Specifically, the "headers_charset"
306 global which is visible via the API is now const and may therefore
307 not be modified by local-scan code.
309 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
310 sendfile() didi not account for the way the ClamAV driver code called it.
312 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
313 time check was being done for every character; replace that with one
316 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
317 sent was prefixed with a length byte.
319 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
320 RFC 5321. Previously Dovecot (at least) would log errors during
323 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
324 in 4.61 without a reason logged. Binaries will be bigger, which might
325 matter on diskspace-constrained systems, but debug is easier.
327 JH/09 Fix macro-definition during "-be" expansion testing. The move to
328 write-protected store for macros had not accounted for these runtime
329 additions; fix by removing this protection for "-be" mode.
331 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
332 handing out large-numbered file descriptors, violating the usual Unix
333 assumption (and required by Posix) that the lowest possible number will be
334 allocated by the kernel when a new one is needed. In the daemon, and any
335 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
336 are not useable for FD_SET() [and hence select()] and overwrite the stack.
337 Assorted crashes happen.
339 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
340 main-section options or in a connect ACL, the value from the first ever
341 connection was never replaced for subsequent connections. Found by
344 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
345 though only once PCRE2 was introduced: the memory accounting used under
346 debug offset allocations by an int, giving a hard trap in early startup.
347 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
349 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
350 with underbars is given. The write-protection of configuration introduced
351 in 4.95 trapped when normalisation was applied to an option not needing
354 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
356 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
357 failing to close the certificates directory, every hour or any time it
360 JH/16 Debugging initiated by an ACL control now continues through into routing
361 and transport processes. Previously debugging stopped any time Exim
362 re-execs, or for processing a queued message.
364 JH/17 The "expand" debug selector now gives more detail, specifically on the
365 result of expansion operators and items.
367 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
368 bad comparison between the option value and the name of the file to
369 be included was done, and a mismatch was wrongly identified.
370 4.88 to 4.95 are affected.
372 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
374 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
375 supplied as the db file-pair basename. Previously, if a directory
376 path was given, for example via the autoreply "once" option, the DB
377 file.pag and file.dir files would be created in that directory's
380 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
381 "taint" log_selector. These were previously deprecated.
383 JH/22 Fix static address-list lookups to properly return the matched item.
384 Previously only the domain part was returned.
386 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
387 the call into OpenSSL to send a TLS Close was being repeated; this
388 resulted in the library waiting for the peer's Close. If that was never
389 sent we waited forever. Fix by tracking send calls.
391 JH/24 The ${run} expansion item now expands its command string elements after
392 splitting. Previously it was before; the new ordering makes handling
393 zero-length arguments simpler. The old ordering can be obtained by
394 appending a new option "preexpand", after a comma, to the "run".
396 JH/25 Taint-check exec arguments for transport-initiated external processes.
397 Previously, tainted values could be used. This affects "pipe", "lmtp" and
398 "queryprogram" transport, transport-filter, and ETRN commands.
399 The ${run} expansion is also affected: in "preexpand" mode no part of
400 the command line may be tainted, in default mode the executable name
403 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
404 the the facility was not passed across execs, and only the first message
405 passed over a connection could use BDAT; any further ones using DATA.
407 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
408 uses $sending_ip_address and an interface is specified.
409 Previously any use of the local address in the EHLO name disabled
410 PIPECONNECT, the common case being to use the rDNS of it.
412 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
413 resumption. Previously verify failed because no certificate status is
414 passed on the wire for the restarted session. Fix by using the recorded
415 ocsp status of the stored session for the new connection.
417 JH/29 TLS resumption: the key for session lookup in the client now includes
418 more info that a server could potentially use in configuring a TLS
419 session, avoiding oferring mismatching sessions to such a server.
420 Previously only the server IP was used.
422 JH/30 Fix string_copyn() for limit greater than actual string length.
423 Previously the copied amount was the limit, which could result in a
424 overlapping memcpy for newly allocated destination soon after a
425 source string shorter than the limit. Found/investigated by KM.
427 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
428 close; it may be needed for a subsequent connection. This caused a
429 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
431 JH/32 Fix CHUNKING for a second message on a connection when the first was
432 rejected. Previously we did not reset the chunking-offered state, and
433 erroneously rejected the BDAT command. Investigation help from
436 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
437 an empty address. Previously the expansion returned an error.
439 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
440 proxy. Previously these were misparsed, leading to paniclog entries.
446 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
447 related applications. Previously an "H" was used where available info
448 says that "M" should be, so change to match.
450 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
451 as arguments, so an implementation trying to copy these into a local
452 buffer was taking a taint-enforcement trap. Fix by using dynamically
453 created buffers. Similar fix for radius expansion condition.
455 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
456 reasonable, eg. to count headers. Fix by using dynamically created
457 buffers rather than a local. Do similar fixes for ACL actions "dcc",
458 "log_reject_target", "malware" and "spam"; the arguments are expanded
459 so could be handling tainted values.
461 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
462 broken the (no-op) support for this sendmail command. Restore it
463 to doing nothing, silently, and returning good status.
465 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
466 record path was given (or the default used) without a leading directory
467 path, an error occurred on trying to open it. Use the transport's working
470 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
471 transport. Previously it was the name on the DNS A-record; use instead
472 the head of the CNAME chain leading there (if there is one). This seems
473 to align better with RFC 6125.
475 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
476 smtp_accept_max_per_host allocated resources which were not released
477 when the limit was exceeded. This eventually crashed the daemon. Fix
478 by adding a release action in that path.
480 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
481 expanded; previously using tainted values was rejected. Fix by using
482 dynamically-created buffers.
484 JH/09 Relax restrictions on ACL verify condition needing access to message
485 headers. Previously they were only permitted in data and non-smtp ACLs;
486 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
487 not_blind, header_sender and header_names_ascii verification.
489 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
490 Previously a macro used one argument twice; when called with the
491 argument as an expression having side-effects, incorrect operation
492 resulted. Use an inlineable function.
494 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
495 held open for a verify callout. Previously this wan not accounted for
496 and a corrupt onward SMTP conversation resulted.
498 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
499 Previously they were embedded naively in the constructed address; when
500 needed, strip the quoting and quote the entire local_part.
501 Also make the inbound_srs expansion condition handle quoting.
503 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
504 excluded, not matching the documentation.
506 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
507 was given for the sqlite_dbfile a trap resulted.
509 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
510 "name" argument resulted in a trap. There is no reason to disallow such;
511 this was a coding error.
513 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
514 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
515 spent suspended, ignoring the POSIX definition. Previously we assumed
516 it did and a constant offset from real time could be used as a correction.
517 Change to using the same clock source for the start-of-message and the
518 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
519 exists, just to get a clock slightly more aligned to reality.
521 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
522 RFC says it is optional some validators care. The missing char was not
523 intended but triggered by a line-wrap alignment. Discovery and fix by
524 Guillaume Outters, hacked on by JH.
526 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
527 name being quoted was tainted a trap would be taken. Fix by using
528 dynamically created buffers. The routine could have been called by a
529 rewrite with the "h" flag, by using the "-F" command-line option, or
530 by using a "name=" option on a control=submission ACL modifier.
532 JH/19 SPF: change the Authentication-Results expansion component to give
533 smtp.helo when the sender domain is empty. Previously it gave
536 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
537 not in 127.0.0.0/8 to help in spotting list domains taken over by a
538 domain-parking registrar.
540 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
541 Previously when a whitespace character was specified it was not inserted
542 after removing the newline.
544 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
545 the domain part of the recipient address. This overrides any tls_sni
546 option set, which was previously used.
548 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
551 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
552 is_tainted() had an off-by-one error in the overenthusiastic direction.
553 Find and fix by Gavan. Although NetBSD is not a supported platform for
554 4.94 this bug could affect other platforms.
556 PP/01 Fix default prime selection to be consistent.
557 One path used ike23 still, instead of exim.dev.20160529.3; now both
558 execution flows will use the same DH primes (currently
559 exim.dev.20160529.3).
561 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
562 option in smtp transport, to match the documentation. Previously
563 verification was not being done in some cases where it should have been.
565 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
566 than one server was defined and depending on the platform memory layout
567 details, an internal consistency trap could be hit while walking the list
570 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
571 files. The value is used by the authresults expansion item. Previously
572 if this was used in a router or transport, a crash could result.
574 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
575 have a sufficient incidence of EINTR returns from select that an
576 interaction with other operations done by the main daemon loop exposed
577 a bug in the error-handling. This was benign apart from the log
580 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
581 for consistency with delivered "=>" and failed "**" lines. While we're
582 there, handle PRX and TFO.
584 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
585 applied. This resulted, if any header-line rewrite rules were configured,
586 in a panic-log triggerable by sending a message with a long address in
587 a header. Fix by increasing the arbitrary limit to larger than a single
588 (dewrapped) 5322 header line maximum size.
590 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
591 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
592 RFC 6648 which deprecates X- options in protocols as a general practice.
593 Changeover between the implementations is handled by the mechanisms
596 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
597 a remote which uses callout/hold. Previously the local was queued.
599 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
602 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
603 log line. The asterisk is now consistently appended to the remote IP
604 (and port, if given), and will also be provided on defer and fail log
605 lines. Previously it could be placed on the local IP if that was being
606 logged, and was only provided on delivery lines.
608 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
610 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
611 authenticator client_send option. Previously the next char, after a pair
612 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
613 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
614 documentation. There is still no way to get a leading ^ immediately
615 after a NUL (ie. for the password of a PLAIN method authenticator.
617 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
618 files. For bad sizes read, delete the record and whine to paniclog.
620 JH/38 When logging an AUTH failure, as server, do not include sensitive
621 information. Previously, the credentials would be included if given
622 as part of the AUTH command line and an ACL denied authentication.
624 JH/39 Bug 2691: fix $local_part_data. When the matching list element
625 referred to a file, bad data was returned. This likely also affected
628 JH/40 The gsasl authenticator now supports caching of the salted password
629 generated by the client-side implementation. This required the addition
630 of a new variable: $auth4.
632 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
633 left undeleted; the attempt to re-create it then failed - resulting in
634 the usual "SIGHUP tp have daemon reload configuration" to not work.
635 This affected any platform not supporting "abstract" Unix-domain
636 sockets (i.e. not Linux).
638 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
639 recipients" response to RCPT in a later response, with a 250. The
640 previous coding assumed this would not happen, and under PIPELINING
641 would result in both lost and duplicate recipients for a message.
643 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
644 Previously the weighting was incorrectly applied. Similar fix for socks
645 proxies. Found and fixed by Heiko Schlichting.
647 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
648 not handle sub-lists included using the +namedlist syntax. While
649 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
650 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
653 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
654 Previously the permanent pool was used, so the sore could not be freed.
655 This meant a connection with many messages would use continually-growing
658 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
659 per-pool so as not to waste too much space. Previously a constant size
660 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
661 DOS attacks harder. The cost is wasted memory use in the larger blocks.
663 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
664 same space repeatedly, and a smaller process footprint.
666 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
667 for GnuTLS. Previously the name started "%s" which, while not a bug,
668 looked as if if might be one.
670 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
671 "continued connection") make the $proxy_* variables available. Previously
672 the information was not passed across the exec() call for subsequent
673 transport executions. This also mean that the log lines for the
674 messages can show the proxy information.
676 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
677 receive time. With modern systems the difference is significant.
678 The historical behaviour can be restored by disabling (a new) log_selector
679 "queue_time_exclusive".
681 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
682 content) were specifically tested for. Now, also cover expansions
683 resulting in ACL names and inline ACL content.
685 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
686 making it unusable in complex expressions.
688 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
689 fail with a taint-check on the spoolfile name, and leave the message
692 HS/01 Enforce absolute PID file path name.
694 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
696 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
698 PP/02 Bug 2643: Correct TLS DH constants.
699 A missing NUL termination in our code-generation tool had led to some
700 incorrect Diffie-Hellman constants in the Exim source.
701 Reported by kylon94, code-gen tool fix by Simon Arlott.
703 PP/03 Impose security length checks on various command-line options.
704 Fixes CVE-2020-SPRSS reported by Qualys.
706 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
707 better. Reported by Qualys.
709 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
710 providing a particularly obnoxious sender full name.
713 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
715 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
718 PP/08 Change default for recipients_max from unlimited to 50,000.
720 PP/09 Fix security issue with too many recipients on a message (to remove a
721 known security problem if someone does set recipients_max to unlimited,
722 or if local additions add to the recipient list).
723 Fixes CVE-2020-RCPTL reported by Qualys.
725 PP/10 Fix security issue in SMTP verb option parsing
726 Fixes CVE-2020-EXOPT reported by Qualys.
728 PP/11 Fix security issue in BDAT state confusion.
729 Ensure we reset known-good where we know we need to not be reading BDAT
730 data, as a general case fix, and move the places where we switch to BDAT
731 mode until after various protocol state checks.
732 Fixes CVE-2020-BDATA reported by Qualys.
734 HS/03 Die on "/../" in msglog file names
736 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
737 the Exim runtime user are allowed to create files.
739 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
742 QS/03 When reading the output from interpreted forward files we do not
743 pass the pipe between the parent and the interpreting process to
744 executed child processes (if any).
746 QS/04 Always die if requested from internal logging, even is logging is
749 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
750 the API; compilation noo longer completes with DMARC support included.
751 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
752 been reported on other platforms.
754 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
756 JH/56 Make the majority of info read from config files readonly, for defence-in-
757 depth against exploits. Suggestion by Qualys.
758 Not supported on Solaris 10.
760 JH/57 Fix control=fakreject for a custom message containing tainted data.
761 Previously this resulted in a log complaint, due to a re-expansion present
762 since fakereject was originally introduced.
764 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
765 modifier was given, a loop resulted.
767 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
768 pipelined SMTP command followed the BDAT LAST then it would be
769 incorrectly treated as part of the message body, causing a verification
772 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
773 require looser validation rules than those for 821-level addresses,
774 which only permit IP addresses.
780 JH/01 Avoid costly startup code when not strictly needed. This reduces time
781 for some exim process initialisations. It does mean that the logging
782 of TLS configuration problems is only done for the daemon startup.
784 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
786 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
787 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
790 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
791 DKIM signing is being done. Previously a transport_filter would always
792 disable CHUNKING, falling back to traditional DATA.
794 JH/05 Regard command-line recipients as tainted.
796 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
798 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
799 PAM library frees one of the arguments given to it, despite the
800 documentation. Therefore a plain malloc must be used.
802 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
803 on-stack buffers were used, resulting in a taint trap when DSN information
804 copied from a received message was written into the buffer.
806 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
807 the ordering of its ARC headers. This caused a crash.
809 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
810 a new record was being constructed with information from the peer, a trap
813 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
814 installation would get error messages from DMARC verify, when it hit the
815 nonexistent file indicated by the default. Distros wanting DMARC enabled
816 should both provide the file and set the option.
817 Also enforce no DMARC verification for command-line sourced messages.
819 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
820 could, depending on the platform, hang at the STARTTLS response.
822 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
823 message on a connection. Previously if one message had ARC headers and
824 the following one did not, a crash could result when adding an
825 Authentication-Results: header.
827 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
828 Exim main code and Exim-related utities. The introduction of taint
829 tracking also did many adjustments to string handling. Since then, eximon
830 frequently terminated with an assert failure.
832 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
833 check for 452 responses. This slightly helps the inefficieny of doing
834 a large alias-expansion into a recipient-limited target. The max_rcpt
835 transport option still applies (and at the current default, will override
836 the new feature). The check is done for either cause of synch, and forces
837 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
838 connection. The new facility is not tunable at this time.
840 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
841 library live data was being used, so the results became garbage. Make
842 copies while it is still usable.
844 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
845 on delivery deferred (==) and failed (**) lines (if a delivery was
846 attemtped). Previously it was only on completion (=>) lines.
848 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
849 for the expansion of the server_scram_iter and server_scram_salt options.
851 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
852 are now specifically given a NO_DATA response without hitting the system
853 resolver. The library goes on to do the now-standard TXT lookup.
854 Use of dnsdb lookups is not affected.
856 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
857 only retrieve the errormessage once. Previously two calls to dlerror()
858 were used, and the second one (for mainlog/paniclog) retrieved null
861 JH/20 Taint checking: disallow use of tainted data for
862 - the appendfile transport file and directory options
863 - the pipe transport command
864 - the autoreply transport file, log and once options
865 - file names used by the redirect router (including filter files)
867 - paths used by single-key lookups
868 Previously this was permitted.
870 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
871 adjusted the size of a major service buffer; this failed because the
872 buffer was in use at the time. Change to a compile-time increase in the
873 buffer size, when this authenticator is compiled into exim.
875 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
876 previous fast-mode was untenable in the face of glibs using mmap to
877 support larger malloc requests.
879 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
880 New values supported, if defined on system where compiled:
881 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
882 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
884 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
885 running a limited number of proceses in parallel, a benefit is gained. The
886 amount varies with the platform hardware and load. The use of the option
887 queue_run_in_order means we cannot do this, as ordering becomes
890 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
891 had introduced a string-copy (for ensuring NUL-termination) which was not
892 appropriate for that case, which can include embedded NUL bytes in the
893 block of data. Investigation showed the copy to actually be needless, the
894 data being length-specified.
896 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
897 done during a receiving connection, and both used TLS, global info was
898 used rather than per-connection info for tracking the state of data
899 queued for transmission. This could result in a connection hang.
901 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
902 Previously, when delivering serveral messages down a single connection
903 only the first would provide a SIZE. This was due to the size information
904 not being properly tracked.
906 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
907 TAI (at 37 seconds currently), pretend to be in UTC for time-related
908 expansion and logging. Previously, spurious values such as a future
909 minute could be seen.
911 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
912 it could crash from a null-deref. This could also affect the
913 ${addresses: } operator and ${readsock } item.
915 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
916 message following a mime one, the variable was not reset.
918 JH/30 When an pipelined-connect fails at the first response, assume incorrect
919 cached capability (perhaps the peer reneged?) and immediately retry in
922 JH/31 Fix spurious detection of timeout while writing to transport filter.
924 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
925 an attempt to copy the string was made before checking it.
927 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
928 taint of the lookup key was maintained; we now regard the presence in the
929 filesystem as sufficient validation.
931 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
932 argument is supplied.
934 JH/35 The dsearch lookup now requires that the directory is an absolute path.
935 Previously this was not checked, and nonempty relative paths made an
936 access under Exim's current working directory.
938 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
939 Previously no event was raised.
941 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
942 parameter supplied by the sender MAIL FROM command. Previously it was
943 ignored, and only the check_spool_space option value for the required
946 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
947 the size of the signing public-key. Previously it was instead giving
948 the size of the signature hash.
950 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
951 the default. See the (new) dkim_verify_min_keysizes option.
953 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
954 and an ACL use a lookup for checking either the local_part or domain,
955 stale data could be accessed. Ensure that variable references are
956 dropped between messages.
958 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
959 by the client was not checked as pointing within response data before
960 being used. A malicious client could thus cause an out-of-bounds read and
961 possibly gain authentication. Fix by adding the check.
963 JH/42 Internationalisation: change the default for downconversion in the smtp
964 transport to be "if needed". Previously it was "as previously set" for
965 the message, which usually meant "if needed" for message-submission but
966 "no" for everything else. However, MTAs have been seen using SMTPUTF8
967 even when the envelope addresses did not need it, resulting in forwarding
968 failures to non-supporting MTAs. A downconvert in such cases will be
969 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
970 transport. The change does mean that addresses needing conversion will
971 be converted when previously a delivery failure would occur.
973 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
974 response was received it would be used unchecked in a fail-DSN, violating
975 standards on line-length limits. Truncate if needed.
977 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
978 doesn't work. (Additionally add a new main config option to configure the
985 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
986 side, to decode a TLS 1.3 packet capture.
988 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
989 Previously the default library behaviour applied, sending two, each in
992 JH/03 Debug output for ACL now gives the config file name and line number for
995 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
997 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
999 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1000 buffer overrun for (non-chunking) other transports.
1002 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1003 TLS1.3, means that a server rejecting a client certificate is not visible
1004 to the client until the first read of encrypted data (typically the
1005 response to EHLO). Add detection for that case and treat it as a failed
1006 TLS connection attempt, so that the normal retry-in-clear can work (if
1007 suitably configured).
1009 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1010 and/or domain. Found and fixed by Jason Betts.
1012 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1013 configuration). If a CNAME target was not a wellformed name pattern, a
1016 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1017 the OS reports them interleaved with other addresses.
1019 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1020 used both for input and for a verify callout, both encrypted, SMTP
1021 responses being sent by the server could be lost. This resulted in
1022 dropped connections and sometimes bounces generated by a peer sending
1025 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1026 string. Previously it was possible to cause undefined behaviour in a
1027 library routine (usually a crash). Found by "zerons".
1029 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1032 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1033 API was removed, so update to use the newer ones.
1035 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1036 any timeout set, is taking a long time. Previously we would hang on to a
1037 rotated logfile "forever" if the input was arriving with long gaps
1038 (a previous attempt to fix addressed lack, for a long time, of initial
1041 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1042 shared (NFS) environment. The length of the tempfile name is now
1043 4 + 16 ("hdr.$message_exim_id") which might break on file
1044 systems which restrict the file name length to lower values.
1045 (It was "hdr.$pid".)
1047 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1048 shared (NFS) environment.
1050 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1051 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1054 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1055 on some platforms for bit 31.
1057 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1058 to changes apparently associated with TLS1.3 handling some of the APIs
1059 previously used were either nonfunctional or inappropriate. Strings
1060 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1061 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1062 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1063 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1064 and the use of specific cipher names in the encrypted= ACL condition.
1066 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1068 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1069 verification result was not updated unless hosts_require_ocsp applied.
1071 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1072 queue_list_requires_admin set to false, non-admin users were denied the
1075 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1076 directory-of-certs mode. Previously they were advertised despite the
1079 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1080 A single TCP connection by a client will now hold a TLS connection open
1081 for multiple message deliveries, by default. Previously the default was to
1084 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1085 default. If built with the facility, DANE will be used. The facility
1086 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1088 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1089 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1090 must be defined and you must still, unless you define DISABLE_TLS, manage
1091 the the include-dir and library-file requirements that go with that
1092 choice. Non-TLS builds are still supported.
1094 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1097 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1098 default. If the platform supports and has the facility enabled, it will
1099 be requested on all coneections.
1101 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1102 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1104 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1106 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1107 success-DSN messages. Previously the From: header was always the default
1108 one for these; the option was ignored.
1110 JH/28 Fix the timeout on smtp response to apply to the whole response.
1111 Previously it was reset for every read, so a teergrubing peer sending
1112 single bytes within the time limit could extend the connection for a
1113 long time. Credit to Qualsys Security Advisory Team for the discovery.
1115 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1116 delivery address, which leaked information of the results of local
1117 forwarding. Change to the original envelope recipient address, per
1120 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1121 requested. Previously not bounce was generated and a log entry of
1122 error ignored was made.
1124 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1126 JH/32 Introduce a general tainting mechanism for values read from the input
1127 channel, and values derived from them. Refuse to expand any tainted
1128 values, to catch one form of exploit.
1130 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1131 was unused and the unexpanded text used for the test. Found and
1132 fixed by Ruben Jenster.
1134 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1135 an attempt to use a TLS library read routine dereffed a nul pointer,
1138 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1139 them in our smtp response.
1141 JH/36 Have the general router option retry_use_local_part default to true when
1142 any of the restrictive preconditions are set (to anything). Previously it
1143 was only for check_local user. The change removes one item of manual
1144 configuration which is required for proper retries when a remote router
1145 handles a subset of addresses for a domain.
1147 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1148 link count into consideration.
1150 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1151 caused the extension of big_buffer, the following lines were ignored.
1153 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1154 accordance with RFC 2308. Previously there was no expiry, so a longlived
1155 receive process (eg. due to ACL delays) versus a short SOA value could
1158 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1160 JH/39 Promote DMARC support to mainline.
1162 JH/40 Bug 2452: Add a References: header to DSNs.
1164 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1165 parameters. The relevant library call is documented as "Deprecated: This
1166 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1167 3.6.0, DH parameters are negotiated following RFC7919."
1169 HS/06 Change the default of dnssec_request_domains to "*"
1171 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1172 carried on and emitted a BDAT command, even when PIPELINING was not
1175 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1176 buffer was used for the filename, resulting in a trap when tainted
1177 arguments (eg. $domain) were used.
1179 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1180 recommended to avoid a possible server-load attack. The feature can be
1181 re-enabled via the openssl_options main cofiguration option.
1183 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1184 for version 4.90 - adding a "more data" boolean to the arguments.
1185 Bumped the ABI version number also, this having been missed previously;
1186 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1187 respect of usage of smtp_printf() by either local_scan code or libraries
1188 accessed via the ${dlfunc } expansion item. Both will need coding
1189 adjustment for any calls to smtp_printf() to match the new function
1190 signature; a FALSE value for the new argument is always safe.
1192 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1193 the file-offset (which the Linux syscall does, and exim expects); this
1194 resulted in an indefinite loop.
1196 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1197 to do ARC verification. The Authentication-Results: header line added
1198 by the configuration then had no ARC item.
1200 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1201 an incorrect mode was used for reading the result, resulting in it being
1208 JH/01 Remove code calling the customisable local_scan function, unless a new
1209 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1211 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1212 non-signal-safe functions being used.
1214 JH/03 Bug 2269: When presented with a received message having a stupidly large
1215 number of DKIM-Signature headers, disable DKIM verification to avoid
1216 a resource-consumption attack. The limit is set at twenty.
1218 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1219 report of oldest_pass in ${authres } in consequence, and separate out
1220 some descriptions of reasons for verification fail.
1222 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1223 files in the spool were present and unlocked. A queue-runner could spot
1224 them, resulting in a duplicate delivery. Fix that by doing the unlock
1225 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1226 add more error-checking on spoolfile handling while that code is being
1229 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1230 No known attacks, no CVE, this is defensive hardening.
1232 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1233 a queue-runner could start a delivery while other operations were ongoing.
1234 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1235 Found and investigated by Tim Stewart. Fix by using the open message data
1236 file handle rather than opening another, and not locally closing it (which
1237 releases a lock) for that case, while creating the temporary .eml format
1238 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1240 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1241 $sender_verify_failure/$recipient_verify_failure to "random".
1243 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1246 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1247 Previously this would segfault.
1249 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1252 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1253 like zero, since the resolver should be doing this for us, But we need one
1254 as a CNAME but no MX presence gets the CNAME returned; we need to check
1255 that doesn't point to an MX to declare it "no MX returned" rather than
1256 "error, loop". A new main option is added so the older capability of
1257 following some limited number of chain links is maintained.
1259 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1261 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1262 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1263 now-deprecated earlier definitions used only specified the range up to TLS
1264 1.2 (in the older-version library docs).
1266 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1268 JH/15 Rework TLS client-side context management. Stop using a global, and
1269 explicitly pass a context around. This enables future use of TLS for
1270 connections to service-daemons (eg. malware scanning) while a client smtp
1271 connection is using TLS; with cutthrough connections this is quite likely.
1273 JH/16 Fix ARC verification to do AS checks in reverse order.
1275 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1277 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1278 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1279 Previously the "utf8" would be re-prepended for every additional message.
1281 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1282 Previously thery were accepted, resulting in issues when attempting to
1283 forward messages to a non-supporting MTA.
1285 PP/02 Let -n work with printing macros too, not just options.
1287 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1288 one parent address was copied, and bogus data was used at delivery-logging
1289 time. Either a crash (after delivery) or bogus log data could result.
1290 Discovery and analysis by Tim Stewart.
1292 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1293 Previously if the string ended mid-character, we did not insert the
1294 promised '?' replacement.
1296 PP/04 Documentation: current string operators work on bytes, not codepoints.
1298 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1299 should pack well giving a smaller memory footprint so better caching and
1300 therefore performance. Group the declarations where this can't be done so
1301 that the byte-sized flag variables are not interspersed among pointer
1302 variables, giving a better chance of good packing by the compiler.
1304 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1305 non-null, to avoid issues with sites running BATV. Previously reports were
1306 sent with an empty envelope sender so looked like bounces.
1308 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1309 The ignore_error flag wasn't being returned from the filter subprocess so
1310 was not set for later routers. Investigation and fix by Matthias Kurz.
1312 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1313 and a msg:complete for the whole, when a message is manually removed using
1314 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1316 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1317 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1318 hope that is portable enough.
1320 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1321 requiring a known-CA anchor certificate; make it now rely entirely on the
1322 TLSA as an anchor. Checking the name on the leaf cert against the name
1323 on the A-record for the host is still done for TA (but not for EE mode).
1325 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1326 included in delivery lines for non-proxied connections, when compiled with
1327 SUPPORT_SOCKS and running with proxy logging enabled.
1329 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1330 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1331 move the existing event to fire before the normal logging of message
1332 failure so that custom logging is bracketed by normal logging.
1334 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1335 msg:fail:internal event. Developement by Matthias Kurz.
1337 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1338 far too small for todays use of crypto signatures stored there. Go all
1339 the way to the max DNS message size of 64kB, even though this might be
1340 overmuch for IOT constrained device use.
1342 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1343 copy a string over itself. The library routine is documented as not
1344 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1346 JH/32 For main options check_spool_space and check_inode_space, where the
1347 platform supports 64b integers, support more than the previous 2^31 kB
1348 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1349 the previous G, M, k.
1351 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1352 $authenticated_fail_id variable on authentication failure. Previously
1355 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1356 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1357 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1358 GNUTLS_SEC_PARAM_MEDIUM.
1360 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1361 side. Previously we would continue as if no SNI had been received.
1363 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1364 "<" character, which should be a list-separator specification, we walked
1365 off past the nul-terimation.
1367 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1368 causes) even when the retry time is not yet met. Previously they were
1369 not, meaning that when (say) an account was over-quota and temp-rejecting,
1370 and multiple senders' messages were queued, only one sender would get
1371 notified on each configured delay_warning cycle.
1373 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1375 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1376 cipher-suites, an error can be left on the stack even for a succeeding
1377 accept; this results in impossible error messages when a later operation
1380 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1381 return error codes indicating retry. Under TLS1.3 this becomes required.
1383 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1384 it only wrote the new authenticators, resulting in a lack of tracking of
1385 peer changes of ESMTP extensions until the next cache flush.
1387 JH/41 Fix the loop reading a message header line to check for integer overflow,
1388 and more-often against header_maxsize. Previously a crafted message could
1389 induce a crash of the recive process; now the message is cleanly rejected.
1391 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1392 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1398 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1399 When redis_servers is set to a list of > 1 element, and the Redis servers
1400 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1401 case of MOVED into a DEFER case instead, thus moving the query onto the
1402 next server in the list. For a cluster of N elements, all N servers must
1403 be defined in redis_servers.
1405 GF/02 Catch and remove uninitialized value warning in exiqsumm
1406 Check for existence of @ARGV before looking at $ARGV[0]
1408 JH/01 Replace the store_release() internal interface with store_newblock(),
1409 which internalises the check required to safely use the old one, plus
1410 the allocate and data copy operations duplicated in both (!) of the
1411 extant use locations.
1413 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1414 modifier. This matches the restriction on the commandline.
1416 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1417 Previously only the last row was returned.
1419 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1420 we assumed that tags in the header were well-formed, and parsed the
1421 element content after inspecting only the first char of the tag.
1422 Assumptions at that stage could crash the receive process on malformed
1425 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1426 While running the DKIM ACL we operate on the Permanent memory pool so that
1427 variables created with "set" persist to the DATA ACL. Also (at any time)
1428 DNS lookups that fail create cache records using the Permanent pool. But
1429 expansions release any allocations made on the current pool - so a dnsdb
1430 lookup expansion done in the DKIM ACL releases the memory used for the
1431 DNS negative-cache, and bad things result. Solution is to switch to the
1432 Main pool for expansions.
1433 While we're in that code, add checks on the DNS cache during store_reset,
1434 active in the testsuite.
1435 Problem spotted, and debugging aided, by Wolfgang Breyha.
1437 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1438 When none of the hosts presented to a transport match an already-open
1439 connection, close it and proceed with the list. Previously we would
1440 queue the message. Spotted by Lena with Yahoo, probably involving
1443 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1444 Previously a spurious "250 OK id=" response was appended to the proper
1447 JH/08 The "support for" informational output now, which built with Content
1448 Scanning support, has a line for the malware scanner interfaces compiled
1449 in. Interface can be individually included or not at build time.
1451 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1452 by the template makefile "src/EDITME". The "STREAM" support for an older
1453 ClamAV interface method is removed.
1455 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1456 rows affected is given instead).
1458 JH/11 The runtime Berkeley DB library version is now additionally output by
1459 "exim -d -bV". Previously only the compile-time version was shown.
1461 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1462 SMTP connection. Previously, when one had more recipients than the
1463 first, an abortive onward connection was made. Move to full support for
1464 multiple onward connections in sequence, handling cutthrough connection
1465 for all multi-message initiating connections.
1467 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1468 routers. Previously, a multi-recipient message would fail to match the
1469 onward-connection opened for the first recipient, and cause its closure.
1471 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1472 a timeout on read on a GnuTLS initiating connection, resulting in the
1473 initiating connection being dropped. This mattered most when the callout
1474 was marked defer_ok. Fix to keep the two timeout-detection methods
1477 JH/15 Relax results from ACL control request to enable cutthrough, in
1478 unsupported situations, from error to silently (except under debug)
1479 ignoring. This covers use with PRDR, frozen messages, queue-only and
1482 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1484 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1485 metadata, resulting in a crash in free().
1487 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1488 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1489 Broken also in d185889f4, with init system revamp.
1491 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1492 Previously we abruptly closed the connection after reading a malware-
1493 found indication; now we go on to read the "scan ok" response line,
1496 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1497 ACL. Previously, a crash would result.
1499 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1500 macro text after a replacement (previously it was only once per line) and
1501 by skipping builtin macros when searching for an uppercase lead character.
1503 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1504 for the build is renamed.
1506 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1507 was allocated for every new TLS startup, meaning one per message. Fix
1508 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1510 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1511 reported the original. Fix to report (as far as possible) the ACL
1512 result replacing the original.
1514 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1515 OpenSSL. Certificate information is loaded for every new TLS startup,
1516 and the resources needed to be freed.
1518 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1520 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1523 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1524 DATA response info to the (existing) per-recipient response info for
1525 the "C=" log element. It can have useful tracking info from the
1526 destination system. Patch from Simon Arlott.
1528 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1529 length value. Previously this would segfault.
1531 HS/02 Support Avast multiline protoocol, this allows passing flags to
1532 newer versions of the scanner.
1534 JH/28 Ensure that variables possibly set during message acceptance are marked
1535 dead before release of memory in the daemon loop. This stops complaints
1536 about them when the debug_store option is enabled. Discovered specifically
1537 for sender_rate_period, but applies to a whole set of variables.
1538 Do the same for the queue-runner and queue-list loops, for variables set
1539 from spool message files. Do the same for the SMTP per-message loop, for
1540 certain variables indirectly set in ACL operations.
1542 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1543 as a multi-recipient message from a mailinglist manager). The coding had
1544 an arbitrary cutoff number of characters while checking for more input;
1545 enforced by writing a NUL into the buffer. This corrupted long / fast
1546 input. The problem was exposed more widely when more pipelineing of SMTP
1547 responses was introduced, and one Exim system was feeding another.
1548 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1549 receiving system, and refused recipients seen by the sending system
1550 (propating to people being dropped from mailing lists).
1551 Discovered and pinpointed by David Carter.
1553 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1554 replaced by the ${authresults } expansion.
1556 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1558 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1559 allows proper process termination in container environments.
1561 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1562 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1564 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1565 and "err_perm", deprecated since 4.83 when the RFC-defined words
1566 "temperror" and "permerror" were introduced.
1568 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1569 transport-filters or DKIM-signing. The restriction was lost in the
1570 consolidation of verify-callout and delivery SMTP handling.
1571 Extend the restriction to also cover ARC-signing.
1573 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1574 in defer=pass mode supply a 450 to the initiator. Previously the message
1577 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1578 tls_require_ciphers is used as before.
1580 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1581 "pass_unscanned". Only tmpfails from the scanner are written to
1582 the paniclog, as they may require admin intervention (permission
1583 denied, license issues). Other scanner errors (like decompression
1584 bombs) do not cause a paniclog entry.
1586 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1587 Previously it was possible to log spurious information in receive log
1590 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1591 triggered odd behaviour from Outlook Express clients.
1593 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1596 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1597 since the IETF WG has not yet settled on that versus the original
1598 "bare" representation.
1600 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1601 Previously the millisecond value corrupted the output.
1602 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1603 corrupted the output.
1609 JH/01 Rework error string handling in TLS interface so that the caller in
1610 more cases is responsible for logging. This permits library-sourced
1611 string to be attached to addresses during delivery, and collapses
1612 pairs of long lines into single ones.
1614 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1615 during configuration. Wildcards are allowed and expanded.
1617 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1618 This permits better logging.
1620 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1621 do not close it down and have the child transport start it up again on
1622 the passed-on TCP connection. Instead, proxy the child (and any
1623 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1624 affected: the continued delivery log lines do not have any DNSSEC, TLS
1625 Certificate or OCSP information. TLS cipher information is still logged.
1627 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1628 identical IP addresses on different listening ports. Will also affect
1631 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1632 add noisy ifdef guards to special-case this sillyness.
1633 Patch from Bernd Kuhls.
1635 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1636 than 255 are no longer allowed.
1638 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1639 Disable the session-cache too, which might reduce our load. Since we
1640 currrectly use a new context for every connection, both as server and
1641 client, there is no benefit for these.
1642 GnuTLS appears to not support tickets server-side by default (we don't
1643 call gnutls_session_ticket_enable_server()) but client side is enabled
1644 by default on recent versions (3.1.3 +) unless the PFS priority string
1647 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1648 <https://reproducible-builds.org/specs/source-date-epoch/>.
1650 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1651 the check for any unsuccessful recipients did not notice the limit, and
1652 erroneously found still-pending ones.
1654 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1655 MSG_MORE. Only in-clear (not on TLS connections).
1657 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1658 transport-filter is involved we can buffer the headers in memory for
1659 creating the signature, and read the spool data file once for the
1660 signature and again for transmission.
1662 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1663 4.77 as the kernel support then wasn't solid, having issues in 64bit
1664 mode. Now, it's been long enough. Add support for FreeBSD also.
1666 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1667 case where the routing stage had gathered several addresses to send to
1668 a host before calling the transport for the first, we previously failed
1669 to close down TLS in the old transport process before passing the TCP
1670 connection to the new process. The new one sent a STARTTLS command
1671 which naturally failed, giving a failed delivery and bloating the retry
1672 database. Investigation and fix prototype from Wolfgang Breyha.
1674 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1675 false-negatives in the check that the sender had not preempted a response
1676 or prompt from Exim (running as a server), due to that code's lack of
1677 awareness of the SMTP input buffering.
1679 PP/04 Add commandline_checks_require_admin option.
1680 Exim drops privileges sanely, various checks such as -be aren't a
1681 security problem, as long as you trust local users with access to their
1682 own account. When invoked by services which pass untrusted data to
1683 Exim, this might be an issue. Set this option in main configuration
1684 AND make fixes to the calling application, such as using `--` to stop
1687 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1688 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1689 responses to those, into a single TLS record each way (this usually means
1690 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1693 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1694 affects you only if you're dancing at the edge of the param size limits.
1695 If you are, and this message makes sense to you, then: raise the
1696 configured limit or use OpenSSL 1.1. Nothing we can do for older
1699 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1700 cmdline element to get the documented default one. Previously it was
1703 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1704 are unsuable for non-message-specific reasons - eg. connection timeout,
1705 banner-time rejection.
1707 JH/16 Fix logging of delivery remote port, when specified by router, under
1710 PP/06 Repair manualroute's ability to take options in any order, even if one
1711 is the name of a transport.
1714 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1716 JH/17 Change the list-building routines interface to use the expanding-string
1717 triplet model, for better allocation and copying behaviour.
1719 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1720 Previously it was constructed the first time a possibly-matching string
1721 was met in the configuration file input during startup; now it is done
1724 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1725 compatible one, to avoid the (poorly documented) possibility of a config
1726 file in the working directory redirecting the DB files, possibly correpting
1727 some existing file. CVE-2017-10140 assigned for BDB.
1729 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1730 cache-hot. Previously, although the result was properly cached, the
1731 initial verify call returned a defer.
1733 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1734 the main verify for receipient in uncached-mode.
1736 JH/22 Retire historical build files to an "unsupported" subdir. These are
1737 defined as "ones for which we have no current evidence of testing".
1739 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1740 if present. Previously it was ignored.
1742 JH/24 Start using specified-initialisers in C structure init coding. This is
1743 a C99 feature (it's 2017, so now considered safe).
1745 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1746 if was a fixed-sized field and bitmask ops via macros; it is now more
1749 PP/07 GitHub PR 56: Apply MariaDB build fix.
1750 Patch provided by Jaroslav Škarvada.
1752 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1753 during Coverity cleanups [4.87 JH/47]
1754 Diagnosis and fix provided by Michael Fischer v. Mollard.
1756 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1757 the right size to place the terminating semicolon on its own folded
1758 line, the header hash was calculated to an incorrect value thanks to
1759 the (relaxed) space the fold became.
1761 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1762 and confused the parent.
1764 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1765 which could crash as a result. This could lead to undeliverable messages.
1767 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1770 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1771 matters on fast-turnover and PID-randomising systems, which were getting
1772 out-of-order delivery.
1774 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1775 a possibly-overlapping copy. The symptom was that "Remote host closed
1776 connection in response to HELO" was logged instead of the actual 4xx
1779 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1780 Previously only that bufferd was discarded, resulting in SYMTP command
1783 JH/32 DKIM: when a message has multiple signatures matching an identity given
1784 in dkim_verify_signers, run the dkim acl once for each. Previously only
1785 one run was done. Bug 2189.
1787 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1788 "panic the current process" to "deliberately defer". The panic log is
1789 still written with the problem list name; the mail and reject logs now
1790 get a temp-reject line for the message that was being handled, saying
1791 something like "domains check lookup or other defer". The SMTP 451
1792 message is still "Temporary local problem".
1794 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1795 A crafted sequence of BDAT commands could result in in-use memory beeing
1796 freed. CVE-2017-16943.
1798 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1799 from SMTP input. Previously it was always done; now only done for DATA
1800 and not BDAT commands. CVE-2017-16944.
1802 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1803 to the message (such as an overlong header line). Previously this was
1804 not done and we did not exit BDAT mode. Followon from the previous item
1805 though a different problem.
1811 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1812 than -2003 did; needs libidn2 in addition to libidn.
1814 JH/02 The path option on a pipe transport is now expanded before use.
1816 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1817 Patch provided by "Björn", documentation fix added too.
1819 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1820 missing a wire-to-host endian conversion.
1822 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1823 close after a BDAT command line could be taken as a following command,
1824 giving a synch failure. Fix by only checking for synch immediately
1825 before acknowledging the chunk.
1827 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1828 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1829 macro. Patches provided by Josh Soref.
1831 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1832 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1833 by IANA as a service extension. However, John Klensin suggests that we
1836 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1837 the dkim code may be unix-mode line endings rather than smtp wire-format
1838 CRLF, so prepend a CR to any bare LF.
1840 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1841 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1843 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1844 the first were themselves being wrongly included in the feed into dkim
1845 processing; with most chunk sizes in use this resulted in an incorrect
1846 body hash calculated value.
1848 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1849 DKIM signature block, for verification. Although advised against by
1850 standards it is specifically not ruled illegal.
1852 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1854 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1855 missing a body hash (the bh= tag).
1857 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1858 It seems that HAProxy sends the Proxy Protocol information in clear and
1859 only then does a TLS startup, so do the same.
1861 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1862 TCP connections (such as for Spamd) unless the daemon successfully set
1863 Fast Open mode on its listening sockets. This fixes breakage seen on
1864 too-old kernels or those not configured for Fast Open, at the cost of
1865 requiring both directions being enabled for TFO, and TFO never being used
1866 by non-daemon-related Exim processes.
1868 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1869 endings, at least on the first header line. Try to canonify any that get
1870 past that check, despite the cost.
1872 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1873 now limited to an arbitrary five deep, while parsing addresses with the
1874 strip_excess_angle_brackets option enabled.
1876 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1877 instead leave the unprompted TLS handshake in socket buffer for the
1878 TLS library to consume.
1880 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1882 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1884 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1885 region in one operation, for speed, and this leaves assigned pointers
1886 dangling. Add checks run only under the testsuite which checks all
1887 variables at a store-reset and panics on a dangling pointer; add code
1888 explicitly nulling out all the variables discovered. Fixes one known
1889 bug: a transport crash, where a dangling pointer for $sending_ip_address
1890 originally assigned in a verify callout, is re-used.
1892 PP/06 Drop '.' from @INC in various Perl scripts.
1894 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1896 PP/08 Reduce a number of compilation warnings under clang; building with
1897 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1898 should be warning-free.
1900 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1902 HS/01 Fix portability problems introduced by PP/08 for platforms where
1903 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1905 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1906 chunk. This allows us to accept broken chunked messages. We need a more
1907 general solution here.
1909 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1910 already-broken messages in the queue.
1912 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1914 JH/19 Fix reference counting bug in routing-generated-address tracking.
1920 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1921 supports it and a size is available (ie. the sending peer gave us one).
1923 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1924 years of being deprecated). The replacements are the ACLs
1925 acl_smtp_mime and acl_not_smtp_mime.
1927 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1928 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1929 or even an in-clear connection were permitted. Now, if the host lookup
1930 was dnssec and dane was requested then the host is only used if the
1931 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1932 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1933 if one fails this test.
1934 This means that a poorly-configured remote DNS will make it incommunicado;
1935 but it protects against a DNS-interception attack on it.
1937 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1938 non-noisy when a race steals the message being considered.
1940 JH/05 If main configuration option tls_certificate is unset, generate a
1941 self-signed certificate for inbound TLS connections.
1943 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1944 in rewrites and routers.
1946 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1947 and logged a warning sing 4.83; now they are a configuration file error.
1949 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1950 (lacking @domain). Apply the same qualification processing as RCPT.
1952 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1954 JH/10 Support ${sha256:} applied to a string (as well as the previous
1957 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1958 a cutthrough deliver is pending, as we always want to make a connection.
1959 This also avoids re-routing the message when later placing the cutthrough
1960 connection after a verify cache hit.
1961 Do not update it with the verify result either.
1963 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1964 when routing results in more than one destination address.
1966 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1967 signing (which inhibits the cutthrough capability). Previously only
1968 the presence of an option was tested; now an expansion evaluating as
1969 empty is permissible (obviously it should depend only on data available
1970 when the cutthrough connection is made).
1972 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1973 the relevant preceding SMTP command did not note the pipelining mode.
1975 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1976 Previously they were not counted.
1978 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1979 as one having no matching records. Previously we deferred the message
1980 that needed the lookup.
1982 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1983 distinguished as "(=".
1985 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1986 for missing MX records. Previously it only worked for missing A records.
1988 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1990 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1991 after the data-go-ahead and data-ack. Patch from Jason Betts.
1993 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1994 even for a "none" policy. Patch from Tony Meyer.
1996 JH/22 Fix continued use of a connection for further deliveries. If a port was
1997 specified by a router, it must also match for the delivery to be
2000 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2001 When one of the recipients of a message was unsuitable for the connection
2002 (has no matching addresses), we lost track of needing to mark it
2003 deferred. As a result mail would be lost.
2005 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2007 JH/25 Decoding ACL controls is now done using a binary search; the source code
2008 takes up less space and should be simpler to maintain. Merge the ACL
2009 condition decode tables also, with similar effect.
2011 JH/26 Fix problem with one_time used on a redirect router which returned the
2012 parent address unchanged. A retry would see the parent address marked as
2013 delivered, so not attempt the (identical) child. As a result mail would
2016 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2017 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2018 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2021 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2022 Main config options check_{log,spool}_{inodes,space} are now
2023 100 inodes, 10MB unless set otherwise in the configuration.
2025 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2026 Previously it only applied to the main-section connection policy
2029 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2031 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2032 by me. Added RFC7919 DH primes as an alternative.
2034 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2035 are not in the system include path.
2037 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2038 GnuTLS, when a session startup failed (eg because the client disconnected)
2039 Exim did stdio operations after fclose. This was exposed by a recent
2040 change which nulled out the file handle after the fclose.
2042 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2043 signed directly by the cert-signing cert, rather than an intermediate
2044 OCSP-signing cert. This is the model used by LetsEncrypt.
2046 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2048 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2049 an incoming connection.
2051 HS/02 Bug 1802: Do not half-close the connection after sending a request
2054 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2055 fallback to "prime256v1".
2057 JH/34 SECURITY: Use proper copy of DATA command in error message.
2058 Could leak key material. Remotely exploitable. CVE-2016-9963.
2064 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2065 and 3.4.4 - once the server is enabled to respond to an OCSP request
2066 it does even when not requested, resulting in a stapling non-aware
2067 client dropping the TLS connection.
2069 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2070 support variable-length bit vectors. No functional change.
2072 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2073 The I= interface field on outgoing lines is now after the H= remote
2074 host field, same as incoming lines. There is a separate
2075 outgoing_interface log selector which allows you to disable the
2078 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2079 If not running log_selector +smtp_connection the mainlog would be held
2080 open indefinitely after a "too many connections" event, including to a
2081 deleted file after a log rotate. Leave the per net connection logging
2082 leaving it open for efficiency as that will be quickly detected by the
2083 check on the next write.
2085 HS/01 Bug 1671: Fix post transport crash.
2086 Processing the wait-<transport> messages could crash the delivery
2087 process if the message IDs didn't exist for some reason. When
2088 using 'split_spool_directory=yes' the construction of the spool
2089 file name failed already, exposing the same netto behaviour.
2091 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2092 mime_regex ACL conditions.
2094 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2095 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2096 to HELO, local diagnostic string.
2098 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2099 log line to debug. Even when configured with a tls authenticator many
2100 client connections are expected to not authenticate in this way, so
2101 an authenticate fail is not an error.
2103 HS/02 Add the Exim version string to the process info. This way exiwhat
2104 gives some more detail about the running daemon.
2106 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2107 matter for fast-change records such as DNSBLs.
2109 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2110 retry record, even if constant. There may be multiple transports with
2111 different interface settings and the retry behaviour needs to be kept
2114 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2116 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2118 JH/10 Bug 840: fix log_defer_output option of pipe transport
2120 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2121 hosts_randomize. This matters a lot when combined with mua_wrapper.
2123 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2124 ${quote_pgsql:<string>} operator.
2126 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2128 JH/14 Change header file name for freeradius-client. Relevant if compiling
2129 with Radius support; from the Gentoo tree and checked under Fedora.
2131 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2133 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2134 option transport_filter, meaning no filtering.
2136 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2138 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2139 defaults to "*" (all hosts). The variable is now available when not built
2140 with TLS, default unset, mainly to enable keeping the testsuite sane.
2141 If a server certificate is not supplied (via tls_certificate) an error is
2142 logged, and clients will find TLS connections fail on startup. Presumably
2143 they will retry in-clear.
2144 Packagers of Exim are strongly encouraged to create a server certificate
2145 at installation time.
2147 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2148 with the $config_file variable.
2150 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2151 in transport context, after the attempt, and per-recipient. The latter type
2152 is per host attempted. The event data is the error message, and the errno
2153 information encodes the lookup type (A vs. MX) used for the (first) host,
2154 and the trailing two digits of the smtp 4xx response.
2156 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2157 to write to mainlog (or rejectlog, paniclog) in the window between file
2158 creation and permissions/ownership being changed. Particularly affects
2159 installations where exicyclog is run as root, rather than exim user;
2160 result is that the running daemon panics and dies.
2162 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2164 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2165 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2166 "pri" and "weight". Note that the previous implicit priority given by the
2167 list order is no longer honoured.
2169 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2170 for DKIM processing.
2172 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2173 by defining SUPPORT_SOCKS.
2175 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2176 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2177 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2178 variables are renamed to proxy_{local,external}_{address,port}.
2180 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2181 for a build by defining SUPPORT_I18N
2183 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2184 of the query string, and make ${quote_redis:} do that quoting.
2186 JH/29 Move Events support from Experimental to mainline, enabled by default
2187 and removable for a build by defining DISABLE_EVENT.
2189 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2191 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2192 cached by the daemon.
2194 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2195 by defining LOOKUP_REDIS. The libhiredis library is required.
2197 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2198 keys are given for lookup.
2200 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2201 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2202 only supported when built with TLS support. The PolarSSL SHA routines
2203 are still used when the TLS library is too old for convenient support.
2205 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2206 openssl_options), for security. OpenSSL forces this from version 1.1.0
2207 server-side so match that on older versions.
2209 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2210 allocation for $value could be released as the expansion processing
2211 concluded, but leaving the global pointer active for it.
2213 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2214 and to use the domains and local_parts ACL conditions.
2216 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2217 incorrectly not doubled on cutthrough transmission, hence seen as a
2218 body-termination at the receiving system - resulting in truncated mails.
2219 Commonly the sender saw a TCP-level error, and retransmitted the message
2220 via the normal store-and-forward channel. This could result in duplicates
2221 received - but deduplicating mailstores were liable to retain only the
2222 initial truncated version.
2224 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2226 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2228 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2229 we're in there, support oversigning also; bug 1309.
2231 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2233 HS/04 Add support for keep_environment and add_environment options.
2235 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2236 either intentional arithmetic overflow during PRNG, or testing config-
2239 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2240 delivery resulted in actual delivery. Cancel cutthrough before DATA
2243 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2244 rejects a recipient: pass the reject to the originator.
2246 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2247 Many were false-positives and ignorable, but it's worth fixing the
2250 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2251 for the new environment-manipulation done at startup. Move the routines
2252 from being local to tls.c to being global via the os.c file.
2254 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2255 an extract embedded as result-arg for a map, the first arg for extract
2256 is unavailable so we cannot tell if this is a numbered or keyed
2257 extraction. Accept either.
2263 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2266 JH/02 The smtp transport option "multi_domain" is now expanded.
2268 JH/03 The smtp transport now requests PRDR by default, if the server offers
2271 JH/04 Certificate name checking on server certificates, when exim is a client,
2272 is now done by default. The transport option tls_verify_cert_hostnames
2273 can be used to disable this per-host. The build option
2274 EXPERIMENTAL_CERTNAMES is withdrawn.
2276 JH/05 The value of the tls_verify_certificates smtp transport and main options
2277 default to the word "system" to access the system default CA bundle.
2278 For GnuTLS, only version 3.0.20 or later.
2280 JH/06 Verification of the server certificate for a TLS connection is now tried
2281 (but not required) by default. The verification status is now logged by
2282 default, for both outbound TLS and client-certificate supplying inbound
2285 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2288 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2289 Status Notification (bounce) messages are now MIME format per RFC 3464.
2290 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2291 under the control of the dsn_advertise_hosts option, and routers may
2292 have a dsn_lasthop option.
2294 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2295 default, modifiable by a malware= option. The list separator for
2296 the options can now be changed in the usual way. Bug 68.
2298 JH/10 The smtp_receive_timeout main option is now expanded before use.
2300 JH/11 The incoming_interface log option now also enables logging of the
2301 local interface on delivery outgoing connections.
2303 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2304 if the interface and destination host and port all match.
2306 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2309 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2310 Patch from Andrew Lewis.
2312 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2313 now supports optional time-restrictions, weighting, and priority
2314 modifiers per server. Patch originally by <rommer@active.by>.
2316 JH/16 The spamd_address main option now supports a mixed list of local
2317 and remote servers. Remote servers can be IPv6 addresses, and
2318 specify a port-range.
2320 JH/17 Bug 68: The spamd_address main option now supports an optional
2321 timeout value per server.
2323 JH/18 Bug 1581: Router and transport options headers_add/remove can
2324 now have the list separator specified.
2326 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2329 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2332 JH/21 Support for the A6 type of dns record is withdrawn.
2334 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2335 rather than the verbs used.
2337 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2338 from 255 to 1024 chars.
2340 JH/24 Verification callouts now attempt to use TLS by default.
2342 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2343 are generic router options now. The defaults didn't change.
2345 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2346 Original patch from Alexander Shikoff, worked over by JH.
2348 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2349 files if ZCAT_COMMAND is not executable.
2351 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2353 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2355 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2356 Normally benign, it bites when the pair was led to by a CNAME;
2357 modern usage is to not canonicalize the domain to a CNAME target
2358 (and we were inconsistent anyway for A-only vs AAAA+A).
2360 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2362 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2363 when evaluating $sender_host_dnssec.
2365 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2366 $sender_helo_dnssec variable.
2368 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2370 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2372 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2374 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2375 documented as working, but never had. Support all but $spam_report.
2377 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2378 added for tls authenticator.
2380 HS/03 Add perl_taintmode main config option
2386 TL/01 When running the test suite, the README says that variables such as
2387 no_msglog_check are global and can be placed anywhere in a specific
2388 test's script, however it was observed that placement needed to be near
2389 the beginning for it to behave that way. Changed the runtest perl
2390 script to read through the entire script once to detect and set these
2391 variables, reset to the beginning of the script, and then run through
2392 the script parsing/test process like normal.
2394 TL/02 The BSD's have an arc4random API. One of the functions to induce
2395 adding randomness was arc4random_stir(), but it has been removed in
2396 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2397 function when detected.
2399 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2400 cause callback expansion.
2402 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2403 syntax errors in an expansion can be treated as a string instead of
2404 logging or causing an error, due to the internal use of bool_lax
2405 instead of bool when processing it.
2407 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2408 server certificates when making smtp deliveries.
2410 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2412 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2414 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2416 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2417 Merged patch from Sebastian Wiedenroth.
2419 JH/05 Fix results-pipe from transport process. Several recipients, combined
2420 with certificate use, exposed issues where response data items split
2421 over buffer boundaries were not parsed properly. This eventually
2422 resulted in duplicates being sent. This issue only became common enough
2423 to notice due to the introduction of connection certificate information,
2424 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2426 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2427 size buffer was used, resulting in syntax errors when an expansion
2430 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2431 version 3.3.6 or later.
2433 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2434 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2435 both become "event_action", the variables become $event_name, $event_data
2436 and $event_defer_errno. There is a new variable $verify_mode, usable in
2437 routers, transports and related events. The tls:cert event is now also
2438 raised for inbound connections, if the main configuration event_action
2441 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2442 early OCSP support, but no stapling (appears to be less than 1.0.0).
2444 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2445 server certificate names available under the smtp transport option
2446 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2449 JH/10 Time-related extraction expansions from certificates now use the main
2450 option "timezone" setting for output formatting, and are consistent
2451 between OpenSSL and GnuTLS compilations. Bug 1541.
2453 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2454 encoded parameter in the incoming message. Bug 1558.
2456 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2457 include certificate info, eximon was claiming there were spoolfile
2460 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2462 JH/14 Log delivery-related information more consistently, using the sequence
2463 "H=<name> [<ip>]" wherever possible.
2465 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2466 are problematic for Debian distribution, omit them from the release
2469 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2471 JH/16 Fix string representation of time values on 64bit time_t architectures.
2474 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2475 output list separator was used.
2480 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2481 checkers that were complaining about end of non-void function with no
2484 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2485 This was a regression introduced in 4.83 by another bugfix.
2487 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2489 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2490 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2496 TF/01 Correctly close the server side of TLS when forking for delivery.
2498 When a message was received over SMTP with TLS, Exim failed to clear up
2499 the incoming connection properly after forking off the child process to
2500 deliver the message. In some situations the subsequent outgoing
2501 delivery connection happened to have the same fd number as the incoming
2502 connection previously had. Exim would try to use TLS and fail, logging
2503 a "Bad file descriptor" error.
2505 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2506 utilities have not been installed.
2508 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2509 temporary space as the ACL may create new global variables.
2511 TL/01 LDAP support uses per connection or global context settings, depending
2512 upon the detected version of the libraries at build time.
2514 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2515 to extract and use the src ip:port in logging and expansions as if it
2516 were a direct connection from the outside internet. PPv2 support was
2517 updated based on HAProxy spec change in May 2014.
2519 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2521 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2522 Properly escape header and check for NULL return.
2524 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2527 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2529 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2530 characters in header names, implemented as a verify condition.
2531 Contributed by Michael Fischer v. Mollard.
2533 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2534 results permerror and temperror. Previous values are deprecated but
2535 still accepted. In a future release, err_perm and err_temp will be
2536 completely removed, which will be a backward incompatibility if the
2537 ACL tests for either of these two old results. Patch contributed by
2538 user bes-internal on the mailing list.
2540 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2542 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2543 selectors, in both main and reject logs.
2545 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2548 JH/07 Add malware type "sock" for talking to simple daemon.
2550 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2552 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2553 routers/transports under cutthrough routing.
2555 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2556 numbers. Touch up "bool" conditional to keep the same definition.
2558 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2560 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2562 JH/12 Expand items in router/transport headers_add or headers_remove lists
2563 individually rather than the list as a whole. Bug 1452.
2565 Required for reasonable handling of multiple headers_ options when
2566 they may be empty; requires that headers_remove items with embedded
2567 colons must have them doubled (or the list-separator changed).
2569 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2570 view the policy declared in the DMARC record. Currently, $dmarc_status
2571 is a combined value of both the record presence and the result of the
2574 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2576 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2577 dnslookup router and the smtp transport (applying to the forward
2580 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2581 of ldap servers used for a specific lookup. Patch provided by Heiko
2584 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2585 New variable $lookup_dnssec_authenticated for observability.
2587 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2588 Patch submitted by Lars Timman.
2590 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2592 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2593 Requires trusted mode and valid format message id, aborts otherwise.
2594 Patch contributed by Heiko Schlichting.
2596 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2597 certextract with support for various fields. Bug 1358.
2599 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2600 is requested by default, modifiable by smtp transport option
2603 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2604 operate on certificate variables to give certificate fingerprints
2605 Also new ${sha256:cert_variable}.
2607 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2609 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2610 Christian Aistleitner.
2612 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2614 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2615 file. Patch from Wolfgang Breyha.
2617 JH/25 Expand the coverage of the delivery $host and $host_address to
2618 client authenticators run in verify callout. Bug 1476.
2620 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2621 align with daemon_smtp_ports. Bug 72.
2623 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2624 support and error reporting did not work properly.
2626 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2627 and is readable. Patch from Andrew Colin Kissa.
2629 TL/14 Enhance documentation of ${run expansion and how it parses the
2630 commandline after expansion, particularly in the case when an
2631 unquoted variable expansion results in an empty value.
2633 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2635 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2636 by renaming away from it. Addresses GH issue 15, reported by
2639 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2641 TL/15 SECURITY: prevent double expansion in math comparison functions
2642 (can expand unsanitized data). Not remotely exploitable.
2649 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2651 PP/02 Make -n do something, by making it not do something.
2652 When combined with -bP, the name of an option is not output.
2654 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2657 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2658 $sender_host_name and config options to manage this, and basic check
2661 PP/05 DSCP support for outbound connections and control modifier for inbound.
2663 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2664 (Only plugin which currently uses this is kerberos4, which nobody should
2665 be using, but we should make it available and other future plugins might
2666 conceivably use it, even though it would break NAT; stuff *should* be
2667 using channel bindings instead).
2669 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2670 name; added for Sendmail compatibility; requires admin caller.
2671 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2672 just ignore it); requires trusted caller.
2673 Also parse but ignore: -Ac -Am -X<logfile>
2676 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2678 TL/02 Add +smtp_confirmation as a default logging option.
2680 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2681 Patch by Magnus Holmgren from 2007-02-20.
2683 TL/04 Bugzilla 1281 - Spec typo.
2684 Bugzilla 1283 - Spec typo.
2685 Bugzilla 1290 - Spec grammar fixes.
2687 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2689 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2691 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2692 mailing list by by Dmitry Isaikin.
2694 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2696 JH/02 Support "G" suffix to numbers in ${if comparisons.
2698 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2700 NM/01 Bugzilla 1197 - Spec typo
2701 Bugzilla 1196 - Spec examples corrections
2703 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2705 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2706 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2709 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2710 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2712 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2713 "acl {{name}{arg}...}", and optional args on acl condition
2716 JH/05 Permit multiple router/transport headers_add/remove lines.
2718 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2720 JH/07 Avoid using a waiting database for a single-message-only transport.
2721 Performance patch from Paul Fisher. Bugzilla 1262.
2723 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2726 JH/09 Add $headers_added variable, with content from use of ACL modifier
2727 add_header (but not yet added to the message). Bugzilla 199.
2729 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2730 Pulled from Bugzilla 817 by Wolfgang Breyha.
2732 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2734 (nb: this is the same fix as in Exim 4.80.1)
2736 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2739 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2742 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2744 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2745 advertise SMTP AUTH mechanism to us, instead of a generic
2746 protocol violation error. Also, make Exim more robust to bad
2747 data from the Dovecot auth socket.
2749 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2751 When a queue runner is handling a message, Exim first routes the
2752 recipient addresses, during which it prunes them based on the retry
2753 hints database. After that it attempts to deliver the message to
2754 any remaining recipients. It then updates the hints database using
2757 So if a recipient address works intermittently, it can get repeatedly
2758 deferred at routing time. The retry hints record remains fresh so the
2759 address never reaches the final cutoff time.
2761 This is a fairly common occurrence when a user is bumping up against
2762 their storage quota. Exim had some logic in its local delivery code
2763 to deal with this. However it did not apply to per-recipient defers
2764 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2766 This change adds a proper retry rule check during routing so that the
2767 final cutoff time is checked against the message's age. We only do
2768 this check if there is an address retry record and there is not a
2769 domain retry record; this implies that previous attempts to handle
2770 the address had the retry_use_local_parts option turned on. We use
2771 this as an approximation for the destination being like a local
2772 delivery, as in LMTP.
2774 I suspect this new check makes the old local delivery cutoff check
2775 redundant, but I have not verified this so I left the code in place.
2777 TF/02 Correct gecos expansion when From: is a prefix of the username.
2779 Test 0254 submits a message to Exim with the header
2783 When I ran the test suite under the user fanf2, Exim expanded
2784 the header to contain my full name, whereas it should have added
2785 a Resent-Sender: header. It erroneously treats any prefix of the
2786 username as equal to the username.
2788 This change corrects that bug.
2790 GF/01 DCC debug and logging tidyup
2791 Error conditions log to paniclog rather than rejectlog.
2792 Debug lines prefixed by "DCC: " to remove any ambiguity.
2794 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2796 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2797 Bug spotted by Jeremy Harris; was flawed since initial commit.
2798 Would have resulted in OCSP responses post-SNI triggering an Exim
2799 NULL dereference and crash.
2801 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2803 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2804 Bug detection, analysis and fix by Samuel Thibault.
2805 Bugzilla 1331, Debian bug #698092.
2807 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2809 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2810 Server implementation by Todd Lyons, client by JH.
2811 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2812 config variable "prdr_enable" controls whether the server
2813 advertises the facility. If the client requests PRDR a new
2814 acl_data_smtp_prdr ACL is called once for each recipient, after
2815 the body content is received and before the acl_smtp_data ACL.
2816 The client is controlled by both of: a hosts_try_prdr option
2817 on the smtp transport, and the server advertisement.
2818 Default client logging of deliveries and rejections involving
2819 PRDR are flagged with the string "PRDR".
2821 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2822 fclose(). Diagnosis by Todd Lyons.
2824 PP/17 Update configure.default to handle IPv6 localhost better.
2825 Patch by Alain Williams (plus minor tweaks).
2828 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2829 This is now consistent with GnuTLS, and is now documented: the
2830 previous undocumented portable approach to treating the option as
2831 unset was to force an expansion failure. That still works, and
2832 an empty string is now equivalent.
2834 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2835 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2836 not performing validation itself.
2838 PP/20 Added force_command boolean option to pipe transport.
2839 Patch from Nick Koston, of cPanel Inc.
2841 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2844 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2846 PP/21 Fix eximon continuous updating with timestamped log-files.
2847 Broken in a format-string cleanup in 4.80, missed when I repaired the
2848 other false fix of the same issue.
2849 Report and fix from Heiko Schlichting.
2852 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2853 Report from Prashanth Katuri.
2855 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2856 It's SecureTransport, so affects any MacOS clients which use the
2857 system-integrated TLS libraries, including email clients.
2859 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2860 using a MIME ACL for non-SMTP local injection.
2861 Report and assistance in diagnosis by Warren Baker.
2863 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2865 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2867 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2868 last id that failed so it may be referenced in subsequent ACL's.
2870 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2873 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2874 ldap library initialization, allowing self-signed CA's to be
2875 used. Also properly sets require_cert option later in code by
2876 using NULL (global ldap config) instead of ldap handle (per
2877 session). Bug diagnosis and testing by alxgomz.
2879 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2880 the src/util/ subdirectory.
2882 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2883 renamed to Transport Post Delivery Action by Jeremy Harris, as
2886 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2887 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2888 redis_servers = needs to be configured which will be used by the redis
2889 lookup. Patch from Warren Baker, of The Packet Hub.
2891 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2893 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2894 hostname or reverse DNS when processing a host list. Used suggestions
2895 from multiple comments on this bug.
2897 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2899 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2900 Missed a few lines, added it to make the runtest require no keyboard
2903 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2904 contains upper case chars. Make router use caseful_local_part.
2906 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2907 support when GnuTLS has been built with p11-kit.
2913 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2915 This, or similar/improved, will also be change PP/11 of 4.82.
2921 PP/01 Handle short writes when writing local log-files.
2922 In practice, only affects FreeBSD (8 onwards).
2923 Bugzilla 1053, with thanks to Dmitry Isaikin.
2925 NM/01 Bugzilla 949 - Documentation tweak
2927 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2930 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2932 PP/02 Implemented gsasl authenticator.
2934 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2936 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2937 `pkg-config foo` for cflags/libs.
2939 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2940 with rest of GSASL and with heimdal_gssapi.
2942 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2943 `pkg-config foo` for cflags/libs for the TLS implementation.
2945 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2946 properties get this fed in as external SSF. A number of robustness
2947 and debugging improvements to the cyrus_sasl authenticator.
2949 PP/08 cyrus_sasl server now expands the server_realm option.
2951 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2952 Patch by Jeremy Harris.
2954 PP/10 Added dbmjz lookup type.
2956 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2958 PP/12 MAIL args handles TAB as well as SP, for better interop with
2959 non-compliant senders.
2960 Analysis and variant patch by Todd Lyons.
2962 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2963 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2964 Patch from Dirk Mueller <dmueller@suse.com>
2966 PP/13 tls_peerdn now print-escaped for spool files.
2967 Observed some $tls_peerdn in wild which contained \n, which resulted
2968 in spool file corruption.
2970 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2971 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2972 or write after TLS renegotiation, which otherwise led to messages
2975 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2976 as a tracking header (ie: a signed header comes before the signature).
2977 Patch from Wolfgang Breyha.
2979 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2980 comma-sep list; embedded commas doubled.
2982 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2984 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2986 Report and patch from Dmitry Banschikov.
2988 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2989 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2990 failures. We appear to now support TLS1.1+ with Exim.
2992 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2993 lets Exim select keys and certificates based upon TLS SNI from client.
2994 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2995 before an outbound SMTP session. New log_selector, +tls_sni.
2997 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2998 NULL dereference. Report and patch from Alun Jones.
3000 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3001 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3002 Not seeing resolver debug output on NetBSD, but suspect this is a
3003 resolver implementation change.
3005 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3006 Left warnings. Added "eximon gdb" invocation mode.
3008 PP/21 Defaulting "accept_8bitmime" to true, not false.
3010 PP/22 Added -bw for inetd wait mode support.
3012 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3013 locate the relevant includes and libraries. Made this the default.
3015 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3016 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3018 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3019 This may cause build issues on older platforms.
3021 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3022 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3023 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3024 Added SNI support via GnuTLS too.
3025 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3027 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3029 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3030 Applied second patch from Janne, implementing suggestion to default
3031 multiple-strings-in-record handling to match SPF spec.
3033 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3035 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3036 read-only, out of scope).
3037 Patch from Wolfgang Breyha, report from Stuart Northfield.
3039 PP/29 Fix three issues highlighted by clang analyser static analysis.
3040 Only crash-plausible issue would require the Cambridge-specific
3041 iplookup router and a misconfiguration.
3042 Report from Marcin Mirosław.
3044 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3046 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3047 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3048 As part of this, removing so much warning spew let me fix some minor
3049 real issues in debug logging.
3051 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3052 assignment on my part. Fixed.
3054 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3055 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3056 Janne Snabb (who went above and beyond: thank you).
3058 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3059 string otherwise requires a connection and a bunch more work and it's
3060 relatively easy to get wrong. Should also expose TLS library linkage
3063 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3064 64-bit ${eval} (JH/03).
3066 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3067 GNU libc to support some of the 64-bit stuff, should not lead to
3068 conflicts. Defined before os.h is pulled in, so if a given platform
3069 needs to override this, it can.
3071 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3072 protection layer was required, which is not implemented.
3073 Bugzilla 1254, patch from Wolfgang Breyha.
3075 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3076 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3077 tls_dhparam take prime identifiers. Also unbreak combination of
3078 OpenSSL+DH_params+TLSSNI.
3080 PP/39 Disable SSLv2 by default in OpenSSL support.
3086 PP/01 Solaris build fix for Oracle's LDAP libraries.
3087 Bugzilla 1109, patch from Stephen Usher.
3089 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3091 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3094 TF/02 Fix a couple more cases where we did not log the error message
3095 when unlink() failed. See also change 4.74-TF/03.
3097 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3098 lock up or crash if it happened to be inside a call to libc when it
3099 got a SIGUSR1 from exiwhat.
3101 The SIGUSR1 handler appends the current process status to the process
3102 log which is later printed by exiwhat. It used to use the general
3103 purpose logging code to do this, but several functions it calls are
3104 not safe for signals.
3106 The new output code in the SIGUSR1 handler is specific to the process
3107 log, and simple enough that it's easy to inspect for signal safety.
3108 Removing some special cases also simplifies the general logging code.
3109 Removing the spurious timestamps from the process log simplifies
3112 TF/04 Improved ratelimit ACL condition.
3114 The /noupdate option has been deprecated in favour of /readonly which
3115 has clearer semantics. The /leaky, /strict, and /readonly update modes
3116 are mutually exclusive. The update mode is no longer included in the
3117 database key; it just determines when the database is updated. (This
3118 means that when you upgrade Exim will forget old rate measurements.)
3120 Exim now checks that the per_* options are used with an update mode that
3121 makes sense for the current ACL. For example, when Exim is processing a
3122 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3123 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3124 must specify per_mail/readonly. If you omit the update mode it defaults to
3125 /leaky where that makes sense (as before) or /readonly where required.
3127 The /noupdate option is now undocumented but still supported for
3128 backwards compatibility. It is equivalent to /readonly except that in
3129 ACLs where /readonly is required you may specify /leaky/noupdate or
3130 /strict/noupdate which are treated the same as /readonly.
3132 A useful new feature is the /count= option. This is a generalization
3133 of the per_byte option, so that you can measure the throughput of other
3134 aggregate values. For example, the per_byte option is now equivalent
3135 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3137 The per_rcpt option has been generalized using the /count= mechanism
3138 (though it's more complicated than the per_byte equivalence). When it is
3139 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3140 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3141 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3142 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3143 Note that using per_rcpt with a non-readonly update mode in more than
3144 one ACL will cause the recipients to be double-counted. (The per_mail
3145 and per_byte options don't have this problem.)
3147 The handling of very low rates has changed slightly. If the computed rate
3148 is less than the event's count (usually one) then this event is the first
3149 after a long gap. In this case the rate is set to the same as this event's
3150 count, so that the first message of a spam run is counted properly.
3152 The major new feature is a mechanism for counting the rate of unique
3153 events. The new per_addr option counts the number of different
3154 recipients that someone has sent messages to in the last time period. It
3155 behaves like per_rcpt if all the recipient addresses are different, but
3156 duplicate recipient addresses do not increase the measured rate. Like
3157 the /count= option this is a general mechanism, so the per_addr option
3158 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3159 example, measure the rate that a client uses different sender addresses
3160 with the options per_mail/unique=$sender_address. There are further
3161 details in the main documentation.
3163 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3165 TF/06 Removed a few PCRE remnants.
3167 TF/07 Automatically extract Exim's version number from tags in the git
3168 repository when doing development or release builds.
3170 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3171 Bugzilla 879. Patch from Paul Fisher.
3173 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3174 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3177 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3179 PP/05 Variable $av_failed, true if the AV scanner deferred.
3180 Bugzilla 1078. Patch from John Horne.
3182 PP/06 Stop make process more reliably on build failure.
3183 Bugzilla 1087. Patch from Heiko Schlittermann.
3185 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3186 Bugzilla 1089. Patch from Heiko Schlittermann.
3188 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3189 Bugzilla 1131. Patch from Holger Weiß.
3191 PP/09 Handle IPv6 addresses with SPF.
3192 Bugzilla 860. Patch from Wolfgang Breyha.
3194 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3196 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3199 PP/11 match_* no longer expand right-hand-side by default.
3200 New compile-time build option, EXPAND_LISTMATCH_RHS.
3201 New expansion conditions, "inlist", "inlisti".
3203 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3205 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3207 PP/14 fix log_write() format string regression from TF/03.
3208 Bugzilla 1152. Patch from Dmitry Isaikin.
3214 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3216 PP/02 Harmonised TLS library version reporting; only show if debugging.
3217 Layout now matches that introduced for other libraries in 4.74 PP/03.
3219 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3221 PP/04 New "dns_use_edns0" global option.
3223 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3226 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3227 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3229 TK/01 Updated PolarSSL code to 0.14.2.
3230 Bugzilla 1097. Patch from Andreas Metzler.
3232 PP/07 Catch divide-by-zero in ${eval:...}.
3233 Fixes bugzilla 1102.
3235 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3238 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3239 format-string attack -- SECURITY: remote arbitrary code execution.
3241 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3242 time unintentionally subject to list matching rules, letting the header
3243 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3244 arbitrary string expansion). This allowed for information disclosure.
3246 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3247 INT_MIN/-1 -- value coerced to INT_MAX.
3253 NM/01 Workaround for PCRE version dependency in version reporting
3256 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3257 This fixes portability to compilers other than gcc, notably
3258 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3260 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3261 makefiles for portability to HP-UX and POSIX correctness.
3263 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3264 Also via indented variable definition in the Makefile.
3265 (Debugging by Oliver Heesakkers).
3267 PP/02 Restore caching of spamd results with expanded spamd_address.
3268 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3270 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3271 Improves build reliability. Fix from: Frank Elsner
3273 NM/02 Fix wide character breakage in the rfc2047 coding
3274 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3276 NM/03 Allow underscore in dnslist lookups
3277 Fixes bug 1026. Patch from Graeme Fowler
3279 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3280 Code patches from Adam Ciarcinski of NetBSD.
3282 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3285 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3286 is logged, to avoid truncation. Patch from John Horne.
3288 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3289 Patch from Jakob Hirsch.
3291 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3292 SQL string expansion failure details.
3293 Patch from Andrey Oktyabrski.
3295 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3296 Patch from Simon Arlott.
3298 PP/09 New lookups functionality failed to compile on old gcc which rejects
3299 extern declarations in function scope.
3300 Patch from Oliver Fleischmann
3302 PP/10 Use sig_atomic_t for flags set from signal handlers.
3303 Check getgroups() return and improve debugging.
3304 Fixed developed for diagnosis in bug 927 (which turned out to be
3307 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3308 Patch from Mark Zealey.
3310 PP/12 Bugzilla 1056: Improved spamd server selection.
3311 Patch from Mark Zealey.
3313 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3314 Based on patch from Heiko Schlittermann.
3316 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3317 Patch from Uwe Doering, sign-off by Michael Haardt.
3319 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3320 variable declaration deep within a block. Bug and patch from
3323 PP/15 lookups-Makefile IRIX compatibility coercion.
3325 PP/16 Make DISABLE_DKIM build knob functional.
3327 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3328 Patch by Simon Arlott
3330 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3331 variable argument macros. Our copy now differs from upstream.
3337 TF/01 Failure to get a lock on a hints database can have serious
3338 consequences so log it to the panic log.
3340 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3341 controlled using the smtp_confirmation log selector.
3343 TF/03 Include the error message when we fail to unlink a spool file.
3345 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3346 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3347 for maintaining out-of-tree patches for some time.
3349 PP/01 Bugzilla 139: Documentation and portability issues.
3350 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3351 Handle per-OS dynamic-module compilation flags.
3353 PP/02 Let /dev/null have normal permissions.
3354 The 4.73 fixes were a little too stringent and complained about the
3355 permissions on /dev/null. Exempt it from some checks.
3356 Reported by Andreas M. Kirchwitz.
3358 PP/03 Report version information for many libraries, including
3359 Exim version information for dynamically loaded libraries. Created
3360 version.h, now support a version extension string for distributors
3361 who patch heavily. Dynamic module ABI change.
3363 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3364 privilege escalation vulnerability whereby the Exim run-time user
3365 can cause root to append content of the attacker's choosing to
3368 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3371 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3372 If dropping privileges for untrusted macros, we disabled normal logging
3373 on the basis that it would fail; for the Exim run-time user, this is not
3374 the case, and it resulted in successful deliveries going unlogged.
3375 Fixed. Reported by Andreas Metzler.
3381 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3382 only prepend for the Resent-* case. Fixes regression introduced in
3383 Exim 4.70 by NM/22 for Bugzilla 607.
3385 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3386 increasing numbers of administrators be bitten by this.
3388 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3390 PP/03 Bugzilla 994: added openssl_options main configuration option.
3392 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3394 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3396 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3398 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3399 without a peer certificate, leading to a segfault because of an
3400 assumption that peers always have certificates. Be a little more
3401 paranoid. Problem reported by Martin Tscholak.
3403 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3404 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3405 NB: ClamAV planning to remove STREAM in "middle of 2010".
3406 CL also introduces -bmalware, various -d+acl logging additions and
3407 more caution in buffer sizes.
3409 PP/09 Implemented reverse_ip expansion operator.
3411 PP/10 Bugzilla 937: provide a "debug" ACL control.
3413 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3415 PP/12 Bugzilla 973: Implement --version.
3417 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3419 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3421 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3423 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3424 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3425 ignore trailing whitespace.
3427 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3429 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3432 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3433 Notification from Dr Andrew Aitchison.
3435 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3436 ExtendedDetectionInfo response format.
3437 Notification from John Horne.
3439 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3442 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3443 XSL and documented dependency on system catalogs, with examples of how
3446 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3449 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3450 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3451 configuration file which is writeable by the Exim user or group.
3453 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3454 of configuration files to cover files specified with the -C option if
3455 they are going to be used with root privileges, not just the default
3458 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3459 option (effectively making it always true).
3461 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3462 files to be used while preserving root privileges.
3464 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3465 that rogue child processes cannot use them.
3467 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3468 run-time user, instead of root.
3470 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3471 Exim run-time user without dropping privileges.
3473 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3474 result string, instead of calling string_vformat() twice with the same
3477 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3478 for other users. Others should always drop root privileges if they use
3479 -C on the command line, even for a whitelisted configure file.
3481 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3483 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3489 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3490 $data_path, and $header_path variables; fixed documentation bugs and
3493 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3494 exipick to access non-standard spools, including the "frozen" queue
3497 NM/01 Bugzilla 965: Support mysql stored procedures.
3498 Patch from Alain Williams
3500 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3502 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3503 Patch from Andreas Metzler
3505 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3506 Patch from Kirill Miazine
3508 NM/05 Bugzilla 671: Added umask to procmail example.
3510 JJ/03 installed exipick 20100323.0, fixing doc bug
3512 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3513 directory. Notification and patch from Dan Rosenberg.
3515 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3517 TK/02 Improve log output when DKIM signing operation fails.
3519 MH/01 Treat the transport option dkim_domain as a colon separated
3520 list, not as a single string, and sign the message with each element,
3521 omitting multiple occurences of the same signer.
3523 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3524 Bugzilla 985, 986. Patch by Simon Arlott
3526 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3527 Patch by Simon Arlott
3529 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3530 MBX locking. Notification from Dan Rosenberg.
3536 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3538 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3540 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3542 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3544 NM/04 Bugzilla 915: Flag broken perl installation during build.
3550 TK/01 Added patch by Johannes Berg that expands the main option
3551 "spamd_address" if it starts with a dollar sign.
3553 TK/02 Write list of recipients to X-Envelope-Sender header when building
3554 the mbox-format spool file for content scanning (suggested by Jakob
3557 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3558 (http://www.dcc-servers.net/) support via dccifd. Activated by
3559 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3561 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3562 by Mark Daniel Reidel <mr@df.eu>.
3564 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3565 When building exim an external PCRE library is now needed -
3566 PCRE is a system library on the majority of modern systems.
3567 See entry on PCRE_LIBS in EDITME file.
3569 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3570 conversation. Added nologin parameter to request.
3571 Patch contributed by Kirill Miazine.
3573 TF/01 Do not log submission mode rewrites if they do not change the address.
3575 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3577 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3578 log files in place. Contributed by Roberto Lima.
3580 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3582 TF/03 Bugzilla 615: When checking the local_parts router precondition
3583 after a local_part_suffix or local_part_prefix option, Exim now
3584 does not use the address's named list lookup cache, since this
3585 contains cached lookups for the whole local part.
3587 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3588 Robert Millan. Documentation is in experimental-spec.txt.
3590 TF/04 Bugzilla 668: Fix parallel build (make -j).
3592 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3594 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3595 Patch provided by Jan Srzednicki.
3597 TF/05 Leading white space used to be stripped from $spam_report which
3598 wrecked the formatting. Now it is preserved.
3600 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3601 that they are available at delivery time.
3603 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3605 TF/08 TLS error reporting now respects the incoming_interface and
3606 incoming_port log selectors.
3608 TF/09 Produce a more useful error message if an SMTP transport's hosts
3609 setting expands to an empty string.
3611 NM/06 Bugzilla 744: EXPN did not work under TLS.
3612 Patch provided by Phil Pennock.
3614 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3615 Patch provided by Richard Godbee.
3617 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3618 acl_smtp_notquit, added index entry.
3620 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3621 Patch provided by Eugene Bujak.
3623 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3624 accept(). Patch provided by Maxim Dounin.
3626 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3627 Patch provided by Phil Pennock.
3629 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3631 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3632 Patch provided by Brad "anomie" Jorsch.
3634 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3635 Patch provided by Dean Brooks.
3637 TK/05 Add native DKIM support (does not depend on external libraries).
3639 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3640 Patch provided by Graeme Fowler.
3642 NM/16 Bugzilla 851: Documentation example syntax fix.
3644 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3646 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3649 NM/19 Bugzilla 745: TLS version reporting.
3650 Patch provided by Phil Pennock.
3652 NM/20 Bugzilla 167: bool: condition support.
3653 Patch provided by Phil Pennock.
3655 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3656 clients. Patch provided by Phil Pennock.
3658 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3659 Patch provided by Brad "anomie" Jorsch.
3661 NM/23 Bugzilla 687: Fix misparses in eximstats.
3662 Patch provided by Heiko Schlittermann.
3664 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3665 Patch provided by Heiko Schlittermann.
3667 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3668 plus update to original patch.
3670 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3672 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3673 Patch provided by David Brownlee.
3675 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3677 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3679 NM/30 Bugzilla 888: TLS documentation bugfixes.
3681 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3683 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3684 Unlike the original bugzilla I have changed all shell scripts in src tree.
3686 NM/33 Bugzilla 898: Transport filter timeout fix.
3687 Patch by Todd Rinaldo.
3689 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3690 Patch by Serge Demonchaux.
3692 NM/35 Bugzilla 39: Base64 decode bug fixes.
3693 Patch by Jakob Hirsch.
3695 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3697 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3699 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3701 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3707 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3708 ALT-N's libdkim that I have put here:
3709 http://duncanthrax.net/exim-experimental/
3711 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3712 were called 'true' and it seems that C99 defines that as a reserved
3713 keyword to be used with 'bool' variable types. That means you could
3714 not include C99-style headers which use bools without triggering
3715 build errors in sieve.c.
3717 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3718 as mailq or other aliases. Changed the --help handling significantly
3719 to do whats expected. exim_usage() emits usage/help information.
3721 SC/01 Added the -bylocaldomain option to eximstats.
3723 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3725 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3727 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3733 PH/01 Another patch from the Sieve maintainer.
3735 PH/02 When an IPv6 address is converted to a string for single-key lookup
3736 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3737 dots are used instead of colons so that keys in lsearch files need not
3738 contain colons. This was done some time before quoting was made available
3739 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3740 (notated using the quote facility) so as to distinguish them from IPv4
3741 keys. This meant that lookups for IP addresses in host lists did not work
3742 for iplsearch lookups.
3744 This has been fixed by arranging for IPv6 addresses to be expressed with
3745 colons if the lookup type is iplsearch. This is not incompatible, because
3746 previously such lookups could never work.
3748 The situation is now rather anomalous, since one *can* have colons in
3749 ordinary lsearch keys. However, making the change in all cases is
3750 incompatible and would probably break a number of configurations.
3752 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3755 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3756 conversion specification without a maximum field width, thereby enabling
3757 a rogue spamd server to cause a buffer overflow. While nobody in their
3758 right mind would setup Exim to query an untrusted spamd server, an
3759 attacker that gains access to a server running spamd could potentially
3760 exploit this vulnerability to run arbitrary code as the Exim user.
3762 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3763 $primary_hostname instead of what libspf2 thinks the hosts name is.
3765 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3766 a directory entry by the name of the lookup key. Previously, if a
3767 symlink pointed to a non-existing file or a file in a directory that
3768 Exim lacked permissions to read, a lookup for a key matching that
3769 symlink would fail. Now it is enough that a matching directory entry
3770 exists, symlink or not. (Bugzilla 503.)
3772 PH/03 The body_linecount and body_zerocount variables are now exported in the
3775 PH/04 Added the $dnslist_matched variable.
3777 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3778 This means they are set thereafter only if the connection becomes
3781 PH/06 Added the client_condition to authenticators so that some can be skipped
3782 by clients under certain conditions.
3784 PH/07 The error message for a badly-placed control=no_multiline_responses left
3785 "_responses" off the end of the name.
3787 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3789 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3790 (without spaces) instead of just copying the configuration text.
3792 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3794 PH/11 Added $max_received_linelength.
3796 PH/12 Added +ignore_defer and +include_defer to host lists.
3798 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3799 way in which PCRE > 7.0 is built.
3801 PH/14 Implemented queue_only_load_latch.
3803 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3804 MAIL command. The effect was to mangle the value on 64-bit systems.
3806 PH/16 Another patch from the Sieve maintainer.
3808 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3810 PH/18 If a system quota error occurred while trying to create the file for
3811 a maildir delivery, the message "Mailbox is full" was not appended to the
3812 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3813 only to a quota excession during the actual writing of the file.
3815 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3816 characters?) which causes problems in log lines. The DN values are now
3817 passed through string_printing() before being added to log lines.
3819 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3820 and InterBase are left for another time.)
3822 PH/21 Added message_body_newlines option.
3824 PH/22 Guard against possible overflow in moan_check_errorcopy().
3826 PH/23 POSIX allows open() to be a macro; guard against that.
3828 PH/24 If the recipient of an error message contained an @ in the local part
3829 (suitably quoted, of course), incorrect values were put in $domain and
3830 $local_part during the evaluation of errors_copy.
3836 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3837 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3840 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3841 issue a MAIL command.
3843 PH/02 In an ACL statement such as
3845 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3847 if a client was not listed at all, or was listed with a value other than
3848 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3849 the condition was not true (as it should be), so access was not denied.
3850 The bug was that the ! inversion was incorrectly passed on to the second
3851 item. This has been fixed.
3853 PH/03 Added additional dnslists conditions == and =& which are different from
3854 = and & when the dns lookup returns more than one IP address.
3856 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3857 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3859 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3860 FSYNC, which compiles an option called disable_fsync that allows for
3861 bypassing fsync(). The documentation is heavily laced with warnings.
3863 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3865 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3866 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3867 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3868 including adding "make clean"; (3) Added -fPIC when compiling the test
3869 dynamically loaded module, to get rid of a warning.
3871 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3872 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3873 = 0s. The bug is otherwise harmless.
3875 PH/07 There was a bug in the dovecot authenticator such that the value of
3876 $auth1 could be overwritten, and so not correctly preserved, after a
3877 successful authentication. This usually meant that the value preserved by
3878 the server_setid option was incorrect.
3880 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3882 PH/09 Installed PCRE release 7.0.
3884 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3885 run for batched SMTP input. It is now run at the start of every message
3886 in the batch. While fixing this I discovered that the process information
3887 (output by running exiwhat) was not always getting set for -bs and -bS
3888 input. This is fixed, and it now also says "batched" for BSMTP.
3890 PH/11 Added control=no_pipelining.
3892 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3893 patch, slightly modified), and move the expansion of helo_data till after
3894 the connection is made in the smtp transport (so it can use these
3897 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3899 PH/14 Added log_selector = +pid.
3901 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3903 PH/16 Add ${if forany and ${if forall.
3905 PH/17 Added dsn_from option to vary the From: line in DSNs.
3907 PH/18 Flush SMTP output before performing a callout, unless control =
3908 no_callout_flush is set.
3910 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3911 was true (the default) a successful delivery failed to delete the retry
3912 item, thus causing premature timeout of the address. The bug is now
3915 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3917 PH/21 Long custom messages for fakedefer and fakereject are now split up
3918 into multiline responses in the same way that messages for "deny" and
3919 other ACL rejections are.
3921 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3922 with slight modification.
3924 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3925 draft, changing the syntax and factoring some duplicate code".
3927 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3928 for deliveries of the second and subsequent messages over the same SMTP
3931 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3932 ${reduce, with only minor "tidies".
3934 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3936 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3937 expansion side effects.
3939 PH/27 When a message times out after an over-quota error from an Exim-imposed
3940 quota, the bounce message says "mailbox is full". This message was not
3941 being given when it was a system quota that was exceeded. It now should
3944 MH/03 Made $recipients available in local_scan(). local_scan() already has
3945 better access to the recipient list through recipients_list[], but
3946 $recipients can be useful in postmaster-provided expansion strings.
3948 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3949 in the case of a MAIL command with additional options following the
3950 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3951 were accidentally chopped off.
3953 PH/29 SMTP synchronization checks are implemented when a command is read -
3954 there is a check that no more input is waiting when there shouldn't be
3955 any. However, for some commands, a delay in an ACL can mean that it is
3956 some time before the response is written. In this time, more input might
3957 arrive, invalidly. So now there are extra checks after an ACL has run for
3958 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3959 pipelining has not been advertised.
3961 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3963 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3964 returned in the SMTP response when smtp_return_error_details was false.
3965 This has been fixed.
3967 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3968 instead of the C I/O that was originally supplied, because problems were
3969 reported on Solaris.
3971 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3972 Exim which did not show up earlier: it was assuming that a call to
3973 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3974 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3975 was a macro that became an assignment, so it seemed to work. This has
3976 changed to a proper function call with a void return, hence the compile
3977 error. Exim's code has been fixed.
3979 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3982 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3984 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3986 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3987 because it turns out to be too big or there is a timeout) there is no
3988 "Completed" line in the log. When some messages of this type were
3989 selected by exigrep, they were listed as "not completed". Others were
3990 picked up by some special patterns. I have improved the selection
3991 criteria to be more general.
3993 PH/38 The host_find_failed option in the manualroute router can now be set
3994 to "ignore", to completely ignore a host whose IP address cannot be
3995 found. If all hosts are ignored, the behaviour is controlled by the new
3996 host_all_ignored option.
3998 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3999 homing or because of multiple MX records with /mx) generated more than
4000 one IP address, and the following item turned out to be the local host,
4001 all the secondary addresses of the first item were incorrectly removed
4002 from the list, along with the local host and any following hosts (which
4003 is what is supposed to happen).
4005 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4006 whoever called Exim into the -H file. In the case of the daemon it was
4007 behaving confusingly. When first started, it used values for whoever
4008 started the daemon, but after a SIGHUP it used the Exim user (because it
4009 calls itself on a restart). I have changed the code so that it now always
4012 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4013 message are rejected with the same error (e.g. no authentication or bad
4014 sender address), and a DATA command is nevertheless sent (as can happen
4015 with PIPELINING or a stupid MUA), the error message that was given to the
4016 RCPT commands is included in the rejection of the DATA command. This is
4017 intended to be helpful for MUAs that show only the final error to their
4020 PH/42 Another patch from the Sieve maintainer.
4022 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4023 Eximstats - Fixed some broken HTML links and added missing column headers
4025 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4026 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4028 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4030 PH/43 Yet another patch from the Sieve maintainer.
4032 PH/44 I found a way to check for a TCP/IP connection going away before sending
4033 the response to the final '.' that terminates a message, but only in the
4034 case where the client has not sent further data following the '.'
4035 (unfortunately, this is allowed). However, in many cases there won't be
4036 any further data because there won't be any more messages to send. A call
4037 to select() can be used: if it shows that the input is "ready", there is
4038 either input waiting, or the socket has been closed. An attempt to read
4039 the next input character can distinguish the two cases. Previously, Exim
4040 would have sent an OK response which the client would never have see.
4041 This could lead to message repetition. This fix should cure that, at
4042 least in a lot of common cases.
4044 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4045 advertised in response to EHLO.
4051 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4052 fixed by 4.65/MH/01 (is this a record?) are fixed:
4054 (i) An empty string was always treated as zero by the numeric comparison
4055 operators. This behaviour has been restored.
4057 (ii) It is documented that the numeric comparison operators always treat
4058 their arguments as decimal numbers. This was broken in that numbers
4059 starting with 0 were being interpreted as octal.
4061 While fixing these problems I realized that there was another issue that
4062 hadn't been noticed. Values of message_size_limit (both the global option
4063 and the transport option) were treated as octal if they started with 0.
4064 The documentation was vague. These values are now always treated as
4065 decimal, and I will make that clear in the documentation.
4071 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4072 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4075 MH/01 Don't check that the operands of numeric comparison operators are
4076 integers when their expansion is in "skipping" mode (fixes bug
4077 introduced by 4.64-PH/07).
4079 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4080 child addresses, Exim now panics and dies. Previously, because the count
4081 is held in a short int, deliveries were likely to be lost. As such a
4082 large number of recipients for a single message is ridiculous
4083 (performance will be very, very poor), I have chosen to impose a limit
4084 rather than extend the field.
4090 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4091 leftover -K file (the existence of which was triggered by #402).
4092 While we were at it, introduced process PID as part of the -K
4093 filename. This should rule out race conditions when creating
4096 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4097 processing considerably. Previous code took too long for large mails,
4098 triggering a timeout which in turn triggers #401.
4100 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4101 in the DK code in transports.c. sendfile() is not really portable,
4102 hence the _LINUX specificness.
4104 TF/01 In the add_headers option to the mail command in an Exim filter,
4105 there was a bug that Exim would claim a syntax error in any
4106 header after the first one which had an odd number of characters
4109 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4110 callout verification, Exim cached a "reject" for the entire domain. This
4111 is correct for most verifications, but it is not correct for a recipient
4112 verification with use_sender or use_postmaster set, because in that case
4113 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4114 case of MAIL FROM:<> rejection from other early rejections (e.g.
4115 rejection of HELO). When verifying a recipient using a non-null MAIL
4116 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4117 Whatever the result of the callout, the value of the domain cache is
4118 left unchanged (for any other kind of callout, getting as far as trying
4119 RCPT means that the domain itself is ok).
4121 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4124 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4125 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4126 session, the daemon ignores ECONNECT errors and logs others; it now
4127 ignores EPIPE as well.
4129 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4130 (quoted-printable decoding).
4132 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4133 later the small subsequent patch to fix an introduced bug.
4135 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4137 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4139 PH/08 An error is now given if message_size_limit is specified negative.
4141 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4142 to be given (somewhat) arbitrary names.
4144 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4147 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4148 miscellaneous code fixes
4150 PH/10 Added the log_reject_target ACL modifier to specify where to log
4153 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4154 hostname. This is wrong, because it relates to the incoming message (and
4155 probably the interface on which it is arriving) and not to the outgoing
4156 callout (which could be using a different interface). This has been
4157 changed to use the value of the helo_data option from the smtp transport
4158 instead - this is what is used when a message is actually being sent. If
4159 there is no remote transport (possible with a router that sets up host
4160 addresses), $smtp_active_hostname is used.
4162 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4163 tweaks were necessary in order to get it to work (see also 21 below):
4164 (a) The code assumed that strncpy() returns a negative number on buffer
4165 overflow, which isn't the case. Replaced with Exim's string_format()
4167 (b) There were several signed/unsigned issues. I just did the minimum
4168 hacking in of casts. There is scope for a larger refactoring.
4169 (c) The code used strcasecmp() which is not a standard C function.
4170 Replaced with Exim's strcmpic() function.
4171 (d) The code set only $1; it now sets $auth1 as well.
4172 (e) A simple test gave the error "authentication client didn't specify
4173 service in request". It would seem that Dovecot has changed its
4174 interface. Fortunately there's a specification; I followed it and
4175 changed what the client sends and it appears to be working now.
4177 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4180 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4181 address A is aliased to B and C, where B exists and C does not. Without
4182 -v the output is "A verified" because verification stops after a
4183 successful redirection if more than one address is generated. However,
4184 with -v the child addresses are also verified. Exim was outputting "A
4185 failed to verify" and then showing the successful verification for C,
4186 with its parentage. It now outputs "B failed to verify", showing B's
4187 parentage before showing the successful verification of C.
4189 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4190 look up a TXT record in a specific list after matching in a combined
4193 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4194 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4195 they consult the DNS. I had assumed they would set it the way they
4196 wanted; and indeed my experiments on Linux seem to show that in some
4197 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4198 To be on the safe side, however, I have now made the interface to
4199 host_find_byname() similar to host_find_bydns(), with an argument
4200 containing the DNS resolver options. The host_find_byname() function now
4201 sets these options at its start, just as host_find_bydns() does. The smtp
4202 transport options dns_qualify_single and dns_search_parents are passed to
4203 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4204 of host_find_byname() use the default settings of RES_DEFNAMES
4205 (qualify_single) but not RES_DNSRCH (search_parents).
4207 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4208 spool_read_header() do less string testing, by means of a preliminary
4209 switch on the second character of optional "-foo" lines. (This is
4210 overdue, caused by the large number of possibilities that now exist.
4211 Originally there were few.) While I was there, I also converted the
4212 str(n)cmp tests so they don't re-test the leading "-" and the first
4213 character, in the hope this might squeeze out yet more improvement.
4215 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4216 flag allowing group syntax was set by the header_syntax check but not
4217 turned off, possible causing trouble later; (2) The flag was not being
4218 set at all for the header_verify test, causing "group"-style headers to
4219 be rejected. I have now set it in this case, and also caused header_
4220 verify to ignore an empty address taken from a group. While doing this, I
4221 came across some other cases where the code for allowing group syntax
4222 while scanning a header line wasn't quite right (mostly, not resetting
4223 the flag correctly in the right place). These bugs could have caused
4224 trouble for malformed header lines. I hope it is now all correct.
4226 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4227 with the "reply" argument non-NULL. The code, however (which originally
4228 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4229 but it didn't always do it. This confused somebody who was copying the
4230 code for some other use. I have removed all the tests.
4232 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4233 feature that was used to support insecure browsers during the U.S. crypto
4234 embargo. It requires special client support, and Exim is probably the
4235 only MTA that supported it -- and would never use it because real RSA is
4236 always available. This code has been removed, because it had the bad
4237 effect of slowing Exim down by computing (never used) parameters for the
4238 RSA_EXPORT functionality.
4240 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4241 authenticator to fail if there's a tab character in the incoming data
4242 (there should never be unless someone is messing about, as it's supposed
4243 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4244 if the connection is using TLS or if the remote IP is the same as the
4245 local IP, and the "valid-client-cert option" if a client certificate has
4248 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4249 authenticators. This can be used for authorization after authentication
4250 succeeds. (In the case of plaintext, it servers for both authentication
4253 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4254 if any retry times were supplied.
4256 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4257 connection, where there is no client IP address to check. In this
4258 situation, the verify now always succeeds.
4260 PH/25 Applied John Jetmore's -Mset patch.
4262 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4264 PH/27 In a string expansion for a processed (not raw) header when multiple
4265 headers of the same name were present, leading whitespace was being
4266 removed from all of them, but trailing whitespace was being removed only
4267 from the last one. Now trailing whitespace is removed from each header
4268 before concatenation. Completely empty headers in a concatenation (as
4269 before) are ignored.
4271 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4272 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4274 PH/29 [Removed. This was a change that I later backed out, and forgot to
4275 correct the ChangeLog entry (that I had efficiently created) before
4276 committing the later change.]
4278 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4279 address errors (4xx response to RCPT) over the same connection as other
4280 messages routed to the same hosts. Such deliveries are always "forced",
4281 so retry times are not inspected. This resulted in far too many retries
4282 for the affected addresses. The effect occurred only when there were more
4283 hosts than the hosts_max_try setting in the smtp transport when it had
4284 the 4xx errors. Those hosts that it had tried were not added to the list
4285 of hosts for which the message was waiting, so if all were tried, there
4286 was no problem. Two fixes have been applied:
4288 (i) If there are any address or message errors in an SMTP delivery, none
4289 of the hosts (tried or untried) are now added to the list of hosts
4290 for which the message is waiting, so the message should not be a
4291 candidate for sending over the same connection that was used for a
4292 successful delivery of some other message. This seems entirely
4293 reasonable: after all the message is NOT "waiting for some host".
4294 This is so "obvious" that I'm not sure why it wasn't done
4295 previously. Hope I haven't missed anything, but it can't do any
4296 harm, as the worst effect is to miss an optimization.
4298 (ii) If, despite (i), such a delivery is accidentally attempted, the
4299 routing retry time is respected, so at least it doesn't keep
4300 hammering the server.
4302 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4303 in ${readsocket because some servers need this prod.
4305 PH/32 Added some extra debug output when updating a wait-xxx database.
4307 PH/33 The hint "could be header name not terminated by colon", which has been
4308 given for certain expansion errors for a long time, was not being given
4309 for the ${if def:h_colon_omitted{... case.
4311 PH/34 The spec says: "With one important exception, whenever a domain list is
4312 being scanned, $domain contains the subject domain." There was at least
4313 one case where this was not true.
4315 PH/35 The error "getsockname() failed: connection reset by peer" was being
4316 written to the panic log as well as the main log, but it isn't really
4317 panic-worthy as it just means the connection died rather early on. I have
4318 removed the panic log writing for the ECONNRESET error when getsockname()
4321 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4322 runs only) independently of the message's sender address. This meant
4323 that, if the 4xx error was in fact related to the sender, a different
4324 message to the same recipient with a different sender could confuse
4325 things. In particular, this can happen when sending to a greylisting
4326 server, but other circumstances could also provoke similar problems.
4327 I have changed the default so that the retry time for these errors is now
4328 based a combination of the sender and recipient addresses. This change
4329 can be overridden by setting address_retry_include_sender=false in the
4332 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4333 remote server are returned as part of bounce messages. This was not
4334 happening for LMTP over a pipe (the lmtp transport), but now it is the
4335 same for both kinds of LMTP.
4337 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4338 in header lines that were in fact CNAMEs. This is no longer the case.
4340 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4341 and queue runs started by the daemon processed all messages. This has
4342 been fixed so that -R and -S can now usefully be given with -q<time>.
4344 PH/40 Import PCRE release 6.7 (fixes some bugs).
4346 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4348 PH/42 Give an error if -q is specified more than once.
4350 PH/43 Renamed the variables $interface_address and $interface_port as
4351 $received_ip_address and $received_port, to make it clear that these
4352 values apply to message reception, and not to the outgoing interface when
4353 a message is delivered. (The old names remain recognized, of course.)
4355 PH/44 There was no timeout on the connect() call when using a Unix domain
4356 socket in the ${readsocket expansion. There now is.
4358 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4359 be meaningful with "accept".
4361 SC/01 Eximstats V1.43
4362 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4364 SC/02 Eximstats V1.44
4365 Use a glob alias rather than an array ref in the generated
4366 parser. This improves both readability and performance.
4368 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4369 Collect SpamAssassin and rejection statistics.
4370 Don't display local sender or destination tables unless
4371 there is data to show.
4372 Added average volumes into the top table text output.
4374 SC/04 Eximstats V1.46
4375 Collect data on the number of addresses (recipients)
4376 as well as the number of messages.
4378 SC/05 Eximstats V1.47
4379 Added 'Message too big' to the list of mail rejection
4380 reasons (thanks to Marco Gaiarin).
4382 SC/06 Eximstats V1.48
4383 Mainlog lines which have GMT offsets and are too short to
4384 have a flag are now skipped.
4386 SC/07 Eximstats V1.49 (Alain Williams)
4387 Added the -emptyok flag.
4389 SC/08 Eximstats V1.50
4390 Fixes for obtaining the IP address from reject messages.
4392 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4393 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4394 whitespace changes from 4.64-PH/27
4396 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4399 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4400 are found, allow negative numbers in numeric criteria)
4402 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4404 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4405 to match changes made in 4.64-PH/43
4407 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4409 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4410 "domain" is actually an IP address, and force a failure. This locks out
4411 those revolvers/nameservers that support "A-for-A" lookups, in
4412 contravention of the specifications.
4414 PH/48 When a host name was looked up from an IP address, and the subsequent
4415 forward lookup of the name timed out, the host name was left in
4416 $sender_host_name, contrary to the specification.
4418 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4419 restricted to single-key lookups, Exim was not diagnosing an error if
4420 * or *@ was used with a query-style lookup.
4422 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4424 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4425 long ago, but noone interested enough thought of it. Let's just say that
4426 the "1.1" means that there are some new functions that weren't there at
4427 some point in the past.
4429 PH/51 Error processing for expansion failure of helo_data from an smtp
4430 transport during callout processing was broken.
4432 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4433 tested/used via the -bh/-bhc/-bs options.
4435 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4436 bug, fixed in subsequent PCRE releases).
4438 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4439 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4441 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4447 SC/01 Use a glob alias rather than an array ref in eximstats generated
4448 parser. This improves both readability and performance.
4450 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4451 Don't display local sender or destination tables in eximstats unless
4452 there is data to show.
4453 Added average volumes into the eximstats top table text output.
4455 SC/03 Collect data on the number of addresses (recipients) as well
4456 as the number of messages in eximstats.
4458 TF/01 Correct an error in the documentation for the redirect router. Exim
4459 does (usually) call initgroups() when daemonizing.
4461 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4462 with consistent privilege compared to when running as a daemon.
4464 TF/03 Note in the spec that $authenticated_id is not set for local
4465 submissions from trusted users.
4467 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4468 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4470 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4471 by adding some example configuration directives to the default
4472 configuration file. A little bit of work is required to uncomment the
4473 directives and define how usernames and passwords are checked, but
4474 there is now a framework to start from.
4476 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4477 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4478 without this. I don't know how relevant this is to other LDAP libraries.
4480 PH/02 Add the verb name to the "unknown ACL verb" error.
4482 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4484 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4486 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4487 directory not expanded when it should be if an expanded home directory
4488 was set for the address (which is overridden by the transport).
4490 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4493 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4494 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4495 because it is too late at that time, and has no effect.
4497 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4498 security issue with \' (bugzilla #107). I could not use the
4499 PQescapeStringConn() function, because it needs a PGconn value as one of
4502 PH/08 When testing addresses using -bt, indicate those final addresses that
4503 are duplicates that would not cause an additional delivery. At least one
4504 person was confused, thinking that -bt output corresponded to deliveries.
4505 (Suppressing duplicates isn't a good idea as you lose the information
4506 about possibly different redirections that led to the duplicates.)
4508 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4509 systems where poll() doesn't work, in particular OS X.
4511 PH/10 Added more information to debugging output for retry time not reached.
4513 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4514 operations in malware.c.
4516 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4519 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4520 syslog_duplication set false, log lines that would normally be written
4521 both the the main log and to the reject log were not written to syslog at
4524 PH/14 In the default configuration, change the use of "message" in ACL warn
4525 statements to "add_header".
4527 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4528 not followed by a command (e.g. "seen endif").
4530 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4531 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4534 PH/17 Added extra conditions to the default value of delay_warning_condition
4538 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4539 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4540 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4543 The Auto-Submitted: and various List- headers are standardised, whereas I
4544 don't think Precedence: ever was.
4546 PH/18 Refactored debugging code in route_finduser() to show more information,
4547 in particular, the error code if getpwnam() issues one.
4549 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4550 This is apparently needed in addition to the PH/07 change above to avoid
4551 any possible encoding problems.
4553 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4554 but not after initializing Perl.
4556 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4557 output them only if debugging. By default they are written stderr,
4558 apparently, which is not desirable.
4560 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4563 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4566 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4568 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4569 authenticated or an ident call has been made. Suppress the default
4570 values for $authenticated_id and $authenticated_sender (but permit -oMai
4571 and -oMas) when testing with -bh.
4573 PH/24 Re-jigged the order of the tests in the default configuration so that the
4574 tests for valid domains and recipients precede the DNS black list and CSA
4575 tests, on the grounds that those ones are more expensive.
4577 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4578 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4579 This bug exists in every version of Exim that I still have, right back to
4582 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4583 However, an attempt to turn on case-sensitivity in a regex key by
4584 including (?-i) didn't work because the subject string was already
4585 lowercased, and the effects were non-intuitive. It turns out that a
4586 one-line patch can be used to allow (?-i) to work as expected.
4592 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4593 other effects) broke the use of negated acl sub-conditions.
4595 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4598 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4599 "Deny" causes Exim to reject the incoming connection with a 554 error.
4600 Unfortunately, if there is a major crisis, such as a disk failure,
4601 tcp-wrappers gives "deny", whereas what one would like would be some
4602 kind of temporary error. A kludge has been added to help with this.
4603 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4604 554 error is used if errno is still zero or contains ENOENT (which occurs
4605 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4608 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4610 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4611 errors. Otherwise a message that provokes a temporary error (when other
4612 messages do not) can cause a whole host to time out.
4614 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4615 addresses were routed directly to files or pipes from a redirect router.
4616 File deliveries just didn't batch; pipe deliveries might have suffered
4619 PH/06 A failure to get a lock for a hints database would erroneously always say
4620 "Failed to get write lock", even when it was really a read lock.
4622 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4623 of 0600. This has been changed to use the value of the lockfile_mode
4624 option (which defaults to 0600).
4626 PH/08 Applied small patch from the Sieve maintainer.
4628 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4629 folder from quota calculations, a direct delivery into this folder messed
4630 up the contents of the maildirsize file. This was because the regex was
4631 used only to exclude .Trash (or whatever) when the size of the mailbox
4632 was calculated. There was no check that a delivery was happening into an
4633 excluded directory. This bug has been fixed by ignoring all quota
4634 processing for deliveries into excluded directories.
4636 PH/10 Added the maildirfolder_create_regex option to appendfile.
4642 PH/01 The code for finding all the local interface addresses on a FreeBSD
4643 system running IPv6 was broken. This may well have applied to all BSD
4644 systems, as well as to others that have similar system calls. The broken
4645 code found IPv4 interfaces correctly, but gave incorrect values for the
4646 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4647 that it would not match correctly against @[] and not recognize the IPv6
4650 PH/02 The ipliteral router was not recognizing addresses of the form user@
4651 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4653 PH/03 Added disable_ipv6.
4655 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4656 decoded form, because it is most often used to construct To: headers
4657 lines in autoreplies, and the decoded form may well be syntactically
4658 invalid. However, $reply_address has leading white space removed, and all
4659 newlines turned into spaces so that the autoreply transport does not
4662 PH/05 If group was specified without a user on a router, and no group or user
4663 was specified on a transport, the group from the router was ignored.
4665 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4666 for visible compile-time settings that can be used to change these
4667 numbers, for those that want even more. Backwards compatibility with old
4668 spool files has been maintained. However, going back to a previous Exim
4669 release will lost any variables that are in spool files.
4671 PH/07 Two small changes when running in the test harness: increase delay when
4672 passing a TCP/IP connection to a new process, in case the original
4673 process has to generate a bounce, and remove special handling of
4674 127.0.0.2 (sic), which is no longer necessary.
4676 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4677 be the same on different OS.
4679 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4682 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4683 whether --show-vars was specified or not
4685 JJ/02 exipick: Added support for new ACL variable spool format introduced
4688 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4689 syntactically invalid From: or Reply-to: line, and a filter used this to
4690 generate an autoreply, and therefore failed to obtain an address for the
4691 autoreply, Exim could try to deliver to a non-existent relative file
4692 name, causing unrelated and misleading errors. What now happens is that
4693 it logs this as a hard delivery error, but does not attempt to create a
4696 PH/11 The exinext utility has a -C option for testing purposes, but although
4697 the given file was scanned by exinext itself; it wasn't being passed on
4698 when Exim was called.
4700 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4701 an end-of-file indication when reading a command response.
4703 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4704 compiled. In many other places in Exim, IPv6 addresses are always
4705 recognized, so I have changed this. It also means that IPv4 domain
4706 literals of the form [IPV4:n.n.n.n] are now always recognized.
4708 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4709 used if the router is not running as root, for example, when verifying at
4710 ACL time, or when using -bh. The debugging output from this situation was
4711 non-existent - all you got was a failure to exec. I have made two
4714 (a) Failures to set uid/gid, the current directory, or a process leader
4715 in a subprocess such as that created by queryprogram now generate
4716 suitable debugging output when -d is set.
4718 (b) The queryprogram router detects when it is not running as root,
4719 outputs suitable debugging information if -d is set, and then runs
4720 the subprocess without attempting to change uid/gid.
4722 PH/15 Minor change to Makefile for building test_host (undocumented testing
4725 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4726 additional section of a DNS packet that returns MX or SRV records.
4727 Instead, it always explicitly searches for A/AAAA records. This avoids
4728 major problems that occur when a DNS server includes only records of one
4729 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4730 fixed another bug: if SRV records were looked up and the corresponding
4731 address records were *not* found in the additional section, the port
4732 values from the SRV records were lost.
4734 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4735 using the correct key (the original address) when searching the retry
4736 rules in order to find which one to use for generating the retry hint.
4738 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4739 adding the default one. Similarly, if it contains a Reply-To: header, the
4740 errors_reply_to option, if set, is not used.
4742 PH/19 When calculating a retry time, Exim used to measure the "time since
4743 failure" by looking at the "first failed" field in the retry record. Now
4744 it does not use this if it is later than than the arrival time of the
4745 message. Instead it uses the arrival time. This makes for better
4746 behaviour in cases where some deliveries succeed, thus re-setting the
4747 "first failed" field. An example is a quota failure for a huge message
4748 when small messages continue to be delivered. Without this change, the
4749 "time since failure" will always be short, possible causing more frequent
4750 delivery attempts for the huge message than are intended.
4751 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4753 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4754 $1, $2, $3) because the numerical variables can be reset during some
4755 expansion items (e.g. "match"), thereby losing the authentication data.
4757 PH/21 Make -bV show the size of off_t variables so that the test suite can
4758 decide whether to run tests for quotas > 2G.
4760 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4761 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4762 filecount value is greater than 2G or if a quota value is greater than 2G
4763 on a system where the size of off_t is not greater than 4, a panic error
4766 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4767 never match. The debug and -bh output now contains an explicit error
4768 message indicating a malformed IPv4 address or mask.
4770 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4771 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4772 PH/23 above applies.
4774 PH/25 Do not write to syslog when running in the test harness. The only
4775 occasion when this arises is a failure to open the main or panic logs
4776 (for which there is an explicit test).
4778 PH/26 Added the /no_tell option to "control=freeze".
4780 PH/27 If a host name lookup failed very early in a connection, for example, if
4781 the IP address matched host_lookup and the reverse lookup yielded a name
4782 that did not have a forward lookup, an error message of the form "no IP
4783 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4784 could be logged. Now it outputs the IP address instead of "NULL".
4786 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4787 allows the sender and the authenticated sender to be set when
4788 submitting a message from within Exim. Since child_open_exim() is
4789 documented for local_scan(), the new function should be too.
4791 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4792 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4793 results in an empty string is now treated as unset.
4795 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4797 PH/31 Added sender_verify_fail logging option.
4799 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4800 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4801 tidied the source and removed it altogether.
4803 PH/33 When a queue run was abandoned because the load average was too high, a
4804 log line was always written; now it is written only if the queue_run log
4805 selector is set. In addition, the log line for abandonment now contains
4806 information about the queue run such as the pid. This is always present
4807 in "start" and "stop" lines but was omitted from the "abandon" line.
4809 PH/34 Omit spaces between a header name and the colon in the error message that
4810 is given when verify = headers_syntax fails (if there are lots of them,
4811 the message gets confusing).
4813 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4814 names, as there are now some PTR records that contain slashes. This check
4815 is only to protect against broken name servers that fall over on strange
4816 characters, so the fact that it applies to all lookups doesn't matter.
4818 PH/36 Now that the new test suite is complete, we can remove some of the
4819 special code in Exim that was needed for the old test suite. For example,
4820 sorting DNS records because real resolvers return them in an arbitrary
4821 order. The new test suite's fake resolver always returns records in the
4824 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4825 bounces) except when queue_only is set, to avoid logging races between
4826 the different processes.
4828 PH/38 Panic-die if .include specifies a non-absolute path.
4830 PH/39 A tweak to the "H" retry rule from its user.
4832 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4833 a label. They prevented compilation on older perls.
4835 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4836 a warning to be raised on newish perls.
4838 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4839 on queue. Changes to match documented behaviour of showing count of
4840 messages matching specified criteria.
4842 PH/40 Changed the default ident timeout from 30s to 5s.
4844 PH/41 Added support for the use of login_cap features, on those BSD systems
4845 that have them, for controlling the resources used by pipe deliveries.
4847 PH/42 The content-scanning code uses fopen() to create files in which to put
4848 message data. Previously it was not paying any attention to the mode of
4849 the files. Exim runs with umask(0) because the rest of the code creates
4850 files with open(), and sets the required mode explicitly. Thus, these
4851 files were ending up world-writeable. This was not a big issue, because,
4852 being within the spool directory, they were not world-accessible. I have
4853 created a function called modefopen, which takes an additional mode
4854 argument. It sets umask(777), creates the file, chmods it to the required
4855 mode, then resets the umask. All the relevant calls to fopen() in the
4856 content scanning code have been changed to use this function.
4858 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4859 to 24 hours. This avoids potential overflow problems when processing G
4860 and H retry rules. I suspect nobody ever tinkers with this value.
4862 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4864 PH/45 When the plaintext authenticator is running as a client, the server's
4865 challenges are checked to ensure they are valid base64 strings. By
4866 default, the authentication attempt is cancelled if an invalid string is
4867 received. Setting client_ignore_invalid_base64 true ignores these errors.
4868 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4869 they are received. Thus, the responses can be made to depend on the
4870 challenges. If an invalid string is ignored, an empty string is placed in
4873 PH/46 Messages that are created by the autoreply transport now contains a
4874 References: header, in accordance with RFCs 2822 and 3834.
4876 PH/47 Added authenticated_sender_force to the smtp transport.
4878 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4880 PH/49 Installed latest patch from the Sieve maintainer.
4882 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4883 was also set, the appendfile transport was unnecessarily scanning a
4884 directory of message files (e.g. for maildir delivery) to find the count
4885 of files (along with the size), even though it did not need this
4886 information. It now does the scan only if it needs to find either the
4887 size of the count of files.
4889 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4891 PH/52 Two bugs concerned with error handling when the smtp transport is
4894 (i) Exim was not creating retry information for temporary errors given
4895 for individual recipients after the DATA command when the smtp transport
4896 was used in LMTP mode. This meant that they could be retried too
4897 frequently, and not timed out correctly.
4899 (ii) Exim was setting the flag that allows error details to be returned
4900 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4901 recipients that were returned after the DATA command.
4903 PH/53 This is related to PH/52, but is more general: for any failing address,
4904 when detailed error information was permitted to be returned to the
4905 sender, but the error was temporary, then after the final timeout, only
4906 "retry timeout exceeded" was returned. Now it returns the full error as
4907 well as "retry timeout exceeded".
4909 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4910 do this, and (what is worse) MTAs that accept it.
4912 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4913 will now be deprecated.
4915 PH/56 New os.c-cygwin from the Cygwin maintainer.
4917 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4918 formats (previously only available in exim formats via -bpr, -bpru,
4919 and -bpra. Now also available in native and exiqgrep formats)
4921 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4922 with very large, slow to parse queues
4924 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4926 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4928 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4929 responses to authentication challenges, though it was showing the
4930 challenges; (ii) I've removed the CR characters from the debug output for
4933 PH/58 Allow for the insertion of a newline as well as a space when a string
4934 is turned into more than one encoded-word during RFC 2047 encoding. The
4935 Sieve code now uses this.
4937 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4938 data_4xx, lost_connection, tls_required.
4940 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4941 message was being sent as an SMTP response.
4943 PH/61 Add -l and -k options to exicyclog.
4945 PH/62 When verifying, if an address was redirected to one new address, so that
4946 verification continued, and the new address failed or deferred after
4947 having set something in $address_data, the value of $address_data was not
4948 passed back to the ACL. This was different to the case when no
4949 redirection occurred. The value is now passed back in both cases.
4951 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4952 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4953 use login_cap.h, so on its own it isn't the distinguishing feature. The
4954 new name refers directly to the setclassresources() function.
4956 PH/65 Added configuration files for NetBSD3.
4958 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4960 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4961 is preferred over IPv4.
4963 PH/68 The bounce_return_message and bounce_return_body options were not being
4964 honoured for bounces generated during the reception of non-SMTP messages.
4965 In particular, this applied to messages rejected by the ACL. This bug has
4966 been fixed. However, if bounce_return_message is true and bounce_return_
4967 body is false, the headers that are returned for a non-SMTP message
4968 include only those that have been read before the error was detected.
4969 (In the case of an ACL rejection, they have all been read.)
4971 PH/69 The HTML version of the specification is now built in a directory called
4972 spec_html instead of spec.html, because the latter looks like a path with
4973 a MIME-type, and this confuses some software.
4975 PH/70 Catch two compiler warnings in sieve.c.
4977 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4978 function verify_get_ident() calls ip_connect() to connect a socket, but
4979 if the "connect()" function timed out, ip_connect() used to close the
4980 socket. However, verify_get_ident() also closes the socket later, and in
4981 between Exim writes to the log, which may get opened at this point. When
4982 the socket was closed in ip_connect(), the log could get the same file
4983 descriptor number as the socket. This naturally causes chaos. The fix is
4984 not to close the socket in ip_connect(); the socket should be closed by
4985 the function that creates it. There was only one place in the code where
4986 this was missing, in the iplookup router, which I don't think anybody now
4987 uses, but I've fixed it anyway.
4989 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4990 well as to direct DNS lookups. Otherwise the handling of names in host
4991 lists is inconsistent and therefore confusing.
4997 PH/01 Two changes to the default runtime configuration:
4999 (1) Move the checks for relay_from_hosts and authenticated clients from
5000 after to before the (commented out) DNS black list checks.
5002 (2) Add control=submission to the relay_from_hosts and authenticated
5003 clients checks, on the grounds that messages accepted by these
5004 statements are most likely to be submissions.
5006 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5008 (1) Generate an error if the third argument for the ${prvs expansion is
5011 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5014 (3) Reset the variables that are obtained from the first argument of
5015 ${prvscheck and used in the second argument before leaving the code,
5016 because their memory is reclaimed, so using them afterwards may do
5019 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5020 one (it's much easier than Tom thought :-).
5022 (5) Because of (4), we can now allow for the use of $prvscheck_result
5023 inside the third argument.
5025 PH/03 For some reason, the default setting of PATH when running a command from
5026 a pipe transport was just "/usr/bin". I have changed it to
5029 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5030 anything to be listed in the output from -bV.
5032 PH/05 When a filter generated an autoreply, the entire To: header line was
5033 quoted in the delivery log line, like this:
5035 => >A.N.Other <ano@some.domain> <original@ddress> ...
5037 This has been changed so that it extracts the operative address. There
5038 may be more than one such address. If so, they are comma-separated, like
5041 => >ano@some.domain,ona@other.domain <original@ddress> ...
5043 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5044 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5045 not being looked up in the rDNS to get a host name, Exim was showing the
5046 IP address twice in Received: lines, even though the IP addresses were
5047 identical. For example:
5049 Received: from [1.2.3.4] (helo=[1.2.3.4])
5051 However, if the real host name was known, it was omitting the HELO data
5052 if it matched the actual IP address. This has been tidied up so that it
5053 doesn't show the same IP address twice.
5055 PH/07 When both +timestamp and +memory debugging was on, the value given by
5056 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5057 called by the debug printing, thereby overwriting the timestamp buffer.
5058 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5060 PH/08 When the original message was included in an autoreply transport, it
5061 always said "this is a copy of the message, including all the headers",
5062 even if body_only or headers_only was set. It now gives an appropriate
5065 PH/09 Applied a patch from the Sieve maintainer which:
5067 o fixes some comments
5068 o adds the (disabled) notify extension core
5069 o adds some debug output for the result of if/elsif tests
5070 o points to the current vacation draft in the documentation
5071 and documents the missing references header update
5075 o fixes a bug in processing the envelope test (when testing
5076 multiple envelope elements, the last element determined the
5079 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5080 Electronic Mail") by including:
5082 Auto-submitted: auto-generated
5084 in the messages that it generates (bounce messages and others, such as
5085 warnings). In the case of bounce messages for non-SMTP messages, there was
5086 also a typo: it was using "Auto_submitted" (underscore instead of
5087 hyphen). Since every message generated by Exim is necessarily in response
5088 to another message, thes have all been changed to:
5090 Auto-Submitted: auto-replied
5092 in accordance with these statements in the RFC:
5094 The auto-replied keyword:
5096 - SHOULD be used on messages sent in direct response to another
5097 message by an automatic process,
5099 - MUST NOT be used on manually-generated messages,
5101 - MAY be used on Delivery Status Notifications (DSNs) and Message
5102 Disposition Notifications (MDNs),
5104 - MUST NOT be used on messages generated by automatic or periodic
5105 processes, except for messages which are automatic responses to
5108 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5109 to the default Received: header definition.
5111 PH/12 Added log selector acl_warn_skipped (default on).
5113 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5114 variables because (a) they are in the wrong storage pool and (b) even if
5115 they were copied, it wouldn't work properly because of the caching.
5117 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5118 checking when decoding. Apparently there are clients that generate
5119 overlong encoded strings. Why am I not surprised?
5121 PH/15 If the first argument of "${if match_address" was not empty, but did not
5122 contain an "@" character, Exim crashed. Now it writes a panic log message
5123 and treats the condition as false.
5125 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5127 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5128 extension "envelope-auth". The code is finished and in agreement with
5129 other implementations, but there is no documentation so far and in fact,
5130 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5131 not changing the active code.
5133 Print executed "if" and "elsif" statements when debugging is used. This
5134 helps a great deal to understand what a filter does.
5136 Document more things not specified clearly in RFC3028. I had all this
5137 sorted out, when out of a sudden new issues came to my mind. Oops."
5139 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5142 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5143 canonical form (as documented). However, after a host name lookup from
5144 the IP address, check_host() was doing a simple string comparison with
5145 addresses acquired from the DNS when checking that the found name did
5146 have the original IP as one of its addresses. Since any found IPv6
5147 addresses are likely to be in abbreviated form, the comparison could
5148 fail. Luckily, there already exists a function for doing the comparison
5149 by converting both addresses to binary, so now that is used instead of
5150 the text comparison.
5152 PH/20 There was another similar case to PH/19, when a complete host name was
5153 given in a host list; looking up its IP address could give an abbreviated
5154 form, whereas the current host's name might or might not be abbreviated.
5155 The same fix has been applied.
5161 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5162 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5165 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5166 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5168 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5170 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5171 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5172 Also, for "Auto-Submitted", treat anything other than "no" as
5173 non-personal, in accordance with RFC 3834. (Previously it treated
5174 anything starting "auto-" as non-personal.)
5176 TF/01 The control=submission/name=... option had a problem with syntax
5177 errors if the name included a slash character. The /name= option
5178 now slurps the rest of the string, so it can include any characters
5179 but it must come last in the list of options (after /sender_retain
5182 PH/05 Some modifications to the interface to the fake nameserver for the new
5190 TK/01 Added the "success_on_redirect" address verification option. See
5191 NewStuff for rationale and an example.
5193 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5195 PH/02 Patch to exigrep to allow it to work on syslog lines.
5197 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5198 fread() to skip over the body file's header line, because in Cygwin the
5199 header line is locked and is inaccessible.
5201 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5202 co-exist for some time) to make it clear that it is the Exim ID that is
5203 referenced, not the Message-ID: header line.
5205 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5206 string_format() function, because snprintf() does not exist on all
5209 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5210 ${stat: expansion item.
5212 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5213 protocol synchronization error", to keep the pedants happy.
5215 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5216 well as for IRIX systems, when gcc is being used. See the host.c source
5219 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5221 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5224 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5225 transport and to the smtp transport in LMTP mode.
5227 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5229 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5230 run a filter in a subprocess. This could lead to confusion in subsequent
5231 lookups in the parent process. There should also be a search_tidyup() at
5232 the end of the subprocess.
5234 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5235 only if the host matched helo_try_verify_hosts, which caused the
5236 verification to occur when the EHLO/HELO command was issued. The ACL just
5237 tested the remembered result. Now, if a previous verification attempt has
5238 not happened, "verify = helo" does it there and then.
5240 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5242 TK/03 Fix log output including CR from clamd.
5244 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5245 exist provoked a memory error which could cause a segfault.
5247 PH/15 Installed PCRE 6.2
5249 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5251 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5252 of the problem. Specifically, suggested +O2 rather than +O1 for the
5255 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5257 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5258 record, the port was not correctly logged when the outgoing_port log
5259 selector was set (it logged the transort's default port).
5261 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5262 fallback_hosts, and "hosts" in the smtp transport.
5264 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5265 host errors such as "Connection refused".
5267 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5268 authentication with radiusclient 0.4.9:
5270 - Error returned from rc_read_config was caught wrongly
5271 - Username/password not passed on to radius server due to wrong length.
5273 The presumption is that some radiusclient API changes for 4.51/PH/17
5274 were not taken care of correctly. The code is still untested by me (my
5275 Linux distribution still has 0.3.2 of radiusclient), but it was
5276 contributed by a Radius user.
5278 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5279 expanding the "port" option of the smtp transport.
5281 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5282 while reading a MIME header. Thanks to Tom Hughes for a patch.
5284 PH/24 Include config.h inside local_scan.h so that configuration settings are
5287 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5288 that in an ACL for RCPT (for example), you can examine exactly what was
5291 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5292 commands, but it was not correctly comparing the address with the actual
5293 client host address. Thus, it would show the EHLO address in Received:
5294 header lines when this was not necessary.
5296 PH/27 Added the % operator to ${eval:}.
5298 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5299 it should be ignoring failures (because with -C, for example, it has lost
5300 privilege). It wasn't ignoring creation failures other than "already
5303 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5306 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5307 because an input line was too long, either on its own, or by virtue of
5308 too many continuations, the temporary file was not being removed, and the
5309 return code was incorrect.
5311 PH/31 Missing "BOOL" in function definition in filtertest.c.
5313 PH/32 Applied Sieve patches from the maintainer.
5315 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5317 PH/33 Added "verify = not_blind".
5319 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5320 Local/Makefile (with some defaults set). These are used in built scripts
5321 such as exicyclog, but they have never been used in the exim_install
5322 script (though there are many overriding facilities there). I have
5323 arranged that the exim_install script now takes note of these two
5326 PH/35 Installed configuration files for Dragonfly.
5328 PH/36 When a locally submitted message by a trusted user did not contain a
5329 From: header, and the sender address was obtained from -f or from an SMTP
5330 MAIL command, and the trusted user did not use -F to supply a sender
5331 name, $originator_name was incorrectly used when constructing a From:
5332 header. Furthermore, $originator_name was used for submission mode
5333 messages from external hosts without From: headers in a similar way,
5334 which is clearly wrong.
5336 PH/37 Added control=suppress_local_fixups.
5338 PH/38 When log_selector = +received_sender was set, and the addition of the
5339 sender made the log line's construction buffer exactly full, or one byte
5340 less than full, an overflow happened when the terminating "\n" was
5343 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5344 when the result of a list match is failure because a DNS lookup failed.
5346 PH/40 RM_COMMAND is now used in the building process.
5348 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5349 the "build-* directories that it finds.
5351 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5352 address in a domain literal was a prefix of an interface address.
5354 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5355 when verifying a sender address, unless rewrite_headers is false.
5357 PH/44 Wrote a long comment about why errors_to addresses are verified as
5358 recipients, not senders.
5360 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5361 the ratelimit ACL was added.
5363 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5365 PH/46 Added extra information about PostgreSQL errors to the error string.
5367 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5368 avoiding the need to install special zones in a real server. This is
5369 backwards compatible; if it can't find the fake resolver, it drops back.
5370 Thus, both old and new test suites can be run.
5372 TF/02 Added util/ratelimit.pl
5374 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5375 clock is set back in time.
5377 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5378 Candler <B.Candler@pobox.com>.
5380 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5381 OK for addresses that are the result of redirections.
5383 PH/48 A number of further additions for the benefit of the new test suite,
5384 including a fake gethostbyname() that interfaces to the fake DNS resolver
5387 TF/06 The fix for widen_domains has also been applied to qualify_single and
5388 search_parents which are the other dnslookup options that can cause
5391 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5394 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5396 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5397 service" if it encountered an MX record with an empty target hostname.
5398 The message is now "an MX or SRV record indicated no SMTP service".
5400 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5401 if the DNS of the sending site is misconfigured. This is quite a
5402 common situation. This change restores the behaviour of treating a
5403 helo verification defer as a failure.
5405 PH/51 If self=fail was set on a router, the bounce message did not include the
5406 actual error message.
5412 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5414 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5415 command itself ended in error, the underlying message about the transport
5416 filter timeout was being overwritten with the pipe command error. Now the
5417 underlying error message should be appended to the second error message.
5419 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5421 PH/02 Reduce the amount of output that "make" produces by default. Full output
5422 can still be requested.
5424 PH/03 The warning log line about a condition test deferring for a "warn" verb
5425 was being output only once per connection, rather than after each
5426 occurrence (because it was using the same function as for successful
5427 "warn" verbs). This seems wrong, so I have changed it.
5429 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5430 it should not have, which might have caused a crash in the right
5431 circumstances, but probably never did.
5433 PH/04 Installed a modified version of Tony Finch's patch to make submission
5434 mode fix the return path as well as the Sender: header line, and to
5435 add a /name= option so that you can make the user's friendly name appear
5438 TF/03 Added the control = fakedefer ACL modifier.
5440 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5441 Mark Lowes for thorough testing.
5443 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5445 TK/03 Merged latest SRS patch from Miles Wilton.
5447 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5448 with the definition in sysexits.h (which is #included earlier).
5449 Fortunately, Exim does not actually use EX_OK. The code used to try to
5450 preserve the sysexits.h value, by assuming that macro definitions were
5451 scanned for macro replacements. I have been disabused of this notion,
5452 so now the code just undefines EX_OK before #including unistd.h.
5454 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5455 in the smtp transport. When a block could not be written in a single
5456 write() function, the timeout was being re-applied to each part-write.
5457 This seems wrong - if the receiver was accepting one byte at a time it
5458 would take for ever. The timeout is now adjusted when this happens. It
5459 doesn't have to be particularly precise.
5461 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5462 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5464 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5465 without a domain if the check to <postmaster@domain> fails.
5467 SC/01 Eximstats: added -xls and the ability to specify output files
5468 (patch written by Frank Heydlauf).
5470 SC/02 Eximstats: use FileHandles for outputting results.
5472 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5474 SC/04 Eximstats: fixed display of large numbers with -nvr option
5476 SC/05 Eximstats: fixed merging of reports with empty tables.
5478 SC/06 Eximstats: added the -include_original_destination flag
5480 SC/07 Eximstats: removed tabs and trailing whitespace.
5482 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5484 TK/06 MBOX spool code: Add real "From " MBOX separator line
5485 so the .eml file is really in mbox format (even though
5486 most programs do not really care). Patch from Alex Miller.
5488 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5489 The latter is generated from $received_to and is only set if the
5490 message has one envelope recipient. SA can use these headers,
5491 obviously out-of-the-box. Patch from Alex Miller.
5493 PH/08 The ${def test on a variable was returning false if the variable's
5494 value was "0", contrary to what the specification has always said!
5495 The result should be true unless the variable is empty.
5497 PH/09 The syntax error of a character other than { following "${if
5498 def:variable_name" (after optional whitespace) was not being diagnosed.
5499 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5500 accidental colon was present, for example, could give incorrect results.
5502 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5503 result is used (not including appendfile, where other changes are about
5506 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5507 This involved changing a lot of size variables from int to off_t. It
5508 should work with maildirs and everything.
5510 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5511 spamd dying while we are connected to it.
5513 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5516 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5517 function for BDB 4.3.
5519 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5521 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5522 the sources for PCRE have been re-arranged and more files are now
5525 PH/15 The code I had for printing potentially long long variables in PH/11
5526 above was not the best (it lost precision). The length of off_t variables
5527 is now inspected at build time, and an appropriate printing format (%ld
5528 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5529 to be "long long int" or "long int". This is needed for the internal
5530 formatting function string_vformat().
5532 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5533 the configuration file to be ":syslog", then the script "guesses" where
5534 the logs files are, rather than using the compiled in default. In our
5535 case the guess is not the same as the compiled default, so the script
5536 suddenly stopped working when I started to use syslog. The patch checks
5537 to see if log_file_path is "". If so, it attempts to read it from exim
5538 with no configuration file to get the compiled in version, before it
5539 falls back to the previous guessing code."
5541 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5542 implementing BATV in an Exim configuration. See NewStuff for the gory
5545 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5546 Makefile that are specific to HP-UX.
5548 PH/18 If the "use_postmaster" option was set for a recipient callout together
5549 with the "random" option, the postmaster address was used as the MAIL
5550 FROM address for the random test, but not for the subsequent recipient
5551 test. It is now used for both.
5553 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5554 patch removes a few documentation additions to RFC 3028, because the
5555 latest draft now contains them. It adds the new en;ascii-case comparator
5556 and a new error check for 8bit text in MIME parts. Comparator and
5557 require names are now matched exactly. I enabled the subaddress
5558 extension, but it is not well tested yet (read: it works for me)."
5560 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5561 rework some of the code of TK/09 above to avoid the hardwired use of
5562 "%lld" and "long long". Replaced the call to snprintf() with a call to
5565 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5566 records point to non-existent hosts", "retry timeout exceeded", and
5567 "retry time not reached for any host after a long failure period".
5569 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5570 experimental DomainKeys support:
5572 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5573 (2) On an error such as an illegally used "control", the wrong name for
5574 the control was given.
5576 These problems did NOT occur unless DomainKeys support was compiled.
5578 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5580 PH/24 Added ${if match_ip condition.
5582 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5583 databases so that it will be absolutely obvious if a crash occurs in the
5584 DB library. This is a regular occurrence (often caused by mis-matched
5587 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5588 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5589 picked up on a user's system that detects such things. There doesn't seem
5590 to be a gcc warning option for this - only an attribute that has to be
5591 put on the function's prototype. It seems that in Fedora Core 4 they have
5592 set this on a number of new functions. No doubt there will be more in due
5595 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5596 specify a transport. However, if an address that was verified by such a
5597 router was the subject of a callout, Exim crashed because it tried to
5598 read the rcpt_include_affixes from the non-existent transport. Now it
5599 just assumes that the setting of that option is false. This bug was
5600 introduced by 4.51/PH/31.
5602 PH/28 Changed -d+all to exclude +memory, because that information is very
5603 rarely of interest, but it makes the output a lot bigger. People tend to
5604 do -d+all out of habit.
5606 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5607 code in os-type was giving problems when libc.so lives in lib64, like on
5610 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5611 aren't the modern standard, and it seems that some systems' include files
5612 don't always have them. Exim was already checking for some of the newer
5613 ones like T_AAAA, and defining it itself. I've added checks for all the
5614 record types that Exim uses.
5616 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5617 not automatically generating a new one, as it is supposed to. This
5618 prevented TLS from working. If the file did exist, but contained invalid
5619 data, a new version was generated, as expected. It was only the case of a
5620 non-existent file that was broken.
5622 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5623 with a change in libdomainkeys > 0.64.
5625 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5626 from DNS. If the selector record carries the flag, it now has
5627 precedence over the domain-wide flag.
5629 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5631 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5632 the use of a port name that isn't defined in /etc/services) occurred, the
5633 message was deferred as in a normal delivery, and thus remained on the
5634 spool, instead of being failed because of the mua_wrapper setting. This
5635 is now fixed, and I tidied up some of the mua_wrapper messages at the
5638 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5639 the messages in a hash of arrays rather than using individual hashes.
5640 This is a bit cleaner and results in dramatic memory savings, albeit
5641 at a slight CPU cost.
5643 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5644 as requested by Marc Sherman.
5646 SC/10 Eximstats: added histograms for user specified patterns as requested
5649 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5651 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5652 fopen() in the content-scanning modules that did not already have it.
5658 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5659 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5661 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5663 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5665 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5666 to test Sieve filters that use "vacation".
5668 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5669 that changes the way the GnuTLS parameters are stored in the cache file.
5670 The new format can be generated externally. For backward compatibility,
5671 if the data in the cache doesn't make sense, Exim assumes it has read an
5672 old-format file, and it generates new data and writes a new file. This
5673 means that you can't go back to an older release without removing the
5676 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5677 work if there are any delivery delays because "one_time" forces the
5678 parent to be marked "delivered", so its unseen clone is never tried
5679 again. For this reason, Exim now forbids the simultaneous setting of
5682 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5683 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5684 entry if you want to know the details. The fix, however, neglected to
5685 consider the case when local delivery batching is involved. The test for
5686 "previously delivered" was not happening when checking to see if an
5687 address could be batched with a previous (undelivered) one; under
5688 certain circumstances this could lead to multiple deliveries to the same
5691 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5692 in its include files, and this causes problems building Exim.
5694 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5695 header_syntax) but Exim was just ignoring anything given after a slash.
5696 In particular, this caused confusion with an attempt to use "verify =
5697 reverse_host_lookup/defer_ok". An error is now given when options are
5698 supplied for verify items that do not have them. (Maybe reverse_host_
5699 lookup should have a defer_ok option, but that's a different point.)
5701 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5702 defined by RFC 821) to 2048, because there were problems with some AUTH
5703 commands, and RFC 1869 says the size should be increased for extended
5704 SMTP commands that take arguments.
5706 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5709 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5710 "unknown" error; now it says that the functionality isn't in the binary.
5712 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5713 an address' error message when a string expansion fails (syntax or
5714 whatever). Otherwise the password may appear in the log. Following change
5715 PH/42 below, there is no longer a chance of it appearing in a bounce
5718 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5720 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5721 of its addresses was ever tried. (Bugzilla bug #2.)
5723 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5724 the result incorrectly in the debug output. (It correctly added a newline
5725 to what was transported.)
5727 TF/01 Added $received_time.
5729 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5730 commented out examples of how to interface to a virus scanner and to
5731 SpamAssassin. Also added commented examples of av_scanner and
5732 spamd_address settings.
5734 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5735 and controls are allowed in which ACLs. There were a couple of minor
5736 errors. Some of the entries in the conditions table (which is a table of
5737 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5738 negation of where the condition IS allowed.
5740 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5742 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5743 header file does not have a version number, so I've had to invent a new
5744 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5745 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5746 radiusclient), but it was contributed by a Radius user.
5748 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5749 files or directories, for OpenSSL.
5751 PH/19 When an Exim process that is running as root has to create an Exim log
5752 file, it does so in a subprocess that runs as exim:exim so as to get the
5753 ownership right at creation (otherwise, other Exim processes might see
5754 the file with the wrong ownership). There was no test for failure of this
5755 fork() call, which would lead to the process getting stuck as it waited
5756 for a non-existent subprocess. Forks do occasionally fail when resources
5757 run out. I reviewed all the other calls to fork(); they all seem to check
5760 PH/20 When checking for unexpected SMTP input at connect time (before writing
5761 the banner), Exim was not dealing correctly with a non-positive return
5762 from the read() function. If the client had disconnected by this time,
5763 the result was a log entry for a synchronization error with an empty
5764 string after "input=" when read() returned zero. If read() returned -1
5765 (an event I could not check), uninitialized data bytes were printed.
5766 There were reports of junk text (parts of files, etc) appearing after
5769 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5771 PH/22 Added support for macro redefinition, and (re)definition in between
5772 driver and ACL definitions.
5774 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5775 forgetting to use the resulting value; it was using the unexpanded value.
5777 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5778 hadn't been configured. The fix is from Juergen Kreileder, who
5779 understands it better than I do:
5781 "Here's what I see happening with three configured cyrus_sasl
5782 authenticators configured (plain, login, cram-md5):
5784 On startup auth_cyrus_sasl_init() gets called for each of these.
5785 This means three calls to sasl_listmech() without a specified mech_list.
5786 => SASL tests which mechs of all available mechs actually work
5787 => three warnings about OTP not working
5788 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5790 With the patch, sasl_listmech() also gets called three times. But now
5791 SASL's mech_list option is set to the server_mech specified in the the
5792 authenticator. Or in other words, the answer from sasl_listmech()
5793 gets limited to just the mech you're testing for (which is different
5795 => the return list contains just 'plain' or 'login', 'cram-md5' or
5796 nothing depending on the value of ob->server_mech.
5798 I've just tested the patch: Authentication still works fine,
5799 unavailable mechs specified in the exim configuration are still
5800 caught, and the auth.log warnings about OTP are gone."
5802 PH/25 When debugging is enabled, the contents of the command line are added
5803 to the debugging output, even when log_selector=+arguments is not
5806 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5807 answer is "GNU", and only if the return is "GNU/something" is the answer
5810 PH/27 $acl_verify_message is now set immediately after the failure of a
5811 verification in an ACL, and so is available in subsequent modifiers. In
5812 particular, the message can be preserved by coding like this:
5814 warn !verify = sender
5815 set acl_m0 = $acl_verify_message
5817 Previously, $acl_verify_message was set only while expanding "message"
5818 and "log_message" when a very denied access.
5820 PH/28 Modified OS/os.c-Linux with
5822 -#ifndef OS_LOAD_AVERAGE
5823 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5825 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5826 nomenclature these days.)
5828 PH/29 Installed patch from the Sieve maintainer that adds the options
5829 sieve_useraddress and sieve_subaddress to the redirect router.
5831 PH/30 In these circumstances:
5832 . Two addresses routed to the same list of hosts;
5833 . First host does not offer TLS;
5834 . First host accepts first address;
5835 . First host gives temporary error to second address;
5836 . Second host offers TLS and a TLS session is established;
5837 . Second host accepts second address.
5838 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5839 and peerdn, if requested) that were in fact used only for the second
5842 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5843 attention to any local part prefix or suffix that was matched by the
5844 router that accepted the address. It now behaves in the same way as it
5845 does for delivery: the affixes are removed from the local part unless
5846 rcpt_include_affixes is set on the transport.
5848 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5849 timeout during the DATA phase of an incoming message.
5851 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5852 applied a patch sanctioned by the Sieve maintainer.
5854 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5855 the uid or gid is negative. A case of a negative gid caused this to be
5856 noticed. The fix allows for either to be negative.
5858 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5859 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5862 PH/36 The change PH/12 above was broken. Fixed it.
5864 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5865 the grounds that routing the same address twice would always produce the
5866 same answer. This might have been true once, but it is certainly no
5867 longer true now. Routing a child address may depend on the previous
5868 routing that produced that child. Some complicated redirection strategies
5869 went wrong when messages had multiple recipients, and made Exim's
5870 behaviour dependent on the order in which the addresses were given.
5872 I have moved the duplicate checking until after the routing is complete.
5873 Exim scans the addresses that are assigned to local and remote
5874 transports, and removes any duplicates. This means that more work will be
5875 done, as duplicates will always all be routed, but duplicates are
5876 presumably rare, so I don't expect this is of any significance.
5878 For deliveries to pipes, files, and autoreplies, the duplicate checking
5879 still happens during the routing process, since they are not going to be
5882 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5883 It corrects a timeout issue with spamd. This is Ian's comment: "The
5884 background is that sometimes spamd either never reads data from a
5885 connection it has accepted, or it never writes response data. The exiscan
5886 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5887 blindly assumes that writes won't block so it may never time out."
5889 PH/39 Allow G after quota size as well as K and M.
5891 PH/40 The value set for $authenticated_id in an authenticator may not contain
5892 binary zeroes or newlines because the value is written to log lines and
5893 to spool files. There was no check on this. Now the value is run through
5894 the string_printing() function so that such characters are converted to
5895 printable escape sequences.
5897 PH/41 $message_linecount is a new variable that contains the total number of
5898 lines in the message. Compare $body_linecount, which is the count for the
5901 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5902 bounce and delay warning messages, except in certain special cases, which
5905 (a) An SMTP error message from a remote host;
5906 (b) A message specified in a :fail: redirection;
5907 (c) A message specified in a "fail" command in a system filter;
5908 (d) A message specified in a FAIL return from the queryprogram router;
5909 (e) A message specified by the cannot_route_message router option.
5911 In these cases only, Exim does include the error details in bounce and
5912 warning messages. There are also a few cases where bland messages such
5913 as "unrouteable address" or "local delivery error" are given.
5915 PH/43 $value is now also set for the "else" part of a ${run expansion.
5917 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5918 being worked on, but at least Exim now implements the latest version to
5921 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5922 process to complete was treated as a delivery failure, a timeout while
5923 writing the message to the pipe was logged, but erroneously treated as a
5924 successful delivery. Such timeouts include transport filter timeouts. For
5925 consistency with the overall process timeout, these timeouts are now
5926 treated as errors, giving rise to delivery failures by default. However,
5927 there is now a new Boolean option for the pipe transport called
5928 timeout_defer, which, if set TRUE, converts the failures into defers for
5929 both kinds of timeout. A transport filter timeout is now identified in
5932 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5933 systems where "make" and "gmake" are different, calling "gmake" at top
5934 level broke things. I've arranged for the value of $(MAKE) to be passed
5935 from the Makefile to this script so that it can call the same version of
5939 A note about Exim versions 4.44 and 4.50
5940 ----------------------------------------
5942 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5943 changes of various kinds. As a consequence, a big documentation update was
5944 needed. This delayed the release for rather longer than seemed good, especially
5945 in the light of a couple of (minor) security issues. Therefore, the changes
5946 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5947 release. So 4.44 and 4.50 are in effect two different branches that both start
5950 I have left the 4.50 change log unchanged; it contains all the changes since
5951 4.43. The change log for 4.44 is below; many of its items are identical to
5952 those for 4.50. This seems to be the most sensible way to preserve the
5953 historical information.
5959 1. Minor wording change to the doc/README.SIEVE file.
5961 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5962 computation of the current number of files was incorrect.
5964 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5965 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5968 4. Give more explanation in the error message when the command for a transport
5969 filter fails to execute.
5971 5. There are several places where Exim runs a non-Exim command in a
5972 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5973 was being done only for the command run by the queryprogram router. It is
5974 now done for all such subprocesses. The other cases are: ${run, transport
5975 filters, and the commands run by the lmtp and pipe transports.
5977 6. Added CONFIGURE_GROUP build-time option.
5979 7. Some older OS have a limit of 256 on the maximum number of file
5980 descriptors. Exim was using setrlimit() to set 1000 as a large value
5981 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5982 systems. I've change it so that if it can't get 1000, it tries for 256.
5984 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5985 was an oversight, and furthermore, ever since the addition of extra
5986 controls (e.g. 4.43/32), the checks on when to allow different forms of
5987 "control" were broken. There should now be diagnostics for all cases when a
5988 control that does not make sense is encountered.
5990 9. Added the /retain_sender option to "control=submission".
5992 10. $recipients is now available in the predata ACL (oversight).
5994 11. Tidy the search cache before the fork to do a delivery from a message
5995 received from the command line. Otherwise the child will trigger a lookup
5996 failure and thereby defer the delivery if it tries to use (for example) a
5997 cached ldap connection that the parent has called unbind on.
5999 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6000 of $address_data from the recipient verification was clobbered by the
6001 sender verification.
6003 13. The value of address_data from a sender verification is now available in
6004 $sender_address_data in subsequent conditions in the ACL statement.
6006 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6008 15. Added a new option "connect=<time>" to callout options, to set a different
6011 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6012 was its contents. (It was OK if the option was not defined at all.)
6014 17. A "Completed" log line is now written for messages that are removed from
6015 the spool by the -Mrm option.
6017 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6018 information about exactly what failed.
6020 19. Added -dd to debug only the daemon process.
6022 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6023 handles timeouts, both on the server side and network timeouts. Renamed the
6024 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6026 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6027 It is now set to "smtps".
6029 22. $host_address is now set to the target address during the checking of
6030 ignore_target_hosts.
6032 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6033 being passed; this would have caused $sender_host_name to have been used if
6034 matching the list had actually called for a host name (not very likely,
6035 since this list is usually IP addresses). A host name is now passed as
6038 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6039 code that specifies a non-restarting handler (typically sigaction() in
6040 modern systems) in an attempt to fix a rare and obscure crash bug.
6042 25. Narrowed the window for a race in the daemon that could cause it to ignore
6043 SIGCHLD signals. This is not a major problem, because they are used only to
6044 wake it up if nothing else does.
6046 26. A malformed maildirsize file could cause Exim to calculate negative values
6047 for the mailbox size or file count. Odd effects could occur as a result.
6048 The maildirsize information is now recalculated if the size or filecount
6051 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6052 support for a long time. Removed HAVE_SYS_VFS_H.
6054 28. Installed the latest version of exipick from John Jetmore.
6056 29. In an address list, if the pattern was not a regular expression, an empty
6057 subject address (from a bounce message) matched only if the pattern was an
6058 empty string. Non-empty patterns were not even tested. This was the wrong
6059 because it is perfectly reasonable to use an empty address as part of a
6060 database query. An empty address is now tested by patterns that are
6061 lookups. However, all the other forms of pattern expect the subject to
6062 contain a local part and a domain, and therefore, for them, an empty
6063 address still always fails if the pattern is not itself empty.
6065 30. Exim went into a mad DNS loop when attempting to do a callout where the
6066 host was specified on an smtp transport, and looking it up yielded more
6067 than one IP address.
6069 31. Re-factored the code for checking spool and log partition space into a
6070 function that finds that data and another that does the check. The former
6071 is then used to implement four new variables: $spool_space, $log_space,
6072 $spool_inodes, and $log_inodes.
6074 32. The RFC2047 encoding function was originally intended for short strings
6075 such as real names; it was not keeping to the 75-character limit for
6076 encoded words that the RFC imposes. It now respects the limit, and
6077 generates multiple encoded words if necessary. To be on the safe side, I
6078 have increased the buffer size for the ${rfc2047: expansion operator from
6081 33. It is now permitted to omit both strings after an "if" condition; if the
6082 condition is true, the result is "true". As before, when the second string
6083 is omitted, a false condition yields an empty string. This makes it less
6084 cumbersome to write custom ACL and router conditions.
6086 34. Failure to deliver a bounce message always caused it to be frozen, even if
6087 there was an errors_to setting on the router. The errors_to setting is now
6090 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6091 canonical form (fully expanded) before being placed in
6092 $sender_host_address.
6094 36. The table in the code that translates DNS record types into text (T_A to
6095 "A" for instance) was missing entries for NS and CNAME. It is just possible
6096 that this could have caused confusion if both these types were looked up
6097 for the same domain, because the text type is used as part of Exim's
6098 per-process caching. But the chance of anyone hitting this buglet seems
6101 37. The dnsdb lookup has been extended in a number of ways.
6103 (1) There is a new type, "zns", which walks up the domain tree until it
6104 finds some nameserver records. It should be used with care.
6106 (2) There is a new type, "mxh", which is like "mx" except that it returns
6107 just the host names, not the priorities.
6109 (3) It is now possible to give a list of domains (or IP addresses) to be
6110 looked up. The behaviour when one of the lookups defers can be
6111 controlled by a keyword.
6113 (4) It is now possible to specify the separator character for use when
6114 multiple records are returned.
6116 38. The dnslists ACL condition has been extended: it is now possible to supply
6117 a list of IP addresses and/or domains to be looked up in a particular DNS
6120 39. Added log_selector=+queue_time_overall.
6122 40. When running the queue in the test harness, wait just a tad after forking a
6123 delivery process, to get repeatability of debugging output.
6125 41. Include certificate and key file names in error message when GnuTLS fails
6126 to set them up, because the GnuTLS error message doesn't include the name
6127 of the failing file when there is a problem reading it.
6129 42. Allow both -bf and -bF in the same test run.
6131 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6133 44. The "Exiscan patch" is now merged into the mainline Exim source.
6135 45. Sometimes the final signoff response after QUIT could fail to get
6136 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6137 < 0 before doing a fflush(). This bug looks as though it goes back to the
6138 introduction of TLS in release 3.20, but "sometimes" must have been rare
6139 because the tests only now provoked it.
6141 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6142 (this can affect the format of dates).
6144 47. exim_tidydb, when checking for the continued existence of a message for
6145 which it has found a message-specific retry record, was not finding
6146 messages that were in split spool directories. Consequently, it was
6147 deleting retry records that should have stayed in existence.
6149 48. Steve fixed some bugs in eximstats.
6151 49. The SPA authentication driver was not abandoning authentication and moving
6152 on to the next authenticator when an expansion was forced to fail,
6153 contradicting the general specification for all authenticators. Instead it
6154 was generating a temporary error. It now behaves as specified.
6156 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6157 (the order specifies the preference for clients). The order is now AES256,
6158 AES128, 3DES, ARCFOUR128.
6160 51. Small patch to Sieve code - explicitly set From: when generating an
6163 52. Exim crashed if a remote delivery caused a very long error message to be
6164 recorded - for instance if somebody sent an entire SpamAssassin report back
6165 as a large number of 550 error lines. This bug was coincidentally fixed by
6166 increasing the size of one of Exim's internal buffers (big_buffer) that
6167 happened as part of the Exiscan merge. However, to be on the safe side, I
6168 have made the code more robust (and fixed the comments that describe what
6171 53. Now that there can be additional text after "Completed" in log lines (if
6172 the queue_time_overall log selector is set), a one-byte patch to exigrep
6173 was needed to allow it to recognize "Completed" as not the last thing in
6176 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6177 patch that reportedly fixes this has been added. I am not expert enough to
6178 create a test for it. This is what the patch creator wrote:
6180 "I found a little strange behaviour of ldap code when working with
6181 Windows 2003 AD Domain, where users was placed in more than one
6182 Organization Units. When I tried to give exim partial DN, the exit code
6183 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6184 But simultaneously result of request was absolutely normal ldap result,
6185 so I produce this patch..."
6187 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6188 REFERENCE, so I have modified the code to exclude the patch when that macro
6191 55. Some experimental protocols are using DNS PTR records for new purposes. The
6192 keys for these records are domain names, not reversed IP addresses. The
6193 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6194 leaves it alone. Component reversal etc. now happens only for IP addresses.
6197 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6199 57. Double the size of the debug message buffer (to 2048) so that more of very
6200 long debug lines gets shown.
6202 58. The exicyclog utility now does better if the number of log files to keep
6203 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6205 59. Two changes related to the smtp_active_hostname option:
6207 (1) $smtp_active_hostname is now available as a variable.
6208 (2) The default for smtp_banner uses $smtp_active_hostname instead
6209 of $primary_hostname.
6211 60. The host_aton() function is supposed to be passed a string that is known
6212 to be a valid IP address. However, in the case of IPv6 addresses, it was
6213 not checking this. This is a hostage to fortune. Exim now panics and dies
6214 if the condition is not met. A case was found where this could be provoked
6215 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6216 components; fortuitously, this particular loophole had already been fixed
6217 by change 4.50/55 above.
6219 If there are any other similar loopholes, the new check in host_aton()
6220 itself should stop them being exploited. The report I received stated that
6221 data on the command line could provoke the exploit when Exim was running as
6222 exim, but did not say which command line option was involved. All I could
6223 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6224 running as the user.
6227 61. There was a buffer overflow vulnerability in the SPA authentication code
6228 (which came originally from the Samba project). I have added a test to the
6229 spa_base64_to_bits() function which I hope fixes it.
6232 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6233 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6235 63. The daemon start-up calls getloadavg() while still root for those OS that
6236 need the first call to be done as root, but it missed one case: when
6237 deliver_queue_load_max is set with deliver_drop_privilege. This is
6238 necessary for the benefit of the queue runner, because there is no re-exec
6239 when deliver_drop_privilege is set.
6241 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6242 This has been fixed.
6244 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6245 was in use, was not putting the data itself into the right store pool;
6246 consequently, it could be overwritten for a subsequent message in the same
6247 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6250 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6252 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6253 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6254 the function were treating the return as a boolean value, which happened to
6255 work because 0=false and not-0=true, but is not correct code.
6257 68. The host_aton() function was not handling scoped IPv6 addresses (those
6258 with, for example, "%eth0" on the end) correctly.
6260 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6261 negated items (that is, ~something) in unsigned ints. Some compilers
6262 apparently mutter when there is no cast.
6264 70. If an address verification called from an ACL failed, and did not produce a
6265 user-specific message (i.e. there was only a "system" message), nothing was
6266 put in $acl_verify_message. In this situation, it now puts the system
6269 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6270 session; change 4.31/43 added the unwanted input to the log line - except
6271 that it did not do this in the start of session case. It now does.
6273 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6274 This is wrong and can cause the other end to generate a synchronization
6275 error if it is another Exim or anything else that does the synchronization
6276 check. A QUIT command is no longer sent after a timeout.
6278 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6279 during host lookups.
6281 74. The defer_ok option of callout verification was not working if it was used
6282 when verifying addresses in header lines, that is, for this case:
6284 verify = header_sender/callout=defer_ok
6286 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6287 those file descriptors could be used for SMTP connections. If anything
6288 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6289 could be sent to the SMTP client, causing chaos. The daemon now opens
6290 stdin, stdout, and stderr to /dev/null when it puts itself into the
6293 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6294 log by default. The user can override this with suitable Perl magic.
6296 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6297 the log message when discard triggers, was not working for the DATA ACL or
6298 for the non-SMTP ACL.
6300 78. Error message wording change in sieve.c.
6302 79. If smtp_accept_max_per_host was set, the number of connections could be
6303 restricted to fewer than expected, because the daemon was trying to set up
6304 a new connection before checking whether the processes handling previous
6305 connections had finished. The check for completed processes is now done
6306 earlier. On busy systems, this bug wouldn't be noticed because something
6307 else would have woken the daemon, and it would have reaped the completed
6310 80. If a message was submitted locally by a user whose login name contained one
6311 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6312 It caused a spool format error. I have fixed the spool reading code. A
6313 related problem was that the "from" clause in the Received: line became
6314 illegal because of the space(s). It is now covered by ${quote_local_part.
6316 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6319 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6321 83. Patch from Sieve maintainer to fix unterminated string problem in
6322 "vacation" handling.
6324 84. Some minor changes to the Linux configuration files to help with other
6325 OS variants using glibc.
6327 85. One more patch for Sieve to update vacation handling to latest spec.
6330 ----------------------------------------------------
6331 See the note above about the 4.44 and 4.50 releases.
6332 ----------------------------------------------------
6338 1. Change 4.43/35 introduced a bug that caused file counts to be
6339 incorrectly computed when quota_filecount was set in an appendfile
6342 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6343 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6346 3. Give more explanation in the error message when the command for a transport
6347 filter fails to execute.
6349 4. There are several places where Exim runs a non-Exim command in a
6350 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6351 was being done only for the command run by the queryprogram router. It is
6352 now done for all such subprocesses. The other cases are: ${run, transport
6353 filters, and the commands run by the lmtp and pipe transports.
6355 5. Some older OS have a limit of 256 on the maximum number of file
6356 descriptors. Exim was using setrlimit() to set 1000 as a large value
6357 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6358 systems. I've change it so that if it can't get 1000, it tries for 256.
6360 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6361 was an oversight, and furthermore, ever since the addition of extra
6362 controls (e.g. 4.43/32), the checks on when to allow different forms of
6363 "control" were broken. There should now be diagnostics for all cases when a
6364 control that does not make sense is encountered.
6366 7. $recipients is now available in the predata ACL (oversight).
6368 8. Tidy the search cache before the fork to do a delivery from a message
6369 received from the command line. Otherwise the child will trigger a lookup
6370 failure and thereby defer the delivery if it tries to use (for example) a
6371 cached ldap connection that the parent has called unbind on.
6373 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6374 of $address_data from the recipient verification was clobbered by the
6375 sender verification.
6377 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6378 was its contents. (It was OK if the option was not defined at all.)
6380 11. A "Completed" log line is now written for messages that are removed from
6381 the spool by the -Mrm option.
6383 12. $host_address is now set to the target address during the checking of
6384 ignore_target_hosts.
6386 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6387 being passed; this would have caused $sender_host_name to have been used if
6388 matching the list had actually called for a host name (not very likely,
6389 since this list is usually IP addresses). A host name is now passed as
6392 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6393 code that specifies a non-restarting handler (typically sigaction() in
6394 modern systems) in an attempt to fix a rare and obscure crash bug.
6396 15. Narrowed the window for a race in the daemon that could cause it to ignore
6397 SIGCHLD signals. This is not a major problem, because they are used only to
6398 wake it up if nothing else does.
6400 16. A malformed maildirsize file could cause Exim to calculate negative values
6401 for the mailbox size or file count. Odd effects could occur as a result.
6402 The maildirsize information is now recalculated if the size or filecount
6405 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6406 support for a long time. Removed HAVE_SYS_VFS_H.
6408 18. Updated exipick to current release from John Jetmore.
6410 19. Allow an empty sender to be matched against a lookup in an address list.
6411 Previously the only cases considered were a regular expression, or an
6414 20. Exim went into a mad DNS lookup loop when doing a callout where the
6415 host was specified on the transport, if the DNS lookup yielded more than
6418 21. The RFC2047 encoding function was originally intended for short strings
6419 such as real names; it was not keeping to the 75-character limit for
6420 encoded words that the RFC imposes. It now respects the limit, and
6421 generates multiple encoded words if necessary. To be on the safe side, I
6422 have increased the buffer size for the ${rfc2047: expansion operator from
6425 22. Failure to deliver a bounce message always caused it to be frozen, even if
6426 there was an errors_to setting on the router. The errors_to setting is now
6429 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6430 canonical form (fully expanded) before being placed in
6431 $sender_host_address.
6433 24. Updated eximstats to version 1.33
6435 25. Include certificate and key file names in error message when GnuTLS fails
6436 to set them up, because the GnuTLS error message doesn't include the name
6437 of the failing file when there is a problem reading it.
6439 26. Expand error message when OpenSSL has problems setting up cert/key files.
6442 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6443 (this can affect the format of dates).
6445 28. exim_tidydb, when checking for the continued existence of a message for
6446 which it has found a message-specific retry record, was not finding
6447 messages that were in split spool directories. Consequently, it was
6448 deleting retry records that should have stayed in existence.
6450 29. eximstats updated to version 1.35
6451 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6452 1.35 - bugfix such that pie charts by volume are generated correctly
6454 30. The SPA authentication driver was not abandoning authentication and moving
6455 on to the next authenticator when an expansion was forced to fail,
6456 contradicting the general specification for all authenticators. Instead it
6457 was generating a temporary error. It now behaves as specified.
6459 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6460 (the order specifies the preference for clients). The order is now AES256,
6461 AES128, 3DES, ARCFOUR128.
6463 31. Small patch to Sieve code - explicitly set From: when generating an
6466 32. Exim crashed if a remote delivery caused a very long error message to be
6467 recorded - for instance if somebody sent an entire SpamAssassin report back
6468 as a large number of 550 error lines. This bug was coincidentally fixed by
6469 increasing the size of one of Exim's internal buffers (big_buffer) that
6470 happened as part of the Exiscan merge. However, to be on the safe side, I
6471 have made the code more robust (and fixed the comments that describe what
6474 33. Some experimental protocols are using DNS PTR records for new purposes. The
6475 keys for these records are domain names, not reversed IP addresses. The
6476 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6477 leaves it alone. Component reversal etc. now happens only for IP addresses.
6480 34. The host_aton() function is supposed to be passed a string that is known
6481 to be a valid IP address. However, in the case of IPv6 addresses, it was
6482 not checking this. This is a hostage to fortune. Exim now panics and dies
6483 if the condition is not met. A case was found where this could be provoked
6484 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6485 components; fortuitously, this particular loophole had already been fixed
6486 by change 4.50/55 or 4.44/33 above.
6488 If there are any other similar loopholes, the new check in host_aton()
6489 itself should stop them being exploited. The report I received stated that
6490 data on the command line could provoke the exploit when Exim was running as
6491 exim, but did not say which command line option was involved. All I could
6492 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6493 running as the user.
6496 35. There was a buffer overflow vulnerability in the SPA authentication code
6497 (which came originally from the Samba project). I have added a test to the
6498 spa_base64_to_bits() function which I hope fixes it.
6501 36. The daemon start-up calls getloadavg() while still root for those OS that
6502 need the first call to be done as root, but it missed one case: when
6503 deliver_queue_load_max is set with deliver_drop_privilege. This is
6504 necessary for the benefit of the queue runner, because there is no re-exec
6505 when deliver_drop_privilege is set.
6507 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6508 was in use, was not putting the data itself into the right store pool;
6509 consequently, it could be overwritten for a subsequent message in the same
6510 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6513 38. Sometimes the final signoff response after QUIT could fail to get
6514 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6515 < 0 before doing a fflush(). This bug looks as though it goes back to the
6516 introduction of TLS in release 3.20, but "sometimes" must have been rare
6517 because the tests only now provoked it.
6523 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6524 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6525 Now it returns an integer. A number of calls were still expecting a T/F
6526 return. Fortuitously, in all cases, the tests worked in OK situations,
6527 which is the norm. However, things would have gone wrong on any write
6528 failures on the smtp file descriptor. This function is used when sending
6529 messages over SMTP and also when doing verify callouts.
6531 2. When Exim is called to do synchronous delivery of a locally submitted
6532 message (the -odf or -odi options), it no longer closes stderr before doing
6535 3. Implemented the mua_wrapper option.
6537 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6539 5. Implemented the functions header_remove(), header_testname(),
6540 header_add_at_position(), and receive_remove_recipient(), and exported them
6543 6. If an ACL "warn" statement specified the addition of headers, Exim already
6544 inserted X-ACL-Warn: at the start if there was no header name. However, it
6545 was not making this test for the second and subsequent header lines if
6546 there were newlines in the string. This meant that an invalid header could
6547 be inserted if Exim was badly configured.
6549 7. Allow an ACL "warn" statement to add header lines at the start or after all
6550 the Received: headers, as well as at the end.
6552 8. Added the rcpt_4xx retry error code.
6554 9. Added postmaster_mailfrom=xxx to callout verification option.
6556 10. Added mailfrom=xxxx to the callout verification option, for verify=
6559 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6560 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6562 12. Inserted some casts to stop certain compilers warning when using pointer
6563 differences as field lengths or precisions in printf-type calls (mostly
6564 affecting debugging statements).
6566 13. Added optional readline() support for -be (dynamically loaded).
6568 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6569 same clock tick as a message's arrival, so that its received time was the
6570 same as the "first fail" time on the retry record, and that message
6571 remained on the queue past the ultimate address timeout, every queue runner
6572 would try a delivery (because it was past the ultimate address timeout) but
6573 after another failure, the ultimate address timeout, which should have then
6574 bounced the address, did not kick in. This was a "< instead of <=" error;
6575 in most cases the first failure would have been in the next clock tick
6576 after the received time, and all would be well.
6578 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6579 being recognized when the domain list was tested by the match_domain
6580 condition in an expansion string.
6582 16. Added the ${str2b64: operator.
6584 17. Exim was always calling setrlimit() to set a large limit for the number of
6585 processes, without checking whether the existing limit was already
6586 adequate. (It did check for the limit on file descriptors.) Furthermore,
6587 errors from getrlimit() and setrlimit() were being ignored. Now they are
6588 logged to the main and panic logs, but Exim does carry on, to try to do its
6589 job under whatever limits there are.
6591 18. Imported PCRE 5.0.
6593 19. Trivial typo in log message " temporarily refused connection" (the leading
6596 20. If the log selector return_path_on_delivery was set and an address was
6597 redirected to /dev/null, the delivery process crashed because it assumed
6598 that a return path would always be set for a "successful" delivery. In this
6599 case, the whole delivery is bypassed as an optimization, and therefore no
6602 21. Internal re-arrangement: the function for sending a challenge and reading
6603 a response while authentication was assuming a zero-terminated challenge
6604 string. It's now changed to take a pointer and a length, to allow for
6605 binary data in such strings.
6607 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6609 23. Exim was not respecting finduser_retries when seeking the login of the
6610 uid under which it was called; it was always trying 10 times. (The default
6611 setting of finduser_retries is zero.) Also, it was sleeping after the final
6612 failure, which is pointless.
6614 24. Implemented tls_on_connect_ports.
6616 25. Implemented acl_smtp_predata.
6618 26. If the domain in control=submission is set empty, Exim assumes that the
6619 authenticated id is a complete email address when it generates From: or
6620 Sender: header lines.
6622 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6623 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6624 chown and chgrp in /bin and hostname in /usr/bin.
6626 28. Exim was keeping the "process log" file open after each use, just as it
6627 does for the main log. This opens the possibility of it remaining open for
6628 long periods when the USR1 signal hits a daemon. Occasional processlog
6629 errors were reported, that could have been caused by this. Anyway, it seems
6630 much more sensible not to leave this file open at all, so that is what now
6633 29. The long-running daemon process does not normally write to the log once it
6634 has entered its main loop, and it closes the log before doing so. This is
6635 so that log files can straightforwardly be renamed and moved. However,
6636 there are a couple of unusual error situations where the daemon does write
6637 log entries, and I had neglected to close the log afterwards.
6639 30. The text of an SMTP error response that was received during a remote
6640 delivery was being truncated at 512 bytes. This is too short for some of
6641 the long messages that one sometimes sees. I've increased the limit to
6644 31. It is now possible to make retry rules that apply only when a message has a
6645 specific sender, in particular, an empty sender.
6647 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6648 it possible to be selective about when SMTP synchronization is enforced.
6650 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6652 32. Implemented hosts_connection_nolog.
6654 33. Added an ACL for QUIT.
6656 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6659 35. Added mailbox_size and mailbox_filecount to appendfile.
6661 36. Added control = no_multiline_responses to ACLs.
6663 37. There was a bug in the logic of the code that waits for the clock to tick
6664 in the case where the clock went backwards by a substantial amount such
6665 that the microsecond fraction of "now" was more than the microsecond
6666 fraction of "then" (but the whole seconds number was less).
6668 38. Added support for the libradius Radius client library this is found on
6669 FreeBSD (previously only the radiusclient library was supported).
6675 1. When certain lookups returned multiple values in the form name=value, the
6676 quoting of the values was not always being done properly. Specifically:
6677 (a) If the value started with a double quote, but contained no whitespace,
6679 (b) If the value contained whitespace other than a space character (i.e.
6680 tabs or newlines or carriage returns) it was not quoted.
6681 This fix has been applied to the mysql and pgsql lookups by writing a
6682 separate quoting function and calling it from the lookup code. The fix
6683 should probably also be applied to nisplus, ibase and oracle lookups, but
6684 since I cannot test any of those, I have not disturbed their existing code.
6686 2. A hit in the callout cache for a specific address caused a log line with no
6687 reason for rejecting RCPT. Now it says "Previous (cached) callout
6688 verification failure".
6690 3. There was an off-by-one bug in the queryprogram router. An over-long
6691 return line was truncated at 256 instead of 255 characters, thereby
6692 overflowing its buffer with the terminating zero. As well as fixing this, I
6693 have increased the buffer size to 1024 (and made a note to document this).
6695 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6696 when Exim is waiting for an SMTP response from a remote server, Exim
6697 restarts its select() call on the socket, thereby resetting its timeout.
6698 This is not a problem when such interrupts are rare. Somebody set up a cron
6699 job to run exiwhat every 2 minutes, which is less than the normal select()
6700 timeout (5 or 10 minutes). This meant that the select() timeout never
6701 kicked in because it was always reset. I have fixed this by comparing the
6702 time when an interrupt arrives with the time at the start of the first call
6703 to select(). If more time than the timeout has elapsed, the interrupt is
6704 treated as a timeout.
6706 5. Some internal re-factoring in preparation for the addition of Sieve
6707 extensions (by MH). In particular, the "personal" test is moved to a
6708 separate function, and given an option for scanning Cc: and Bcc: (which is
6709 not set for Exim filters).
6711 6. When Exim created an email address using the login of the caller as the
6712 local part (e.g. when creating a From: or Sender: header line), it was not
6713 quoting the local part when it contained special characters such as @.
6715 7. Installed new OpenBSD configuration files.
6717 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6718 try to make them clearer.
6720 9. Callout options, other than the timeout value, were being ignored when
6721 verifying sender addresses in header lines. For example, when using
6723 verify = header_sender/callout=no_cache
6725 the cache was (incorrectly) being used.
6727 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6728 only the Cygwin environment.
6730 11. When return_path_on_delivery was set as a log selector, if different remote
6731 addresses in the same message used different return paths and parallel
6732 remote delivery occurred, the wrong values would sometimes be logged.
6733 (Whenever a remote delivery process finished, the return path value from
6734 the most recently started remote delivery process was logged.)
6736 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6737 lines when AUTH and/or TLS are in use. This is the "received protocol"
6738 field. Exim used to use "asmtp" for authenticated SMTP, without any
6739 indication (in the protocol name) for TLS use. Now it follows the RFC and
6740 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6741 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6742 names appear in log lines as well as in Received: header lines.
6744 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6745 extensions, and comparison tests, and to fix some bugs.
6747 14. Changes to the "personal" filter test:
6749 (1) The test was buggy in that it was just doing the equivalent of
6750 "contains" tests on header lines. For example, if a user's address was
6751 anne@some.where, the "personal" test would incorrectly be true for
6753 To: susanne@some.where
6755 This test is now done by extracting each address from the header in turn,
6756 and checking the entire address. Other tests that are part of "personal"
6757 are now done using regular expressions (for example, to check local parts
6758 of addresses in From: header lines).
6760 (2) The list of non-personal local parts in From: addresses has been
6761 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6762 taken from the Sieve specification recommendations.
6764 (3) If the message contains any header line starting with "List-" it is
6765 treated as non-personal.
6767 (4) The test for "circular" in the Subject: header line has been removed
6768 because it now seems ill-conceived.
6770 15. Minor typos in src/EDITME comments corrected.
6772 16. Installed latest exipick from John Jetmore.
6774 17. If headers_add on a router specified a text string that was too long for
6775 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6776 of string_sprintf() is now avoided.
6778 18. $message_body_size was not set (it was always zero) when running the DATA
6779 ACL and the local_scan() function.
6781 19. For the "mail" command in an Exim filter, no default was being set for
6782 the once_repeat time, causing a random time value to be used if "once" was
6783 specified. (If the value happened to be <= 0, no repeat happened.) The
6784 default is now 0s, meaning "never repeat". The "vacation" command was OK
6785 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6786 (I found it when inspecting the code).
6788 20. There is now an overall timeout for performing a callout verification. It
6789 defaults to 4 times the callout timeout, which applies to individual SMTP
6790 commands during the callout. The overall timeout applies when there is more
6791 than one host that can be tried. The timeout is checked before trying the
6792 next host. This prevents very long delays if there are a large number of
6793 hosts and all are timing out (e.g. when the network connections are timing
6794 out). The value of the overall timeout can be changed by specifying an
6795 additional sub-option for "callout", called "maxwait". For example:
6797 verify = sender/callout=5s,maxwait=20s
6799 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6800 to the end before writing, but this should make it even safer).
6802 22. Exim was forgetting that it had advertised PIPELINING for the second and
6803 subsequent messages on an SMTP connection. It was also not resetting its
6804 memory on STARTTLS and an internal HELO.
6806 23. When Exim logs an SMTP synchronization error within a session, it now
6807 records whether PIPELINING has been advertised or not.
6809 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6810 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6811 rather than long int.
6813 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6815 26. Added the never_mail option to autoreply.
6821 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6822 crash if the getsockname() call failed; this can happen if a connection is
6823 closed very soon after it is established. The problem was simply in the
6824 order in which certain operations were done, causing Exim to try to write
6825 to the SMTP stream before it had set up the file descriptor. The bug has
6826 been fixed by making things happen in the correct order.
6832 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6833 before the connection was closed, thus losing the rejection response.
6835 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6836 some early Solaris releases, but causes trouble in current releases where
6837 socklen_t is defined.
6839 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6842 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6845 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6846 that releasing the top of it at the end releases what was used for sub-
6847 expansions (unless the block got too big). However, discard this block if
6848 the first thing is a variable or header, so that we can use its block when
6849 it is dynamic (useful for very large $message_headers, for example).
6851 6. Lookups now cache *every* query, not just the most recent. A new, separate
6852 store pool is used for this. It can be recovered when all lookup caches are
6853 flushed. Lookups now release memory at the end of their result strings.
6854 This has involved some general refactoring of the lookup sources.
6856 7. Some code has been added to the store_xxx() functions to reduce the amount
6857 of flapping under certain conditions.
6859 8. log_incoming_interface used to affect only the <= reception log lines. Now
6860 it causes the local interface and port to be added to several more SMTP log
6861 lines, for example "SMTP connection from", and rejection lines.
6863 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6865 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6867 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6868 could be overwritten at the end of the current message (or the start of a
6869 new message if it was set in a HELO ACL). The value is now preserved for
6870 the duration of the SMTP connection.
6872 12. If a transport had a headers_rewrite setting, and a matching header line
6873 contained an unqualified address, that address was qualified, even if it
6874 did not match any rewriting rules. The underlying bug was that the values
6875 of the flags that permit the existence of unqualified sender and recipient
6876 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6877 non-local messages, and by -bnq for local messages) were not being
6878 preserved with the message after it was received.
6880 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6881 "next input=" as part of the text comprising the host identity instead of
6882 the correct text. The code was using the same buffer for two different
6883 strings. However, depending on which order the printing function evaluated
6884 its arguments, the bug did not always show up. Under Linux, for example, my
6885 test suite worked just fine.
6887 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6888 doesn't work with some older versions of Perl. It has been changed to "my",
6889 which in any case is probably the better facility to use.
6891 15. A really picky compiler found some instances of statements for creating
6892 error messages that either had too many or two few arguments for the format
6895 16. The size of the buffer for calls to the DNS resolver has been increased
6896 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6897 for addresses that have a lot of PTR records. This alleviates a problem; it
6898 does not fully solve it.
6900 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6901 buffer now truncates the list and logs the incident, which is the same
6902 action as happens when Exim is looking up a host name and its aliases.
6903 Previously in this situation something unpredictable would happen;
6904 sometimes it was "internal error: store_reset failed".
6906 18. If a server dropped the connection unexpectedly when an Exim client was
6907 using GnuTLS and trying to read a response, the client delivery process
6908 crashed while trying to generate an error log message.
6910 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6911 string, for example:
6913 warn message = H1: something\nH2: something
6915 the text was added as a single header line from Exim's point of view
6916 though it ended up OK in the delivered message. However, searching for the
6917 second and subsequent header lines using $h_h2: did not work. This has been
6918 fixed. Similarly, if a system filter added multiple headers in this way,
6919 the routers could not see them.
6921 20. Expanded the error message when iplsearch is called with an invalid key to
6922 suggest using net-iplsearch in a host list.
6924 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6925 ACLs are no longer actually imposed (and a message to that effect is
6928 22. If a "gecos" field in a passwd entry contained escaped characters, in
6929 particular, if it contained a \" sequence, Exim got it wrong when building
6930 a From: or a Sender: header from that name. A second bug also caused
6931 incorrect handling when an unquoted " was present following a character
6932 that needed quoting.
6934 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6935 was not being matched caselessly.
6937 24. Arranged for all hyphens in the exim.8 source to be escaped with
6940 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6941 properly. Recipient callout cache records were still being keyed to include
6942 the sender, even when use_sender was set false. This led to far more
6943 callouts that were necessary. The sender is no longer included in the key
6944 when use_sender is false.
6946 26. Added "control = submission" modifier to ACLs.
6948 27. Added the ${base62d: operator to decode base 62 numbers.
6950 28. dnsdb lookups can now access SRV records.
6952 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6953 the configuration file.
6955 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6956 (-v) mode. This makes the output for a verbose queue run more intelligible.
6958 31. Added a use_postmaster feature to recipient callouts.
6960 32. Added the $body_zerocount variable, containing the number of binary zero
6961 bytes in the message body.
6963 33. The time of last modification of the "new" subdirectory is now used as the
6964 "mailbox time last read" when there is a quota error for a maildir
6967 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6969 35. Added +ignore_unknown as a special item in host lists.
6971 36. Code for decoding IPv6 addresses in host lists is now included, even if
6972 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6973 address was recognized as an IP address, but was then not correctly decoded
6974 into binary, causing unexpected and incorrect effects when compared with
6981 1. Very minor rewording of debugging text in manualroute to say "list of
6982 hosts" instead of "hostlist".
6984 2. If verify=header_syntax was set, and a header line with an unqualified
6985 address (no domain) and a large number of spaces between the end of the
6986 name and the colon was received, the reception process suffered a buffer
6987 overflow, and (when I tested it) crashed. This was caused by some obsolete
6988 code that should have been removed. The fix is to remove it!
6990 3. When running in the test harness, delay a bit after writing a bounce
6991 message to get a bit more predictability in the log output.
6993 4. Added a call to search_tidyup() just before forking a reception process. In
6994 theory, someone could use a lookup in the expansion of smtp_accept_max_
6995 per_host which, without the tidyup, could leave open a database connection.
6997 5. Added the variables $recipient_data and $sender_data which get set from a
6998 lookup success in an ACL "recipients" or "senders" condition, or a router
6999 "senders" option, similar to $domain_data and $local_part_data.
7001 6. Moved the writing of debug_print from before to after the "senders" test
7004 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7005 problems for message scanning, either using a data ACL, or using
7006 local_scan() because the Received: header was not generated till after they
7007 were called (in order to set the time as the time of reception completion).
7008 I have revised the way this works. The header is now generated after the
7009 body is received, but before the ACL or local_scan() are called. After they
7010 are run, the timestamp in the header is updated.
7016 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7017 before starting a queue runner without re-exec. This happened only when
7018 deliver_drop_privilege was set or when the Exim user was set to root. The
7019 effect of the bug was that timeouts during subsequent deliveries caused
7020 crashes instead of being properly handled. The handler is now left at its
7021 default (and expected) setting.
7023 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7024 message, again when deliver_drop_privilege is set or Exim is run as root.
7025 The bug described in (1) was not present in this case, but the tidying up
7026 of the other signals was missing. I have made the two cases consistent.
7028 3. The ignore_target_hosts setting on a manualroute router was being ignored
7029 for hosts that were looked up using the /MX notation.
7031 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7034 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7035 operated on the sender address. After changing the $sender_address to <>
7036 for the sender address verify, Exim was re-instated it as the original
7037 (before rewriting) address, but remembering that it had rewritten it, so it
7038 wasn't rewriting it again. This bug also had the effect of breaking the
7039 sender address verification caching when the sender address was rewritten.
7041 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7042 This has been changed so that if the ip literal address matches
7043 ignore_target_hosts, the router declines.
7045 7. Added expansion conditions match_domain, match_address, and match_local_
7046 part (NOT match_host).
7048 8. The placeholder for the Received: header didn't have a length field set.
7050 9. Added code to Exim itself and to exim_lock to test for a specific race
7051 condition that could lead to file corruption when using MBX delivery. The
7052 issue is with the lockfile that is created in /tmp. If this file is removed
7053 after a process has opened it but before that process has acquired a lock,
7054 there is the potential for a second process to recreate the file and also
7055 acquire a lock. This could lead to two Exim processes writing to the file
7056 at the same time. The added code performs the same test as UW imapd; it
7057 checks after acquiring the lock that its file descriptor still refers to
7058 the same named file.
7060 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7061 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7064 11. Added the smtp_active_hostname option. If used, this will typically be made
7065 to depend on the incoming interface address. Because $interface_address is
7066 not set up until the daemon has forked a reception process, error responses
7067 that can happen earlier (such as "too many connections") no longer contain
7070 12. If an expansion in a condition on a "warn" statement fails because a lookup
7071 defers, the "warn" statement is abandoned, and the next ACL statement is
7072 processed. Previously this caused the whole ACL to be aborted.
7074 13. Added the iplsearch lookup type.
7076 14. Added ident_timeout as a log selector.
7078 15. Added tls_certificate_verified as a log selector.
7080 16. Added a global option tls_require_ciphers (compare the smtp transport
7081 option of the same name). This controls incoming TLS connections.
7083 17. I finally figured out how to make tls_require_ciphers do a similar thing
7084 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7085 before starting the TLS session.
7087 18. Tabs are now shown as \t in -bP output.
7089 19. If the log selector return_path_on_delivery was set, Exim crashed when
7090 bouncing a message because it had too many Received: header lines.
7092 20. If two routers both had headers_remove settings, and the first one included
7093 a superfluous trailing colon, the final name in the first list and the
7094 first name in the second list were incorrectly joined into one item (with a
7095 colon in the middle).
7101 1. Added -C and -D options to the exinext utility, mainly to make it easier
7102 to include in the automated testing, but these could be helpful when
7103 multiple configurations are in use.
7105 2. The exinext utility was not formatting the output nicely when there was
7106 an alternate port involved in the retry record key, nor when there was a
7107 message id as well (for retries that were specific to a specific message
7108 and a specific host). It was also confused by IPv6 addresses, because of
7109 the additional colons they contain. I have fixed the IPv4 problem, and
7110 patched it up to do a reasonable job for IPv6.
7112 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7113 delivery, the log line now contains "pipelined" if PIPELINING was used.
7115 4. An SMTP transport process used to panic and die if the bind() call to set
7116 an explicit outgoing interface failed. This has been changed; it is now
7117 treated in the same way as a connect() failure.
7119 5. A reference to $sender_host_name in the part of a conditional expansion
7120 that was being skipped was still causing a DNS lookup. This no longer
7123 6. The def: expansion condition was not recognizing references to header lines
7124 that used bh_ and bheader_.
7126 7. Added the _cache feature to named lists.
7128 8. The code for checking quota_filecount in the appendfile transport was
7129 allowing one more file than it should have been.
7131 9. For compatibility with Sendmail, the command line option
7139 and sets the incoming protocol and host name (for trusted callers). The
7140 host name and its colon can be omitted when only the protocol is to be set.
7141 Note the Exim already has two private options, -pd and -ps, that refer to
7142 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7143 "s", but I don't think that's a major issue.
7145 10. A number of refactoring changes to the code, none of which should affect
7148 (a) The number of logging options was getting close to filling up the
7149 32-bit word that was used as a bit map. I have split them into two classes:
7150 those that are passed in the argument to log_write(), and those that are
7151 only ever tested independently outside of that function. These are now in
7152 separate 32-bit words, so there is plenty of room for expansion again.
7153 There is no change in the user interface or the logging behaviour.
7155 (b) When building, for example, log lines, the code previously used a
7156 macro that called string_cat() twice, in order to add two strings. This is
7157 not really sufficiently general. Furthermore, there was one instance where
7158 it was actually wrong because one of the argument was used twice, and in
7159 one call a function was used. (As it happened, calling the function twice
7160 did not affect the overall behaviour.) The macro has been replaced by a
7161 function that can join an arbitrary number of extra strings onto a growing
7164 (c) The code for expansion conditions now uses a table and a binary chop
7165 instead of a serial search (which was left over from when there were very
7166 few conditions). Also, it now recognizes conditions like "pam" even when
7167 the relevant support is not compiled in: a suitably worded error message is
7168 given if an attempt is made to use such a condition.
7170 11. Added ${time_interval:xxxxx}.
7172 12. A bug was causing one of the ddress fields not to be passed back correctly
7173 from remote delivery subprocesses. The field in question was not being
7174 subsequently used, so this caused to problems in practice.
7176 13. Added new log selectors queue_time and deliver_time.
7178 14. Might have fixed a bug in maildirsizefile handling that threw up
7179 "unexpected character" debug warnings, and recalculated the data
7180 unnecessarily. In any case, I expanded the warning message to give more
7183 15. Added the message "Restricted characters in address" to the statements in
7184 the default ACL that block characters like @ and % in local parts.
7186 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7187 Three changes have been made:
7189 (a) There was a serious bug; a negative response to MAIL caused the whole
7190 recipient domain to be cached as invalid, thereby blocking all messages
7191 to all local parts at the same domain, from all senders. This bug has
7192 been fixed. The domain is no longer cached after a negative response to
7193 MAIL if the sender used is not empty.
7195 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7198 (c) A new callout option, "use_sender" has been added for people who want
7199 the modified behaviour.
7205 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7208 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7209 indeed breaks things for older releases.
7211 3. Added additional logging to the case where there is a problem reading data
7212 from a filter that is running in a subprocess using a pipe, in order to
7213 try to track down a specific problem.
7215 4. Testing facility fudge: when running in the test harness and attempting
7216 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7217 getting "No route to host". Convert this to a timeout.
7219 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7222 6. Some OS don't have socklen_t but use size_t instead. This affects the
7223 fifth argument of getsockopt() amongst other things. This is now
7224 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7225 can be set for individual OS. I have set it for SunOS5, OSF1, and
7226 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7227 some earlier ones do not.
7229 7. Change 4.30/15 was not doing the test caselessly.
7231 8. The standard form for an IPv6 address literal was being rejected by address
7232 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7233 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7234 this, as well as the form without the "IPv6" on the front (but only when
7235 address literals are enabled, of course).
7237 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7239 10. Exim crashed if a message with an empty sender address specified by -f
7240 encountered a router with an errors_to setting. This could be provoked only
7241 by a command such as
7245 where an empty string was supplied; "<>" did not hit this bug.
7247 11. Installed PCRE release 4.5.
7249 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7250 remained set. It is now erased.
7252 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7253 times from message ids (which are base 36 rather than the normal 62).
7255 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7256 were being counted as actual protocol errors, and logged if the log
7257 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7258 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7259 and DATA following a set of rejected RCPTs do not count as protocol errors.
7260 In other words, Exim assumes they were pipelined, though this may not
7261 actually be the case. Of course, in all cases the client gets an
7262 appropriate error code.
7264 15. If a lookup fails in an ACL condition, a message about the failure may
7265 be available; it is used if testing the ACL cannot continue, because most
7266 such messages specify what the cause of the deferral is. However, some
7267 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7268 that caused an old message to be retained and used if a later statement
7269 caused a defer, replacing the real cause of the deferral.
7271 16. If an IP address had so many PTR records that the DNS lookup buffer
7272 was not large enough to hold them, Exim could crash while trying to process
7273 the truncated data. It now detects and logs this case.
7275 17. Further to 4.21/58, another change has been made: if (and only if) the
7276 first line of a message (the first header line) ends with CRLF, a bare LF
7277 in a subsequent header line has a space inserted after it, so as not to
7278 terminate the header.
7280 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7281 unnecessarily, used atoi() instead of strtol(), and didn't check the
7282 termination when getting file sizes from file names by regex.
7284 19. Completely re-implemented the support for maildirsize files, in the light
7285 of a number of problems with the previous contributed implementation
7286 (4.30/29). In particular:
7288 . If the quota is zero, the maildirsize file is maintained, but no quota is
7291 . If the maildir directory does not exist, it is created before any attempt
7292 to write a maildirsize file.
7294 . The quota value in the file is just a cache; if the quota is changed in
7295 the transport, the new value overrides.
7297 . A regular expression is available for excluding directories from the
7300 20. The autoreply transport checks the characters in options that define the
7301 message's headers; it allows continued headers, but it was checking with
7302 isspace() after an embedded newline instead of explicitly looking for a
7305 21. If all the "regular" hosts to which an address was routed had passed their
7306 expiry times, and had not reached their retry times, the address was
7307 bounced, even if fallback hosts were defined. Now Exim should go on to try
7310 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7311 equivalent code in the SMTP transport. Some hosts send humungous responses
7312 to HELO/EHLO, more than 1024 it seems.
7314 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7315 gives compiler warnings in some environments. I've now done it "properly",
7318 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7319 (because of problems with the built-in one) was declared to return uschar *
7320 instead of char *, causing compiler failure.
7322 25. Fixed a file descriptor leak when processing alias/forward files.
7324 26. Fixed a minor format string issue in dbfn.c.
7326 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7328 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7329 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7332 29. When a sender address is verified, it is cached, to save repeating the test
7333 when there is more than one recipient in a message. However, when the
7334 verification involves a callout, it is possible for different callout
7335 options to be set for different recipients. It is too complicated to keep
7336 track of this in the cache, so now Exim always runs a verification when a
7337 callout is required, relying on the callout cache for the optimization.
7338 The overhead is duplication of the address routing, but this should not be
7341 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7342 to be verified with callout=postmaster, and the main callout worked but the
7343 postmaster check failed, the verification correctly failed. However, if a
7344 subsequent RCPT command asked for sender verification *without* the
7345 postmaster check, incorrect caching caused this verification also to fail,
7348 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7349 it was not caching the DNS options (qualify_single, search_parents) that
7350 were used when the lookup failed. A subsequent lookup with different
7351 options therefore always gave the same answer, though there were cases
7352 where it should not have. (Example: a "domains = !$mx_any" option on a
7353 dnslookup router: the "domains" option is always processed without any
7354 widening, but the router might have qualify_single set.) Now Exim uses the
7355 cached value only when the same options are set.
7357 32. Added John Jetmore's "exipick" utility to the distribution.
7359 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7360 than a timeout (e.g. a certificate is required, and is not provided), an
7361 Exim server now closes the connection immediately. Previously it waited for
7362 the client to close - but if the client is SSL, it seems that they each
7363 wait for each other, leading to a delay before one of them times out.
7365 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7366 maintained 0.8.x compatibility because I don't think many are using it, and
7367 it is clearly obsolete.
7369 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7372 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7373 client certificate was expired. A simple patch fixes this, though I don't
7374 understand the full logic of why the verify callback is called multiple
7377 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7378 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7379 which causes problems with some clients (such as the Certicom SSL Plus
7380 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7381 disables the coutermeasure allowing Eudora to connect."
7383 38. Exim was not checking that a write() to a log file succeeded. This could
7384 lead to Bad Things if a log got too big, in particular if it hit a file
7385 size limit. Exim now panics and dies if it cannot write to a log file, just
7386 as it does if it cannot open a log file.
7388 39. Modified OS/Makefile-Linux so that it now contains
7390 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7392 The two -D definitions ensure that Exim is compiled with large file
7393 support, which makes it possible to handle log files that are bigger than
7396 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7397 instance) a domain was checked against a named list that involved a lookup,
7398 causing $domain_data to be set, then another domain was checked against the
7399 same list, then the first domain was re-checked, the value of $domain_data
7400 after the final check could be wrong. In particular, if the second check
7401 failed, it could be set empty. This bug probably also applied to
7404 41. The strip_trailing_dot option was not being applied to the address given
7405 with the -f command-line option.
7407 42. The code for reading a message's header from the spool was incrementing
7408 $received_count, but never initializing it. This meant that the value was
7409 incorrect (doubled) while delivering a message in the same process in which
7410 it was received. In the most common configuration of Exim, this never
7411 happens - a fresh exec is done - but it can happen when
7412 deliver_drop_privilege is set.
7414 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7415 it now includes up to 150 characters of the unexpected data in the log
7418 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7419 and building data strings. The size of both of these buffers was 10 000
7420 bytes - far larger than anybody would *ever* want, thought I. Needless to
7421 say, somebody hit the limit. I have increased the maximum line length to
7422 20 000 and the maximum data length of concatenated lines to 100 000. I have
7423 also fixed two bugs, because there was no checking on these buffers. Tsk,
7424 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7425 buffer is too small.
7427 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7428 lsearch lookups. Now it does.
7430 46. When parsing a route_list item in a manualroute router, a fixed-length
7431 buffer was used for the list of hosts. I made this 1024 bytes long,
7432 thinking that nobody would ever have a list of hosts that long. Wrong.
7433 Somebody had a whole pile of complicated expansion conditions, and the
7434 string was silently truncated, leading to an expansion error. It turns out
7435 that it is easier to change to an unlimited length (owing to other changes
7436 that have happened since this code was originally written) than to build
7437 structure for giving a limitation error. The length of the item that
7438 expands into the list of hosts is now unlimited.
7440 47. The lsearch lookup could not handle data where the length of text line was
7441 more than 4095 characters. Such lines were truncated, leading to shortened
7442 data being returned. It should now handle lines of any length.
7444 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7445 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7448 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7449 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7450 no longer happens in comment lines. A list of replacements is now placed
7451 at the head of all of the source files, except those whose only change is
7452 to replace PERL_COMMAND in the very first #! line.
7454 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7455 messages on the queue, with a bottom-up merge sort, using code contributed
7456 by Michael Haardt. This should make operations like -bp somewhat faster on
7457 large queues. It won't affect queue runners, except when queue_run_in_order
7460 51. Installed eximstats 1.31 in the distribution.
7462 52. Added support for SRV lookups to the dnslookup router.
7464 53. If an ACL referred to $message_body or $message_body_end, the value was not
7465 reset for any messages that followed in the same SMTP session.
7467 54. The store-handling optimization for building very long strings was not
7468 differentiating between the different store pools. I don't think this
7469 actually made any difference in practice, but I've tidied it.
7471 55. While running the routers to verify a sender address, $sender_address
7472 was still set to the sender address. This is wrong, because when routing to
7473 send a bounce to the sender, it would be empty. Therefore, I have changed
7474 it so that, while verifying a sender address, $sender_address is set to <>.
7475 (There is no change to what happens when verifying a recipient address.)
7477 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7478 target A or AAAA records (if not already returned) without resetting the
7479 qualify_single or search_parents options of the DNS resolver. These are
7480 inappropriate in this case because the targets of MX and SRV records must
7481 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7482 target that, when qualified, matched something in the local domain. These
7483 two options are now turned off when doing these lookups.
7485 57. It seems that at least some releases of Reiserfs (which does not have the
7486 concept of a fixed number of inodes) returns zero and not -1 for the
7487 number of available inodes. This interacted badly with check_spool_inodes,
7488 which assumed that -1 was the "no such thing" setting. What I have done is
7489 to check that the total number of inodes is greater than zero before doing
7490 the test of how many are available.
7492 58. When a "warn" ACL statement has a log_message modifier, the message is
7493 remembered, and not repeated. This is to avoid a lot of repetition when a
7494 message has many recipients that cause the same warning to be written.
7495 However, Exim was preserving the list of already written lines for an
7496 entire SMTP session, which doesn't seem right. The memory is now reset if a
7497 new message is started.
7499 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7500 debugging output unless log_rewrite was also set.
7502 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7503 of (int)(handle) when we know that handle contains (void *)(-1).
7505 61. The Exim daemon panic-logs an error return when it closes the incoming
7506 connection. However "connection reset by peer" seems to be common, and
7507 isn't really an error worthy of noting specially, so that particular error
7510 62. When Exim is trying to find all the local interfaces, it used to panic and
7511 die if the ioctl to get the interface flags failed. However, it seems that
7512 on at least one OS (Solaris 9) it is possible to have an interface that is
7513 included in the list of interfaces, but for which you get a failure error
7514 for this call. This happens when the interface is not "plumbed" into a
7515 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7516 failure of the "get flags" call assumes that the interface is down.
7518 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7519 makes life easier for people who are doing arithmetic on fields extracted
7520 from dates, where you often get leading zeros that should not be
7521 interpreted as octal.
7523 64. Added qualify_domain to the redirect router, to override the global
7526 65. If a pathologically long header line contained very many addresses (the
7527 report of this problem mentioned 10 000) and each of them was rewritten,
7528 Exim could use up a very large amount of memory. (It kept on making new
7529 copies of the header line as it rewrote, and never released the old ones.)
7530 At the expense of a bit more processing, the header rewriting function has
7531 been changed so that it no longer eats memory in this way.
7533 66. The generation of the Received: header has been moved from the time that a
7534 message starts to be received, to the time that it finishes. The timestamp
7535 in the Received: header should now be very close to that of the <= log
7536 line. There are two side-effects of this change:
7538 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7539 logged header lines no longer include the local Received: line, because
7540 it has not yet been created. The same applies to a copy of the message
7541 that is returned to a non-SMTP sender when a message is rejected.
7543 (b) When a filter file is tested using -bf, no additional Received: header
7544 is added to the test message. After some thought, I decided that this
7547 This change does not affect the value of $received_for. It is still set
7548 after address rewriting, but before local_scan() is called.
7550 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7552 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7553 gave an unhelpful panic error message, and a defer error. I have managed to
7554 change this behaviour so that it now rejects any supplied certificate,
7555 which seems right, as the list of acceptable certificates is empty.
7557 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7558 gave an unhelpful defer error. I have not managed to make this reject any
7559 supplied certificates, but the error message it gives is "no certificate
7560 supplied", which is not helpful.
7562 70. exigrep's output now also includes lines that are not associated with any
7563 message, but which match the given pattern. Implemented by a patch from
7564 Martin Sluka, which also tidied up the Perl a bit.
7566 71. Recipient callout verification, like sender verification, was using <> in
7567 the MAIL FROM command. This isn't really the right thing, since the actual
7568 sender may affect whether the remote host accepts the recipient or not. I
7569 have changed it to use the actual sender in the callout; this means that
7570 the cache record is now keyed on a recipient/sender pair, not just the
7571 recipient address. There doesn't seem to be a real danger of callout loops,
7572 since a callout by the remote host to check the sender would use <>.
7573 [SEE ABOVE: changed after hitting problems.]
7575 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7576 temporary errors. However, in the case of such a code being given after
7577 the end of a data transmission (i.e. after ".") Exim was failing to write
7578 a retry record for the message. (Yes, there was some broken host that was
7579 actually sending 8xx at this point.)
7581 73. An unknown lookup type in a host list could cause Exim to panic-die when
7582 the list was checked. (An example that provoked this was putting <; in the
7583 middle of a list instead of at the start.) If this happened during a DATA
7584 ACL check, a -D file could be left lying around. This kind of configuration
7585 error no longer causes Exim to die; instead it causes a defer error. The
7586 incident is still logged to the main and panic logs.
7588 74. Buglet left over from Exim 3 conversion. The message "too many messages
7589 in one connection" was written to the rejectlog but not the mainlog, except
7590 when address rewriting (yes!) was being logged.
7592 75. Added write_rejectlog option.
7594 76. When a system filter was run not as root (that is, when system_filter_user
7595 was set), the values of the $n variables were not being returned to the
7596 main process; thus, they were not subsequently available in the $sn
7599 77. Added +return_path_on_delivery log selector.
7601 78. A connection timeout was being treated differently from recipients deferred
7602 when testing hosts_max_try with a message that was older than the host's
7603 retry timeout. (The host should not be counted, thus allowing all hosts to
7604 be tried at least once before bouncing.) This may have been the cause of an
7605 occasionally reported bug whereby a message would remain on the queue
7606 longer than the retry timeout, but would be bounced if a delivery was
7607 forced. I say "may" because I never totally pinned down the problem;
7608 setting up timeout/retry tests is difficult. See also the next item.
7610 79. The ultimate address timeout was not being applied to errors that involved
7611 a combination of host plus message (for example, a timeout on a MAIL
7612 command). When an address resolved to a number of possible hosts, and they
7613 were not all tried for each delivery (e.g. because of hosts_max_try), a
7614 message could remain on the queue longer than the retry timeout.
7616 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7619 81. Fixed an obscure SMTP outgoing bug which required at least the following
7620 conditions: (a) there was another message waiting for the same server;
7621 (b) the server returned 5xx to all RCPT commands in the first message so
7622 that the message was not completed; (c) the server dropped the connection
7623 or gave a negative response to the RSET that Exim sends to abort the
7624 transaction. The observed case was a dropped connection after DATA that had
7625 been sent in pipelining mode. That is, the server had advertised PIPELINING
7626 but was not implementing it correctly. The effect of the bug was incorrect
7627 behaviour, such as trying another host, and this could lead to a crash.
7633 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7634 and daemon.c were passed as pointers to ints; they should have been
7635 pointers to socklen_t variables (which are typically unsigned ints).
7637 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7640 3. Fixed a really odd bug that affected only the testing scheme; patching a
7641 certain fixed string in the binary changed the value of another string that
7642 happened to be identical to the end of the original first string.
7644 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7645 name", it returns that address as the IP address. On some operating
7646 systems (e.g. Solaris), it also passes back the IP address string as the
7647 "host name". However, on others (e.g. Linux), it passes back an empty
7648 string. Exim wasn't checking for this, and was changing the host name to an
7649 empty string, assuming it had been canonicalized.
7651 5. Although rare, it is permitted to have more than one PTR record for a given
7652 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7653 all the names associated with an address, because they do in Solaris.
7654 However, it seems that they do not in Linux for data that comes from the
7655 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7656 I found this out when I moved to a new Linux workstation and tried to run
7657 the Exim test suite.
7659 To get round this problem I have changed the code so that it now does its
7660 own call to the DNS to look up PTR records when searching for a host name.
7661 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7662 addresses that are only in /etc/hosts are still found.
7664 This behaviour is, however, controlled by an option called host_lookup_
7665 order, which defaults to "bydns:byaddr". If people want to use the other
7666 order, or indeed, just use one or the other means of lookup, they can
7667 specify it in this variable.
7669 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7670 some operating systems, this comes back from gethostbyaddr() as an empty
7671 string, and this is what Exim used to test for. However, it seems that in
7672 other systems, "." is yielded. Exim now tests for this case too.
7674 7. The values of check_spool_space and check_log_space are now held internally
7675 as a number of kilobytes instead of an absolute number of bytes. If a
7676 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7677 kilobyte. This means that much larger values can be stored.
7679 8. Exim monitor: an attempt to get the action menu when not actually pointing
7680 at a message produces an empty menu entitled "No message selected". This
7681 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7682 no entries in it ("Shell widget menu has zero width and/or height"). So I
7683 have added a single, blank menu entry in this case.
7685 9. Added ${quote_local_part.
7687 10. MIME decoding is now applied to the contents of Subject: header lines when
7690 11. Now that a reference to $sender_host_address automatically causes a reverse
7691 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7692 host lookup before query-style lookups in lists that might use this
7693 variable. This has therefore been abolished, and the "net-" prefix is no
7694 longer necessary for query-style lookups.
7696 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7697 have been a typo for LFLAGS, so it has been changed.
7699 13. The install script calls Exim with "-C /dev/null" in order to find the
7700 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7701 to be output. However, since Exim outputs its version number before the
7702 error, it didn't break the script. It just looked ugly. I fixed this by
7703 always allowing "-C /dev/null" if the caller is root.
7705 14. Ignore overlarge ACL variable number when reading spool file - insurance
7706 against a later release with more variables having written the file.
7708 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7709 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7710 this, as well as the form without the "IPv6" on the front.
7712 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7713 OS/Makefile-Darwin file.
7715 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7716 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7718 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7719 variable. However, some OS use ints for inodes. I've added cast to long int
7720 to get rid of the compiler warning.
7722 19. I had forgotten to lock out "/../" in configuration file names when
7723 ALT_CONFIG_PREFIX was set.
7725 20. Routers used for verification do not need to specify transports. However,
7726 if such a router generated a host list, and callout was configured, Exim
7727 crashed, because it could not find a port number from the (non-existent)
7728 transport. It now assumes port 25 in this circumstance.
7730 21. Added the -t option to exigrep.
7732 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7733 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7734 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7735 them is set, LOOKUP_LSEARCH is forced.
7737 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7739 24. Added sender and host information to the "rejected by local_scan()" log
7740 line; previously there was no indication of these.
7742 25. Added .include_if_exists.
7744 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7745 Linux. It turns out that not all file systems support this. Apparently some
7746 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7747 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7748 not supported on the file descriptor, is now ignored when Exim is trying to
7749 sync a directory. This applies only to Linux.
7751 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7753 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7754 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7757 29. Added support for maildirsize files from supplied patch (modified a bit).
7759 30. The use of :fail: followed by an empty string could lead Exim to respond to
7760 sender verification failures with (e.g.):
7762 550 Verification failed for <xxx>
7763 550 Sender verify failed
7765 where the first response line was missing the '-' that indicates it is not
7766 the final line of the response.
7768 31. The loop for finding the name of the user that called Exim had a hardwired
7769 limit of 10; it now uses the value of finduser_retries, which is used for
7770 all other user lookups.
7772 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7775 33. Exim was neglecting to zero errno before one call of strtol() when
7776 expanding a string and expecting an integer value. On some systems this
7777 resulted in spurious "integer overflow" errors. Also, it was casting the
7778 result into an int without checking.
7780 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7781 did not work. The code looks as if it has *never* worked, though it appears
7782 to have been documented since at least release 1.62. I have made it work.
7784 35. The "timeout_DNS" error in retry rules, also documented since at least
7785 1.62, also never worked. As it isn't clear exactly what this means, and
7786 clearly it isn't a major issue, I have abolished the feature by treating it
7787 as "timeout", and writing a warning to the main and panic logs.
7789 36. The display of retry rules for -brt wasn't always showing the error code
7792 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7793 timeout_connect_A, timeout_connect_MX.
7795 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7796 to the empty sender.
7798 39. The daemon was not analysing the content of -oX till after it had closed
7799 stderr and disconnected from the controlling terminal. This meant that any
7800 syntax errors were only noted on the panic log, and the return code from
7801 the command was 0. By re-arranging the code a little, I've made the
7802 decoding happen first, so such errors now appear on stderr, and the return
7803 code is 1. However, the actual setting up of the sockets still happens in
7804 the disconnected process, so errors there are still only recorded on the
7807 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7808 connections (as happens on some IP stacks) was logged at start up time as
7809 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7810 it from "IPv6 and IPv4", which means that two separate sockets are being
7813 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7814 says whether AF_INET or AF_INET6 was passed as an argument.
7816 42. Exiwhat output was messed up when time zones were included in log
7819 43. Exiwhat now gives more information about the daemon's listening ports,
7820 and whether -tls-on-connect was used.
7822 44. The "port" option of the smtp transport is now expanded.
7824 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7825 silently ignored. Now an error message is written to the main and panic
7828 46. There's a new ACL modifier called "logwrite" which writes to a log file
7829 as soon as it is encountered.
7831 47. Added $local_user_uid and $local_user_gid at routing time.
7833 48. Exim crashed when trying to verify a sender address that was being
7836 49. Exim was recognizing only a space character after ".include". It now also
7837 recognizes a tab character.
7839 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7840 extracting the relevant information from the specification. The man page no
7841 longer contains scrambled data for the -d option, and I've added a section
7842 at the front about calling Exim under different names.
7844 51. Added "extra_headers" argument to the "mail" command in filter files.
7846 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7849 53. Installed eximstats 1.29.
7851 54. Added transport_filter_timeout as a generic transport option.
7853 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7854 Cc: header lines. This was required by RFC 822, but it not required by RFC
7857 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7858 incoming messages that did not have them. Now it does so only if the
7859 message originates locally, that is, if there is no associated remote host
7860 address. When Resent- header lines are present, this applies to the Resent-
7861 lines rather than the non-Resent- lines.
7863 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7864 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7866 58. Messages for configuration errors now include the name of the main
7867 configuration file - useful now that there may be more than one file in a
7868 list (.included file names were always shown).
7870 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7871 for those rare installations that do not start the daemon as root or run it
7872 setuid root. I've cut out the call to initgroups() if the daemon is not
7875 60. The Exim user and group can now be bound into the binary as text strings
7876 that are looked up at the start of Exim's processing.
7878 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7880 62. Added $mailstore_basename variable.
7882 63. Installed patch to sieve.c from Michael Haardt.
7884 64. When Exim failed to open the panic log after failing to open the main log,
7885 the original message it was trying to log was written to stderr and debug
7886 output, but if they were not available (the usual case in production), it
7887 was lost. Now it is written to syslog before the two lines that record the
7888 failures to open the logs.
7890 65. Users' Exim filters run in subprocesses under the user's uid. It is
7891 possible for a "deliver" command or an alias in a "personal" command to
7892 provoke an address rewrite. If logging of address rewriting is configured,
7893 this fails because the process is not running as root or exim. There may be
7894 a better way of dealing with this, but for the moment (because 4.30 needs
7895 to be released), I have disabled address rewrite logging when running a
7896 filter in a non-root, non-exim process.
7902 1. The buildconfig auxiliary program wasn't quoting the value set for
7903 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7904 not defined. This bug was masked in 4.22 by the effect that was fixed in
7907 2. Some messages that were rejected after a message id was allocated were
7908 shown as "incomplete" by exigrep. It no longer does this for messages that
7909 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7911 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7912 have allow_domain_literals set, the ID did not get logged in the <= line.
7913 Domain literals are now always recognized in Message-ID: header lines.
7915 4. The first argument for a ${extract expansion item is the key name or field
7916 number. Leading and trailing spaces in this item were not being ignored,
7917 causing some misleading effects.
7919 5. When deliver_drop_privilege was set, single queue runner processes started
7920 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7921 same command in the process it spins off) were not dropping privilege.
7923 6. When the daemon running as "exim" started a queue runner, it always
7924 re-executed Exim in the spun-off process. This is a waste of effort when
7925 deliver_drop_privilege is set. The new process now just calls the
7926 queue-runner function directly.
7932 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7935 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7936 ignored. Though the use of -oP was forcing the writing of a pid file, it
7937 was always written to the default place.
7939 3. If the message "no IP address found for host xxxx" is generated during
7940 incoming verification, it is now followed by identification of the incoming
7941 connection (so you can more easily find what provoked it).
7943 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7945 5. Added some features to "harden" Exim a bit more against certain attacks:
7947 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7948 be put in Local/Makefile. This is like the never_users runtime option,
7949 but it cannot be overridden. The default setting is "root".
7951 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7952 prefix string with which any file named in a -C command line option
7955 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7956 is retained for -C and -D only if the caller of Exim is root. Without
7957 it, the exim user may also use -C and -D and retain privilege.
7959 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7960 command line option is disabled.
7962 6. Macro names set by the -D option must start with an upper case letter, just
7963 like macro names defined in the configuration file.
7965 7. Added "dereference=" facility to LDAP.
7967 8. Two instances of the typo "uknown" in the source files are fixed.
7969 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7970 the Configure-Makefile script screwed up while processing it.
7972 10. Incorporated PCRE 4.4.
7974 11. The SMTP synchronization check was not operating right at the start of an
7975 SMTP session. For example, it could not catch a HELO sent before the client
7976 waited for the greeting. There is now a check for outstanding input at the
7977 point when the greeting is written. Because of the duplex, asynchronous
7978 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7979 way, but not yet received, when the check is performed.
7981 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7982 on TCP/IP sockets, because this apparently causes some broken clients to
7985 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7986 unchanged) from the Cygwin maintainer.
7988 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7989 was supported instead of testing for mbx. Effectively a typo.
7991 15. The spa authenticator server code was not checking that the input it
7992 received was valid base64.
7994 16. The debug output line for the "set" modifier in ACLs was not showing the
7995 name of the variable that was being set.
7997 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7999 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8000 lookup on its own. Something else was needed to trigger the lookup. For
8001 example, a match in host_lookup or the need for a host name in a host list.
8002 Now, if $sender_host_name is referenced and the host name has not yet been
8003 looked up, a lookup is performed. If the lookup fails, the variable remains
8004 empty, and $host_lookup_failed is set to "1".
8006 19. Added "eqi" as a case-independent comparison operator.
8008 20. The saslauthd authentication condition could segfault if neither service
8009 nor realm was specified.
8011 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8012 error message that was logged was misleading, and incoming SMTP
8013 connections were dropped. The message is now more accurate, and temporary
8014 errors are given to SMTP connections.
8016 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8017 (except RSET) with a 421 error, until QUIT is received. However, it was
8018 failing to send a response to QUIT.
8020 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8021 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8022 ACL is now run after the helo_try_verify_hosts code.
8024 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8025 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8026 case-independent, so other case variants are also recognized). Apparently
8027 some systems use these upper case variants.
8029 25. If more than two messages were waiting for the same host, and a transport
8030 filter was specified for the transport, Exim sent two messages over the
8031 same TCP/IP connection, and then failed with "socket operation on non-
8032 socket" when it tried to send the third.
8034 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8036 27. The extern definition of crypt16() in expand.c was not being excluded when
8037 the OS had its own crypt16() function.
8039 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8040 as a preferred synonym for return_size_limit, both as an option and as an
8043 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8045 30. Changed the default configuration ACL to relax the local part checking rule
8046 for addresses that are not in any local domains. For these addresses,
8047 slashes and pipe symbols are allowed within local parts, but the sequence
8048 /../ is explicitly forbidden.
8050 31. SPA server authentication was not clearing the challenge buffer before
8053 32. log_message in a "warn" ACL statement was writing to the reject log as
8054 well as to the main log, which contradicts the documentation and doesn't
8055 seem right (because no rejection is happening). So I have stopped it.
8057 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8058 However, I am unable to do any testing of this.
8060 34. Fixed an infelicity in the appendfile transport. When checking directories
8061 for a mailbox, to see if any needed to be created, it was accidentally
8062 using path names with one or more superfluous leading slashes; tracing
8063 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8065 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8066 contents are added to the log line that is written for every discarded
8067 recipient. (Previously a log_message setting was ignored.)
8069 36. The ${quote: operator now quotes the string if it is empty.
8071 37. The install script runs exim in order to find its version number. If for
8072 some reason other than non-existence or emptiness, which it checks, it
8073 could not run './exim', it was installing it with an empty version number,
8074 i.e. as "exim-". This error state is now caught, and the installation is
8077 38. An argument was missing from the function that creates an error message
8078 when Exim fails to connect to the socket for saslauthd authentication.
8079 This could cause Exim to crash, or give a corrupted message.
8081 39. Added isip, isip4, and isip6 to ${if conditions.
8083 40. The ACL variables $acl_xx are now saved with the message, and can be
8084 accessed later in routers, transports, and filters.
8086 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8087 strings in the file are not string-expanded.
8089 42. If a MAIL command specified a SIZE value that was too large to fit into an
8090 int variable, the check against message_size_limit failed. Such values are
8091 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8092 day this will have to be increased, but I don't think I want to be around
8093 when emails are that large.
8100 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8101 iconv() is not standard in FreeBSD.
8103 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8104 IPv6 enabled. The observed symptom was a segmentation fault on return from
8105 the function os_common_find_running_interfaces() in src/os.c.
8107 3. In the check_special_case() function in daemon.c I had used "errno" as an
8108 argument name, which causes warnings on some systems. This was basically a
8109 typo, since it was named "eno" in the comments!
8111 4. The code that waits for the clock to tick (at a resolution of some fraction
8112 of a second) so as to ensure message-id uniqueness was always waiting for
8113 at least one whole tick, when it could have waited for less. [This is
8114 almost certainly not relevant at current processor speeds, where it is
8115 unlikely to ever wait at all. But we try to future-proof.]
8117 5. The function that sleeps for a time interval that includes fractions of a
8118 second contained a race. It did not block SIGALRM between setting the
8119 timer, and suspending (a couple of lines later). If the interval was short
8120 and the sigsuspend() was delayed until after it had expired, the suspension
8121 never ended. On busy systems this could lead to processes getting stuck for
8124 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8125 process, before it forks any delivery processes. The open lookup caching
8126 mechanism meant that the open file or database connection was passed into
8127 the delivery process. The problem was that delivery processes always tidy
8128 up cached lookup data. This could cause a problem for the next delivery
8129 process started by the queue runner, because the external queue runner
8130 process does not know about the closure. So the next delivery process
8131 still has data in the lookup cache. In the case of a file lookup, there was
8132 no problem because closing a file descriptor in a subprocess doesn't affect
8133 the parent. However, if the lookup was caching a connection to a database,
8134 the connection was closed, and the second delivery process was likely to
8135 see errors such as "PGSQL: query failed: server closed the connection
8136 unexpectedly". The problem has been fixed by closing all cached lookups
8137 in a queue runner before running a delivery process.
8139 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8140 seem to have the "const" qualifier which it has on other OS. I've
8143 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8144 *of the same type* (client or server) with the same public name that an
8145 error should be diagnosed.
8147 9. When Exim looked up a host name for an IP address, but failed to find the
8148 original IP address when looking up the host name (a safety check), it
8149 output the message "<ip address> does not match any IP for NULL", which was
8150 confusing, to say the least. The bug was that the host name should have
8151 appeared instead of "NULL".
8153 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8154 user that is built into the binary, and the -C or -D options is used, root
8155 privilege is dropped before the configuration file is read. In addition,
8156 logging is switched to stderr instead of the normal log files. If the
8157 configuration then re-defines the Exim user, the unprivileged environment
8158 is probably not what is expected, so Exim logs a panic warning message (but
8161 However, if deliver_drop_privilege is set, the unprivileged state may well
8162 be exactly what is intended, so the warning has been cut out in that case,
8163 and Exim is allowed to try to write to its normal log files.
8169 1. smtp_return_error_details was not giving details for temporary sender
8170 or receiver verification errors.
8172 2. Diagnose a configuration error if two authenticators have the same public
8175 3. Exim used not to create the message log file for a message until the first
8176 delivery attempt. This could be confusing when incoming messages were held
8177 for policy or load reasons. The message log file is now created at the time
8178 the message is received, and an initial "Received" line is written to it.
8180 4. The automatically generated man page for command line options had a minor
8181 bug that caused no ill effects; however, a more serious problem was that
8182 the procedure for building the man page automatically didn't always
8183 operate. Consequently, release 4.20 contains an out-of-date version. This
8184 shouldn't happen again.
8186 5. When building Exim with embedded Perl support, the script that builds the
8187 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8188 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8190 6. The freeze_tell option was not being used for messages that were frozen on
8191 arrival, either by an ACL or by local_scan().
8193 7. Added the smtp_incomplete_transaction log selector.
8195 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8196 was accepting AUTH without a new EHLO.
8198 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8199 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8202 10. Logging of TCP/IP connections (when configured) now happens in the main
8203 daemon process instead of the child process, so that the TCP/IP connection
8204 count is more accurate (but it can never be perfect).
8206 11. The use of "drop" in a nested ACL was not being handled correctly in the
8207 outer ACL. Now, if condition failure induced by the nested "drop" causes
8208 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8209 or "require"), the connection is dropped.
8211 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8212 that yield "discard" can now be used with an "accept" or a "discard" verb,
8213 but an error is generated for any others (because I can't see a useful way
8214 to define what should happen).
8216 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8217 are now processed in the same way as lines in the Exim configuration file.
8218 In particular, continuation lines are supported.
8220 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8222 15. Added -ti meaning -t -i.
8224 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8225 domains, and warn by logging if others are found.
8227 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8228 returned from GIFCONF when Exim is trying to find the list of interfaces on
8229 a host. The code in os.c has been modified to copy each ifreq to an aligned
8230 structure in all cases.
8232 Also, in some cases, the returned ifreq's were being copied to a 'struct
8233 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8234 means the last couple of bytes of an IPv6 address could be chopped if the
8235 ifreq contained only a normal sockaddr (14 bytes storage).
8237 18. Named domain lists were not supported in the hosts_treat_as_local option.
8238 An entry such as +xxxx was not recognized, and was treated as a literal
8241 19. Ensure that header lines added by a DATA ACL are included in the reject log
8242 if the ACL subsequently rejects the message.
8244 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8245 MD5 (which is deprecated).
8247 21. When testing a filter file using -bf, Exim was writing a message when it
8248 took the sender from a "From " line in the message, but it was not doing so
8249 when it took $return_path from a Return-Path: header line. It now does.
8251 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8252 with a valid header line field name (a series of printing characters
8253 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8255 23. Changed "disc" in the source to "disk" to conform to the documentation and
8256 the book and for uniformity.
8258 24. Ignore Sendmail's -Ooption=value command line item.
8260 25. When execve() failed while trying to run a command in a pipe transport,
8261 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8262 could be confused with a return value of 69 from the command itself. This
8263 has been changed to 127, the value the shell returns if it is asked to run
8264 a non-existent command. The wording for the related log line suggests a
8265 non-existent command as the problem.
8267 26. If received_header_text expands to an empty string, do not add a Received:
8268 header line to the message. (Well, it adds a token one on the spool, but
8269 marks it "old" so that it doesn't get used or transmitted.)
8271 27. Installed eximstats 1.28 (addition of -nt option).
8273 28. There was no check for failure on the call to getsockname() in the daemon
8274 code. This can fail if there is a shortage of resources on the system, with
8275 ENOMEM, for example. A temporary error is now given on failure.
8277 29. Contrary to the C standard, it seems that in some environments, the
8278 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8279 program. Exim now does this explicitly; it affects the formatting of
8280 timestamps using strftime().
8282 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8283 complaints. I've now initialized all the variables, to avoid this.
8285 32. Header lines added by a system filter were not being "seen" during
8286 transport-time rewrites.
8288 33. The info_callback() function passed to OpenSSL is set up with type void
8289 (*)(SSL *, int, int), as described somewhere. However, when calling the
8290 function (actually a macro) that sets it up, the type void(*)() is
8291 expected. I've put in a cast to prevent warnings from picky compilers.
8293 34. If a DNS black list lookup found a CNAME record, but there were no A
8294 records associated with the domain it pointed at, Exim crashed.
8296 35. If a DNS black list lookup returned more than one A record, Exim ignored
8297 all but the first. It now scans all returned addresses if a particular IP
8298 value is being sought. In this situation, the contents of the
8299 $dnslist_value variable are a list of all the addresses, separated by a
8302 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8303 to accept a host name and all its aliases if the forward lookup for any of
8304 them yielded the IP address of the incoming connection. Now it accepts only
8305 those names whose forward lookup yields the correct IP address. Any other
8306 names are discarded. This closes a loophole whereby a rogue DNS
8307 administrator could create reverse DNS records to break through a
8308 wildcarded host restriction in an ACL.
8310 37. If a user filter or a system filter that ran in a subprocess used any of
8311 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8312 the wrong values were passed to the pipe command ($thisaddress had the
8313 value of $0, $0 had the value of $1, etc). This bug was introduced by
8314 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8316 38. Improved the line breaking for long SMTP error messages from ACLs.
8317 Previously, if there was no break point between 40 and 75 characters, Exim
8318 left the rest of the message alone. Two changes have been made: (a) I've
8319 reduced the minimum length to 35 characters; (b) if it can't find a break
8320 point between 35 and 75 characters, it looks ahead and uses the first one
8321 that it finds. This may give the occasional overlong line, but at least the
8322 remaining text gets split now.
8324 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8325 file descriptors might be low, and that setting 1000 would always raise it.
8326 It turns out that in some environments, the limit is already over 1000 and
8327 that lowering it causes trouble. So now Exim takes care not to decrease it.
8329 40. When delivering a message, the value of $return_path is set to $sender_
8330 address at the start of routing (routers may change the value). By an
8331 oversight, this default was not being set up when an address was tested by
8332 -bt or -bv, which affected the outcome if any router or filter referred to
8335 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8336 log only when "message" or "log_message" are set. However, if one of the
8337 conditions was an address verification, or a call to a nested ACL, the
8338 messages generated by the underlying test were being passed through. This
8339 no longer happens. The underlying message is available in $acl_verify_
8340 message for both "message" and "log_message" expansions, so it can be
8341 passed through if needed.
8343 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8344 new expansion $bh_ to give the encoded byte string without charset
8345 translation. Translation happens only if iconv() is available; HAVE_ICONV
8346 indicates this at build time. HEADERS_CHARSET gives the charset to
8347 translate to; headers_charset can change it in the configuration, and
8348 "headers charset" can change it in an individual filter file.
8350 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8351 that creates RFC 2047 encoded "words" labels them as that charset instead
8352 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8353 expansion operator; (ii) when Exim creates a From: line for a local
8354 message; (iii) when a header line is rewritten to include a "phrase" part.
8356 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8357 buggy, causing it to skip the first lines of messages whose message ID
8358 ended in 'D'. This would not have bitten before Exim release 4.14, because
8359 message IDs were unlikely to end in 'D' before then. The effect was to have
8360 incorrect size information for certain domains.
8362 45. #include "config.h" was missing at the start of the crypt16.c module. This
8363 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8366 46. If there was a timeout during a "random" callout check, Exim treated it as
8367 a failure of the random address, and carried on sending RSET and the real
8368 address. If the delay was just some slowness somewhere, the response to the
8369 original RCPT would be taken as a response to RSET and so on, causing
8370 mayhem of various kinds.
8372 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8373 when I implemented it. It didn't allow for the fact that some option values
8374 may legitimately be negative (e.g. size_addition), and it didn't even do
8375 the right test for positive values.
8377 48. Domain names in DNS records are case-independent. Exim always looks them up
8378 in lower case. Some resolvers return domain names in exactly the case they
8379 appear in the zone file, that is, they may contain uppercase letters. Not
8380 all resolvers do this - some return always lower case. Exim was treating a
8381 change of case by a resolver as a change of domain, similar to a widening
8382 of a domain abbreviation. This triggered its re-routing code and so it was
8383 trying to route what was effectively the same domain again. This normally
8384 caused routing to fail (because the router wouldn't handle the domain
8385 twice). Now Exim checks for this case specially, and just changes the
8386 casing of the domain that it ultimately uses when it transmits the message
8389 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8392 50. If a filter generated a file delivery with a non-absolute name (possible if
8393 no home directory exists for the router), the forbid_file option was not
8396 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8397 the existing equality matching.
8399 52. Exim was using ints instead of ino_t variables in some places where it was
8400 dealing with inode numbers.
8402 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8403 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8404 called TMPDIR, and if it finds it is different, it changes its value.
8406 54. The smtp_printf() function is now made available to local_scan() so
8407 additional output lines can be written before returning. There is also an
8408 smtp_fflush() function to enable the detection of a dropped connection.
8409 The variables smtp_input and smtp_batched_input are exported to
8412 55. Changed the default runtime configuration: the message "Unknown user"
8413 has been removed from the ACL, and instead placed on the localuser router,
8414 using the cannot_route_message feature. This means that any verification
8415 failures that generate their own messages won't get overridden. Similarly,
8416 the "Unrouteable address" message that was in the ACL for unverifiable
8417 relay addresses has also been removed.
8419 56. Added hosts_avoid_esmtp to the smtp transport.
8421 57. The exicyclog script was not checking for the esoteric option
8422 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8423 will work only if exicyclog is run under the appropriate euid.
8425 58. Following a discussion on the list, the rules by which Exim recognises line
8426 endings on incoming messages have been changed. The -dropcr and drop_cr
8427 options are now no-ops, retained only for backwards compatibility. The
8428 following line terminators are recognized: LF CRLF CR. However, special
8429 processing applies to CR:
8431 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8432 nor a local message in the state where . is a terminator.
8434 (ii) If a bare CR is encountered in a header line, an extra space is added
8435 after the line terminator so as not to end the header. The reasoning
8436 behind this is that bare CRs in header lines are most likely either
8437 to be mistakes, or people trying to play silly games.
8439 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8440 was being incorrectly given as 18 bytes larger than it should have been.
8441 This is a VOB (very old bug).
8443 60. This may never have affected anything current, but just in case it has:
8444 When the local host is found other than at the start of a list of hosts,
8445 the local host, those with the same MX, and any that follow, are discarded.
8446 When the list in question was part of a longer list of hosts, the following
8447 hosts (not currently being processed) were also being discarded. This no
8448 longer happens. I'm not sure if this situation could ever has previously
8451 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8454 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8455 header. This is something that is recommended in a new Internet Draft, and
8456 is something that is documented as being done by Sendmail. There are two
8457 possible values. For messages generated by the autoreply transport, Exim
8460 Auto-Submitted: auto-replied
8462 whereas for all other generated messages (e.g. bounces) it adds
8464 Auto-Submitted: auto-generated
8466 63. The "personal" condition in filters now includes a test for the
8467 Auto-Submitted: header. If it contains the string "auto-" the message it
8468 not considered personal.
8470 64. Added rcpt_include_affixes as a generic transport option.
8472 65. Added queue_only_override (default true).
8474 66. Added the syslog_duplication option.
8476 67. If what should have been the first header line of a message consisted of
8477 a space followed by a colon, Exim was mis-interpreting it as a header line.
8478 It isn't of course - it is syntactically invalid and should therefore be
8479 treated as the start of the message body. The misbehaviour could have
8480 caused a number of strange effects, including loss of data in subsequent
8481 header lines, and spool format errors.
8483 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8484 client host had authenticated. This control can now be exercised by an ACL
8485 for more flexibility.
8487 69. By default, callouts do not happen when testing with -bh. There is now a
8488 variant, -bhc, which does actually run the callout code, including
8489 consulting and updating the callout cache.
8491 70. Added support for saslauthd authentication, courtesy of Alexander
8494 71. If statvfs() failed on the spool or log directories while checking their
8495 size for availability, Exim confusingly gave the error "space shortage".
8496 Furthermore, in debugging mode it crashed with a floating point exception.
8497 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8498 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8499 serious problem, Exim now writes to the main and panic logs when this
8500 happens, with details of the failure. It then refuses to accept the
8501 incoming message, giving the message "spool directory problem" or "log
8502 directory problem" with a 421 code for SMTP messages.
8504 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8505 0, 1, and 2 exist, because some OS complain for execs without them (see
8506 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8507 descriptors. However, the code omitted to check that the open succeeded,
8508 causing mysterious errors if for some reason the permissions on /dev/null
8509 got screwed. Now Exim writes a message to the main and panic logs, and
8510 bombs out if it can't open /dev/null.
8512 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8513 interact so that it is all more flexible. It is supposed to remain
8514 backwards compatible. Also added extra_local_interfaces.
8516 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8517 to bomb out with an assertion failure - to the client this appears as a
8518 connection drop. This problem occurs in the part of the code that was taken
8519 from the Samba project. Fortunately, the assertion is in a very simple
8520 function, so I have fixed this by reproducing the function inline in the
8521 one place where it is called, and arranging for authentication to fail
8522 instead of killing the process with assert().
8524 75. The SPA client code was not working when the server requested OEM rather
8525 than Unicode encoding.
8527 76. Added code to make require_files with a specific uid setting more usable in
8528 the case where statting the file as root fails - usually a non-root-mounted
8529 NFS file system. When this happens and the failure is EACCES, Exim now
8530 forks a subprocess and does the per-uid checking as the relevant uid.
8532 77. Added process_log_path.
8534 78. If log_file_path was not explicitly set, a setting of check_log_space or
8535 check_log_inodes was ignored.
8537 79. If a space check for the spool or log partitions fails, the incident is now
8538 logged. Of course, in the latter case the data may get lost...
8540 80. Added the %p formatting code to string_format() so that it can be used to
8541 print addresses in debug_print(). Adjusted all the address printing in the
8542 debugging in store.c to use %p rather than %d.
8544 81. There was a concern that a line of code in smtp_in.c could overflow a
8545 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8546 initially expressed, the concern was not well-founded, because trailing
8547 spaces are removed early. However, if the trailing spaces were followed by
8548 a NULL, they did not get removed, so the overflow was possible. Two fixes
8551 (a) I re-wrote the offending code in a cleaner fashion.
8552 (b) If an incoming SMTP command contains a NULL character, it is rejected
8555 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8556 runs initgroups(), so that if the Exim user is in any additional groups,
8557 they will be used during message reception.
8563 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob