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 The hosts_connection_nolog main option now also controls "no MAIL in
9 SMTP connection" log lines.
11 JH/02 Option default value updates:
12 - queue_fast_ramp (main) true (was false)
13 - remote_max_parallel (main) 4 (was 2)
15 JH/03 Cache static regex pattern compilations, for use by ACLs.
17 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
18 Make the rewrite never match and keep the logging. Trust the
19 admin to be using verify=header-syntax (to actually reject the message).
21 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
22 (under Linux) we watch the dir containing the final file; previously
23 it would be the dir with the first symlink. We still do not monitor
26 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
27 dn_expand() is happy to pass them through.
29 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
30 the file with the proof had an unchanged name, the new proof(s) were
31 loaded on top of the old ones (and nover used; the old ones were stapled).
33 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
34 more than one message arrived in a single connection a reference from
35 the earlier message could be re-used. Often a sigsegv resulted.
36 These variables were introduced in Exim 4.87.
37 Debug help from Graeme Fowler.
39 JH/09 Fix ${filter } for conditions that modify $value. Previously the
40 modified version would be used in construction the result, and a memory
43 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
44 Find and fix by Jasen Betts.
46 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
47 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
48 the systemwide configuration to override the Exim config.
50 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
51 API changes in libopendmarc.
53 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
54 pid 1, in the normal "background daemon" mode, having to drop process-
55 group leadership also lost track of needing to create listener sockets.
57 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
58 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
60 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
61 a capture group which obtained no text (eg. "(abc)*" matching zero
62 occurrences) could cause a segfault if the corresponding $<n> was
65 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
66 included a close-brace character (eg. it itself used an expansion) an
69 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
70 starting TLS. Previously it was after, meaning that attackers on such
71 ports had to be screened using the host_reject_connection main config
72 option. The new sequence aligns better with the STARTTLS behaviour, and
73 permits defences against crypto-processing load attacks, even though it
74 is strictly an incompatible change.
75 Also, avoid sending any SMTP fail response for either the connect ACL
76 or host_reject_connection, for TLS-on-connect ports.
78 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
79 Previously this was not permitted, but it makes reasonable sense.
80 While there, restore a restriction on using it from a connect ACL; given
81 the change JH/16 it could only return false (and before 4.91 was not
84 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
85 was exactly sized compared to the log buffer, a crash occurred with the
86 misleading message "bad memory reference; pool not found".
87 Found and traced by Jasen Betts.
89 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
90 dns_again_means_nonexist included an element causing a DNS lookup which
91 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
92 included (though probably not limited to) a process crash from stack
93 memory limit, or from excessive open files. Replace this with a paniclog
94 whine (as this is likely a configuration error), and returning
97 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
98 this always failed, probably leading to the usual downgrade to in-clear
101 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
102 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
103 to not checking that list for these lookups.
105 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
108 JH/23 Fix crash in string expansions. Previously, if an empty variable was
109 immediately followed by an expansion operator, a null-indirection read
110 was done, killing the process.
112 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
113 include an SMTP response string which is longer than that supported
114 by the delivering transport. Alleviate by wrapping such lines before
117 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
118 chars (RFC limit). Previously a limit of 12 items was made, which with
119 a not-impossible References: in the message being bounced could still
120 be over-large and get stopped in the transport.
122 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
123 close. Previously a bare socket close was done.
125 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
128 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
129 message being considered for delivery was already being handled by
130 another process, and having an SMTP connection already open, the function
131 to close it tried to use an uninitialized variable. This would afftect
132 high-volume sites more, especially when running mailing-list-style loads.
133 Pollution of logs was the major effect, as the other process delivered
134 the message. Found and partly investigated by Graeme Fowler.
136 JH/29 Change format of the internal ID used for message identification. The old
137 version only supported 31 bits for a PID element; the new 64 (on systems
138 which can use Base-62 encoding, which is all currently supported ones
139 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
140 and must use Base-36). The new ID is 23 characters rather than 16, and is
141 visible in various places - notably logs, message headers, and spool file
142 names. Various of the ancillary utilities also have to know the format.
143 As well as the expanded PID portion, the sub-second part of the time
144 recorded in the ID is expanded to support finer precision. Theoretically
145 this permits a receive rate from a single comms channel of better than the
147 The major timestamp part of the ID is not changed; at 6 characters it is
148 usable until about year 3700.
149 Updating from previously releases is fully supported: old-format spool
150 files are still usable, and the utilities support both formats. New
151 message will use the new format. The one hints-DB file type which uses
152 message-IDs (the transport wait- DB) will be discarded if an old-format ID
153 is seen; new ones will be built with only new-format IDs.
154 Optionally, a utility can be used to convert spool files from old to new,
155 but this is only an efficiency measure not a requirement for operation
156 Downgrading from new to old requires running a provided utility, having
157 first stopped all operations. This will convert any spool files from new
158 back to old (losing time-precision and PID information) and remove any
159 wait- hints databases.
161 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
162 we treated them as item separators when parsing for a list item, but they
163 need to be protected by the doublequotes. While there, add handling for
166 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
167 Found and fixed by Jasen Betts. No testcase for this as my usual text
168 editor insists on emitting only valid UTF-8.
170 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
171 a null-indirection SIGSEGV for the receive process.
173 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
174 -bh use. Internal checking would spot this and take a panic.
176 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
177 In 4.96 this would expand to empty.
179 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
180 certificate. Find and fix by Andreas Metzler.
182 JH/36 Add ARC info to DMARC hostory records.
184 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
185 or fakedefer. Previously the sender could discover that the message
186 had in fact been accepted.
188 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
189 sequences. Previously the input was not noted as being tainted; notably
190 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
191 bad coding of authenticators.
193 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
194 and ${tr...}. Found and diagnosed by Heiko Schlichting.
196 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
197 could be triggered by externally-supplied input. Found by Trend Micro.
200 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
201 be triggered by externally-controlled input. Found by Trend Micro.
204 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
205 be triggered by externally-controlled input. Found by Trend Micro.
208 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
209 Make the rewrite never match and keep the logging. Trust the
210 admin to be using verify=header-syntax (to actually reject the message).
212 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
215 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
220 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
221 after reception to before a subsequent reception. This should
222 mean slightly faster delivery, and also confirmation of reception
225 JH/02 Move from using the pcre library to pcre2. The former is no longer
226 being developed or supported (by the original developer).
228 JH/03 Constification work in the filters module required a major version
229 bump for the local-scan API. Specifically, the "headers_charset"
230 global which is visible via the API is now const and may therefore
231 not be modified by local-scan code.
233 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
234 sendfile() didi not account for the way the ClamAV driver code called it.
236 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
237 time check was being done for every character; replace that with one
240 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
241 sent was prefixed with a length byte.
243 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
244 RFC 5321. Previously Dovecot (at least) would log errors during
247 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
248 in 4.61 without a reason logged. Binaries will be bigger, which might
249 matter on diskspace-constrained systems, but debug is easier.
251 JH/09 Fix macro-definition during "-be" expansion testing. The move to
252 write-protected store for macros had not accounted for these runtime
253 additions; fix by removing this protection for "-be" mode.
255 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
256 handing out large-numbered file descriptors, violating the usual Unix
257 assumption (and required by Posix) that the lowest possible number will be
258 allocated by the kernel when a new one is needed. In the daemon, and any
259 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
260 are not useable for FD_SET() [and hence select()] and overwrite the stack.
261 Assorted crashes happen.
263 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
264 main-section options or in a connect ACL, the value from the first ever
265 connection was never replaced for subsequent connections. Found by
268 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
269 though only once PCRE2 was introduced: the memory accounting used under
270 debug offset allocations by an int, giving a hard trap in early startup.
271 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
273 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
274 with underbars is given. The write-protection of configuration introduced
275 in 4.95 trapped when normalisation was applied to an option not needing
278 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
280 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
281 failing to close the certificates directory, every hour or any time it
284 JH/16 Debugging initiated by an ACL control now continues through into routing
285 and transport processes. Previously debugging stopped any time Exim
286 re-execs, or for processing a queued message.
288 JH/17 The "expand" debug selector now gives more detail, specifically on the
289 result of expansion operators and items.
291 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
292 bad comparison between the option value and the name of the file to
293 be included was done, and a mismatch was wrongly identified.
294 4.88 to 4.95 are affected.
296 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
298 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
299 supplied as the db file-pair basename. Previously, if a directory
300 path was given, for example via the autoreply "once" option, the DB
301 file.pag and file.dir files would be created in that directory's
304 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
305 "taint" log_selector. These were previously deprecated.
307 JH/22 Fix static address-list lookups to properly return the matched item.
308 Previously only the domain part was returned.
310 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
311 the call into OpenSSL to send a TLS Close was being repeated; this
312 resulted in the library waiting for the peer's Close. If that was never
313 sent we waited forever. Fix by tracking send calls.
315 JH/24 The ${run} expansion item now expands its command string elements after
316 splitting. Previously it was before; the new ordering makes handling
317 zero-length arguments simpler. The old ordering can be obtained by
318 appending a new option "preexpand", after a comma, to the "run".
320 JH/25 Taint-check exec arguments for transport-initiated external processes.
321 Previously, tainted values could be used. This affects "pipe", "lmtp" and
322 "queryprogram" transport, transport-filter, and ETRN commands.
323 The ${run} expansion is also affected: in "preexpand" mode no part of
324 the command line may be tainted, in default mode the executable name
327 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
328 the the facility was not passed across execs, and only the first message
329 passed over a connection could use BDAT; any further ones using DATA.
331 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
332 uses $sending_ip_address and an interface is specified.
333 Previously any use of the local address in the EHLO name disabled
334 PIPECONNECT, the common case being to use the rDNS of it.
336 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
337 resumption. Previously verify failed because no certificate status is
338 passed on the wire for the restarted session. Fix by using the recorded
339 ocsp status of the stored session for the new connection.
341 JH/29 TLS resumption: the key for session lookup in the client now includes
342 more info that a server could potentially use in configuring a TLS
343 session, avoiding oferring mismatching sessions to such a server.
344 Previously only the server IP was used.
346 JH/30 Fix string_copyn() for limit greater than actual string length.
347 Previously the copied amount was the limit, which could result in a
348 overlapping memcpy for newly allocated destination soon after a
349 source string shorter than the limit. Found/investigated by KM.
351 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
352 close; it may be needed for a subsequent connection. This caused a
353 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
355 JH/32 Fix CHUNKING for a second message on a connection when the first was
356 rejected. Previously we did not reset the chunking-offered state, and
357 erroneously rejected the BDAT command. Investigation help from
360 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
361 an empty address. Previously the expansion returned an error.
363 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
364 proxy. Previously these were misparsed, leading to paniclog entries.
370 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
371 related applications. Previously an "H" was used where available info
372 says that "M" should be, so change to match.
374 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
375 as arguments, so an implementation trying to copy these into a local
376 buffer was taking a taint-enforcement trap. Fix by using dynamically
377 created buffers. Similar fix for radius expansion condition.
379 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
380 reasonable, eg. to count headers. Fix by using dynamically created
381 buffers rather than a local. Do similar fixes for ACL actions "dcc",
382 "log_reject_target", "malware" and "spam"; the arguments are expanded
383 so could be handling tainted values.
385 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
386 broken the (no-op) support for this sendmail command. Restore it
387 to doing nothing, silently, and returning good status.
389 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
390 record path was given (or the default used) without a leading directory
391 path, an error occurred on trying to open it. Use the transport's working
394 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
395 transport. Previously it was the name on the DNS A-record; use instead
396 the head of the CNAME chain leading there (if there is one). This seems
397 to align better with RFC 6125.
399 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
400 smtp_accept_max_per_host allocated resources which were not released
401 when the limit was exceeded. This eventually crashed the daemon. Fix
402 by adding a release action in that path.
404 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
405 expanded; previously using tainted values was rejected. Fix by using
406 dynamically-created buffers.
408 JH/09 Relax restrictions on ACL verify condition needing access to message
409 headers. Previously they were only permitted in data and non-smtp ACLs;
410 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
411 not_blind, header_sender and header_names_ascii verification.
413 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
414 Previously a macro used one argument twice; when called with the
415 argument as an expression having side-effects, incorrect operation
416 resulted. Use an inlineable function.
418 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
419 held open for a verify callout. Previously this wan not accounted for
420 and a corrupt onward SMTP conversation resulted.
422 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
423 Previously they were embedded naively in the constructed address; when
424 needed, strip the quoting and quote the entire local_part.
425 Also make the inbound_srs expansion condition handle quoting.
427 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
428 excluded, not matching the documentation.
430 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
431 was given for the sqlite_dbfile a trap resulted.
433 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
434 "name" argument resulted in a trap. There is no reason to disallow such;
435 this was a coding error.
437 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
438 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
439 spent suspended, ignoring the POSIX definition. Previously we assumed
440 it did and a constant offset from real time could be used as a correction.
441 Change to using the same clock source for the start-of-message and the
442 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
443 exists, just to get a clock slightly more aligned to reality.
445 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
446 RFC says it is optional some validators care. The missing char was not
447 intended but triggered by a line-wrap alignment. Discovery and fix by
448 Guillaume Outters, hacked on by JH.
450 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
451 name being quoted was tainted a trap would be taken. Fix by using
452 dynamically created buffers. The routine could have been called by a
453 rewrite with the "h" flag, by using the "-F" command-line option, or
454 by using a "name=" option on a control=submission ACL modifier.
456 JH/19 SPF: change the Authentication-Results expansion component to give
457 smtp.helo when the sender domain is empty. Previously it gave
460 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
461 not in 127.0.0.0/8 to help in spotting list domains taken over by a
462 domain-parking registrar.
464 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
465 Previously when a whitespace character was specified it was not inserted
466 after removing the newline.
468 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
469 the domain part of the recipient address. This overrides any tls_sni
470 option set, which was previously used.
472 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
475 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
476 is_tainted() had an off-by-one error in the overenthusiastic direction.
477 Find and fix by Gavan. Although NetBSD is not a supported platform for
478 4.94 this bug could affect other platforms.
480 PP/01 Fix default prime selection to be consistent.
481 One path used ike23 still, instead of exim.dev.20160529.3; now both
482 execution flows will use the same DH primes (currently
483 exim.dev.20160529.3).
485 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
486 option in smtp transport, to match the documentation. Previously
487 verification was not being done in some cases where it should have been.
489 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
490 than one server was defined and depending on the platform memory layout
491 details, an internal consistency trap could be hit while walking the list
494 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
495 files. The value is used by the authresults expansion item. Previously
496 if this was used in a router or transport, a crash could result.
498 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
499 have a sufficient incidence of EINTR returns from select that an
500 interaction with other operations done by the main daemon loop exposed
501 a bug in the error-handling. This was benign apart from the log
504 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
505 for consistency with delivered "=>" and failed "**" lines. While we're
506 there, handle PRX and TFO.
508 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
509 applied. This resulted, if any header-line rewrite rules were configured,
510 in a panic-log triggerable by sending a message with a long address in
511 a header. Fix by increasing the arbitrary limit to larger than a single
512 (dewrapped) 5322 header line maximum size.
514 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
515 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
516 RFC 6648 which deprecates X- options in protocols as a general practice.
517 Changeover between the implementations is handled by the mechanisms
520 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
521 a remote which uses callout/hold. Previously the local was queued.
523 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
526 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
527 log line. The asterisk is now consistently appended to the remote IP
528 (and port, if given), and will also be provided on defer and fail log
529 lines. Previously it could be placed on the local IP if that was being
530 logged, and was only provided on delivery lines.
532 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
534 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
535 authenticator client_send option. Previously the next char, after a pair
536 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
537 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
538 documentation. There is still no way to get a leading ^ immediately
539 after a NUL (ie. for the password of a PLAIN method authenticator.
541 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
542 files. For bad sizes read, delete the record and whine to paniclog.
544 JH/38 When logging an AUTH failure, as server, do not include sensitive
545 information. Previously, the credentials would be included if given
546 as part of the AUTH command line and an ACL denied authentication.
548 JH/39 Bug 2691: fix $local_part_data. When the matching list element
549 referred to a file, bad data was returned. This likely also affected
552 JH/40 The gsasl authenticator now supports caching of the salted password
553 generated by the client-side implementation. This required the addition
554 of a new variable: $auth4.
556 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
557 left undeleted; the attempt to re-create it then failed - resulting in
558 the usual "SIGHUP tp have daemon reload configuration" to not work.
559 This affected any platform not supporting "abstract" Unix-domain
560 sockets (i.e. not Linux).
562 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
563 recipients" response to RCPT in a later response, with a 250. The
564 previous coding assumed this would not happen, and under PIPELINING
565 would result in both lost and duplicate recipients for a message.
567 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
568 Previously the weighting was incorrectly applied. Similar fix for socks
569 proxies. Found and fixed by Heiko Schlichting.
571 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
572 not handle sub-lists included using the +namedlist syntax. While
573 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
574 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
577 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
578 Previously the permanent pool was used, so the sore could not be freed.
579 This meant a connection with many messages would use continually-growing
582 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
583 per-pool so as not to waste too much space. Previously a constant size
584 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
585 DOS attacks harder. The cost is wasted memory use in the larger blocks.
587 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
588 same space repeatedly, and a smaller process footprint.
590 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
591 for GnuTLS. Previously the name started "%s" which, while not a bug,
592 looked as if if might be one.
594 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
595 "continued connection") make the $proxy_* variables available. Previously
596 the information was not passed across the exec() call for subsequent
597 transport executions. This also mean that the log lines for the
598 messages can show the proxy information.
600 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
601 receive time. With modern systems the difference is significant.
602 The historical behaviour can be restored by disabling (a new) log_selector
603 "queue_time_exclusive".
605 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
606 content) were specifically tested for. Now, also cover expansions
607 resulting in ACL names and inline ACL content.
609 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
610 making it unusable in complex expressions.
612 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
613 fail with a taint-check on the spoolfile name, and leave the message
616 HS/01 Enforce absolute PID file path name.
618 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
620 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
622 PP/02 Bug 2643: Correct TLS DH constants.
623 A missing NUL termination in our code-generation tool had led to some
624 incorrect Diffie-Hellman constants in the Exim source.
625 Reported by kylon94, code-gen tool fix by Simon Arlott.
627 PP/03 Impose security length checks on various command-line options.
628 Fixes CVE-2020-SPRSS reported by Qualys.
630 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
631 better. Reported by Qualys.
633 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
634 providing a particularly obnoxious sender full name.
637 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
639 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
642 PP/08 Change default for recipients_max from unlimited to 50,000.
644 PP/09 Fix security issue with too many recipients on a message (to remove a
645 known security problem if someone does set recipients_max to unlimited,
646 or if local additions add to the recipient list).
647 Fixes CVE-2020-RCPTL reported by Qualys.
649 PP/10 Fix security issue in SMTP verb option parsing
650 Fixes CVE-2020-EXOPT reported by Qualys.
652 PP/11 Fix security issue in BDAT state confusion.
653 Ensure we reset known-good where we know we need to not be reading BDAT
654 data, as a general case fix, and move the places where we switch to BDAT
655 mode until after various protocol state checks.
656 Fixes CVE-2020-BDATA reported by Qualys.
658 HS/03 Die on "/../" in msglog file names
660 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
661 the Exim runtime user are allowed to create files.
663 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
666 QS/03 When reading the output from interpreted forward files we do not
667 pass the pipe between the parent and the interpreting process to
668 executed child processes (if any).
670 QS/04 Always die if requested from internal logging, even is logging is
673 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
674 the API; compilation noo longer completes with DMARC support included.
675 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
676 been reported on other platforms.
678 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
680 JH/56 Make the majority of info read from config files readonly, for defence-in-
681 depth against exploits. Suggestion by Qualys.
682 Not supported on Solaris 10.
684 JH/57 Fix control=fakreject for a custom message containing tainted data.
685 Previously this resulted in a log complaint, due to a re-expansion present
686 since fakereject was originally introduced.
688 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
689 modifier was given, a loop resulted.
691 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
692 pipelined SMTP command followed the BDAT LAST then it would be
693 incorrectly treated as part of the message body, causing a verification
696 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
697 require looser validation rules than those for 821-level addresses,
698 which only permit IP addresses.
704 JH/01 Avoid costly startup code when not strictly needed. This reduces time
705 for some exim process initialisations. It does mean that the logging
706 of TLS configuration problems is only done for the daemon startup.
708 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
710 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
711 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
714 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
715 DKIM signing is being done. Previously a transport_filter would always
716 disable CHUNKING, falling back to traditional DATA.
718 JH/05 Regard command-line recipients as tainted.
720 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
722 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
723 PAM library frees one of the arguments given to it, despite the
724 documentation. Therefore a plain malloc must be used.
726 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
727 on-stack buffers were used, resulting in a taint trap when DSN information
728 copied from a received message was written into the buffer.
730 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
731 the ordering of its ARC headers. This caused a crash.
733 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
734 a new record was being constructed with information from the peer, a trap
737 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
738 installation would get error messages from DMARC verify, when it hit the
739 nonexistent file indicated by the default. Distros wanting DMARC enabled
740 should both provide the file and set the option.
741 Also enforce no DMARC verification for command-line sourced messages.
743 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
744 could, depending on the platform, hang at the STARTTLS response.
746 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
747 message on a connection. Previously if one message had ARC headers and
748 the following one did not, a crash could result when adding an
749 Authentication-Results: header.
751 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
752 Exim main code and Exim-related utities. The introduction of taint
753 tracking also did many adjustments to string handling. Since then, eximon
754 frequently terminated with an assert failure.
756 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
757 check for 452 responses. This slightly helps the inefficieny of doing
758 a large alias-expansion into a recipient-limited target. The max_rcpt
759 transport option still applies (and at the current default, will override
760 the new feature). The check is done for either cause of synch, and forces
761 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
762 connection. The new facility is not tunable at this time.
764 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
765 library live data was being used, so the results became garbage. Make
766 copies while it is still usable.
768 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
769 on delivery deferred (==) and failed (**) lines (if a delivery was
770 attemtped). Previously it was only on completion (=>) lines.
772 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
773 for the expansion of the server_scram_iter and server_scram_salt options.
775 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
776 are now specifically given a NO_DATA response without hitting the system
777 resolver. The library goes on to do the now-standard TXT lookup.
778 Use of dnsdb lookups is not affected.
780 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
781 only retrieve the errormessage once. Previously two calls to dlerror()
782 were used, and the second one (for mainlog/paniclog) retrieved null
785 JH/20 Taint checking: disallow use of tainted data for
786 - the appendfile transport file and directory options
787 - the pipe transport command
788 - the autoreply transport file, log and once options
789 - file names used by the redirect router (including filter files)
791 - paths used by single-key lookups
792 Previously this was permitted.
794 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
795 adjusted the size of a major service buffer; this failed because the
796 buffer was in use at the time. Change to a compile-time increase in the
797 buffer size, when this authenticator is compiled into exim.
799 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
800 previous fast-mode was untenable in the face of glibs using mmap to
801 support larger malloc requests.
803 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
804 New values supported, if defined on system where compiled:
805 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
806 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
808 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
809 running a limited number of proceses in parallel, a benefit is gained. The
810 amount varies with the platform hardware and load. The use of the option
811 queue_run_in_order means we cannot do this, as ordering becomes
814 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
815 had introduced a string-copy (for ensuring NUL-termination) which was not
816 appropriate for that case, which can include embedded NUL bytes in the
817 block of data. Investigation showed the copy to actually be needless, the
818 data being length-specified.
820 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
821 done during a receiving connection, and both used TLS, global info was
822 used rather than per-connection info for tracking the state of data
823 queued for transmission. This could result in a connection hang.
825 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
826 Previously, when delivering serveral messages down a single connection
827 only the first would provide a SIZE. This was due to the size information
828 not being properly tracked.
830 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
831 TAI (at 37 seconds currently), pretend to be in UTC for time-related
832 expansion and logging. Previously, spurious values such as a future
833 minute could be seen.
835 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
836 it could crash from a null-deref. This could also affect the
837 ${addresses: } operator and ${readsock } item.
839 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
840 message following a mime one, the variable was not reset.
842 JH/30 When an pipelined-connect fails at the first response, assume incorrect
843 cached capability (perhaps the peer reneged?) and immediately retry in
846 JH/31 Fix spurious detection of timeout while writing to transport filter.
848 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
849 an attempt to copy the string was made before checking it.
851 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
852 taint of the lookup key was maintained; we now regard the presence in the
853 filesystem as sufficient validation.
855 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
856 argument is supplied.
858 JH/35 The dsearch lookup now requires that the directory is an absolute path.
859 Previously this was not checked, and nonempty relative paths made an
860 access under Exim's current working directory.
862 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
863 Previously no event was raised.
865 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
866 parameter supplied by the sender MAIL FROM command. Previously it was
867 ignored, and only the check_spool_space option value for the required
870 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
871 the size of the signing public-key. Previously it was instead giving
872 the size of the signature hash.
874 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
875 the default. See the (new) dkim_verify_min_keysizes option.
877 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
878 and an ACL use a lookup for checking either the local_part or domain,
879 stale data could be accessed. Ensure that variable references are
880 dropped between messages.
882 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
883 by the client was not checked as pointing within response data before
884 being used. A malicious client could thus cause an out-of-bounds read and
885 possibly gain authentication. Fix by adding the check.
887 JH/42 Internationalisation: change the default for downconversion in the smtp
888 transport to be "if needed". Previously it was "as previously set" for
889 the message, which usually meant "if needed" for message-submission but
890 "no" for everything else. However, MTAs have been seen using SMTPUTF8
891 even when the envelope addresses did not need it, resulting in forwarding
892 failures to non-supporting MTAs. A downconvert in such cases will be
893 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
894 transport. The change does mean that addresses needing conversion will
895 be converted when previously a delivery failure would occur.
897 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
898 response was received it would be used unchecked in a fail-DSN, violating
899 standards on line-length limits. Truncate if needed.
901 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
902 doesn't work. (Additionally add a new main config option to configure the
909 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
910 side, to decode a TLS 1.3 packet capture.
912 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
913 Previously the default library behaviour applied, sending two, each in
916 JH/03 Debug output for ACL now gives the config file name and line number for
919 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
921 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
923 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
924 buffer overrun for (non-chunking) other transports.
926 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
927 TLS1.3, means that a server rejecting a client certificate is not visible
928 to the client until the first read of encrypted data (typically the
929 response to EHLO). Add detection for that case and treat it as a failed
930 TLS connection attempt, so that the normal retry-in-clear can work (if
931 suitably configured).
933 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
934 and/or domain. Found and fixed by Jason Betts.
936 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
937 configuration). If a CNAME target was not a wellformed name pattern, a
940 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
941 the OS reports them interleaved with other addresses.
943 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
944 used both for input and for a verify callout, both encrypted, SMTP
945 responses being sent by the server could be lost. This resulted in
946 dropped connections and sometimes bounces generated by a peer sending
949 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
950 string. Previously it was possible to cause undefined behaviour in a
951 library routine (usually a crash). Found by "zerons".
953 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
956 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
957 API was removed, so update to use the newer ones.
959 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
960 any timeout set, is taking a long time. Previously we would hang on to a
961 rotated logfile "forever" if the input was arriving with long gaps
962 (a previous attempt to fix addressed lack, for a long time, of initial
965 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
966 shared (NFS) environment. The length of the tempfile name is now
967 4 + 16 ("hdr.$message_exim_id") which might break on file
968 systems which restrict the file name length to lower values.
971 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
972 shared (NFS) environment.
974 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
975 did for all versions <4.90). Notably -M, -m, --invert, -I may be
978 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
979 on some platforms for bit 31.
981 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
982 to changes apparently associated with TLS1.3 handling some of the APIs
983 previously used were either nonfunctional or inappropriate. Strings
984 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
985 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
986 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
987 This affects log line X= elements, the $tls_{in,out}_cipher variables,
988 and the use of specific cipher names in the encrypted= ACL condition.
990 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
992 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
993 verification result was not updated unless hosts_require_ocsp applied.
995 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
996 queue_list_requires_admin set to false, non-admin users were denied the
999 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1000 directory-of-certs mode. Previously they were advertised despite the
1003 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1004 A single TCP connection by a client will now hold a TLS connection open
1005 for multiple message deliveries, by default. Previously the default was to
1008 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1009 default. If built with the facility, DANE will be used. The facility
1010 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1012 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1013 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1014 must be defined and you must still, unless you define DISABLE_TLS, manage
1015 the the include-dir and library-file requirements that go with that
1016 choice. Non-TLS builds are still supported.
1018 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1021 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1022 default. If the platform supports and has the facility enabled, it will
1023 be requested on all coneections.
1025 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1026 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1028 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1030 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1031 success-DSN messages. Previously the From: header was always the default
1032 one for these; the option was ignored.
1034 JH/28 Fix the timeout on smtp response to apply to the whole response.
1035 Previously it was reset for every read, so a teergrubing peer sending
1036 single bytes within the time limit could extend the connection for a
1037 long time. Credit to Qualsys Security Advisory Team for the discovery.
1039 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1040 delivery address, which leaked information of the results of local
1041 forwarding. Change to the original envelope recipient address, per
1044 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1045 requested. Previously not bounce was generated and a log entry of
1046 error ignored was made.
1048 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1050 JH/32 Introduce a general tainting mechanism for values read from the input
1051 channel, and values derived from them. Refuse to expand any tainted
1052 values, to catch one form of exploit.
1054 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1055 was unused and the unexpanded text used for the test. Found and
1056 fixed by Ruben Jenster.
1058 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1059 an attempt to use a TLS library read routine dereffed a nul pointer,
1062 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1063 them in our smtp response.
1065 JH/36 Have the general router option retry_use_local_part default to true when
1066 any of the restrictive preconditions are set (to anything). Previously it
1067 was only for check_local user. The change removes one item of manual
1068 configuration which is required for proper retries when a remote router
1069 handles a subset of addresses for a domain.
1071 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1072 link count into consideration.
1074 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1075 caused the extension of big_buffer, the following lines were ignored.
1077 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1078 accordance with RFC 2308. Previously there was no expiry, so a longlived
1079 receive process (eg. due to ACL delays) versus a short SOA value could
1082 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1084 JH/39 Promote DMARC support to mainline.
1086 JH/40 Bug 2452: Add a References: header to DSNs.
1088 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1089 parameters. The relevant library call is documented as "Deprecated: This
1090 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1091 3.6.0, DH parameters are negotiated following RFC7919."
1093 HS/06 Change the default of dnssec_request_domains to "*"
1095 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1096 carried on and emitted a BDAT command, even when PIPELINING was not
1099 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1100 buffer was used for the filename, resulting in a trap when tainted
1101 arguments (eg. $domain) were used.
1103 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1104 recommended to avoid a possible server-load attack. The feature can be
1105 re-enabled via the openssl_options main cofiguration option.
1107 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1108 for version 4.90 - adding a "more data" boolean to the arguments.
1109 Bumped the ABI version number also, this having been missed previously;
1110 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1111 respect of usage of smtp_printf() by either local_scan code or libraries
1112 accessed via the ${dlfunc } expansion item. Both will need coding
1113 adjustment for any calls to smtp_printf() to match the new function
1114 signature; a FALSE value for the new argument is always safe.
1116 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1117 the file-offset (which the Linux syscall does, and exim expects); this
1118 resulted in an indefinite loop.
1120 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1121 to do ARC verification. The Authentication-Results: header line added
1122 by the configuration then had no ARC item.
1124 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1125 an incorrect mode was used for reading the result, resulting in it being
1132 JH/01 Remove code calling the customisable local_scan function, unless a new
1133 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1135 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1136 non-signal-safe functions being used.
1138 JH/03 Bug 2269: When presented with a received message having a stupidly large
1139 number of DKIM-Signature headers, disable DKIM verification to avoid
1140 a resource-consumption attack. The limit is set at twenty.
1142 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1143 report of oldest_pass in ${authres } in consequence, and separate out
1144 some descriptions of reasons for verification fail.
1146 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1147 files in the spool were present and unlocked. A queue-runner could spot
1148 them, resulting in a duplicate delivery. Fix that by doing the unlock
1149 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1150 add more error-checking on spoolfile handling while that code is being
1153 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1154 No known attacks, no CVE, this is defensive hardening.
1156 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1157 a queue-runner could start a delivery while other operations were ongoing.
1158 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1159 Found and investigated by Tim Stewart. Fix by using the open message data
1160 file handle rather than opening another, and not locally closing it (which
1161 releases a lock) for that case, while creating the temporary .eml format
1162 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1164 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1165 $sender_verify_failure/$recipient_verify_failure to "random".
1167 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1170 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1171 Previously this would segfault.
1173 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1176 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1177 like zero, since the resolver should be doing this for us, But we need one
1178 as a CNAME but no MX presence gets the CNAME returned; we need to check
1179 that doesn't point to an MX to declare it "no MX returned" rather than
1180 "error, loop". A new main option is added so the older capability of
1181 following some limited number of chain links is maintained.
1183 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1185 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1186 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1187 now-deprecated earlier definitions used only specified the range up to TLS
1188 1.2 (in the older-version library docs).
1190 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1192 JH/15 Rework TLS client-side context management. Stop using a global, and
1193 explicitly pass a context around. This enables future use of TLS for
1194 connections to service-daemons (eg. malware scanning) while a client smtp
1195 connection is using TLS; with cutthrough connections this is quite likely.
1197 JH/16 Fix ARC verification to do AS checks in reverse order.
1199 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1201 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1202 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1203 Previously the "utf8" would be re-prepended for every additional message.
1205 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1206 Previously thery were accepted, resulting in issues when attempting to
1207 forward messages to a non-supporting MTA.
1209 PP/02 Let -n work with printing macros too, not just options.
1211 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1212 one parent address was copied, and bogus data was used at delivery-logging
1213 time. Either a crash (after delivery) or bogus log data could result.
1214 Discovery and analysis by Tim Stewart.
1216 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1217 Previously if the string ended mid-character, we did not insert the
1218 promised '?' replacement.
1220 PP/04 Documentation: current string operators work on bytes, not codepoints.
1222 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1223 should pack well giving a smaller memory footprint so better caching and
1224 therefore performance. Group the declarations where this can't be done so
1225 that the byte-sized flag variables are not interspersed among pointer
1226 variables, giving a better chance of good packing by the compiler.
1228 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1229 non-null, to avoid issues with sites running BATV. Previously reports were
1230 sent with an empty envelope sender so looked like bounces.
1232 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1233 The ignore_error flag wasn't being returned from the filter subprocess so
1234 was not set for later routers. Investigation and fix by Matthias Kurz.
1236 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1237 and a msg:complete for the whole, when a message is manually removed using
1238 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1240 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1241 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1242 hope that is portable enough.
1244 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1245 requiring a known-CA anchor certificate; make it now rely entirely on the
1246 TLSA as an anchor. Checking the name on the leaf cert against the name
1247 on the A-record for the host is still done for TA (but not for EE mode).
1249 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1250 included in delivery lines for non-proxied connections, when compiled with
1251 SUPPORT_SOCKS and running with proxy logging enabled.
1253 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1254 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1255 move the existing event to fire before the normal logging of message
1256 failure so that custom logging is bracketed by normal logging.
1258 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1259 msg:fail:internal event. Developement by Matthias Kurz.
1261 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1262 far too small for todays use of crypto signatures stored there. Go all
1263 the way to the max DNS message size of 64kB, even though this might be
1264 overmuch for IOT constrained device use.
1266 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1267 copy a string over itself. The library routine is documented as not
1268 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1270 JH/32 For main options check_spool_space and check_inode_space, where the
1271 platform supports 64b integers, support more than the previous 2^31 kB
1272 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1273 the previous G, M, k.
1275 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1276 $authenticated_fail_id variable on authentication failure. Previously
1279 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1280 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1281 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1282 GNUTLS_SEC_PARAM_MEDIUM.
1284 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1285 side. Previously we would continue as if no SNI had been received.
1287 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1288 "<" character, which should be a list-separator specification, we walked
1289 off past the nul-terimation.
1291 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1292 causes) even when the retry time is not yet met. Previously they were
1293 not, meaning that when (say) an account was over-quota and temp-rejecting,
1294 and multiple senders' messages were queued, only one sender would get
1295 notified on each configured delay_warning cycle.
1297 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1299 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1300 cipher-suites, an error can be left on the stack even for a succeeding
1301 accept; this results in impossible error messages when a later operation
1304 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1305 return error codes indicating retry. Under TLS1.3 this becomes required.
1307 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1308 it only wrote the new authenticators, resulting in a lack of tracking of
1309 peer changes of ESMTP extensions until the next cache flush.
1311 JH/41 Fix the loop reading a message header line to check for integer overflow,
1312 and more-often against header_maxsize. Previously a crafted message could
1313 induce a crash of the recive process; now the message is cleanly rejected.
1315 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1316 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1322 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1323 When redis_servers is set to a list of > 1 element, and the Redis servers
1324 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1325 case of MOVED into a DEFER case instead, thus moving the query onto the
1326 next server in the list. For a cluster of N elements, all N servers must
1327 be defined in redis_servers.
1329 GF/02 Catch and remove uninitialized value warning in exiqsumm
1330 Check for existence of @ARGV before looking at $ARGV[0]
1332 JH/01 Replace the store_release() internal interface with store_newblock(),
1333 which internalises the check required to safely use the old one, plus
1334 the allocate and data copy operations duplicated in both (!) of the
1335 extant use locations.
1337 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1338 modifier. This matches the restriction on the commandline.
1340 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1341 Previously only the last row was returned.
1343 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1344 we assumed that tags in the header were well-formed, and parsed the
1345 element content after inspecting only the first char of the tag.
1346 Assumptions at that stage could crash the receive process on malformed
1349 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1350 While running the DKIM ACL we operate on the Permanent memory pool so that
1351 variables created with "set" persist to the DATA ACL. Also (at any time)
1352 DNS lookups that fail create cache records using the Permanent pool. But
1353 expansions release any allocations made on the current pool - so a dnsdb
1354 lookup expansion done in the DKIM ACL releases the memory used for the
1355 DNS negative-cache, and bad things result. Solution is to switch to the
1356 Main pool for expansions.
1357 While we're in that code, add checks on the DNS cache during store_reset,
1358 active in the testsuite.
1359 Problem spotted, and debugging aided, by Wolfgang Breyha.
1361 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1362 When none of the hosts presented to a transport match an already-open
1363 connection, close it and proceed with the list. Previously we would
1364 queue the message. Spotted by Lena with Yahoo, probably involving
1367 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1368 Previously a spurious "250 OK id=" response was appended to the proper
1371 JH/08 The "support for" informational output now, which built with Content
1372 Scanning support, has a line for the malware scanner interfaces compiled
1373 in. Interface can be individually included or not at build time.
1375 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1376 by the template makefile "src/EDITME". The "STREAM" support for an older
1377 ClamAV interface method is removed.
1379 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1380 rows affected is given instead).
1382 JH/11 The runtime Berkeley DB library version is now additionally output by
1383 "exim -d -bV". Previously only the compile-time version was shown.
1385 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1386 SMTP connection. Previously, when one had more recipients than the
1387 first, an abortive onward connection was made. Move to full support for
1388 multiple onward connections in sequence, handling cutthrough connection
1389 for all multi-message initiating connections.
1391 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1392 routers. Previously, a multi-recipient message would fail to match the
1393 onward-connection opened for the first recipient, and cause its closure.
1395 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1396 a timeout on read on a GnuTLS initiating connection, resulting in the
1397 initiating connection being dropped. This mattered most when the callout
1398 was marked defer_ok. Fix to keep the two timeout-detection methods
1401 JH/15 Relax results from ACL control request to enable cutthrough, in
1402 unsupported situations, from error to silently (except under debug)
1403 ignoring. This covers use with PRDR, frozen messages, queue-only and
1406 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1408 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1409 metadata, resulting in a crash in free().
1411 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1412 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1413 Broken also in d185889f4, with init system revamp.
1415 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1416 Previously we abruptly closed the connection after reading a malware-
1417 found indication; now we go on to read the "scan ok" response line,
1420 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1421 ACL. Previously, a crash would result.
1423 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1424 macro text after a replacement (previously it was only once per line) and
1425 by skipping builtin macros when searching for an uppercase lead character.
1427 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1428 for the build is renamed.
1430 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1431 was allocated for every new TLS startup, meaning one per message. Fix
1432 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1434 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1435 reported the original. Fix to report (as far as possible) the ACL
1436 result replacing the original.
1438 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1439 OpenSSL. Certificate information is loaded for every new TLS startup,
1440 and the resources needed to be freed.
1442 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1444 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1447 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1448 DATA response info to the (existing) per-recipient response info for
1449 the "C=" log element. It can have useful tracking info from the
1450 destination system. Patch from Simon Arlott.
1452 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1453 length value. Previously this would segfault.
1455 HS/02 Support Avast multiline protoocol, this allows passing flags to
1456 newer versions of the scanner.
1458 JH/28 Ensure that variables possibly set during message acceptance are marked
1459 dead before release of memory in the daemon loop. This stops complaints
1460 about them when the debug_store option is enabled. Discovered specifically
1461 for sender_rate_period, but applies to a whole set of variables.
1462 Do the same for the queue-runner and queue-list loops, for variables set
1463 from spool message files. Do the same for the SMTP per-message loop, for
1464 certain variables indirectly set in ACL operations.
1466 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1467 as a multi-recipient message from a mailinglist manager). The coding had
1468 an arbitrary cutoff number of characters while checking for more input;
1469 enforced by writing a NUL into the buffer. This corrupted long / fast
1470 input. The problem was exposed more widely when more pipelineing of SMTP
1471 responses was introduced, and one Exim system was feeding another.
1472 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1473 receiving system, and refused recipients seen by the sending system
1474 (propating to people being dropped from mailing lists).
1475 Discovered and pinpointed by David Carter.
1477 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1478 replaced by the ${authresults } expansion.
1480 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1482 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1483 allows proper process termination in container environments.
1485 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1486 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1488 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1489 and "err_perm", deprecated since 4.83 when the RFC-defined words
1490 "temperror" and "permerror" were introduced.
1492 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1493 transport-filters or DKIM-signing. The restriction was lost in the
1494 consolidation of verify-callout and delivery SMTP handling.
1495 Extend the restriction to also cover ARC-signing.
1497 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1498 in defer=pass mode supply a 450 to the initiator. Previously the message
1501 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1502 tls_require_ciphers is used as before.
1504 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1505 "pass_unscanned". Only tmpfails from the scanner are written to
1506 the paniclog, as they may require admin intervention (permission
1507 denied, license issues). Other scanner errors (like decompression
1508 bombs) do not cause a paniclog entry.
1510 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1511 Previously it was possible to log spurious information in receive log
1514 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1515 triggered odd behaviour from Outlook Express clients.
1517 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1520 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1521 since the IETF WG has not yet settled on that versus the original
1522 "bare" representation.
1524 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1525 Previously the millisecond value corrupted the output.
1526 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1527 corrupted the output.
1533 JH/01 Rework error string handling in TLS interface so that the caller in
1534 more cases is responsible for logging. This permits library-sourced
1535 string to be attached to addresses during delivery, and collapses
1536 pairs of long lines into single ones.
1538 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1539 during configuration. Wildcards are allowed and expanded.
1541 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1542 This permits better logging.
1544 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1545 do not close it down and have the child transport start it up again on
1546 the passed-on TCP connection. Instead, proxy the child (and any
1547 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1548 affected: the continued delivery log lines do not have any DNSSEC, TLS
1549 Certificate or OCSP information. TLS cipher information is still logged.
1551 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1552 identical IP addresses on different listening ports. Will also affect
1555 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1556 add noisy ifdef guards to special-case this sillyness.
1557 Patch from Bernd Kuhls.
1559 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1560 than 255 are no longer allowed.
1562 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1563 Disable the session-cache too, which might reduce our load. Since we
1564 currrectly use a new context for every connection, both as server and
1565 client, there is no benefit for these.
1566 GnuTLS appears to not support tickets server-side by default (we don't
1567 call gnutls_session_ticket_enable_server()) but client side is enabled
1568 by default on recent versions (3.1.3 +) unless the PFS priority string
1571 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1572 <https://reproducible-builds.org/specs/source-date-epoch/>.
1574 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1575 the check for any unsuccessful recipients did not notice the limit, and
1576 erroneously found still-pending ones.
1578 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1579 MSG_MORE. Only in-clear (not on TLS connections).
1581 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1582 transport-filter is involved we can buffer the headers in memory for
1583 creating the signature, and read the spool data file once for the
1584 signature and again for transmission.
1586 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1587 4.77 as the kernel support then wasn't solid, having issues in 64bit
1588 mode. Now, it's been long enough. Add support for FreeBSD also.
1590 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1591 case where the routing stage had gathered several addresses to send to
1592 a host before calling the transport for the first, we previously failed
1593 to close down TLS in the old transport process before passing the TCP
1594 connection to the new process. The new one sent a STARTTLS command
1595 which naturally failed, giving a failed delivery and bloating the retry
1596 database. Investigation and fix prototype from Wolfgang Breyha.
1598 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1599 false-negatives in the check that the sender had not preempted a response
1600 or prompt from Exim (running as a server), due to that code's lack of
1601 awareness of the SMTP input buffering.
1603 PP/04 Add commandline_checks_require_admin option.
1604 Exim drops privileges sanely, various checks such as -be aren't a
1605 security problem, as long as you trust local users with access to their
1606 own account. When invoked by services which pass untrusted data to
1607 Exim, this might be an issue. Set this option in main configuration
1608 AND make fixes to the calling application, such as using `--` to stop
1611 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1612 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1613 responses to those, into a single TLS record each way (this usually means
1614 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1617 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1618 affects you only if you're dancing at the edge of the param size limits.
1619 If you are, and this message makes sense to you, then: raise the
1620 configured limit or use OpenSSL 1.1. Nothing we can do for older
1623 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1624 cmdline element to get the documented default one. Previously it was
1627 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1628 are unsuable for non-message-specific reasons - eg. connection timeout,
1629 banner-time rejection.
1631 JH/16 Fix logging of delivery remote port, when specified by router, under
1634 PP/06 Repair manualroute's ability to take options in any order, even if one
1635 is the name of a transport.
1638 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1640 JH/17 Change the list-building routines interface to use the expanding-string
1641 triplet model, for better allocation and copying behaviour.
1643 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1644 Previously it was constructed the first time a possibly-matching string
1645 was met in the configuration file input during startup; now it is done
1648 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1649 compatible one, to avoid the (poorly documented) possibility of a config
1650 file in the working directory redirecting the DB files, possibly correpting
1651 some existing file. CVE-2017-10140 assigned for BDB.
1653 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1654 cache-hot. Previously, although the result was properly cached, the
1655 initial verify call returned a defer.
1657 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1658 the main verify for receipient in uncached-mode.
1660 JH/22 Retire historical build files to an "unsupported" subdir. These are
1661 defined as "ones for which we have no current evidence of testing".
1663 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1664 if present. Previously it was ignored.
1666 JH/24 Start using specified-initialisers in C structure init coding. This is
1667 a C99 feature (it's 2017, so now considered safe).
1669 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1670 if was a fixed-sized field and bitmask ops via macros; it is now more
1673 PP/07 GitHub PR 56: Apply MariaDB build fix.
1674 Patch provided by Jaroslav Škarvada.
1676 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1677 during Coverity cleanups [4.87 JH/47]
1678 Diagnosis and fix provided by Michael Fischer v. Mollard.
1680 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1681 the right size to place the terminating semicolon on its own folded
1682 line, the header hash was calculated to an incorrect value thanks to
1683 the (relaxed) space the fold became.
1685 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1686 and confused the parent.
1688 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1689 which could crash as a result. This could lead to undeliverable messages.
1691 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1694 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1695 matters on fast-turnover and PID-randomising systems, which were getting
1696 out-of-order delivery.
1698 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1699 a possibly-overlapping copy. The symptom was that "Remote host closed
1700 connection in response to HELO" was logged instead of the actual 4xx
1703 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1704 Previously only that bufferd was discarded, resulting in SYMTP command
1707 JH/32 DKIM: when a message has multiple signatures matching an identity given
1708 in dkim_verify_signers, run the dkim acl once for each. Previously only
1709 one run was done. Bug 2189.
1711 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1712 "panic the current process" to "deliberately defer". The panic log is
1713 still written with the problem list name; the mail and reject logs now
1714 get a temp-reject line for the message that was being handled, saying
1715 something like "domains check lookup or other defer". The SMTP 451
1716 message is still "Temporary local problem".
1718 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1719 A crafted sequence of BDAT commands could result in in-use memory beeing
1720 freed. CVE-2017-16943.
1722 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1723 from SMTP input. Previously it was always done; now only done for DATA
1724 and not BDAT commands. CVE-2017-16944.
1726 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1727 to the message (such as an overlong header line). Previously this was
1728 not done and we did not exit BDAT mode. Followon from the previous item
1729 though a different problem.
1735 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1736 than -2003 did; needs libidn2 in addition to libidn.
1738 JH/02 The path option on a pipe transport is now expanded before use.
1740 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1741 Patch provided by "Björn", documentation fix added too.
1743 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1744 missing a wire-to-host endian conversion.
1746 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1747 close after a BDAT command line could be taken as a following command,
1748 giving a synch failure. Fix by only checking for synch immediately
1749 before acknowledging the chunk.
1751 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1752 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1753 macro. Patches provided by Josh Soref.
1755 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1756 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1757 by IANA as a service extension. However, John Klensin suggests that we
1760 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1761 the dkim code may be unix-mode line endings rather than smtp wire-format
1762 CRLF, so prepend a CR to any bare LF.
1764 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1765 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1767 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1768 the first were themselves being wrongly included in the feed into dkim
1769 processing; with most chunk sizes in use this resulted in an incorrect
1770 body hash calculated value.
1772 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1773 DKIM signature block, for verification. Although advised against by
1774 standards it is specifically not ruled illegal.
1776 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1778 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1779 missing a body hash (the bh= tag).
1781 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1782 It seems that HAProxy sends the Proxy Protocol information in clear and
1783 only then does a TLS startup, so do the same.
1785 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1786 TCP connections (such as for Spamd) unless the daemon successfully set
1787 Fast Open mode on its listening sockets. This fixes breakage seen on
1788 too-old kernels or those not configured for Fast Open, at the cost of
1789 requiring both directions being enabled for TFO, and TFO never being used
1790 by non-daemon-related Exim processes.
1792 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1793 endings, at least on the first header line. Try to canonify any that get
1794 past that check, despite the cost.
1796 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1797 now limited to an arbitrary five deep, while parsing addresses with the
1798 strip_excess_angle_brackets option enabled.
1800 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1801 instead leave the unprompted TLS handshake in socket buffer for the
1802 TLS library to consume.
1804 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1806 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1808 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1809 region in one operation, for speed, and this leaves assigned pointers
1810 dangling. Add checks run only under the testsuite which checks all
1811 variables at a store-reset and panics on a dangling pointer; add code
1812 explicitly nulling out all the variables discovered. Fixes one known
1813 bug: a transport crash, where a dangling pointer for $sending_ip_address
1814 originally assigned in a verify callout, is re-used.
1816 PP/06 Drop '.' from @INC in various Perl scripts.
1818 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1820 PP/08 Reduce a number of compilation warnings under clang; building with
1821 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1822 should be warning-free.
1824 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1826 HS/01 Fix portability problems introduced by PP/08 for platforms where
1827 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1829 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1830 chunk. This allows us to accept broken chunked messages. We need a more
1831 general solution here.
1833 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1834 already-broken messages in the queue.
1836 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1838 JH/19 Fix reference counting bug in routing-generated-address tracking.
1844 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1845 supports it and a size is available (ie. the sending peer gave us one).
1847 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1848 years of being deprecated). The replacements are the ACLs
1849 acl_smtp_mime and acl_not_smtp_mime.
1851 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1852 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1853 or even an in-clear connection were permitted. Now, if the host lookup
1854 was dnssec and dane was requested then the host is only used if the
1855 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1856 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1857 if one fails this test.
1858 This means that a poorly-configured remote DNS will make it incommunicado;
1859 but it protects against a DNS-interception attack on it.
1861 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1862 non-noisy when a race steals the message being considered.
1864 JH/05 If main configuration option tls_certificate is unset, generate a
1865 self-signed certificate for inbound TLS connections.
1867 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1868 in rewrites and routers.
1870 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1871 and logged a warning sing 4.83; now they are a configuration file error.
1873 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1874 (lacking @domain). Apply the same qualification processing as RCPT.
1876 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1878 JH/10 Support ${sha256:} applied to a string (as well as the previous
1881 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1882 a cutthrough deliver is pending, as we always want to make a connection.
1883 This also avoids re-routing the message when later placing the cutthrough
1884 connection after a verify cache hit.
1885 Do not update it with the verify result either.
1887 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1888 when routing results in more than one destination address.
1890 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1891 signing (which inhibits the cutthrough capability). Previously only
1892 the presence of an option was tested; now an expansion evaluating as
1893 empty is permissible (obviously it should depend only on data available
1894 when the cutthrough connection is made).
1896 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1897 the relevant preceding SMTP command did not note the pipelining mode.
1899 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1900 Previously they were not counted.
1902 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1903 as one having no matching records. Previously we deferred the message
1904 that needed the lookup.
1906 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1907 distinguished as "(=".
1909 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1910 for missing MX records. Previously it only worked for missing A records.
1912 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1914 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1915 after the data-go-ahead and data-ack. Patch from Jason Betts.
1917 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1918 even for a "none" policy. Patch from Tony Meyer.
1920 JH/22 Fix continued use of a connection for further deliveries. If a port was
1921 specified by a router, it must also match for the delivery to be
1924 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1925 When one of the recipients of a message was unsuitable for the connection
1926 (has no matching addresses), we lost track of needing to mark it
1927 deferred. As a result mail would be lost.
1929 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1931 JH/25 Decoding ACL controls is now done using a binary search; the source code
1932 takes up less space and should be simpler to maintain. Merge the ACL
1933 condition decode tables also, with similar effect.
1935 JH/26 Fix problem with one_time used on a redirect router which returned the
1936 parent address unchanged. A retry would see the parent address marked as
1937 delivered, so not attempt the (identical) child. As a result mail would
1940 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1941 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1942 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1945 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1946 Main config options check_{log,spool}_{inodes,space} are now
1947 100 inodes, 10MB unless set otherwise in the configuration.
1949 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1950 Previously it only applied to the main-section connection policy
1953 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1955 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1956 by me. Added RFC7919 DH primes as an alternative.
1958 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1959 are not in the system include path.
1961 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1962 GnuTLS, when a session startup failed (eg because the client disconnected)
1963 Exim did stdio operations after fclose. This was exposed by a recent
1964 change which nulled out the file handle after the fclose.
1966 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1967 signed directly by the cert-signing cert, rather than an intermediate
1968 OCSP-signing cert. This is the model used by LetsEncrypt.
1970 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1972 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1973 an incoming connection.
1975 HS/02 Bug 1802: Do not half-close the connection after sending a request
1978 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1979 fallback to "prime256v1".
1981 JH/34 SECURITY: Use proper copy of DATA command in error message.
1982 Could leak key material. Remotely exploitable. CVE-2016-9963.
1988 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1989 and 3.4.4 - once the server is enabled to respond to an OCSP request
1990 it does even when not requested, resulting in a stapling non-aware
1991 client dropping the TLS connection.
1993 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1994 support variable-length bit vectors. No functional change.
1996 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1997 The I= interface field on outgoing lines is now after the H= remote
1998 host field, same as incoming lines. There is a separate
1999 outgoing_interface log selector which allows you to disable the
2002 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2003 If not running log_selector +smtp_connection the mainlog would be held
2004 open indefinitely after a "too many connections" event, including to a
2005 deleted file after a log rotate. Leave the per net connection logging
2006 leaving it open for efficiency as that will be quickly detected by the
2007 check on the next write.
2009 HS/01 Bug 1671: Fix post transport crash.
2010 Processing the wait-<transport> messages could crash the delivery
2011 process if the message IDs didn't exist for some reason. When
2012 using 'split_spool_directory=yes' the construction of the spool
2013 file name failed already, exposing the same netto behaviour.
2015 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2016 mime_regex ACL conditions.
2018 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2019 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2020 to HELO, local diagnostic string.
2022 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2023 log line to debug. Even when configured with a tls authenticator many
2024 client connections are expected to not authenticate in this way, so
2025 an authenticate fail is not an error.
2027 HS/02 Add the Exim version string to the process info. This way exiwhat
2028 gives some more detail about the running daemon.
2030 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2031 matter for fast-change records such as DNSBLs.
2033 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2034 retry record, even if constant. There may be multiple transports with
2035 different interface settings and the retry behaviour needs to be kept
2038 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2040 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2042 JH/10 Bug 840: fix log_defer_output option of pipe transport
2044 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2045 hosts_randomize. This matters a lot when combined with mua_wrapper.
2047 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2048 ${quote_pgsql:<string>} operator.
2050 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2052 JH/14 Change header file name for freeradius-client. Relevant if compiling
2053 with Radius support; from the Gentoo tree and checked under Fedora.
2055 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2057 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2058 option transport_filter, meaning no filtering.
2060 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2062 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2063 defaults to "*" (all hosts). The variable is now available when not built
2064 with TLS, default unset, mainly to enable keeping the testsuite sane.
2065 If a server certificate is not supplied (via tls_certificate) an error is
2066 logged, and clients will find TLS connections fail on startup. Presumably
2067 they will retry in-clear.
2068 Packagers of Exim are strongly encouraged to create a server certificate
2069 at installation time.
2071 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2072 with the $config_file variable.
2074 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2075 in transport context, after the attempt, and per-recipient. The latter type
2076 is per host attempted. The event data is the error message, and the errno
2077 information encodes the lookup type (A vs. MX) used for the (first) host,
2078 and the trailing two digits of the smtp 4xx response.
2080 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2081 to write to mainlog (or rejectlog, paniclog) in the window between file
2082 creation and permissions/ownership being changed. Particularly affects
2083 installations where exicyclog is run as root, rather than exim user;
2084 result is that the running daemon panics and dies.
2086 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2088 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2089 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2090 "pri" and "weight". Note that the previous implicit priority given by the
2091 list order is no longer honoured.
2093 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2094 for DKIM processing.
2096 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2097 by defining SUPPORT_SOCKS.
2099 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2100 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2101 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2102 variables are renamed to proxy_{local,external}_{address,port}.
2104 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2105 for a build by defining SUPPORT_I18N
2107 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2108 of the query string, and make ${quote_redis:} do that quoting.
2110 JH/29 Move Events support from Experimental to mainline, enabled by default
2111 and removable for a build by defining DISABLE_EVENT.
2113 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2115 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2116 cached by the daemon.
2118 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2119 by defining LOOKUP_REDIS. The libhiredis library is required.
2121 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2122 keys are given for lookup.
2124 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2125 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2126 only supported when built with TLS support. The PolarSSL SHA routines
2127 are still used when the TLS library is too old for convenient support.
2129 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2130 openssl_options), for security. OpenSSL forces this from version 1.1.0
2131 server-side so match that on older versions.
2133 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2134 allocation for $value could be released as the expansion processing
2135 concluded, but leaving the global pointer active for it.
2137 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2138 and to use the domains and local_parts ACL conditions.
2140 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2141 incorrectly not doubled on cutthrough transmission, hence seen as a
2142 body-termination at the receiving system - resulting in truncated mails.
2143 Commonly the sender saw a TCP-level error, and retransmitted the message
2144 via the normal store-and-forward channel. This could result in duplicates
2145 received - but deduplicating mailstores were liable to retain only the
2146 initial truncated version.
2148 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2150 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2152 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2153 we're in there, support oversigning also; bug 1309.
2155 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2157 HS/04 Add support for keep_environment and add_environment options.
2159 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2160 either intentional arithmetic overflow during PRNG, or testing config-
2163 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2164 delivery resulted in actual delivery. Cancel cutthrough before DATA
2167 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2168 rejects a recipient: pass the reject to the originator.
2170 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2171 Many were false-positives and ignorable, but it's worth fixing the
2174 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2175 for the new environment-manipulation done at startup. Move the routines
2176 from being local to tls.c to being global via the os.c file.
2178 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2179 an extract embedded as result-arg for a map, the first arg for extract
2180 is unavailable so we cannot tell if this is a numbered or keyed
2181 extraction. Accept either.
2187 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2190 JH/02 The smtp transport option "multi_domain" is now expanded.
2192 JH/03 The smtp transport now requests PRDR by default, if the server offers
2195 JH/04 Certificate name checking on server certificates, when exim is a client,
2196 is now done by default. The transport option tls_verify_cert_hostnames
2197 can be used to disable this per-host. The build option
2198 EXPERIMENTAL_CERTNAMES is withdrawn.
2200 JH/05 The value of the tls_verify_certificates smtp transport and main options
2201 default to the word "system" to access the system default CA bundle.
2202 For GnuTLS, only version 3.0.20 or later.
2204 JH/06 Verification of the server certificate for a TLS connection is now tried
2205 (but not required) by default. The verification status is now logged by
2206 default, for both outbound TLS and client-certificate supplying inbound
2209 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2212 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2213 Status Notification (bounce) messages are now MIME format per RFC 3464.
2214 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2215 under the control of the dsn_advertise_hosts option, and routers may
2216 have a dsn_lasthop option.
2218 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2219 default, modifiable by a malware= option. The list separator for
2220 the options can now be changed in the usual way. Bug 68.
2222 JH/10 The smtp_receive_timeout main option is now expanded before use.
2224 JH/11 The incoming_interface log option now also enables logging of the
2225 local interface on delivery outgoing connections.
2227 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2228 if the interface and destination host and port all match.
2230 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2233 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2234 Patch from Andrew Lewis.
2236 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2237 now supports optional time-restrictions, weighting, and priority
2238 modifiers per server. Patch originally by <rommer@active.by>.
2240 JH/16 The spamd_address main option now supports a mixed list of local
2241 and remote servers. Remote servers can be IPv6 addresses, and
2242 specify a port-range.
2244 JH/17 Bug 68: The spamd_address main option now supports an optional
2245 timeout value per server.
2247 JH/18 Bug 1581: Router and transport options headers_add/remove can
2248 now have the list separator specified.
2250 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2253 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2256 JH/21 Support for the A6 type of dns record is withdrawn.
2258 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2259 rather than the verbs used.
2261 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2262 from 255 to 1024 chars.
2264 JH/24 Verification callouts now attempt to use TLS by default.
2266 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2267 are generic router options now. The defaults didn't change.
2269 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2270 Original patch from Alexander Shikoff, worked over by JH.
2272 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2273 files if ZCAT_COMMAND is not executable.
2275 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2277 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2279 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2280 Normally benign, it bites when the pair was led to by a CNAME;
2281 modern usage is to not canonicalize the domain to a CNAME target
2282 (and we were inconsistent anyway for A-only vs AAAA+A).
2284 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2286 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2287 when evaluating $sender_host_dnssec.
2289 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2290 $sender_helo_dnssec variable.
2292 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2294 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2296 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2298 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2299 documented as working, but never had. Support all but $spam_report.
2301 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2302 added for tls authenticator.
2304 HS/03 Add perl_taintmode main config option
2310 TL/01 When running the test suite, the README says that variables such as
2311 no_msglog_check are global and can be placed anywhere in a specific
2312 test's script, however it was observed that placement needed to be near
2313 the beginning for it to behave that way. Changed the runtest perl
2314 script to read through the entire script once to detect and set these
2315 variables, reset to the beginning of the script, and then run through
2316 the script parsing/test process like normal.
2318 TL/02 The BSD's have an arc4random API. One of the functions to induce
2319 adding randomness was arc4random_stir(), but it has been removed in
2320 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2321 function when detected.
2323 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2324 cause callback expansion.
2326 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2327 syntax errors in an expansion can be treated as a string instead of
2328 logging or causing an error, due to the internal use of bool_lax
2329 instead of bool when processing it.
2331 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2332 server certificates when making smtp deliveries.
2334 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2336 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2338 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2340 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2341 Merged patch from Sebastian Wiedenroth.
2343 JH/05 Fix results-pipe from transport process. Several recipients, combined
2344 with certificate use, exposed issues where response data items split
2345 over buffer boundaries were not parsed properly. This eventually
2346 resulted in duplicates being sent. This issue only became common enough
2347 to notice due to the introduction of connection certificate information,
2348 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2350 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2351 size buffer was used, resulting in syntax errors when an expansion
2354 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2355 version 3.3.6 or later.
2357 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2358 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2359 both become "event_action", the variables become $event_name, $event_data
2360 and $event_defer_errno. There is a new variable $verify_mode, usable in
2361 routers, transports and related events. The tls:cert event is now also
2362 raised for inbound connections, if the main configuration event_action
2365 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2366 early OCSP support, but no stapling (appears to be less than 1.0.0).
2368 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2369 server certificate names available under the smtp transport option
2370 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2373 JH/10 Time-related extraction expansions from certificates now use the main
2374 option "timezone" setting for output formatting, and are consistent
2375 between OpenSSL and GnuTLS compilations. Bug 1541.
2377 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2378 encoded parameter in the incoming message. Bug 1558.
2380 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2381 include certificate info, eximon was claiming there were spoolfile
2384 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2386 JH/14 Log delivery-related information more consistently, using the sequence
2387 "H=<name> [<ip>]" wherever possible.
2389 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2390 are problematic for Debian distribution, omit them from the release
2393 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2395 JH/16 Fix string representation of time values on 64bit time_t architectures.
2398 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2399 output list separator was used.
2404 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2405 checkers that were complaining about end of non-void function with no
2408 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2409 This was a regression introduced in 4.83 by another bugfix.
2411 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2413 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2414 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2420 TF/01 Correctly close the server side of TLS when forking for delivery.
2422 When a message was received over SMTP with TLS, Exim failed to clear up
2423 the incoming connection properly after forking off the child process to
2424 deliver the message. In some situations the subsequent outgoing
2425 delivery connection happened to have the same fd number as the incoming
2426 connection previously had. Exim would try to use TLS and fail, logging
2427 a "Bad file descriptor" error.
2429 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2430 utilities have not been installed.
2432 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2433 temporary space as the ACL may create new global variables.
2435 TL/01 LDAP support uses per connection or global context settings, depending
2436 upon the detected version of the libraries at build time.
2438 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2439 to extract and use the src ip:port in logging and expansions as if it
2440 were a direct connection from the outside internet. PPv2 support was
2441 updated based on HAProxy spec change in May 2014.
2443 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2445 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2446 Properly escape header and check for NULL return.
2448 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2451 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2453 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2454 characters in header names, implemented as a verify condition.
2455 Contributed by Michael Fischer v. Mollard.
2457 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2458 results permerror and temperror. Previous values are deprecated but
2459 still accepted. In a future release, err_perm and err_temp will be
2460 completely removed, which will be a backward incompatibility if the
2461 ACL tests for either of these two old results. Patch contributed by
2462 user bes-internal on the mailing list.
2464 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2466 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2467 selectors, in both main and reject logs.
2469 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2472 JH/07 Add malware type "sock" for talking to simple daemon.
2474 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2476 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2477 routers/transports under cutthrough routing.
2479 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2480 numbers. Touch up "bool" conditional to keep the same definition.
2482 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2484 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2486 JH/12 Expand items in router/transport headers_add or headers_remove lists
2487 individually rather than the list as a whole. Bug 1452.
2489 Required for reasonable handling of multiple headers_ options when
2490 they may be empty; requires that headers_remove items with embedded
2491 colons must have them doubled (or the list-separator changed).
2493 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2494 view the policy declared in the DMARC record. Currently, $dmarc_status
2495 is a combined value of both the record presence and the result of the
2498 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2500 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2501 dnslookup router and the smtp transport (applying to the forward
2504 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2505 of ldap servers used for a specific lookup. Patch provided by Heiko
2508 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2509 New variable $lookup_dnssec_authenticated for observability.
2511 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2512 Patch submitted by Lars Timman.
2514 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2516 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2517 Requires trusted mode and valid format message id, aborts otherwise.
2518 Patch contributed by Heiko Schlichting.
2520 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2521 certextract with support for various fields. Bug 1358.
2523 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2524 is requested by default, modifiable by smtp transport option
2527 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2528 operate on certificate variables to give certificate fingerprints
2529 Also new ${sha256:cert_variable}.
2531 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2533 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2534 Christian Aistleitner.
2536 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2538 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2539 file. Patch from Wolfgang Breyha.
2541 JH/25 Expand the coverage of the delivery $host and $host_address to
2542 client authenticators run in verify callout. Bug 1476.
2544 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2545 align with daemon_smtp_ports. Bug 72.
2547 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2548 support and error reporting did not work properly.
2550 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2551 and is readable. Patch from Andrew Colin Kissa.
2553 TL/14 Enhance documentation of ${run expansion and how it parses the
2554 commandline after expansion, particularly in the case when an
2555 unquoted variable expansion results in an empty value.
2557 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2559 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2560 by renaming away from it. Addresses GH issue 15, reported by
2563 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2565 TL/15 SECURITY: prevent double expansion in math comparison functions
2566 (can expand unsanitized data). Not remotely exploitable.
2573 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2575 PP/02 Make -n do something, by making it not do something.
2576 When combined with -bP, the name of an option is not output.
2578 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2581 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2582 $sender_host_name and config options to manage this, and basic check
2585 PP/05 DSCP support for outbound connections and control modifier for inbound.
2587 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2588 (Only plugin which currently uses this is kerberos4, which nobody should
2589 be using, but we should make it available and other future plugins might
2590 conceivably use it, even though it would break NAT; stuff *should* be
2591 using channel bindings instead).
2593 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2594 name; added for Sendmail compatibility; requires admin caller.
2595 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2596 just ignore it); requires trusted caller.
2597 Also parse but ignore: -Ac -Am -X<logfile>
2600 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2602 TL/02 Add +smtp_confirmation as a default logging option.
2604 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2605 Patch by Magnus Holmgren from 2007-02-20.
2607 TL/04 Bugzilla 1281 - Spec typo.
2608 Bugzilla 1283 - Spec typo.
2609 Bugzilla 1290 - Spec grammar fixes.
2611 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2613 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2615 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2616 mailing list by by Dmitry Isaikin.
2618 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2620 JH/02 Support "G" suffix to numbers in ${if comparisons.
2622 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2624 NM/01 Bugzilla 1197 - Spec typo
2625 Bugzilla 1196 - Spec examples corrections
2627 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2629 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2630 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2633 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2634 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2636 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2637 "acl {{name}{arg}...}", and optional args on acl condition
2640 JH/05 Permit multiple router/transport headers_add/remove lines.
2642 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2644 JH/07 Avoid using a waiting database for a single-message-only transport.
2645 Performance patch from Paul Fisher. Bugzilla 1262.
2647 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2650 JH/09 Add $headers_added variable, with content from use of ACL modifier
2651 add_header (but not yet added to the message). Bugzilla 199.
2653 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2654 Pulled from Bugzilla 817 by Wolfgang Breyha.
2656 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2658 (nb: this is the same fix as in Exim 4.80.1)
2660 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2663 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2666 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2668 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2669 advertise SMTP AUTH mechanism to us, instead of a generic
2670 protocol violation error. Also, make Exim more robust to bad
2671 data from the Dovecot auth socket.
2673 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2675 When a queue runner is handling a message, Exim first routes the
2676 recipient addresses, during which it prunes them based on the retry
2677 hints database. After that it attempts to deliver the message to
2678 any remaining recipients. It then updates the hints database using
2681 So if a recipient address works intermittently, it can get repeatedly
2682 deferred at routing time. The retry hints record remains fresh so the
2683 address never reaches the final cutoff time.
2685 This is a fairly common occurrence when a user is bumping up against
2686 their storage quota. Exim had some logic in its local delivery code
2687 to deal with this. However it did not apply to per-recipient defers
2688 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2690 This change adds a proper retry rule check during routing so that the
2691 final cutoff time is checked against the message's age. We only do
2692 this check if there is an address retry record and there is not a
2693 domain retry record; this implies that previous attempts to handle
2694 the address had the retry_use_local_parts option turned on. We use
2695 this as an approximation for the destination being like a local
2696 delivery, as in LMTP.
2698 I suspect this new check makes the old local delivery cutoff check
2699 redundant, but I have not verified this so I left the code in place.
2701 TF/02 Correct gecos expansion when From: is a prefix of the username.
2703 Test 0254 submits a message to Exim with the header
2707 When I ran the test suite under the user fanf2, Exim expanded
2708 the header to contain my full name, whereas it should have added
2709 a Resent-Sender: header. It erroneously treats any prefix of the
2710 username as equal to the username.
2712 This change corrects that bug.
2714 GF/01 DCC debug and logging tidyup
2715 Error conditions log to paniclog rather than rejectlog.
2716 Debug lines prefixed by "DCC: " to remove any ambiguity.
2718 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2720 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2721 Bug spotted by Jeremy Harris; was flawed since initial commit.
2722 Would have resulted in OCSP responses post-SNI triggering an Exim
2723 NULL dereference and crash.
2725 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2727 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2728 Bug detection, analysis and fix by Samuel Thibault.
2729 Bugzilla 1331, Debian bug #698092.
2731 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2733 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2734 Server implementation by Todd Lyons, client by JH.
2735 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2736 config variable "prdr_enable" controls whether the server
2737 advertises the facility. If the client requests PRDR a new
2738 acl_data_smtp_prdr ACL is called once for each recipient, after
2739 the body content is received and before the acl_smtp_data ACL.
2740 The client is controlled by both of: a hosts_try_prdr option
2741 on the smtp transport, and the server advertisement.
2742 Default client logging of deliveries and rejections involving
2743 PRDR are flagged with the string "PRDR".
2745 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2746 fclose(). Diagnosis by Todd Lyons.
2748 PP/17 Update configure.default to handle IPv6 localhost better.
2749 Patch by Alain Williams (plus minor tweaks).
2752 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2753 This is now consistent with GnuTLS, and is now documented: the
2754 previous undocumented portable approach to treating the option as
2755 unset was to force an expansion failure. That still works, and
2756 an empty string is now equivalent.
2758 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2759 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2760 not performing validation itself.
2762 PP/20 Added force_command boolean option to pipe transport.
2763 Patch from Nick Koston, of cPanel Inc.
2765 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2768 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2770 PP/21 Fix eximon continuous updating with timestamped log-files.
2771 Broken in a format-string cleanup in 4.80, missed when I repaired the
2772 other false fix of the same issue.
2773 Report and fix from Heiko Schlichting.
2776 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2777 Report from Prashanth Katuri.
2779 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2780 It's SecureTransport, so affects any MacOS clients which use the
2781 system-integrated TLS libraries, including email clients.
2783 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2784 using a MIME ACL for non-SMTP local injection.
2785 Report and assistance in diagnosis by Warren Baker.
2787 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2789 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2791 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2792 last id that failed so it may be referenced in subsequent ACL's.
2794 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2797 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2798 ldap library initialization, allowing self-signed CA's to be
2799 used. Also properly sets require_cert option later in code by
2800 using NULL (global ldap config) instead of ldap handle (per
2801 session). Bug diagnosis and testing by alxgomz.
2803 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2804 the src/util/ subdirectory.
2806 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2807 renamed to Transport Post Delivery Action by Jeremy Harris, as
2810 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2811 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2812 redis_servers = needs to be configured which will be used by the redis
2813 lookup. Patch from Warren Baker, of The Packet Hub.
2815 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2817 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2818 hostname or reverse DNS when processing a host list. Used suggestions
2819 from multiple comments on this bug.
2821 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2823 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2824 Missed a few lines, added it to make the runtest require no keyboard
2827 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2828 contains upper case chars. Make router use caseful_local_part.
2830 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2831 support when GnuTLS has been built with p11-kit.
2837 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2839 This, or similar/improved, will also be change PP/11 of 4.82.
2845 PP/01 Handle short writes when writing local log-files.
2846 In practice, only affects FreeBSD (8 onwards).
2847 Bugzilla 1053, with thanks to Dmitry Isaikin.
2849 NM/01 Bugzilla 949 - Documentation tweak
2851 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2854 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2856 PP/02 Implemented gsasl authenticator.
2858 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2860 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2861 `pkg-config foo` for cflags/libs.
2863 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2864 with rest of GSASL and with heimdal_gssapi.
2866 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2867 `pkg-config foo` for cflags/libs for the TLS implementation.
2869 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2870 properties get this fed in as external SSF. A number of robustness
2871 and debugging improvements to the cyrus_sasl authenticator.
2873 PP/08 cyrus_sasl server now expands the server_realm option.
2875 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2876 Patch by Jeremy Harris.
2878 PP/10 Added dbmjz lookup type.
2880 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2882 PP/12 MAIL args handles TAB as well as SP, for better interop with
2883 non-compliant senders.
2884 Analysis and variant patch by Todd Lyons.
2886 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2887 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2888 Patch from Dirk Mueller <dmueller@suse.com>
2890 PP/13 tls_peerdn now print-escaped for spool files.
2891 Observed some $tls_peerdn in wild which contained \n, which resulted
2892 in spool file corruption.
2894 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2895 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2896 or write after TLS renegotiation, which otherwise led to messages
2899 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2900 as a tracking header (ie: a signed header comes before the signature).
2901 Patch from Wolfgang Breyha.
2903 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2904 comma-sep list; embedded commas doubled.
2906 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2908 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2910 Report and patch from Dmitry Banschikov.
2912 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2913 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2914 failures. We appear to now support TLS1.1+ with Exim.
2916 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2917 lets Exim select keys and certificates based upon TLS SNI from client.
2918 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2919 before an outbound SMTP session. New log_selector, +tls_sni.
2921 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2922 NULL dereference. Report and patch from Alun Jones.
2924 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2925 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2926 Not seeing resolver debug output on NetBSD, but suspect this is a
2927 resolver implementation change.
2929 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2930 Left warnings. Added "eximon gdb" invocation mode.
2932 PP/21 Defaulting "accept_8bitmime" to true, not false.
2934 PP/22 Added -bw for inetd wait mode support.
2936 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2937 locate the relevant includes and libraries. Made this the default.
2939 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2940 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2942 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2943 This may cause build issues on older platforms.
2945 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2946 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2947 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2948 Added SNI support via GnuTLS too.
2949 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2951 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2953 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2954 Applied second patch from Janne, implementing suggestion to default
2955 multiple-strings-in-record handling to match SPF spec.
2957 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2959 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2960 read-only, out of scope).
2961 Patch from Wolfgang Breyha, report from Stuart Northfield.
2963 PP/29 Fix three issues highlighted by clang analyser static analysis.
2964 Only crash-plausible issue would require the Cambridge-specific
2965 iplookup router and a misconfiguration.
2966 Report from Marcin Mirosław.
2968 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2970 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2971 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2972 As part of this, removing so much warning spew let me fix some minor
2973 real issues in debug logging.
2975 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2976 assignment on my part. Fixed.
2978 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2979 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2980 Janne Snabb (who went above and beyond: thank you).
2982 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2983 string otherwise requires a connection and a bunch more work and it's
2984 relatively easy to get wrong. Should also expose TLS library linkage
2987 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2988 64-bit ${eval} (JH/03).
2990 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2991 GNU libc to support some of the 64-bit stuff, should not lead to
2992 conflicts. Defined before os.h is pulled in, so if a given platform
2993 needs to override this, it can.
2995 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2996 protection layer was required, which is not implemented.
2997 Bugzilla 1254, patch from Wolfgang Breyha.
2999 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3000 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3001 tls_dhparam take prime identifiers. Also unbreak combination of
3002 OpenSSL+DH_params+TLSSNI.
3004 PP/39 Disable SSLv2 by default in OpenSSL support.
3010 PP/01 Solaris build fix for Oracle's LDAP libraries.
3011 Bugzilla 1109, patch from Stephen Usher.
3013 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3015 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3018 TF/02 Fix a couple more cases where we did not log the error message
3019 when unlink() failed. See also change 4.74-TF/03.
3021 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3022 lock up or crash if it happened to be inside a call to libc when it
3023 got a SIGUSR1 from exiwhat.
3025 The SIGUSR1 handler appends the current process status to the process
3026 log which is later printed by exiwhat. It used to use the general
3027 purpose logging code to do this, but several functions it calls are
3028 not safe for signals.
3030 The new output code in the SIGUSR1 handler is specific to the process
3031 log, and simple enough that it's easy to inspect for signal safety.
3032 Removing some special cases also simplifies the general logging code.
3033 Removing the spurious timestamps from the process log simplifies
3036 TF/04 Improved ratelimit ACL condition.
3038 The /noupdate option has been deprecated in favour of /readonly which
3039 has clearer semantics. The /leaky, /strict, and /readonly update modes
3040 are mutually exclusive. The update mode is no longer included in the
3041 database key; it just determines when the database is updated. (This
3042 means that when you upgrade Exim will forget old rate measurements.)
3044 Exim now checks that the per_* options are used with an update mode that
3045 makes sense for the current ACL. For example, when Exim is processing a
3046 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3047 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3048 must specify per_mail/readonly. If you omit the update mode it defaults to
3049 /leaky where that makes sense (as before) or /readonly where required.
3051 The /noupdate option is now undocumented but still supported for
3052 backwards compatibility. It is equivalent to /readonly except that in
3053 ACLs where /readonly is required you may specify /leaky/noupdate or
3054 /strict/noupdate which are treated the same as /readonly.
3056 A useful new feature is the /count= option. This is a generalization
3057 of the per_byte option, so that you can measure the throughput of other
3058 aggregate values. For example, the per_byte option is now equivalent
3059 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3061 The per_rcpt option has been generalized using the /count= mechanism
3062 (though it's more complicated than the per_byte equivalence). When it is
3063 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3064 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3065 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3066 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3067 Note that using per_rcpt with a non-readonly update mode in more than
3068 one ACL will cause the recipients to be double-counted. (The per_mail
3069 and per_byte options don't have this problem.)
3071 The handling of very low rates has changed slightly. If the computed rate
3072 is less than the event's count (usually one) then this event is the first
3073 after a long gap. In this case the rate is set to the same as this event's
3074 count, so that the first message of a spam run is counted properly.
3076 The major new feature is a mechanism for counting the rate of unique
3077 events. The new per_addr option counts the number of different
3078 recipients that someone has sent messages to in the last time period. It
3079 behaves like per_rcpt if all the recipient addresses are different, but
3080 duplicate recipient addresses do not increase the measured rate. Like
3081 the /count= option this is a general mechanism, so the per_addr option
3082 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3083 example, measure the rate that a client uses different sender addresses
3084 with the options per_mail/unique=$sender_address. There are further
3085 details in the main documentation.
3087 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3089 TF/06 Removed a few PCRE remnants.
3091 TF/07 Automatically extract Exim's version number from tags in the git
3092 repository when doing development or release builds.
3094 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3095 Bugzilla 879. Patch from Paul Fisher.
3097 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3098 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3101 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3103 PP/05 Variable $av_failed, true if the AV scanner deferred.
3104 Bugzilla 1078. Patch from John Horne.
3106 PP/06 Stop make process more reliably on build failure.
3107 Bugzilla 1087. Patch from Heiko Schlittermann.
3109 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3110 Bugzilla 1089. Patch from Heiko Schlittermann.
3112 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3113 Bugzilla 1131. Patch from Holger Weiß.
3115 PP/09 Handle IPv6 addresses with SPF.
3116 Bugzilla 860. Patch from Wolfgang Breyha.
3118 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3120 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3123 PP/11 match_* no longer expand right-hand-side by default.
3124 New compile-time build option, EXPAND_LISTMATCH_RHS.
3125 New expansion conditions, "inlist", "inlisti".
3127 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3129 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3131 PP/14 fix log_write() format string regression from TF/03.
3132 Bugzilla 1152. Patch from Dmitry Isaikin.
3138 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3140 PP/02 Harmonised TLS library version reporting; only show if debugging.
3141 Layout now matches that introduced for other libraries in 4.74 PP/03.
3143 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3145 PP/04 New "dns_use_edns0" global option.
3147 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3150 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3151 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3153 TK/01 Updated PolarSSL code to 0.14.2.
3154 Bugzilla 1097. Patch from Andreas Metzler.
3156 PP/07 Catch divide-by-zero in ${eval:...}.
3157 Fixes bugzilla 1102.
3159 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3162 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3163 format-string attack -- SECURITY: remote arbitrary code execution.
3165 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3166 time unintentionally subject to list matching rules, letting the header
3167 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3168 arbitrary string expansion). This allowed for information disclosure.
3170 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3171 INT_MIN/-1 -- value coerced to INT_MAX.
3177 NM/01 Workaround for PCRE version dependency in version reporting
3180 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3181 This fixes portability to compilers other than gcc, notably
3182 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3184 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3185 makefiles for portability to HP-UX and POSIX correctness.
3187 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3188 Also via indented variable definition in the Makefile.
3189 (Debugging by Oliver Heesakkers).
3191 PP/02 Restore caching of spamd results with expanded spamd_address.
3192 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3194 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3195 Improves build reliability. Fix from: Frank Elsner
3197 NM/02 Fix wide character breakage in the rfc2047 coding
3198 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3200 NM/03 Allow underscore in dnslist lookups
3201 Fixes bug 1026. Patch from Graeme Fowler
3203 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3204 Code patches from Adam Ciarcinski of NetBSD.
3206 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3209 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3210 is logged, to avoid truncation. Patch from John Horne.
3212 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3213 Patch from Jakob Hirsch.
3215 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3216 SQL string expansion failure details.
3217 Patch from Andrey Oktyabrski.
3219 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3220 Patch from Simon Arlott.
3222 PP/09 New lookups functionality failed to compile on old gcc which rejects
3223 extern declarations in function scope.
3224 Patch from Oliver Fleischmann
3226 PP/10 Use sig_atomic_t for flags set from signal handlers.
3227 Check getgroups() return and improve debugging.
3228 Fixed developed for diagnosis in bug 927 (which turned out to be
3231 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3232 Patch from Mark Zealey.
3234 PP/12 Bugzilla 1056: Improved spamd server selection.
3235 Patch from Mark Zealey.
3237 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3238 Based on patch from Heiko Schlittermann.
3240 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3241 Patch from Uwe Doering, sign-off by Michael Haardt.
3243 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3244 variable declaration deep within a block. Bug and patch from
3247 PP/15 lookups-Makefile IRIX compatibility coercion.
3249 PP/16 Make DISABLE_DKIM build knob functional.
3251 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3252 Patch by Simon Arlott
3254 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3255 variable argument macros. Our copy now differs from upstream.
3261 TF/01 Failure to get a lock on a hints database can have serious
3262 consequences so log it to the panic log.
3264 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3265 controlled using the smtp_confirmation log selector.
3267 TF/03 Include the error message when we fail to unlink a spool file.
3269 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3270 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3271 for maintaining out-of-tree patches for some time.
3273 PP/01 Bugzilla 139: Documentation and portability issues.
3274 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3275 Handle per-OS dynamic-module compilation flags.
3277 PP/02 Let /dev/null have normal permissions.
3278 The 4.73 fixes were a little too stringent and complained about the
3279 permissions on /dev/null. Exempt it from some checks.
3280 Reported by Andreas M. Kirchwitz.
3282 PP/03 Report version information for many libraries, including
3283 Exim version information for dynamically loaded libraries. Created
3284 version.h, now support a version extension string for distributors
3285 who patch heavily. Dynamic module ABI change.
3287 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3288 privilege escalation vulnerability whereby the Exim run-time user
3289 can cause root to append content of the attacker's choosing to
3292 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3295 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3296 If dropping privileges for untrusted macros, we disabled normal logging
3297 on the basis that it would fail; for the Exim run-time user, this is not
3298 the case, and it resulted in successful deliveries going unlogged.
3299 Fixed. Reported by Andreas Metzler.
3305 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3306 only prepend for the Resent-* case. Fixes regression introduced in
3307 Exim 4.70 by NM/22 for Bugzilla 607.
3309 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3310 increasing numbers of administrators be bitten by this.
3312 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3314 PP/03 Bugzilla 994: added openssl_options main configuration option.
3316 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3318 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3320 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3322 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3323 without a peer certificate, leading to a segfault because of an
3324 assumption that peers always have certificates. Be a little more
3325 paranoid. Problem reported by Martin Tscholak.
3327 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3328 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3329 NB: ClamAV planning to remove STREAM in "middle of 2010".
3330 CL also introduces -bmalware, various -d+acl logging additions and
3331 more caution in buffer sizes.
3333 PP/09 Implemented reverse_ip expansion operator.
3335 PP/10 Bugzilla 937: provide a "debug" ACL control.
3337 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3339 PP/12 Bugzilla 973: Implement --version.
3341 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3343 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3345 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3347 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3348 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3349 ignore trailing whitespace.
3351 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3353 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3356 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3357 Notification from Dr Andrew Aitchison.
3359 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3360 ExtendedDetectionInfo response format.
3361 Notification from John Horne.
3363 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3366 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3367 XSL and documented dependency on system catalogs, with examples of how
3370 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3373 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3374 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3375 configuration file which is writeable by the Exim user or group.
3377 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3378 of configuration files to cover files specified with the -C option if
3379 they are going to be used with root privileges, not just the default
3382 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3383 option (effectively making it always true).
3385 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3386 files to be used while preserving root privileges.
3388 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3389 that rogue child processes cannot use them.
3391 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3392 run-time user, instead of root.
3394 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3395 Exim run-time user without dropping privileges.
3397 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3398 result string, instead of calling string_vformat() twice with the same
3401 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3402 for other users. Others should always drop root privileges if they use
3403 -C on the command line, even for a whitelisted configure file.
3405 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3407 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3413 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3414 $data_path, and $header_path variables; fixed documentation bugs and
3417 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3418 exipick to access non-standard spools, including the "frozen" queue
3421 NM/01 Bugzilla 965: Support mysql stored procedures.
3422 Patch from Alain Williams
3424 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3426 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3427 Patch from Andreas Metzler
3429 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3430 Patch from Kirill Miazine
3432 NM/05 Bugzilla 671: Added umask to procmail example.
3434 JJ/03 installed exipick 20100323.0, fixing doc bug
3436 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3437 directory. Notification and patch from Dan Rosenberg.
3439 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3441 TK/02 Improve log output when DKIM signing operation fails.
3443 MH/01 Treat the transport option dkim_domain as a colon separated
3444 list, not as a single string, and sign the message with each element,
3445 omitting multiple occurences of the same signer.
3447 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3448 Bugzilla 985, 986. Patch by Simon Arlott
3450 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3451 Patch by Simon Arlott
3453 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3454 MBX locking. Notification from Dan Rosenberg.
3460 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3462 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3464 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3466 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3468 NM/04 Bugzilla 915: Flag broken perl installation during build.
3474 TK/01 Added patch by Johannes Berg that expands the main option
3475 "spamd_address" if it starts with a dollar sign.
3477 TK/02 Write list of recipients to X-Envelope-Sender header when building
3478 the mbox-format spool file for content scanning (suggested by Jakob
3481 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3482 (http://www.dcc-servers.net/) support via dccifd. Activated by
3483 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3485 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3486 by Mark Daniel Reidel <mr@df.eu>.
3488 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3489 When building exim an external PCRE library is now needed -
3490 PCRE is a system library on the majority of modern systems.
3491 See entry on PCRE_LIBS in EDITME file.
3493 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3494 conversation. Added nologin parameter to request.
3495 Patch contributed by Kirill Miazine.
3497 TF/01 Do not log submission mode rewrites if they do not change the address.
3499 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3501 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3502 log files in place. Contributed by Roberto Lima.
3504 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3506 TF/03 Bugzilla 615: When checking the local_parts router precondition
3507 after a local_part_suffix or local_part_prefix option, Exim now
3508 does not use the address's named list lookup cache, since this
3509 contains cached lookups for the whole local part.
3511 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3512 Robert Millan. Documentation is in experimental-spec.txt.
3514 TF/04 Bugzilla 668: Fix parallel build (make -j).
3516 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3518 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3519 Patch provided by Jan Srzednicki.
3521 TF/05 Leading white space used to be stripped from $spam_report which
3522 wrecked the formatting. Now it is preserved.
3524 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3525 that they are available at delivery time.
3527 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3529 TF/08 TLS error reporting now respects the incoming_interface and
3530 incoming_port log selectors.
3532 TF/09 Produce a more useful error message if an SMTP transport's hosts
3533 setting expands to an empty string.
3535 NM/06 Bugzilla 744: EXPN did not work under TLS.
3536 Patch provided by Phil Pennock.
3538 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3539 Patch provided by Richard Godbee.
3541 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3542 acl_smtp_notquit, added index entry.
3544 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3545 Patch provided by Eugene Bujak.
3547 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3548 accept(). Patch provided by Maxim Dounin.
3550 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3551 Patch provided by Phil Pennock.
3553 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3555 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3556 Patch provided by Brad "anomie" Jorsch.
3558 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3559 Patch provided by Dean Brooks.
3561 TK/05 Add native DKIM support (does not depend on external libraries).
3563 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3564 Patch provided by Graeme Fowler.
3566 NM/16 Bugzilla 851: Documentation example syntax fix.
3568 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3570 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3573 NM/19 Bugzilla 745: TLS version reporting.
3574 Patch provided by Phil Pennock.
3576 NM/20 Bugzilla 167: bool: condition support.
3577 Patch provided by Phil Pennock.
3579 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3580 clients. Patch provided by Phil Pennock.
3582 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3583 Patch provided by Brad "anomie" Jorsch.
3585 NM/23 Bugzilla 687: Fix misparses in eximstats.
3586 Patch provided by Heiko Schlittermann.
3588 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3589 Patch provided by Heiko Schlittermann.
3591 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3592 plus update to original patch.
3594 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3596 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3597 Patch provided by David Brownlee.
3599 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3601 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3603 NM/30 Bugzilla 888: TLS documentation bugfixes.
3605 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3607 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3608 Unlike the original bugzilla I have changed all shell scripts in src tree.
3610 NM/33 Bugzilla 898: Transport filter timeout fix.
3611 Patch by Todd Rinaldo.
3613 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3614 Patch by Serge Demonchaux.
3616 NM/35 Bugzilla 39: Base64 decode bug fixes.
3617 Patch by Jakob Hirsch.
3619 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3621 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3623 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3625 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3631 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3632 ALT-N's libdkim that I have put here:
3633 http://duncanthrax.net/exim-experimental/
3635 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3636 were called 'true' and it seems that C99 defines that as a reserved
3637 keyword to be used with 'bool' variable types. That means you could
3638 not include C99-style headers which use bools without triggering
3639 build errors in sieve.c.
3641 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3642 as mailq or other aliases. Changed the --help handling significantly
3643 to do whats expected. exim_usage() emits usage/help information.
3645 SC/01 Added the -bylocaldomain option to eximstats.
3647 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3649 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3651 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3657 PH/01 Another patch from the Sieve maintainer.
3659 PH/02 When an IPv6 address is converted to a string for single-key lookup
3660 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3661 dots are used instead of colons so that keys in lsearch files need not
3662 contain colons. This was done some time before quoting was made available
3663 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3664 (notated using the quote facility) so as to distinguish them from IPv4
3665 keys. This meant that lookups for IP addresses in host lists did not work
3666 for iplsearch lookups.
3668 This has been fixed by arranging for IPv6 addresses to be expressed with
3669 colons if the lookup type is iplsearch. This is not incompatible, because
3670 previously such lookups could never work.
3672 The situation is now rather anomalous, since one *can* have colons in
3673 ordinary lsearch keys. However, making the change in all cases is
3674 incompatible and would probably break a number of configurations.
3676 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3679 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3680 conversion specification without a maximum field width, thereby enabling
3681 a rogue spamd server to cause a buffer overflow. While nobody in their
3682 right mind would setup Exim to query an untrusted spamd server, an
3683 attacker that gains access to a server running spamd could potentially
3684 exploit this vulnerability to run arbitrary code as the Exim user.
3686 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3687 $primary_hostname instead of what libspf2 thinks the hosts name is.
3689 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3690 a directory entry by the name of the lookup key. Previously, if a
3691 symlink pointed to a non-existing file or a file in a directory that
3692 Exim lacked permissions to read, a lookup for a key matching that
3693 symlink would fail. Now it is enough that a matching directory entry
3694 exists, symlink or not. (Bugzilla 503.)
3696 PH/03 The body_linecount and body_zerocount variables are now exported in the
3699 PH/04 Added the $dnslist_matched variable.
3701 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3702 This means they are set thereafter only if the connection becomes
3705 PH/06 Added the client_condition to authenticators so that some can be skipped
3706 by clients under certain conditions.
3708 PH/07 The error message for a badly-placed control=no_multiline_responses left
3709 "_responses" off the end of the name.
3711 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3713 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3714 (without spaces) instead of just copying the configuration text.
3716 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3718 PH/11 Added $max_received_linelength.
3720 PH/12 Added +ignore_defer and +include_defer to host lists.
3722 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3723 way in which PCRE > 7.0 is built.
3725 PH/14 Implemented queue_only_load_latch.
3727 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3728 MAIL command. The effect was to mangle the value on 64-bit systems.
3730 PH/16 Another patch from the Sieve maintainer.
3732 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3734 PH/18 If a system quota error occurred while trying to create the file for
3735 a maildir delivery, the message "Mailbox is full" was not appended to the
3736 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3737 only to a quota excession during the actual writing of the file.
3739 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3740 characters?) which causes problems in log lines. The DN values are now
3741 passed through string_printing() before being added to log lines.
3743 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3744 and InterBase are left for another time.)
3746 PH/21 Added message_body_newlines option.
3748 PH/22 Guard against possible overflow in moan_check_errorcopy().
3750 PH/23 POSIX allows open() to be a macro; guard against that.
3752 PH/24 If the recipient of an error message contained an @ in the local part
3753 (suitably quoted, of course), incorrect values were put in $domain and
3754 $local_part during the evaluation of errors_copy.
3760 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3761 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3764 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3765 issue a MAIL command.
3767 PH/02 In an ACL statement such as
3769 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3771 if a client was not listed at all, or was listed with a value other than
3772 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3773 the condition was not true (as it should be), so access was not denied.
3774 The bug was that the ! inversion was incorrectly passed on to the second
3775 item. This has been fixed.
3777 PH/03 Added additional dnslists conditions == and =& which are different from
3778 = and & when the dns lookup returns more than one IP address.
3780 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3781 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3783 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3784 FSYNC, which compiles an option called disable_fsync that allows for
3785 bypassing fsync(). The documentation is heavily laced with warnings.
3787 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3789 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3790 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3791 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3792 including adding "make clean"; (3) Added -fPIC when compiling the test
3793 dynamically loaded module, to get rid of a warning.
3795 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3796 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3797 = 0s. The bug is otherwise harmless.
3799 PH/07 There was a bug in the dovecot authenticator such that the value of
3800 $auth1 could be overwritten, and so not correctly preserved, after a
3801 successful authentication. This usually meant that the value preserved by
3802 the server_setid option was incorrect.
3804 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3806 PH/09 Installed PCRE release 7.0.
3808 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3809 run for batched SMTP input. It is now run at the start of every message
3810 in the batch. While fixing this I discovered that the process information
3811 (output by running exiwhat) was not always getting set for -bs and -bS
3812 input. This is fixed, and it now also says "batched" for BSMTP.
3814 PH/11 Added control=no_pipelining.
3816 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3817 patch, slightly modified), and move the expansion of helo_data till after
3818 the connection is made in the smtp transport (so it can use these
3821 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3823 PH/14 Added log_selector = +pid.
3825 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3827 PH/16 Add ${if forany and ${if forall.
3829 PH/17 Added dsn_from option to vary the From: line in DSNs.
3831 PH/18 Flush SMTP output before performing a callout, unless control =
3832 no_callout_flush is set.
3834 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3835 was true (the default) a successful delivery failed to delete the retry
3836 item, thus causing premature timeout of the address. The bug is now
3839 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3841 PH/21 Long custom messages for fakedefer and fakereject are now split up
3842 into multiline responses in the same way that messages for "deny" and
3843 other ACL rejections are.
3845 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3846 with slight modification.
3848 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3849 draft, changing the syntax and factoring some duplicate code".
3851 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3852 for deliveries of the second and subsequent messages over the same SMTP
3855 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3856 ${reduce, with only minor "tidies".
3858 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3860 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3861 expansion side effects.
3863 PH/27 When a message times out after an over-quota error from an Exim-imposed
3864 quota, the bounce message says "mailbox is full". This message was not
3865 being given when it was a system quota that was exceeded. It now should
3868 MH/03 Made $recipients available in local_scan(). local_scan() already has
3869 better access to the recipient list through recipients_list[], but
3870 $recipients can be useful in postmaster-provided expansion strings.
3872 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3873 in the case of a MAIL command with additional options following the
3874 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3875 were accidentally chopped off.
3877 PH/29 SMTP synchronization checks are implemented when a command is read -
3878 there is a check that no more input is waiting when there shouldn't be
3879 any. However, for some commands, a delay in an ACL can mean that it is
3880 some time before the response is written. In this time, more input might
3881 arrive, invalidly. So now there are extra checks after an ACL has run for
3882 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3883 pipelining has not been advertised.
3885 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3887 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3888 returned in the SMTP response when smtp_return_error_details was false.
3889 This has been fixed.
3891 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3892 instead of the C I/O that was originally supplied, because problems were
3893 reported on Solaris.
3895 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3896 Exim which did not show up earlier: it was assuming that a call to
3897 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3898 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3899 was a macro that became an assignment, so it seemed to work. This has
3900 changed to a proper function call with a void return, hence the compile
3901 error. Exim's code has been fixed.
3903 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3906 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3908 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3910 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3911 because it turns out to be too big or there is a timeout) there is no
3912 "Completed" line in the log. When some messages of this type were
3913 selected by exigrep, they were listed as "not completed". Others were
3914 picked up by some special patterns. I have improved the selection
3915 criteria to be more general.
3917 PH/38 The host_find_failed option in the manualroute router can now be set
3918 to "ignore", to completely ignore a host whose IP address cannot be
3919 found. If all hosts are ignored, the behaviour is controlled by the new
3920 host_all_ignored option.
3922 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3923 homing or because of multiple MX records with /mx) generated more than
3924 one IP address, and the following item turned out to be the local host,
3925 all the secondary addresses of the first item were incorrectly removed
3926 from the list, along with the local host and any following hosts (which
3927 is what is supposed to happen).
3929 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3930 whoever called Exim into the -H file. In the case of the daemon it was
3931 behaving confusingly. When first started, it used values for whoever
3932 started the daemon, but after a SIGHUP it used the Exim user (because it
3933 calls itself on a restart). I have changed the code so that it now always
3936 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3937 message are rejected with the same error (e.g. no authentication or bad
3938 sender address), and a DATA command is nevertheless sent (as can happen
3939 with PIPELINING or a stupid MUA), the error message that was given to the
3940 RCPT commands is included in the rejection of the DATA command. This is
3941 intended to be helpful for MUAs that show only the final error to their
3944 PH/42 Another patch from the Sieve maintainer.
3946 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3947 Eximstats - Fixed some broken HTML links and added missing column headers
3949 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3950 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3952 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3954 PH/43 Yet another patch from the Sieve maintainer.
3956 PH/44 I found a way to check for a TCP/IP connection going away before sending
3957 the response to the final '.' that terminates a message, but only in the
3958 case where the client has not sent further data following the '.'
3959 (unfortunately, this is allowed). However, in many cases there won't be
3960 any further data because there won't be any more messages to send. A call
3961 to select() can be used: if it shows that the input is "ready", there is
3962 either input waiting, or the socket has been closed. An attempt to read
3963 the next input character can distinguish the two cases. Previously, Exim
3964 would have sent an OK response which the client would never have see.
3965 This could lead to message repetition. This fix should cure that, at
3966 least in a lot of common cases.
3968 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3969 advertised in response to EHLO.
3975 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3976 fixed by 4.65/MH/01 (is this a record?) are fixed:
3978 (i) An empty string was always treated as zero by the numeric comparison
3979 operators. This behaviour has been restored.
3981 (ii) It is documented that the numeric comparison operators always treat
3982 their arguments as decimal numbers. This was broken in that numbers
3983 starting with 0 were being interpreted as octal.
3985 While fixing these problems I realized that there was another issue that
3986 hadn't been noticed. Values of message_size_limit (both the global option
3987 and the transport option) were treated as octal if they started with 0.
3988 The documentation was vague. These values are now always treated as
3989 decimal, and I will make that clear in the documentation.
3995 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3996 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3999 MH/01 Don't check that the operands of numeric comparison operators are
4000 integers when their expansion is in "skipping" mode (fixes bug
4001 introduced by 4.64-PH/07).
4003 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4004 child addresses, Exim now panics and dies. Previously, because the count
4005 is held in a short int, deliveries were likely to be lost. As such a
4006 large number of recipients for a single message is ridiculous
4007 (performance will be very, very poor), I have chosen to impose a limit
4008 rather than extend the field.
4014 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4015 leftover -K file (the existence of which was triggered by #402).
4016 While we were at it, introduced process PID as part of the -K
4017 filename. This should rule out race conditions when creating
4020 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4021 processing considerably. Previous code took too long for large mails,
4022 triggering a timeout which in turn triggers #401.
4024 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4025 in the DK code in transports.c. sendfile() is not really portable,
4026 hence the _LINUX specificness.
4028 TF/01 In the add_headers option to the mail command in an Exim filter,
4029 there was a bug that Exim would claim a syntax error in any
4030 header after the first one which had an odd number of characters
4033 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4034 callout verification, Exim cached a "reject" for the entire domain. This
4035 is correct for most verifications, but it is not correct for a recipient
4036 verification with use_sender or use_postmaster set, because in that case
4037 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4038 case of MAIL FROM:<> rejection from other early rejections (e.g.
4039 rejection of HELO). When verifying a recipient using a non-null MAIL
4040 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4041 Whatever the result of the callout, the value of the domain cache is
4042 left unchanged (for any other kind of callout, getting as far as trying
4043 RCPT means that the domain itself is ok).
4045 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4048 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4049 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4050 session, the daemon ignores ECONNECT errors and logs others; it now
4051 ignores EPIPE as well.
4053 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4054 (quoted-printable decoding).
4056 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4057 later the small subsequent patch to fix an introduced bug.
4059 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4061 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4063 PH/08 An error is now given if message_size_limit is specified negative.
4065 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4066 to be given (somewhat) arbitrary names.
4068 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4071 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4072 miscellaneous code fixes
4074 PH/10 Added the log_reject_target ACL modifier to specify where to log
4077 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4078 hostname. This is wrong, because it relates to the incoming message (and
4079 probably the interface on which it is arriving) and not to the outgoing
4080 callout (which could be using a different interface). This has been
4081 changed to use the value of the helo_data option from the smtp transport
4082 instead - this is what is used when a message is actually being sent. If
4083 there is no remote transport (possible with a router that sets up host
4084 addresses), $smtp_active_hostname is used.
4086 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4087 tweaks were necessary in order to get it to work (see also 21 below):
4088 (a) The code assumed that strncpy() returns a negative number on buffer
4089 overflow, which isn't the case. Replaced with Exim's string_format()
4091 (b) There were several signed/unsigned issues. I just did the minimum
4092 hacking in of casts. There is scope for a larger refactoring.
4093 (c) The code used strcasecmp() which is not a standard C function.
4094 Replaced with Exim's strcmpic() function.
4095 (d) The code set only $1; it now sets $auth1 as well.
4096 (e) A simple test gave the error "authentication client didn't specify
4097 service in request". It would seem that Dovecot has changed its
4098 interface. Fortunately there's a specification; I followed it and
4099 changed what the client sends and it appears to be working now.
4101 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4104 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4105 address A is aliased to B and C, where B exists and C does not. Without
4106 -v the output is "A verified" because verification stops after a
4107 successful redirection if more than one address is generated. However,
4108 with -v the child addresses are also verified. Exim was outputting "A
4109 failed to verify" and then showing the successful verification for C,
4110 with its parentage. It now outputs "B failed to verify", showing B's
4111 parentage before showing the successful verification of C.
4113 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4114 look up a TXT record in a specific list after matching in a combined
4117 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4118 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4119 they consult the DNS. I had assumed they would set it the way they
4120 wanted; and indeed my experiments on Linux seem to show that in some
4121 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4122 To be on the safe side, however, I have now made the interface to
4123 host_find_byname() similar to host_find_bydns(), with an argument
4124 containing the DNS resolver options. The host_find_byname() function now
4125 sets these options at its start, just as host_find_bydns() does. The smtp
4126 transport options dns_qualify_single and dns_search_parents are passed to
4127 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4128 of host_find_byname() use the default settings of RES_DEFNAMES
4129 (qualify_single) but not RES_DNSRCH (search_parents).
4131 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4132 spool_read_header() do less string testing, by means of a preliminary
4133 switch on the second character of optional "-foo" lines. (This is
4134 overdue, caused by the large number of possibilities that now exist.
4135 Originally there were few.) While I was there, I also converted the
4136 str(n)cmp tests so they don't re-test the leading "-" and the first
4137 character, in the hope this might squeeze out yet more improvement.
4139 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4140 flag allowing group syntax was set by the header_syntax check but not
4141 turned off, possible causing trouble later; (2) The flag was not being
4142 set at all for the header_verify test, causing "group"-style headers to
4143 be rejected. I have now set it in this case, and also caused header_
4144 verify to ignore an empty address taken from a group. While doing this, I
4145 came across some other cases where the code for allowing group syntax
4146 while scanning a header line wasn't quite right (mostly, not resetting
4147 the flag correctly in the right place). These bugs could have caused
4148 trouble for malformed header lines. I hope it is now all correct.
4150 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4151 with the "reply" argument non-NULL. The code, however (which originally
4152 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4153 but it didn't always do it. This confused somebody who was copying the
4154 code for some other use. I have removed all the tests.
4156 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4157 feature that was used to support insecure browsers during the U.S. crypto
4158 embargo. It requires special client support, and Exim is probably the
4159 only MTA that supported it -- and would never use it because real RSA is
4160 always available. This code has been removed, because it had the bad
4161 effect of slowing Exim down by computing (never used) parameters for the
4162 RSA_EXPORT functionality.
4164 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4165 authenticator to fail if there's a tab character in the incoming data
4166 (there should never be unless someone is messing about, as it's supposed
4167 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4168 if the connection is using TLS or if the remote IP is the same as the
4169 local IP, and the "valid-client-cert option" if a client certificate has
4172 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4173 authenticators. This can be used for authorization after authentication
4174 succeeds. (In the case of plaintext, it servers for both authentication
4177 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4178 if any retry times were supplied.
4180 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4181 connection, where there is no client IP address to check. In this
4182 situation, the verify now always succeeds.
4184 PH/25 Applied John Jetmore's -Mset patch.
4186 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4188 PH/27 In a string expansion for a processed (not raw) header when multiple
4189 headers of the same name were present, leading whitespace was being
4190 removed from all of them, but trailing whitespace was being removed only
4191 from the last one. Now trailing whitespace is removed from each header
4192 before concatenation. Completely empty headers in a concatenation (as
4193 before) are ignored.
4195 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4196 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4198 PH/29 [Removed. This was a change that I later backed out, and forgot to
4199 correct the ChangeLog entry (that I had efficiently created) before
4200 committing the later change.]
4202 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4203 address errors (4xx response to RCPT) over the same connection as other
4204 messages routed to the same hosts. Such deliveries are always "forced",
4205 so retry times are not inspected. This resulted in far too many retries
4206 for the affected addresses. The effect occurred only when there were more
4207 hosts than the hosts_max_try setting in the smtp transport when it had
4208 the 4xx errors. Those hosts that it had tried were not added to the list
4209 of hosts for which the message was waiting, so if all were tried, there
4210 was no problem. Two fixes have been applied:
4212 (i) If there are any address or message errors in an SMTP delivery, none
4213 of the hosts (tried or untried) are now added to the list of hosts
4214 for which the message is waiting, so the message should not be a
4215 candidate for sending over the same connection that was used for a
4216 successful delivery of some other message. This seems entirely
4217 reasonable: after all the message is NOT "waiting for some host".
4218 This is so "obvious" that I'm not sure why it wasn't done
4219 previously. Hope I haven't missed anything, but it can't do any
4220 harm, as the worst effect is to miss an optimization.
4222 (ii) If, despite (i), such a delivery is accidentally attempted, the
4223 routing retry time is respected, so at least it doesn't keep
4224 hammering the server.
4226 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4227 in ${readsocket because some servers need this prod.
4229 PH/32 Added some extra debug output when updating a wait-xxx database.
4231 PH/33 The hint "could be header name not terminated by colon", which has been
4232 given for certain expansion errors for a long time, was not being given
4233 for the ${if def:h_colon_omitted{... case.
4235 PH/34 The spec says: "With one important exception, whenever a domain list is
4236 being scanned, $domain contains the subject domain." There was at least
4237 one case where this was not true.
4239 PH/35 The error "getsockname() failed: connection reset by peer" was being
4240 written to the panic log as well as the main log, but it isn't really
4241 panic-worthy as it just means the connection died rather early on. I have
4242 removed the panic log writing for the ECONNRESET error when getsockname()
4245 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4246 runs only) independently of the message's sender address. This meant
4247 that, if the 4xx error was in fact related to the sender, a different
4248 message to the same recipient with a different sender could confuse
4249 things. In particular, this can happen when sending to a greylisting
4250 server, but other circumstances could also provoke similar problems.
4251 I have changed the default so that the retry time for these errors is now
4252 based a combination of the sender and recipient addresses. This change
4253 can be overridden by setting address_retry_include_sender=false in the
4256 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4257 remote server are returned as part of bounce messages. This was not
4258 happening for LMTP over a pipe (the lmtp transport), but now it is the
4259 same for both kinds of LMTP.
4261 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4262 in header lines that were in fact CNAMEs. This is no longer the case.
4264 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4265 and queue runs started by the daemon processed all messages. This has
4266 been fixed so that -R and -S can now usefully be given with -q<time>.
4268 PH/40 Import PCRE release 6.7 (fixes some bugs).
4270 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4272 PH/42 Give an error if -q is specified more than once.
4274 PH/43 Renamed the variables $interface_address and $interface_port as
4275 $received_ip_address and $received_port, to make it clear that these
4276 values apply to message reception, and not to the outgoing interface when
4277 a message is delivered. (The old names remain recognized, of course.)
4279 PH/44 There was no timeout on the connect() call when using a Unix domain
4280 socket in the ${readsocket expansion. There now is.
4282 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4283 be meaningful with "accept".
4285 SC/01 Eximstats V1.43
4286 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4288 SC/02 Eximstats V1.44
4289 Use a glob alias rather than an array ref in the generated
4290 parser. This improves both readability and performance.
4292 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4293 Collect SpamAssassin and rejection statistics.
4294 Don't display local sender or destination tables unless
4295 there is data to show.
4296 Added average volumes into the top table text output.
4298 SC/04 Eximstats V1.46
4299 Collect data on the number of addresses (recipients)
4300 as well as the number of messages.
4302 SC/05 Eximstats V1.47
4303 Added 'Message too big' to the list of mail rejection
4304 reasons (thanks to Marco Gaiarin).
4306 SC/06 Eximstats V1.48
4307 Mainlog lines which have GMT offsets and are too short to
4308 have a flag are now skipped.
4310 SC/07 Eximstats V1.49 (Alain Williams)
4311 Added the -emptyok flag.
4313 SC/08 Eximstats V1.50
4314 Fixes for obtaining the IP address from reject messages.
4316 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4317 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4318 whitespace changes from 4.64-PH/27
4320 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4323 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4324 are found, allow negative numbers in numeric criteria)
4326 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4328 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4329 to match changes made in 4.64-PH/43
4331 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4333 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4334 "domain" is actually an IP address, and force a failure. This locks out
4335 those revolvers/nameservers that support "A-for-A" lookups, in
4336 contravention of the specifications.
4338 PH/48 When a host name was looked up from an IP address, and the subsequent
4339 forward lookup of the name timed out, the host name was left in
4340 $sender_host_name, contrary to the specification.
4342 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4343 restricted to single-key lookups, Exim was not diagnosing an error if
4344 * or *@ was used with a query-style lookup.
4346 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4348 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4349 long ago, but noone interested enough thought of it. Let's just say that
4350 the "1.1" means that there are some new functions that weren't there at
4351 some point in the past.
4353 PH/51 Error processing for expansion failure of helo_data from an smtp
4354 transport during callout processing was broken.
4356 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4357 tested/used via the -bh/-bhc/-bs options.
4359 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4360 bug, fixed in subsequent PCRE releases).
4362 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4363 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4365 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4371 SC/01 Use a glob alias rather than an array ref in eximstats generated
4372 parser. This improves both readability and performance.
4374 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4375 Don't display local sender or destination tables in eximstats unless
4376 there is data to show.
4377 Added average volumes into the eximstats top table text output.
4379 SC/03 Collect data on the number of addresses (recipients) as well
4380 as the number of messages in eximstats.
4382 TF/01 Correct an error in the documentation for the redirect router. Exim
4383 does (usually) call initgroups() when daemonizing.
4385 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4386 with consistent privilege compared to when running as a daemon.
4388 TF/03 Note in the spec that $authenticated_id is not set for local
4389 submissions from trusted users.
4391 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4392 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4394 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4395 by adding some example configuration directives to the default
4396 configuration file. A little bit of work is required to uncomment the
4397 directives and define how usernames and passwords are checked, but
4398 there is now a framework to start from.
4400 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4401 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4402 without this. I don't know how relevant this is to other LDAP libraries.
4404 PH/02 Add the verb name to the "unknown ACL verb" error.
4406 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4408 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4410 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4411 directory not expanded when it should be if an expanded home directory
4412 was set for the address (which is overridden by the transport).
4414 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4417 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4418 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4419 because it is too late at that time, and has no effect.
4421 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4422 security issue with \' (bugzilla #107). I could not use the
4423 PQescapeStringConn() function, because it needs a PGconn value as one of
4426 PH/08 When testing addresses using -bt, indicate those final addresses that
4427 are duplicates that would not cause an additional delivery. At least one
4428 person was confused, thinking that -bt output corresponded to deliveries.
4429 (Suppressing duplicates isn't a good idea as you lose the information
4430 about possibly different redirections that led to the duplicates.)
4432 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4433 systems where poll() doesn't work, in particular OS X.
4435 PH/10 Added more information to debugging output for retry time not reached.
4437 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4438 operations in malware.c.
4440 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4443 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4444 syslog_duplication set false, log lines that would normally be written
4445 both the the main log and to the reject log were not written to syslog at
4448 PH/14 In the default configuration, change the use of "message" in ACL warn
4449 statements to "add_header".
4451 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4452 not followed by a command (e.g. "seen endif").
4454 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4455 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4458 PH/17 Added extra conditions to the default value of delay_warning_condition
4462 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4463 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4464 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4467 The Auto-Submitted: and various List- headers are standardised, whereas I
4468 don't think Precedence: ever was.
4470 PH/18 Refactored debugging code in route_finduser() to show more information,
4471 in particular, the error code if getpwnam() issues one.
4473 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4474 This is apparently needed in addition to the PH/07 change above to avoid
4475 any possible encoding problems.
4477 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4478 but not after initializing Perl.
4480 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4481 output them only if debugging. By default they are written stderr,
4482 apparently, which is not desirable.
4484 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4487 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4490 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4492 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4493 authenticated or an ident call has been made. Suppress the default
4494 values for $authenticated_id and $authenticated_sender (but permit -oMai
4495 and -oMas) when testing with -bh.
4497 PH/24 Re-jigged the order of the tests in the default configuration so that the
4498 tests for valid domains and recipients precede the DNS black list and CSA
4499 tests, on the grounds that those ones are more expensive.
4501 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4502 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4503 This bug exists in every version of Exim that I still have, right back to
4506 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4507 However, an attempt to turn on case-sensitivity in a regex key by
4508 including (?-i) didn't work because the subject string was already
4509 lowercased, and the effects were non-intuitive. It turns out that a
4510 one-line patch can be used to allow (?-i) to work as expected.
4516 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4517 other effects) broke the use of negated acl sub-conditions.
4519 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4522 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4523 "Deny" causes Exim to reject the incoming connection with a 554 error.
4524 Unfortunately, if there is a major crisis, such as a disk failure,
4525 tcp-wrappers gives "deny", whereas what one would like would be some
4526 kind of temporary error. A kludge has been added to help with this.
4527 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4528 554 error is used if errno is still zero or contains ENOENT (which occurs
4529 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4532 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4534 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4535 errors. Otherwise a message that provokes a temporary error (when other
4536 messages do not) can cause a whole host to time out.
4538 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4539 addresses were routed directly to files or pipes from a redirect router.
4540 File deliveries just didn't batch; pipe deliveries might have suffered
4543 PH/06 A failure to get a lock for a hints database would erroneously always say
4544 "Failed to get write lock", even when it was really a read lock.
4546 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4547 of 0600. This has been changed to use the value of the lockfile_mode
4548 option (which defaults to 0600).
4550 PH/08 Applied small patch from the Sieve maintainer.
4552 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4553 folder from quota calculations, a direct delivery into this folder messed
4554 up the contents of the maildirsize file. This was because the regex was
4555 used only to exclude .Trash (or whatever) when the size of the mailbox
4556 was calculated. There was no check that a delivery was happening into an
4557 excluded directory. This bug has been fixed by ignoring all quota
4558 processing for deliveries into excluded directories.
4560 PH/10 Added the maildirfolder_create_regex option to appendfile.
4566 PH/01 The code for finding all the local interface addresses on a FreeBSD
4567 system running IPv6 was broken. This may well have applied to all BSD
4568 systems, as well as to others that have similar system calls. The broken
4569 code found IPv4 interfaces correctly, but gave incorrect values for the
4570 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4571 that it would not match correctly against @[] and not recognize the IPv6
4574 PH/02 The ipliteral router was not recognizing addresses of the form user@
4575 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4577 PH/03 Added disable_ipv6.
4579 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4580 decoded form, because it is most often used to construct To: headers
4581 lines in autoreplies, and the decoded form may well be syntactically
4582 invalid. However, $reply_address has leading white space removed, and all
4583 newlines turned into spaces so that the autoreply transport does not
4586 PH/05 If group was specified without a user on a router, and no group or user
4587 was specified on a transport, the group from the router was ignored.
4589 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4590 for visible compile-time settings that can be used to change these
4591 numbers, for those that want even more. Backwards compatibility with old
4592 spool files has been maintained. However, going back to a previous Exim
4593 release will lost any variables that are in spool files.
4595 PH/07 Two small changes when running in the test harness: increase delay when
4596 passing a TCP/IP connection to a new process, in case the original
4597 process has to generate a bounce, and remove special handling of
4598 127.0.0.2 (sic), which is no longer necessary.
4600 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4601 be the same on different OS.
4603 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4606 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4607 whether --show-vars was specified or not
4609 JJ/02 exipick: Added support for new ACL variable spool format introduced
4612 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4613 syntactically invalid From: or Reply-to: line, and a filter used this to
4614 generate an autoreply, and therefore failed to obtain an address for the
4615 autoreply, Exim could try to deliver to a non-existent relative file
4616 name, causing unrelated and misleading errors. What now happens is that
4617 it logs this as a hard delivery error, but does not attempt to create a
4620 PH/11 The exinext utility has a -C option for testing purposes, but although
4621 the given file was scanned by exinext itself; it wasn't being passed on
4622 when Exim was called.
4624 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4625 an end-of-file indication when reading a command response.
4627 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4628 compiled. In many other places in Exim, IPv6 addresses are always
4629 recognized, so I have changed this. It also means that IPv4 domain
4630 literals of the form [IPV4:n.n.n.n] are now always recognized.
4632 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4633 used if the router is not running as root, for example, when verifying at
4634 ACL time, or when using -bh. The debugging output from this situation was
4635 non-existent - all you got was a failure to exec. I have made two
4638 (a) Failures to set uid/gid, the current directory, or a process leader
4639 in a subprocess such as that created by queryprogram now generate
4640 suitable debugging output when -d is set.
4642 (b) The queryprogram router detects when it is not running as root,
4643 outputs suitable debugging information if -d is set, and then runs
4644 the subprocess without attempting to change uid/gid.
4646 PH/15 Minor change to Makefile for building test_host (undocumented testing
4649 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4650 additional section of a DNS packet that returns MX or SRV records.
4651 Instead, it always explicitly searches for A/AAAA records. This avoids
4652 major problems that occur when a DNS server includes only records of one
4653 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4654 fixed another bug: if SRV records were looked up and the corresponding
4655 address records were *not* found in the additional section, the port
4656 values from the SRV records were lost.
4658 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4659 using the correct key (the original address) when searching the retry
4660 rules in order to find which one to use for generating the retry hint.
4662 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4663 adding the default one. Similarly, if it contains a Reply-To: header, the
4664 errors_reply_to option, if set, is not used.
4666 PH/19 When calculating a retry time, Exim used to measure the "time since
4667 failure" by looking at the "first failed" field in the retry record. Now
4668 it does not use this if it is later than than the arrival time of the
4669 message. Instead it uses the arrival time. This makes for better
4670 behaviour in cases where some deliveries succeed, thus re-setting the
4671 "first failed" field. An example is a quota failure for a huge message
4672 when small messages continue to be delivered. Without this change, the
4673 "time since failure" will always be short, possible causing more frequent
4674 delivery attempts for the huge message than are intended.
4675 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4677 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4678 $1, $2, $3) because the numerical variables can be reset during some
4679 expansion items (e.g. "match"), thereby losing the authentication data.
4681 PH/21 Make -bV show the size of off_t variables so that the test suite can
4682 decide whether to run tests for quotas > 2G.
4684 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4685 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4686 filecount value is greater than 2G or if a quota value is greater than 2G
4687 on a system where the size of off_t is not greater than 4, a panic error
4690 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4691 never match. The debug and -bh output now contains an explicit error
4692 message indicating a malformed IPv4 address or mask.
4694 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4695 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4696 PH/23 above applies.
4698 PH/25 Do not write to syslog when running in the test harness. The only
4699 occasion when this arises is a failure to open the main or panic logs
4700 (for which there is an explicit test).
4702 PH/26 Added the /no_tell option to "control=freeze".
4704 PH/27 If a host name lookup failed very early in a connection, for example, if
4705 the IP address matched host_lookup and the reverse lookup yielded a name
4706 that did not have a forward lookup, an error message of the form "no IP
4707 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4708 could be logged. Now it outputs the IP address instead of "NULL".
4710 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4711 allows the sender and the authenticated sender to be set when
4712 submitting a message from within Exim. Since child_open_exim() is
4713 documented for local_scan(), the new function should be too.
4715 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4716 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4717 results in an empty string is now treated as unset.
4719 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4721 PH/31 Added sender_verify_fail logging option.
4723 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4724 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4725 tidied the source and removed it altogether.
4727 PH/33 When a queue run was abandoned because the load average was too high, a
4728 log line was always written; now it is written only if the queue_run log
4729 selector is set. In addition, the log line for abandonment now contains
4730 information about the queue run such as the pid. This is always present
4731 in "start" and "stop" lines but was omitted from the "abandon" line.
4733 PH/34 Omit spaces between a header name and the colon in the error message that
4734 is given when verify = headers_syntax fails (if there are lots of them,
4735 the message gets confusing).
4737 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4738 names, as there are now some PTR records that contain slashes. This check
4739 is only to protect against broken name servers that fall over on strange
4740 characters, so the fact that it applies to all lookups doesn't matter.
4742 PH/36 Now that the new test suite is complete, we can remove some of the
4743 special code in Exim that was needed for the old test suite. For example,
4744 sorting DNS records because real resolvers return them in an arbitrary
4745 order. The new test suite's fake resolver always returns records in the
4748 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4749 bounces) except when queue_only is set, to avoid logging races between
4750 the different processes.
4752 PH/38 Panic-die if .include specifies a non-absolute path.
4754 PH/39 A tweak to the "H" retry rule from its user.
4756 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4757 a label. They prevented compilation on older perls.
4759 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4760 a warning to be raised on newish perls.
4762 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4763 on queue. Changes to match documented behaviour of showing count of
4764 messages matching specified criteria.
4766 PH/40 Changed the default ident timeout from 30s to 5s.
4768 PH/41 Added support for the use of login_cap features, on those BSD systems
4769 that have them, for controlling the resources used by pipe deliveries.
4771 PH/42 The content-scanning code uses fopen() to create files in which to put
4772 message data. Previously it was not paying any attention to the mode of
4773 the files. Exim runs with umask(0) because the rest of the code creates
4774 files with open(), and sets the required mode explicitly. Thus, these
4775 files were ending up world-writeable. This was not a big issue, because,
4776 being within the spool directory, they were not world-accessible. I have
4777 created a function called modefopen, which takes an additional mode
4778 argument. It sets umask(777), creates the file, chmods it to the required
4779 mode, then resets the umask. All the relevant calls to fopen() in the
4780 content scanning code have been changed to use this function.
4782 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4783 to 24 hours. This avoids potential overflow problems when processing G
4784 and H retry rules. I suspect nobody ever tinkers with this value.
4786 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4788 PH/45 When the plaintext authenticator is running as a client, the server's
4789 challenges are checked to ensure they are valid base64 strings. By
4790 default, the authentication attempt is cancelled if an invalid string is
4791 received. Setting client_ignore_invalid_base64 true ignores these errors.
4792 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4793 they are received. Thus, the responses can be made to depend on the
4794 challenges. If an invalid string is ignored, an empty string is placed in
4797 PH/46 Messages that are created by the autoreply transport now contains a
4798 References: header, in accordance with RFCs 2822 and 3834.
4800 PH/47 Added authenticated_sender_force to the smtp transport.
4802 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4804 PH/49 Installed latest patch from the Sieve maintainer.
4806 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4807 was also set, the appendfile transport was unnecessarily scanning a
4808 directory of message files (e.g. for maildir delivery) to find the count
4809 of files (along with the size), even though it did not need this
4810 information. It now does the scan only if it needs to find either the
4811 size of the count of files.
4813 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4815 PH/52 Two bugs concerned with error handling when the smtp transport is
4818 (i) Exim was not creating retry information for temporary errors given
4819 for individual recipients after the DATA command when the smtp transport
4820 was used in LMTP mode. This meant that they could be retried too
4821 frequently, and not timed out correctly.
4823 (ii) Exim was setting the flag that allows error details to be returned
4824 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4825 recipients that were returned after the DATA command.
4827 PH/53 This is related to PH/52, but is more general: for any failing address,
4828 when detailed error information was permitted to be returned to the
4829 sender, but the error was temporary, then after the final timeout, only
4830 "retry timeout exceeded" was returned. Now it returns the full error as
4831 well as "retry timeout exceeded".
4833 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4834 do this, and (what is worse) MTAs that accept it.
4836 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4837 will now be deprecated.
4839 PH/56 New os.c-cygwin from the Cygwin maintainer.
4841 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4842 formats (previously only available in exim formats via -bpr, -bpru,
4843 and -bpra. Now also available in native and exiqgrep formats)
4845 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4846 with very large, slow to parse queues
4848 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4850 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4852 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4853 responses to authentication challenges, though it was showing the
4854 challenges; (ii) I've removed the CR characters from the debug output for
4857 PH/58 Allow for the insertion of a newline as well as a space when a string
4858 is turned into more than one encoded-word during RFC 2047 encoding. The
4859 Sieve code now uses this.
4861 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4862 data_4xx, lost_connection, tls_required.
4864 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4865 message was being sent as an SMTP response.
4867 PH/61 Add -l and -k options to exicyclog.
4869 PH/62 When verifying, if an address was redirected to one new address, so that
4870 verification continued, and the new address failed or deferred after
4871 having set something in $address_data, the value of $address_data was not
4872 passed back to the ACL. This was different to the case when no
4873 redirection occurred. The value is now passed back in both cases.
4875 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4876 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4877 use login_cap.h, so on its own it isn't the distinguishing feature. The
4878 new name refers directly to the setclassresources() function.
4880 PH/65 Added configuration files for NetBSD3.
4882 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4884 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4885 is preferred over IPv4.
4887 PH/68 The bounce_return_message and bounce_return_body options were not being
4888 honoured for bounces generated during the reception of non-SMTP messages.
4889 In particular, this applied to messages rejected by the ACL. This bug has
4890 been fixed. However, if bounce_return_message is true and bounce_return_
4891 body is false, the headers that are returned for a non-SMTP message
4892 include only those that have been read before the error was detected.
4893 (In the case of an ACL rejection, they have all been read.)
4895 PH/69 The HTML version of the specification is now built in a directory called
4896 spec_html instead of spec.html, because the latter looks like a path with
4897 a MIME-type, and this confuses some software.
4899 PH/70 Catch two compiler warnings in sieve.c.
4901 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4902 function verify_get_ident() calls ip_connect() to connect a socket, but
4903 if the "connect()" function timed out, ip_connect() used to close the
4904 socket. However, verify_get_ident() also closes the socket later, and in
4905 between Exim writes to the log, which may get opened at this point. When
4906 the socket was closed in ip_connect(), the log could get the same file
4907 descriptor number as the socket. This naturally causes chaos. The fix is
4908 not to close the socket in ip_connect(); the socket should be closed by
4909 the function that creates it. There was only one place in the code where
4910 this was missing, in the iplookup router, which I don't think anybody now
4911 uses, but I've fixed it anyway.
4913 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4914 well as to direct DNS lookups. Otherwise the handling of names in host
4915 lists is inconsistent and therefore confusing.
4921 PH/01 Two changes to the default runtime configuration:
4923 (1) Move the checks for relay_from_hosts and authenticated clients from
4924 after to before the (commented out) DNS black list checks.
4926 (2) Add control=submission to the relay_from_hosts and authenticated
4927 clients checks, on the grounds that messages accepted by these
4928 statements are most likely to be submissions.
4930 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4932 (1) Generate an error if the third argument for the ${prvs expansion is
4935 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4938 (3) Reset the variables that are obtained from the first argument of
4939 ${prvscheck and used in the second argument before leaving the code,
4940 because their memory is reclaimed, so using them afterwards may do
4943 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4944 one (it's much easier than Tom thought :-).
4946 (5) Because of (4), we can now allow for the use of $prvscheck_result
4947 inside the third argument.
4949 PH/03 For some reason, the default setting of PATH when running a command from
4950 a pipe transport was just "/usr/bin". I have changed it to
4953 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4954 anything to be listed in the output from -bV.
4956 PH/05 When a filter generated an autoreply, the entire To: header line was
4957 quoted in the delivery log line, like this:
4959 => >A.N.Other <ano@some.domain> <original@ddress> ...
4961 This has been changed so that it extracts the operative address. There
4962 may be more than one such address. If so, they are comma-separated, like
4965 => >ano@some.domain,ona@other.domain <original@ddress> ...
4967 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4968 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4969 not being looked up in the rDNS to get a host name, Exim was showing the
4970 IP address twice in Received: lines, even though the IP addresses were
4971 identical. For example:
4973 Received: from [1.2.3.4] (helo=[1.2.3.4])
4975 However, if the real host name was known, it was omitting the HELO data
4976 if it matched the actual IP address. This has been tidied up so that it
4977 doesn't show the same IP address twice.
4979 PH/07 When both +timestamp and +memory debugging was on, the value given by
4980 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4981 called by the debug printing, thereby overwriting the timestamp buffer.
4982 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4984 PH/08 When the original message was included in an autoreply transport, it
4985 always said "this is a copy of the message, including all the headers",
4986 even if body_only or headers_only was set. It now gives an appropriate
4989 PH/09 Applied a patch from the Sieve maintainer which:
4991 o fixes some comments
4992 o adds the (disabled) notify extension core
4993 o adds some debug output for the result of if/elsif tests
4994 o points to the current vacation draft in the documentation
4995 and documents the missing references header update
4999 o fixes a bug in processing the envelope test (when testing
5000 multiple envelope elements, the last element determined the
5003 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5004 Electronic Mail") by including:
5006 Auto-submitted: auto-generated
5008 in the messages that it generates (bounce messages and others, such as
5009 warnings). In the case of bounce messages for non-SMTP messages, there was
5010 also a typo: it was using "Auto_submitted" (underscore instead of
5011 hyphen). Since every message generated by Exim is necessarily in response
5012 to another message, thes have all been changed to:
5014 Auto-Submitted: auto-replied
5016 in accordance with these statements in the RFC:
5018 The auto-replied keyword:
5020 - SHOULD be used on messages sent in direct response to another
5021 message by an automatic process,
5023 - MUST NOT be used on manually-generated messages,
5025 - MAY be used on Delivery Status Notifications (DSNs) and Message
5026 Disposition Notifications (MDNs),
5028 - MUST NOT be used on messages generated by automatic or periodic
5029 processes, except for messages which are automatic responses to
5032 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5033 to the default Received: header definition.
5035 PH/12 Added log selector acl_warn_skipped (default on).
5037 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5038 variables because (a) they are in the wrong storage pool and (b) even if
5039 they were copied, it wouldn't work properly because of the caching.
5041 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5042 checking when decoding. Apparently there are clients that generate
5043 overlong encoded strings. Why am I not surprised?
5045 PH/15 If the first argument of "${if match_address" was not empty, but did not
5046 contain an "@" character, Exim crashed. Now it writes a panic log message
5047 and treats the condition as false.
5049 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5051 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5052 extension "envelope-auth". The code is finished and in agreement with
5053 other implementations, but there is no documentation so far and in fact,
5054 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5055 not changing the active code.
5057 Print executed "if" and "elsif" statements when debugging is used. This
5058 helps a great deal to understand what a filter does.
5060 Document more things not specified clearly in RFC3028. I had all this
5061 sorted out, when out of a sudden new issues came to my mind. Oops."
5063 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5066 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5067 canonical form (as documented). However, after a host name lookup from
5068 the IP address, check_host() was doing a simple string comparison with
5069 addresses acquired from the DNS when checking that the found name did
5070 have the original IP as one of its addresses. Since any found IPv6
5071 addresses are likely to be in abbreviated form, the comparison could
5072 fail. Luckily, there already exists a function for doing the comparison
5073 by converting both addresses to binary, so now that is used instead of
5074 the text comparison.
5076 PH/20 There was another similar case to PH/19, when a complete host name was
5077 given in a host list; looking up its IP address could give an abbreviated
5078 form, whereas the current host's name might or might not be abbreviated.
5079 The same fix has been applied.
5085 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5086 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5089 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5090 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5092 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5094 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5095 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5096 Also, for "Auto-Submitted", treat anything other than "no" as
5097 non-personal, in accordance with RFC 3834. (Previously it treated
5098 anything starting "auto-" as non-personal.)
5100 TF/01 The control=submission/name=... option had a problem with syntax
5101 errors if the name included a slash character. The /name= option
5102 now slurps the rest of the string, so it can include any characters
5103 but it must come last in the list of options (after /sender_retain
5106 PH/05 Some modifications to the interface to the fake nameserver for the new
5114 TK/01 Added the "success_on_redirect" address verification option. See
5115 NewStuff for rationale and an example.
5117 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5119 PH/02 Patch to exigrep to allow it to work on syslog lines.
5121 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5122 fread() to skip over the body file's header line, because in Cygwin the
5123 header line is locked and is inaccessible.
5125 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5126 co-exist for some time) to make it clear that it is the Exim ID that is
5127 referenced, not the Message-ID: header line.
5129 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5130 string_format() function, because snprintf() does not exist on all
5133 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5134 ${stat: expansion item.
5136 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5137 protocol synchronization error", to keep the pedants happy.
5139 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5140 well as for IRIX systems, when gcc is being used. See the host.c source
5143 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5145 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5148 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5149 transport and to the smtp transport in LMTP mode.
5151 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5153 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5154 run a filter in a subprocess. This could lead to confusion in subsequent
5155 lookups in the parent process. There should also be a search_tidyup() at
5156 the end of the subprocess.
5158 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5159 only if the host matched helo_try_verify_hosts, which caused the
5160 verification to occur when the EHLO/HELO command was issued. The ACL just
5161 tested the remembered result. Now, if a previous verification attempt has
5162 not happened, "verify = helo" does it there and then.
5164 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5166 TK/03 Fix log output including CR from clamd.
5168 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5169 exist provoked a memory error which could cause a segfault.
5171 PH/15 Installed PCRE 6.2
5173 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5175 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5176 of the problem. Specifically, suggested +O2 rather than +O1 for the
5179 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5181 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5182 record, the port was not correctly logged when the outgoing_port log
5183 selector was set (it logged the transort's default port).
5185 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5186 fallback_hosts, and "hosts" in the smtp transport.
5188 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5189 host errors such as "Connection refused".
5191 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5192 authentication with radiusclient 0.4.9:
5194 - Error returned from rc_read_config was caught wrongly
5195 - Username/password not passed on to radius server due to wrong length.
5197 The presumption is that some radiusclient API changes for 4.51/PH/17
5198 were not taken care of correctly. The code is still untested by me (my
5199 Linux distribution still has 0.3.2 of radiusclient), but it was
5200 contributed by a Radius user.
5202 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5203 expanding the "port" option of the smtp transport.
5205 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5206 while reading a MIME header. Thanks to Tom Hughes for a patch.
5208 PH/24 Include config.h inside local_scan.h so that configuration settings are
5211 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5212 that in an ACL for RCPT (for example), you can examine exactly what was
5215 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5216 commands, but it was not correctly comparing the address with the actual
5217 client host address. Thus, it would show the EHLO address in Received:
5218 header lines when this was not necessary.
5220 PH/27 Added the % operator to ${eval:}.
5222 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5223 it should be ignoring failures (because with -C, for example, it has lost
5224 privilege). It wasn't ignoring creation failures other than "already
5227 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5230 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5231 because an input line was too long, either on its own, or by virtue of
5232 too many continuations, the temporary file was not being removed, and the
5233 return code was incorrect.
5235 PH/31 Missing "BOOL" in function definition in filtertest.c.
5237 PH/32 Applied Sieve patches from the maintainer.
5239 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5241 PH/33 Added "verify = not_blind".
5243 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5244 Local/Makefile (with some defaults set). These are used in built scripts
5245 such as exicyclog, but they have never been used in the exim_install
5246 script (though there are many overriding facilities there). I have
5247 arranged that the exim_install script now takes note of these two
5250 PH/35 Installed configuration files for Dragonfly.
5252 PH/36 When a locally submitted message by a trusted user did not contain a
5253 From: header, and the sender address was obtained from -f or from an SMTP
5254 MAIL command, and the trusted user did not use -F to supply a sender
5255 name, $originator_name was incorrectly used when constructing a From:
5256 header. Furthermore, $originator_name was used for submission mode
5257 messages from external hosts without From: headers in a similar way,
5258 which is clearly wrong.
5260 PH/37 Added control=suppress_local_fixups.
5262 PH/38 When log_selector = +received_sender was set, and the addition of the
5263 sender made the log line's construction buffer exactly full, or one byte
5264 less than full, an overflow happened when the terminating "\n" was
5267 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5268 when the result of a list match is failure because a DNS lookup failed.
5270 PH/40 RM_COMMAND is now used in the building process.
5272 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5273 the "build-* directories that it finds.
5275 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5276 address in a domain literal was a prefix of an interface address.
5278 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5279 when verifying a sender address, unless rewrite_headers is false.
5281 PH/44 Wrote a long comment about why errors_to addresses are verified as
5282 recipients, not senders.
5284 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5285 the ratelimit ACL was added.
5287 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5289 PH/46 Added extra information about PostgreSQL errors to the error string.
5291 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5292 avoiding the need to install special zones in a real server. This is
5293 backwards compatible; if it can't find the fake resolver, it drops back.
5294 Thus, both old and new test suites can be run.
5296 TF/02 Added util/ratelimit.pl
5298 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5299 clock is set back in time.
5301 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5302 Candler <B.Candler@pobox.com>.
5304 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5305 OK for addresses that are the result of redirections.
5307 PH/48 A number of further additions for the benefit of the new test suite,
5308 including a fake gethostbyname() that interfaces to the fake DNS resolver
5311 TF/06 The fix for widen_domains has also been applied to qualify_single and
5312 search_parents which are the other dnslookup options that can cause
5315 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5318 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5320 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5321 service" if it encountered an MX record with an empty target hostname.
5322 The message is now "an MX or SRV record indicated no SMTP service".
5324 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5325 if the DNS of the sending site is misconfigured. This is quite a
5326 common situation. This change restores the behaviour of treating a
5327 helo verification defer as a failure.
5329 PH/51 If self=fail was set on a router, the bounce message did not include the
5330 actual error message.
5336 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5338 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5339 command itself ended in error, the underlying message about the transport
5340 filter timeout was being overwritten with the pipe command error. Now the
5341 underlying error message should be appended to the second error message.
5343 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5345 PH/02 Reduce the amount of output that "make" produces by default. Full output
5346 can still be requested.
5348 PH/03 The warning log line about a condition test deferring for a "warn" verb
5349 was being output only once per connection, rather than after each
5350 occurrence (because it was using the same function as for successful
5351 "warn" verbs). This seems wrong, so I have changed it.
5353 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5354 it should not have, which might have caused a crash in the right
5355 circumstances, but probably never did.
5357 PH/04 Installed a modified version of Tony Finch's patch to make submission
5358 mode fix the return path as well as the Sender: header line, and to
5359 add a /name= option so that you can make the user's friendly name appear
5362 TF/03 Added the control = fakedefer ACL modifier.
5364 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5365 Mark Lowes for thorough testing.
5367 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5369 TK/03 Merged latest SRS patch from Miles Wilton.
5371 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5372 with the definition in sysexits.h (which is #included earlier).
5373 Fortunately, Exim does not actually use EX_OK. The code used to try to
5374 preserve the sysexits.h value, by assuming that macro definitions were
5375 scanned for macro replacements. I have been disabused of this notion,
5376 so now the code just undefines EX_OK before #including unistd.h.
5378 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5379 in the smtp transport. When a block could not be written in a single
5380 write() function, the timeout was being re-applied to each part-write.
5381 This seems wrong - if the receiver was accepting one byte at a time it
5382 would take for ever. The timeout is now adjusted when this happens. It
5383 doesn't have to be particularly precise.
5385 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5386 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5388 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5389 without a domain if the check to <postmaster@domain> fails.
5391 SC/01 Eximstats: added -xls and the ability to specify output files
5392 (patch written by Frank Heydlauf).
5394 SC/02 Eximstats: use FileHandles for outputting results.
5396 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5398 SC/04 Eximstats: fixed display of large numbers with -nvr option
5400 SC/05 Eximstats: fixed merging of reports with empty tables.
5402 SC/06 Eximstats: added the -include_original_destination flag
5404 SC/07 Eximstats: removed tabs and trailing whitespace.
5406 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5408 TK/06 MBOX spool code: Add real "From " MBOX separator line
5409 so the .eml file is really in mbox format (even though
5410 most programs do not really care). Patch from Alex Miller.
5412 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5413 The latter is generated from $received_to and is only set if the
5414 message has one envelope recipient. SA can use these headers,
5415 obviously out-of-the-box. Patch from Alex Miller.
5417 PH/08 The ${def test on a variable was returning false if the variable's
5418 value was "0", contrary to what the specification has always said!
5419 The result should be true unless the variable is empty.
5421 PH/09 The syntax error of a character other than { following "${if
5422 def:variable_name" (after optional whitespace) was not being diagnosed.
5423 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5424 accidental colon was present, for example, could give incorrect results.
5426 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5427 result is used (not including appendfile, where other changes are about
5430 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5431 This involved changing a lot of size variables from int to off_t. It
5432 should work with maildirs and everything.
5434 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5435 spamd dying while we are connected to it.
5437 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5440 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5441 function for BDB 4.3.
5443 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5445 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5446 the sources for PCRE have been re-arranged and more files are now
5449 PH/15 The code I had for printing potentially long long variables in PH/11
5450 above was not the best (it lost precision). The length of off_t variables
5451 is now inspected at build time, and an appropriate printing format (%ld
5452 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5453 to be "long long int" or "long int". This is needed for the internal
5454 formatting function string_vformat().
5456 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5457 the configuration file to be ":syslog", then the script "guesses" where
5458 the logs files are, rather than using the compiled in default. In our
5459 case the guess is not the same as the compiled default, so the script
5460 suddenly stopped working when I started to use syslog. The patch checks
5461 to see if log_file_path is "". If so, it attempts to read it from exim
5462 with no configuration file to get the compiled in version, before it
5463 falls back to the previous guessing code."
5465 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5466 implementing BATV in an Exim configuration. See NewStuff for the gory
5469 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5470 Makefile that are specific to HP-UX.
5472 PH/18 If the "use_postmaster" option was set for a recipient callout together
5473 with the "random" option, the postmaster address was used as the MAIL
5474 FROM address for the random test, but not for the subsequent recipient
5475 test. It is now used for both.
5477 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5478 patch removes a few documentation additions to RFC 3028, because the
5479 latest draft now contains them. It adds the new en;ascii-case comparator
5480 and a new error check for 8bit text in MIME parts. Comparator and
5481 require names are now matched exactly. I enabled the subaddress
5482 extension, but it is not well tested yet (read: it works for me)."
5484 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5485 rework some of the code of TK/09 above to avoid the hardwired use of
5486 "%lld" and "long long". Replaced the call to snprintf() with a call to
5489 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5490 records point to non-existent hosts", "retry timeout exceeded", and
5491 "retry time not reached for any host after a long failure period".
5493 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5494 experimental DomainKeys support:
5496 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5497 (2) On an error such as an illegally used "control", the wrong name for
5498 the control was given.
5500 These problems did NOT occur unless DomainKeys support was compiled.
5502 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5504 PH/24 Added ${if match_ip condition.
5506 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5507 databases so that it will be absolutely obvious if a crash occurs in the
5508 DB library. This is a regular occurrence (often caused by mis-matched
5511 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5512 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5513 picked up on a user's system that detects such things. There doesn't seem
5514 to be a gcc warning option for this - only an attribute that has to be
5515 put on the function's prototype. It seems that in Fedora Core 4 they have
5516 set this on a number of new functions. No doubt there will be more in due
5519 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5520 specify a transport. However, if an address that was verified by such a
5521 router was the subject of a callout, Exim crashed because it tried to
5522 read the rcpt_include_affixes from the non-existent transport. Now it
5523 just assumes that the setting of that option is false. This bug was
5524 introduced by 4.51/PH/31.
5526 PH/28 Changed -d+all to exclude +memory, because that information is very
5527 rarely of interest, but it makes the output a lot bigger. People tend to
5528 do -d+all out of habit.
5530 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5531 code in os-type was giving problems when libc.so lives in lib64, like on
5534 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5535 aren't the modern standard, and it seems that some systems' include files
5536 don't always have them. Exim was already checking for some of the newer
5537 ones like T_AAAA, and defining it itself. I've added checks for all the
5538 record types that Exim uses.
5540 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5541 not automatically generating a new one, as it is supposed to. This
5542 prevented TLS from working. If the file did exist, but contained invalid
5543 data, a new version was generated, as expected. It was only the case of a
5544 non-existent file that was broken.
5546 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5547 with a change in libdomainkeys > 0.64.
5549 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5550 from DNS. If the selector record carries the flag, it now has
5551 precedence over the domain-wide flag.
5553 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5555 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5556 the use of a port name that isn't defined in /etc/services) occurred, the
5557 message was deferred as in a normal delivery, and thus remained on the
5558 spool, instead of being failed because of the mua_wrapper setting. This
5559 is now fixed, and I tidied up some of the mua_wrapper messages at the
5562 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5563 the messages in a hash of arrays rather than using individual hashes.
5564 This is a bit cleaner and results in dramatic memory savings, albeit
5565 at a slight CPU cost.
5567 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5568 as requested by Marc Sherman.
5570 SC/10 Eximstats: added histograms for user specified patterns as requested
5573 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5575 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5576 fopen() in the content-scanning modules that did not already have it.
5582 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5583 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5585 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5587 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5589 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5590 to test Sieve filters that use "vacation".
5592 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5593 that changes the way the GnuTLS parameters are stored in the cache file.
5594 The new format can be generated externally. For backward compatibility,
5595 if the data in the cache doesn't make sense, Exim assumes it has read an
5596 old-format file, and it generates new data and writes a new file. This
5597 means that you can't go back to an older release without removing the
5600 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5601 work if there are any delivery delays because "one_time" forces the
5602 parent to be marked "delivered", so its unseen clone is never tried
5603 again. For this reason, Exim now forbids the simultaneous setting of
5606 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5607 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5608 entry if you want to know the details. The fix, however, neglected to
5609 consider the case when local delivery batching is involved. The test for
5610 "previously delivered" was not happening when checking to see if an
5611 address could be batched with a previous (undelivered) one; under
5612 certain circumstances this could lead to multiple deliveries to the same
5615 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5616 in its include files, and this causes problems building Exim.
5618 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5619 header_syntax) but Exim was just ignoring anything given after a slash.
5620 In particular, this caused confusion with an attempt to use "verify =
5621 reverse_host_lookup/defer_ok". An error is now given when options are
5622 supplied for verify items that do not have them. (Maybe reverse_host_
5623 lookup should have a defer_ok option, but that's a different point.)
5625 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5626 defined by RFC 821) to 2048, because there were problems with some AUTH
5627 commands, and RFC 1869 says the size should be increased for extended
5628 SMTP commands that take arguments.
5630 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5633 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5634 "unknown" error; now it says that the functionality isn't in the binary.
5636 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5637 an address' error message when a string expansion fails (syntax or
5638 whatever). Otherwise the password may appear in the log. Following change
5639 PH/42 below, there is no longer a chance of it appearing in a bounce
5642 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5644 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5645 of its addresses was ever tried. (Bugzilla bug #2.)
5647 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5648 the result incorrectly in the debug output. (It correctly added a newline
5649 to what was transported.)
5651 TF/01 Added $received_time.
5653 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5654 commented out examples of how to interface to a virus scanner and to
5655 SpamAssassin. Also added commented examples of av_scanner and
5656 spamd_address settings.
5658 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5659 and controls are allowed in which ACLs. There were a couple of minor
5660 errors. Some of the entries in the conditions table (which is a table of
5661 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5662 negation of where the condition IS allowed.
5664 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5666 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5667 header file does not have a version number, so I've had to invent a new
5668 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5669 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5670 radiusclient), but it was contributed by a Radius user.
5672 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5673 files or directories, for OpenSSL.
5675 PH/19 When an Exim process that is running as root has to create an Exim log
5676 file, it does so in a subprocess that runs as exim:exim so as to get the
5677 ownership right at creation (otherwise, other Exim processes might see
5678 the file with the wrong ownership). There was no test for failure of this
5679 fork() call, which would lead to the process getting stuck as it waited
5680 for a non-existent subprocess. Forks do occasionally fail when resources
5681 run out. I reviewed all the other calls to fork(); they all seem to check
5684 PH/20 When checking for unexpected SMTP input at connect time (before writing
5685 the banner), Exim was not dealing correctly with a non-positive return
5686 from the read() function. If the client had disconnected by this time,
5687 the result was a log entry for a synchronization error with an empty
5688 string after "input=" when read() returned zero. If read() returned -1
5689 (an event I could not check), uninitialized data bytes were printed.
5690 There were reports of junk text (parts of files, etc) appearing after
5693 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5695 PH/22 Added support for macro redefinition, and (re)definition in between
5696 driver and ACL definitions.
5698 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5699 forgetting to use the resulting value; it was using the unexpanded value.
5701 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5702 hadn't been configured. The fix is from Juergen Kreileder, who
5703 understands it better than I do:
5705 "Here's what I see happening with three configured cyrus_sasl
5706 authenticators configured (plain, login, cram-md5):
5708 On startup auth_cyrus_sasl_init() gets called for each of these.
5709 This means three calls to sasl_listmech() without a specified mech_list.
5710 => SASL tests which mechs of all available mechs actually work
5711 => three warnings about OTP not working
5712 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5714 With the patch, sasl_listmech() also gets called three times. But now
5715 SASL's mech_list option is set to the server_mech specified in the the
5716 authenticator. Or in other words, the answer from sasl_listmech()
5717 gets limited to just the mech you're testing for (which is different
5719 => the return list contains just 'plain' or 'login', 'cram-md5' or
5720 nothing depending on the value of ob->server_mech.
5722 I've just tested the patch: Authentication still works fine,
5723 unavailable mechs specified in the exim configuration are still
5724 caught, and the auth.log warnings about OTP are gone."
5726 PH/25 When debugging is enabled, the contents of the command line are added
5727 to the debugging output, even when log_selector=+arguments is not
5730 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5731 answer is "GNU", and only if the return is "GNU/something" is the answer
5734 PH/27 $acl_verify_message is now set immediately after the failure of a
5735 verification in an ACL, and so is available in subsequent modifiers. In
5736 particular, the message can be preserved by coding like this:
5738 warn !verify = sender
5739 set acl_m0 = $acl_verify_message
5741 Previously, $acl_verify_message was set only while expanding "message"
5742 and "log_message" when a very denied access.
5744 PH/28 Modified OS/os.c-Linux with
5746 -#ifndef OS_LOAD_AVERAGE
5747 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5749 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5750 nomenclature these days.)
5752 PH/29 Installed patch from the Sieve maintainer that adds the options
5753 sieve_useraddress and sieve_subaddress to the redirect router.
5755 PH/30 In these circumstances:
5756 . Two addresses routed to the same list of hosts;
5757 . First host does not offer TLS;
5758 . First host accepts first address;
5759 . First host gives temporary error to second address;
5760 . Second host offers TLS and a TLS session is established;
5761 . Second host accepts second address.
5762 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5763 and peerdn, if requested) that were in fact used only for the second
5766 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5767 attention to any local part prefix or suffix that was matched by the
5768 router that accepted the address. It now behaves in the same way as it
5769 does for delivery: the affixes are removed from the local part unless
5770 rcpt_include_affixes is set on the transport.
5772 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5773 timeout during the DATA phase of an incoming message.
5775 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5776 applied a patch sanctioned by the Sieve maintainer.
5778 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5779 the uid or gid is negative. A case of a negative gid caused this to be
5780 noticed. The fix allows for either to be negative.
5782 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5783 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5786 PH/36 The change PH/12 above was broken. Fixed it.
5788 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5789 the grounds that routing the same address twice would always produce the
5790 same answer. This might have been true once, but it is certainly no
5791 longer true now. Routing a child address may depend on the previous
5792 routing that produced that child. Some complicated redirection strategies
5793 went wrong when messages had multiple recipients, and made Exim's
5794 behaviour dependent on the order in which the addresses were given.
5796 I have moved the duplicate checking until after the routing is complete.
5797 Exim scans the addresses that are assigned to local and remote
5798 transports, and removes any duplicates. This means that more work will be
5799 done, as duplicates will always all be routed, but duplicates are
5800 presumably rare, so I don't expect this is of any significance.
5802 For deliveries to pipes, files, and autoreplies, the duplicate checking
5803 still happens during the routing process, since they are not going to be
5806 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5807 It corrects a timeout issue with spamd. This is Ian's comment: "The
5808 background is that sometimes spamd either never reads data from a
5809 connection it has accepted, or it never writes response data. The exiscan
5810 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5811 blindly assumes that writes won't block so it may never time out."
5813 PH/39 Allow G after quota size as well as K and M.
5815 PH/40 The value set for $authenticated_id in an authenticator may not contain
5816 binary zeroes or newlines because the value is written to log lines and
5817 to spool files. There was no check on this. Now the value is run through
5818 the string_printing() function so that such characters are converted to
5819 printable escape sequences.
5821 PH/41 $message_linecount is a new variable that contains the total number of
5822 lines in the message. Compare $body_linecount, which is the count for the
5825 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5826 bounce and delay warning messages, except in certain special cases, which
5829 (a) An SMTP error message from a remote host;
5830 (b) A message specified in a :fail: redirection;
5831 (c) A message specified in a "fail" command in a system filter;
5832 (d) A message specified in a FAIL return from the queryprogram router;
5833 (e) A message specified by the cannot_route_message router option.
5835 In these cases only, Exim does include the error details in bounce and
5836 warning messages. There are also a few cases where bland messages such
5837 as "unrouteable address" or "local delivery error" are given.
5839 PH/43 $value is now also set for the "else" part of a ${run expansion.
5841 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5842 being worked on, but at least Exim now implements the latest version to
5845 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5846 process to complete was treated as a delivery failure, a timeout while
5847 writing the message to the pipe was logged, but erroneously treated as a
5848 successful delivery. Such timeouts include transport filter timeouts. For
5849 consistency with the overall process timeout, these timeouts are now
5850 treated as errors, giving rise to delivery failures by default. However,
5851 there is now a new Boolean option for the pipe transport called
5852 timeout_defer, which, if set TRUE, converts the failures into defers for
5853 both kinds of timeout. A transport filter timeout is now identified in
5856 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5857 systems where "make" and "gmake" are different, calling "gmake" at top
5858 level broke things. I've arranged for the value of $(MAKE) to be passed
5859 from the Makefile to this script so that it can call the same version of
5863 A note about Exim versions 4.44 and 4.50
5864 ----------------------------------------
5866 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5867 changes of various kinds. As a consequence, a big documentation update was
5868 needed. This delayed the release for rather longer than seemed good, especially
5869 in the light of a couple of (minor) security issues. Therefore, the changes
5870 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5871 release. So 4.44 and 4.50 are in effect two different branches that both start
5874 I have left the 4.50 change log unchanged; it contains all the changes since
5875 4.43. The change log for 4.44 is below; many of its items are identical to
5876 those for 4.50. This seems to be the most sensible way to preserve the
5877 historical information.
5883 1. Minor wording change to the doc/README.SIEVE file.
5885 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5886 computation of the current number of files was incorrect.
5888 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5889 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5892 4. Give more explanation in the error message when the command for a transport
5893 filter fails to execute.
5895 5. There are several places where Exim runs a non-Exim command in a
5896 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5897 was being done only for the command run by the queryprogram router. It is
5898 now done for all such subprocesses. The other cases are: ${run, transport
5899 filters, and the commands run by the lmtp and pipe transports.
5901 6. Added CONFIGURE_GROUP build-time option.
5903 7. Some older OS have a limit of 256 on the maximum number of file
5904 descriptors. Exim was using setrlimit() to set 1000 as a large value
5905 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5906 systems. I've change it so that if it can't get 1000, it tries for 256.
5908 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5909 was an oversight, and furthermore, ever since the addition of extra
5910 controls (e.g. 4.43/32), the checks on when to allow different forms of
5911 "control" were broken. There should now be diagnostics for all cases when a
5912 control that does not make sense is encountered.
5914 9. Added the /retain_sender option to "control=submission".
5916 10. $recipients is now available in the predata ACL (oversight).
5918 11. Tidy the search cache before the fork to do a delivery from a message
5919 received from the command line. Otherwise the child will trigger a lookup
5920 failure and thereby defer the delivery if it tries to use (for example) a
5921 cached ldap connection that the parent has called unbind on.
5923 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5924 of $address_data from the recipient verification was clobbered by the
5925 sender verification.
5927 13. The value of address_data from a sender verification is now available in
5928 $sender_address_data in subsequent conditions in the ACL statement.
5930 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5932 15. Added a new option "connect=<time>" to callout options, to set a different
5935 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5936 was its contents. (It was OK if the option was not defined at all.)
5938 17. A "Completed" log line is now written for messages that are removed from
5939 the spool by the -Mrm option.
5941 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5942 information about exactly what failed.
5944 19. Added -dd to debug only the daemon process.
5946 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5947 handles timeouts, both on the server side and network timeouts. Renamed the
5948 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5950 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5951 It is now set to "smtps".
5953 22. $host_address is now set to the target address during the checking of
5954 ignore_target_hosts.
5956 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5957 being passed; this would have caused $sender_host_name to have been used if
5958 matching the list had actually called for a host name (not very likely,
5959 since this list is usually IP addresses). A host name is now passed as
5962 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5963 code that specifies a non-restarting handler (typically sigaction() in
5964 modern systems) in an attempt to fix a rare and obscure crash bug.
5966 25. Narrowed the window for a race in the daemon that could cause it to ignore
5967 SIGCHLD signals. This is not a major problem, because they are used only to
5968 wake it up if nothing else does.
5970 26. A malformed maildirsize file could cause Exim to calculate negative values
5971 for the mailbox size or file count. Odd effects could occur as a result.
5972 The maildirsize information is now recalculated if the size or filecount
5975 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5976 support for a long time. Removed HAVE_SYS_VFS_H.
5978 28. Installed the latest version of exipick from John Jetmore.
5980 29. In an address list, if the pattern was not a regular expression, an empty
5981 subject address (from a bounce message) matched only if the pattern was an
5982 empty string. Non-empty patterns were not even tested. This was the wrong
5983 because it is perfectly reasonable to use an empty address as part of a
5984 database query. An empty address is now tested by patterns that are
5985 lookups. However, all the other forms of pattern expect the subject to
5986 contain a local part and a domain, and therefore, for them, an empty
5987 address still always fails if the pattern is not itself empty.
5989 30. Exim went into a mad DNS loop when attempting to do a callout where the
5990 host was specified on an smtp transport, and looking it up yielded more
5991 than one IP address.
5993 31. Re-factored the code for checking spool and log partition space into a
5994 function that finds that data and another that does the check. The former
5995 is then used to implement four new variables: $spool_space, $log_space,
5996 $spool_inodes, and $log_inodes.
5998 32. The RFC2047 encoding function was originally intended for short strings
5999 such as real names; it was not keeping to the 75-character limit for
6000 encoded words that the RFC imposes. It now respects the limit, and
6001 generates multiple encoded words if necessary. To be on the safe side, I
6002 have increased the buffer size for the ${rfc2047: expansion operator from
6005 33. It is now permitted to omit both strings after an "if" condition; if the
6006 condition is true, the result is "true". As before, when the second string
6007 is omitted, a false condition yields an empty string. This makes it less
6008 cumbersome to write custom ACL and router conditions.
6010 34. Failure to deliver a bounce message always caused it to be frozen, even if
6011 there was an errors_to setting on the router. The errors_to setting is now
6014 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6015 canonical form (fully expanded) before being placed in
6016 $sender_host_address.
6018 36. The table in the code that translates DNS record types into text (T_A to
6019 "A" for instance) was missing entries for NS and CNAME. It is just possible
6020 that this could have caused confusion if both these types were looked up
6021 for the same domain, because the text type is used as part of Exim's
6022 per-process caching. But the chance of anyone hitting this buglet seems
6025 37. The dnsdb lookup has been extended in a number of ways.
6027 (1) There is a new type, "zns", which walks up the domain tree until it
6028 finds some nameserver records. It should be used with care.
6030 (2) There is a new type, "mxh", which is like "mx" except that it returns
6031 just the host names, not the priorities.
6033 (3) It is now possible to give a list of domains (or IP addresses) to be
6034 looked up. The behaviour when one of the lookups defers can be
6035 controlled by a keyword.
6037 (4) It is now possible to specify the separator character for use when
6038 multiple records are returned.
6040 38. The dnslists ACL condition has been extended: it is now possible to supply
6041 a list of IP addresses and/or domains to be looked up in a particular DNS
6044 39. Added log_selector=+queue_time_overall.
6046 40. When running the queue in the test harness, wait just a tad after forking a
6047 delivery process, to get repeatability of debugging output.
6049 41. Include certificate and key file names in error message when GnuTLS fails
6050 to set them up, because the GnuTLS error message doesn't include the name
6051 of the failing file when there is a problem reading it.
6053 42. Allow both -bf and -bF in the same test run.
6055 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6057 44. The "Exiscan patch" is now merged into the mainline Exim source.
6059 45. Sometimes the final signoff response after QUIT could fail to get
6060 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6061 < 0 before doing a fflush(). This bug looks as though it goes back to the
6062 introduction of TLS in release 3.20, but "sometimes" must have been rare
6063 because the tests only now provoked it.
6065 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6066 (this can affect the format of dates).
6068 47. exim_tidydb, when checking for the continued existence of a message for
6069 which it has found a message-specific retry record, was not finding
6070 messages that were in split spool directories. Consequently, it was
6071 deleting retry records that should have stayed in existence.
6073 48. Steve fixed some bugs in eximstats.
6075 49. The SPA authentication driver was not abandoning authentication and moving
6076 on to the next authenticator when an expansion was forced to fail,
6077 contradicting the general specification for all authenticators. Instead it
6078 was generating a temporary error. It now behaves as specified.
6080 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6081 (the order specifies the preference for clients). The order is now AES256,
6082 AES128, 3DES, ARCFOUR128.
6084 51. Small patch to Sieve code - explicitly set From: when generating an
6087 52. Exim crashed if a remote delivery caused a very long error message to be
6088 recorded - for instance if somebody sent an entire SpamAssassin report back
6089 as a large number of 550 error lines. This bug was coincidentally fixed by
6090 increasing the size of one of Exim's internal buffers (big_buffer) that
6091 happened as part of the Exiscan merge. However, to be on the safe side, I
6092 have made the code more robust (and fixed the comments that describe what
6095 53. Now that there can be additional text after "Completed" in log lines (if
6096 the queue_time_overall log selector is set), a one-byte patch to exigrep
6097 was needed to allow it to recognize "Completed" as not the last thing in
6100 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6101 patch that reportedly fixes this has been added. I am not expert enough to
6102 create a test for it. This is what the patch creator wrote:
6104 "I found a little strange behaviour of ldap code when working with
6105 Windows 2003 AD Domain, where users was placed in more than one
6106 Organization Units. When I tried to give exim partial DN, the exit code
6107 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6108 But simultaneously result of request was absolutely normal ldap result,
6109 so I produce this patch..."
6111 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6112 REFERENCE, so I have modified the code to exclude the patch when that macro
6115 55. Some experimental protocols are using DNS PTR records for new purposes. The
6116 keys for these records are domain names, not reversed IP addresses. The
6117 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6118 leaves it alone. Component reversal etc. now happens only for IP addresses.
6121 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6123 57. Double the size of the debug message buffer (to 2048) so that more of very
6124 long debug lines gets shown.
6126 58. The exicyclog utility now does better if the number of log files to keep
6127 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6129 59. Two changes related to the smtp_active_hostname option:
6131 (1) $smtp_active_hostname is now available as a variable.
6132 (2) The default for smtp_banner uses $smtp_active_hostname instead
6133 of $primary_hostname.
6135 60. The host_aton() function is supposed to be passed a string that is known
6136 to be a valid IP address. However, in the case of IPv6 addresses, it was
6137 not checking this. This is a hostage to fortune. Exim now panics and dies
6138 if the condition is not met. A case was found where this could be provoked
6139 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6140 components; fortuitously, this particular loophole had already been fixed
6141 by change 4.50/55 above.
6143 If there are any other similar loopholes, the new check in host_aton()
6144 itself should stop them being exploited. The report I received stated that
6145 data on the command line could provoke the exploit when Exim was running as
6146 exim, but did not say which command line option was involved. All I could
6147 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6148 running as the user.
6151 61. There was a buffer overflow vulnerability in the SPA authentication code
6152 (which came originally from the Samba project). I have added a test to the
6153 spa_base64_to_bits() function which I hope fixes it.
6156 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6157 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6159 63. The daemon start-up calls getloadavg() while still root for those OS that
6160 need the first call to be done as root, but it missed one case: when
6161 deliver_queue_load_max is set with deliver_drop_privilege. This is
6162 necessary for the benefit of the queue runner, because there is no re-exec
6163 when deliver_drop_privilege is set.
6165 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6166 This has been fixed.
6168 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6169 was in use, was not putting the data itself into the right store pool;
6170 consequently, it could be overwritten for a subsequent message in the same
6171 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6174 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6176 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6177 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6178 the function were treating the return as a boolean value, which happened to
6179 work because 0=false and not-0=true, but is not correct code.
6181 68. The host_aton() function was not handling scoped IPv6 addresses (those
6182 with, for example, "%eth0" on the end) correctly.
6184 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6185 negated items (that is, ~something) in unsigned ints. Some compilers
6186 apparently mutter when there is no cast.
6188 70. If an address verification called from an ACL failed, and did not produce a
6189 user-specific message (i.e. there was only a "system" message), nothing was
6190 put in $acl_verify_message. In this situation, it now puts the system
6193 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6194 session; change 4.31/43 added the unwanted input to the log line - except
6195 that it did not do this in the start of session case. It now does.
6197 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6198 This is wrong and can cause the other end to generate a synchronization
6199 error if it is another Exim or anything else that does the synchronization
6200 check. A QUIT command is no longer sent after a timeout.
6202 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6203 during host lookups.
6205 74. The defer_ok option of callout verification was not working if it was used
6206 when verifying addresses in header lines, that is, for this case:
6208 verify = header_sender/callout=defer_ok
6210 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6211 those file descriptors could be used for SMTP connections. If anything
6212 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6213 could be sent to the SMTP client, causing chaos. The daemon now opens
6214 stdin, stdout, and stderr to /dev/null when it puts itself into the
6217 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6218 log by default. The user can override this with suitable Perl magic.
6220 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6221 the log message when discard triggers, was not working for the DATA ACL or
6222 for the non-SMTP ACL.
6224 78. Error message wording change in sieve.c.
6226 79. If smtp_accept_max_per_host was set, the number of connections could be
6227 restricted to fewer than expected, because the daemon was trying to set up
6228 a new connection before checking whether the processes handling previous
6229 connections had finished. The check for completed processes is now done
6230 earlier. On busy systems, this bug wouldn't be noticed because something
6231 else would have woken the daemon, and it would have reaped the completed
6234 80. If a message was submitted locally by a user whose login name contained one
6235 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6236 It caused a spool format error. I have fixed the spool reading code. A
6237 related problem was that the "from" clause in the Received: line became
6238 illegal because of the space(s). It is now covered by ${quote_local_part.
6240 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6243 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6245 83. Patch from Sieve maintainer to fix unterminated string problem in
6246 "vacation" handling.
6248 84. Some minor changes to the Linux configuration files to help with other
6249 OS variants using glibc.
6251 85. One more patch for Sieve to update vacation handling to latest spec.
6254 ----------------------------------------------------
6255 See the note above about the 4.44 and 4.50 releases.
6256 ----------------------------------------------------
6262 1. Change 4.43/35 introduced a bug that caused file counts to be
6263 incorrectly computed when quota_filecount was set in an appendfile
6266 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6267 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6270 3. Give more explanation in the error message when the command for a transport
6271 filter fails to execute.
6273 4. There are several places where Exim runs a non-Exim command in a
6274 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6275 was being done only for the command run by the queryprogram router. It is
6276 now done for all such subprocesses. The other cases are: ${run, transport
6277 filters, and the commands run by the lmtp and pipe transports.
6279 5. Some older OS have a limit of 256 on the maximum number of file
6280 descriptors. Exim was using setrlimit() to set 1000 as a large value
6281 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6282 systems. I've change it so that if it can't get 1000, it tries for 256.
6284 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6285 was an oversight, and furthermore, ever since the addition of extra
6286 controls (e.g. 4.43/32), the checks on when to allow different forms of
6287 "control" were broken. There should now be diagnostics for all cases when a
6288 control that does not make sense is encountered.
6290 7. $recipients is now available in the predata ACL (oversight).
6292 8. Tidy the search cache before the fork to do a delivery from a message
6293 received from the command line. Otherwise the child will trigger a lookup
6294 failure and thereby defer the delivery if it tries to use (for example) a
6295 cached ldap connection that the parent has called unbind on.
6297 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6298 of $address_data from the recipient verification was clobbered by the
6299 sender verification.
6301 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6302 was its contents. (It was OK if the option was not defined at all.)
6304 11. A "Completed" log line is now written for messages that are removed from
6305 the spool by the -Mrm option.
6307 12. $host_address is now set to the target address during the checking of
6308 ignore_target_hosts.
6310 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6311 being passed; this would have caused $sender_host_name to have been used if
6312 matching the list had actually called for a host name (not very likely,
6313 since this list is usually IP addresses). A host name is now passed as
6316 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6317 code that specifies a non-restarting handler (typically sigaction() in
6318 modern systems) in an attempt to fix a rare and obscure crash bug.
6320 15. Narrowed the window for a race in the daemon that could cause it to ignore
6321 SIGCHLD signals. This is not a major problem, because they are used only to
6322 wake it up if nothing else does.
6324 16. A malformed maildirsize file could cause Exim to calculate negative values
6325 for the mailbox size or file count. Odd effects could occur as a result.
6326 The maildirsize information is now recalculated if the size or filecount
6329 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6330 support for a long time. Removed HAVE_SYS_VFS_H.
6332 18. Updated exipick to current release from John Jetmore.
6334 19. Allow an empty sender to be matched against a lookup in an address list.
6335 Previously the only cases considered were a regular expression, or an
6338 20. Exim went into a mad DNS lookup loop when doing a callout where the
6339 host was specified on the transport, if the DNS lookup yielded more than
6342 21. The RFC2047 encoding function was originally intended for short strings
6343 such as real names; it was not keeping to the 75-character limit for
6344 encoded words that the RFC imposes. It now respects the limit, and
6345 generates multiple encoded words if necessary. To be on the safe side, I
6346 have increased the buffer size for the ${rfc2047: expansion operator from
6349 22. Failure to deliver a bounce message always caused it to be frozen, even if
6350 there was an errors_to setting on the router. The errors_to setting is now
6353 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6354 canonical form (fully expanded) before being placed in
6355 $sender_host_address.
6357 24. Updated eximstats to version 1.33
6359 25. Include certificate and key file names in error message when GnuTLS fails
6360 to set them up, because the GnuTLS error message doesn't include the name
6361 of the failing file when there is a problem reading it.
6363 26. Expand error message when OpenSSL has problems setting up cert/key files.
6366 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6367 (this can affect the format of dates).
6369 28. exim_tidydb, when checking for the continued existence of a message for
6370 which it has found a message-specific retry record, was not finding
6371 messages that were in split spool directories. Consequently, it was
6372 deleting retry records that should have stayed in existence.
6374 29. eximstats updated to version 1.35
6375 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6376 1.35 - bugfix such that pie charts by volume are generated correctly
6378 30. The SPA authentication driver was not abandoning authentication and moving
6379 on to the next authenticator when an expansion was forced to fail,
6380 contradicting the general specification for all authenticators. Instead it
6381 was generating a temporary error. It now behaves as specified.
6383 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6384 (the order specifies the preference for clients). The order is now AES256,
6385 AES128, 3DES, ARCFOUR128.
6387 31. Small patch to Sieve code - explicitly set From: when generating an
6390 32. Exim crashed if a remote delivery caused a very long error message to be
6391 recorded - for instance if somebody sent an entire SpamAssassin report back
6392 as a large number of 550 error lines. This bug was coincidentally fixed by
6393 increasing the size of one of Exim's internal buffers (big_buffer) that
6394 happened as part of the Exiscan merge. However, to be on the safe side, I
6395 have made the code more robust (and fixed the comments that describe what
6398 33. Some experimental protocols are using DNS PTR records for new purposes. The
6399 keys for these records are domain names, not reversed IP addresses. The
6400 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6401 leaves it alone. Component reversal etc. now happens only for IP addresses.
6404 34. The host_aton() function is supposed to be passed a string that is known
6405 to be a valid IP address. However, in the case of IPv6 addresses, it was
6406 not checking this. This is a hostage to fortune. Exim now panics and dies
6407 if the condition is not met. A case was found where this could be provoked
6408 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6409 components; fortuitously, this particular loophole had already been fixed
6410 by change 4.50/55 or 4.44/33 above.
6412 If there are any other similar loopholes, the new check in host_aton()
6413 itself should stop them being exploited. The report I received stated that
6414 data on the command line could provoke the exploit when Exim was running as
6415 exim, but did not say which command line option was involved. All I could
6416 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6417 running as the user.
6420 35. There was a buffer overflow vulnerability in the SPA authentication code
6421 (which came originally from the Samba project). I have added a test to the
6422 spa_base64_to_bits() function which I hope fixes it.
6425 36. The daemon start-up calls getloadavg() while still root for those OS that
6426 need the first call to be done as root, but it missed one case: when
6427 deliver_queue_load_max is set with deliver_drop_privilege. This is
6428 necessary for the benefit of the queue runner, because there is no re-exec
6429 when deliver_drop_privilege is set.
6431 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6432 was in use, was not putting the data itself into the right store pool;
6433 consequently, it could be overwritten for a subsequent message in the same
6434 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6437 38. Sometimes the final signoff response after QUIT could fail to get
6438 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6439 < 0 before doing a fflush(). This bug looks as though it goes back to the
6440 introduction of TLS in release 3.20, but "sometimes" must have been rare
6441 because the tests only now provoked it.
6447 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6448 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6449 Now it returns an integer. A number of calls were still expecting a T/F
6450 return. Fortuitously, in all cases, the tests worked in OK situations,
6451 which is the norm. However, things would have gone wrong on any write
6452 failures on the smtp file descriptor. This function is used when sending
6453 messages over SMTP and also when doing verify callouts.
6455 2. When Exim is called to do synchronous delivery of a locally submitted
6456 message (the -odf or -odi options), it no longer closes stderr before doing
6459 3. Implemented the mua_wrapper option.
6461 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6463 5. Implemented the functions header_remove(), header_testname(),
6464 header_add_at_position(), and receive_remove_recipient(), and exported them
6467 6. If an ACL "warn" statement specified the addition of headers, Exim already
6468 inserted X-ACL-Warn: at the start if there was no header name. However, it
6469 was not making this test for the second and subsequent header lines if
6470 there were newlines in the string. This meant that an invalid header could
6471 be inserted if Exim was badly configured.
6473 7. Allow an ACL "warn" statement to add header lines at the start or after all
6474 the Received: headers, as well as at the end.
6476 8. Added the rcpt_4xx retry error code.
6478 9. Added postmaster_mailfrom=xxx to callout verification option.
6480 10. Added mailfrom=xxxx to the callout verification option, for verify=
6483 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6484 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6486 12. Inserted some casts to stop certain compilers warning when using pointer
6487 differences as field lengths or precisions in printf-type calls (mostly
6488 affecting debugging statements).
6490 13. Added optional readline() support for -be (dynamically loaded).
6492 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6493 same clock tick as a message's arrival, so that its received time was the
6494 same as the "first fail" time on the retry record, and that message
6495 remained on the queue past the ultimate address timeout, every queue runner
6496 would try a delivery (because it was past the ultimate address timeout) but
6497 after another failure, the ultimate address timeout, which should have then
6498 bounced the address, did not kick in. This was a "< instead of <=" error;
6499 in most cases the first failure would have been in the next clock tick
6500 after the received time, and all would be well.
6502 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6503 being recognized when the domain list was tested by the match_domain
6504 condition in an expansion string.
6506 16. Added the ${str2b64: operator.
6508 17. Exim was always calling setrlimit() to set a large limit for the number of
6509 processes, without checking whether the existing limit was already
6510 adequate. (It did check for the limit on file descriptors.) Furthermore,
6511 errors from getrlimit() and setrlimit() were being ignored. Now they are
6512 logged to the main and panic logs, but Exim does carry on, to try to do its
6513 job under whatever limits there are.
6515 18. Imported PCRE 5.0.
6517 19. Trivial typo in log message " temporarily refused connection" (the leading
6520 20. If the log selector return_path_on_delivery was set and an address was
6521 redirected to /dev/null, the delivery process crashed because it assumed
6522 that a return path would always be set for a "successful" delivery. In this
6523 case, the whole delivery is bypassed as an optimization, and therefore no
6526 21. Internal re-arrangement: the function for sending a challenge and reading
6527 a response while authentication was assuming a zero-terminated challenge
6528 string. It's now changed to take a pointer and a length, to allow for
6529 binary data in such strings.
6531 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6533 23. Exim was not respecting finduser_retries when seeking the login of the
6534 uid under which it was called; it was always trying 10 times. (The default
6535 setting of finduser_retries is zero.) Also, it was sleeping after the final
6536 failure, which is pointless.
6538 24. Implemented tls_on_connect_ports.
6540 25. Implemented acl_smtp_predata.
6542 26. If the domain in control=submission is set empty, Exim assumes that the
6543 authenticated id is a complete email address when it generates From: or
6544 Sender: header lines.
6546 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6547 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6548 chown and chgrp in /bin and hostname in /usr/bin.
6550 28. Exim was keeping the "process log" file open after each use, just as it
6551 does for the main log. This opens the possibility of it remaining open for
6552 long periods when the USR1 signal hits a daemon. Occasional processlog
6553 errors were reported, that could have been caused by this. Anyway, it seems
6554 much more sensible not to leave this file open at all, so that is what now
6557 29. The long-running daemon process does not normally write to the log once it
6558 has entered its main loop, and it closes the log before doing so. This is
6559 so that log files can straightforwardly be renamed and moved. However,
6560 there are a couple of unusual error situations where the daemon does write
6561 log entries, and I had neglected to close the log afterwards.
6563 30. The text of an SMTP error response that was received during a remote
6564 delivery was being truncated at 512 bytes. This is too short for some of
6565 the long messages that one sometimes sees. I've increased the limit to
6568 31. It is now possible to make retry rules that apply only when a message has a
6569 specific sender, in particular, an empty sender.
6571 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6572 it possible to be selective about when SMTP synchronization is enforced.
6574 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6576 32. Implemented hosts_connection_nolog.
6578 33. Added an ACL for QUIT.
6580 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6583 35. Added mailbox_size and mailbox_filecount to appendfile.
6585 36. Added control = no_multiline_responses to ACLs.
6587 37. There was a bug in the logic of the code that waits for the clock to tick
6588 in the case where the clock went backwards by a substantial amount such
6589 that the microsecond fraction of "now" was more than the microsecond
6590 fraction of "then" (but the whole seconds number was less).
6592 38. Added support for the libradius Radius client library this is found on
6593 FreeBSD (previously only the radiusclient library was supported).
6599 1. When certain lookups returned multiple values in the form name=value, the
6600 quoting of the values was not always being done properly. Specifically:
6601 (a) If the value started with a double quote, but contained no whitespace,
6603 (b) If the value contained whitespace other than a space character (i.e.
6604 tabs or newlines or carriage returns) it was not quoted.
6605 This fix has been applied to the mysql and pgsql lookups by writing a
6606 separate quoting function and calling it from the lookup code. The fix
6607 should probably also be applied to nisplus, ibase and oracle lookups, but
6608 since I cannot test any of those, I have not disturbed their existing code.
6610 2. A hit in the callout cache for a specific address caused a log line with no
6611 reason for rejecting RCPT. Now it says "Previous (cached) callout
6612 verification failure".
6614 3. There was an off-by-one bug in the queryprogram router. An over-long
6615 return line was truncated at 256 instead of 255 characters, thereby
6616 overflowing its buffer with the terminating zero. As well as fixing this, I
6617 have increased the buffer size to 1024 (and made a note to document this).
6619 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6620 when Exim is waiting for an SMTP response from a remote server, Exim
6621 restarts its select() call on the socket, thereby resetting its timeout.
6622 This is not a problem when such interrupts are rare. Somebody set up a cron
6623 job to run exiwhat every 2 minutes, which is less than the normal select()
6624 timeout (5 or 10 minutes). This meant that the select() timeout never
6625 kicked in because it was always reset. I have fixed this by comparing the
6626 time when an interrupt arrives with the time at the start of the first call
6627 to select(). If more time than the timeout has elapsed, the interrupt is
6628 treated as a timeout.
6630 5. Some internal re-factoring in preparation for the addition of Sieve
6631 extensions (by MH). In particular, the "personal" test is moved to a
6632 separate function, and given an option for scanning Cc: and Bcc: (which is
6633 not set for Exim filters).
6635 6. When Exim created an email address using the login of the caller as the
6636 local part (e.g. when creating a From: or Sender: header line), it was not
6637 quoting the local part when it contained special characters such as @.
6639 7. Installed new OpenBSD configuration files.
6641 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6642 try to make them clearer.
6644 9. Callout options, other than the timeout value, were being ignored when
6645 verifying sender addresses in header lines. For example, when using
6647 verify = header_sender/callout=no_cache
6649 the cache was (incorrectly) being used.
6651 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6652 only the Cygwin environment.
6654 11. When return_path_on_delivery was set as a log selector, if different remote
6655 addresses in the same message used different return paths and parallel
6656 remote delivery occurred, the wrong values would sometimes be logged.
6657 (Whenever a remote delivery process finished, the return path value from
6658 the most recently started remote delivery process was logged.)
6660 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6661 lines when AUTH and/or TLS are in use. This is the "received protocol"
6662 field. Exim used to use "asmtp" for authenticated SMTP, without any
6663 indication (in the protocol name) for TLS use. Now it follows the RFC and
6664 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6665 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6666 names appear in log lines as well as in Received: header lines.
6668 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6669 extensions, and comparison tests, and to fix some bugs.
6671 14. Changes to the "personal" filter test:
6673 (1) The test was buggy in that it was just doing the equivalent of
6674 "contains" tests on header lines. For example, if a user's address was
6675 anne@some.where, the "personal" test would incorrectly be true for
6677 To: susanne@some.where
6679 This test is now done by extracting each address from the header in turn,
6680 and checking the entire address. Other tests that are part of "personal"
6681 are now done using regular expressions (for example, to check local parts
6682 of addresses in From: header lines).
6684 (2) The list of non-personal local parts in From: addresses has been
6685 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6686 taken from the Sieve specification recommendations.
6688 (3) If the message contains any header line starting with "List-" it is
6689 treated as non-personal.
6691 (4) The test for "circular" in the Subject: header line has been removed
6692 because it now seems ill-conceived.
6694 15. Minor typos in src/EDITME comments corrected.
6696 16. Installed latest exipick from John Jetmore.
6698 17. If headers_add on a router specified a text string that was too long for
6699 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6700 of string_sprintf() is now avoided.
6702 18. $message_body_size was not set (it was always zero) when running the DATA
6703 ACL and the local_scan() function.
6705 19. For the "mail" command in an Exim filter, no default was being set for
6706 the once_repeat time, causing a random time value to be used if "once" was
6707 specified. (If the value happened to be <= 0, no repeat happened.) The
6708 default is now 0s, meaning "never repeat". The "vacation" command was OK
6709 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6710 (I found it when inspecting the code).
6712 20. There is now an overall timeout for performing a callout verification. It
6713 defaults to 4 times the callout timeout, which applies to individual SMTP
6714 commands during the callout. The overall timeout applies when there is more
6715 than one host that can be tried. The timeout is checked before trying the
6716 next host. This prevents very long delays if there are a large number of
6717 hosts and all are timing out (e.g. when the network connections are timing
6718 out). The value of the overall timeout can be changed by specifying an
6719 additional sub-option for "callout", called "maxwait". For example:
6721 verify = sender/callout=5s,maxwait=20s
6723 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6724 to the end before writing, but this should make it even safer).
6726 22. Exim was forgetting that it had advertised PIPELINING for the second and
6727 subsequent messages on an SMTP connection. It was also not resetting its
6728 memory on STARTTLS and an internal HELO.
6730 23. When Exim logs an SMTP synchronization error within a session, it now
6731 records whether PIPELINING has been advertised or not.
6733 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6734 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6735 rather than long int.
6737 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6739 26. Added the never_mail option to autoreply.
6745 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6746 crash if the getsockname() call failed; this can happen if a connection is
6747 closed very soon after it is established. The problem was simply in the
6748 order in which certain operations were done, causing Exim to try to write
6749 to the SMTP stream before it had set up the file descriptor. The bug has
6750 been fixed by making things happen in the correct order.
6756 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6757 before the connection was closed, thus losing the rejection response.
6759 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6760 some early Solaris releases, but causes trouble in current releases where
6761 socklen_t is defined.
6763 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6766 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6769 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6770 that releasing the top of it at the end releases what was used for sub-
6771 expansions (unless the block got too big). However, discard this block if
6772 the first thing is a variable or header, so that we can use its block when
6773 it is dynamic (useful for very large $message_headers, for example).
6775 6. Lookups now cache *every* query, not just the most recent. A new, separate
6776 store pool is used for this. It can be recovered when all lookup caches are
6777 flushed. Lookups now release memory at the end of their result strings.
6778 This has involved some general refactoring of the lookup sources.
6780 7. Some code has been added to the store_xxx() functions to reduce the amount
6781 of flapping under certain conditions.
6783 8. log_incoming_interface used to affect only the <= reception log lines. Now
6784 it causes the local interface and port to be added to several more SMTP log
6785 lines, for example "SMTP connection from", and rejection lines.
6787 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6789 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6791 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6792 could be overwritten at the end of the current message (or the start of a
6793 new message if it was set in a HELO ACL). The value is now preserved for
6794 the duration of the SMTP connection.
6796 12. If a transport had a headers_rewrite setting, and a matching header line
6797 contained an unqualified address, that address was qualified, even if it
6798 did not match any rewriting rules. The underlying bug was that the values
6799 of the flags that permit the existence of unqualified sender and recipient
6800 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6801 non-local messages, and by -bnq for local messages) were not being
6802 preserved with the message after it was received.
6804 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6805 "next input=" as part of the text comprising the host identity instead of
6806 the correct text. The code was using the same buffer for two different
6807 strings. However, depending on which order the printing function evaluated
6808 its arguments, the bug did not always show up. Under Linux, for example, my
6809 test suite worked just fine.
6811 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6812 doesn't work with some older versions of Perl. It has been changed to "my",
6813 which in any case is probably the better facility to use.
6815 15. A really picky compiler found some instances of statements for creating
6816 error messages that either had too many or two few arguments for the format
6819 16. The size of the buffer for calls to the DNS resolver has been increased
6820 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6821 for addresses that have a lot of PTR records. This alleviates a problem; it
6822 does not fully solve it.
6824 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6825 buffer now truncates the list and logs the incident, which is the same
6826 action as happens when Exim is looking up a host name and its aliases.
6827 Previously in this situation something unpredictable would happen;
6828 sometimes it was "internal error: store_reset failed".
6830 18. If a server dropped the connection unexpectedly when an Exim client was
6831 using GnuTLS and trying to read a response, the client delivery process
6832 crashed while trying to generate an error log message.
6834 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6835 string, for example:
6837 warn message = H1: something\nH2: something
6839 the text was added as a single header line from Exim's point of view
6840 though it ended up OK in the delivered message. However, searching for the
6841 second and subsequent header lines using $h_h2: did not work. This has been
6842 fixed. Similarly, if a system filter added multiple headers in this way,
6843 the routers could not see them.
6845 20. Expanded the error message when iplsearch is called with an invalid key to
6846 suggest using net-iplsearch in a host list.
6848 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6849 ACLs are no longer actually imposed (and a message to that effect is
6852 22. If a "gecos" field in a passwd entry contained escaped characters, in
6853 particular, if it contained a \" sequence, Exim got it wrong when building
6854 a From: or a Sender: header from that name. A second bug also caused
6855 incorrect handling when an unquoted " was present following a character
6856 that needed quoting.
6858 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6859 was not being matched caselessly.
6861 24. Arranged for all hyphens in the exim.8 source to be escaped with
6864 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6865 properly. Recipient callout cache records were still being keyed to include
6866 the sender, even when use_sender was set false. This led to far more
6867 callouts that were necessary. The sender is no longer included in the key
6868 when use_sender is false.
6870 26. Added "control = submission" modifier to ACLs.
6872 27. Added the ${base62d: operator to decode base 62 numbers.
6874 28. dnsdb lookups can now access SRV records.
6876 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6877 the configuration file.
6879 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6880 (-v) mode. This makes the output for a verbose queue run more intelligible.
6882 31. Added a use_postmaster feature to recipient callouts.
6884 32. Added the $body_zerocount variable, containing the number of binary zero
6885 bytes in the message body.
6887 33. The time of last modification of the "new" subdirectory is now used as the
6888 "mailbox time last read" when there is a quota error for a maildir
6891 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6893 35. Added +ignore_unknown as a special item in host lists.
6895 36. Code for decoding IPv6 addresses in host lists is now included, even if
6896 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6897 address was recognized as an IP address, but was then not correctly decoded
6898 into binary, causing unexpected and incorrect effects when compared with
6905 1. Very minor rewording of debugging text in manualroute to say "list of
6906 hosts" instead of "hostlist".
6908 2. If verify=header_syntax was set, and a header line with an unqualified
6909 address (no domain) and a large number of spaces between the end of the
6910 name and the colon was received, the reception process suffered a buffer
6911 overflow, and (when I tested it) crashed. This was caused by some obsolete
6912 code that should have been removed. The fix is to remove it!
6914 3. When running in the test harness, delay a bit after writing a bounce
6915 message to get a bit more predictability in the log output.
6917 4. Added a call to search_tidyup() just before forking a reception process. In
6918 theory, someone could use a lookup in the expansion of smtp_accept_max_
6919 per_host which, without the tidyup, could leave open a database connection.
6921 5. Added the variables $recipient_data and $sender_data which get set from a
6922 lookup success in an ACL "recipients" or "senders" condition, or a router
6923 "senders" option, similar to $domain_data and $local_part_data.
6925 6. Moved the writing of debug_print from before to after the "senders" test
6928 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6929 problems for message scanning, either using a data ACL, or using
6930 local_scan() because the Received: header was not generated till after they
6931 were called (in order to set the time as the time of reception completion).
6932 I have revised the way this works. The header is now generated after the
6933 body is received, but before the ACL or local_scan() are called. After they
6934 are run, the timestamp in the header is updated.
6940 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6941 before starting a queue runner without re-exec. This happened only when
6942 deliver_drop_privilege was set or when the Exim user was set to root. The
6943 effect of the bug was that timeouts during subsequent deliveries caused
6944 crashes instead of being properly handled. The handler is now left at its
6945 default (and expected) setting.
6947 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6948 message, again when deliver_drop_privilege is set or Exim is run as root.
6949 The bug described in (1) was not present in this case, but the tidying up
6950 of the other signals was missing. I have made the two cases consistent.
6952 3. The ignore_target_hosts setting on a manualroute router was being ignored
6953 for hosts that were looked up using the /MX notation.
6955 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6958 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6959 operated on the sender address. After changing the $sender_address to <>
6960 for the sender address verify, Exim was re-instated it as the original
6961 (before rewriting) address, but remembering that it had rewritten it, so it
6962 wasn't rewriting it again. This bug also had the effect of breaking the
6963 sender address verification caching when the sender address was rewritten.
6965 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6966 This has been changed so that if the ip literal address matches
6967 ignore_target_hosts, the router declines.
6969 7. Added expansion conditions match_domain, match_address, and match_local_
6970 part (NOT match_host).
6972 8. The placeholder for the Received: header didn't have a length field set.
6974 9. Added code to Exim itself and to exim_lock to test for a specific race
6975 condition that could lead to file corruption when using MBX delivery. The
6976 issue is with the lockfile that is created in /tmp. If this file is removed
6977 after a process has opened it but before that process has acquired a lock,
6978 there is the potential for a second process to recreate the file and also
6979 acquire a lock. This could lead to two Exim processes writing to the file
6980 at the same time. The added code performs the same test as UW imapd; it
6981 checks after acquiring the lock that its file descriptor still refers to
6982 the same named file.
6984 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6985 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6988 11. Added the smtp_active_hostname option. If used, this will typically be made
6989 to depend on the incoming interface address. Because $interface_address is
6990 not set up until the daemon has forked a reception process, error responses
6991 that can happen earlier (such as "too many connections") no longer contain
6994 12. If an expansion in a condition on a "warn" statement fails because a lookup
6995 defers, the "warn" statement is abandoned, and the next ACL statement is
6996 processed. Previously this caused the whole ACL to be aborted.
6998 13. Added the iplsearch lookup type.
7000 14. Added ident_timeout as a log selector.
7002 15. Added tls_certificate_verified as a log selector.
7004 16. Added a global option tls_require_ciphers (compare the smtp transport
7005 option of the same name). This controls incoming TLS connections.
7007 17. I finally figured out how to make tls_require_ciphers do a similar thing
7008 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7009 before starting the TLS session.
7011 18. Tabs are now shown as \t in -bP output.
7013 19. If the log selector return_path_on_delivery was set, Exim crashed when
7014 bouncing a message because it had too many Received: header lines.
7016 20. If two routers both had headers_remove settings, and the first one included
7017 a superfluous trailing colon, the final name in the first list and the
7018 first name in the second list were incorrectly joined into one item (with a
7019 colon in the middle).
7025 1. Added -C and -D options to the exinext utility, mainly to make it easier
7026 to include in the automated testing, but these could be helpful when
7027 multiple configurations are in use.
7029 2. The exinext utility was not formatting the output nicely when there was
7030 an alternate port involved in the retry record key, nor when there was a
7031 message id as well (for retries that were specific to a specific message
7032 and a specific host). It was also confused by IPv6 addresses, because of
7033 the additional colons they contain. I have fixed the IPv4 problem, and
7034 patched it up to do a reasonable job for IPv6.
7036 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7037 delivery, the log line now contains "pipelined" if PIPELINING was used.
7039 4. An SMTP transport process used to panic and die if the bind() call to set
7040 an explicit outgoing interface failed. This has been changed; it is now
7041 treated in the same way as a connect() failure.
7043 5. A reference to $sender_host_name in the part of a conditional expansion
7044 that was being skipped was still causing a DNS lookup. This no longer
7047 6. The def: expansion condition was not recognizing references to header lines
7048 that used bh_ and bheader_.
7050 7. Added the _cache feature to named lists.
7052 8. The code for checking quota_filecount in the appendfile transport was
7053 allowing one more file than it should have been.
7055 9. For compatibility with Sendmail, the command line option
7063 and sets the incoming protocol and host name (for trusted callers). The
7064 host name and its colon can be omitted when only the protocol is to be set.
7065 Note the Exim already has two private options, -pd and -ps, that refer to
7066 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7067 "s", but I don't think that's a major issue.
7069 10. A number of refactoring changes to the code, none of which should affect
7072 (a) The number of logging options was getting close to filling up the
7073 32-bit word that was used as a bit map. I have split them into two classes:
7074 those that are passed in the argument to log_write(), and those that are
7075 only ever tested independently outside of that function. These are now in
7076 separate 32-bit words, so there is plenty of room for expansion again.
7077 There is no change in the user interface or the logging behaviour.
7079 (b) When building, for example, log lines, the code previously used a
7080 macro that called string_cat() twice, in order to add two strings. This is
7081 not really sufficiently general. Furthermore, there was one instance where
7082 it was actually wrong because one of the argument was used twice, and in
7083 one call a function was used. (As it happened, calling the function twice
7084 did not affect the overall behaviour.) The macro has been replaced by a
7085 function that can join an arbitrary number of extra strings onto a growing
7088 (c) The code for expansion conditions now uses a table and a binary chop
7089 instead of a serial search (which was left over from when there were very
7090 few conditions). Also, it now recognizes conditions like "pam" even when
7091 the relevant support is not compiled in: a suitably worded error message is
7092 given if an attempt is made to use such a condition.
7094 11. Added ${time_interval:xxxxx}.
7096 12. A bug was causing one of the ddress fields not to be passed back correctly
7097 from remote delivery subprocesses. The field in question was not being
7098 subsequently used, so this caused to problems in practice.
7100 13. Added new log selectors queue_time and deliver_time.
7102 14. Might have fixed a bug in maildirsizefile handling that threw up
7103 "unexpected character" debug warnings, and recalculated the data
7104 unnecessarily. In any case, I expanded the warning message to give more
7107 15. Added the message "Restricted characters in address" to the statements in
7108 the default ACL that block characters like @ and % in local parts.
7110 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7111 Three changes have been made:
7113 (a) There was a serious bug; a negative response to MAIL caused the whole
7114 recipient domain to be cached as invalid, thereby blocking all messages
7115 to all local parts at the same domain, from all senders. This bug has
7116 been fixed. The domain is no longer cached after a negative response to
7117 MAIL if the sender used is not empty.
7119 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7122 (c) A new callout option, "use_sender" has been added for people who want
7123 the modified behaviour.
7129 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7132 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7133 indeed breaks things for older releases.
7135 3. Added additional logging to the case where there is a problem reading data
7136 from a filter that is running in a subprocess using a pipe, in order to
7137 try to track down a specific problem.
7139 4. Testing facility fudge: when running in the test harness and attempting
7140 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7141 getting "No route to host". Convert this to a timeout.
7143 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7146 6. Some OS don't have socklen_t but use size_t instead. This affects the
7147 fifth argument of getsockopt() amongst other things. This is now
7148 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7149 can be set for individual OS. I have set it for SunOS5, OSF1, and
7150 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7151 some earlier ones do not.
7153 7. Change 4.30/15 was not doing the test caselessly.
7155 8. The standard form for an IPv6 address literal was being rejected by address
7156 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7157 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7158 this, as well as the form without the "IPv6" on the front (but only when
7159 address literals are enabled, of course).
7161 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7163 10. Exim crashed if a message with an empty sender address specified by -f
7164 encountered a router with an errors_to setting. This could be provoked only
7165 by a command such as
7169 where an empty string was supplied; "<>" did not hit this bug.
7171 11. Installed PCRE release 4.5.
7173 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7174 remained set. It is now erased.
7176 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7177 times from message ids (which are base 36 rather than the normal 62).
7179 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7180 were being counted as actual protocol errors, and logged if the log
7181 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7182 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7183 and DATA following a set of rejected RCPTs do not count as protocol errors.
7184 In other words, Exim assumes they were pipelined, though this may not
7185 actually be the case. Of course, in all cases the client gets an
7186 appropriate error code.
7188 15. If a lookup fails in an ACL condition, a message about the failure may
7189 be available; it is used if testing the ACL cannot continue, because most
7190 such messages specify what the cause of the deferral is. However, some
7191 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7192 that caused an old message to be retained and used if a later statement
7193 caused a defer, replacing the real cause of the deferral.
7195 16. If an IP address had so many PTR records that the DNS lookup buffer
7196 was not large enough to hold them, Exim could crash while trying to process
7197 the truncated data. It now detects and logs this case.
7199 17. Further to 4.21/58, another change has been made: if (and only if) the
7200 first line of a message (the first header line) ends with CRLF, a bare LF
7201 in a subsequent header line has a space inserted after it, so as not to
7202 terminate the header.
7204 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7205 unnecessarily, used atoi() instead of strtol(), and didn't check the
7206 termination when getting file sizes from file names by regex.
7208 19. Completely re-implemented the support for maildirsize files, in the light
7209 of a number of problems with the previous contributed implementation
7210 (4.30/29). In particular:
7212 . If the quota is zero, the maildirsize file is maintained, but no quota is
7215 . If the maildir directory does not exist, it is created before any attempt
7216 to write a maildirsize file.
7218 . The quota value in the file is just a cache; if the quota is changed in
7219 the transport, the new value overrides.
7221 . A regular expression is available for excluding directories from the
7224 20. The autoreply transport checks the characters in options that define the
7225 message's headers; it allows continued headers, but it was checking with
7226 isspace() after an embedded newline instead of explicitly looking for a
7229 21. If all the "regular" hosts to which an address was routed had passed their
7230 expiry times, and had not reached their retry times, the address was
7231 bounced, even if fallback hosts were defined. Now Exim should go on to try
7234 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7235 equivalent code in the SMTP transport. Some hosts send humungous responses
7236 to HELO/EHLO, more than 1024 it seems.
7238 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7239 gives compiler warnings in some environments. I've now done it "properly",
7242 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7243 (because of problems with the built-in one) was declared to return uschar *
7244 instead of char *, causing compiler failure.
7246 25. Fixed a file descriptor leak when processing alias/forward files.
7248 26. Fixed a minor format string issue in dbfn.c.
7250 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7252 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7253 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7256 29. When a sender address is verified, it is cached, to save repeating the test
7257 when there is more than one recipient in a message. However, when the
7258 verification involves a callout, it is possible for different callout
7259 options to be set for different recipients. It is too complicated to keep
7260 track of this in the cache, so now Exim always runs a verification when a
7261 callout is required, relying on the callout cache for the optimization.
7262 The overhead is duplication of the address routing, but this should not be
7265 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7266 to be verified with callout=postmaster, and the main callout worked but the
7267 postmaster check failed, the verification correctly failed. However, if a
7268 subsequent RCPT command asked for sender verification *without* the
7269 postmaster check, incorrect caching caused this verification also to fail,
7272 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7273 it was not caching the DNS options (qualify_single, search_parents) that
7274 were used when the lookup failed. A subsequent lookup with different
7275 options therefore always gave the same answer, though there were cases
7276 where it should not have. (Example: a "domains = !$mx_any" option on a
7277 dnslookup router: the "domains" option is always processed without any
7278 widening, but the router might have qualify_single set.) Now Exim uses the
7279 cached value only when the same options are set.
7281 32. Added John Jetmore's "exipick" utility to the distribution.
7283 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7284 than a timeout (e.g. a certificate is required, and is not provided), an
7285 Exim server now closes the connection immediately. Previously it waited for
7286 the client to close - but if the client is SSL, it seems that they each
7287 wait for each other, leading to a delay before one of them times out.
7289 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7290 maintained 0.8.x compatibility because I don't think many are using it, and
7291 it is clearly obsolete.
7293 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7296 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7297 client certificate was expired. A simple patch fixes this, though I don't
7298 understand the full logic of why the verify callback is called multiple
7301 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7302 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7303 which causes problems with some clients (such as the Certicom SSL Plus
7304 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7305 disables the coutermeasure allowing Eudora to connect."
7307 38. Exim was not checking that a write() to a log file succeeded. This could
7308 lead to Bad Things if a log got too big, in particular if it hit a file
7309 size limit. Exim now panics and dies if it cannot write to a log file, just
7310 as it does if it cannot open a log file.
7312 39. Modified OS/Makefile-Linux so that it now contains
7314 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7316 The two -D definitions ensure that Exim is compiled with large file
7317 support, which makes it possible to handle log files that are bigger than
7320 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7321 instance) a domain was checked against a named list that involved a lookup,
7322 causing $domain_data to be set, then another domain was checked against the
7323 same list, then the first domain was re-checked, the value of $domain_data
7324 after the final check could be wrong. In particular, if the second check
7325 failed, it could be set empty. This bug probably also applied to
7328 41. The strip_trailing_dot option was not being applied to the address given
7329 with the -f command-line option.
7331 42. The code for reading a message's header from the spool was incrementing
7332 $received_count, but never initializing it. This meant that the value was
7333 incorrect (doubled) while delivering a message in the same process in which
7334 it was received. In the most common configuration of Exim, this never
7335 happens - a fresh exec is done - but it can happen when
7336 deliver_drop_privilege is set.
7338 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7339 it now includes up to 150 characters of the unexpected data in the log
7342 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7343 and building data strings. The size of both of these buffers was 10 000
7344 bytes - far larger than anybody would *ever* want, thought I. Needless to
7345 say, somebody hit the limit. I have increased the maximum line length to
7346 20 000 and the maximum data length of concatenated lines to 100 000. I have
7347 also fixed two bugs, because there was no checking on these buffers. Tsk,
7348 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7349 buffer is too small.
7351 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7352 lsearch lookups. Now it does.
7354 46. When parsing a route_list item in a manualroute router, a fixed-length
7355 buffer was used for the list of hosts. I made this 1024 bytes long,
7356 thinking that nobody would ever have a list of hosts that long. Wrong.
7357 Somebody had a whole pile of complicated expansion conditions, and the
7358 string was silently truncated, leading to an expansion error. It turns out
7359 that it is easier to change to an unlimited length (owing to other changes
7360 that have happened since this code was originally written) than to build
7361 structure for giving a limitation error. The length of the item that
7362 expands into the list of hosts is now unlimited.
7364 47. The lsearch lookup could not handle data where the length of text line was
7365 more than 4095 characters. Such lines were truncated, leading to shortened
7366 data being returned. It should now handle lines of any length.
7368 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7369 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7372 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7373 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7374 no longer happens in comment lines. A list of replacements is now placed
7375 at the head of all of the source files, except those whose only change is
7376 to replace PERL_COMMAND in the very first #! line.
7378 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7379 messages on the queue, with a bottom-up merge sort, using code contributed
7380 by Michael Haardt. This should make operations like -bp somewhat faster on
7381 large queues. It won't affect queue runners, except when queue_run_in_order
7384 51. Installed eximstats 1.31 in the distribution.
7386 52. Added support for SRV lookups to the dnslookup router.
7388 53. If an ACL referred to $message_body or $message_body_end, the value was not
7389 reset for any messages that followed in the same SMTP session.
7391 54. The store-handling optimization for building very long strings was not
7392 differentiating between the different store pools. I don't think this
7393 actually made any difference in practice, but I've tidied it.
7395 55. While running the routers to verify a sender address, $sender_address
7396 was still set to the sender address. This is wrong, because when routing to
7397 send a bounce to the sender, it would be empty. Therefore, I have changed
7398 it so that, while verifying a sender address, $sender_address is set to <>.
7399 (There is no change to what happens when verifying a recipient address.)
7401 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7402 target A or AAAA records (if not already returned) without resetting the
7403 qualify_single or search_parents options of the DNS resolver. These are
7404 inappropriate in this case because the targets of MX and SRV records must
7405 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7406 target that, when qualified, matched something in the local domain. These
7407 two options are now turned off when doing these lookups.
7409 57. It seems that at least some releases of Reiserfs (which does not have the
7410 concept of a fixed number of inodes) returns zero and not -1 for the
7411 number of available inodes. This interacted badly with check_spool_inodes,
7412 which assumed that -1 was the "no such thing" setting. What I have done is
7413 to check that the total number of inodes is greater than zero before doing
7414 the test of how many are available.
7416 58. When a "warn" ACL statement has a log_message modifier, the message is
7417 remembered, and not repeated. This is to avoid a lot of repetition when a
7418 message has many recipients that cause the same warning to be written.
7419 However, Exim was preserving the list of already written lines for an
7420 entire SMTP session, which doesn't seem right. The memory is now reset if a
7421 new message is started.
7423 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7424 debugging output unless log_rewrite was also set.
7426 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7427 of (int)(handle) when we know that handle contains (void *)(-1).
7429 61. The Exim daemon panic-logs an error return when it closes the incoming
7430 connection. However "connection reset by peer" seems to be common, and
7431 isn't really an error worthy of noting specially, so that particular error
7434 62. When Exim is trying to find all the local interfaces, it used to panic and
7435 die if the ioctl to get the interface flags failed. However, it seems that
7436 on at least one OS (Solaris 9) it is possible to have an interface that is
7437 included in the list of interfaces, but for which you get a failure error
7438 for this call. This happens when the interface is not "plumbed" into a
7439 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7440 failure of the "get flags" call assumes that the interface is down.
7442 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7443 makes life easier for people who are doing arithmetic on fields extracted
7444 from dates, where you often get leading zeros that should not be
7445 interpreted as octal.
7447 64. Added qualify_domain to the redirect router, to override the global
7450 65. If a pathologically long header line contained very many addresses (the
7451 report of this problem mentioned 10 000) and each of them was rewritten,
7452 Exim could use up a very large amount of memory. (It kept on making new
7453 copies of the header line as it rewrote, and never released the old ones.)
7454 At the expense of a bit more processing, the header rewriting function has
7455 been changed so that it no longer eats memory in this way.
7457 66. The generation of the Received: header has been moved from the time that a
7458 message starts to be received, to the time that it finishes. The timestamp
7459 in the Received: header should now be very close to that of the <= log
7460 line. There are two side-effects of this change:
7462 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7463 logged header lines no longer include the local Received: line, because
7464 it has not yet been created. The same applies to a copy of the message
7465 that is returned to a non-SMTP sender when a message is rejected.
7467 (b) When a filter file is tested using -bf, no additional Received: header
7468 is added to the test message. After some thought, I decided that this
7471 This change does not affect the value of $received_for. It is still set
7472 after address rewriting, but before local_scan() is called.
7474 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7476 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7477 gave an unhelpful panic error message, and a defer error. I have managed to
7478 change this behaviour so that it now rejects any supplied certificate,
7479 which seems right, as the list of acceptable certificates is empty.
7481 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7482 gave an unhelpful defer error. I have not managed to make this reject any
7483 supplied certificates, but the error message it gives is "no certificate
7484 supplied", which is not helpful.
7486 70. exigrep's output now also includes lines that are not associated with any
7487 message, but which match the given pattern. Implemented by a patch from
7488 Martin Sluka, which also tidied up the Perl a bit.
7490 71. Recipient callout verification, like sender verification, was using <> in
7491 the MAIL FROM command. This isn't really the right thing, since the actual
7492 sender may affect whether the remote host accepts the recipient or not. I
7493 have changed it to use the actual sender in the callout; this means that
7494 the cache record is now keyed on a recipient/sender pair, not just the
7495 recipient address. There doesn't seem to be a real danger of callout loops,
7496 since a callout by the remote host to check the sender would use <>.
7497 [SEE ABOVE: changed after hitting problems.]
7499 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7500 temporary errors. However, in the case of such a code being given after
7501 the end of a data transmission (i.e. after ".") Exim was failing to write
7502 a retry record for the message. (Yes, there was some broken host that was
7503 actually sending 8xx at this point.)
7505 73. An unknown lookup type in a host list could cause Exim to panic-die when
7506 the list was checked. (An example that provoked this was putting <; in the
7507 middle of a list instead of at the start.) If this happened during a DATA
7508 ACL check, a -D file could be left lying around. This kind of configuration
7509 error no longer causes Exim to die; instead it causes a defer error. The
7510 incident is still logged to the main and panic logs.
7512 74. Buglet left over from Exim 3 conversion. The message "too many messages
7513 in one connection" was written to the rejectlog but not the mainlog, except
7514 when address rewriting (yes!) was being logged.
7516 75. Added write_rejectlog option.
7518 76. When a system filter was run not as root (that is, when system_filter_user
7519 was set), the values of the $n variables were not being returned to the
7520 main process; thus, they were not subsequently available in the $sn
7523 77. Added +return_path_on_delivery log selector.
7525 78. A connection timeout was being treated differently from recipients deferred
7526 when testing hosts_max_try with a message that was older than the host's
7527 retry timeout. (The host should not be counted, thus allowing all hosts to
7528 be tried at least once before bouncing.) This may have been the cause of an
7529 occasionally reported bug whereby a message would remain on the queue
7530 longer than the retry timeout, but would be bounced if a delivery was
7531 forced. I say "may" because I never totally pinned down the problem;
7532 setting up timeout/retry tests is difficult. See also the next item.
7534 79. The ultimate address timeout was not being applied to errors that involved
7535 a combination of host plus message (for example, a timeout on a MAIL
7536 command). When an address resolved to a number of possible hosts, and they
7537 were not all tried for each delivery (e.g. because of hosts_max_try), a
7538 message could remain on the queue longer than the retry timeout.
7540 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7543 81. Fixed an obscure SMTP outgoing bug which required at least the following
7544 conditions: (a) there was another message waiting for the same server;
7545 (b) the server returned 5xx to all RCPT commands in the first message so
7546 that the message was not completed; (c) the server dropped the connection
7547 or gave a negative response to the RSET that Exim sends to abort the
7548 transaction. The observed case was a dropped connection after DATA that had
7549 been sent in pipelining mode. That is, the server had advertised PIPELINING
7550 but was not implementing it correctly. The effect of the bug was incorrect
7551 behaviour, such as trying another host, and this could lead to a crash.
7557 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7558 and daemon.c were passed as pointers to ints; they should have been
7559 pointers to socklen_t variables (which are typically unsigned ints).
7561 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7564 3. Fixed a really odd bug that affected only the testing scheme; patching a
7565 certain fixed string in the binary changed the value of another string that
7566 happened to be identical to the end of the original first string.
7568 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7569 name", it returns that address as the IP address. On some operating
7570 systems (e.g. Solaris), it also passes back the IP address string as the
7571 "host name". However, on others (e.g. Linux), it passes back an empty
7572 string. Exim wasn't checking for this, and was changing the host name to an
7573 empty string, assuming it had been canonicalized.
7575 5. Although rare, it is permitted to have more than one PTR record for a given
7576 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7577 all the names associated with an address, because they do in Solaris.
7578 However, it seems that they do not in Linux for data that comes from the
7579 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7580 I found this out when I moved to a new Linux workstation and tried to run
7581 the Exim test suite.
7583 To get round this problem I have changed the code so that it now does its
7584 own call to the DNS to look up PTR records when searching for a host name.
7585 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7586 addresses that are only in /etc/hosts are still found.
7588 This behaviour is, however, controlled by an option called host_lookup_
7589 order, which defaults to "bydns:byaddr". If people want to use the other
7590 order, or indeed, just use one or the other means of lookup, they can
7591 specify it in this variable.
7593 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7594 some operating systems, this comes back from gethostbyaddr() as an empty
7595 string, and this is what Exim used to test for. However, it seems that in
7596 other systems, "." is yielded. Exim now tests for this case too.
7598 7. The values of check_spool_space and check_log_space are now held internally
7599 as a number of kilobytes instead of an absolute number of bytes. If a
7600 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7601 kilobyte. This means that much larger values can be stored.
7603 8. Exim monitor: an attempt to get the action menu when not actually pointing
7604 at a message produces an empty menu entitled "No message selected". This
7605 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7606 no entries in it ("Shell widget menu has zero width and/or height"). So I
7607 have added a single, blank menu entry in this case.
7609 9. Added ${quote_local_part.
7611 10. MIME decoding is now applied to the contents of Subject: header lines when
7614 11. Now that a reference to $sender_host_address automatically causes a reverse
7615 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7616 host lookup before query-style lookups in lists that might use this
7617 variable. This has therefore been abolished, and the "net-" prefix is no
7618 longer necessary for query-style lookups.
7620 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7621 have been a typo for LFLAGS, so it has been changed.
7623 13. The install script calls Exim with "-C /dev/null" in order to find the
7624 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7625 to be output. However, since Exim outputs its version number before the
7626 error, it didn't break the script. It just looked ugly. I fixed this by
7627 always allowing "-C /dev/null" if the caller is root.
7629 14. Ignore overlarge ACL variable number when reading spool file - insurance
7630 against a later release with more variables having written the file.
7632 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7633 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7634 this, as well as the form without the "IPv6" on the front.
7636 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7637 OS/Makefile-Darwin file.
7639 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7640 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7642 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7643 variable. However, some OS use ints for inodes. I've added cast to long int
7644 to get rid of the compiler warning.
7646 19. I had forgotten to lock out "/../" in configuration file names when
7647 ALT_CONFIG_PREFIX was set.
7649 20. Routers used for verification do not need to specify transports. However,
7650 if such a router generated a host list, and callout was configured, Exim
7651 crashed, because it could not find a port number from the (non-existent)
7652 transport. It now assumes port 25 in this circumstance.
7654 21. Added the -t option to exigrep.
7656 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7657 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7658 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7659 them is set, LOOKUP_LSEARCH is forced.
7661 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7663 24. Added sender and host information to the "rejected by local_scan()" log
7664 line; previously there was no indication of these.
7666 25. Added .include_if_exists.
7668 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7669 Linux. It turns out that not all file systems support this. Apparently some
7670 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7671 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7672 not supported on the file descriptor, is now ignored when Exim is trying to
7673 sync a directory. This applies only to Linux.
7675 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7677 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7678 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7681 29. Added support for maildirsize files from supplied patch (modified a bit).
7683 30. The use of :fail: followed by an empty string could lead Exim to respond to
7684 sender verification failures with (e.g.):
7686 550 Verification failed for <xxx>
7687 550 Sender verify failed
7689 where the first response line was missing the '-' that indicates it is not
7690 the final line of the response.
7692 31. The loop for finding the name of the user that called Exim had a hardwired
7693 limit of 10; it now uses the value of finduser_retries, which is used for
7694 all other user lookups.
7696 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7699 33. Exim was neglecting to zero errno before one call of strtol() when
7700 expanding a string and expecting an integer value. On some systems this
7701 resulted in spurious "integer overflow" errors. Also, it was casting the
7702 result into an int without checking.
7704 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7705 did not work. The code looks as if it has *never* worked, though it appears
7706 to have been documented since at least release 1.62. I have made it work.
7708 35. The "timeout_DNS" error in retry rules, also documented since at least
7709 1.62, also never worked. As it isn't clear exactly what this means, and
7710 clearly it isn't a major issue, I have abolished the feature by treating it
7711 as "timeout", and writing a warning to the main and panic logs.
7713 36. The display of retry rules for -brt wasn't always showing the error code
7716 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7717 timeout_connect_A, timeout_connect_MX.
7719 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7720 to the empty sender.
7722 39. The daemon was not analysing the content of -oX till after it had closed
7723 stderr and disconnected from the controlling terminal. This meant that any
7724 syntax errors were only noted on the panic log, and the return code from
7725 the command was 0. By re-arranging the code a little, I've made the
7726 decoding happen first, so such errors now appear on stderr, and the return
7727 code is 1. However, the actual setting up of the sockets still happens in
7728 the disconnected process, so errors there are still only recorded on the
7731 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7732 connections (as happens on some IP stacks) was logged at start up time as
7733 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7734 it from "IPv6 and IPv4", which means that two separate sockets are being
7737 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7738 says whether AF_INET or AF_INET6 was passed as an argument.
7740 42. Exiwhat output was messed up when time zones were included in log
7743 43. Exiwhat now gives more information about the daemon's listening ports,
7744 and whether -tls-on-connect was used.
7746 44. The "port" option of the smtp transport is now expanded.
7748 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7749 silently ignored. Now an error message is written to the main and panic
7752 46. There's a new ACL modifier called "logwrite" which writes to a log file
7753 as soon as it is encountered.
7755 47. Added $local_user_uid and $local_user_gid at routing time.
7757 48. Exim crashed when trying to verify a sender address that was being
7760 49. Exim was recognizing only a space character after ".include". It now also
7761 recognizes a tab character.
7763 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7764 extracting the relevant information from the specification. The man page no
7765 longer contains scrambled data for the -d option, and I've added a section
7766 at the front about calling Exim under different names.
7768 51. Added "extra_headers" argument to the "mail" command in filter files.
7770 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7773 53. Installed eximstats 1.29.
7775 54. Added transport_filter_timeout as a generic transport option.
7777 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7778 Cc: header lines. This was required by RFC 822, but it not required by RFC
7781 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7782 incoming messages that did not have them. Now it does so only if the
7783 message originates locally, that is, if there is no associated remote host
7784 address. When Resent- header lines are present, this applies to the Resent-
7785 lines rather than the non-Resent- lines.
7787 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7788 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7790 58. Messages for configuration errors now include the name of the main
7791 configuration file - useful now that there may be more than one file in a
7792 list (.included file names were always shown).
7794 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7795 for those rare installations that do not start the daemon as root or run it
7796 setuid root. I've cut out the call to initgroups() if the daemon is not
7799 60. The Exim user and group can now be bound into the binary as text strings
7800 that are looked up at the start of Exim's processing.
7802 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7804 62. Added $mailstore_basename variable.
7806 63. Installed patch to sieve.c from Michael Haardt.
7808 64. When Exim failed to open the panic log after failing to open the main log,
7809 the original message it was trying to log was written to stderr and debug
7810 output, but if they were not available (the usual case in production), it
7811 was lost. Now it is written to syslog before the two lines that record the
7812 failures to open the logs.
7814 65. Users' Exim filters run in subprocesses under the user's uid. It is
7815 possible for a "deliver" command or an alias in a "personal" command to
7816 provoke an address rewrite. If logging of address rewriting is configured,
7817 this fails because the process is not running as root or exim. There may be
7818 a better way of dealing with this, but for the moment (because 4.30 needs
7819 to be released), I have disabled address rewrite logging when running a
7820 filter in a non-root, non-exim process.
7826 1. The buildconfig auxiliary program wasn't quoting the value set for
7827 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7828 not defined. This bug was masked in 4.22 by the effect that was fixed in
7831 2. Some messages that were rejected after a message id was allocated were
7832 shown as "incomplete" by exigrep. It no longer does this for messages that
7833 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7835 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7836 have allow_domain_literals set, the ID did not get logged in the <= line.
7837 Domain literals are now always recognized in Message-ID: header lines.
7839 4. The first argument for a ${extract expansion item is the key name or field
7840 number. Leading and trailing spaces in this item were not being ignored,
7841 causing some misleading effects.
7843 5. When deliver_drop_privilege was set, single queue runner processes started
7844 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7845 same command in the process it spins off) were not dropping privilege.
7847 6. When the daemon running as "exim" started a queue runner, it always
7848 re-executed Exim in the spun-off process. This is a waste of effort when
7849 deliver_drop_privilege is set. The new process now just calls the
7850 queue-runner function directly.
7856 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7859 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7860 ignored. Though the use of -oP was forcing the writing of a pid file, it
7861 was always written to the default place.
7863 3. If the message "no IP address found for host xxxx" is generated during
7864 incoming verification, it is now followed by identification of the incoming
7865 connection (so you can more easily find what provoked it).
7867 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7869 5. Added some features to "harden" Exim a bit more against certain attacks:
7871 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7872 be put in Local/Makefile. This is like the never_users runtime option,
7873 but it cannot be overridden. The default setting is "root".
7875 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7876 prefix string with which any file named in a -C command line option
7879 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7880 is retained for -C and -D only if the caller of Exim is root. Without
7881 it, the exim user may also use -C and -D and retain privilege.
7883 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7884 command line option is disabled.
7886 6. Macro names set by the -D option must start with an upper case letter, just
7887 like macro names defined in the configuration file.
7889 7. Added "dereference=" facility to LDAP.
7891 8. Two instances of the typo "uknown" in the source files are fixed.
7893 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7894 the Configure-Makefile script screwed up while processing it.
7896 10. Incorporated PCRE 4.4.
7898 11. The SMTP synchronization check was not operating right at the start of an
7899 SMTP session. For example, it could not catch a HELO sent before the client
7900 waited for the greeting. There is now a check for outstanding input at the
7901 point when the greeting is written. Because of the duplex, asynchronous
7902 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7903 way, but not yet received, when the check is performed.
7905 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7906 on TCP/IP sockets, because this apparently causes some broken clients to
7909 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7910 unchanged) from the Cygwin maintainer.
7912 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7913 was supported instead of testing for mbx. Effectively a typo.
7915 15. The spa authenticator server code was not checking that the input it
7916 received was valid base64.
7918 16. The debug output line for the "set" modifier in ACLs was not showing the
7919 name of the variable that was being set.
7921 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7923 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7924 lookup on its own. Something else was needed to trigger the lookup. For
7925 example, a match in host_lookup or the need for a host name in a host list.
7926 Now, if $sender_host_name is referenced and the host name has not yet been
7927 looked up, a lookup is performed. If the lookup fails, the variable remains
7928 empty, and $host_lookup_failed is set to "1".
7930 19. Added "eqi" as a case-independent comparison operator.
7932 20. The saslauthd authentication condition could segfault if neither service
7933 nor realm was specified.
7935 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7936 error message that was logged was misleading, and incoming SMTP
7937 connections were dropped. The message is now more accurate, and temporary
7938 errors are given to SMTP connections.
7940 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7941 (except RSET) with a 421 error, until QUIT is received. However, it was
7942 failing to send a response to QUIT.
7944 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7945 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7946 ACL is now run after the helo_try_verify_hosts code.
7948 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7949 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7950 case-independent, so other case variants are also recognized). Apparently
7951 some systems use these upper case variants.
7953 25. If more than two messages were waiting for the same host, and a transport
7954 filter was specified for the transport, Exim sent two messages over the
7955 same TCP/IP connection, and then failed with "socket operation on non-
7956 socket" when it tried to send the third.
7958 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7960 27. The extern definition of crypt16() in expand.c was not being excluded when
7961 the OS had its own crypt16() function.
7963 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7964 as a preferred synonym for return_size_limit, both as an option and as an
7967 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7969 30. Changed the default configuration ACL to relax the local part checking rule
7970 for addresses that are not in any local domains. For these addresses,
7971 slashes and pipe symbols are allowed within local parts, but the sequence
7972 /../ is explicitly forbidden.
7974 31. SPA server authentication was not clearing the challenge buffer before
7977 32. log_message in a "warn" ACL statement was writing to the reject log as
7978 well as to the main log, which contradicts the documentation and doesn't
7979 seem right (because no rejection is happening). So I have stopped it.
7981 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7982 However, I am unable to do any testing of this.
7984 34. Fixed an infelicity in the appendfile transport. When checking directories
7985 for a mailbox, to see if any needed to be created, it was accidentally
7986 using path names with one or more superfluous leading slashes; tracing
7987 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7989 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7990 contents are added to the log line that is written for every discarded
7991 recipient. (Previously a log_message setting was ignored.)
7993 36. The ${quote: operator now quotes the string if it is empty.
7995 37. The install script runs exim in order to find its version number. If for
7996 some reason other than non-existence or emptiness, which it checks, it
7997 could not run './exim', it was installing it with an empty version number,
7998 i.e. as "exim-". This error state is now caught, and the installation is
8001 38. An argument was missing from the function that creates an error message
8002 when Exim fails to connect to the socket for saslauthd authentication.
8003 This could cause Exim to crash, or give a corrupted message.
8005 39. Added isip, isip4, and isip6 to ${if conditions.
8007 40. The ACL variables $acl_xx are now saved with the message, and can be
8008 accessed later in routers, transports, and filters.
8010 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8011 strings in the file are not string-expanded.
8013 42. If a MAIL command specified a SIZE value that was too large to fit into an
8014 int variable, the check against message_size_limit failed. Such values are
8015 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8016 day this will have to be increased, but I don't think I want to be around
8017 when emails are that large.
8024 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8025 iconv() is not standard in FreeBSD.
8027 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8028 IPv6 enabled. The observed symptom was a segmentation fault on return from
8029 the function os_common_find_running_interfaces() in src/os.c.
8031 3. In the check_special_case() function in daemon.c I had used "errno" as an
8032 argument name, which causes warnings on some systems. This was basically a
8033 typo, since it was named "eno" in the comments!
8035 4. The code that waits for the clock to tick (at a resolution of some fraction
8036 of a second) so as to ensure message-id uniqueness was always waiting for
8037 at least one whole tick, when it could have waited for less. [This is
8038 almost certainly not relevant at current processor speeds, where it is
8039 unlikely to ever wait at all. But we try to future-proof.]
8041 5. The function that sleeps for a time interval that includes fractions of a
8042 second contained a race. It did not block SIGALRM between setting the
8043 timer, and suspending (a couple of lines later). If the interval was short
8044 and the sigsuspend() was delayed until after it had expired, the suspension
8045 never ended. On busy systems this could lead to processes getting stuck for
8048 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8049 process, before it forks any delivery processes. The open lookup caching
8050 mechanism meant that the open file or database connection was passed into
8051 the delivery process. The problem was that delivery processes always tidy
8052 up cached lookup data. This could cause a problem for the next delivery
8053 process started by the queue runner, because the external queue runner
8054 process does not know about the closure. So the next delivery process
8055 still has data in the lookup cache. In the case of a file lookup, there was
8056 no problem because closing a file descriptor in a subprocess doesn't affect
8057 the parent. However, if the lookup was caching a connection to a database,
8058 the connection was closed, and the second delivery process was likely to
8059 see errors such as "PGSQL: query failed: server closed the connection
8060 unexpectedly". The problem has been fixed by closing all cached lookups
8061 in a queue runner before running a delivery process.
8063 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8064 seem to have the "const" qualifier which it has on other OS. I've
8067 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8068 *of the same type* (client or server) with the same public name that an
8069 error should be diagnosed.
8071 9. When Exim looked up a host name for an IP address, but failed to find the
8072 original IP address when looking up the host name (a safety check), it
8073 output the message "<ip address> does not match any IP for NULL", which was
8074 confusing, to say the least. The bug was that the host name should have
8075 appeared instead of "NULL".
8077 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8078 user that is built into the binary, and the -C or -D options is used, root
8079 privilege is dropped before the configuration file is read. In addition,
8080 logging is switched to stderr instead of the normal log files. If the
8081 configuration then re-defines the Exim user, the unprivileged environment
8082 is probably not what is expected, so Exim logs a panic warning message (but
8085 However, if deliver_drop_privilege is set, the unprivileged state may well
8086 be exactly what is intended, so the warning has been cut out in that case,
8087 and Exim is allowed to try to write to its normal log files.
8093 1. smtp_return_error_details was not giving details for temporary sender
8094 or receiver verification errors.
8096 2. Diagnose a configuration error if two authenticators have the same public
8099 3. Exim used not to create the message log file for a message until the first
8100 delivery attempt. This could be confusing when incoming messages were held
8101 for policy or load reasons. The message log file is now created at the time
8102 the message is received, and an initial "Received" line is written to it.
8104 4. The automatically generated man page for command line options had a minor
8105 bug that caused no ill effects; however, a more serious problem was that
8106 the procedure for building the man page automatically didn't always
8107 operate. Consequently, release 4.20 contains an out-of-date version. This
8108 shouldn't happen again.
8110 5. When building Exim with embedded Perl support, the script that builds the
8111 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8112 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8114 6. The freeze_tell option was not being used for messages that were frozen on
8115 arrival, either by an ACL or by local_scan().
8117 7. Added the smtp_incomplete_transaction log selector.
8119 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8120 was accepting AUTH without a new EHLO.
8122 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8123 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8126 10. Logging of TCP/IP connections (when configured) now happens in the main
8127 daemon process instead of the child process, so that the TCP/IP connection
8128 count is more accurate (but it can never be perfect).
8130 11. The use of "drop" in a nested ACL was not being handled correctly in the
8131 outer ACL. Now, if condition failure induced by the nested "drop" causes
8132 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8133 or "require"), the connection is dropped.
8135 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8136 that yield "discard" can now be used with an "accept" or a "discard" verb,
8137 but an error is generated for any others (because I can't see a useful way
8138 to define what should happen).
8140 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8141 are now processed in the same way as lines in the Exim configuration file.
8142 In particular, continuation lines are supported.
8144 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8146 15. Added -ti meaning -t -i.
8148 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8149 domains, and warn by logging if others are found.
8151 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8152 returned from GIFCONF when Exim is trying to find the list of interfaces on
8153 a host. The code in os.c has been modified to copy each ifreq to an aligned
8154 structure in all cases.
8156 Also, in some cases, the returned ifreq's were being copied to a 'struct
8157 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8158 means the last couple of bytes of an IPv6 address could be chopped if the
8159 ifreq contained only a normal sockaddr (14 bytes storage).
8161 18. Named domain lists were not supported in the hosts_treat_as_local option.
8162 An entry such as +xxxx was not recognized, and was treated as a literal
8165 19. Ensure that header lines added by a DATA ACL are included in the reject log
8166 if the ACL subsequently rejects the message.
8168 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8169 MD5 (which is deprecated).
8171 21. When testing a filter file using -bf, Exim was writing a message when it
8172 took the sender from a "From " line in the message, but it was not doing so
8173 when it took $return_path from a Return-Path: header line. It now does.
8175 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8176 with a valid header line field name (a series of printing characters
8177 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8179 23. Changed "disc" in the source to "disk" to conform to the documentation and
8180 the book and for uniformity.
8182 24. Ignore Sendmail's -Ooption=value command line item.
8184 25. When execve() failed while trying to run a command in a pipe transport,
8185 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8186 could be confused with a return value of 69 from the command itself. This
8187 has been changed to 127, the value the shell returns if it is asked to run
8188 a non-existent command. The wording for the related log line suggests a
8189 non-existent command as the problem.
8191 26. If received_header_text expands to an empty string, do not add a Received:
8192 header line to the message. (Well, it adds a token one on the spool, but
8193 marks it "old" so that it doesn't get used or transmitted.)
8195 27. Installed eximstats 1.28 (addition of -nt option).
8197 28. There was no check for failure on the call to getsockname() in the daemon
8198 code. This can fail if there is a shortage of resources on the system, with
8199 ENOMEM, for example. A temporary error is now given on failure.
8201 29. Contrary to the C standard, it seems that in some environments, the
8202 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8203 program. Exim now does this explicitly; it affects the formatting of
8204 timestamps using strftime().
8206 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8207 complaints. I've now initialized all the variables, to avoid this.
8209 32. Header lines added by a system filter were not being "seen" during
8210 transport-time rewrites.
8212 33. The info_callback() function passed to OpenSSL is set up with type void
8213 (*)(SSL *, int, int), as described somewhere. However, when calling the
8214 function (actually a macro) that sets it up, the type void(*)() is
8215 expected. I've put in a cast to prevent warnings from picky compilers.
8217 34. If a DNS black list lookup found a CNAME record, but there were no A
8218 records associated with the domain it pointed at, Exim crashed.
8220 35. If a DNS black list lookup returned more than one A record, Exim ignored
8221 all but the first. It now scans all returned addresses if a particular IP
8222 value is being sought. In this situation, the contents of the
8223 $dnslist_value variable are a list of all the addresses, separated by a
8226 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8227 to accept a host name and all its aliases if the forward lookup for any of
8228 them yielded the IP address of the incoming connection. Now it accepts only
8229 those names whose forward lookup yields the correct IP address. Any other
8230 names are discarded. This closes a loophole whereby a rogue DNS
8231 administrator could create reverse DNS records to break through a
8232 wildcarded host restriction in an ACL.
8234 37. If a user filter or a system filter that ran in a subprocess used any of
8235 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8236 the wrong values were passed to the pipe command ($thisaddress had the
8237 value of $0, $0 had the value of $1, etc). This bug was introduced by
8238 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8240 38. Improved the line breaking for long SMTP error messages from ACLs.
8241 Previously, if there was no break point between 40 and 75 characters, Exim
8242 left the rest of the message alone. Two changes have been made: (a) I've
8243 reduced the minimum length to 35 characters; (b) if it can't find a break
8244 point between 35 and 75 characters, it looks ahead and uses the first one
8245 that it finds. This may give the occasional overlong line, but at least the
8246 remaining text gets split now.
8248 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8249 file descriptors might be low, and that setting 1000 would always raise it.
8250 It turns out that in some environments, the limit is already over 1000 and
8251 that lowering it causes trouble. So now Exim takes care not to decrease it.
8253 40. When delivering a message, the value of $return_path is set to $sender_
8254 address at the start of routing (routers may change the value). By an
8255 oversight, this default was not being set up when an address was tested by
8256 -bt or -bv, which affected the outcome if any router or filter referred to
8259 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8260 log only when "message" or "log_message" are set. However, if one of the
8261 conditions was an address verification, or a call to a nested ACL, the
8262 messages generated by the underlying test were being passed through. This
8263 no longer happens. The underlying message is available in $acl_verify_
8264 message for both "message" and "log_message" expansions, so it can be
8265 passed through if needed.
8267 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8268 new expansion $bh_ to give the encoded byte string without charset
8269 translation. Translation happens only if iconv() is available; HAVE_ICONV
8270 indicates this at build time. HEADERS_CHARSET gives the charset to
8271 translate to; headers_charset can change it in the configuration, and
8272 "headers charset" can change it in an individual filter file.
8274 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8275 that creates RFC 2047 encoded "words" labels them as that charset instead
8276 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8277 expansion operator; (ii) when Exim creates a From: line for a local
8278 message; (iii) when a header line is rewritten to include a "phrase" part.
8280 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8281 buggy, causing it to skip the first lines of messages whose message ID
8282 ended in 'D'. This would not have bitten before Exim release 4.14, because
8283 message IDs were unlikely to end in 'D' before then. The effect was to have
8284 incorrect size information for certain domains.
8286 45. #include "config.h" was missing at the start of the crypt16.c module. This
8287 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8290 46. If there was a timeout during a "random" callout check, Exim treated it as
8291 a failure of the random address, and carried on sending RSET and the real
8292 address. If the delay was just some slowness somewhere, the response to the
8293 original RCPT would be taken as a response to RSET and so on, causing
8294 mayhem of various kinds.
8296 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8297 when I implemented it. It didn't allow for the fact that some option values
8298 may legitimately be negative (e.g. size_addition), and it didn't even do
8299 the right test for positive values.
8301 48. Domain names in DNS records are case-independent. Exim always looks them up
8302 in lower case. Some resolvers return domain names in exactly the case they
8303 appear in the zone file, that is, they may contain uppercase letters. Not
8304 all resolvers do this - some return always lower case. Exim was treating a
8305 change of case by a resolver as a change of domain, similar to a widening
8306 of a domain abbreviation. This triggered its re-routing code and so it was
8307 trying to route what was effectively the same domain again. This normally
8308 caused routing to fail (because the router wouldn't handle the domain
8309 twice). Now Exim checks for this case specially, and just changes the
8310 casing of the domain that it ultimately uses when it transmits the message
8313 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8316 50. If a filter generated a file delivery with a non-absolute name (possible if
8317 no home directory exists for the router), the forbid_file option was not
8320 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8321 the existing equality matching.
8323 52. Exim was using ints instead of ino_t variables in some places where it was
8324 dealing with inode numbers.
8326 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8327 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8328 called TMPDIR, and if it finds it is different, it changes its value.
8330 54. The smtp_printf() function is now made available to local_scan() so
8331 additional output lines can be written before returning. There is also an
8332 smtp_fflush() function to enable the detection of a dropped connection.
8333 The variables smtp_input and smtp_batched_input are exported to
8336 55. Changed the default runtime configuration: the message "Unknown user"
8337 has been removed from the ACL, and instead placed on the localuser router,
8338 using the cannot_route_message feature. This means that any verification
8339 failures that generate their own messages won't get overridden. Similarly,
8340 the "Unrouteable address" message that was in the ACL for unverifiable
8341 relay addresses has also been removed.
8343 56. Added hosts_avoid_esmtp to the smtp transport.
8345 57. The exicyclog script was not checking for the esoteric option
8346 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8347 will work only if exicyclog is run under the appropriate euid.
8349 58. Following a discussion on the list, the rules by which Exim recognises line
8350 endings on incoming messages have been changed. The -dropcr and drop_cr
8351 options are now no-ops, retained only for backwards compatibility. The
8352 following line terminators are recognized: LF CRLF CR. However, special
8353 processing applies to CR:
8355 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8356 nor a local message in the state where . is a terminator.
8358 (ii) If a bare CR is encountered in a header line, an extra space is added
8359 after the line terminator so as not to end the header. The reasoning
8360 behind this is that bare CRs in header lines are most likely either
8361 to be mistakes, or people trying to play silly games.
8363 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8364 was being incorrectly given as 18 bytes larger than it should have been.
8365 This is a VOB (very old bug).
8367 60. This may never have affected anything current, but just in case it has:
8368 When the local host is found other than at the start of a list of hosts,
8369 the local host, those with the same MX, and any that follow, are discarded.
8370 When the list in question was part of a longer list of hosts, the following
8371 hosts (not currently being processed) were also being discarded. This no
8372 longer happens. I'm not sure if this situation could ever has previously
8375 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8378 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8379 header. This is something that is recommended in a new Internet Draft, and
8380 is something that is documented as being done by Sendmail. There are two
8381 possible values. For messages generated by the autoreply transport, Exim
8384 Auto-Submitted: auto-replied
8386 whereas for all other generated messages (e.g. bounces) it adds
8388 Auto-Submitted: auto-generated
8390 63. The "personal" condition in filters now includes a test for the
8391 Auto-Submitted: header. If it contains the string "auto-" the message it
8392 not considered personal.
8394 64. Added rcpt_include_affixes as a generic transport option.
8396 65. Added queue_only_override (default true).
8398 66. Added the syslog_duplication option.
8400 67. If what should have been the first header line of a message consisted of
8401 a space followed by a colon, Exim was mis-interpreting it as a header line.
8402 It isn't of course - it is syntactically invalid and should therefore be
8403 treated as the start of the message body. The misbehaviour could have
8404 caused a number of strange effects, including loss of data in subsequent
8405 header lines, and spool format errors.
8407 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8408 client host had authenticated. This control can now be exercised by an ACL
8409 for more flexibility.
8411 69. By default, callouts do not happen when testing with -bh. There is now a
8412 variant, -bhc, which does actually run the callout code, including
8413 consulting and updating the callout cache.
8415 70. Added support for saslauthd authentication, courtesy of Alexander
8418 71. If statvfs() failed on the spool or log directories while checking their
8419 size for availability, Exim confusingly gave the error "space shortage".
8420 Furthermore, in debugging mode it crashed with a floating point exception.
8421 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8422 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8423 serious problem, Exim now writes to the main and panic logs when this
8424 happens, with details of the failure. It then refuses to accept the
8425 incoming message, giving the message "spool directory problem" or "log
8426 directory problem" with a 421 code for SMTP messages.
8428 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8429 0, 1, and 2 exist, because some OS complain for execs without them (see
8430 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8431 descriptors. However, the code omitted to check that the open succeeded,
8432 causing mysterious errors if for some reason the permissions on /dev/null
8433 got screwed. Now Exim writes a message to the main and panic logs, and
8434 bombs out if it can't open /dev/null.
8436 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8437 interact so that it is all more flexible. It is supposed to remain
8438 backwards compatible. Also added extra_local_interfaces.
8440 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8441 to bomb out with an assertion failure - to the client this appears as a
8442 connection drop. This problem occurs in the part of the code that was taken
8443 from the Samba project. Fortunately, the assertion is in a very simple
8444 function, so I have fixed this by reproducing the function inline in the
8445 one place where it is called, and arranging for authentication to fail
8446 instead of killing the process with assert().
8448 75. The SPA client code was not working when the server requested OEM rather
8449 than Unicode encoding.
8451 76. Added code to make require_files with a specific uid setting more usable in
8452 the case where statting the file as root fails - usually a non-root-mounted
8453 NFS file system. When this happens and the failure is EACCES, Exim now
8454 forks a subprocess and does the per-uid checking as the relevant uid.
8456 77. Added process_log_path.
8458 78. If log_file_path was not explicitly set, a setting of check_log_space or
8459 check_log_inodes was ignored.
8461 79. If a space check for the spool or log partitions fails, the incident is now
8462 logged. Of course, in the latter case the data may get lost...
8464 80. Added the %p formatting code to string_format() so that it can be used to
8465 print addresses in debug_print(). Adjusted all the address printing in the
8466 debugging in store.c to use %p rather than %d.
8468 81. There was a concern that a line of code in smtp_in.c could overflow a
8469 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8470 initially expressed, the concern was not well-founded, because trailing
8471 spaces are removed early. However, if the trailing spaces were followed by
8472 a NULL, they did not get removed, so the overflow was possible. Two fixes
8475 (a) I re-wrote the offending code in a cleaner fashion.
8476 (b) If an incoming SMTP command contains a NULL character, it is rejected
8479 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8480 runs initgroups(), so that if the Exim user is in any additional groups,
8481 they will be used during message reception.
8487 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob