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.
9 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
16 created buffers. Similar fix for radius expansion condition.
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
52 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
53 Previously a macro used one argument twice; when called with the
54 argument as an expression having side-effects, incorrect operation
55 resulted. Use an inlineable function.
57 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
58 held open for a verify callout. Previously this wan not accounted for
59 and a corrupt onward SMTP conversation resulted.
61 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
62 Previously they were embedded naively in the constructed address; when
63 needed, strip the quoting and quote the entire local_part.
64 Also make the inbound_srs expansion condition handle quoting.
66 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
67 excluded, not matching the documentation.
69 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
70 was given for the sqlite_dbfile a trap resulted.
72 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
73 "name" argument resulted in a trap. There is no reason to disallow such;
74 this was a coding error.
76 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
77 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
78 spent suspended, ignoring the Posix definition. Previously we assumed
79 it did and a constant offset from real time could be used as a correction.
80 Change to using the same clock source for the start-of-message and the
81 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
82 exists, just to get a clock slightly more aligned to reality.
84 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
85 RFC says it is optional some validators care. The missing char was not
86 intended but triggered by a line-wrap alignement. Discovery and fix by
87 Guillaume Outters, hacked on by JH.
89 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
90 name being quoted was tainted a trap would be taken. Fix by using
91 dynamicaly created buffers. The routine could have been called by a
92 rewrite with the "h" flag, by using the "-F" command-line option, or
93 by using a "name=" option on a control=submission ACL modifier.
95 JH/19 SPF: change the Authentication-Results expansion component to give
96 smtp.helo when the sender domain is empty. Previously it gave
99 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
100 not in 127.0.0.0/8 to help in spotting list domains taken over by a
101 domain-parking registrar.
103 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
104 Previously when a whitespace character was specified it was not inserted
105 after removing the newline.
107 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
108 the domain part of the recipient address. This overrides any tls_sni
109 option set, which was previously used.
111 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
114 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
115 is_tainted() had an off-by-one error in the overenthusiastic direction.
116 Find and fix by Gavan. Although NetBSD is not a supported platform for
117 4.94 this bug could affect other platforms.
119 PP/01 Fix default prime selection to be consistent.
120 One path used ike23 still, instead of exim.dev.20160529.3; now both
121 execution flows will use the same DH primes (currently
122 exim.dev.20160529.3).
124 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
125 option in smtp transport, to match the documentation. Previously
126 verification was not being done in some cases where it should have been.
128 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
129 than one server was defined and depending on the platform memory layout
130 details, an internal consistency trap could be hit while walking the list
133 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
134 files. The value is used by the authresults expansion item. Previously
135 if this was used in a router or transport, a crash could result.
137 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
138 have a sufficient incidence of EINTR returns from select that an
139 interaction with other operations done by the main daemon loop exposed
140 a bug in the error-handling. This was benign apart from the log
143 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
144 for consistency with delivered "=>" and failed "**" lines. While we're
145 there, handle PRX and TFO.
147 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
148 applied. This resulted, if any header-line rewrite rules were configured,
149 in a panic-log trigerrable by sending a message with a long address in
150 a header. Fix by increaing the arbitrary limit to larger than a single
151 (dewrapped) 5322 header line maximum size.
153 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
154 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
155 RFC 6648 which deprecates X- options in protocols as a general practice.
156 Changeover between the implementations is handled by the mechanisms
159 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
160 a remote which uses callout/hold. Previously the local was queued.
162 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
165 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
166 log line. The asterisk is now consistently appended to the remote IP
167 (and port, if given), and will also be provided on defer and fail log
168 lines. Previously it could be placed on the local IP if that was being
169 logged, and was only provided on delivery lines.
171 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
173 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
174 authenticator client_send option. Previously the next char, after a pair
175 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
176 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
177 documentation. There is still no way to get a leading ^ immediately
178 after a NUL (ie. for the password of a PLAIN method authenticator.
180 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
181 files. For bad sizes read, delete the record and whine to paniclog.
183 JH/38 When logging an AUTH failure, as server, do not include sensitive
184 information. Previously, the credentials would be included if given
185 as part of the AUTH command line and an ACL denied authentidcation.
187 JH/39 Bug 2691: fix $local_part_data. When the matching list element
188 referred to a file, bad data was returned. This likely also affected
191 JH/40 The gsasl authenticator now supports caching of the salted password
192 generated by the client-side implementation. This required the addition
193 of a new variable: $auth4.
195 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
196 left undeleted; the attempt to re-create it then failed - resulting in
197 the usual "SIGHUP tp have daemon reload configuration" to not work.
198 This affected any platform not supporting "abstract" Unix-domain
199 sockets (i.e. not Linux).
201 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
202 recipients" response to RCPT in a later response, with a 250. The
203 previous coding assumed this would not happen, and under PIPELINING
204 would result in both lost and duplicate recipients for a message.
206 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
207 Previously the weighting was incorrectly applied. Similar fix for socks
208 proxies. Found and fixed by Heiko Schlichting.
210 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
211 not handle sub-lists included using the +namedlist syntax. While
212 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
213 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
216 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
217 Previously the permanent pool was used, so the sore could not be freed.
218 This meant a connection with many messages would use continually-growing
221 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
222 per-pool so as not to waste too much space. Previously a constant size
223 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
224 DOS attacks harder. The cost is wasted memory use in the larger blocks.
226 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
227 same space repeatedly, and a smaller process footprint.
229 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
230 for GnuTLS. Previously the name started "%s" which, while not a bug,
231 looked as if if might be one.
233 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
234 "continued connection") make the $proxy_* variables available. Previously
235 the information was not passed across the exec() call for subsequent
236 transport executions. This also mean that the log lines for the
237 messages can show the proxy information.
239 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
240 receive time. With modern systems the difference is significant.
241 The historical behaviour can be restored by disabling (a new) log_selector
242 "queue_time_exclusive".
244 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
245 content) were specifically tested for. Now, also cover epxansions
246 rerulting in acl names and inline ACL content.
248 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
249 making it unusable in complex expressions.
251 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
252 fail with a taint-check on the spoolfile name, and leave the message
255 HS/01 Enforce absolute PID file path name.
257 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
259 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
265 JH/01 Avoid costly startup code when not strictly needed. This reduces time
266 for some exim process initialisations. It does mean that the logging
267 of TLS configuration problems is only done for the daemon startup.
269 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
271 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
272 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
275 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
276 DKIM signing is being done. Previously a transport_filter would always
277 disable CHUNKING, falling back to traditional DATA.
279 JH/05 Regard command-line recipients as tainted.
281 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
283 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
284 PAM library frees one of the arguments given to it, despite the
285 documentation. Therefore a plain malloc must be used.
287 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
288 on-stack buffers were used, resulting in a taint trap when DSN information
289 copied from a received message was written into the buffer.
291 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
292 the ordering of its ARC headers. This caused a crash.
294 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
295 a new record was being constructed with information from the peer, a trap
298 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
299 installation would get error messages from DMARC verify, when it hit the
300 nonexistent file indicated by the default. Distros wanting DMARC enabled
301 should both provide the file and set the option.
302 Also enforce no DMARC verification for command-line sourced messages.
304 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
305 could, depending on the platform, hang at the STARTTLS response.
307 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
308 message on a connection. Previously if one message had ARC headers and
309 the following one did not, a crash could result when adding an
310 Authentication-Results: header.
312 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
313 Exim main code and Exim-related utities. The introduction of taint
314 tracking also did many adjustments to string handling. Since then, eximon
315 frequently terminated with an assert failure.
317 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
318 check for 452 responses. This slightly helps the inefficieny of doing
319 a large alias-expansion into a recipient-limited target. The max_rcpt
320 transport option still applies (and at the current default, will override
321 the new feature). The check is done for either cause of synch, and forces
322 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
323 connection. The new facility is not tunable at this time.
325 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
326 library live data was being used, so the results became garbage. Make
327 copies while it is still usable.
329 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
330 on delivery deferred (==) and failed (**) lines (if a delivery was
331 attemtped). Previously it was only on completion (=>) lines.
333 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
334 for the expansion of the server_scram_iter and server_scram_salt options.
336 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
337 are now specifically given a NO_DATA response without hitting the system
338 resolver. The library goes on to do the now-standard TXT lookup.
339 Use of dnsdb lookups is not affected.
341 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
342 only retrieve the errormessage once. Previously two calls to dlerror()
343 were used, and the second one (for mainlog/paniclog) retrieved null
346 JH/20 Taint checking: disallow use of tainted data for
347 - the appendfile transport file and directory options
348 - the pipe transport command
349 - the autoreply transport file, log and once options
350 - file names used by the redirect router (including filter files)
352 - paths used by single-key lookups
353 Previously this was permitted.
355 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
356 adjusted the size of a major service buffer; this failed because the
357 buffer was in use at the time. Change to a compile-time increase in the
358 buffer size, when this authenticator is compiled into exim.
360 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
361 previous fast-mode was untenable in the face of glibs using mmap to
362 support larger malloc requests.
364 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
365 New values supported, if defined on system where compiled:
366 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
367 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
369 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
370 running a limited number of proceses in parallel, a benefit is gained. The
371 amount varies with the platform hardware and load. The use of the option
372 queue_run_in_order means we cannot do this, as ordering becomes
375 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
376 had introduced a string-copy (for ensuring NUL-termination) which was not
377 appropriate for that case, which can include embedded NUL bytes in the
378 block of data. Investigation showed the copy to actually be needless, the
379 data being length-specified.
381 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
382 done during a receiving connection, and both used TLS, global info was
383 used rather than per-connection info for tracking the state of data
384 queued for transmission. This could result in a connection hang.
386 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
387 Previously, when delivering serveral messages down a single connection
388 only the first would provide a SIZE. This was due to the size information
389 not being properly tracked.
391 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
392 TAI (at 37 seconds currently), pretend to be in UTC for time-related
393 expansion and logging. Previously, spurious values such as a future
394 minute could be seen.
396 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
397 it could crash from a null-deref. This could also affect the
398 ${addresses: } operator and ${readsock } item.
400 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
401 message following a mime one, the variable was not reset.
403 JH/30 When an pipelined-connect fails at the first response, assume incorrect
404 cached capability (perhaps the peer reneged?) and immediately retry in
407 JH/31 Fix spurious detection of timeout while writing to transport filter.
409 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
410 an attempt to copy the string was made before checking it.
412 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
413 taint of the lookup key was maintained; we now regard the presence in the
414 filesystem as sufficient validation.
416 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
417 argument is supplied.
419 JH/35 The dsearch lookup now requires that the directory is an absolute path.
420 Previously this was not checked, and nonempty relative paths made an
421 access under Exim's current working directory.
423 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
424 Previously no event was raised.
426 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
427 parameter supplied by the sender MAIL FROM command. Previously it was
428 ignored, and only the check_spool_space option value for the required
431 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
432 the size of the signing public-key. Previously it was instead giving
433 the size of the signature hash.
435 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
436 the default. See the (new) dkim_verify_min_keysizes option.
438 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
439 and an ACL use a lookup for checking either the local_part or domain,
440 stale data could be accessed. Ensure that variable references are
441 dropped between messages.
443 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
444 by the client was not checked as pointing within response data before
445 being used. A malicious client could thus cause an out-of-bounds read and
446 possibly gain authentication. Fix by adding the check.
448 JH/42 Internationalisation: change the default for downconversion in the smtp
449 transport to be "if needed". Previously it was "as previously set" for
450 the message, which usually meant "if needed" for message-submission but
451 "no" for everything else. However, MTAs have been seen using SMTPUTF8
452 even when the envelope addresses did not need it, resulting in forwarding
453 failures to non-supporting MTAs. A downconvert in such cases will be
454 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
455 transport. The change does mean that addresses needing conversion will
456 be converted when previously a delivery failure would occur.
458 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
459 response was received it would be used unchecked in a fail-DSN, violating
460 standards on line-length limits. Truncate if needed.
462 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
463 doesn't work. (Additionally add a new main config option to configure the
470 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
471 side, to decode a TLS 1.3 packet capture.
473 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
474 Previously the default library behaviour applied, sending two, each in
477 JH/03 Debug output for ACL now gives the config file name and line number for
480 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
482 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
484 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
485 buffer overrun for (non-chunking) other transports.
487 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
488 TLS1.3, means that a server rejecting a client certificate is not visible
489 to the client until the first read of encrypted data (typically the
490 response to EHLO). Add detection for that case and treat it as a failed
491 TLS connection attempt, so that the normal retry-in-clear can work (if
492 suitably configured).
494 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
495 and/or domain. Found and fixed by Jason Betts.
497 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
498 configuration). If a CNAME target was not a wellformed name pattern, a
501 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
502 the OS reports them interleaved with other addresses.
504 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
505 used both for input and for a verify callout, both encrypted, SMTP
506 responses being sent by the server could be lost. This resulted in
507 dropped connections and sometimes bounces generated by a peer sending
510 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
511 string. Previously it was possible to cause undefined behaviour in a
512 library routine (usually a crash). Found by "zerons".
514 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
517 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
518 API was removed, so update to use the newer ones.
520 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
521 any timeout set, is taking a long time. Previously we would hang on to a
522 rotated logfile "forever" if the input was arriving with long gaps
523 (a previous attempt to fix addressed lack, for a long time, of initial
526 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
527 shared (NFS) environment. The length of the tempfile name is now
528 4 + 16 ("hdr.$message_exim_id") which might break on file
529 systems which restrict the file name length to lower values.
532 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
533 shared (NFS) environment.
535 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
536 did for all versions <4.90). Notably -M, -m, --invert, -I may be
539 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
540 on some platforms for bit 31.
542 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
543 to changes apparently associated with TLS1.3 handling some of the APIs
544 previously used were either nonfunctional or inappropriate. Strings
545 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
546 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
547 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
548 This affects log line X= elements, the $tls_{in,out}_cipher variables,
549 and the use of specific cipher names in the encrypted= ACL condition.
551 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
553 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
554 verification result was not updated unless hosts_require_ocsp applied.
556 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
557 queue_list_requires_admin set to false, non-admin users were denied the
560 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
561 directory-of-certs mode. Previously they were advertised despite the
564 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
565 A single TCP connection by a client will now hold a TLS connection open
566 for multiple message deliveries, by default. Previously the default was to
569 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
570 default. If built with the facility, DANE will be used. The facility
571 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
573 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
574 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
575 must be defined and you must still, unless you define DISABLE_TLS, manage
576 the the include-dir and library-file requirements that go with that
577 choice. Non-TLS builds are still supported.
579 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
582 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
583 default. If the platform supports and has the facility enabled, it will
584 be requested on all coneections.
586 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
587 controlled by the build-time option SUPPORT_PIPE_CONNECT.
589 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
591 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
592 success-DSN messages. Previously the From: header was always the default
593 one for these; the option was ignored.
595 JH/28 Fix the timeout on smtp response to apply to the whole response.
596 Previously it was reset for every read, so a teergrubing peer sending
597 single bytes within the time limit could extend the connection for a
598 long time. Credit to Qualsys Security Advisory Team for the discovery.
600 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
601 delivery address, which leaked information of the results of local
602 forwarding. Change to the original envelope recipient address, per
605 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
606 requested. Previously not bounce was generated and a log entry of
607 error ignored was made.
609 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
611 JH/32 Introduce a general tainting mechanism for values read from the input
612 channel, and values derived from them. Refuse to expand any tainted
613 values, to catch one form of exploit.
615 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
616 was unused and the unexpanded text used for the test. Found and
617 fixed by Ruben Jenster.
619 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
620 an attempt to use a TLS library read routine dereffed a nul pointer,
623 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
624 them in our smtp response.
626 JH/36 Have the general router option retry_use_local_part default to true when
627 any of the restrictive preconditions are set (to anything). Previously it
628 was only for check_local user. The change removes one item of manual
629 configuration which is required for proper retries when a remote router
630 handles a subset of addresses for a domain.
632 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
633 link count into consideration.
635 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
636 caused the extension of big_buffer, the following lines were ignored.
638 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
639 accordance with RFC 2308. Previously there was no expiry, so a longlived
640 receive process (eg. due to ACL delays) versus a short SOA value could
643 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
645 JH/39 Promote DMARC support to mainline.
647 JH/40 Bug 2452: Add a References: header to DSNs.
649 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
650 parameters. The relevant library call is documented as "Deprecated: This
651 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
652 3.6.0, DH parameters are negotiated following RFC7919."
654 HS/06 Change the default of dnssec_request_domains to "*"
656 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
657 carried on and emitted a BDAT command, even when PIPELINING was not
660 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
661 buffer was used for the filename, resulting in a trap when tainted
662 arguments (eg. $domain) were used.
664 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
665 recommended to avoid a possible server-load attack. The feature can be
666 re-enabled via the openssl_options main cofiguration option.
668 JH/45 local_scan API: documented the current smtp_printf() call. This changed
669 for version 4.90 - adding a "more data" boolean to the arguments.
670 Bumped the ABI version number also, this having been missed previously;
671 release versions 4.90 to 4.92.3 inclusive were effectively broken in
672 respect of usage of smtp_printf() by either local_scan code or libraries
673 accessed via the ${dlfunc } expansion item. Both will need coding
674 adjustment for any calls to smtp_printf() to match the new function
675 signature; a FALSE value for the new argument is always safe.
677 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
678 the file-offset (which the Linux syscall does, and exim expects); this
679 resulted in an indefinite loop.
681 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
682 to do ARC verification. The Authentication-Results: header line added
683 by the configuration then had no ARC item.
689 JH/01 Remove code calling the customisable local_scan function, unless a new
690 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
692 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
693 non-signal-safe functions being used.
695 JH/03 Bug 2269: When presented with a received message having a stupidly large
696 number of DKIM-Signature headers, disable DKIM verification to avoid
697 a resource-consumption attack. The limit is set at twenty.
699 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
700 report of oldest_pass in ${authres } in consequence, and separate out
701 some descriptions of reasons for verification fail.
703 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
704 files in the spool were present and unlocked. A queue-runner could spot
705 them, resulting in a duplicate delivery. Fix that by doing the unlock
706 after the unlink. Investigation by Tim Stewart. Take the opportunity to
707 add more error-checking on spoolfile handling while that code is being
710 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
711 No known attacks, no CVE, this is defensive hardening.
713 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
714 a queue-runner could start a delivery while other operations were ongoing.
715 Cutthrough delivery was a common victim, resulting in duplicate delivery.
716 Found and investigated by Tim Stewart. Fix by using the open message data
717 file handle rather than opening another, and not locally closing it (which
718 releases a lock) for that case, while creating the temporary .eml format
719 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
721 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
722 $sender_verify_failure/$recipient_verify_failure to "random".
724 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
727 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
728 Previously this would segfault.
730 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
733 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
734 like zero, since the resolver should be doing this for us, But we need one
735 as a CNAME but no MX presence gets the CNAME returned; we need to check
736 that doesn't point to an MX to declare it "no MX returned" rather than
737 "error, loop". A new main option is added so the older capability of
738 following some limited number of chain links is maintained.
740 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
742 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
743 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
744 now-deprecated earlier definitions used only specified the range up to TLS
745 1.2 (in the older-version library docs).
747 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
749 JH/15 Rework TLS client-side context management. Stop using a global, and
750 explicitly pass a context around. This enables future use of TLS for
751 connections to service-daemons (eg. malware scanning) while a client smtp
752 connection is using TLS; with cutthrough connections this is quite likely.
754 JH/16 Fix ARC verification to do AS checks in reverse order.
756 JH/17 Support a "tls" option on the ${readsocket } expansion item.
758 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
759 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
760 Previously the "utf8" would be re-prepended for every additional message.
762 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
763 Previously thery were accepted, resulting in issues when attempting to
764 forward messages to a non-supporting MTA.
766 PP/02 Let -n work with printing macros too, not just options.
768 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
769 one parent address was copied, and bogus data was used at delivery-logging
770 time. Either a crash (after delivery) or bogus log data could result.
771 Discovery and analysis by Tim Stewart.
773 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
774 Previously if the string ended mid-character, we did not insert the
775 promised '?' replacement.
777 PP/04 Documentation: current string operators work on bytes, not codepoints.
779 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
780 should pack well giving a smaller memory footprint so better caching and
781 therefore performance. Group the declarations where this can't be done so
782 that the byte-sized flag variables are not interspersed among pointer
783 variables, giving a better chance of good packing by the compiler.
785 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
786 non-null, to avoid issues with sites running BATV. Previously reports were
787 sent with an empty envelope sender so looked like bounces.
789 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
790 The ignore_error flag wasn't being returned from the filter subprocess so
791 was not set for later routers. Investigation and fix by Matthias Kurz.
793 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
794 and a msg:complete for the whole, when a message is manually removed using
795 -Mrm. Developement by Matthias Kurz, hacked on by JH.
797 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
798 a "Gnu special" function, asprintf() in the DB utility binary builds; I
799 hope that is portable enough.
801 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
802 requiring a known-CA anchor certificate; make it now rely entirely on the
803 TLSA as an anchor. Checking the name on the leaf cert against the name
804 on the A-record for the host is still done for TA (but not for EE mode).
806 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
807 included in delivery lines for non-proxied connections, when compiled with
808 SUPPORT_SOCKS and running with proxy logging enabled.
810 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
811 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
812 move the existing event to fire before the normal logging of message
813 failure so that custom logging is bracketed by normal logging.
815 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
816 msg:fail:internal event. Developement by Matthias Kurz.
818 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
819 far too small for todays use of crypto signatures stored there. Go all
820 the way to the max DNS message size of 64kB, even though this might be
821 overmuch for IOT constrained device use.
823 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
824 copy a string over itself. The library routine is documented as not
825 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
827 JH/32 For main options check_spool_space and check_inode_space, where the
828 platform supports 64b integers, support more than the previous 2^31 kB
829 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
830 the previous G, M, k.
832 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
833 $authenticated_fail_id variable on authentication failure. Previously
836 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
837 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
838 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
839 GNUTLS_SEC_PARAM_MEDIUM.
841 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
842 side. Previously we would continue as if no SNI had been received.
844 JH/36 Harden the handling of string-lists. When a list consisted of a sole
845 "<" character, which should be a list-separator specification, we walked
846 off past the nul-terimation.
848 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
849 causes) even when the retry time is not yet met. Previously they were
850 not, meaning that when (say) an account was over-quota and temp-rejecting,
851 and multiple senders' messages were queued, only one sender would get
852 notified on each configured delay_warning cycle.
854 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
856 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
857 cipher-suites, an error can be left on the stack even for a succeeding
858 accept; this results in impossible error messages when a later operation
861 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
862 return error codes indicating retry. Under TLS1.3 this becomes required.
864 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
865 it only wrote the new authenticators, resulting in a lack of tracking of
866 peer changes of ESMTP extensions until the next cache flush.
868 JH/41 Fix the loop reading a message header line to check for integer overflow,
869 and more-often against header_maxsize. Previously a crafted message could
870 induce a crash of the recive process; now the message is cleanly rejected.
872 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
873 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
879 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
880 When redis_servers is set to a list of > 1 element, and the Redis servers
881 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
882 case of MOVED into a DEFER case instead, thus moving the query onto the
883 next server in the list. For a cluster of N elements, all N servers must
884 be defined in redis_servers.
886 GF/02 Catch and remove uninitialized value warning in exiqsumm
887 Check for existence of @ARGV before looking at $ARGV[0]
889 JH/01 Replace the store_release() internal interface with store_newblock(),
890 which internalises the check required to safely use the old one, plus
891 the allocate and data copy operations duplicated in both (!) of the
892 extant use locations.
894 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
895 modifier. This matches the restriction on the commandline.
897 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
898 Previously only the last row was returned.
900 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
901 we assumed that tags in the header were well-formed, and parsed the
902 element content after inspecting only the first char of the tag.
903 Assumptions at that stage could crash the receive process on malformed
906 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
907 While running the DKIM ACL we operate on the Permanent memory pool so that
908 variables created with "set" persist to the DATA ACL. Also (at any time)
909 DNS lookups that fail create cache records using the Permanent pool. But
910 expansions release any allocations made on the current pool - so a dnsdb
911 lookup expansion done in the DKIM ACL releases the memory used for the
912 DNS negative-cache, and bad things result. Solution is to switch to the
913 Main pool for expansions.
914 While we're in that code, add checks on the DNS cache during store_reset,
915 active in the testsuite.
916 Problem spotted, and debugging aided, by Wolfgang Breyha.
918 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
919 When none of the hosts presented to a transport match an already-open
920 connection, close it and proceed with the list. Previously we would
921 queue the message. Spotted by Lena with Yahoo, probably involving
924 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
925 Previously a spurious "250 OK id=" response was appended to the proper
928 JH/08 The "support for" informational output now, which built with Content
929 Scanning support, has a line for the malware scanner interfaces compiled
930 in. Interface can be individually included or not at build time.
932 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
933 by the template makefile "src/EDITME". The "STREAM" support for an older
934 ClamAV interface method is removed.
936 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
937 rows affected is given instead).
939 JH/11 The runtime Berkeley DB library version is now additionally output by
940 "exim -d -bV". Previously only the compile-time version was shown.
942 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
943 SMTP connection. Previously, when one had more recipients than the
944 first, an abortive onward connection was made. Move to full support for
945 multiple onward connections in sequence, handling cutthrough connection
946 for all multi-message initiating connections.
948 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
949 routers. Previously, a multi-recipient message would fail to match the
950 onward-connection opened for the first recipient, and cause its closure.
952 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
953 a timeout on read on a GnuTLS initiating connection, resulting in the
954 initiating connection being dropped. This mattered most when the callout
955 was marked defer_ok. Fix to keep the two timeout-detection methods
958 JH/15 Relax results from ACL control request to enable cutthrough, in
959 unsupported situations, from error to silently (except under debug)
960 ignoring. This covers use with PRDR, frozen messages, queue-only and
963 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
965 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
966 metadata, resulting in a crash in free().
968 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
969 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
970 Broken also in d185889f4, with init system revamp.
972 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
973 Previously we abruptly closed the connection after reading a malware-
974 found indication; now we go on to read the "scan ok" response line,
977 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
978 ACL. Previously, a crash would result.
980 JH/19 Speed up macro lookups during configuration file read, by skipping non-
981 macro text after a replacement (previously it was only once per line) and
982 by skipping builtin macros when searching for an uppercase lead character.
984 JH/20 DANE support moved from Experimental to mainline. The Makefile control
985 for the build is renamed.
987 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
988 was allocated for every new TLS startup, meaning one per message. Fix
989 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
991 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
992 reported the original. Fix to report (as far as possible) the ACL
993 result replacing the original.
995 JH/23 Fix memory leak during multi-message connections using STARTTLS under
996 OpenSSL. Certificate information is loaded for every new TLS startup,
997 and the resources needed to be freed.
999 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1001 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1004 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1005 DATA response info to the (existing) per-recipient response info for
1006 the "C=" log element. It can have useful tracking info from the
1007 destination system. Patch from Simon Arlott.
1009 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1010 length value. Previously this would segfault.
1012 HS/02 Support Avast multiline protoocol, this allows passing flags to
1013 newer versions of the scanner.
1015 JH/28 Ensure that variables possibly set during message acceptance are marked
1016 dead before release of memory in the daemon loop. This stops complaints
1017 about them when the debug_store option is enabled. Discovered specifically
1018 for sender_rate_period, but applies to a whole set of variables.
1019 Do the same for the queue-runner and queue-list loops, for variables set
1020 from spool message files. Do the same for the SMTP per-message loop, for
1021 certain variables indirectly set in ACL operations.
1023 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1024 as a multi-recipient message from a mailinglist manager). The coding had
1025 an arbitrary cutoff number of characters while checking for more input;
1026 enforced by writing a NUL into the buffer. This corrupted long / fast
1027 input. The problem was exposed more widely when more pipelineing of SMTP
1028 responses was introduced, and one Exim system was feeding another.
1029 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1030 receiving system, and refused recipients seen by the sending system
1031 (propating to people being dropped from mailing lists).
1032 Discovered and pinpointed by David Carter.
1034 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1035 replaced by the ${authresults } expansion.
1037 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1039 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1040 allows proper process termination in container environments.
1042 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1043 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1045 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1046 and "err_perm", deprecated since 4.83 when the RFC-defined words
1047 "temperror" and "permerror" were introduced.
1049 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1050 transport-filters or DKIM-signing. The restriction was lost in the
1051 consolidation of verify-callout and delivery SMTP handling.
1052 Extend the restriction to also cover ARC-signing.
1054 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1055 in defer=pass mode supply a 450 to the initiator. Previously the message
1058 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1059 tls_require_ciphers is used as before.
1061 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1062 "pass_unscanned". Only tmpfails from the scanner are written to
1063 the paniclog, as they may require admin intervention (permission
1064 denied, license issues). Other scanner errors (like decompression
1065 bombs) do not cause a paniclog entry.
1067 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1068 Previously it was possible to log spurious information in receive log
1071 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1072 triggered odd behaviour from Outlook Express clients.
1074 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1077 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1078 since the IETF WG has not yet settled on that versus the original
1079 "bare" representation.
1081 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1082 Previously the millisecond value corrupted the output.
1083 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1084 corrupted the output.
1090 JH/01 Rework error string handling in TLS interface so that the caller in
1091 more cases is responsible for logging. This permits library-sourced
1092 string to be attached to addresses during delivery, and collapses
1093 pairs of long lines into single ones.
1095 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1096 during configuration. Wildcards are allowed and expanded.
1098 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1099 This permits better logging.
1101 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1102 do not close it down and have the child transport start it up again on
1103 the passed-on TCP connection. Instead, proxy the child (and any
1104 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1105 affected: the continued delivery log lines do not have any DNSSEC, TLS
1106 Certificate or OCSP information. TLS cipher information is still logged.
1108 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1109 identical IP addresses on different listening ports. Will also affect
1112 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1113 add noisy ifdef guards to special-case this sillyness.
1114 Patch from Bernd Kuhls.
1116 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1117 than 255 are no longer allowed.
1119 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1120 Disable the session-cache too, which might reduce our load. Since we
1121 currrectly use a new context for every connection, both as server and
1122 client, there is no benefit for these.
1123 GnuTLS appears to not support tickets server-side by default (we don't
1124 call gnutls_session_ticket_enable_server()) but client side is enabled
1125 by default on recent versions (3.1.3 +) unless the PFS priority string
1128 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1129 <https://reproducible-builds.org/specs/source-date-epoch/>.
1131 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1132 the check for any unsuccessful recipients did not notice the limit, and
1133 erroneously found still-pending ones.
1135 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1136 MSG_MORE. Only in-clear (not on TLS connections).
1138 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1139 transport-filter is involved we can buffer the headers in memory for
1140 creating the signature, and read the spool data file once for the
1141 signature and again for transmission.
1143 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1144 4.77 as the kernel support then wasn't solid, having issues in 64bit
1145 mode. Now, it's been long enough. Add support for FreeBSD also.
1147 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1148 case where the routing stage had gathered several addresses to send to
1149 a host before calling the transport for the first, we previously failed
1150 to close down TLS in the old transport process before passing the TCP
1151 connection to the new process. The new one sent a STARTTLS command
1152 which naturally failed, giving a failed delivery and bloating the retry
1153 database. Investigation and fix prototype from Wolfgang Breyha.
1155 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1156 false-negatives in the check that the sender had not preempted a response
1157 or prompt from Exim (running as a server), due to that code's lack of
1158 awareness of the SMTP input buffering.
1160 PP/04 Add commandline_checks_require_admin option.
1161 Exim drops privileges sanely, various checks such as -be aren't a
1162 security problem, as long as you trust local users with access to their
1163 own account. When invoked by services which pass untrusted data to
1164 Exim, this might be an issue. Set this option in main configuration
1165 AND make fixes to the calling application, such as using `--` to stop
1168 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1169 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1170 responses to those, into a single TLS record each way (this usually means
1171 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1174 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1175 affects you only if you're dancing at the edge of the param size limits.
1176 If you are, and this message makes sense to you, then: raise the
1177 configured limit or use OpenSSL 1.1. Nothing we can do for older
1180 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1181 cmdline element to get the documented default one. Previously it was
1184 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1185 are unsuable for non-message-specific reasons - eg. connection timeout,
1186 banner-time rejection.
1188 JH/16 Fix logging of delivery remote port, when specified by router, under
1191 PP/06 Repair manualroute's ability to take options in any order, even if one
1192 is the name of a transport.
1195 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1197 JH/17 Change the list-building routines interface to use the expanding-string
1198 triplet model, for better allocation and copying behaviour.
1200 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1201 Previously it was constructed the first time a possibly-matching string
1202 was met in the configuration file input during startup; now it is done
1205 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1206 compatible one, to avoid the (poorly documented) possibility of a config
1207 file in the working directory redirecting the DB files, possibly correpting
1208 some existing file. CVE-2017-10140 assigned for BDB.
1210 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1211 cache-hot. Previously, although the result was properly cached, the
1212 initial verify call returned a defer.
1214 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1215 the main verify for receipient in uncached-mode.
1217 JH/22 Retire historical build files to an "unsupported" subdir. These are
1218 defined as "ones for which we have no current evidence of testing".
1220 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1221 if present. Previously it was ignored.
1223 JH/24 Start using specified-initialisers in C structure init coding. This is
1224 a C99 feature (it's 2017, so now considered safe).
1226 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1227 if was a fixed-sized field and bitmask ops via macros; it is now more
1230 PP/07 GitHub PR 56: Apply MariaDB build fix.
1231 Patch provided by Jaroslav Škarvada.
1233 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1234 during Coverity cleanups [4.87 JH/47]
1235 Diagnosis and fix provided by Michael Fischer v. Mollard.
1237 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1238 the right size to place the terminating semicolon on its own folded
1239 line, the header hash was calculated to an incorrect value thanks to
1240 the (relaxed) space the fold became.
1242 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1243 and confused the parent.
1245 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1246 which could crash as a result. This could lead to undeliverable messages.
1248 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1251 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1252 matters on fast-turnover and PID-randomising systems, which were getting
1253 out-of-order delivery.
1255 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1256 a possibly-overlapping copy. The symptom was that "Remote host closed
1257 connection in response to HELO" was logged instead of the actual 4xx
1260 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1261 Previously only that bufferd was discarded, resulting in SYMTP command
1264 JH/32 DKIM: when a message has multiple signatures matching an identity given
1265 in dkim_verify_signers, run the dkim acl once for each. Previously only
1266 one run was done. Bug 2189.
1268 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1269 "panic the current process" to "deliberately defer". The panic log is
1270 still written with the problem list name; the mail and reject logs now
1271 get a temp-reject line for the message that was being handled, saying
1272 something like "domains check lookup or other defer". The SMTP 451
1273 message is still "Temporary local problem".
1275 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1276 A crafted sequence of BDAT commands could result in in-use memory beeing
1277 freed. CVE-2017-16943.
1279 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1280 from SMTP input. Previously it was always done; now only done for DATA
1281 and not BDAT commands. CVE-2017-16944.
1283 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1284 to the message (such as an overlong header line). Previously this was
1285 not done and we did not exit BDAT mode. Followon from the previous item
1286 though a different problem.
1292 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1293 than -2003 did; needs libidn2 in addition to libidn.
1295 JH/02 The path option on a pipe transport is now expanded before use.
1297 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1298 Patch provided by "Björn", documentation fix added too.
1300 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1301 missing a wire-to-host endian conversion.
1303 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1304 close after a BDAT command line could be taken as a following command,
1305 giving a synch failure. Fix by only checking for synch immediately
1306 before acknowledging the chunk.
1308 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1309 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1310 macro. Patches provided by Josh Soref.
1312 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1313 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1314 by IANA as a service extension. However, John Klensin suggests that we
1317 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1318 the dkim code may be unix-mode line endings rather than smtp wire-format
1319 CRLF, so prepend a CR to any bare LF.
1321 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1322 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1324 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1325 the first were themselves being wrongly included in the feed into dkim
1326 processing; with most chunk sizes in use this resulted in an incorrect
1327 body hash calculated value.
1329 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1330 DKIM signature block, for verification. Although advised against by
1331 standards it is specifically not ruled illegal.
1333 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1335 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1336 missing a body hash (the bh= tag).
1338 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1339 It seems that HAProxy sends the Proxy Protocol information in clear and
1340 only then does a TLS startup, so do the same.
1342 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1343 TCP connections (such as for Spamd) unless the daemon successfully set
1344 Fast Open mode on its listening sockets. This fixes breakage seen on
1345 too-old kernels or those not configured for Fast Open, at the cost of
1346 requiring both directions being enabled for TFO, and TFO never being used
1347 by non-daemon-related Exim processes.
1349 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1350 endings, at least on the first header line. Try to canonify any that get
1351 past that check, despite the cost.
1353 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1354 now limited to an arbitrary five deep, while parsing addresses with the
1355 strip_excess_angle_brackets option enabled.
1357 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1358 instead leave the unprompted TLS handshake in socket buffer for the
1359 TLS library to consume.
1361 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1363 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1365 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1366 region in one operation, for speed, and this leaves assigned pointers
1367 dangling. Add checks run only under the testsuite which checks all
1368 variables at a store-reset and panics on a dangling pointer; add code
1369 explicitly nulling out all the variables discovered. Fixes one known
1370 bug: a transport crash, where a dangling pointer for $sending_ip_address
1371 originally assigned in a verify callout, is re-used.
1373 PP/06 Drop '.' from @INC in various Perl scripts.
1375 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1377 PP/08 Reduce a number of compilation warnings under clang; building with
1378 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1379 should be warning-free.
1381 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1383 HS/01 Fix portability problems introduced by PP/08 for platforms where
1384 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1386 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1387 chunk. This allows us to accept broken chunked messages. We need a more
1388 general solution here.
1390 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1391 already-broken messages in the queue.
1393 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1395 JH/19 Fix reference counting bug in routing-generated-address tracking.
1401 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1402 supports it and a size is available (ie. the sending peer gave us one).
1404 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1405 years of being deprecated). The replacements are the ACLs
1406 acl_smtp_mime and acl_not_smtp_mime.
1408 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1409 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1410 or even an in-clear connection were permitted. Now, if the host lookup
1411 was dnssec and dane was requested then the host is only used if the
1412 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1413 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1414 if one fails this test.
1415 This means that a poorly-configured remote DNS will make it incommunicado;
1416 but it protects against a DNS-interception attack on it.
1418 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1419 non-noisy when a race steals the message being considered.
1421 JH/05 If main configuration option tls_certificate is unset, generate a
1422 self-signed certificate for inbound TLS connections.
1424 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1425 in rewrites and routers.
1427 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1428 and logged a warning sing 4.83; now they are a configuration file error.
1430 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1431 (lacking @domain). Apply the same qualification processing as RCPT.
1433 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1435 JH/10 Support ${sha256:} applied to a string (as well as the previous
1438 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1439 a cutthrough deliver is pending, as we always want to make a connection.
1440 This also avoids re-routing the message when later placing the cutthrough
1441 connection after a verify cache hit.
1442 Do not update it with the verify result either.
1444 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1445 when routing results in more than one destination address.
1447 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1448 signing (which inhibits the cutthrough capability). Previously only
1449 the presence of an option was tested; now an expansion evaluating as
1450 empty is permissible (obviously it should depend only on data available
1451 when the cutthrough connection is made).
1453 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1454 the relevant preceding SMTP command did not note the pipelining mode.
1456 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1457 Previously they were not counted.
1459 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1460 as one having no matching records. Previously we deferred the message
1461 that needed the lookup.
1463 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1464 distinguished as "(=".
1466 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1467 for missing MX records. Previously it only worked for missing A records.
1469 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1471 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1472 after the data-go-ahead and data-ack. Patch from Jason Betts.
1474 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1475 even for a "none" policy. Patch from Tony Meyer.
1477 JH/22 Fix continued use of a connection for further deliveries. If a port was
1478 specified by a router, it must also match for the delivery to be
1481 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1482 When one of the recipients of a message was unsuitable for the connection
1483 (has no matching addresses), we lost track of needing to mark it
1484 deferred. As a result mail would be lost.
1486 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1488 JH/25 Decoding ACL controls is now done using a binary search; the source code
1489 takes up less space and should be simpler to maintain. Merge the ACL
1490 condition decode tables also, with similar effect.
1492 JH/26 Fix problem with one_time used on a redirect router which returned the
1493 parent address unchanged. A retry would see the parent address marked as
1494 delivered, so not attempt the (identical) child. As a result mail would
1497 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1498 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1499 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1502 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1503 Main config options check_{log,spool}_{inodes,space} are now
1504 100 inodes, 10MB unless set otherwise in the configuration.
1506 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1507 Previously it only applied to the main-section connection policy
1510 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1512 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1513 by me. Added RFC7919 DH primes as an alternative.
1515 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1516 are not in the system include path.
1518 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1519 GnuTLS, when a session startup failed (eg because the client disconnected)
1520 Exim did stdio operations after fclose. This was exposed by a recent
1521 change which nulled out the file handle after the fclose.
1523 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1524 signed directly by the cert-signing cert, rather than an intermediate
1525 OCSP-signing cert. This is the model used by LetsEncrypt.
1527 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1529 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1530 an incoming connection.
1532 HS/02 Bug 1802: Do not half-close the connection after sending a request
1535 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1536 fallback to "prime256v1".
1538 JH/34 SECURITY: Use proper copy of DATA command in error message.
1539 Could leak key material. Remotely exploitable. CVE-2016-9963.
1545 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1546 and 3.4.4 - once the server is enabled to respond to an OCSP request
1547 it does even when not requested, resulting in a stapling non-aware
1548 client dropping the TLS connection.
1550 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1551 support variable-length bit vectors. No functional change.
1553 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1554 The I= interface field on outgoing lines is now after the H= remote
1555 host field, same as incoming lines. There is a separate
1556 outgoing_interface log selector which allows you to disable the
1559 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1560 If not running log_selector +smtp_connection the mainlog would be held
1561 open indefinitely after a "too many connections" event, including to a
1562 deleted file after a log rotate. Leave the per net connection logging
1563 leaving it open for efficiency as that will be quickly detected by the
1564 check on the next write.
1566 HS/01 Bug 1671: Fix post transport crash.
1567 Processing the wait-<transport> messages could crash the delivery
1568 process if the message IDs didn't exist for some reason. When
1569 using 'split_spool_directory=yes' the construction of the spool
1570 file name failed already, exposing the same netto behaviour.
1572 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1573 mime_regex ACL conditions.
1575 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1576 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1577 to HELO, local diagnostic string.
1579 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1580 log line to debug. Even when configured with a tls authenticator many
1581 client connections are expected to not authenticate in this way, so
1582 an authenticate fail is not an error.
1584 HS/02 Add the Exim version string to the process info. This way exiwhat
1585 gives some more detail about the running daemon.
1587 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1588 matter for fast-change records such as DNSBLs.
1590 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1591 retry record, even if constant. There may be multiple transports with
1592 different interface settings and the retry behaviour needs to be kept
1595 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1597 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1599 JH/10 Bug 840: fix log_defer_output option of pipe transport
1601 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1602 hosts_randomize. This matters a lot when combined with mua_wrapper.
1604 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1605 ${quote_pgsql:<string>} operator.
1607 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1609 JH/14 Change header file name for freeradius-client. Relevant if compiling
1610 with Radius support; from the Gentoo tree and checked under Fedora.
1612 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1614 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1615 option transport_filter, meaning no filtering.
1617 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1619 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1620 defaults to "*" (all hosts). The variable is now available when not built
1621 with TLS, default unset, mainly to enable keeping the testsuite sane.
1622 If a server certificate is not supplied (via tls_certificate) an error is
1623 logged, and clients will find TLS connections fail on startup. Presumably
1624 they will retry in-clear.
1625 Packagers of Exim are strongly encouraged to create a server certificate
1626 at installation time.
1628 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1629 with the $config_file variable.
1631 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1632 in transport context, after the attempt, and per-recipient. The latter type
1633 is per host attempted. The event data is the error message, and the errno
1634 information encodes the lookup type (A vs. MX) used for the (first) host,
1635 and the trailing two digits of the smtp 4xx response.
1637 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1638 to write to mainlog (or rejectlog, paniclog) in the window between file
1639 creation and permissions/ownership being changed. Particularly affects
1640 installations where exicyclog is run as root, rather than exim user;
1641 result is that the running daemon panics and dies.
1643 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1645 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1646 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1647 "pri" and "weight". Note that the previous implicit priority given by the
1648 list order is no longer honoured.
1650 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1651 for DKIM processing.
1653 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1654 by defining SUPPORT_SOCKS.
1656 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1657 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1658 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1659 variables are renamed to proxy_{local,external}_{address,port}.
1661 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1662 for a build by defining SUPPORT_I18N
1664 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1665 of the query string, and make ${quote_redis:} do that quoting.
1667 JH/29 Move Events support from Experimental to mainline, enabled by default
1668 and removable for a build by defining DISABLE_EVENT.
1670 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1672 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1673 cached by the daemon.
1675 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1676 by defining LOOKUP_REDIS. The libhiredis library is required.
1678 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1679 keys are given for lookup.
1681 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1682 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1683 only supported when built with TLS support. The PolarSSL SHA routines
1684 are still used when the TLS library is too old for convenient support.
1686 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1687 openssl_options), for security. OpenSSL forces this from version 1.1.0
1688 server-side so match that on older versions.
1690 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1691 allocation for $value could be released as the expansion processing
1692 concluded, but leaving the global pointer active for it.
1694 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1695 and to use the domains and local_parts ACL conditions.
1697 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1698 incorrectly not doubled on cutthrough transmission, hence seen as a
1699 body-termination at the receiving system - resulting in truncated mails.
1700 Commonly the sender saw a TCP-level error, and retransmitted the message
1701 via the normal store-and-forward channel. This could result in duplicates
1702 received - but deduplicating mailstores were liable to retain only the
1703 initial truncated version.
1705 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1707 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1709 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1710 we're in there, support oversigning also; bug 1309.
1712 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1714 HS/04 Add support for keep_environment and add_environment options.
1716 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1717 either intentional arithmetic overflow during PRNG, or testing config-
1720 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1721 delivery resulted in actual delivery. Cancel cutthrough before DATA
1724 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1725 rejects a recipient: pass the reject to the originator.
1727 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1728 Many were false-positives and ignorable, but it's worth fixing the
1731 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1732 for the new environment-manipulation done at startup. Move the routines
1733 from being local to tls.c to being global via the os.c file.
1735 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1736 an extract embedded as result-arg for a map, the first arg for extract
1737 is unavailable so we cannot tell if this is a numbered or keyed
1738 extraction. Accept either.
1744 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1747 JH/02 The smtp transport option "multi_domain" is now expanded.
1749 JH/03 The smtp transport now requests PRDR by default, if the server offers
1752 JH/04 Certificate name checking on server certificates, when exim is a client,
1753 is now done by default. The transport option tls_verify_cert_hostnames
1754 can be used to disable this per-host. The build option
1755 EXPERIMENTAL_CERTNAMES is withdrawn.
1757 JH/05 The value of the tls_verify_certificates smtp transport and main options
1758 default to the word "system" to access the system default CA bundle.
1759 For GnuTLS, only version 3.0.20 or later.
1761 JH/06 Verification of the server certificate for a TLS connection is now tried
1762 (but not required) by default. The verification status is now logged by
1763 default, for both outbound TLS and client-certificate supplying inbound
1766 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1769 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1770 Status Notification (bounce) messages are now MIME format per RFC 3464.
1771 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1772 under the control of the dsn_advertise_hosts option, and routers may
1773 have a dsn_lasthop option.
1775 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1776 default, modifiable by a malware= option. The list separator for
1777 the options can now be changed in the usual way. Bug 68.
1779 JH/10 The smtp_receive_timeout main option is now expanded before use.
1781 JH/11 The incoming_interface log option now also enables logging of the
1782 local interface on delivery outgoing connections.
1784 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1785 if the interface and destination host and port all match.
1787 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1790 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1791 Patch from Andrew Lewis.
1793 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1794 now supports optional time-restrictions, weighting, and priority
1795 modifiers per server. Patch originally by <rommer@active.by>.
1797 JH/16 The spamd_address main option now supports a mixed list of local
1798 and remote servers. Remote servers can be IPv6 addresses, and
1799 specify a port-range.
1801 JH/17 Bug 68: The spamd_address main option now supports an optional
1802 timeout value per server.
1804 JH/18 Bug 1581: Router and transport options headers_add/remove can
1805 now have the list separator specified.
1807 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1810 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1813 JH/21 Support for the A6 type of dns record is withdrawn.
1815 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1816 rather than the verbs used.
1818 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1819 from 255 to 1024 chars.
1821 JH/24 Verification callouts now attempt to use TLS by default.
1823 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1824 are generic router options now. The defaults didn't change.
1826 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1827 Original patch from Alexander Shikoff, worked over by JH.
1829 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1830 files if ZCAT_COMMAND is not executable.
1832 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1834 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1836 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1837 Normally benign, it bites when the pair was led to by a CNAME;
1838 modern usage is to not canonicalize the domain to a CNAME target
1839 (and we were inconsistent anyway for A-only vs AAAA+A).
1841 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1843 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1844 when evaluating $sender_host_dnssec.
1846 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1847 $sender_helo_dnssec variable.
1849 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1851 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1853 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1855 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1856 documented as working, but never had. Support all but $spam_report.
1858 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1859 added for tls authenticator.
1861 HS/03 Add perl_taintmode main config option
1867 TL/01 When running the test suite, the README says that variables such as
1868 no_msglog_check are global and can be placed anywhere in a specific
1869 test's script, however it was observed that placement needed to be near
1870 the beginning for it to behave that way. Changed the runtest perl
1871 script to read through the entire script once to detect and set these
1872 variables, reset to the beginning of the script, and then run through
1873 the script parsing/test process like normal.
1875 TL/02 The BSD's have an arc4random API. One of the functions to induce
1876 adding randomness was arc4random_stir(), but it has been removed in
1877 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1878 function when detected.
1880 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1881 cause callback expansion.
1883 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1884 syntax errors in an expansion can be treated as a string instead of
1885 logging or causing an error, due to the internal use of bool_lax
1886 instead of bool when processing it.
1888 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1889 server certificates when making smtp deliveries.
1891 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1893 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1895 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1897 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1898 Merged patch from Sebastian Wiedenroth.
1900 JH/05 Fix results-pipe from transport process. Several recipients, combined
1901 with certificate use, exposed issues where response data items split
1902 over buffer boundaries were not parsed properly. This eventually
1903 resulted in duplicates being sent. This issue only became common enough
1904 to notice due to the introduction of connection certificate information,
1905 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1907 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1908 size buffer was used, resulting in syntax errors when an expansion
1911 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1912 version 3.3.6 or later.
1914 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1915 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1916 both become "event_action", the variables become $event_name, $event_data
1917 and $event_defer_errno. There is a new variable $verify_mode, usable in
1918 routers, transports and related events. The tls:cert event is now also
1919 raised for inbound connections, if the main configuration event_action
1922 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1923 early OCSP support, but no stapling (appears to be less than 1.0.0).
1925 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1926 server certificate names available under the smtp transport option
1927 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1930 JH/10 Time-related extraction expansions from certificates now use the main
1931 option "timezone" setting for output formatting, and are consistent
1932 between OpenSSL and GnuTLS compilations. Bug 1541.
1934 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1935 encoded parameter in the incoming message. Bug 1558.
1937 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1938 include certificate info, eximon was claiming there were spoolfile
1941 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1943 JH/14 Log delivery-related information more consistently, using the sequence
1944 "H=<name> [<ip>]" wherever possible.
1946 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1947 are problematic for Debian distribution, omit them from the release
1950 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1952 JH/16 Fix string representation of time values on 64bit time_t architectures.
1955 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1956 output list separator was used.
1961 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1962 checkers that were complaining about end of non-void function with no
1965 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1966 This was a regression introduced in 4.83 by another bugfix.
1968 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1970 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1971 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1977 TF/01 Correctly close the server side of TLS when forking for delivery.
1979 When a message was received over SMTP with TLS, Exim failed to clear up
1980 the incoming connection properly after forking off the child process to
1981 deliver the message. In some situations the subsequent outgoing
1982 delivery connection happened to have the same fd number as the incoming
1983 connection previously had. Exim would try to use TLS and fail, logging
1984 a "Bad file descriptor" error.
1986 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1987 utilities have not been installed.
1989 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1990 temporary space as the ACL may create new global variables.
1992 TL/01 LDAP support uses per connection or global context settings, depending
1993 upon the detected version of the libraries at build time.
1995 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1996 to extract and use the src ip:port in logging and expansions as if it
1997 were a direct connection from the outside internet. PPv2 support was
1998 updated based on HAProxy spec change in May 2014.
2000 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2002 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2003 Properly escape header and check for NULL return.
2005 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2008 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2010 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2011 characters in header names, implemented as a verify condition.
2012 Contributed by Michael Fischer v. Mollard.
2014 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2015 results permerror and temperror. Previous values are deprecated but
2016 still accepted. In a future release, err_perm and err_temp will be
2017 completely removed, which will be a backward incompatibility if the
2018 ACL tests for either of these two old results. Patch contributed by
2019 user bes-internal on the mailing list.
2021 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2023 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2024 selectors, in both main and reject logs.
2026 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2029 JH/07 Add malware type "sock" for talking to simple daemon.
2031 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2033 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2034 routers/transports under cutthrough routing.
2036 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2037 numbers. Touch up "bool" conditional to keep the same definition.
2039 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2041 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2043 JH/12 Expand items in router/transport headers_add or headers_remove lists
2044 individually rather than the list as a whole. Bug 1452.
2046 Required for reasonable handling of multiple headers_ options when
2047 they may be empty; requires that headers_remove items with embedded
2048 colons must have them doubled (or the list-separator changed).
2050 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2051 view the policy declared in the DMARC record. Currently, $dmarc_status
2052 is a combined value of both the record presence and the result of the
2055 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2057 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2058 dnslookup router and the smtp transport (applying to the forward
2061 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2062 of ldap servers used for a specific lookup. Patch provided by Heiko
2065 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2066 New variable $lookup_dnssec_authenticated for observability.
2068 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2069 Patch submitted by Lars Timman.
2071 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2073 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2074 Requires trusted mode and valid format message id, aborts otherwise.
2075 Patch contributed by Heiko Schlichting.
2077 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2078 certextract with support for various fields. Bug 1358.
2080 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2081 is requested by default, modifiable by smtp transport option
2084 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2085 operate on certificate variables to give certificate fingerprints
2086 Also new ${sha256:cert_variable}.
2088 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2090 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2091 Christian Aistleitner.
2093 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2095 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2096 file. Patch from Wolfgang Breyha.
2098 JH/25 Expand the coverage of the delivery $host and $host_address to
2099 client authenticators run in verify callout. Bug 1476.
2101 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2102 align with daemon_smtp_ports. Bug 72.
2104 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2105 support and error reporting did not work properly.
2107 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2108 and is readable. Patch from Andrew Colin Kissa.
2110 TL/14 Enhance documentation of ${run expansion and how it parses the
2111 commandline after expansion, particularly in the case when an
2112 unquoted variable expansion results in an empty value.
2114 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2116 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2117 by renaming away from it. Addresses GH issue 15, reported by
2120 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2122 TL/15 SECURITY: prevent double expansion in math comparison functions
2123 (can expand unsanitized data). Not remotely exploitable.
2130 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2132 PP/02 Make -n do something, by making it not do something.
2133 When combined with -bP, the name of an option is not output.
2135 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2138 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2139 $sender_host_name and config options to manage this, and basic check
2142 PP/05 DSCP support for outbound connections and control modifier for inbound.
2144 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2145 (Only plugin which currently uses this is kerberos4, which nobody should
2146 be using, but we should make it available and other future plugins might
2147 conceivably use it, even though it would break NAT; stuff *should* be
2148 using channel bindings instead).
2150 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2151 name; added for Sendmail compatibility; requires admin caller.
2152 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2153 just ignore it); requires trusted caller.
2154 Also parse but ignore: -Ac -Am -X<logfile>
2157 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2159 TL/02 Add +smtp_confirmation as a default logging option.
2161 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2162 Patch by Magnus Holmgren from 2007-02-20.
2164 TL/04 Bugzilla 1281 - Spec typo.
2165 Bugzilla 1283 - Spec typo.
2166 Bugzilla 1290 - Spec grammar fixes.
2168 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2170 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2172 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2173 mailing list by by Dmitry Isaikin.
2175 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2177 JH/02 Support "G" suffix to numbers in ${if comparisons.
2179 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2181 NM/01 Bugzilla 1197 - Spec typo
2182 Bugzilla 1196 - Spec examples corrections
2184 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2186 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2187 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2190 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2191 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2193 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2194 "acl {{name}{arg}...}", and optional args on acl condition
2197 JH/05 Permit multiple router/transport headers_add/remove lines.
2199 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2201 JH/07 Avoid using a waiting database for a single-message-only transport.
2202 Performance patch from Paul Fisher. Bugzilla 1262.
2204 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2207 JH/09 Add $headers_added variable, with content from use of ACL modifier
2208 add_header (but not yet added to the message). Bugzilla 199.
2210 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2211 Pulled from Bugzilla 817 by Wolfgang Breyha.
2213 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2215 (nb: this is the same fix as in Exim 4.80.1)
2217 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2220 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2223 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2225 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2226 advertise SMTP AUTH mechanism to us, instead of a generic
2227 protocol violation error. Also, make Exim more robust to bad
2228 data from the Dovecot auth socket.
2230 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2232 When a queue runner is handling a message, Exim first routes the
2233 recipient addresses, during which it prunes them based on the retry
2234 hints database. After that it attempts to deliver the message to
2235 any remaining recipients. It then updates the hints database using
2238 So if a recipient address works intermittently, it can get repeatedly
2239 deferred at routing time. The retry hints record remains fresh so the
2240 address never reaches the final cutoff time.
2242 This is a fairly common occurrence when a user is bumping up against
2243 their storage quota. Exim had some logic in its local delivery code
2244 to deal with this. However it did not apply to per-recipient defers
2245 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2247 This change adds a proper retry rule check during routing so that the
2248 final cutoff time is checked against the message's age. We only do
2249 this check if there is an address retry record and there is not a
2250 domain retry record; this implies that previous attempts to handle
2251 the address had the retry_use_local_parts option turned on. We use
2252 this as an approximation for the destination being like a local
2253 delivery, as in LMTP.
2255 I suspect this new check makes the old local delivery cutoff check
2256 redundant, but I have not verified this so I left the code in place.
2258 TF/02 Correct gecos expansion when From: is a prefix of the username.
2260 Test 0254 submits a message to Exim with the header
2264 When I ran the test suite under the user fanf2, Exim expanded
2265 the header to contain my full name, whereas it should have added
2266 a Resent-Sender: header. It erroneously treats any prefix of the
2267 username as equal to the username.
2269 This change corrects that bug.
2271 GF/01 DCC debug and logging tidyup
2272 Error conditions log to paniclog rather than rejectlog.
2273 Debug lines prefixed by "DCC: " to remove any ambiguity.
2275 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2277 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2278 Bug spotted by Jeremy Harris; was flawed since initial commit.
2279 Would have resulted in OCSP responses post-SNI triggering an Exim
2280 NULL dereference and crash.
2282 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2284 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2285 Bug detection, analysis and fix by Samuel Thibault.
2286 Bugzilla 1331, Debian bug #698092.
2288 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2290 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2291 Server implementation by Todd Lyons, client by JH.
2292 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2293 config variable "prdr_enable" controls whether the server
2294 advertises the facility. If the client requests PRDR a new
2295 acl_data_smtp_prdr ACL is called once for each recipient, after
2296 the body content is received and before the acl_smtp_data ACL.
2297 The client is controlled by both of: a hosts_try_prdr option
2298 on the smtp transport, and the server advertisement.
2299 Default client logging of deliveries and rejections involving
2300 PRDR are flagged with the string "PRDR".
2302 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2303 fclose(). Diagnosis by Todd Lyons.
2305 PP/17 Update configure.default to handle IPv6 localhost better.
2306 Patch by Alain Williams (plus minor tweaks).
2309 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2310 This is now consistent with GnuTLS, and is now documented: the
2311 previous undocumented portable approach to treating the option as
2312 unset was to force an expansion failure. That still works, and
2313 an empty string is now equivalent.
2315 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2316 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2317 not performing validation itself.
2319 PP/20 Added force_command boolean option to pipe transport.
2320 Patch from Nick Koston, of cPanel Inc.
2322 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2325 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2327 PP/21 Fix eximon continuous updating with timestamped log-files.
2328 Broken in a format-string cleanup in 4.80, missed when I repaired the
2329 other false fix of the same issue.
2330 Report and fix from Heiko Schlichting.
2333 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2334 Report from Prashanth Katuri.
2336 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2337 It's SecureTransport, so affects any MacOS clients which use the
2338 system-integrated TLS libraries, including email clients.
2340 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2341 using a MIME ACL for non-SMTP local injection.
2342 Report and assistance in diagnosis by Warren Baker.
2344 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2346 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2348 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2349 last id that failed so it may be referenced in subsequent ACL's.
2351 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2354 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2355 ldap library initialization, allowing self-signed CA's to be
2356 used. Also properly sets require_cert option later in code by
2357 using NULL (global ldap config) instead of ldap handle (per
2358 session). Bug diagnosis and testing by alxgomz.
2360 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2361 the src/util/ subdirectory.
2363 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2364 renamed to Transport Post Delivery Action by Jeremy Harris, as
2367 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2368 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2369 redis_servers = needs to be configured which will be used by the redis
2370 lookup. Patch from Warren Baker, of The Packet Hub.
2372 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2374 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2375 hostname or reverse DNS when processing a host list. Used suggestions
2376 from multiple comments on this bug.
2378 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2380 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2381 Missed a few lines, added it to make the runtest require no keyboard
2384 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2385 contains upper case chars. Make router use caseful_local_part.
2387 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2388 support when GnuTLS has been built with p11-kit.
2394 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2396 This, or similar/improved, will also be change PP/11 of 4.82.
2402 PP/01 Handle short writes when writing local log-files.
2403 In practice, only affects FreeBSD (8 onwards).
2404 Bugzilla 1053, with thanks to Dmitry Isaikin.
2406 NM/01 Bugzilla 949 - Documentation tweak
2408 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2411 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2413 PP/02 Implemented gsasl authenticator.
2415 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2417 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2418 `pkg-config foo` for cflags/libs.
2420 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2421 with rest of GSASL and with heimdal_gssapi.
2423 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2424 `pkg-config foo` for cflags/libs for the TLS implementation.
2426 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2427 properties get this fed in as external SSF. A number of robustness
2428 and debugging improvements to the cyrus_sasl authenticator.
2430 PP/08 cyrus_sasl server now expands the server_realm option.
2432 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2433 Patch by Jeremy Harris.
2435 PP/10 Added dbmjz lookup type.
2437 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2439 PP/12 MAIL args handles TAB as well as SP, for better interop with
2440 non-compliant senders.
2441 Analysis and variant patch by Todd Lyons.
2443 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2444 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2445 Patch from Dirk Mueller <dmueller@suse.com>
2447 PP/13 tls_peerdn now print-escaped for spool files.
2448 Observed some $tls_peerdn in wild which contained \n, which resulted
2449 in spool file corruption.
2451 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2452 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2453 or write after TLS renegotiation, which otherwise led to messages
2456 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2457 as a tracking header (ie: a signed header comes before the signature).
2458 Patch from Wolfgang Breyha.
2460 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2461 comma-sep list; embedded commas doubled.
2463 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2465 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2467 Report and patch from Dmitry Banschikov.
2469 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2470 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2471 failures. We appear to now support TLS1.1+ with Exim.
2473 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2474 lets Exim select keys and certificates based upon TLS SNI from client.
2475 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2476 before an outbound SMTP session. New log_selector, +tls_sni.
2478 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2479 NULL dereference. Report and patch from Alun Jones.
2481 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2482 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2483 Not seeing resolver debug output on NetBSD, but suspect this is a
2484 resolver implementation change.
2486 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2487 Left warnings. Added "eximon gdb" invocation mode.
2489 PP/21 Defaulting "accept_8bitmime" to true, not false.
2491 PP/22 Added -bw for inetd wait mode support.
2493 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2494 locate the relevant includes and libraries. Made this the default.
2496 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2497 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2499 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2500 This may cause build issues on older platforms.
2502 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2503 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2504 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2505 Added SNI support via GnuTLS too.
2506 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2508 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2510 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2511 Applied second patch from Janne, implementing suggestion to default
2512 multiple-strings-in-record handling to match SPF spec.
2514 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2516 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2517 read-only, out of scope).
2518 Patch from Wolfgang Breyha, report from Stuart Northfield.
2520 PP/29 Fix three issues highlighted by clang analyser static analysis.
2521 Only crash-plausible issue would require the Cambridge-specific
2522 iplookup router and a misconfiguration.
2523 Report from Marcin Mirosław.
2525 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2527 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2528 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2529 As part of this, removing so much warning spew let me fix some minor
2530 real issues in debug logging.
2532 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2533 assignment on my part. Fixed.
2535 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2536 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2537 Janne Snabb (who went above and beyond: thank you).
2539 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2540 string otherwise requires a connection and a bunch more work and it's
2541 relatively easy to get wrong. Should also expose TLS library linkage
2544 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2545 64-bit ${eval} (JH/03).
2547 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2548 GNU libc to support some of the 64-bit stuff, should not lead to
2549 conflicts. Defined before os.h is pulled in, so if a given platform
2550 needs to override this, it can.
2552 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2553 protection layer was required, which is not implemented.
2554 Bugzilla 1254, patch from Wolfgang Breyha.
2556 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2557 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2558 tls_dhparam take prime identifiers. Also unbreak combination of
2559 OpenSSL+DH_params+TLSSNI.
2561 PP/39 Disable SSLv2 by default in OpenSSL support.
2567 PP/01 Solaris build fix for Oracle's LDAP libraries.
2568 Bugzilla 1109, patch from Stephen Usher.
2570 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2572 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2575 TF/02 Fix a couple more cases where we did not log the error message
2576 when unlink() failed. See also change 4.74-TF/03.
2578 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2579 lock up or crash if it happened to be inside a call to libc when it
2580 got a SIGUSR1 from exiwhat.
2582 The SIGUSR1 handler appends the current process status to the process
2583 log which is later printed by exiwhat. It used to use the general
2584 purpose logging code to do this, but several functions it calls are
2585 not safe for signals.
2587 The new output code in the SIGUSR1 handler is specific to the process
2588 log, and simple enough that it's easy to inspect for signal safety.
2589 Removing some special cases also simplifies the general logging code.
2590 Removing the spurious timestamps from the process log simplifies
2593 TF/04 Improved ratelimit ACL condition.
2595 The /noupdate option has been deprecated in favour of /readonly which
2596 has clearer semantics. The /leaky, /strict, and /readonly update modes
2597 are mutually exclusive. The update mode is no longer included in the
2598 database key; it just determines when the database is updated. (This
2599 means that when you upgrade Exim will forget old rate measurements.)
2601 Exim now checks that the per_* options are used with an update mode that
2602 makes sense for the current ACL. For example, when Exim is processing a
2603 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2604 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2605 must specify per_mail/readonly. If you omit the update mode it defaults to
2606 /leaky where that makes sense (as before) or /readonly where required.
2608 The /noupdate option is now undocumented but still supported for
2609 backwards compatibility. It is equivalent to /readonly except that in
2610 ACLs where /readonly is required you may specify /leaky/noupdate or
2611 /strict/noupdate which are treated the same as /readonly.
2613 A useful new feature is the /count= option. This is a generalization
2614 of the per_byte option, so that you can measure the throughput of other
2615 aggregate values. For example, the per_byte option is now equivalent
2616 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2618 The per_rcpt option has been generalized using the /count= mechanism
2619 (though it's more complicated than the per_byte equivalence). When it is
2620 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2621 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2622 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2623 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2624 Note that using per_rcpt with a non-readonly update mode in more than
2625 one ACL will cause the recipients to be double-counted. (The per_mail
2626 and per_byte options don't have this problem.)
2628 The handling of very low rates has changed slightly. If the computed rate
2629 is less than the event's count (usually one) then this event is the first
2630 after a long gap. In this case the rate is set to the same as this event's
2631 count, so that the first message of a spam run is counted properly.
2633 The major new feature is a mechanism for counting the rate of unique
2634 events. The new per_addr option counts the number of different
2635 recipients that someone has sent messages to in the last time period. It
2636 behaves like per_rcpt if all the recipient addresses are different, but
2637 duplicate recipient addresses do not increase the measured rate. Like
2638 the /count= option this is a general mechanism, so the per_addr option
2639 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2640 example, measure the rate that a client uses different sender addresses
2641 with the options per_mail/unique=$sender_address. There are further
2642 details in the main documentation.
2644 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2646 TF/06 Removed a few PCRE remnants.
2648 TF/07 Automatically extract Exim's version number from tags in the git
2649 repository when doing development or release builds.
2651 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2652 Bugzilla 879. Patch from Paul Fisher.
2654 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2655 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2658 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2660 PP/05 Variable $av_failed, true if the AV scanner deferred.
2661 Bugzilla 1078. Patch from John Horne.
2663 PP/06 Stop make process more reliably on build failure.
2664 Bugzilla 1087. Patch from Heiko Schlittermann.
2666 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2667 Bugzilla 1089. Patch from Heiko Schlittermann.
2669 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2670 Bugzilla 1131. Patch from Holger Weiß.
2672 PP/09 Handle IPv6 addresses with SPF.
2673 Bugzilla 860. Patch from Wolfgang Breyha.
2675 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2677 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2680 PP/11 match_* no longer expand right-hand-side by default.
2681 New compile-time build option, EXPAND_LISTMATCH_RHS.
2682 New expansion conditions, "inlist", "inlisti".
2684 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2686 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2688 PP/14 fix log_write() format string regression from TF/03.
2689 Bugzilla 1152. Patch from Dmitry Isaikin.
2695 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2697 PP/02 Harmonised TLS library version reporting; only show if debugging.
2698 Layout now matches that introduced for other libraries in 4.74 PP/03.
2700 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2702 PP/04 New "dns_use_edns0" global option.
2704 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2707 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2708 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2710 TK/01 Updated PolarSSL code to 0.14.2.
2711 Bugzilla 1097. Patch from Andreas Metzler.
2713 PP/07 Catch divide-by-zero in ${eval:...}.
2714 Fixes bugzilla 1102.
2716 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2719 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2720 format-string attack -- SECURITY: remote arbitrary code execution.
2722 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2723 time unintentionally subject to list matching rules, letting the header
2724 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2725 arbitrary string expansion). This allowed for information disclosure.
2727 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2728 INT_MIN/-1 -- value coerced to INT_MAX.
2734 NM/01 Workaround for PCRE version dependency in version reporting
2737 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2738 This fixes portability to compilers other than gcc, notably
2739 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2741 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2742 makefiles for portability to HP-UX and POSIX correctness.
2744 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2745 Also via indented variable definition in the Makefile.
2746 (Debugging by Oliver Heesakkers).
2748 PP/02 Restore caching of spamd results with expanded spamd_address.
2749 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2751 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2752 Improves build reliability. Fix from: Frank Elsner
2754 NM/02 Fix wide character breakage in the rfc2047 coding
2755 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2757 NM/03 Allow underscore in dnslist lookups
2758 Fixes bug 1026. Patch from Graeme Fowler
2760 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2761 Code patches from Adam Ciarcinski of NetBSD.
2763 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2766 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2767 is logged, to avoid truncation. Patch from John Horne.
2769 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2770 Patch from Jakob Hirsch.
2772 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2773 SQL string expansion failure details.
2774 Patch from Andrey Oktyabrski.
2776 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2777 Patch from Simon Arlott.
2779 PP/09 New lookups functionality failed to compile on old gcc which rejects
2780 extern declarations in function scope.
2781 Patch from Oliver Fleischmann
2783 PP/10 Use sig_atomic_t for flags set from signal handlers.
2784 Check getgroups() return and improve debugging.
2785 Fixed developed for diagnosis in bug 927 (which turned out to be
2788 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2789 Patch from Mark Zealey.
2791 PP/12 Bugzilla 1056: Improved spamd server selection.
2792 Patch from Mark Zealey.
2794 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2795 Based on patch from Heiko Schlittermann.
2797 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2798 Patch from Uwe Doering, sign-off by Michael Haardt.
2800 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2801 variable declaration deep within a block. Bug and patch from
2804 PP/15 lookups-Makefile IRIX compatibility coercion.
2806 PP/16 Make DISABLE_DKIM build knob functional.
2808 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2809 Patch by Simon Arlott
2811 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2812 variable argument macros. Our copy now differs from upstream.
2818 TF/01 Failure to get a lock on a hints database can have serious
2819 consequences so log it to the panic log.
2821 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2822 controlled using the smtp_confirmation log selector.
2824 TF/03 Include the error message when we fail to unlink a spool file.
2826 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2827 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2828 for maintaining out-of-tree patches for some time.
2830 PP/01 Bugzilla 139: Documentation and portability issues.
2831 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2832 Handle per-OS dynamic-module compilation flags.
2834 PP/02 Let /dev/null have normal permissions.
2835 The 4.73 fixes were a little too stringent and complained about the
2836 permissions on /dev/null. Exempt it from some checks.
2837 Reported by Andreas M. Kirchwitz.
2839 PP/03 Report version information for many libraries, including
2840 Exim version information for dynamically loaded libraries. Created
2841 version.h, now support a version extension string for distributors
2842 who patch heavily. Dynamic module ABI change.
2844 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2845 privilege escalation vulnerability whereby the Exim run-time user
2846 can cause root to append content of the attacker's choosing to
2849 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2852 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2853 If dropping privileges for untrusted macros, we disabled normal logging
2854 on the basis that it would fail; for the Exim run-time user, this is not
2855 the case, and it resulted in successful deliveries going unlogged.
2856 Fixed. Reported by Andreas Metzler.
2862 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2863 only prepend for the Resent-* case. Fixes regression introduced in
2864 Exim 4.70 by NM/22 for Bugzilla 607.
2866 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2867 increasing numbers of administrators be bitten by this.
2869 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2871 PP/03 Bugzilla 994: added openssl_options main configuration option.
2873 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2875 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2877 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2879 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2880 without a peer certificate, leading to a segfault because of an
2881 assumption that peers always have certificates. Be a little more
2882 paranoid. Problem reported by Martin Tscholak.
2884 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2885 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2886 NB: ClamAV planning to remove STREAM in "middle of 2010".
2887 CL also introduces -bmalware, various -d+acl logging additions and
2888 more caution in buffer sizes.
2890 PP/09 Implemented reverse_ip expansion operator.
2892 PP/10 Bugzilla 937: provide a "debug" ACL control.
2894 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2896 PP/12 Bugzilla 973: Implement --version.
2898 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2900 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2902 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2904 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2905 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2906 ignore trailing whitespace.
2908 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2910 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2913 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2914 Notification from Dr Andrew Aitchison.
2916 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2917 ExtendedDetectionInfo response format.
2918 Notification from John Horne.
2920 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2923 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2924 XSL and documented dependency on system catalogs, with examples of how
2927 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2930 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2931 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2932 configuration file which is writeable by the Exim user or group.
2934 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2935 of configuration files to cover files specified with the -C option if
2936 they are going to be used with root privileges, not just the default
2939 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2940 option (effectively making it always true).
2942 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2943 files to be used while preserving root privileges.
2945 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2946 that rogue child processes cannot use them.
2948 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2949 run-time user, instead of root.
2951 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2952 Exim run-time user without dropping privileges.
2954 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2955 result string, instead of calling string_vformat() twice with the same
2958 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2959 for other users. Others should always drop root privileges if they use
2960 -C on the command line, even for a whitelisted configure file.
2962 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2964 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2970 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2971 $data_path, and $header_path variables; fixed documentation bugs and
2974 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2975 exipick to access non-standard spools, including the "frozen" queue
2978 NM/01 Bugzilla 965: Support mysql stored procedures.
2979 Patch from Alain Williams
2981 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2983 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2984 Patch from Andreas Metzler
2986 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2987 Patch from Kirill Miazine
2989 NM/05 Bugzilla 671: Added umask to procmail example.
2991 JJ/03 installed exipick 20100323.0, fixing doc bug
2993 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2994 directory. Notification and patch from Dan Rosenberg.
2996 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2998 TK/02 Improve log output when DKIM signing operation fails.
3000 MH/01 Treat the transport option dkim_domain as a colon separated
3001 list, not as a single string, and sign the message with each element,
3002 omitting multiple occurences of the same signer.
3004 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3005 Bugzilla 985, 986. Patch by Simon Arlott
3007 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3008 Patch by Simon Arlott
3010 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3011 MBX locking. Notification from Dan Rosenberg.
3017 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3019 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3021 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3023 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3025 NM/04 Bugzilla 915: Flag broken perl installation during build.
3031 TK/01 Added patch by Johannes Berg that expands the main option
3032 "spamd_address" if it starts with a dollar sign.
3034 TK/02 Write list of recipients to X-Envelope-Sender header when building
3035 the mbox-format spool file for content scanning (suggested by Jakob
3038 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3039 (http://www.dcc-servers.net/) support via dccifd. Activated by
3040 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3042 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3043 by Mark Daniel Reidel <mr@df.eu>.
3045 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3046 When building exim an external PCRE library is now needed -
3047 PCRE is a system library on the majority of modern systems.
3048 See entry on PCRE_LIBS in EDITME file.
3050 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3051 conversation. Added nologin parameter to request.
3052 Patch contributed by Kirill Miazine.
3054 TF/01 Do not log submission mode rewrites if they do not change the address.
3056 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3058 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3059 log files in place. Contributed by Roberto Lima.
3061 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3063 TF/03 Bugzilla 615: When checking the local_parts router precondition
3064 after a local_part_suffix or local_part_prefix option, Exim now
3065 does not use the address's named list lookup cache, since this
3066 contains cached lookups for the whole local part.
3068 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3069 Robert Millan. Documentation is in experimental-spec.txt.
3071 TF/04 Bugzilla 668: Fix parallel build (make -j).
3073 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3075 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3076 Patch provided by Jan Srzednicki.
3078 TF/05 Leading white space used to be stripped from $spam_report which
3079 wrecked the formatting. Now it is preserved.
3081 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3082 that they are available at delivery time.
3084 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3086 TF/08 TLS error reporting now respects the incoming_interface and
3087 incoming_port log selectors.
3089 TF/09 Produce a more useful error message if an SMTP transport's hosts
3090 setting expands to an empty string.
3092 NM/06 Bugzilla 744: EXPN did not work under TLS.
3093 Patch provided by Phil Pennock.
3095 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3096 Patch provided by Richard Godbee.
3098 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3099 acl_smtp_notquit, added index entry.
3101 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3102 Patch provided by Eugene Bujak.
3104 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3105 accept(). Patch provided by Maxim Dounin.
3107 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3108 Patch provided by Phil Pennock.
3110 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3112 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3113 Patch provided by Brad "anomie" Jorsch.
3115 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3116 Patch provided by Dean Brooks.
3118 TK/05 Add native DKIM support (does not depend on external libraries).
3120 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3121 Patch provided by Graeme Fowler.
3123 NM/16 Bugzilla 851: Documentation example syntax fix.
3125 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3127 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3130 NM/19 Bugzilla 745: TLS version reporting.
3131 Patch provided by Phil Pennock.
3133 NM/20 Bugzilla 167: bool: condition support.
3134 Patch provided by Phil Pennock.
3136 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3137 clients. Patch provided by Phil Pennock.
3139 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3140 Patch provided by Brad "anomie" Jorsch.
3142 NM/23 Bugzilla 687: Fix misparses in eximstats.
3143 Patch provided by Heiko Schlittermann.
3145 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3146 Patch provided by Heiko Schlittermann.
3148 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3149 plus update to original patch.
3151 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3153 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3154 Patch provided by David Brownlee.
3156 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3158 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3160 NM/30 Bugzilla 888: TLS documentation bugfixes.
3162 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3164 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3165 Unlike the original bugzilla I have changed all shell scripts in src tree.
3167 NM/33 Bugzilla 898: Transport filter timeout fix.
3168 Patch by Todd Rinaldo.
3170 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3171 Patch by Serge Demonchaux.
3173 NM/35 Bugzilla 39: Base64 decode bug fixes.
3174 Patch by Jakob Hirsch.
3176 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3178 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3180 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3182 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3188 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3189 ALT-N's libdkim that I have put here:
3190 http://duncanthrax.net/exim-experimental/
3192 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3193 were called 'true' and it seems that C99 defines that as a reserved
3194 keyword to be used with 'bool' variable types. That means you could
3195 not include C99-style headers which use bools without triggering
3196 build errors in sieve.c.
3198 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3199 as mailq or other aliases. Changed the --help handling significantly
3200 to do whats expected. exim_usage() emits usage/help information.
3202 SC/01 Added the -bylocaldomain option to eximstats.
3204 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3206 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3208 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3214 PH/01 Another patch from the Sieve maintainer.
3216 PH/02 When an IPv6 address is converted to a string for single-key lookup
3217 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3218 dots are used instead of colons so that keys in lsearch files need not
3219 contain colons. This was done some time before quoting was made available
3220 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3221 (notated using the quote facility) so as to distinguish them from IPv4
3222 keys. This meant that lookups for IP addresses in host lists did not work
3223 for iplsearch lookups.
3225 This has been fixed by arranging for IPv6 addresses to be expressed with
3226 colons if the lookup type is iplsearch. This is not incompatible, because
3227 previously such lookups could never work.
3229 The situation is now rather anomalous, since one *can* have colons in
3230 ordinary lsearch keys. However, making the change in all cases is
3231 incompatible and would probably break a number of configurations.
3233 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3236 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3237 conversion specification without a maximum field width, thereby enabling
3238 a rogue spamd server to cause a buffer overflow. While nobody in their
3239 right mind would setup Exim to query an untrusted spamd server, an
3240 attacker that gains access to a server running spamd could potentially
3241 exploit this vulnerability to run arbitrary code as the Exim user.
3243 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3244 $primary_hostname instead of what libspf2 thinks the hosts name is.
3246 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3247 a directory entry by the name of the lookup key. Previously, if a
3248 symlink pointed to a non-existing file or a file in a directory that
3249 Exim lacked permissions to read, a lookup for a key matching that
3250 symlink would fail. Now it is enough that a matching directory entry
3251 exists, symlink or not. (Bugzilla 503.)
3253 PH/03 The body_linecount and body_zerocount variables are now exported in the
3256 PH/04 Added the $dnslist_matched variable.
3258 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3259 This means they are set thereafter only if the connection becomes
3262 PH/06 Added the client_condition to authenticators so that some can be skipped
3263 by clients under certain conditions.
3265 PH/07 The error message for a badly-placed control=no_multiline_responses left
3266 "_responses" off the end of the name.
3268 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3270 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3271 (without spaces) instead of just copying the configuration text.
3273 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3275 PH/11 Added $max_received_linelength.
3277 PH/12 Added +ignore_defer and +include_defer to host lists.
3279 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3280 way in which PCRE > 7.0 is built.
3282 PH/14 Implemented queue_only_load_latch.
3284 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3285 MAIL command. The effect was to mangle the value on 64-bit systems.
3287 PH/16 Another patch from the Sieve maintainer.
3289 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3291 PH/18 If a system quota error occurred while trying to create the file for
3292 a maildir delivery, the message "Mailbox is full" was not appended to the
3293 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3294 only to a quota excession during the actual writing of the file.
3296 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3297 characters?) which causes problems in log lines. The DN values are now
3298 passed through string_printing() before being added to log lines.
3300 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3301 and InterBase are left for another time.)
3303 PH/21 Added message_body_newlines option.
3305 PH/22 Guard against possible overflow in moan_check_errorcopy().
3307 PH/23 POSIX allows open() to be a macro; guard against that.
3309 PH/24 If the recipient of an error message contained an @ in the local part
3310 (suitably quoted, of course), incorrect values were put in $domain and
3311 $local_part during the evaluation of errors_copy.
3317 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3318 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3321 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3322 issue a MAIL command.
3324 PH/02 In an ACL statement such as
3326 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3328 if a client was not listed at all, or was listed with a value other than
3329 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3330 the condition was not true (as it should be), so access was not denied.
3331 The bug was that the ! inversion was incorrectly passed on to the second
3332 item. This has been fixed.
3334 PH/03 Added additional dnslists conditions == and =& which are different from
3335 = and & when the dns lookup returns more than one IP address.
3337 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3338 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3340 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3341 FSYNC, which compiles an option called disable_fsync that allows for
3342 bypassing fsync(). The documentation is heavily laced with warnings.
3344 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3346 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3347 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3348 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3349 including adding "make clean"; (3) Added -fPIC when compiling the test
3350 dynamically loaded module, to get rid of a warning.
3352 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3353 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3354 = 0s. The bug is otherwise harmless.
3356 PH/07 There was a bug in the dovecot authenticator such that the value of
3357 $auth1 could be overwritten, and so not correctly preserved, after a
3358 successful authentication. This usually meant that the value preserved by
3359 the server_setid option was incorrect.
3361 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3363 PH/09 Installed PCRE release 7.0.
3365 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3366 run for batched SMTP input. It is now run at the start of every message
3367 in the batch. While fixing this I discovered that the process information
3368 (output by running exiwhat) was not always getting set for -bs and -bS
3369 input. This is fixed, and it now also says "batched" for BSMTP.
3371 PH/11 Added control=no_pipelining.
3373 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3374 patch, slightly modified), and move the expansion of helo_data till after
3375 the connection is made in the smtp transport (so it can use these
3378 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3380 PH/14 Added log_selector = +pid.
3382 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3384 PH/16 Add ${if forany and ${if forall.
3386 PH/17 Added dsn_from option to vary the From: line in DSNs.
3388 PH/18 Flush SMTP output before performing a callout, unless control =
3389 no_callout_flush is set.
3391 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3392 was true (the default) a successful delivery failed to delete the retry
3393 item, thus causing premature timeout of the address. The bug is now
3396 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3398 PH/21 Long custom messages for fakedefer and fakereject are now split up
3399 into multiline responses in the same way that messages for "deny" and
3400 other ACL rejections are.
3402 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3403 with slight modification.
3405 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3406 draft, changing the syntax and factoring some duplicate code".
3408 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3409 for deliveries of the second and subsequent messages over the same SMTP
3412 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3413 ${reduce, with only minor "tidies".
3415 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3417 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3418 expansion side effects.
3420 PH/27 When a message times out after an over-quota error from an Exim-imposed
3421 quota, the bounce message says "mailbox is full". This message was not
3422 being given when it was a system quota that was exceeded. It now should
3425 MH/03 Made $recipients available in local_scan(). local_scan() already has
3426 better access to the recipient list through recipients_list[], but
3427 $recipients can be useful in postmaster-provided expansion strings.
3429 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3430 in the case of a MAIL command with additional options following the
3431 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3432 were accidentally chopped off.
3434 PH/29 SMTP synchronization checks are implemented when a command is read -
3435 there is a check that no more input is waiting when there shouldn't be
3436 any. However, for some commands, a delay in an ACL can mean that it is
3437 some time before the response is written. In this time, more input might
3438 arrive, invalidly. So now there are extra checks after an ACL has run for
3439 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3440 pipelining has not been advertised.
3442 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3444 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3445 returned in the SMTP response when smtp_return_error_details was false.
3446 This has been fixed.
3448 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3449 instead of the C I/O that was originally supplied, because problems were
3450 reported on Solaris.
3452 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3453 Exim which did not show up earlier: it was assuming that a call to
3454 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3455 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3456 was a macro that became an assignment, so it seemed to work. This has
3457 changed to a proper function call with a void return, hence the compile
3458 error. Exim's code has been fixed.
3460 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3463 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3465 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3467 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3468 because it turns out to be too big or there is a timeout) there is no
3469 "Completed" line in the log. When some messages of this type were
3470 selected by exigrep, they were listed as "not completed". Others were
3471 picked up by some special patterns. I have improved the selection
3472 criteria to be more general.
3474 PH/38 The host_find_failed option in the manualroute router can now be set
3475 to "ignore", to completely ignore a host whose IP address cannot be
3476 found. If all hosts are ignored, the behaviour is controlled by the new
3477 host_all_ignored option.
3479 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3480 homing or because of multiple MX records with /mx) generated more than
3481 one IP address, and the following item turned out to be the local host,
3482 all the secondary addresses of the first item were incorrectly removed
3483 from the list, along with the local host and any following hosts (which
3484 is what is supposed to happen).
3486 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3487 whoever called Exim into the -H file. In the case of the daemon it was
3488 behaving confusingly. When first started, it used values for whoever
3489 started the daemon, but after a SIGHUP it used the Exim user (because it
3490 calls itself on a restart). I have changed the code so that it now always
3493 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3494 message are rejected with the same error (e.g. no authentication or bad
3495 sender address), and a DATA command is nevertheless sent (as can happen
3496 with PIPELINING or a stupid MUA), the error message that was given to the
3497 RCPT commands is included in the rejection of the DATA command. This is
3498 intended to be helpful for MUAs that show only the final error to their
3501 PH/42 Another patch from the Sieve maintainer.
3503 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3504 Eximstats - Fixed some broken HTML links and added missing column headers
3506 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3507 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3509 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3511 PH/43 Yet another patch from the Sieve maintainer.
3513 PH/44 I found a way to check for a TCP/IP connection going away before sending
3514 the response to the final '.' that terminates a message, but only in the
3515 case where the client has not sent further data following the '.'
3516 (unfortunately, this is allowed). However, in many cases there won't be
3517 any further data because there won't be any more messages to send. A call
3518 to select() can be used: if it shows that the input is "ready", there is
3519 either input waiting, or the socket has been closed. An attempt to read
3520 the next input character can distinguish the two cases. Previously, Exim
3521 would have sent an OK response which the client would never have see.
3522 This could lead to message repetition. This fix should cure that, at
3523 least in a lot of common cases.
3525 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3526 advertised in response to EHLO.
3532 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3533 fixed by 4.65/MH/01 (is this a record?) are fixed:
3535 (i) An empty string was always treated as zero by the numeric comparison
3536 operators. This behaviour has been restored.
3538 (ii) It is documented that the numeric comparison operators always treat
3539 their arguments as decimal numbers. This was broken in that numbers
3540 starting with 0 were being interpreted as octal.
3542 While fixing these problems I realized that there was another issue that
3543 hadn't been noticed. Values of message_size_limit (both the global option
3544 and the transport option) were treated as octal if they started with 0.
3545 The documentation was vague. These values are now always treated as
3546 decimal, and I will make that clear in the documentation.
3552 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3553 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3556 MH/01 Don't check that the operands of numeric comparison operators are
3557 integers when their expansion is in "skipping" mode (fixes bug
3558 introduced by 4.64-PH/07).
3560 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3561 child addresses, Exim now panics and dies. Previously, because the count
3562 is held in a short int, deliveries were likely to be lost. As such a
3563 large number of recipients for a single message is ridiculous
3564 (performance will be very, very poor), I have chosen to impose a limit
3565 rather than extend the field.
3571 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3572 leftover -K file (the existence of which was triggered by #402).
3573 While we were at it, introduced process PID as part of the -K
3574 filename. This should rule out race conditions when creating
3577 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3578 processing considerably. Previous code took too long for large mails,
3579 triggering a timeout which in turn triggers #401.
3581 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3582 in the DK code in transports.c. sendfile() is not really portable,
3583 hence the _LINUX specificness.
3585 TF/01 In the add_headers option to the mail command in an Exim filter,
3586 there was a bug that Exim would claim a syntax error in any
3587 header after the first one which had an odd number of characters
3590 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3591 callout verification, Exim cached a "reject" for the entire domain. This
3592 is correct for most verifications, but it is not correct for a recipient
3593 verification with use_sender or use_postmaster set, because in that case
3594 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3595 case of MAIL FROM:<> rejection from other early rejections (e.g.
3596 rejection of HELO). When verifying a recipient using a non-null MAIL
3597 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3598 Whatever the result of the callout, the value of the domain cache is
3599 left unchanged (for any other kind of callout, getting as far as trying
3600 RCPT means that the domain itself is ok).
3602 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3605 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3606 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3607 session, the daemon ignores ECONNECT errors and logs others; it now
3608 ignores EPIPE as well.
3610 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3611 (quoted-printable decoding).
3613 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3614 later the small subsequent patch to fix an introduced bug.
3616 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3618 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3620 PH/08 An error is now given if message_size_limit is specified negative.
3622 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3623 to be given (somewhat) arbitrary names.
3625 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3628 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3629 miscellaneous code fixes
3631 PH/10 Added the log_reject_target ACL modifier to specify where to log
3634 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3635 hostname. This is wrong, because it relates to the incoming message (and
3636 probably the interface on which it is arriving) and not to the outgoing
3637 callout (which could be using a different interface). This has been
3638 changed to use the value of the helo_data option from the smtp transport
3639 instead - this is what is used when a message is actually being sent. If
3640 there is no remote transport (possible with a router that sets up host
3641 addresses), $smtp_active_hostname is used.
3643 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3644 tweaks were necessary in order to get it to work (see also 21 below):
3645 (a) The code assumed that strncpy() returns a negative number on buffer
3646 overflow, which isn't the case. Replaced with Exim's string_format()
3648 (b) There were several signed/unsigned issues. I just did the minimum
3649 hacking in of casts. There is scope for a larger refactoring.
3650 (c) The code used strcasecmp() which is not a standard C function.
3651 Replaced with Exim's strcmpic() function.
3652 (d) The code set only $1; it now sets $auth1 as well.
3653 (e) A simple test gave the error "authentication client didn't specify
3654 service in request". It would seem that Dovecot has changed its
3655 interface. Fortunately there's a specification; I followed it and
3656 changed what the client sends and it appears to be working now.
3658 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3661 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3662 address A is aliased to B and C, where B exists and C does not. Without
3663 -v the output is "A verified" because verification stops after a
3664 successful redirection if more than one address is generated. However,
3665 with -v the child addresses are also verified. Exim was outputting "A
3666 failed to verify" and then showing the successful verification for C,
3667 with its parentage. It now outputs "B failed to verify", showing B's
3668 parentage before showing the successful verification of C.
3670 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3671 look up a TXT record in a specific list after matching in a combined
3674 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3675 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3676 they consult the DNS. I had assumed they would set it the way they
3677 wanted; and indeed my experiments on Linux seem to show that in some
3678 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3679 To be on the safe side, however, I have now made the interface to
3680 host_find_byname() similar to host_find_bydns(), with an argument
3681 containing the DNS resolver options. The host_find_byname() function now
3682 sets these options at its start, just as host_find_bydns() does. The smtp
3683 transport options dns_qualify_single and dns_search_parents are passed to
3684 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3685 of host_find_byname() use the default settings of RES_DEFNAMES
3686 (qualify_single) but not RES_DNSRCH (search_parents).
3688 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3689 spool_read_header() do less string testing, by means of a preliminary
3690 switch on the second character of optional "-foo" lines. (This is
3691 overdue, caused by the large number of possibilities that now exist.
3692 Originally there were few.) While I was there, I also converted the
3693 str(n)cmp tests so they don't re-test the leading "-" and the first
3694 character, in the hope this might squeeze out yet more improvement.
3696 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3697 flag allowing group syntax was set by the header_syntax check but not
3698 turned off, possible causing trouble later; (2) The flag was not being
3699 set at all for the header_verify test, causing "group"-style headers to
3700 be rejected. I have now set it in this case, and also caused header_
3701 verify to ignore an empty address taken from a group. While doing this, I
3702 came across some other cases where the code for allowing group syntax
3703 while scanning a header line wasn't quite right (mostly, not resetting
3704 the flag correctly in the right place). These bugs could have caused
3705 trouble for malformed header lines. I hope it is now all correct.
3707 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3708 with the "reply" argument non-NULL. The code, however (which originally
3709 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3710 but it didn't always do it. This confused somebody who was copying the
3711 code for some other use. I have removed all the tests.
3713 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3714 feature that was used to support insecure browsers during the U.S. crypto
3715 embargo. It requires special client support, and Exim is probably the
3716 only MTA that supported it -- and would never use it because real RSA is
3717 always available. This code has been removed, because it had the bad
3718 effect of slowing Exim down by computing (never used) parameters for the
3719 RSA_EXPORT functionality.
3721 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3722 authenticator to fail if there's a tab character in the incoming data
3723 (there should never be unless someone is messing about, as it's supposed
3724 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3725 if the connection is using TLS or if the remote IP is the same as the
3726 local IP, and the "valid-client-cert option" if a client certificate has
3729 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3730 authenticators. This can be used for authorization after authentication
3731 succeeds. (In the case of plaintext, it servers for both authentication
3734 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3735 if any retry times were supplied.
3737 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3738 connection, where there is no client IP address to check. In this
3739 situation, the verify now always succeeds.
3741 PH/25 Applied John Jetmore's -Mset patch.
3743 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3745 PH/27 In a string expansion for a processed (not raw) header when multiple
3746 headers of the same name were present, leading whitespace was being
3747 removed from all of them, but trailing whitespace was being removed only
3748 from the last one. Now trailing whitespace is removed from each header
3749 before concatenation. Completely empty headers in a concatenation (as
3750 before) are ignored.
3752 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3753 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3755 PH/29 [Removed. This was a change that I later backed out, and forgot to
3756 correct the ChangeLog entry (that I had efficiently created) before
3757 committing the later change.]
3759 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3760 address errors (4xx response to RCPT) over the same connection as other
3761 messages routed to the same hosts. Such deliveries are always "forced",
3762 so retry times are not inspected. This resulted in far too many retries
3763 for the affected addresses. The effect occurred only when there were more
3764 hosts than the hosts_max_try setting in the smtp transport when it had
3765 the 4xx errors. Those hosts that it had tried were not added to the list
3766 of hosts for which the message was waiting, so if all were tried, there
3767 was no problem. Two fixes have been applied:
3769 (i) If there are any address or message errors in an SMTP delivery, none
3770 of the hosts (tried or untried) are now added to the list of hosts
3771 for which the message is waiting, so the message should not be a
3772 candidate for sending over the same connection that was used for a
3773 successful delivery of some other message. This seems entirely
3774 reasonable: after all the message is NOT "waiting for some host".
3775 This is so "obvious" that I'm not sure why it wasn't done
3776 previously. Hope I haven't missed anything, but it can't do any
3777 harm, as the worst effect is to miss an optimization.
3779 (ii) If, despite (i), such a delivery is accidentally attempted, the
3780 routing retry time is respected, so at least it doesn't keep
3781 hammering the server.
3783 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3784 in ${readsocket because some servers need this prod.
3786 PH/32 Added some extra debug output when updating a wait-xxx database.
3788 PH/33 The hint "could be header name not terminated by colon", which has been
3789 given for certain expansion errors for a long time, was not being given
3790 for the ${if def:h_colon_omitted{... case.
3792 PH/34 The spec says: "With one important exception, whenever a domain list is
3793 being scanned, $domain contains the subject domain." There was at least
3794 one case where this was not true.
3796 PH/35 The error "getsockname() failed: connection reset by peer" was being
3797 written to the panic log as well as the main log, but it isn't really
3798 panic-worthy as it just means the connection died rather early on. I have
3799 removed the panic log writing for the ECONNRESET error when getsockname()
3802 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3803 runs only) independently of the message's sender address. This meant
3804 that, if the 4xx error was in fact related to the sender, a different
3805 message to the same recipient with a different sender could confuse
3806 things. In particular, this can happen when sending to a greylisting
3807 server, but other circumstances could also provoke similar problems.
3808 I have changed the default so that the retry time for these errors is now
3809 based a combination of the sender and recipient addresses. This change
3810 can be overridden by setting address_retry_include_sender=false in the
3813 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3814 remote server are returned as part of bounce messages. This was not
3815 happening for LMTP over a pipe (the lmtp transport), but now it is the
3816 same for both kinds of LMTP.
3818 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3819 in header lines that were in fact CNAMEs. This is no longer the case.
3821 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3822 and queue runs started by the daemon processed all messages. This has
3823 been fixed so that -R and -S can now usefully be given with -q<time>.
3825 PH/40 Import PCRE release 6.7 (fixes some bugs).
3827 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3829 PH/42 Give an error if -q is specified more than once.
3831 PH/43 Renamed the variables $interface_address and $interface_port as
3832 $received_ip_address and $received_port, to make it clear that these
3833 values apply to message reception, and not to the outgoing interface when
3834 a message is delivered. (The old names remain recognized, of course.)
3836 PH/44 There was no timeout on the connect() call when using a Unix domain
3837 socket in the ${readsocket expansion. There now is.
3839 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3840 be meaningful with "accept".
3842 SC/01 Eximstats V1.43
3843 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3845 SC/02 Eximstats V1.44
3846 Use a glob alias rather than an array ref in the generated
3847 parser. This improves both readability and performance.
3849 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3850 Collect SpamAssassin and rejection statistics.
3851 Don't display local sender or destination tables unless
3852 there is data to show.
3853 Added average volumes into the top table text output.
3855 SC/04 Eximstats V1.46
3856 Collect data on the number of addresses (recipients)
3857 as well as the number of messages.
3859 SC/05 Eximstats V1.47
3860 Added 'Message too big' to the list of mail rejection
3861 reasons (thanks to Marco Gaiarin).
3863 SC/06 Eximstats V1.48
3864 Mainlog lines which have GMT offsets and are too short to
3865 have a flag are now skipped.
3867 SC/07 Eximstats V1.49 (Alain Williams)
3868 Added the -emptyok flag.
3870 SC/08 Eximstats V1.50
3871 Fixes for obtaining the IP address from reject messages.
3873 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3874 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3875 whitespace changes from 4.64-PH/27
3877 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3880 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3881 are found, allow negative numbers in numeric criteria)
3883 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3885 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3886 to match changes made in 4.64-PH/43
3888 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3890 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3891 "domain" is actually an IP address, and force a failure. This locks out
3892 those revolvers/nameservers that support "A-for-A" lookups, in
3893 contravention of the specifications.
3895 PH/48 When a host name was looked up from an IP address, and the subsequent
3896 forward lookup of the name timed out, the host name was left in
3897 $sender_host_name, contrary to the specification.
3899 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3900 restricted to single-key lookups, Exim was not diagnosing an error if
3901 * or *@ was used with a query-style lookup.
3903 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3905 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3906 long ago, but noone interested enough thought of it. Let's just say that
3907 the "1.1" means that there are some new functions that weren't there at
3908 some point in the past.
3910 PH/51 Error processing for expansion failure of helo_data from an smtp
3911 transport during callout processing was broken.
3913 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3914 tested/used via the -bh/-bhc/-bs options.
3916 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3917 bug, fixed in subsequent PCRE releases).
3919 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3920 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3922 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3928 SC/01 Use a glob alias rather than an array ref in eximstats generated
3929 parser. This improves both readability and performance.
3931 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3932 Don't display local sender or destination tables in eximstats unless
3933 there is data to show.
3934 Added average volumes into the eximstats top table text output.
3936 SC/03 Collect data on the number of addresses (recipients) as well
3937 as the number of messages in eximstats.
3939 TF/01 Correct an error in the documentation for the redirect router. Exim
3940 does (usually) call initgroups() when daemonizing.
3942 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3943 with consistent privilege compared to when running as a daemon.
3945 TF/03 Note in the spec that $authenticated_id is not set for local
3946 submissions from trusted users.
3948 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3949 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3951 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3952 by adding some example configuration directives to the default
3953 configuration file. A little bit of work is required to uncomment the
3954 directives and define how usernames and passwords are checked, but
3955 there is now a framework to start from.
3957 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3958 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3959 without this. I don't know how relevant this is to other LDAP libraries.
3961 PH/02 Add the verb name to the "unknown ACL verb" error.
3963 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3965 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3967 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3968 directory not expanded when it should be if an expanded home directory
3969 was set for the address (which is overridden by the transport).
3971 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3974 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3975 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3976 because it is too late at that time, and has no effect.
3978 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3979 security issue with \' (bugzilla #107). I could not use the
3980 PQescapeStringConn() function, because it needs a PGconn value as one of
3983 PH/08 When testing addresses using -bt, indicate those final addresses that
3984 are duplicates that would not cause an additional delivery. At least one
3985 person was confused, thinking that -bt output corresponded to deliveries.
3986 (Suppressing duplicates isn't a good idea as you lose the information
3987 about possibly different redirections that led to the duplicates.)
3989 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3990 systems where poll() doesn't work, in particular OS X.
3992 PH/10 Added more information to debugging output for retry time not reached.
3994 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3995 operations in malware.c.
3997 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4000 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4001 syslog_duplication set false, log lines that would normally be written
4002 both the the main log and to the reject log were not written to syslog at
4005 PH/14 In the default configuration, change the use of "message" in ACL warn
4006 statements to "add_header".
4008 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4009 not followed by a command (e.g. "seen endif").
4011 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4012 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4015 PH/17 Added extra conditions to the default value of delay_warning_condition
4019 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4020 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4021 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4024 The Auto-Submitted: and various List- headers are standardised, whereas I
4025 don't think Precedence: ever was.
4027 PH/18 Refactored debugging code in route_finduser() to show more information,
4028 in particular, the error code if getpwnam() issues one.
4030 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4031 This is apparently needed in addition to the PH/07 change above to avoid
4032 any possible encoding problems.
4034 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4035 but not after initializing Perl.
4037 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4038 output them only if debugging. By default they are written stderr,
4039 apparently, which is not desirable.
4041 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4044 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4047 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4049 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4050 authenticated or an ident call has been made. Suppress the default
4051 values for $authenticated_id and $authenticated_sender (but permit -oMai
4052 and -oMas) when testing with -bh.
4054 PH/24 Re-jigged the order of the tests in the default configuration so that the
4055 tests for valid domains and recipients precede the DNS black list and CSA
4056 tests, on the grounds that those ones are more expensive.
4058 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4059 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4060 This bug exists in every version of Exim that I still have, right back to
4063 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4064 However, an attempt to turn on case-sensitivity in a regex key by
4065 including (?-i) didn't work because the subject string was already
4066 lowercased, and the effects were non-intuitive. It turns out that a
4067 one-line patch can be used to allow (?-i) to work as expected.
4073 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4074 other effects) broke the use of negated acl sub-conditions.
4076 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4079 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4080 "Deny" causes Exim to reject the incoming connection with a 554 error.
4081 Unfortunately, if there is a major crisis, such as a disk failure,
4082 tcp-wrappers gives "deny", whereas what one would like would be some
4083 kind of temporary error. A kludge has been added to help with this.
4084 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4085 554 error is used if errno is still zero or contains ENOENT (which occurs
4086 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4089 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4091 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4092 errors. Otherwise a message that provokes a temporary error (when other
4093 messages do not) can cause a whole host to time out.
4095 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4096 addresses were routed directly to files or pipes from a redirect router.
4097 File deliveries just didn't batch; pipe deliveries might have suffered
4100 PH/06 A failure to get a lock for a hints database would erroneously always say
4101 "Failed to get write lock", even when it was really a read lock.
4103 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4104 of 0600. This has been changed to use the value of the lockfile_mode
4105 option (which defaults to 0600).
4107 PH/08 Applied small patch from the Sieve maintainer.
4109 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4110 folder from quota calculations, a direct delivery into this folder messed
4111 up the contents of the maildirsize file. This was because the regex was
4112 used only to exclude .Trash (or whatever) when the size of the mailbox
4113 was calculated. There was no check that a delivery was happening into an
4114 excluded directory. This bug has been fixed by ignoring all quota
4115 processing for deliveries into excluded directories.
4117 PH/10 Added the maildirfolder_create_regex option to appendfile.
4123 PH/01 The code for finding all the local interface addresses on a FreeBSD
4124 system running IPv6 was broken. This may well have applied to all BSD
4125 systems, as well as to others that have similar system calls. The broken
4126 code found IPv4 interfaces correctly, but gave incorrect values for the
4127 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4128 that it would not match correctly against @[] and not recognize the IPv6
4131 PH/02 The ipliteral router was not recognizing addresses of the form user@
4132 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4134 PH/03 Added disable_ipv6.
4136 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4137 decoded form, because it is most often used to construct To: headers
4138 lines in autoreplies, and the decoded form may well be syntactically
4139 invalid. However, $reply_address has leading white space removed, and all
4140 newlines turned into spaces so that the autoreply transport does not
4143 PH/05 If group was specified without a user on a router, and no group or user
4144 was specified on a transport, the group from the router was ignored.
4146 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4147 for visible compile-time settings that can be used to change these
4148 numbers, for those that want even more. Backwards compatibility with old
4149 spool files has been maintained. However, going back to a previous Exim
4150 release will lost any variables that are in spool files.
4152 PH/07 Two small changes when running in the test harness: increase delay when
4153 passing a TCP/IP connection to a new process, in case the original
4154 process has to generate a bounce, and remove special handling of
4155 127.0.0.2 (sic), which is no longer necessary.
4157 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4158 be the same on different OS.
4160 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4163 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4164 whether --show-vars was specified or not
4166 JJ/02 exipick: Added support for new ACL variable spool format introduced
4169 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4170 syntactically invalid From: or Reply-to: line, and a filter used this to
4171 generate an autoreply, and therefore failed to obtain an address for the
4172 autoreply, Exim could try to deliver to a non-existent relative file
4173 name, causing unrelated and misleading errors. What now happens is that
4174 it logs this as a hard delivery error, but does not attempt to create a
4177 PH/11 The exinext utility has a -C option for testing purposes, but although
4178 the given file was scanned by exinext itself; it wasn't being passed on
4179 when Exim was called.
4181 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4182 an end-of-file indication when reading a command response.
4184 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4185 compiled. In many other places in Exim, IPv6 addresses are always
4186 recognized, so I have changed this. It also means that IPv4 domain
4187 literals of the form [IPV4:n.n.n.n] are now always recognized.
4189 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4190 used if the router is not running as root, for example, when verifying at
4191 ACL time, or when using -bh. The debugging output from this situation was
4192 non-existent - all you got was a failure to exec. I have made two
4195 (a) Failures to set uid/gid, the current directory, or a process leader
4196 in a subprocess such as that created by queryprogram now generate
4197 suitable debugging output when -d is set.
4199 (b) The queryprogram router detects when it is not running as root,
4200 outputs suitable debugging information if -d is set, and then runs
4201 the subprocess without attempting to change uid/gid.
4203 PH/15 Minor change to Makefile for building test_host (undocumented testing
4206 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4207 additional section of a DNS packet that returns MX or SRV records.
4208 Instead, it always explicitly searches for A/AAAA records. This avoids
4209 major problems that occur when a DNS server includes only records of one
4210 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4211 fixed another bug: if SRV records were looked up and the corresponding
4212 address records were *not* found in the additional section, the port
4213 values from the SRV records were lost.
4215 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4216 using the correct key (the original address) when searching the retry
4217 rules in order to find which one to use for generating the retry hint.
4219 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4220 adding the default one. Similarly, if it contains a Reply-To: header, the
4221 errors_reply_to option, if set, is not used.
4223 PH/19 When calculating a retry time, Exim used to measure the "time since
4224 failure" by looking at the "first failed" field in the retry record. Now
4225 it does not use this if it is later than than the arrival time of the
4226 message. Instead it uses the arrival time. This makes for better
4227 behaviour in cases where some deliveries succeed, thus re-setting the
4228 "first failed" field. An example is a quota failure for a huge message
4229 when small messages continue to be delivered. Without this change, the
4230 "time since failure" will always be short, possible causing more frequent
4231 delivery attempts for the huge message than are intended.
4232 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4234 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4235 $1, $2, $3) because the numerical variables can be reset during some
4236 expansion items (e.g. "match"), thereby losing the authentication data.
4238 PH/21 Make -bV show the size of off_t variables so that the test suite can
4239 decide whether to run tests for quotas > 2G.
4241 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4242 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4243 filecount value is greater than 2G or if a quota value is greater than 2G
4244 on a system where the size of off_t is not greater than 4, a panic error
4247 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4248 never match. The debug and -bh output now contains an explicit error
4249 message indicating a malformed IPv4 address or mask.
4251 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4252 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4253 PH/23 above applies.
4255 PH/25 Do not write to syslog when running in the test harness. The only
4256 occasion when this arises is a failure to open the main or panic logs
4257 (for which there is an explicit test).
4259 PH/26 Added the /no_tell option to "control=freeze".
4261 PH/27 If a host name lookup failed very early in a connection, for example, if
4262 the IP address matched host_lookup and the reverse lookup yielded a name
4263 that did not have a forward lookup, an error message of the form "no IP
4264 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4265 could be logged. Now it outputs the IP address instead of "NULL".
4267 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4268 allows the sender and the authenticated sender to be set when
4269 submitting a message from within Exim. Since child_open_exim() is
4270 documented for local_scan(), the new function should be too.
4272 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4273 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4274 results in an empty string is now treated as unset.
4276 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4278 PH/31 Added sender_verify_fail logging option.
4280 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4281 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4282 tidied the source and removed it altogether.
4284 PH/33 When a queue run was abandoned because the load average was too high, a
4285 log line was always written; now it is written only if the queue_run log
4286 selector is set. In addition, the log line for abandonment now contains
4287 information about the queue run such as the pid. This is always present
4288 in "start" and "stop" lines but was omitted from the "abandon" line.
4290 PH/34 Omit spaces between a header name and the colon in the error message that
4291 is given when verify = headers_syntax fails (if there are lots of them,
4292 the message gets confusing).
4294 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4295 names, as there are now some PTR records that contain slashes. This check
4296 is only to protect against broken name servers that fall over on strange
4297 characters, so the fact that it applies to all lookups doesn't matter.
4299 PH/36 Now that the new test suite is complete, we can remove some of the
4300 special code in Exim that was needed for the old test suite. For example,
4301 sorting DNS records because real resolvers return them in an arbitrary
4302 order. The new test suite's fake resolver always returns records in the
4305 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4306 bounces) except when queue_only is set, to avoid logging races between
4307 the different processes.
4309 PH/38 Panic-die if .include specifies a non-absolute path.
4311 PH/39 A tweak to the "H" retry rule from its user.
4313 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4314 a label. They prevented compilation on older perls.
4316 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4317 a warning to be raised on newish perls.
4319 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4320 on queue. Changes to match documented behaviour of showing count of
4321 messages matching specified criteria.
4323 PH/40 Changed the default ident timeout from 30s to 5s.
4325 PH/41 Added support for the use of login_cap features, on those BSD systems
4326 that have them, for controlling the resources used by pipe deliveries.
4328 PH/42 The content-scanning code uses fopen() to create files in which to put
4329 message data. Previously it was not paying any attention to the mode of
4330 the files. Exim runs with umask(0) because the rest of the code creates
4331 files with open(), and sets the required mode explicitly. Thus, these
4332 files were ending up world-writeable. This was not a big issue, because,
4333 being within the spool directory, they were not world-accessible. I have
4334 created a function called modefopen, which takes an additional mode
4335 argument. It sets umask(777), creates the file, chmods it to the required
4336 mode, then resets the umask. All the relevant calls to fopen() in the
4337 content scanning code have been changed to use this function.
4339 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4340 to 24 hours. This avoids potential overflow problems when processing G
4341 and H retry rules. I suspect nobody ever tinkers with this value.
4343 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4345 PH/45 When the plaintext authenticator is running as a client, the server's
4346 challenges are checked to ensure they are valid base64 strings. By
4347 default, the authentication attempt is cancelled if an invalid string is
4348 received. Setting client_ignore_invalid_base64 true ignores these errors.
4349 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4350 they are received. Thus, the responses can be made to depend on the
4351 challenges. If an invalid string is ignored, an empty string is placed in
4354 PH/46 Messages that are created by the autoreply transport now contains a
4355 References: header, in accordance with RFCs 2822 and 3834.
4357 PH/47 Added authenticated_sender_force to the smtp transport.
4359 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4361 PH/49 Installed latest patch from the Sieve maintainer.
4363 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4364 was also set, the appendfile transport was unnecessarily scanning a
4365 directory of message files (e.g. for maildir delivery) to find the count
4366 of files (along with the size), even though it did not need this
4367 information. It now does the scan only if it needs to find either the
4368 size of the count of files.
4370 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4372 PH/52 Two bugs concerned with error handling when the smtp transport is
4375 (i) Exim was not creating retry information for temporary errors given
4376 for individual recipients after the DATA command when the smtp transport
4377 was used in LMTP mode. This meant that they could be retried too
4378 frequently, and not timed out correctly.
4380 (ii) Exim was setting the flag that allows error details to be returned
4381 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4382 recipients that were returned after the DATA command.
4384 PH/53 This is related to PH/52, but is more general: for any failing address,
4385 when detailed error information was permitted to be returned to the
4386 sender, but the error was temporary, then after the final timeout, only
4387 "retry timeout exceeded" was returned. Now it returns the full error as
4388 well as "retry timeout exceeded".
4390 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4391 do this, and (what is worse) MTAs that accept it.
4393 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4394 will now be deprecated.
4396 PH/56 New os.c-cygwin from the Cygwin maintainer.
4398 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4399 formats (previously only available in exim formats via -bpr, -bpru,
4400 and -bpra. Now also available in native and exiqgrep formats)
4402 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4403 with very large, slow to parse queues
4405 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4407 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4409 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4410 responses to authentication challenges, though it was showing the
4411 challenges; (ii) I've removed the CR characters from the debug output for
4414 PH/58 Allow for the insertion of a newline as well as a space when a string
4415 is turned into more than one encoded-word during RFC 2047 encoding. The
4416 Sieve code now uses this.
4418 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4419 data_4xx, lost_connection, tls_required.
4421 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4422 message was being sent as an SMTP response.
4424 PH/61 Add -l and -k options to exicyclog.
4426 PH/62 When verifying, if an address was redirected to one new address, so that
4427 verification continued, and the new address failed or deferred after
4428 having set something in $address_data, the value of $address_data was not
4429 passed back to the ACL. This was different to the case when no
4430 redirection occurred. The value is now passed back in both cases.
4432 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4433 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4434 use login_cap.h, so on its own it isn't the distinguishing feature. The
4435 new name refers directly to the setclassresources() function.
4437 PH/65 Added configuration files for NetBSD3.
4439 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4441 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4442 is preferred over IPv4.
4444 PH/68 The bounce_return_message and bounce_return_body options were not being
4445 honoured for bounces generated during the reception of non-SMTP messages.
4446 In particular, this applied to messages rejected by the ACL. This bug has
4447 been fixed. However, if bounce_return_message is true and bounce_return_
4448 body is false, the headers that are returned for a non-SMTP message
4449 include only those that have been read before the error was detected.
4450 (In the case of an ACL rejection, they have all been read.)
4452 PH/69 The HTML version of the specification is now built in a directory called
4453 spec_html instead of spec.html, because the latter looks like a path with
4454 a MIME-type, and this confuses some software.
4456 PH/70 Catch two compiler warnings in sieve.c.
4458 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4459 function verify_get_ident() calls ip_connect() to connect a socket, but
4460 if the "connect()" function timed out, ip_connect() used to close the
4461 socket. However, verify_get_ident() also closes the socket later, and in
4462 between Exim writes to the log, which may get opened at this point. When
4463 the socket was closed in ip_connect(), the log could get the same file
4464 descriptor number as the socket. This naturally causes chaos. The fix is
4465 not to close the socket in ip_connect(); the socket should be closed by
4466 the function that creates it. There was only one place in the code where
4467 this was missing, in the iplookup router, which I don't think anybody now
4468 uses, but I've fixed it anyway.
4470 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4471 well as to direct DNS lookups. Otherwise the handling of names in host
4472 lists is inconsistent and therefore confusing.
4478 PH/01 Two changes to the default runtime configuration:
4480 (1) Move the checks for relay_from_hosts and authenticated clients from
4481 after to before the (commented out) DNS black list checks.
4483 (2) Add control=submission to the relay_from_hosts and authenticated
4484 clients checks, on the grounds that messages accepted by these
4485 statements are most likely to be submissions.
4487 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4489 (1) Generate an error if the third argument for the ${prvs expansion is
4492 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4495 (3) Reset the variables that are obtained from the first argument of
4496 ${prvscheck and used in the second argument before leaving the code,
4497 because their memory is reclaimed, so using them afterwards may do
4500 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4501 one (it's much easier than Tom thought :-).
4503 (5) Because of (4), we can now allow for the use of $prvscheck_result
4504 inside the third argument.
4506 PH/03 For some reason, the default setting of PATH when running a command from
4507 a pipe transport was just "/usr/bin". I have changed it to
4510 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4511 anything to be listed in the output from -bV.
4513 PH/05 When a filter generated an autoreply, the entire To: header line was
4514 quoted in the delivery log line, like this:
4516 => >A.N.Other <ano@some.domain> <original@ddress> ...
4518 This has been changed so that it extracts the operative address. There
4519 may be more than one such address. If so, they are comma-separated, like
4522 => >ano@some.domain,ona@other.domain <original@ddress> ...
4524 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4525 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4526 not being looked up in the rDNS to get a host name, Exim was showing the
4527 IP address twice in Received: lines, even though the IP addresses were
4528 identical. For example:
4530 Received: from [1.2.3.4] (helo=[1.2.3.4])
4532 However, if the real host name was known, it was omitting the HELO data
4533 if it matched the actual IP address. This has been tidied up so that it
4534 doesn't show the same IP address twice.
4536 PH/07 When both +timestamp and +memory debugging was on, the value given by
4537 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4538 called by the debug printing, thereby overwriting the timestamp buffer.
4539 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4541 PH/08 When the original message was included in an autoreply transport, it
4542 always said "this is a copy of the message, including all the headers",
4543 even if body_only or headers_only was set. It now gives an appropriate
4546 PH/09 Applied a patch from the Sieve maintainer which:
4548 o fixes some comments
4549 o adds the (disabled) notify extension core
4550 o adds some debug output for the result of if/elsif tests
4551 o points to the current vacation draft in the documentation
4552 and documents the missing references header update
4556 o fixes a bug in processing the envelope test (when testing
4557 multiple envelope elements, the last element determined the
4560 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4561 Electronic Mail") by including:
4563 Auto-submitted: auto-generated
4565 in the messages that it generates (bounce messages and others, such as
4566 warnings). In the case of bounce messages for non-SMTP messages, there was
4567 also a typo: it was using "Auto_submitted" (underscore instead of
4568 hyphen). Since every message generated by Exim is necessarily in response
4569 to another message, thes have all been changed to:
4571 Auto-Submitted: auto-replied
4573 in accordance with these statements in the RFC:
4575 The auto-replied keyword:
4577 - SHOULD be used on messages sent in direct response to another
4578 message by an automatic process,
4580 - MUST NOT be used on manually-generated messages,
4582 - MAY be used on Delivery Status Notifications (DSNs) and Message
4583 Disposition Notifications (MDNs),
4585 - MUST NOT be used on messages generated by automatic or periodic
4586 processes, except for messages which are automatic responses to
4589 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4590 to the default Received: header definition.
4592 PH/12 Added log selector acl_warn_skipped (default on).
4594 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4595 variables because (a) they are in the wrong storage pool and (b) even if
4596 they were copied, it wouldn't work properly because of the caching.
4598 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4599 checking when decoding. Apparently there are clients that generate
4600 overlong encoded strings. Why am I not surprised?
4602 PH/15 If the first argument of "${if match_address" was not empty, but did not
4603 contain an "@" character, Exim crashed. Now it writes a panic log message
4604 and treats the condition as false.
4606 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4608 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4609 extension "envelope-auth". The code is finished and in agreement with
4610 other implementations, but there is no documentation so far and in fact,
4611 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4612 not changing the active code.
4614 Print executed "if" and "elsif" statements when debugging is used. This
4615 helps a great deal to understand what a filter does.
4617 Document more things not specified clearly in RFC3028. I had all this
4618 sorted out, when out of a sudden new issues came to my mind. Oops."
4620 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4623 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4624 canonical form (as documented). However, after a host name lookup from
4625 the IP address, check_host() was doing a simple string comparison with
4626 addresses acquired from the DNS when checking that the found name did
4627 have the original IP as one of its addresses. Since any found IPv6
4628 addresses are likely to be in abbreviated form, the comparison could
4629 fail. Luckily, there already exists a function for doing the comparison
4630 by converting both addresses to binary, so now that is used instead of
4631 the text comparison.
4633 PH/20 There was another similar case to PH/19, when a complete host name was
4634 given in a host list; looking up its IP address could give an abbreviated
4635 form, whereas the current host's name might or might not be abbreviated.
4636 The same fix has been applied.
4642 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4643 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4646 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4647 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4649 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4651 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4652 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4653 Also, for "Auto-Submitted", treat anything other than "no" as
4654 non-personal, in accordance with RFC 3834. (Previously it treated
4655 anything starting "auto-" as non-personal.)
4657 TF/01 The control=submission/name=... option had a problem with syntax
4658 errors if the name included a slash character. The /name= option
4659 now slurps the rest of the string, so it can include any characters
4660 but it must come last in the list of options (after /sender_retain
4663 PH/05 Some modifications to the interface to the fake nameserver for the new
4671 TK/01 Added the "success_on_redirect" address verification option. See
4672 NewStuff for rationale and an example.
4674 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4676 PH/02 Patch to exigrep to allow it to work on syslog lines.
4678 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4679 fread() to skip over the body file's header line, because in Cygwin the
4680 header line is locked and is inaccessible.
4682 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4683 co-exist for some time) to make it clear that it is the Exim ID that is
4684 referenced, not the Message-ID: header line.
4686 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4687 string_format() function, because snprintf() does not exist on all
4690 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4691 ${stat: expansion item.
4693 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4694 protocol synchronization error", to keep the pedants happy.
4696 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4697 well as for IRIX systems, when gcc is being used. See the host.c source
4700 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4702 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4705 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4706 transport and to the smtp transport in LMTP mode.
4708 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4710 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4711 run a filter in a subprocess. This could lead to confusion in subsequent
4712 lookups in the parent process. There should also be a search_tidyup() at
4713 the end of the subprocess.
4715 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4716 only if the host matched helo_try_verify_hosts, which caused the
4717 verification to occur when the EHLO/HELO command was issued. The ACL just
4718 tested the remembered result. Now, if a previous verification attempt has
4719 not happened, "verify = helo" does it there and then.
4721 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4723 TK/03 Fix log output including CR from clamd.
4725 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4726 exist provoked a memory error which could cause a segfault.
4728 PH/15 Installed PCRE 6.2
4730 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4732 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4733 of the problem. Specifically, suggested +O2 rather than +O1 for the
4736 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4738 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4739 record, the port was not correctly logged when the outgoing_port log
4740 selector was set (it logged the transort's default port).
4742 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4743 fallback_hosts, and "hosts" in the smtp transport.
4745 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4746 host errors such as "Connection refused".
4748 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4749 authentication with radiusclient 0.4.9:
4751 - Error returned from rc_read_config was caught wrongly
4752 - Username/password not passed on to radius server due to wrong length.
4754 The presumption is that some radiusclient API changes for 4.51/PH/17
4755 were not taken care of correctly. The code is still untested by me (my
4756 Linux distribution still has 0.3.2 of radiusclient), but it was
4757 contributed by a Radius user.
4759 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4760 expanding the "port" option of the smtp transport.
4762 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4763 while reading a MIME header. Thanks to Tom Hughes for a patch.
4765 PH/24 Include config.h inside local_scan.h so that configuration settings are
4768 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4769 that in an ACL for RCPT (for example), you can examine exactly what was
4772 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4773 commands, but it was not correctly comparing the address with the actual
4774 client host address. Thus, it would show the EHLO address in Received:
4775 header lines when this was not necessary.
4777 PH/27 Added the % operator to ${eval:}.
4779 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4780 it should be ignoring failures (because with -C, for example, it has lost
4781 privilege). It wasn't ignoring creation failures other than "already
4784 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4787 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4788 because an input line was too long, either on its own, or by virtue of
4789 too many continuations, the temporary file was not being removed, and the
4790 return code was incorrect.
4792 PH/31 Missing "BOOL" in function definition in filtertest.c.
4794 PH/32 Applied Sieve patches from the maintainer.
4796 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4798 PH/33 Added "verify = not_blind".
4800 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4801 Local/Makefile (with some defaults set). These are used in built scripts
4802 such as exicyclog, but they have never been used in the exim_install
4803 script (though there are many overriding facilities there). I have
4804 arranged that the exim_install script now takes note of these two
4807 PH/35 Installed configuration files for Dragonfly.
4809 PH/36 When a locally submitted message by a trusted user did not contain a
4810 From: header, and the sender address was obtained from -f or from an SMTP
4811 MAIL command, and the trusted user did not use -F to supply a sender
4812 name, $originator_name was incorrectly used when constructing a From:
4813 header. Furthermore, $originator_name was used for submission mode
4814 messages from external hosts without From: headers in a similar way,
4815 which is clearly wrong.
4817 PH/37 Added control=suppress_local_fixups.
4819 PH/38 When log_selector = +received_sender was set, and the addition of the
4820 sender made the log line's construction buffer exactly full, or one byte
4821 less than full, an overflow happened when the terminating "\n" was
4824 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4825 when the result of a list match is failure because a DNS lookup failed.
4827 PH/40 RM_COMMAND is now used in the building process.
4829 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4830 the "build-* directories that it finds.
4832 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4833 address in a domain literal was a prefix of an interface address.
4835 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4836 when verifying a sender address, unless rewrite_headers is false.
4838 PH/44 Wrote a long comment about why errors_to addresses are verified as
4839 recipients, not senders.
4841 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4842 the ratelimit ACL was added.
4844 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4846 PH/46 Added extra information about PostgreSQL errors to the error string.
4848 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4849 avoiding the need to install special zones in a real server. This is
4850 backwards compatible; if it can't find the fake resolver, it drops back.
4851 Thus, both old and new test suites can be run.
4853 TF/02 Added util/ratelimit.pl
4855 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4856 clock is set back in time.
4858 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4859 Candler <B.Candler@pobox.com>.
4861 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4862 OK for addresses that are the result of redirections.
4864 PH/48 A number of further additions for the benefit of the new test suite,
4865 including a fake gethostbyname() that interfaces to the fake DNS resolver
4868 TF/06 The fix for widen_domains has also been applied to qualify_single and
4869 search_parents which are the other dnslookup options that can cause
4872 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4875 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4877 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4878 service" if it encountered an MX record with an empty target hostname.
4879 The message is now "an MX or SRV record indicated no SMTP service".
4881 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4882 if the DNS of the sending site is misconfigured. This is quite a
4883 common situation. This change restores the behaviour of treating a
4884 helo verification defer as a failure.
4886 PH/51 If self=fail was set on a router, the bounce message did not include the
4887 actual error message.
4893 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4895 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4896 command itself ended in error, the underlying message about the transport
4897 filter timeout was being overwritten with the pipe command error. Now the
4898 underlying error message should be appended to the second error message.
4900 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4902 PH/02 Reduce the amount of output that "make" produces by default. Full output
4903 can still be requested.
4905 PH/03 The warning log line about a condition test deferring for a "warn" verb
4906 was being output only once per connection, rather than after each
4907 occurrence (because it was using the same function as for successful
4908 "warn" verbs). This seems wrong, so I have changed it.
4910 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4911 it should not have, which might have caused a crash in the right
4912 circumstances, but probably never did.
4914 PH/04 Installed a modified version of Tony Finch's patch to make submission
4915 mode fix the return path as well as the Sender: header line, and to
4916 add a /name= option so that you can make the user's friendly name appear
4919 TF/03 Added the control = fakedefer ACL modifier.
4921 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4922 Mark Lowes for thorough testing.
4924 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4926 TK/03 Merged latest SRS patch from Miles Wilton.
4928 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4929 with the definition in sysexits.h (which is #included earlier).
4930 Fortunately, Exim does not actually use EX_OK. The code used to try to
4931 preserve the sysexits.h value, by assuming that macro definitions were
4932 scanned for macro replacements. I have been disabused of this notion,
4933 so now the code just undefines EX_OK before #including unistd.h.
4935 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4936 in the smtp transport. When a block could not be written in a single
4937 write() function, the timeout was being re-applied to each part-write.
4938 This seems wrong - if the receiver was accepting one byte at a time it
4939 would take for ever. The timeout is now adjusted when this happens. It
4940 doesn't have to be particularly precise.
4942 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4943 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4945 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4946 without a domain if the check to <postmaster@domain> fails.
4948 SC/01 Eximstats: added -xls and the ability to specify output files
4949 (patch written by Frank Heydlauf).
4951 SC/02 Eximstats: use FileHandles for outputting results.
4953 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4955 SC/04 Eximstats: fixed display of large numbers with -nvr option
4957 SC/05 Eximstats: fixed merging of reports with empty tables.
4959 SC/06 Eximstats: added the -include_original_destination flag
4961 SC/07 Eximstats: removed tabs and trailing whitespace.
4963 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4965 TK/06 MBOX spool code: Add real "From " MBOX separator line
4966 so the .eml file is really in mbox format (even though
4967 most programs do not really care). Patch from Alex Miller.
4969 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4970 The latter is generated from $received_to and is only set if the
4971 message has one envelope recipient. SA can use these headers,
4972 obviously out-of-the-box. Patch from Alex Miller.
4974 PH/08 The ${def test on a variable was returning false if the variable's
4975 value was "0", contrary to what the specification has always said!
4976 The result should be true unless the variable is empty.
4978 PH/09 The syntax error of a character other than { following "${if
4979 def:variable_name" (after optional whitespace) was not being diagnosed.
4980 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4981 accidental colon was present, for example, could give incorrect results.
4983 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4984 result is used (not including appendfile, where other changes are about
4987 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4988 This involved changing a lot of size variables from int to off_t. It
4989 should work with maildirs and everything.
4991 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4992 spamd dying while we are connected to it.
4994 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4997 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4998 function for BDB 4.3.
5000 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5002 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5003 the sources for PCRE have been re-arranged and more files are now
5006 PH/15 The code I had for printing potentially long long variables in PH/11
5007 above was not the best (it lost precision). The length of off_t variables
5008 is now inspected at build time, and an appropriate printing format (%ld
5009 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5010 to be "long long int" or "long int". This is needed for the internal
5011 formatting function string_vformat().
5013 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5014 the configuration file to be ":syslog", then the script "guesses" where
5015 the logs files are, rather than using the compiled in default. In our
5016 case the guess is not the same as the compiled default, so the script
5017 suddenly stopped working when I started to use syslog. The patch checks
5018 to see if log_file_path is "". If so, it attempts to read it from exim
5019 with no configuration file to get the compiled in version, before it
5020 falls back to the previous guessing code."
5022 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5023 implementing BATV in an Exim configuration. See NewStuff for the gory
5026 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5027 Makefile that are specific to HP-UX.
5029 PH/18 If the "use_postmaster" option was set for a recipient callout together
5030 with the "random" option, the postmaster address was used as the MAIL
5031 FROM address for the random test, but not for the subsequent recipient
5032 test. It is now used for both.
5034 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5035 patch removes a few documentation additions to RFC 3028, because the
5036 latest draft now contains them. It adds the new en;ascii-case comparator
5037 and a new error check for 8bit text in MIME parts. Comparator and
5038 require names are now matched exactly. I enabled the subaddress
5039 extension, but it is not well tested yet (read: it works for me)."
5041 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5042 rework some of the code of TK/09 above to avoid the hardwired use of
5043 "%lld" and "long long". Replaced the call to snprintf() with a call to
5046 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5047 records point to non-existent hosts", "retry timeout exceeded", and
5048 "retry time not reached for any host after a long failure period".
5050 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5051 experimental DomainKeys support:
5053 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5054 (2) On an error such as an illegally used "control", the wrong name for
5055 the control was given.
5057 These problems did NOT occur unless DomainKeys support was compiled.
5059 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5061 PH/24 Added ${if match_ip condition.
5063 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5064 databases so that it will be absolutely obvious if a crash occurs in the
5065 DB library. This is a regular occurrence (often caused by mis-matched
5068 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5069 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5070 picked up on a user's system that detects such things. There doesn't seem
5071 to be a gcc warning option for this - only an attribute that has to be
5072 put on the function's prototype. It seems that in Fedora Core 4 they have
5073 set this on a number of new functions. No doubt there will be more in due
5076 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5077 specify a transport. However, if an address that was verified by such a
5078 router was the subject of a callout, Exim crashed because it tried to
5079 read the rcpt_include_affixes from the non-existent transport. Now it
5080 just assumes that the setting of that option is false. This bug was
5081 introduced by 4.51/PH/31.
5083 PH/28 Changed -d+all to exclude +memory, because that information is very
5084 rarely of interest, but it makes the output a lot bigger. People tend to
5085 do -d+all out of habit.
5087 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5088 code in os-type was giving problems when libc.so lives in lib64, like on
5091 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5092 aren't the modern standard, and it seems that some systems' include files
5093 don't always have them. Exim was already checking for some of the newer
5094 ones like T_AAAA, and defining it itself. I've added checks for all the
5095 record types that Exim uses.
5097 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5098 not automatically generating a new one, as it is supposed to. This
5099 prevented TLS from working. If the file did exist, but contained invalid
5100 data, a new version was generated, as expected. It was only the case of a
5101 non-existent file that was broken.
5103 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5104 with a change in libdomainkeys > 0.64.
5106 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5107 from DNS. If the selector record carries the flag, it now has
5108 precedence over the domain-wide flag.
5110 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5112 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5113 the use of a port name that isn't defined in /etc/services) occurred, the
5114 message was deferred as in a normal delivery, and thus remained on the
5115 spool, instead of being failed because of the mua_wrapper setting. This
5116 is now fixed, and I tidied up some of the mua_wrapper messages at the
5119 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5120 the messages in a hash of arrays rather than using individual hashes.
5121 This is a bit cleaner and results in dramatic memory savings, albeit
5122 at a slight CPU cost.
5124 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5125 as requested by Marc Sherman.
5127 SC/10 Eximstats: added histograms for user specified patterns as requested
5130 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5132 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5133 fopen() in the content-scanning modules that did not already have it.
5139 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5140 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5142 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5144 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5146 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5147 to test Sieve filters that use "vacation".
5149 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5150 that changes the way the GnuTLS parameters are stored in the cache file.
5151 The new format can be generated externally. For backward compatibility,
5152 if the data in the cache doesn't make sense, Exim assumes it has read an
5153 old-format file, and it generates new data and writes a new file. This
5154 means that you can't go back to an older release without removing the
5157 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5158 work if there are any delivery delays because "one_time" forces the
5159 parent to be marked "delivered", so its unseen clone is never tried
5160 again. For this reason, Exim now forbids the simultaneous setting of
5163 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5164 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5165 entry if you want to know the details. The fix, however, neglected to
5166 consider the case when local delivery batching is involved. The test for
5167 "previously delivered" was not happening when checking to see if an
5168 address could be batched with a previous (undelivered) one; under
5169 certain circumstances this could lead to multiple deliveries to the same
5172 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5173 in its include files, and this causes problems building Exim.
5175 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5176 header_syntax) but Exim was just ignoring anything given after a slash.
5177 In particular, this caused confusion with an attempt to use "verify =
5178 reverse_host_lookup/defer_ok". An error is now given when options are
5179 supplied for verify items that do not have them. (Maybe reverse_host_
5180 lookup should have a defer_ok option, but that's a different point.)
5182 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5183 defined by RFC 821) to 2048, because there were problems with some AUTH
5184 commands, and RFC 1869 says the size should be increased for extended
5185 SMTP commands that take arguments.
5187 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5190 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5191 "unknown" error; now it says that the functionality isn't in the binary.
5193 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5194 an address' error message when a string expansion fails (syntax or
5195 whatever). Otherwise the password may appear in the log. Following change
5196 PH/42 below, there is no longer a chance of it appearing in a bounce
5199 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5201 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5202 of its addresses was ever tried. (Bugzilla bug #2.)
5204 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5205 the result incorrectly in the debug output. (It correctly added a newline
5206 to what was transported.)
5208 TF/01 Added $received_time.
5210 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5211 commented out examples of how to interface to a virus scanner and to
5212 SpamAssassin. Also added commented examples of av_scanner and
5213 spamd_address settings.
5215 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5216 and controls are allowed in which ACLs. There were a couple of minor
5217 errors. Some of the entries in the conditions table (which is a table of
5218 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5219 negation of where the condition IS allowed.
5221 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5223 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5224 header file does not have a version number, so I've had to invent a new
5225 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5226 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5227 radiusclient), but it was contributed by a Radius user.
5229 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5230 files or directories, for OpenSSL.
5232 PH/19 When an Exim process that is running as root has to create an Exim log
5233 file, it does so in a subprocess that runs as exim:exim so as to get the
5234 ownership right at creation (otherwise, other Exim processes might see
5235 the file with the wrong ownership). There was no test for failure of this
5236 fork() call, which would lead to the process getting stuck as it waited
5237 for a non-existent subprocess. Forks do occasionally fail when resources
5238 run out. I reviewed all the other calls to fork(); they all seem to check
5241 PH/20 When checking for unexpected SMTP input at connect time (before writing
5242 the banner), Exim was not dealing correctly with a non-positive return
5243 from the read() function. If the client had disconnected by this time,
5244 the result was a log entry for a synchronization error with an empty
5245 string after "input=" when read() returned zero. If read() returned -1
5246 (an event I could not check), uninitialized data bytes were printed.
5247 There were reports of junk text (parts of files, etc) appearing after
5250 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5252 PH/22 Added support for macro redefinition, and (re)definition in between
5253 driver and ACL definitions.
5255 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5256 forgetting to use the resulting value; it was using the unexpanded value.
5258 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5259 hadn't been configured. The fix is from Juergen Kreileder, who
5260 understands it better than I do:
5262 "Here's what I see happening with three configured cyrus_sasl
5263 authenticators configured (plain, login, cram-md5):
5265 On startup auth_cyrus_sasl_init() gets called for each of these.
5266 This means three calls to sasl_listmech() without a specified mech_list.
5267 => SASL tests which mechs of all available mechs actually work
5268 => three warnings about OTP not working
5269 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5271 With the patch, sasl_listmech() also gets called three times. But now
5272 SASL's mech_list option is set to the server_mech specified in the the
5273 authenticator. Or in other words, the answer from sasl_listmech()
5274 gets limited to just the mech you're testing for (which is different
5276 => the return list contains just 'plain' or 'login', 'cram-md5' or
5277 nothing depending on the value of ob->server_mech.
5279 I've just tested the patch: Authentication still works fine,
5280 unavailable mechs specified in the exim configuration are still
5281 caught, and the auth.log warnings about OTP are gone."
5283 PH/25 When debugging is enabled, the contents of the command line are added
5284 to the debugging output, even when log_selector=+arguments is not
5287 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5288 answer is "GNU", and only if the return is "GNU/something" is the answer
5291 PH/27 $acl_verify_message is now set immediately after the failure of a
5292 verification in an ACL, and so is available in subsequent modifiers. In
5293 particular, the message can be preserved by coding like this:
5295 warn !verify = sender
5296 set acl_m0 = $acl_verify_message
5298 Previously, $acl_verify_message was set only while expanding "message"
5299 and "log_message" when a very denied access.
5301 PH/28 Modified OS/os.c-Linux with
5303 -#ifndef OS_LOAD_AVERAGE
5304 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5306 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5307 nomenclature these days.)
5309 PH/29 Installed patch from the Sieve maintainer that adds the options
5310 sieve_useraddress and sieve_subaddress to the redirect router.
5312 PH/30 In these circumstances:
5313 . Two addresses routed to the same list of hosts;
5314 . First host does not offer TLS;
5315 . First host accepts first address;
5316 . First host gives temporary error to second address;
5317 . Second host offers TLS and a TLS session is established;
5318 . Second host accepts second address.
5319 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5320 and peerdn, if requested) that were in fact used only for the second
5323 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5324 attention to any local part prefix or suffix that was matched by the
5325 router that accepted the address. It now behaves in the same way as it
5326 does for delivery: the affixes are removed from the local part unless
5327 rcpt_include_affixes is set on the transport.
5329 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5330 timeout during the DATA phase of an incoming message.
5332 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5333 applied a patch sanctioned by the Sieve maintainer.
5335 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5336 the uid or gid is negative. A case of a negative gid caused this to be
5337 noticed. The fix allows for either to be negative.
5339 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5340 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5343 PH/36 The change PH/12 above was broken. Fixed it.
5345 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5346 the grounds that routing the same address twice would always produce the
5347 same answer. This might have been true once, but it is certainly no
5348 longer true now. Routing a child address may depend on the previous
5349 routing that produced that child. Some complicated redirection strategies
5350 went wrong when messages had multiple recipients, and made Exim's
5351 behaviour dependent on the order in which the addresses were given.
5353 I have moved the duplicate checking until after the routing is complete.
5354 Exim scans the addresses that are assigned to local and remote
5355 transports, and removes any duplicates. This means that more work will be
5356 done, as duplicates will always all be routed, but duplicates are
5357 presumably rare, so I don't expect this is of any significance.
5359 For deliveries to pipes, files, and autoreplies, the duplicate checking
5360 still happens during the routing process, since they are not going to be
5363 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5364 It corrects a timeout issue with spamd. This is Ian's comment: "The
5365 background is that sometimes spamd either never reads data from a
5366 connection it has accepted, or it never writes response data. The exiscan
5367 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5368 blindly assumes that writes won't block so it may never time out."
5370 PH/39 Allow G after quota size as well as K and M.
5372 PH/40 The value set for $authenticated_id in an authenticator may not contain
5373 binary zeroes or newlines because the value is written to log lines and
5374 to spool files. There was no check on this. Now the value is run through
5375 the string_printing() function so that such characters are converted to
5376 printable escape sequences.
5378 PH/41 $message_linecount is a new variable that contains the total number of
5379 lines in the message. Compare $body_linecount, which is the count for the
5382 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5383 bounce and delay warning messages, except in certain special cases, which
5386 (a) An SMTP error message from a remote host;
5387 (b) A message specified in a :fail: redirection;
5388 (c) A message specified in a "fail" command in a system filter;
5389 (d) A message specified in a FAIL return from the queryprogram router;
5390 (e) A message specified by the cannot_route_message router option.
5392 In these cases only, Exim does include the error details in bounce and
5393 warning messages. There are also a few cases where bland messages such
5394 as "unrouteable address" or "local delivery error" are given.
5396 PH/43 $value is now also set for the "else" part of a ${run expansion.
5398 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5399 being worked on, but at least Exim now implements the latest version to
5402 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5403 process to complete was treated as a delivery failure, a timeout while
5404 writing the message to the pipe was logged, but erroneously treated as a
5405 successful delivery. Such timeouts include transport filter timeouts. For
5406 consistency with the overall process timeout, these timeouts are now
5407 treated as errors, giving rise to delivery failures by default. However,
5408 there is now a new Boolean option for the pipe transport called
5409 timeout_defer, which, if set TRUE, converts the failures into defers for
5410 both kinds of timeout. A transport filter timeout is now identified in
5413 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5414 systems where "make" and "gmake" are different, calling "gmake" at top
5415 level broke things. I've arranged for the value of $(MAKE) to be passed
5416 from the Makefile to this script so that it can call the same version of
5420 A note about Exim versions 4.44 and 4.50
5421 ----------------------------------------
5423 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5424 changes of various kinds. As a consequence, a big documentation update was
5425 needed. This delayed the release for rather longer than seemed good, especially
5426 in the light of a couple of (minor) security issues. Therefore, the changes
5427 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5428 release. So 4.44 and 4.50 are in effect two different branches that both start
5431 I have left the 4.50 change log unchanged; it contains all the changes since
5432 4.43. The change log for 4.44 is below; many of its items are identical to
5433 those for 4.50. This seems to be the most sensible way to preserve the
5434 historical information.
5440 1. Minor wording change to the doc/README.SIEVE file.
5442 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5443 computation of the current number of files was incorrect.
5445 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5446 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5449 4. Give more explanation in the error message when the command for a transport
5450 filter fails to execute.
5452 5. There are several places where Exim runs a non-Exim command in a
5453 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5454 was being done only for the command run by the queryprogram router. It is
5455 now done for all such subprocesses. The other cases are: ${run, transport
5456 filters, and the commands run by the lmtp and pipe transports.
5458 6. Added CONFIGURE_GROUP build-time option.
5460 7. Some older OS have a limit of 256 on the maximum number of file
5461 descriptors. Exim was using setrlimit() to set 1000 as a large value
5462 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5463 systems. I've change it so that if it can't get 1000, it tries for 256.
5465 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5466 was an oversight, and furthermore, ever since the addition of extra
5467 controls (e.g. 4.43/32), the checks on when to allow different forms of
5468 "control" were broken. There should now be diagnostics for all cases when a
5469 control that does not make sense is encountered.
5471 9. Added the /retain_sender option to "control=submission".
5473 10. $recipients is now available in the predata ACL (oversight).
5475 11. Tidy the search cache before the fork to do a delivery from a message
5476 received from the command line. Otherwise the child will trigger a lookup
5477 failure and thereby defer the delivery if it tries to use (for example) a
5478 cached ldap connection that the parent has called unbind on.
5480 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5481 of $address_data from the recipient verification was clobbered by the
5482 sender verification.
5484 13. The value of address_data from a sender verification is now available in
5485 $sender_address_data in subsequent conditions in the ACL statement.
5487 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5489 15. Added a new option "connect=<time>" to callout options, to set a different
5492 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5493 was its contents. (It was OK if the option was not defined at all.)
5495 17. A "Completed" log line is now written for messages that are removed from
5496 the spool by the -Mrm option.
5498 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5499 information about exactly what failed.
5501 19. Added -dd to debug only the daemon process.
5503 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5504 handles timeouts, both on the server side and network timeouts. Renamed the
5505 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5507 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5508 It is now set to "smtps".
5510 22. $host_address is now set to the target address during the checking of
5511 ignore_target_hosts.
5513 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5514 being passed; this would have caused $sender_host_name to have been used if
5515 matching the list had actually called for a host name (not very likely,
5516 since this list is usually IP addresses). A host name is now passed as
5519 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5520 code that specifies a non-restarting handler (typically sigaction() in
5521 modern systems) in an attempt to fix a rare and obscure crash bug.
5523 25. Narrowed the window for a race in the daemon that could cause it to ignore
5524 SIGCHLD signals. This is not a major problem, because they are used only to
5525 wake it up if nothing else does.
5527 26. A malformed maildirsize file could cause Exim to calculate negative values
5528 for the mailbox size or file count. Odd effects could occur as a result.
5529 The maildirsize information is now recalculated if the size or filecount
5532 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5533 support for a long time. Removed HAVE_SYS_VFS_H.
5535 28. Installed the latest version of exipick from John Jetmore.
5537 29. In an address list, if the pattern was not a regular expression, an empty
5538 subject address (from a bounce message) matched only if the pattern was an
5539 empty string. Non-empty patterns were not even tested. This was the wrong
5540 because it is perfectly reasonable to use an empty address as part of a
5541 database query. An empty address is now tested by patterns that are
5542 lookups. However, all the other forms of pattern expect the subject to
5543 contain a local part and a domain, and therefore, for them, an empty
5544 address still always fails if the pattern is not itself empty.
5546 30. Exim went into a mad DNS loop when attempting to do a callout where the
5547 host was specified on an smtp transport, and looking it up yielded more
5548 than one IP address.
5550 31. Re-factored the code for checking spool and log partition space into a
5551 function that finds that data and another that does the check. The former
5552 is then used to implement four new variables: $spool_space, $log_space,
5553 $spool_inodes, and $log_inodes.
5555 32. The RFC2047 encoding function was originally intended for short strings
5556 such as real names; it was not keeping to the 75-character limit for
5557 encoded words that the RFC imposes. It now respects the limit, and
5558 generates multiple encoded words if necessary. To be on the safe side, I
5559 have increased the buffer size for the ${rfc2047: expansion operator from
5562 33. It is now permitted to omit both strings after an "if" condition; if the
5563 condition is true, the result is "true". As before, when the second string
5564 is omitted, a false condition yields an empty string. This makes it less
5565 cumbersome to write custom ACL and router conditions.
5567 34. Failure to deliver a bounce message always caused it to be frozen, even if
5568 there was an errors_to setting on the router. The errors_to setting is now
5571 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5572 canonical form (fully expanded) before being placed in
5573 $sender_host_address.
5575 36. The table in the code that translates DNS record types into text (T_A to
5576 "A" for instance) was missing entries for NS and CNAME. It is just possible
5577 that this could have caused confusion if both these types were looked up
5578 for the same domain, because the text type is used as part of Exim's
5579 per-process caching. But the chance of anyone hitting this buglet seems
5582 37. The dnsdb lookup has been extended in a number of ways.
5584 (1) There is a new type, "zns", which walks up the domain tree until it
5585 finds some nameserver records. It should be used with care.
5587 (2) There is a new type, "mxh", which is like "mx" except that it returns
5588 just the host names, not the priorities.
5590 (3) It is now possible to give a list of domains (or IP addresses) to be
5591 looked up. The behaviour when one of the lookups defers can be
5592 controlled by a keyword.
5594 (4) It is now possible to specify the separator character for use when
5595 multiple records are returned.
5597 38. The dnslists ACL condition has been extended: it is now possible to supply
5598 a list of IP addresses and/or domains to be looked up in a particular DNS
5601 39. Added log_selector=+queue_time_overall.
5603 40. When running the queue in the test harness, wait just a tad after forking a
5604 delivery process, to get repeatability of debugging output.
5606 41. Include certificate and key file names in error message when GnuTLS fails
5607 to set them up, because the GnuTLS error message doesn't include the name
5608 of the failing file when there is a problem reading it.
5610 42. Allow both -bf and -bF in the same test run.
5612 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5614 44. The "Exiscan patch" is now merged into the mainline Exim source.
5616 45. Sometimes the final signoff response after QUIT could fail to get
5617 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5618 < 0 before doing a fflush(). This bug looks as though it goes back to the
5619 introduction of TLS in release 3.20, but "sometimes" must have been rare
5620 because the tests only now provoked it.
5622 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5623 (this can affect the format of dates).
5625 47. exim_tidydb, when checking for the continued existence of a message for
5626 which it has found a message-specific retry record, was not finding
5627 messages that were in split spool directories. Consequently, it was
5628 deleting retry records that should have stayed in existence.
5630 48. Steve fixed some bugs in eximstats.
5632 49. The SPA authentication driver was not abandoning authentication and moving
5633 on to the next authenticator when an expansion was forced to fail,
5634 contradicting the general specification for all authenticators. Instead it
5635 was generating a temporary error. It now behaves as specified.
5637 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5638 (the order specifies the preference for clients). The order is now AES256,
5639 AES128, 3DES, ARCFOUR128.
5641 51. Small patch to Sieve code - explicitly set From: when generating an
5644 52. Exim crashed if a remote delivery caused a very long error message to be
5645 recorded - for instance if somebody sent an entire SpamAssassin report back
5646 as a large number of 550 error lines. This bug was coincidentally fixed by
5647 increasing the size of one of Exim's internal buffers (big_buffer) that
5648 happened as part of the Exiscan merge. However, to be on the safe side, I
5649 have made the code more robust (and fixed the comments that describe what
5652 53. Now that there can be additional text after "Completed" in log lines (if
5653 the queue_time_overall log selector is set), a one-byte patch to exigrep
5654 was needed to allow it to recognize "Completed" as not the last thing in
5657 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5658 patch that reportedly fixes this has been added. I am not expert enough to
5659 create a test for it. This is what the patch creator wrote:
5661 "I found a little strange behaviour of ldap code when working with
5662 Windows 2003 AD Domain, where users was placed in more than one
5663 Organization Units. When I tried to give exim partial DN, the exit code
5664 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5665 But simultaneously result of request was absolutely normal ldap result,
5666 so I produce this patch..."
5668 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5669 REFERENCE, so I have modified the code to exclude the patch when that macro
5672 55. Some experimental protocols are using DNS PTR records for new purposes. The
5673 keys for these records are domain names, not reversed IP addresses. The
5674 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5675 leaves it alone. Component reversal etc. now happens only for IP addresses.
5678 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5680 57. Double the size of the debug message buffer (to 2048) so that more of very
5681 long debug lines gets shown.
5683 58. The exicyclog utility now does better if the number of log files to keep
5684 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5686 59. Two changes related to the smtp_active_hostname option:
5688 (1) $smtp_active_hostname is now available as a variable.
5689 (2) The default for smtp_banner uses $smtp_active_hostname instead
5690 of $primary_hostname.
5692 60. The host_aton() function is supposed to be passed a string that is known
5693 to be a valid IP address. However, in the case of IPv6 addresses, it was
5694 not checking this. This is a hostage to fortune. Exim now panics and dies
5695 if the condition is not met. A case was found where this could be provoked
5696 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5697 components; fortuitously, this particular loophole had already been fixed
5698 by change 4.50/55 above.
5700 If there are any other similar loopholes, the new check in host_aton()
5701 itself should stop them being exploited. The report I received stated that
5702 data on the command line could provoke the exploit when Exim was running as
5703 exim, but did not say which command line option was involved. All I could
5704 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5705 running as the user.
5708 61. There was a buffer overflow vulnerability in the SPA authentication code
5709 (which came originally from the Samba project). I have added a test to the
5710 spa_base64_to_bits() function which I hope fixes it.
5713 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5714 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5716 63. The daemon start-up calls getloadavg() while still root for those OS that
5717 need the first call to be done as root, but it missed one case: when
5718 deliver_queue_load_max is set with deliver_drop_privilege. This is
5719 necessary for the benefit of the queue runner, because there is no re-exec
5720 when deliver_drop_privilege is set.
5722 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5723 This has been fixed.
5725 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5726 was in use, was not putting the data itself into the right store pool;
5727 consequently, it could be overwritten for a subsequent message in the same
5728 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5731 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5733 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5734 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5735 the function were treating the return as a boolean value, which happened to
5736 work because 0=false and not-0=true, but is not correct code.
5738 68. The host_aton() function was not handling scoped IPv6 addresses (those
5739 with, for example, "%eth0" on the end) correctly.
5741 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5742 negated items (that is, ~something) in unsigned ints. Some compilers
5743 apparently mutter when there is no cast.
5745 70. If an address verification called from an ACL failed, and did not produce a
5746 user-specific message (i.e. there was only a "system" message), nothing was
5747 put in $acl_verify_message. In this situation, it now puts the system
5750 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5751 session; change 4.31/43 added the unwanted input to the log line - except
5752 that it did not do this in the start of session case. It now does.
5754 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5755 This is wrong and can cause the other end to generate a synchronization
5756 error if it is another Exim or anything else that does the synchronization
5757 check. A QUIT command is no longer sent after a timeout.
5759 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5760 during host lookups.
5762 74. The defer_ok option of callout verification was not working if it was used
5763 when verifying addresses in header lines, that is, for this case:
5765 verify = header_sender/callout=defer_ok
5767 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5768 those file descriptors could be used for SMTP connections. If anything
5769 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5770 could be sent to the SMTP client, causing chaos. The daemon now opens
5771 stdin, stdout, and stderr to /dev/null when it puts itself into the
5774 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5775 log by default. The user can override this with suitable Perl magic.
5777 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5778 the log message when discard triggers, was not working for the DATA ACL or
5779 for the non-SMTP ACL.
5781 78. Error message wording change in sieve.c.
5783 79. If smtp_accept_max_per_host was set, the number of connections could be
5784 restricted to fewer than expected, because the daemon was trying to set up
5785 a new connection before checking whether the processes handling previous
5786 connections had finished. The check for completed processes is now done
5787 earlier. On busy systems, this bug wouldn't be noticed because something
5788 else would have woken the daemon, and it would have reaped the completed
5791 80. If a message was submitted locally by a user whose login name contained one
5792 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5793 It caused a spool format error. I have fixed the spool reading code. A
5794 related problem was that the "from" clause in the Received: line became
5795 illegal because of the space(s). It is now covered by ${quote_local_part.
5797 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5800 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5802 83. Patch from Sieve maintainer to fix unterminated string problem in
5803 "vacation" handling.
5805 84. Some minor changes to the Linux configuration files to help with other
5806 OS variants using glibc.
5808 85. One more patch for Sieve to update vacation handling to latest spec.
5811 ----------------------------------------------------
5812 See the note above about the 4.44 and 4.50 releases.
5813 ----------------------------------------------------
5819 1. Change 4.43/35 introduced a bug that caused file counts to be
5820 incorrectly computed when quota_filecount was set in an appendfile
5823 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5824 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5827 3. Give more explanation in the error message when the command for a transport
5828 filter fails to execute.
5830 4. There are several places where Exim runs a non-Exim command in a
5831 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5832 was being done only for the command run by the queryprogram router. It is
5833 now done for all such subprocesses. The other cases are: ${run, transport
5834 filters, and the commands run by the lmtp and pipe transports.
5836 5. Some older OS have a limit of 256 on the maximum number of file
5837 descriptors. Exim was using setrlimit() to set 1000 as a large value
5838 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5839 systems. I've change it so that if it can't get 1000, it tries for 256.
5841 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5842 was an oversight, and furthermore, ever since the addition of extra
5843 controls (e.g. 4.43/32), the checks on when to allow different forms of
5844 "control" were broken. There should now be diagnostics for all cases when a
5845 control that does not make sense is encountered.
5847 7. $recipients is now available in the predata ACL (oversight).
5849 8. Tidy the search cache before the fork to do a delivery from a message
5850 received from the command line. Otherwise the child will trigger a lookup
5851 failure and thereby defer the delivery if it tries to use (for example) a
5852 cached ldap connection that the parent has called unbind on.
5854 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5855 of $address_data from the recipient verification was clobbered by the
5856 sender verification.
5858 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5859 was its contents. (It was OK if the option was not defined at all.)
5861 11. A "Completed" log line is now written for messages that are removed from
5862 the spool by the -Mrm option.
5864 12. $host_address is now set to the target address during the checking of
5865 ignore_target_hosts.
5867 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5868 being passed; this would have caused $sender_host_name to have been used if
5869 matching the list had actually called for a host name (not very likely,
5870 since this list is usually IP addresses). A host name is now passed as
5873 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5874 code that specifies a non-restarting handler (typically sigaction() in
5875 modern systems) in an attempt to fix a rare and obscure crash bug.
5877 15. Narrowed the window for a race in the daemon that could cause it to ignore
5878 SIGCHLD signals. This is not a major problem, because they are used only to
5879 wake it up if nothing else does.
5881 16. A malformed maildirsize file could cause Exim to calculate negative values
5882 for the mailbox size or file count. Odd effects could occur as a result.
5883 The maildirsize information is now recalculated if the size or filecount
5886 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5887 support for a long time. Removed HAVE_SYS_VFS_H.
5889 18. Updated exipick to current release from John Jetmore.
5891 19. Allow an empty sender to be matched against a lookup in an address list.
5892 Previously the only cases considered were a regular expression, or an
5895 20. Exim went into a mad DNS lookup loop when doing a callout where the
5896 host was specified on the transport, if the DNS lookup yielded more than
5899 21. The RFC2047 encoding function was originally intended for short strings
5900 such as real names; it was not keeping to the 75-character limit for
5901 encoded words that the RFC imposes. It now respects the limit, and
5902 generates multiple encoded words if necessary. To be on the safe side, I
5903 have increased the buffer size for the ${rfc2047: expansion operator from
5906 22. Failure to deliver a bounce message always caused it to be frozen, even if
5907 there was an errors_to setting on the router. The errors_to setting is now
5910 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5911 canonical form (fully expanded) before being placed in
5912 $sender_host_address.
5914 24. Updated eximstats to version 1.33
5916 25. Include certificate and key file names in error message when GnuTLS fails
5917 to set them up, because the GnuTLS error message doesn't include the name
5918 of the failing file when there is a problem reading it.
5920 26. Expand error message when OpenSSL has problems setting up cert/key files.
5923 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5924 (this can affect the format of dates).
5926 28. exim_tidydb, when checking for the continued existence of a message for
5927 which it has found a message-specific retry record, was not finding
5928 messages that were in split spool directories. Consequently, it was
5929 deleting retry records that should have stayed in existence.
5931 29. eximstats updated to version 1.35
5932 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5933 1.35 - bugfix such that pie charts by volume are generated correctly
5935 30. The SPA authentication driver was not abandoning authentication and moving
5936 on to the next authenticator when an expansion was forced to fail,
5937 contradicting the general specification for all authenticators. Instead it
5938 was generating a temporary error. It now behaves as specified.
5940 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5941 (the order specifies the preference for clients). The order is now AES256,
5942 AES128, 3DES, ARCFOUR128.
5944 31. Small patch to Sieve code - explicitly set From: when generating an
5947 32. Exim crashed if a remote delivery caused a very long error message to be
5948 recorded - for instance if somebody sent an entire SpamAssassin report back
5949 as a large number of 550 error lines. This bug was coincidentally fixed by
5950 increasing the size of one of Exim's internal buffers (big_buffer) that
5951 happened as part of the Exiscan merge. However, to be on the safe side, I
5952 have made the code more robust (and fixed the comments that describe what
5955 33. Some experimental protocols are using DNS PTR records for new purposes. The
5956 keys for these records are domain names, not reversed IP addresses. The
5957 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5958 leaves it alone. Component reversal etc. now happens only for IP addresses.
5961 34. The host_aton() function is supposed to be passed a string that is known
5962 to be a valid IP address. However, in the case of IPv6 addresses, it was
5963 not checking this. This is a hostage to fortune. Exim now panics and dies
5964 if the condition is not met. A case was found where this could be provoked
5965 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5966 components; fortuitously, this particular loophole had already been fixed
5967 by change 4.50/55 or 4.44/33 above.
5969 If there are any other similar loopholes, the new check in host_aton()
5970 itself should stop them being exploited. The report I received stated that
5971 data on the command line could provoke the exploit when Exim was running as
5972 exim, but did not say which command line option was involved. All I could
5973 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5974 running as the user.
5977 35. There was a buffer overflow vulnerability in the SPA authentication code
5978 (which came originally from the Samba project). I have added a test to the
5979 spa_base64_to_bits() function which I hope fixes it.
5982 36. The daemon start-up calls getloadavg() while still root for those OS that
5983 need the first call to be done as root, but it missed one case: when
5984 deliver_queue_load_max is set with deliver_drop_privilege. This is
5985 necessary for the benefit of the queue runner, because there is no re-exec
5986 when deliver_drop_privilege is set.
5988 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5989 was in use, was not putting the data itself into the right store pool;
5990 consequently, it could be overwritten for a subsequent message in the same
5991 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5994 38. Sometimes the final signoff response after QUIT could fail to get
5995 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5996 < 0 before doing a fflush(). This bug looks as though it goes back to the
5997 introduction of TLS in release 3.20, but "sometimes" must have been rare
5998 because the tests only now provoked it.
6004 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6005 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6006 Now it returns an integer. A number of calls were still expecting a T/F
6007 return. Fortuitously, in all cases, the tests worked in OK situations,
6008 which is the norm. However, things would have gone wrong on any write
6009 failures on the smtp file descriptor. This function is used when sending
6010 messages over SMTP and also when doing verify callouts.
6012 2. When Exim is called to do synchronous delivery of a locally submitted
6013 message (the -odf or -odi options), it no longer closes stderr before doing
6016 3. Implemented the mua_wrapper option.
6018 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6020 5. Implemented the functions header_remove(), header_testname(),
6021 header_add_at_position(), and receive_remove_recipient(), and exported them
6024 6. If an ACL "warn" statement specified the addition of headers, Exim already
6025 inserted X-ACL-Warn: at the start if there was no header name. However, it
6026 was not making this test for the second and subsequent header lines if
6027 there were newlines in the string. This meant that an invalid header could
6028 be inserted if Exim was badly configured.
6030 7. Allow an ACL "warn" statement to add header lines at the start or after all
6031 the Received: headers, as well as at the end.
6033 8. Added the rcpt_4xx retry error code.
6035 9. Added postmaster_mailfrom=xxx to callout verification option.
6037 10. Added mailfrom=xxxx to the callout verification option, for verify=
6040 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6041 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6043 12. Inserted some casts to stop certain compilers warning when using pointer
6044 differences as field lengths or precisions in printf-type calls (mostly
6045 affecting debugging statements).
6047 13. Added optional readline() support for -be (dynamically loaded).
6049 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6050 same clock tick as a message's arrival, so that its received time was the
6051 same as the "first fail" time on the retry record, and that message
6052 remained on the queue past the ultimate address timeout, every queue runner
6053 would try a delivery (because it was past the ultimate address timeout) but
6054 after another failure, the ultimate address timeout, which should have then
6055 bounced the address, did not kick in. This was a "< instead of <=" error;
6056 in most cases the first failure would have been in the next clock tick
6057 after the received time, and all would be well.
6059 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6060 being recognized when the domain list was tested by the match_domain
6061 condition in an expansion string.
6063 16. Added the ${str2b64: operator.
6065 17. Exim was always calling setrlimit() to set a large limit for the number of
6066 processes, without checking whether the existing limit was already
6067 adequate. (It did check for the limit on file descriptors.) Furthermore,
6068 errors from getrlimit() and setrlimit() were being ignored. Now they are
6069 logged to the main and panic logs, but Exim does carry on, to try to do its
6070 job under whatever limits there are.
6072 18. Imported PCRE 5.0.
6074 19. Trivial typo in log message " temporarily refused connection" (the leading
6077 20. If the log selector return_path_on_delivery was set and an address was
6078 redirected to /dev/null, the delivery process crashed because it assumed
6079 that a return path would always be set for a "successful" delivery. In this
6080 case, the whole delivery is bypassed as an optimization, and therefore no
6083 21. Internal re-arrangement: the function for sending a challenge and reading
6084 a response while authentication was assuming a zero-terminated challenge
6085 string. It's now changed to take a pointer and a length, to allow for
6086 binary data in such strings.
6088 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6090 23. Exim was not respecting finduser_retries when seeking the login of the
6091 uid under which it was called; it was always trying 10 times. (The default
6092 setting of finduser_retries is zero.) Also, it was sleeping after the final
6093 failure, which is pointless.
6095 24. Implemented tls_on_connect_ports.
6097 25. Implemented acl_smtp_predata.
6099 26. If the domain in control=submission is set empty, Exim assumes that the
6100 authenticated id is a complete email address when it generates From: or
6101 Sender: header lines.
6103 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6104 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6105 chown and chgrp in /bin and hostname in /usr/bin.
6107 28. Exim was keeping the "process log" file open after each use, just as it
6108 does for the main log. This opens the possibility of it remaining open for
6109 long periods when the USR1 signal hits a daemon. Occasional processlog
6110 errors were reported, that could have been caused by this. Anyway, it seems
6111 much more sensible not to leave this file open at all, so that is what now
6114 29. The long-running daemon process does not normally write to the log once it
6115 has entered its main loop, and it closes the log before doing so. This is
6116 so that log files can straightforwardly be renamed and moved. However,
6117 there are a couple of unusual error situations where the daemon does write
6118 log entries, and I had neglected to close the log afterwards.
6120 30. The text of an SMTP error response that was received during a remote
6121 delivery was being truncated at 512 bytes. This is too short for some of
6122 the long messages that one sometimes sees. I've increased the limit to
6125 31. It is now possible to make retry rules that apply only when a message has a
6126 specific sender, in particular, an empty sender.
6128 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6129 it possible to be selective about when SMTP synchronization is enforced.
6131 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6133 32. Implemented hosts_connection_nolog.
6135 33. Added an ACL for QUIT.
6137 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6140 35. Added mailbox_size and mailbox_filecount to appendfile.
6142 36. Added control = no_multiline_responses to ACLs.
6144 37. There was a bug in the logic of the code that waits for the clock to tick
6145 in the case where the clock went backwards by a substantial amount such
6146 that the microsecond fraction of "now" was more than the microsecond
6147 fraction of "then" (but the whole seconds number was less).
6149 38. Added support for the libradius Radius client library this is found on
6150 FreeBSD (previously only the radiusclient library was supported).
6156 1. When certain lookups returned multiple values in the form name=value, the
6157 quoting of the values was not always being done properly. Specifically:
6158 (a) If the value started with a double quote, but contained no whitespace,
6160 (b) If the value contained whitespace other than a space character (i.e.
6161 tabs or newlines or carriage returns) it was not quoted.
6162 This fix has been applied to the mysql and pgsql lookups by writing a
6163 separate quoting function and calling it from the lookup code. The fix
6164 should probably also be applied to nisplus, ibase and oracle lookups, but
6165 since I cannot test any of those, I have not disturbed their existing code.
6167 2. A hit in the callout cache for a specific address caused a log line with no
6168 reason for rejecting RCPT. Now it says "Previous (cached) callout
6169 verification failure".
6171 3. There was an off-by-one bug in the queryprogram router. An over-long
6172 return line was truncated at 256 instead of 255 characters, thereby
6173 overflowing its buffer with the terminating zero. As well as fixing this, I
6174 have increased the buffer size to 1024 (and made a note to document this).
6176 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6177 when Exim is waiting for an SMTP response from a remote server, Exim
6178 restarts its select() call on the socket, thereby resetting its timeout.
6179 This is not a problem when such interrupts are rare. Somebody set up a cron
6180 job to run exiwhat every 2 minutes, which is less than the normal select()
6181 timeout (5 or 10 minutes). This meant that the select() timeout never
6182 kicked in because it was always reset. I have fixed this by comparing the
6183 time when an interrupt arrives with the time at the start of the first call
6184 to select(). If more time than the timeout has elapsed, the interrupt is
6185 treated as a timeout.
6187 5. Some internal re-factoring in preparation for the addition of Sieve
6188 extensions (by MH). In particular, the "personal" test is moved to a
6189 separate function, and given an option for scanning Cc: and Bcc: (which is
6190 not set for Exim filters).
6192 6. When Exim created an email address using the login of the caller as the
6193 local part (e.g. when creating a From: or Sender: header line), it was not
6194 quoting the local part when it contained special characters such as @.
6196 7. Installed new OpenBSD configuration files.
6198 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6199 try to make them clearer.
6201 9. Callout options, other than the timeout value, were being ignored when
6202 verifying sender addresses in header lines. For example, when using
6204 verify = header_sender/callout=no_cache
6206 the cache was (incorrectly) being used.
6208 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6209 only the Cygwin environment.
6211 11. When return_path_on_delivery was set as a log selector, if different remote
6212 addresses in the same message used different return paths and parallel
6213 remote delivery occurred, the wrong values would sometimes be logged.
6214 (Whenever a remote delivery process finished, the return path value from
6215 the most recently started remote delivery process was logged.)
6217 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6218 lines when AUTH and/or TLS are in use. This is the "received protocol"
6219 field. Exim used to use "asmtp" for authenticated SMTP, without any
6220 indication (in the protocol name) for TLS use. Now it follows the RFC and
6221 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6222 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6223 names appear in log lines as well as in Received: header lines.
6225 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6226 extensions, and comparison tests, and to fix some bugs.
6228 14. Changes to the "personal" filter test:
6230 (1) The test was buggy in that it was just doing the equivalent of
6231 "contains" tests on header lines. For example, if a user's address was
6232 anne@some.where, the "personal" test would incorrectly be true for
6234 To: susanne@some.where
6236 This test is now done by extracting each address from the header in turn,
6237 and checking the entire address. Other tests that are part of "personal"
6238 are now done using regular expressions (for example, to check local parts
6239 of addresses in From: header lines).
6241 (2) The list of non-personal local parts in From: addresses has been
6242 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6243 taken from the Sieve specification recommendations.
6245 (3) If the message contains any header line starting with "List-" it is
6246 treated as non-personal.
6248 (4) The test for "circular" in the Subject: header line has been removed
6249 because it now seems ill-conceived.
6251 15. Minor typos in src/EDITME comments corrected.
6253 16. Installed latest exipick from John Jetmore.
6255 17. If headers_add on a router specified a text string that was too long for
6256 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6257 of string_sprintf() is now avoided.
6259 18. $message_body_size was not set (it was always zero) when running the DATA
6260 ACL and the local_scan() function.
6262 19. For the "mail" command in an Exim filter, no default was being set for
6263 the once_repeat time, causing a random time value to be used if "once" was
6264 specified. (If the value happened to be <= 0, no repeat happened.) The
6265 default is now 0s, meaning "never repeat". The "vacation" command was OK
6266 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6267 (I found it when inspecting the code).
6269 20. There is now an overall timeout for performing a callout verification. It
6270 defaults to 4 times the callout timeout, which applies to individual SMTP
6271 commands during the callout. The overall timeout applies when there is more
6272 than one host that can be tried. The timeout is checked before trying the
6273 next host. This prevents very long delays if there are a large number of
6274 hosts and all are timing out (e.g. when the network connections are timing
6275 out). The value of the overall timeout can be changed by specifying an
6276 additional sub-option for "callout", called "maxwait". For example:
6278 verify = sender/callout=5s,maxwait=20s
6280 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6281 to the end before writing, but this should make it even safer).
6283 22. Exim was forgetting that it had advertised PIPELINING for the second and
6284 subsequent messages on an SMTP connection. It was also not resetting its
6285 memory on STARTTLS and an internal HELO.
6287 23. When Exim logs an SMTP synchronization error within a session, it now
6288 records whether PIPELINING has been advertised or not.
6290 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6291 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6292 rather than long int.
6294 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6296 26. Added the never_mail option to autoreply.
6302 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6303 crash if the getsockname() call failed; this can happen if a connection is
6304 closed very soon after it is established. The problem was simply in the
6305 order in which certain operations were done, causing Exim to try to write
6306 to the SMTP stream before it had set up the file descriptor. The bug has
6307 been fixed by making things happen in the correct order.
6313 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6314 before the connection was closed, thus losing the rejection response.
6316 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6317 some early Solaris releases, but causes trouble in current releases where
6318 socklen_t is defined.
6320 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6323 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6326 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6327 that releasing the top of it at the end releases what was used for sub-
6328 expansions (unless the block got too big). However, discard this block if
6329 the first thing is a variable or header, so that we can use its block when
6330 it is dynamic (useful for very large $message_headers, for example).
6332 6. Lookups now cache *every* query, not just the most recent. A new, separate
6333 store pool is used for this. It can be recovered when all lookup caches are
6334 flushed. Lookups now release memory at the end of their result strings.
6335 This has involved some general refactoring of the lookup sources.
6337 7. Some code has been added to the store_xxx() functions to reduce the amount
6338 of flapping under certain conditions.
6340 8. log_incoming_interface used to affect only the <= reception log lines. Now
6341 it causes the local interface and port to be added to several more SMTP log
6342 lines, for example "SMTP connection from", and rejection lines.
6344 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6346 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6348 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6349 could be overwritten at the end of the current message (or the start of a
6350 new message if it was set in a HELO ACL). The value is now preserved for
6351 the duration of the SMTP connection.
6353 12. If a transport had a headers_rewrite setting, and a matching header line
6354 contained an unqualified address, that address was qualified, even if it
6355 did not match any rewriting rules. The underlying bug was that the values
6356 of the flags that permit the existence of unqualified sender and recipient
6357 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6358 non-local messages, and by -bnq for local messages) were not being
6359 preserved with the message after it was received.
6361 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6362 "next input=" as part of the text comprising the host identity instead of
6363 the correct text. The code was using the same buffer for two different
6364 strings. However, depending on which order the printing function evaluated
6365 its arguments, the bug did not always show up. Under Linux, for example, my
6366 test suite worked just fine.
6368 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6369 doesn't work with some older versions of Perl. It has been changed to "my",
6370 which in any case is probably the better facility to use.
6372 15. A really picky compiler found some instances of statements for creating
6373 error messages that either had too many or two few arguments for the format
6376 16. The size of the buffer for calls to the DNS resolver has been increased
6377 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6378 for addresses that have a lot of PTR records. This alleviates a problem; it
6379 does not fully solve it.
6381 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6382 buffer now truncates the list and logs the incident, which is the same
6383 action as happens when Exim is looking up a host name and its aliases.
6384 Previously in this situation something unpredictable would happen;
6385 sometimes it was "internal error: store_reset failed".
6387 18. If a server dropped the connection unexpectedly when an Exim client was
6388 using GnuTLS and trying to read a response, the client delivery process
6389 crashed while trying to generate an error log message.
6391 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6392 string, for example:
6394 warn message = H1: something\nH2: something
6396 the text was added as a single header line from Exim's point of view
6397 though it ended up OK in the delivered message. However, searching for the
6398 second and subsequent header lines using $h_h2: did not work. This has been
6399 fixed. Similarly, if a system filter added multiple headers in this way,
6400 the routers could not see them.
6402 20. Expanded the error message when iplsearch is called with an invalid key to
6403 suggest using net-iplsearch in a host list.
6405 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6406 ACLs are no longer actually imposed (and a message to that effect is
6409 22. If a "gecos" field in a passwd entry contained escaped characters, in
6410 particular, if it contained a \" sequence, Exim got it wrong when building
6411 a From: or a Sender: header from that name. A second bug also caused
6412 incorrect handling when an unquoted " was present following a character
6413 that needed quoting.
6415 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6416 was not being matched caselessly.
6418 24. Arranged for all hyphens in the exim.8 source to be escaped with
6421 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6422 properly. Recipient callout cache records were still being keyed to include
6423 the sender, even when use_sender was set false. This led to far more
6424 callouts that were necessary. The sender is no longer included in the key
6425 when use_sender is false.
6427 26. Added "control = submission" modifier to ACLs.
6429 27. Added the ${base62d: operator to decode base 62 numbers.
6431 28. dnsdb lookups can now access SRV records.
6433 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6434 the configuration file.
6436 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6437 (-v) mode. This makes the output for a verbose queue run more intelligible.
6439 31. Added a use_postmaster feature to recipient callouts.
6441 32. Added the $body_zerocount variable, containing the number of binary zero
6442 bytes in the message body.
6444 33. The time of last modification of the "new" subdirectory is now used as the
6445 "mailbox time last read" when there is a quota error for a maildir
6448 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6450 35. Added +ignore_unknown as a special item in host lists.
6452 36. Code for decoding IPv6 addresses in host lists is now included, even if
6453 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6454 address was recognized as an IP address, but was then not correctly decoded
6455 into binary, causing unexpected and incorrect effects when compared with
6462 1. Very minor rewording of debugging text in manualroute to say "list of
6463 hosts" instead of "hostlist".
6465 2. If verify=header_syntax was set, and a header line with an unqualified
6466 address (no domain) and a large number of spaces between the end of the
6467 name and the colon was received, the reception process suffered a buffer
6468 overflow, and (when I tested it) crashed. This was caused by some obsolete
6469 code that should have been removed. The fix is to remove it!
6471 3. When running in the test harness, delay a bit after writing a bounce
6472 message to get a bit more predictability in the log output.
6474 4. Added a call to search_tidyup() just before forking a reception process. In
6475 theory, someone could use a lookup in the expansion of smtp_accept_max_
6476 per_host which, without the tidyup, could leave open a database connection.
6478 5. Added the variables $recipient_data and $sender_data which get set from a
6479 lookup success in an ACL "recipients" or "senders" condition, or a router
6480 "senders" option, similar to $domain_data and $local_part_data.
6482 6. Moved the writing of debug_print from before to after the "senders" test
6485 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6486 problems for message scanning, either using a data ACL, or using
6487 local_scan() because the Received: header was not generated till after they
6488 were called (in order to set the time as the time of reception completion).
6489 I have revised the way this works. The header is now generated after the
6490 body is received, but before the ACL or local_scan() are called. After they
6491 are run, the timestamp in the header is updated.
6497 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6498 before starting a queue runner without re-exec. This happened only when
6499 deliver_drop_privilege was set or when the Exim user was set to root. The
6500 effect of the bug was that timeouts during subsequent deliveries caused
6501 crashes instead of being properly handled. The handler is now left at its
6502 default (and expected) setting.
6504 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6505 message, again when deliver_drop_privilege is set or Exim is run as root.
6506 The bug described in (1) was not present in this case, but the tidying up
6507 of the other signals was missing. I have made the two cases consistent.
6509 3. The ignore_target_hosts setting on a manualroute router was being ignored
6510 for hosts that were looked up using the /MX notation.
6512 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6515 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6516 operated on the sender address. After changing the $sender_address to <>
6517 for the sender address verify, Exim was re-instated it as the original
6518 (before rewriting) address, but remembering that it had rewritten it, so it
6519 wasn't rewriting it again. This bug also had the effect of breaking the
6520 sender address verification caching when the sender address was rewritten.
6522 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6523 This has been changed so that if the ip literal address matches
6524 ignore_target_hosts, the router declines.
6526 7. Added expansion conditions match_domain, match_address, and match_local_
6527 part (NOT match_host).
6529 8. The placeholder for the Received: header didn't have a length field set.
6531 9. Added code to Exim itself and to exim_lock to test for a specific race
6532 condition that could lead to file corruption when using MBX delivery. The
6533 issue is with the lockfile that is created in /tmp. If this file is removed
6534 after a process has opened it but before that process has acquired a lock,
6535 there is the potential for a second process to recreate the file and also
6536 acquire a lock. This could lead to two Exim processes writing to the file
6537 at the same time. The added code performs the same test as UW imapd; it
6538 checks after acquiring the lock that its file descriptor still refers to
6539 the same named file.
6541 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6542 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6545 11. Added the smtp_active_hostname option. If used, this will typically be made
6546 to depend on the incoming interface address. Because $interface_address is
6547 not set up until the daemon has forked a reception process, error responses
6548 that can happen earlier (such as "too many connections") no longer contain
6551 12. If an expansion in a condition on a "warn" statement fails because a lookup
6552 defers, the "warn" statement is abandoned, and the next ACL statement is
6553 processed. Previously this caused the whole ACL to be aborted.
6555 13. Added the iplsearch lookup type.
6557 14. Added ident_timeout as a log selector.
6559 15. Added tls_certificate_verified as a log selector.
6561 16. Added a global option tls_require_ciphers (compare the smtp transport
6562 option of the same name). This controls incoming TLS connections.
6564 17. I finally figured out how to make tls_require_ciphers do a similar thing
6565 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6566 before starting the TLS session.
6568 18. Tabs are now shown as \t in -bP output.
6570 19. If the log selector return_path_on_delivery was set, Exim crashed when
6571 bouncing a message because it had too many Received: header lines.
6573 20. If two routers both had headers_remove settings, and the first one included
6574 a superfluous trailing colon, the final name in the first list and the
6575 first name in the second list were incorrectly joined into one item (with a
6576 colon in the middle).
6582 1. Added -C and -D options to the exinext utility, mainly to make it easier
6583 to include in the automated testing, but these could be helpful when
6584 multiple configurations are in use.
6586 2. The exinext utility was not formatting the output nicely when there was
6587 an alternate port involved in the retry record key, nor when there was a
6588 message id as well (for retries that were specific to a specific message
6589 and a specific host). It was also confused by IPv6 addresses, because of
6590 the additional colons they contain. I have fixed the IPv4 problem, and
6591 patched it up to do a reasonable job for IPv6.
6593 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6594 delivery, the log line now contains "pipelined" if PIPELINING was used.
6596 4. An SMTP transport process used to panic and die if the bind() call to set
6597 an explicit outgoing interface failed. This has been changed; it is now
6598 treated in the same way as a connect() failure.
6600 5. A reference to $sender_host_name in the part of a conditional expansion
6601 that was being skipped was still causing a DNS lookup. This no longer
6604 6. The def: expansion condition was not recognizing references to header lines
6605 that used bh_ and bheader_.
6607 7. Added the _cache feature to named lists.
6609 8. The code for checking quota_filecount in the appendfile transport was
6610 allowing one more file than it should have been.
6612 9. For compatibility with Sendmail, the command line option
6620 and sets the incoming protocol and host name (for trusted callers). The
6621 host name and its colon can be omitted when only the protocol is to be set.
6622 Note the Exim already has two private options, -pd and -ps, that refer to
6623 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6624 "s", but I don't think that's a major issue.
6626 10. A number of refactoring changes to the code, none of which should affect
6629 (a) The number of logging options was getting close to filling up the
6630 32-bit word that was used as a bit map. I have split them into two classes:
6631 those that are passed in the argument to log_write(), and those that are
6632 only ever tested independently outside of that function. These are now in
6633 separate 32-bit words, so there is plenty of room for expansion again.
6634 There is no change in the user interface or the logging behaviour.
6636 (b) When building, for example, log lines, the code previously used a
6637 macro that called string_cat() twice, in order to add two strings. This is
6638 not really sufficiently general. Furthermore, there was one instance where
6639 it was actually wrong because one of the argument was used twice, and in
6640 one call a function was used. (As it happened, calling the function twice
6641 did not affect the overall behaviour.) The macro has been replaced by a
6642 function that can join an arbitrary number of extra strings onto a growing
6645 (c) The code for expansion conditions now uses a table and a binary chop
6646 instead of a serial search (which was left over from when there were very
6647 few conditions). Also, it now recognizes conditions like "pam" even when
6648 the relevant support is not compiled in: a suitably worded error message is
6649 given if an attempt is made to use such a condition.
6651 11. Added ${time_interval:xxxxx}.
6653 12. A bug was causing one of the ddress fields not to be passed back correctly
6654 from remote delivery subprocesses. The field in question was not being
6655 subsequently used, so this caused to problems in practice.
6657 13. Added new log selectors queue_time and deliver_time.
6659 14. Might have fixed a bug in maildirsizefile handling that threw up
6660 "unexpected character" debug warnings, and recalculated the data
6661 unnecessarily. In any case, I expanded the warning message to give more
6664 15. Added the message "Restricted characters in address" to the statements in
6665 the default ACL that block characters like @ and % in local parts.
6667 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6668 Three changes have been made:
6670 (a) There was a serious bug; a negative response to MAIL caused the whole
6671 recipient domain to be cached as invalid, thereby blocking all messages
6672 to all local parts at the same domain, from all senders. This bug has
6673 been fixed. The domain is no longer cached after a negative response to
6674 MAIL if the sender used is not empty.
6676 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6679 (c) A new callout option, "use_sender" has been added for people who want
6680 the modified behaviour.
6686 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6689 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6690 indeed breaks things for older releases.
6692 3. Added additional logging to the case where there is a problem reading data
6693 from a filter that is running in a subprocess using a pipe, in order to
6694 try to track down a specific problem.
6696 4. Testing facility fudge: when running in the test harness and attempting
6697 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6698 getting "No route to host". Convert this to a timeout.
6700 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6703 6. Some OS don't have socklen_t but use size_t instead. This affects the
6704 fifth argument of getsockopt() amongst other things. This is now
6705 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6706 can be set for individual OS. I have set it for SunOS5, OSF1, and
6707 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6708 some earlier ones do not.
6710 7. Change 4.30/15 was not doing the test caselessly.
6712 8. The standard form for an IPv6 address literal was being rejected by address
6713 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6714 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6715 this, as well as the form without the "IPv6" on the front (but only when
6716 address literals are enabled, of course).
6718 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6720 10. Exim crashed if a message with an empty sender address specified by -f
6721 encountered a router with an errors_to setting. This could be provoked only
6722 by a command such as
6726 where an empty string was supplied; "<>" did not hit this bug.
6728 11. Installed PCRE release 4.5.
6730 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6731 remained set. It is now erased.
6733 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6734 times from message ids (which are base 36 rather than the normal 62).
6736 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6737 were being counted as actual protocol errors, and logged if the log
6738 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6739 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6740 and DATA following a set of rejected RCPTs do not count as protocol errors.
6741 In other words, Exim assumes they were pipelined, though this may not
6742 actually be the case. Of course, in all cases the client gets an
6743 appropriate error code.
6745 15. If a lookup fails in an ACL condition, a message about the failure may
6746 be available; it is used if testing the ACL cannot continue, because most
6747 such messages specify what the cause of the deferral is. However, some
6748 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6749 that caused an old message to be retained and used if a later statement
6750 caused a defer, replacing the real cause of the deferral.
6752 16. If an IP address had so many PTR records that the DNS lookup buffer
6753 was not large enough to hold them, Exim could crash while trying to process
6754 the truncated data. It now detects and logs this case.
6756 17. Further to 4.21/58, another change has been made: if (and only if) the
6757 first line of a message (the first header line) ends with CRLF, a bare LF
6758 in a subsequent header line has a space inserted after it, so as not to
6759 terminate the header.
6761 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6762 unnecessarily, used atoi() instead of strtol(), and didn't check the
6763 termination when getting file sizes from file names by regex.
6765 19. Completely re-implemented the support for maildirsize files, in the light
6766 of a number of problems with the previous contributed implementation
6767 (4.30/29). In particular:
6769 . If the quota is zero, the maildirsize file is maintained, but no quota is
6772 . If the maildir directory does not exist, it is created before any attempt
6773 to write a maildirsize file.
6775 . The quota value in the file is just a cache; if the quota is changed in
6776 the transport, the new value overrides.
6778 . A regular expression is available for excluding directories from the
6781 20. The autoreply transport checks the characters in options that define the
6782 message's headers; it allows continued headers, but it was checking with
6783 isspace() after an embedded newline instead of explicitly looking for a
6786 21. If all the "regular" hosts to which an address was routed had passed their
6787 expiry times, and had not reached their retry times, the address was
6788 bounced, even if fallback hosts were defined. Now Exim should go on to try
6791 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6792 equivalent code in the SMTP transport. Some hosts send humungous responses
6793 to HELO/EHLO, more than 1024 it seems.
6795 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6796 gives compiler warnings in some environments. I've now done it "properly",
6799 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6800 (because of problems with the built-in one) was declared to return uschar *
6801 instead of char *, causing compiler failure.
6803 25. Fixed a file descriptor leak when processing alias/forward files.
6805 26. Fixed a minor format string issue in dbfn.c.
6807 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6809 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6810 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6813 29. When a sender address is verified, it is cached, to save repeating the test
6814 when there is more than one recipient in a message. However, when the
6815 verification involves a callout, it is possible for different callout
6816 options to be set for different recipients. It is too complicated to keep
6817 track of this in the cache, so now Exim always runs a verification when a
6818 callout is required, relying on the callout cache for the optimization.
6819 The overhead is duplication of the address routing, but this should not be
6822 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6823 to be verified with callout=postmaster, and the main callout worked but the
6824 postmaster check failed, the verification correctly failed. However, if a
6825 subsequent RCPT command asked for sender verification *without* the
6826 postmaster check, incorrect caching caused this verification also to fail,
6829 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6830 it was not caching the DNS options (qualify_single, search_parents) that
6831 were used when the lookup failed. A subsequent lookup with different
6832 options therefore always gave the same answer, though there were cases
6833 where it should not have. (Example: a "domains = !$mx_any" option on a
6834 dnslookup router: the "domains" option is always processed without any
6835 widening, but the router might have qualify_single set.) Now Exim uses the
6836 cached value only when the same options are set.
6838 32. Added John Jetmore's "exipick" utility to the distribution.
6840 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6841 than a timeout (e.g. a certificate is required, and is not provided), an
6842 Exim server now closes the connection immediately. Previously it waited for
6843 the client to close - but if the client is SSL, it seems that they each
6844 wait for each other, leading to a delay before one of them times out.
6846 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6847 maintained 0.8.x compatibility because I don't think many are using it, and
6848 it is clearly obsolete.
6850 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6853 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6854 client certificate was expired. A simple patch fixes this, though I don't
6855 understand the full logic of why the verify callback is called multiple
6858 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6859 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6860 which causes problems with some clients (such as the Certicom SSL Plus
6861 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6862 disables the coutermeasure allowing Eudora to connect."
6864 38. Exim was not checking that a write() to a log file succeeded. This could
6865 lead to Bad Things if a log got too big, in particular if it hit a file
6866 size limit. Exim now panics and dies if it cannot write to a log file, just
6867 as it does if it cannot open a log file.
6869 39. Modified OS/Makefile-Linux so that it now contains
6871 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6873 The two -D definitions ensure that Exim is compiled with large file
6874 support, which makes it possible to handle log files that are bigger than
6877 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6878 instance) a domain was checked against a named list that involved a lookup,
6879 causing $domain_data to be set, then another domain was checked against the
6880 same list, then the first domain was re-checked, the value of $domain_data
6881 after the final check could be wrong. In particular, if the second check
6882 failed, it could be set empty. This bug probably also applied to
6885 41. The strip_trailing_dot option was not being applied to the address given
6886 with the -f command-line option.
6888 42. The code for reading a message's header from the spool was incrementing
6889 $received_count, but never initializing it. This meant that the value was
6890 incorrect (doubled) while delivering a message in the same process in which
6891 it was received. In the most common configuration of Exim, this never
6892 happens - a fresh exec is done - but it can happen when
6893 deliver_drop_privilege is set.
6895 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6896 it now includes up to 150 characters of the unexpected data in the log
6899 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6900 and building data strings. The size of both of these buffers was 10 000
6901 bytes - far larger than anybody would *ever* want, thought I. Needless to
6902 say, somebody hit the limit. I have increased the maximum line length to
6903 20 000 and the maximum data length of concatenated lines to 100 000. I have
6904 also fixed two bugs, because there was no checking on these buffers. Tsk,
6905 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6906 buffer is too small.
6908 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6909 lsearch lookups. Now it does.
6911 46. When parsing a route_list item in a manualroute router, a fixed-length
6912 buffer was used for the list of hosts. I made this 1024 bytes long,
6913 thinking that nobody would ever have a list of hosts that long. Wrong.
6914 Somebody had a whole pile of complicated expansion conditions, and the
6915 string was silently truncated, leading to an expansion error. It turns out
6916 that it is easier to change to an unlimited length (owing to other changes
6917 that have happened since this code was originally written) than to build
6918 structure for giving a limitation error. The length of the item that
6919 expands into the list of hosts is now unlimited.
6921 47. The lsearch lookup could not handle data where the length of text line was
6922 more than 4095 characters. Such lines were truncated, leading to shortened
6923 data being returned. It should now handle lines of any length.
6925 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6926 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6929 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6930 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6931 no longer happens in comment lines. A list of replacements is now placed
6932 at the head of all of the source files, except those whose only change is
6933 to replace PERL_COMMAND in the very first #! line.
6935 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6936 messages on the queue, with a bottom-up merge sort, using code contributed
6937 by Michael Haardt. This should make operations like -bp somewhat faster on
6938 large queues. It won't affect queue runners, except when queue_run_in_order
6941 51. Installed eximstats 1.31 in the distribution.
6943 52. Added support for SRV lookups to the dnslookup router.
6945 53. If an ACL referred to $message_body or $message_body_end, the value was not
6946 reset for any messages that followed in the same SMTP session.
6948 54. The store-handling optimization for building very long strings was not
6949 differentiating between the different store pools. I don't think this
6950 actually made any difference in practice, but I've tidied it.
6952 55. While running the routers to verify a sender address, $sender_address
6953 was still set to the sender address. This is wrong, because when routing to
6954 send a bounce to the sender, it would be empty. Therefore, I have changed
6955 it so that, while verifying a sender address, $sender_address is set to <>.
6956 (There is no change to what happens when verifying a recipient address.)
6958 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6959 target A or AAAA records (if not already returned) without resetting the
6960 qualify_single or search_parents options of the DNS resolver. These are
6961 inappropriate in this case because the targets of MX and SRV records must
6962 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6963 target that, when qualified, matched something in the local domain. These
6964 two options are now turned off when doing these lookups.
6966 57. It seems that at least some releases of Reiserfs (which does not have the
6967 concept of a fixed number of inodes) returns zero and not -1 for the
6968 number of available inodes. This interacted badly with check_spool_inodes,
6969 which assumed that -1 was the "no such thing" setting. What I have done is
6970 to check that the total number of inodes is greater than zero before doing
6971 the test of how many are available.
6973 58. When a "warn" ACL statement has a log_message modifier, the message is
6974 remembered, and not repeated. This is to avoid a lot of repetition when a
6975 message has many recipients that cause the same warning to be written.
6976 However, Exim was preserving the list of already written lines for an
6977 entire SMTP session, which doesn't seem right. The memory is now reset if a
6978 new message is started.
6980 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6981 debugging output unless log_rewrite was also set.
6983 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6984 of (int)(handle) when we know that handle contains (void *)(-1).
6986 61. The Exim daemon panic-logs an error return when it closes the incoming
6987 connection. However "connection reset by peer" seems to be common, and
6988 isn't really an error worthy of noting specially, so that particular error
6991 62. When Exim is trying to find all the local interfaces, it used to panic and
6992 die if the ioctl to get the interface flags failed. However, it seems that
6993 on at least one OS (Solaris 9) it is possible to have an interface that is
6994 included in the list of interfaces, but for which you get a failure error
6995 for this call. This happens when the interface is not "plumbed" into a
6996 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6997 failure of the "get flags" call assumes that the interface is down.
6999 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7000 makes life easier for people who are doing arithmetic on fields extracted
7001 from dates, where you often get leading zeros that should not be
7002 interpreted as octal.
7004 64. Added qualify_domain to the redirect router, to override the global
7007 65. If a pathologically long header line contained very many addresses (the
7008 report of this problem mentioned 10 000) and each of them was rewritten,
7009 Exim could use up a very large amount of memory. (It kept on making new
7010 copies of the header line as it rewrote, and never released the old ones.)
7011 At the expense of a bit more processing, the header rewriting function has
7012 been changed so that it no longer eats memory in this way.
7014 66. The generation of the Received: header has been moved from the time that a
7015 message starts to be received, to the time that it finishes. The timestamp
7016 in the Received: header should now be very close to that of the <= log
7017 line. There are two side-effects of this change:
7019 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7020 logged header lines no longer include the local Received: line, because
7021 it has not yet been created. The same applies to a copy of the message
7022 that is returned to a non-SMTP sender when a message is rejected.
7024 (b) When a filter file is tested using -bf, no additional Received: header
7025 is added to the test message. After some thought, I decided that this
7028 This change does not affect the value of $received_for. It is still set
7029 after address rewriting, but before local_scan() is called.
7031 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7033 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7034 gave an unhelpful panic error message, and a defer error. I have managed to
7035 change this behaviour so that it now rejects any supplied certificate,
7036 which seems right, as the list of acceptable certificates is empty.
7038 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7039 gave an unhelpful defer error. I have not managed to make this reject any
7040 supplied certificates, but the error message it gives is "no certificate
7041 supplied", which is not helpful.
7043 70. exigrep's output now also includes lines that are not associated with any
7044 message, but which match the given pattern. Implemented by a patch from
7045 Martin Sluka, which also tidied up the Perl a bit.
7047 71. Recipient callout verification, like sender verification, was using <> in
7048 the MAIL FROM command. This isn't really the right thing, since the actual
7049 sender may affect whether the remote host accepts the recipient or not. I
7050 have changed it to use the actual sender in the callout; this means that
7051 the cache record is now keyed on a recipient/sender pair, not just the
7052 recipient address. There doesn't seem to be a real danger of callout loops,
7053 since a callout by the remote host to check the sender would use <>.
7054 [SEE ABOVE: changed after hitting problems.]
7056 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7057 temporary errors. However, in the case of such a code being given after
7058 the end of a data transmission (i.e. after ".") Exim was failing to write
7059 a retry record for the message. (Yes, there was some broken host that was
7060 actually sending 8xx at this point.)
7062 73. An unknown lookup type in a host list could cause Exim to panic-die when
7063 the list was checked. (An example that provoked this was putting <; in the
7064 middle of a list instead of at the start.) If this happened during a DATA
7065 ACL check, a -D file could be left lying around. This kind of configuration
7066 error no longer causes Exim to die; instead it causes a defer error. The
7067 incident is still logged to the main and panic logs.
7069 74. Buglet left over from Exim 3 conversion. The message "too many messages
7070 in one connection" was written to the rejectlog but not the mainlog, except
7071 when address rewriting (yes!) was being logged.
7073 75. Added write_rejectlog option.
7075 76. When a system filter was run not as root (that is, when system_filter_user
7076 was set), the values of the $n variables were not being returned to the
7077 main process; thus, they were not subsequently available in the $sn
7080 77. Added +return_path_on_delivery log selector.
7082 78. A connection timeout was being treated differently from recipients deferred
7083 when testing hosts_max_try with a message that was older than the host's
7084 retry timeout. (The host should not be counted, thus allowing all hosts to
7085 be tried at least once before bouncing.) This may have been the cause of an
7086 occasionally reported bug whereby a message would remain on the queue
7087 longer than the retry timeout, but would be bounced if a delivery was
7088 forced. I say "may" because I never totally pinned down the problem;
7089 setting up timeout/retry tests is difficult. See also the next item.
7091 79. The ultimate address timeout was not being applied to errors that involved
7092 a combination of host plus message (for example, a timeout on a MAIL
7093 command). When an address resolved to a number of possible hosts, and they
7094 were not all tried for each delivery (e.g. because of hosts_max_try), a
7095 message could remain on the queue longer than the retry timeout.
7097 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7100 81. Fixed an obscure SMTP outgoing bug which required at least the following
7101 conditions: (a) there was another message waiting for the same server;
7102 (b) the server returned 5xx to all RCPT commands in the first message so
7103 that the message was not completed; (c) the server dropped the connection
7104 or gave a negative response to the RSET that Exim sends to abort the
7105 transaction. The observed case was a dropped connection after DATA that had
7106 been sent in pipelining mode. That is, the server had advertised PIPELINING
7107 but was not implementing it correctly. The effect of the bug was incorrect
7108 behaviour, such as trying another host, and this could lead to a crash.
7114 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7115 and daemon.c were passed as pointers to ints; they should have been
7116 pointers to socklen_t variables (which are typically unsigned ints).
7118 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7121 3. Fixed a really odd bug that affected only the testing scheme; patching a
7122 certain fixed string in the binary changed the value of another string that
7123 happened to be identical to the end of the original first string.
7125 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7126 name", it returns that address as the IP address. On some operating
7127 systems (e.g. Solaris), it also passes back the IP address string as the
7128 "host name". However, on others (e.g. Linux), it passes back an empty
7129 string. Exim wasn't checking for this, and was changing the host name to an
7130 empty string, assuming it had been canonicalized.
7132 5. Although rare, it is permitted to have more than one PTR record for a given
7133 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7134 all the names associated with an address, because they do in Solaris.
7135 However, it seems that they do not in Linux for data that comes from the
7136 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7137 I found this out when I moved to a new Linux workstation and tried to run
7138 the Exim test suite.
7140 To get round this problem I have changed the code so that it now does its
7141 own call to the DNS to look up PTR records when searching for a host name.
7142 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7143 addresses that are only in /etc/hosts are still found.
7145 This behaviour is, however, controlled by an option called host_lookup_
7146 order, which defaults to "bydns:byaddr". If people want to use the other
7147 order, or indeed, just use one or the other means of lookup, they can
7148 specify it in this variable.
7150 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7151 some operating systems, this comes back from gethostbyaddr() as an empty
7152 string, and this is what Exim used to test for. However, it seems that in
7153 other systems, "." is yielded. Exim now tests for this case too.
7155 7. The values of check_spool_space and check_log_space are now held internally
7156 as a number of kilobytes instead of an absolute number of bytes. If a
7157 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7158 kilobyte. This means that much larger values can be stored.
7160 8. Exim monitor: an attempt to get the action menu when not actually pointing
7161 at a message produces an empty menu entitled "No message selected". This
7162 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7163 no entries in it ("Shell widget menu has zero width and/or height"). So I
7164 have added a single, blank menu entry in this case.
7166 9. Added ${quote_local_part.
7168 10. MIME decoding is now applied to the contents of Subject: header lines when
7171 11. Now that a reference to $sender_host_address automatically causes a reverse
7172 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7173 host lookup before query-style lookups in lists that might use this
7174 variable. This has therefore been abolished, and the "net-" prefix is no
7175 longer necessary for query-style lookups.
7177 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7178 have been a typo for LFLAGS, so it has been changed.
7180 13. The install script calls Exim with "-C /dev/null" in order to find the
7181 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7182 to be output. However, since Exim outputs its version number before the
7183 error, it didn't break the script. It just looked ugly. I fixed this by
7184 always allowing "-C /dev/null" if the caller is root.
7186 14. Ignore overlarge ACL variable number when reading spool file - insurance
7187 against a later release with more variables having written the file.
7189 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7190 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7191 this, as well as the form without the "IPv6" on the front.
7193 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7194 OS/Makefile-Darwin file.
7196 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7197 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7199 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7200 variable. However, some OS use ints for inodes. I've added cast to long int
7201 to get rid of the compiler warning.
7203 19. I had forgotten to lock out "/../" in configuration file names when
7204 ALT_CONFIG_PREFIX was set.
7206 20. Routers used for verification do not need to specify transports. However,
7207 if such a router generated a host list, and callout was configured, Exim
7208 crashed, because it could not find a port number from the (non-existent)
7209 transport. It now assumes port 25 in this circumstance.
7211 21. Added the -t option to exigrep.
7213 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7214 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7215 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7216 them is set, LOOKUP_LSEARCH is forced.
7218 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7220 24. Added sender and host information to the "rejected by local_scan()" log
7221 line; previously there was no indication of these.
7223 25. Added .include_if_exists.
7225 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7226 Linux. It turns out that not all file systems support this. Apparently some
7227 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7228 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7229 not supported on the file descriptor, is now ignored when Exim is trying to
7230 sync a directory. This applies only to Linux.
7232 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7234 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7235 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7238 29. Added support for maildirsize files from supplied patch (modified a bit).
7240 30. The use of :fail: followed by an empty string could lead Exim to respond to
7241 sender verification failures with (e.g.):
7243 550 Verification failed for <xxx>
7244 550 Sender verify failed
7246 where the first response line was missing the '-' that indicates it is not
7247 the final line of the response.
7249 31. The loop for finding the name of the user that called Exim had a hardwired
7250 limit of 10; it now uses the value of finduser_retries, which is used for
7251 all other user lookups.
7253 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7256 33. Exim was neglecting to zero errno before one call of strtol() when
7257 expanding a string and expecting an integer value. On some systems this
7258 resulted in spurious "integer overflow" errors. Also, it was casting the
7259 result into an int without checking.
7261 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7262 did not work. The code looks as if it has *never* worked, though it appears
7263 to have been documented since at least release 1.62. I have made it work.
7265 35. The "timeout_DNS" error in retry rules, also documented since at least
7266 1.62, also never worked. As it isn't clear exactly what this means, and
7267 clearly it isn't a major issue, I have abolished the feature by treating it
7268 as "timeout", and writing a warning to the main and panic logs.
7270 36. The display of retry rules for -brt wasn't always showing the error code
7273 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7274 timeout_connect_A, timeout_connect_MX.
7276 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7277 to the empty sender.
7279 39. The daemon was not analysing the content of -oX till after it had closed
7280 stderr and disconnected from the controlling terminal. This meant that any
7281 syntax errors were only noted on the panic log, and the return code from
7282 the command was 0. By re-arranging the code a little, I've made the
7283 decoding happen first, so such errors now appear on stderr, and the return
7284 code is 1. However, the actual setting up of the sockets still happens in
7285 the disconnected process, so errors there are still only recorded on the
7288 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7289 connections (as happens on some IP stacks) was logged at start up time as
7290 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7291 it from "IPv6 and IPv4", which means that two separate sockets are being
7294 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7295 says whether AF_INET or AF_INET6 was passed as an argument.
7297 42. Exiwhat output was messed up when time zones were included in log
7300 43. Exiwhat now gives more information about the daemon's listening ports,
7301 and whether -tls-on-connect was used.
7303 44. The "port" option of the smtp transport is now expanded.
7305 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7306 silently ignored. Now an error message is written to the main and panic
7309 46. There's a new ACL modifier called "logwrite" which writes to a log file
7310 as soon as it is encountered.
7312 47. Added $local_user_uid and $local_user_gid at routing time.
7314 48. Exim crashed when trying to verify a sender address that was being
7317 49. Exim was recognizing only a space character after ".include". It now also
7318 recognizes a tab character.
7320 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7321 extracting the relevant information from the specification. The man page no
7322 longer contains scrambled data for the -d option, and I've added a section
7323 at the front about calling Exim under different names.
7325 51. Added "extra_headers" argument to the "mail" command in filter files.
7327 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7330 53. Installed eximstats 1.29.
7332 54. Added transport_filter_timeout as a generic transport option.
7334 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7335 Cc: header lines. This was required by RFC 822, but it not required by RFC
7338 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7339 incoming messages that did not have them. Now it does so only if the
7340 message originates locally, that is, if there is no associated remote host
7341 address. When Resent- header lines are present, this applies to the Resent-
7342 lines rather than the non-Resent- lines.
7344 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7345 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7347 58. Messages for configuration errors now include the name of the main
7348 configuration file - useful now that there may be more than one file in a
7349 list (.included file names were always shown).
7351 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7352 for those rare installations that do not start the daemon as root or run it
7353 setuid root. I've cut out the call to initgroups() if the daemon is not
7356 60. The Exim user and group can now be bound into the binary as text strings
7357 that are looked up at the start of Exim's processing.
7359 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7361 62. Added $mailstore_basename variable.
7363 63. Installed patch to sieve.c from Michael Haardt.
7365 64. When Exim failed to open the panic log after failing to open the main log,
7366 the original message it was trying to log was written to stderr and debug
7367 output, but if they were not available (the usual case in production), it
7368 was lost. Now it is written to syslog before the two lines that record the
7369 failures to open the logs.
7371 65. Users' Exim filters run in subprocesses under the user's uid. It is
7372 possible for a "deliver" command or an alias in a "personal" command to
7373 provoke an address rewrite. If logging of address rewriting is configured,
7374 this fails because the process is not running as root or exim. There may be
7375 a better way of dealing with this, but for the moment (because 4.30 needs
7376 to be released), I have disabled address rewrite logging when running a
7377 filter in a non-root, non-exim process.
7383 1. The buildconfig auxiliary program wasn't quoting the value set for
7384 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7385 not defined. This bug was masked in 4.22 by the effect that was fixed in
7388 2. Some messages that were rejected after a message id was allocated were
7389 shown as "incomplete" by exigrep. It no longer does this for messages that
7390 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7392 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7393 have allow_domain_literals set, the ID did not get logged in the <= line.
7394 Domain literals are now always recognized in Message-ID: header lines.
7396 4. The first argument for a ${extract expansion item is the key name or field
7397 number. Leading and trailing spaces in this item were not being ignored,
7398 causing some misleading effects.
7400 5. When deliver_drop_privilege was set, single queue runner processes started
7401 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7402 same command in the process it spins off) were not dropping privilege.
7404 6. When the daemon running as "exim" started a queue runner, it always
7405 re-executed Exim in the spun-off process. This is a waste of effort when
7406 deliver_drop_privilege is set. The new process now just calls the
7407 queue-runner function directly.
7413 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7416 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7417 ignored. Though the use of -oP was forcing the writing of a pid file, it
7418 was always written to the default place.
7420 3. If the message "no IP address found for host xxxx" is generated during
7421 incoming verification, it is now followed by identification of the incoming
7422 connection (so you can more easily find what provoked it).
7424 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7426 5. Added some features to "harden" Exim a bit more against certain attacks:
7428 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7429 be put in Local/Makefile. This is like the never_users runtime option,
7430 but it cannot be overridden. The default setting is "root".
7432 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7433 prefix string with which any file named in a -C command line option
7436 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7437 is retained for -C and -D only if the caller of Exim is root. Without
7438 it, the exim user may also use -C and -D and retain privilege.
7440 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7441 command line option is disabled.
7443 6. Macro names set by the -D option must start with an upper case letter, just
7444 like macro names defined in the configuration file.
7446 7. Added "dereference=" facility to LDAP.
7448 8. Two instances of the typo "uknown" in the source files are fixed.
7450 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7451 the Configure-Makefile script screwed up while processing it.
7453 10. Incorporated PCRE 4.4.
7455 11. The SMTP synchronization check was not operating right at the start of an
7456 SMTP session. For example, it could not catch a HELO sent before the client
7457 waited for the greeting. There is now a check for outstanding input at the
7458 point when the greeting is written. Because of the duplex, asynchronous
7459 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7460 way, but not yet received, when the check is performed.
7462 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7463 on TCP/IP sockets, because this apparently causes some broken clients to
7466 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7467 unchanged) from the Cygwin maintainer.
7469 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7470 was supported instead of testing for mbx. Effectively a typo.
7472 15. The spa authenticator server code was not checking that the input it
7473 received was valid base64.
7475 16. The debug output line for the "set" modifier in ACLs was not showing the
7476 name of the variable that was being set.
7478 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7480 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7481 lookup on its own. Something else was needed to trigger the lookup. For
7482 example, a match in host_lookup or the need for a host name in a host list.
7483 Now, if $sender_host_name is referenced and the host name has not yet been
7484 looked up, a lookup is performed. If the lookup fails, the variable remains
7485 empty, and $host_lookup_failed is set to "1".
7487 19. Added "eqi" as a case-independent comparison operator.
7489 20. The saslauthd authentication condition could segfault if neither service
7490 nor realm was specified.
7492 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7493 error message that was logged was misleading, and incoming SMTP
7494 connections were dropped. The message is now more accurate, and temporary
7495 errors are given to SMTP connections.
7497 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7498 (except RSET) with a 421 error, until QUIT is received. However, it was
7499 failing to send a response to QUIT.
7501 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7502 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7503 ACL is now run after the helo_try_verify_hosts code.
7505 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7506 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7507 case-independent, so other case variants are also recognized). Apparently
7508 some systems use these upper case variants.
7510 25. If more than two messages were waiting for the same host, and a transport
7511 filter was specified for the transport, Exim sent two messages over the
7512 same TCP/IP connection, and then failed with "socket operation on non-
7513 socket" when it tried to send the third.
7515 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7517 27. The extern definition of crypt16() in expand.c was not being excluded when
7518 the OS had its own crypt16() function.
7520 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7521 as a preferred synonym for return_size_limit, both as an option and as an
7524 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7526 30. Changed the default configuration ACL to relax the local part checking rule
7527 for addresses that are not in any local domains. For these addresses,
7528 slashes and pipe symbols are allowed within local parts, but the sequence
7529 /../ is explicitly forbidden.
7531 31. SPA server authentication was not clearing the challenge buffer before
7534 32. log_message in a "warn" ACL statement was writing to the reject log as
7535 well as to the main log, which contradicts the documentation and doesn't
7536 seem right (because no rejection is happening). So I have stopped it.
7538 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7539 However, I am unable to do any testing of this.
7541 34. Fixed an infelicity in the appendfile transport. When checking directories
7542 for a mailbox, to see if any needed to be created, it was accidentally
7543 using path names with one or more superfluous leading slashes; tracing
7544 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7546 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7547 contents are added to the log line that is written for every discarded
7548 recipient. (Previously a log_message setting was ignored.)
7550 36. The ${quote: operator now quotes the string if it is empty.
7552 37. The install script runs exim in order to find its version number. If for
7553 some reason other than non-existence or emptiness, which it checks, it
7554 could not run './exim', it was installing it with an empty version number,
7555 i.e. as "exim-". This error state is now caught, and the installation is
7558 38. An argument was missing from the function that creates an error message
7559 when Exim fails to connect to the socket for saslauthd authentication.
7560 This could cause Exim to crash, or give a corrupted message.
7562 39. Added isip, isip4, and isip6 to ${if conditions.
7564 40. The ACL variables $acl_xx are now saved with the message, and can be
7565 accessed later in routers, transports, and filters.
7567 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7568 strings in the file are not string-expanded.
7570 42. If a MAIL command specified a SIZE value that was too large to fit into an
7571 int variable, the check against message_size_limit failed. Such values are
7572 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7573 day this will have to be increased, but I don't think I want to be around
7574 when emails are that large.
7581 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7582 iconv() is not standard in FreeBSD.
7584 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7585 IPv6 enabled. The observed symptom was a segmentation fault on return from
7586 the function os_common_find_running_interfaces() in src/os.c.
7588 3. In the check_special_case() function in daemon.c I had used "errno" as an
7589 argument name, which causes warnings on some systems. This was basically a
7590 typo, since it was named "eno" in the comments!
7592 4. The code that waits for the clock to tick (at a resolution of some fraction
7593 of a second) so as to ensure message-id uniqueness was always waiting for
7594 at least one whole tick, when it could have waited for less. [This is
7595 almost certainly not relevant at current processor speeds, where it is
7596 unlikely to ever wait at all. But we try to future-proof.]
7598 5. The function that sleeps for a time interval that includes fractions of a
7599 second contained a race. It did not block SIGALRM between setting the
7600 timer, and suspending (a couple of lines later). If the interval was short
7601 and the sigsuspend() was delayed until after it had expired, the suspension
7602 never ended. On busy systems this could lead to processes getting stuck for
7605 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7606 process, before it forks any delivery processes. The open lookup caching
7607 mechanism meant that the open file or database connection was passed into
7608 the delivery process. The problem was that delivery processes always tidy
7609 up cached lookup data. This could cause a problem for the next delivery
7610 process started by the queue runner, because the external queue runner
7611 process does not know about the closure. So the next delivery process
7612 still has data in the lookup cache. In the case of a file lookup, there was
7613 no problem because closing a file descriptor in a subprocess doesn't affect
7614 the parent. However, if the lookup was caching a connection to a database,
7615 the connection was closed, and the second delivery process was likely to
7616 see errors such as "PGSQL: query failed: server closed the connection
7617 unexpectedly". The problem has been fixed by closing all cached lookups
7618 in a queue runner before running a delivery process.
7620 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7621 seem to have the "const" qualifier which it has on other OS. I've
7624 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7625 *of the same type* (client or server) with the same public name that an
7626 error should be diagnosed.
7628 9. When Exim looked up a host name for an IP address, but failed to find the
7629 original IP address when looking up the host name (a safety check), it
7630 output the message "<ip address> does not match any IP for NULL", which was
7631 confusing, to say the least. The bug was that the host name should have
7632 appeared instead of "NULL".
7634 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7635 user that is built into the binary, and the -C or -D options is used, root
7636 privilege is dropped before the configuration file is read. In addition,
7637 logging is switched to stderr instead of the normal log files. If the
7638 configuration then re-defines the Exim user, the unprivileged environment
7639 is probably not what is expected, so Exim logs a panic warning message (but
7642 However, if deliver_drop_privilege is set, the unprivileged state may well
7643 be exactly what is intended, so the warning has been cut out in that case,
7644 and Exim is allowed to try to write to its normal log files.
7650 1. smtp_return_error_details was not giving details for temporary sender
7651 or receiver verification errors.
7653 2. Diagnose a configuration error if two authenticators have the same public
7656 3. Exim used not to create the message log file for a message until the first
7657 delivery attempt. This could be confusing when incoming messages were held
7658 for policy or load reasons. The message log file is now created at the time
7659 the message is received, and an initial "Received" line is written to it.
7661 4. The automatically generated man page for command line options had a minor
7662 bug that caused no ill effects; however, a more serious problem was that
7663 the procedure for building the man page automatically didn't always
7664 operate. Consequently, release 4.20 contains an out-of-date version. This
7665 shouldn't happen again.
7667 5. When building Exim with embedded Perl support, the script that builds the
7668 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7669 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7671 6. The freeze_tell option was not being used for messages that were frozen on
7672 arrival, either by an ACL or by local_scan().
7674 7. Added the smtp_incomplete_transaction log selector.
7676 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7677 was accepting AUTH without a new EHLO.
7679 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7680 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7683 10. Logging of TCP/IP connections (when configured) now happens in the main
7684 daemon process instead of the child process, so that the TCP/IP connection
7685 count is more accurate (but it can never be perfect).
7687 11. The use of "drop" in a nested ACL was not being handled correctly in the
7688 outer ACL. Now, if condition failure induced by the nested "drop" causes
7689 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7690 or "require"), the connection is dropped.
7692 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7693 that yield "discard" can now be used with an "accept" or a "discard" verb,
7694 but an error is generated for any others (because I can't see a useful way
7695 to define what should happen).
7697 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7698 are now processed in the same way as lines in the Exim configuration file.
7699 In particular, continuation lines are supported.
7701 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7703 15. Added -ti meaning -t -i.
7705 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7706 domains, and warn by logging if others are found.
7708 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7709 returned from GIFCONF when Exim is trying to find the list of interfaces on
7710 a host. The code in os.c has been modified to copy each ifreq to an aligned
7711 structure in all cases.
7713 Also, in some cases, the returned ifreq's were being copied to a 'struct
7714 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7715 means the last couple of bytes of an IPv6 address could be chopped if the
7716 ifreq contained only a normal sockaddr (14 bytes storage).
7718 18. Named domain lists were not supported in the hosts_treat_as_local option.
7719 An entry such as +xxxx was not recognized, and was treated as a literal
7722 19. Ensure that header lines added by a DATA ACL are included in the reject log
7723 if the ACL subsequently rejects the message.
7725 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7726 MD5 (which is deprecated).
7728 21. When testing a filter file using -bf, Exim was writing a message when it
7729 took the sender from a "From " line in the message, but it was not doing so
7730 when it took $return_path from a Return-Path: header line. It now does.
7732 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7733 with a valid header line field name (a series of printing characters
7734 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7736 23. Changed "disc" in the source to "disk" to conform to the documentation and
7737 the book and for uniformity.
7739 24. Ignore Sendmail's -Ooption=value command line item.
7741 25. When execve() failed while trying to run a command in a pipe transport,
7742 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7743 could be confused with a return value of 69 from the command itself. This
7744 has been changed to 127, the value the shell returns if it is asked to run
7745 a non-existent command. The wording for the related log line suggests a
7746 non-existent command as the problem.
7748 26. If received_header_text expands to an empty string, do not add a Received:
7749 header line to the message. (Well, it adds a token one on the spool, but
7750 marks it "old" so that it doesn't get used or transmitted.)
7752 27. Installed eximstats 1.28 (addition of -nt option).
7754 28. There was no check for failure on the call to getsockname() in the daemon
7755 code. This can fail if there is a shortage of resources on the system, with
7756 ENOMEM, for example. A temporary error is now given on failure.
7758 29. Contrary to the C standard, it seems that in some environments, the
7759 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7760 program. Exim now does this explicitly; it affects the formatting of
7761 timestamps using strftime().
7763 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7764 complaints. I've now initialized all the variables, to avoid this.
7766 32. Header lines added by a system filter were not being "seen" during
7767 transport-time rewrites.
7769 33. The info_callback() function passed to OpenSSL is set up with type void
7770 (*)(SSL *, int, int), as described somewhere. However, when calling the
7771 function (actually a macro) that sets it up, the type void(*)() is
7772 expected. I've put in a cast to prevent warnings from picky compilers.
7774 34. If a DNS black list lookup found a CNAME record, but there were no A
7775 records associated with the domain it pointed at, Exim crashed.
7777 35. If a DNS black list lookup returned more than one A record, Exim ignored
7778 all but the first. It now scans all returned addresses if a particular IP
7779 value is being sought. In this situation, the contents of the
7780 $dnslist_value variable are a list of all the addresses, separated by a
7783 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7784 to accept a host name and all its aliases if the forward lookup for any of
7785 them yielded the IP address of the incoming connection. Now it accepts only
7786 those names whose forward lookup yields the correct IP address. Any other
7787 names are discarded. This closes a loophole whereby a rogue DNS
7788 administrator could create reverse DNS records to break through a
7789 wildcarded host restriction in an ACL.
7791 37. If a user filter or a system filter that ran in a subprocess used any of
7792 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7793 the wrong values were passed to the pipe command ($thisaddress had the
7794 value of $0, $0 had the value of $1, etc). This bug was introduced by
7795 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7797 38. Improved the line breaking for long SMTP error messages from ACLs.
7798 Previously, if there was no break point between 40 and 75 characters, Exim
7799 left the rest of the message alone. Two changes have been made: (a) I've
7800 reduced the minimum length to 35 characters; (b) if it can't find a break
7801 point between 35 and 75 characters, it looks ahead and uses the first one
7802 that it finds. This may give the occasional overlong line, but at least the
7803 remaining text gets split now.
7805 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7806 file descriptors might be low, and that setting 1000 would always raise it.
7807 It turns out that in some environments, the limit is already over 1000 and
7808 that lowering it causes trouble. So now Exim takes care not to decrease it.
7810 40. When delivering a message, the value of $return_path is set to $sender_
7811 address at the start of routing (routers may change the value). By an
7812 oversight, this default was not being set up when an address was tested by
7813 -bt or -bv, which affected the outcome if any router or filter referred to
7816 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7817 log only when "message" or "log_message" are set. However, if one of the
7818 conditions was an address verification, or a call to a nested ACL, the
7819 messages generated by the underlying test were being passed through. This
7820 no longer happens. The underlying message is available in $acl_verify_
7821 message for both "message" and "log_message" expansions, so it can be
7822 passed through if needed.
7824 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7825 new expansion $bh_ to give the encoded byte string without charset
7826 translation. Translation happens only if iconv() is available; HAVE_ICONV
7827 indicates this at build time. HEADERS_CHARSET gives the charset to
7828 translate to; headers_charset can change it in the configuration, and
7829 "headers charset" can change it in an individual filter file.
7831 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7832 that creates RFC 2047 encoded "words" labels them as that charset instead
7833 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7834 expansion operator; (ii) when Exim creates a From: line for a local
7835 message; (iii) when a header line is rewritten to include a "phrase" part.
7837 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7838 buggy, causing it to skip the first lines of messages whose message ID
7839 ended in 'D'. This would not have bitten before Exim release 4.14, because
7840 message IDs were unlikely to end in 'D' before then. The effect was to have
7841 incorrect size information for certain domains.
7843 45. #include "config.h" was missing at the start of the crypt16.c module. This
7844 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7847 46. If there was a timeout during a "random" callout check, Exim treated it as
7848 a failure of the random address, and carried on sending RSET and the real
7849 address. If the delay was just some slowness somewhere, the response to the
7850 original RCPT would be taken as a response to RSET and so on, causing
7851 mayhem of various kinds.
7853 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7854 when I implemented it. It didn't allow for the fact that some option values
7855 may legitimately be negative (e.g. size_addition), and it didn't even do
7856 the right test for positive values.
7858 48. Domain names in DNS records are case-independent. Exim always looks them up
7859 in lower case. Some resolvers return domain names in exactly the case they
7860 appear in the zone file, that is, they may contain uppercase letters. Not
7861 all resolvers do this - some return always lower case. Exim was treating a
7862 change of case by a resolver as a change of domain, similar to a widening
7863 of a domain abbreviation. This triggered its re-routing code and so it was
7864 trying to route what was effectively the same domain again. This normally
7865 caused routing to fail (because the router wouldn't handle the domain
7866 twice). Now Exim checks for this case specially, and just changes the
7867 casing of the domain that it ultimately uses when it transmits the message
7870 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7873 50. If a filter generated a file delivery with a non-absolute name (possible if
7874 no home directory exists for the router), the forbid_file option was not
7877 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7878 the existing equality matching.
7880 52. Exim was using ints instead of ino_t variables in some places where it was
7881 dealing with inode numbers.
7883 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7884 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7885 called TMPDIR, and if it finds it is different, it changes its value.
7887 54. The smtp_printf() function is now made available to local_scan() so
7888 additional output lines can be written before returning. There is also an
7889 smtp_fflush() function to enable the detection of a dropped connection.
7890 The variables smtp_input and smtp_batched_input are exported to
7893 55. Changed the default runtime configuration: the message "Unknown user"
7894 has been removed from the ACL, and instead placed on the localuser router,
7895 using the cannot_route_message feature. This means that any verification
7896 failures that generate their own messages won't get overridden. Similarly,
7897 the "Unrouteable address" message that was in the ACL for unverifiable
7898 relay addresses has also been removed.
7900 56. Added hosts_avoid_esmtp to the smtp transport.
7902 57. The exicyclog script was not checking for the esoteric option
7903 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7904 will work only if exicyclog is run under the appropriate euid.
7906 58. Following a discussion on the list, the rules by which Exim recognises line
7907 endings on incoming messages have been changed. The -dropcr and drop_cr
7908 options are now no-ops, retained only for backwards compatibility. The
7909 following line terminators are recognized: LF CRLF CR. However, special
7910 processing applies to CR:
7912 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7913 nor a local message in the state where . is a terminator.
7915 (ii) If a bare CR is encountered in a header line, an extra space is added
7916 after the line terminator so as not to end the header. The reasoning
7917 behind this is that bare CRs in header lines are most likely either
7918 to be mistakes, or people trying to play silly games.
7920 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7921 was being incorrectly given as 18 bytes larger than it should have been.
7922 This is a VOB (very old bug).
7924 60. This may never have affected anything current, but just in case it has:
7925 When the local host is found other than at the start of a list of hosts,
7926 the local host, those with the same MX, and any that follow, are discarded.
7927 When the list in question was part of a longer list of hosts, the following
7928 hosts (not currently being processed) were also being discarded. This no
7929 longer happens. I'm not sure if this situation could ever has previously
7932 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7935 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7936 header. This is something that is recommended in a new Internet Draft, and
7937 is something that is documented as being done by Sendmail. There are two
7938 possible values. For messages generated by the autoreply transport, Exim
7941 Auto-Submitted: auto-replied
7943 whereas for all other generated messages (e.g. bounces) it adds
7945 Auto-Submitted: auto-generated
7947 63. The "personal" condition in filters now includes a test for the
7948 Auto-Submitted: header. If it contains the string "auto-" the message it
7949 not considered personal.
7951 64. Added rcpt_include_affixes as a generic transport option.
7953 65. Added queue_only_override (default true).
7955 66. Added the syslog_duplication option.
7957 67. If what should have been the first header line of a message consisted of
7958 a space followed by a colon, Exim was mis-interpreting it as a header line.
7959 It isn't of course - it is syntactically invalid and should therefore be
7960 treated as the start of the message body. The misbehaviour could have
7961 caused a number of strange effects, including loss of data in subsequent
7962 header lines, and spool format errors.
7964 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7965 client host had authenticated. This control can now be exercised by an ACL
7966 for more flexibility.
7968 69. By default, callouts do not happen when testing with -bh. There is now a
7969 variant, -bhc, which does actually run the callout code, including
7970 consulting and updating the callout cache.
7972 70. Added support for saslauthd authentication, courtesy of Alexander
7975 71. If statvfs() failed on the spool or log directories while checking their
7976 size for availability, Exim confusingly gave the error "space shortage".
7977 Furthermore, in debugging mode it crashed with a floating point exception.
7978 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7979 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7980 serious problem, Exim now writes to the main and panic logs when this
7981 happens, with details of the failure. It then refuses to accept the
7982 incoming message, giving the message "spool directory problem" or "log
7983 directory problem" with a 421 code for SMTP messages.
7985 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7986 0, 1, and 2 exist, because some OS complain for execs without them (see
7987 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7988 descriptors. However, the code omitted to check that the open succeeded,
7989 causing mysterious errors if for some reason the permissions on /dev/null
7990 got screwed. Now Exim writes a message to the main and panic logs, and
7991 bombs out if it can't open /dev/null.
7993 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7994 interact so that it is all more flexible. It is supposed to remain
7995 backwards compatible. Also added extra_local_interfaces.
7997 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7998 to bomb out with an assertion failure - to the client this appears as a
7999 connection drop. This problem occurs in the part of the code that was taken
8000 from the Samba project. Fortunately, the assertion is in a very simple
8001 function, so I have fixed this by reproducing the function inline in the
8002 one place where it is called, and arranging for authentication to fail
8003 instead of killing the process with assert().
8005 75. The SPA client code was not working when the server requested OEM rather
8006 than Unicode encoding.
8008 76. Added code to make require_files with a specific uid setting more usable in
8009 the case where statting the file as root fails - usually a non-root-mounted
8010 NFS file system. When this happens and the failure is EACCES, Exim now
8011 forks a subprocess and does the per-uid checking as the relevant uid.
8013 77. Added process_log_path.
8015 78. If log_file_path was not explicitly set, a setting of check_log_space or
8016 check_log_inodes was ignored.
8018 79. If a space check for the spool or log partitions fails, the incident is now
8019 logged. Of course, in the latter case the data may get lost...
8021 80. Added the %p formatting code to string_format() so that it can be used to
8022 print addresses in debug_print(). Adjusted all the address printing in the
8023 debugging in store.c to use %p rather than %d.
8025 81. There was a concern that a line of code in smtp_in.c could overflow a
8026 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8027 initially expressed, the concern was not well-founded, because trailing
8028 spaces are removed early. However, if the trailing spaces were followed by
8029 a NULL, they did not get removed, so the overflow was possible. Two fixes
8032 (a) I re-wrote the offending code in a cleaner fashion.
8033 (b) If an incoming SMTP command contains a NULL character, it is rejected
8036 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8037 runs initgroups(), so that if the Exim user is in any additional groups,
8038 they will be used during message reception.
8044 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob