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 Avoid costly startup code when not strictly needed. This reduces time
10 for some exim process initialisations. It does mean that the logging
11 of TLS configuration problems is only done for the daemon startup.
13 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
15 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
16 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
19 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
20 DKIM signing is being done. Previously a transport_filter would always
21 disable CHUNKING, falling back to traditional DATA.
23 JH/05 Regard command-line receipients as tainted.
25 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
27 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
28 PAM library frees one of the arguments given to it, despite the
29 documentation. Therefore a plain malloc must be used.
31 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
32 on-stack buffers were used, resulting in a taint trap when DSN information
33 copied from a received message was written into the buffer.
35 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
36 the ordering of its ARC headers. This caused a crash.
38 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
39 a new record was being constructed with information from the peer, a trap
42 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
43 installation would get error messages from DMARC verify, when it hit the
44 nonexistent file indicated by the default. Distros wanting DMARC enabled
45 should both provide the file and set the option.
46 Also enforce no DMARC verification for command-line sourced messages.
48 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
49 could, depending on the platform, hang at the STARTTLS response.
51 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
52 message on a connection. Previously if one message had ARC headers and
53 the following one did not, a crash could result when adding an
54 Authentication-Results: header.
56 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
57 Exim main code and Exim-related utities. The introduction of taint
58 tracking also did many adjustments to string handling. Since then, eximon
59 frequently terminated with an assert failure.
61 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
62 check for 452 responses. This slightly helps the inefficieny of doing
63 a large alias-expansion into a recipient-limited target. The max_rcpt
64 transport option still applies (and at the current default, will override
65 the new feature). The check is done for either cause of synch, and forces
66 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
67 connection. The new facility is not tunable at this time.
69 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
70 library live data was being used, so the results became garbage. Make
71 copies while it is still usable.
73 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
74 on delivery deferred (==) and failed (**) lines (if a delivery was
75 attemtped). Previously it was only on completion (=>) lines.
77 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
78 for the expansion of the server_scram_iter and server_scram_salt options.
80 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
81 are now specifically given a NO_DATA response without hitting the system
82 resolver. The library goes on to do the now-standard TXT lookup.
83 Use of dnsdb lookups is not affected.
85 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
86 only retrieve the errormessage once. Previously two calls to dlerror()
87 were used, and the second one (for mainlog/paniclog) retrieved null
90 JH/20 Taint checking: disallow use of tainted data for
91 - the appendfile transport file and directory options
92 - the pipe transport command
93 - the autoreply transport file, log and once options
94 - file names used by the redirect router (including filter files)
96 - paths used by single-key lookups
97 Previously this was permitted.
99 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
100 adjusted the size of a major service buffer; this failed because the
101 buffer was in use at the time. Change to a compile-time increase in the
102 buffer size, when this authenticator is compiled into exim.
104 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
105 previous fast-mode was untenable in the face of glibs using mmap to
106 support larger malloc requests.
108 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
109 New values supported, if defined on system where compiled:
110 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
111 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
113 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
114 running a limited number of proceses in parallel, a benefit is gained. The
115 amount varies with the platform hardware and load. The use of the option
116 queue_run_in_order means we cannot do this, as ordering becomes
119 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
120 had introduced a string-copy (for ensuring NUL-termination) which was not
121 appropriate for that case, which can include embedded NUL bytes in the
122 block of data. Investigation showed the copy to actually be needless, the
123 data being length-specified.
125 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
126 done during a receiving connection, and both used TLS, global info was
127 used rather than per-connection info for tracking the state of data
128 queued for transmission. This could result in a connection hang.
130 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
131 Previously, when delivering serveral messages down a single connection
132 only the first would provide a SIZE. This was due to the size information
133 not being properly tracked.
135 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
136 TAI (at 37 seconds currently), pretend to be in UTC for time-related
137 expansion and logging. Previously, spurious values such as a future
138 minute could be seen.
140 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
141 it could crash from a null-deref. This could also affect the
142 ${addresses: } operator and ${readsock } item.
144 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
145 message following a mime one, the variable was not reset.
147 JH/30 When an pipelined-connect fails at the first response, assume incorrect
148 cached capability (perhaps the peer reneged?) and immediately retry in
151 JH/31 Fix spurious detection of timeout while writing to transport filter.
153 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
154 an attempt to copy the string was made before checking it.
156 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
157 taint of the lookup key was maintained; we now regard the presence in the
158 filesystem as sufficient validation.
160 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
161 argument is supplied.
163 JH/35 The dsearch lookup now requires that the directory is an absolute path.
164 Previously this was not checked, and nonempty relative paths made an
165 access under Exim's current working directory.
171 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
172 side, to decode a TLS 1.3 packet capture.
174 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
175 Previously the default library behaviour applied, sending two, each in
178 JH/03 Debug output for ACL now gives the config file name and line number for
181 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
183 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
185 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
186 buffer overrun for (non-chunking) other transports.
188 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
189 TLS1.3, means that a server rejecting a client certificate is not visible
190 to the client until the first read of encrypted data (typically the
191 response to EHLO). Add detection for that case and treat it as a failed
192 TLS connection attempt, so that the normal retry-in-clear can work (if
193 suitably configured).
195 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
196 and/or domain. Found and fixed by Jason Betts.
198 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
199 configuration). If a CNAME target was not a wellformed name pattern, a
202 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
203 the OS reports them interleaved with other addresses.
205 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
206 used both for input and for a verify callout, both encrypted, SMTP
207 responses being sent by the server could be lost. This resulted in
208 dropped connections and sometimes bounces generated by a peer sending
211 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
212 string. Previously it was possible to cause undefined behaviour in a
213 library routine (usually a crash). Found by "zerons".
215 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
218 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
219 API was removed, so update to use the newer ones.
221 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
222 any timeout set, is taking a long time. Previously we would hang on to a
223 rotated logfile "forever" if the input was arriving with long gaps
224 (a previous attempt to fix addressed lack, for a long time, of initial
227 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
228 shared (NFS) environment. The length of the tempfile name is now
229 4 + 16 ("hdr.$message_exim_id") which might break on file
230 systems which restrict the file name length to lower values.
233 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
234 shared (NFS) environment.
236 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
237 did for all versions <4.90). Notably -M, -m, --invert, -I may be
240 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
241 on some platforms for bit 31.
243 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
244 to changes apparently associated with TLS1.3 handling some of the APIs
245 previously used were either nonfunctional or inappropriate. Strings
246 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
247 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
248 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
249 This affects log line X= elements, the $tls_{in,out}_cipher variables,
250 and the use of specific cipher names in the encrypted= ACL condition.
252 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
254 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
255 verification result was not updated unless hosts_require_ocsp applied.
257 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
258 queue_list_requires_admin set to false, non-admin users were denied the
261 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
262 directory-of-certs mode. Previously they were advertised despite the
265 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
266 A single TCP connection by a client will now hold a TLS connection open
267 for multiple message deliveries, by default. Previoud the default was to
270 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
271 default. If built with the facility, DANE will be used. The facility
272 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
274 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
275 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
276 must be defined and you must still, unless you define DISABLE_TLS, manage
277 the the include-dir and library-file requirements that go with that
278 choice. Non-TLS builds are still supported.
280 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
283 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
284 default. If the platform supports and has the facility enabled, it will
285 be requested on all coneections.
287 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
288 controlled by the build-time option SUPPORT_PIPE_CONNECT.
290 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
292 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
293 success-DSN messages. Previously the From: header was always the default
294 one for these; the option was ignored.
296 JH/28 Fix the timeout on smtp response to apply to the whole response.
297 Previously it was reset for every read, so a teergrubing peer sending
298 single bytes within the time limit could extend the connection for a
299 long time. Credit to Qualsys Security Advisory Team for the discovery.
301 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
302 delivery address, which leaked information of the results of local
303 forwarding. Change to the original envelope recipient address, per
306 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
307 requested. Previously not bounce was generated and a log entry of
308 error ignored was made.
310 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
312 JH/32 Introduce a general tainting mechanism for values read from the input
313 channel, and values derived from them. Refuse to expand any tainted
314 values, to catch one form of exploit.
316 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
317 was unused and the unexpanded text used for the test. Found and
318 fixed by Ruben Jenster.
320 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
321 an attempt to use a TLS library read routine dereffed a nul pointer,
324 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
325 them in our smtp response.
327 JH/36 Have the general router option retry_use_local_part default to true when
328 any of the restrictive preconditions are set (to anything). Previously it
329 was only for check_local user. The change removes one item of manual
330 configuration which is required for proper retries when a remote router
331 handles a subset of addresses for a domain.
333 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
334 link count into consideration.
336 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
337 caused the extension of big_buffer, the following lines were ignored.
339 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
340 accordance with RFC 2308. Previously there was no expiry, so a longlived
341 receive process (eg. due to ACL delays) versus a short SOA value could
344 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
346 JH/39 Promote DMARC support to mainline.
348 JH/40 Bug 2452: Add a References: header to DSNs.
350 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
351 parameters. The relevant library call is documented as "Deprecated: This
352 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
353 3.6.0, DH parameters are negotiated following RFC7919."
355 HS/06 Change the default of dnssec_request_domains to "*"
357 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
358 carried on and emitted a BDAT command, even when PIPELINING was not
361 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
362 buffer was used for the filename, resulting in a trap when tainted
363 arguments (eg. $domain) were used.
365 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
366 recommended to avoid a possible server-load attack. The feature can be
367 re-enabled via the openssl_options main cofiguration option.
369 JH/45 local_scan API: documented the current smtp_printf() call. This changed
370 for version 4.90 - adding a "more data" boolean to the arguments.
371 Bumped the ABI version number also, this having been missed previously;
372 release versions 4.90 to 4.92.3 inclusive were effectively broken in
373 respect of usage of smtp_printf() by either local_scan code or libraries
374 accessed via the ${dlfunc } expansion item. Both will need coding
375 adjustment for any calls to smtp_printf() to match the new function
376 signature; a FALSE value for the new argument is always safe.
378 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
379 the file-offset (which the Linux syscall does, and exim expects); this
380 resulted in an indefinite loop.
382 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
383 to do ARC verification. The Authentication-Results: header line added
384 by the configuration then had no ARC item.
390 JH/01 Remove code calling the customisable local_scan function, unless a new
391 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
393 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
394 non-signal-safe functions being used.
396 JH/03 Bug 2269: When presented with a received message having a stupidly large
397 number of DKIM-Signature headers, disable DKIM verification to avoid
398 a resource-consumption attack. The limit is set at twenty.
400 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
401 report of oldest_pass in ${authres } in consequence, and separate out
402 some descriptions of reasons for verification fail.
404 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
405 files in the spool were present and unlocked. A queue-runner could spot
406 them, resulting in a duplicate delivery. Fix that by doing the unlock
407 after the unlink. Investigation by Tim Stewart. Take the opportunity to
408 add more error-checking on spoolfile handling while that code is being
411 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
412 No known attacks, no CVE, this is defensive hardening.
414 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
415 a queue-runner could start a delivery while other operations were ongoing.
416 Cutthrough delivery was a common victim, resulting in duplicate delivery.
417 Found and investigated by Tim Stewart. Fix by using the open message data
418 file handle rather than opening another, and not locally closing it (which
419 releases a lock) for that case, while creating the temporary .eml format
420 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
422 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
423 $sender_verify_failure/$recipient_verify_failure to "random".
425 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
428 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
429 Previously this would segfault.
431 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
434 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
435 like zero, since the resolver should be doing this for us, But we need one
436 as a CNAME but no MX presence gets the CNAME returned; we need to check
437 that doesn't point to an MX to declare it "no MX returned" rather than
438 "error, loop". A new main option is added so the older capability of
439 following some limited number of chain links is maintained.
441 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
443 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
444 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
445 now-deprecated earlier definitions used only specified the range up to TLS
446 1.2 (in the older-version library docs).
448 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
450 JH/15 Rework TLS client-side context management. Stop using a global, and
451 explicitly pass a context around. This enables future use of TLS for
452 connections to service-daemons (eg. malware scanning) while a client smtp
453 connection is using TLS; with cutthrough connections this is quite likely.
455 JH/16 Fix ARC verification to do AS checks in reverse order.
457 JH/17 Support a "tls" option on the ${readsocket } expansion item.
459 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
460 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
461 Previously the "utf8" would be re-prepended for every additional message.
463 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
464 Previously thery were accepted, resulting in issues when attempting to
465 forward messages to a non-supporting MTA.
467 PP/02 Let -n work with printing macros too, not just options.
469 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
470 one parent address was copied, and bogus data was used at delivery-logging
471 time. Either a crash (after delivery) or bogus log data could result.
472 Discovery and analysis by Tim Stewart.
474 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
475 Previously if the string ended mid-character, we did not insert the
476 promised '?' replacement.
478 PP/04 Documentation: current string operators work on bytes, not codepoints.
480 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
481 should pack well giving a smaller memory footprint so better caching and
482 therefore performance. Group the declarations where this can't be done so
483 that the byte-sized flag variables are not interspersed among pointer
484 variables, giving a better chance of good packing by the compiler.
486 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
487 non-null, to avoid issues with sites running BATV. Previously reports were
488 sent with an empty envelope sender so looked like bounces.
490 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
491 The ignore_error flag wasn't being returned from the filter subprocess so
492 was not set for later routers. Investigation and fix by Matthias Kurz.
494 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
495 and a msg:complete for the whole, when a message is manually removed using
496 -Mrm. Developement by Matthias Kurz, hacked on by JH.
498 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
499 a "Gnu special" function, asprintf() in the DB utility binary builds; I
500 hope that is portable enough.
502 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
503 requiring a known-CA anchor certificate; make it now rely entirely on the
504 TLSA as an anchor. Checking the name on the leaf cert against the name
505 on the A-record for the host is still done for TA (but not for EE mode).
507 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
508 included in delivery lines for non-proxied connections, when compiled with
509 SUPPORT_SOCKS and running with proxy logging enabled.
511 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
512 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
513 move the existing event to fire before the normal logging of message
514 failure so that custom logging is bracketed by normal logging.
516 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
517 msg:fail:internal event. Developement by Matthias Kurz.
519 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
520 far too small for todays use of crypto signatures stored there. Go all
521 the way to the max DNS message size of 64kB, even though this might be
522 overmuch for IOT constrained device use.
524 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
525 copy a string over itself. The library routine is documented as not
526 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
528 JH/32 For main options check_spool_space and check_inode_space, where the
529 platform supports 64b integers, support more than the previous 2^31 kB
530 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
531 the previous G, M, k.
533 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
534 $authenticated_fail_id variable on authentication failure. Previously
537 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
538 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
539 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
540 GNUTLS_SEC_PARAM_MEDIUM.
542 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
543 side. Previously we would continue as if no SNI had been received.
545 JH/36 Harden the handling of string-lists. When a list consisted of a sole
546 "<" character, which should be a list-separator specification, we walked
547 off past the nul-terimation.
549 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
550 causes) even when the retry time is not yet met. Previously they were
551 not, meaning that when (say) an account was over-quota and temp-rejecting,
552 and multiple senders' messages were queued, only one sender would get
553 notified on each configured delay_warning cycle.
555 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
557 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
558 cipher-suites, an error can be left on the stack even for a succeeding
559 accept; this results in impossible error messages when a later operation
562 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
563 return error codes indicating retry. Under TLS1.3 this becomes required.
565 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
566 it only wrote the new authenticators, resulting in a lack of tracking of
567 peer changes of ESMTP extensions until the next cache flush.
569 JH/41 Fix the loop reading a message header line to check for integer overflow,
570 and more-often against header_maxsize. Previously a crafted message could
571 induce a crash of the recive process; now the message is cleanly rejected.
573 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
574 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
580 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
581 When redis_servers is set to a list of > 1 element, and the Redis servers
582 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
583 case of MOVED into a DEFER case instead, thus moving the query onto the
584 next server in the list. For a cluster of N elements, all N servers must
585 be defined in redis_servers.
587 GF/02 Catch and remove uninitialized value warning in exiqsumm
588 Check for existence of @ARGV before looking at $ARGV[0]
590 JH/01 Replace the store_release() internal interface with store_newblock(),
591 which internalises the check required to safely use the old one, plus
592 the allocate and data copy operations duplicated in both (!) of the
593 extant use locations.
595 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
596 modifier. This matches the restriction on the commandline.
598 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
599 Previously only the last row was returned.
601 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
602 we assumed that tags in the header were well-formed, and parsed the
603 element content after inspecting only the first char of the tag.
604 Assumptions at that stage could crash the receive process on malformed
607 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
608 While running the DKIM ACL we operate on the Permanent memory pool so that
609 variables created with "set" persist to the DATA ACL. Also (at any time)
610 DNS lookups that fail create cache records using the Permanent pool. But
611 expansions release any allocations made on the current pool - so a dnsdb
612 lookup expansion done in the DKIM ACL releases the memory used for the
613 DNS negative-cache, and bad things result. Solution is to switch to the
614 Main pool for expansions.
615 While we're in that code, add checks on the DNS cache during store_reset,
616 active in the testsuite.
617 Problem spotted, and debugging aided, by Wolfgang Breyha.
619 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
620 When none of the hosts presented to a transport match an already-open
621 connection, close it and proceed with the list. Previously we would
622 queue the message. Spotted by Lena with Yahoo, probably involving
625 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
626 Previously a spurious "250 OK id=" response was appended to the proper
629 JH/08 The "support for" informational output now, which built with Content
630 Scanning support, has a line for the malware scanner interfaces compiled
631 in. Interface can be individually included or not at build time.
633 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
634 by the template makefile "src/EDITME". The "STREAM" support for an older
635 ClamAV interface method is removed.
637 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
638 rows affected is given instead).
640 JH/11 The runtime Berkeley DB library version is now additionally output by
641 "exim -d -bV". Previously only the compile-time version was shown.
643 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
644 SMTP connection. Previously, when one had more receipients than the
645 first, an abortive onward connection was made. Move to full support for
646 multiple onward connections in sequence, handling cutthrough connection
647 for all multi-message initiating connections.
649 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
650 routers. Previously, a multi-recipient message would fail to match the
651 onward-connection opened for the first recipient, and cause its closure.
653 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
654 a timeout on read on a GnuTLS initiating connection, resulting in the
655 initiating connection being dropped. This mattered most when the callout
656 was marked defer_ok. Fix to keep the two timeout-detection methods
659 JH/15 Relax results from ACL control request to enable cutthrough, in
660 unsupported situations, from error to silently (except under debug)
661 ignoring. This covers use with PRDR, frozen messages, queue-only and
664 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
666 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
667 metadata, resulting in a crash in free().
669 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
670 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
671 Broken also in d185889f4, with init system revamp.
673 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
674 Previously we abruptly closed the connection after reading a malware-
675 found indication; now we go on to read the "scan ok" response line,
678 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
679 ACL. Previously, a crash would result.
681 JH/19 Speed up macro lookups during configuration file read, by skipping non-
682 macro text after a replacement (previously it was only once per line) and
683 by skipping builtin macros when searching for an uppercase lead character.
685 JH/20 DANE support moved from Experimental to mainline. The Makefile control
686 for the build is renamed.
688 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
689 was allocated for every new TLS startup, meaning one per message. Fix
690 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
692 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
693 reported the original. Fix to report (as far as possible) the ACL
694 result replacing the original.
696 JH/23 Fix memory leak during multi-message connections using STARTTLS under
697 OpenSSL. Certificate information is loaded for every new TLS startup,
698 and the resources needed to be freed.
700 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
702 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
705 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
706 DATA response info to the (existing) per-recipient response info for
707 the "C=" log element. It can have useful tracking info from the
708 destination system. Patch from Simon Arlott.
710 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
711 length value. Previously this would segfault.
713 HS/02 Support Avast multiline protoocol, this allows passing flags to
714 newer versions of the scanner.
716 JH/28 Ensure that variables possibly set during message acceptance are marked
717 dead before release of memory in the daemon loop. This stops complaints
718 about them when the debug_store option is enabled. Discovered specifically
719 for sender_rate_period, but applies to a whole set of variables.
720 Do the same for the queue-runner and queue-list loops, for variables set
721 from spool message files. Do the same for the SMTP per-message loop, for
722 certain variables indirectly set in ACL operations.
724 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
725 as a multi-recipient message from a mailinglist manager). The coding had
726 an arbitrary cutoff number of characters while checking for more input;
727 enforced by writing a NUL into the buffer. This corrupted long / fast
728 input. The problem was exposed more widely when more pipelineing of SMTP
729 responses was introduced, and one Exim system was feeding another.
730 The symptom is log complaints of SMTP syntax error (NUL chars) on the
731 receiving system, and refused recipients seen by the sending system
732 (propating to people being dropped from mailing lists).
733 Discovered and pinpointed by David Carter.
735 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
736 replaced by the ${authresults } expansion.
738 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
740 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
741 allows proper process termination in container environments.
743 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
744 Previously the "final dot" had a newline after it; ensure it is CR,LF.
746 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
747 and "err_perm", deprecated since 4.83 when the RFC-defined words
748 "temperror" and "permerror" were introduced.
750 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
751 transport-filters or DKIM-signing. The restriction was lost in the
752 consolidation of verify-callout and delivery SMTP handling.
753 Extend the restriction to also cover ARC-signing.
755 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
756 in defer=pass mode supply a 450 to the initiator. Previously the message
759 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
760 tls_require_ciphers is used as before.
762 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
763 "pass_unscanned". Only tmpfails from the scanner are written to
764 the paniclog, as they may require admin intervention (permission
765 denied, license issues). Other scanner errors (like decompression
766 bombs) do not cause a paniclog entry.
768 JH/36 Fix reinitialisation of DKIM logging variable between messages.
769 Previously it was possible to log spurious information in receive log
772 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
773 triggered odd behaviour from Outlook Express clients.
775 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
778 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
779 since the IETF WG has not yet settled on that versus the original
780 "bare" representation.
782 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
783 Previously the millisecond value corrupted the output.
784 Fix also for syslog_pid=no and log_selector +pid, for which the pid
785 corrupted the output.
791 JH/01 Rework error string handling in TLS interface so that the caller in
792 more cases is responsible for logging. This permits library-sourced
793 string to be attached to addresses during delivery, and collapses
794 pairs of long lines into single ones.
796 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
797 during configuration. Wildcards are allowed and expanded.
799 JH/02 Rework error string handling in DKIM to pass more info back to callers.
800 This permits better logging.
802 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
803 do not close it down and have the child transport start it up again on
804 the passed-on TCP connection. Instead, proxy the child (and any
805 subsequent ones) for TLS via a unix-domain socket channel. Logging is
806 affected: the continued delivery log lines do not have any DNSSEC, TLS
807 Certificate or OCSP information. TLS cipher information is still logged.
809 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
810 identical IP addresses on different listening ports. Will also affect
813 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
814 add noisy ifdef guards to special-case this sillyness.
815 Patch from Bernd Kuhls.
817 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
818 than 255 are no longer allowed.
820 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
821 Disable the session-cache too, which might reduce our load. Since we
822 currrectly use a new context for every connection, both as server and
823 client, there is no benefit for these.
824 GnuTLS appears to not support tickets server-side by default (we don't
825 call gnutls_session_ticket_enable_server()) but client side is enabled
826 by default on recent versions (3.1.3 +) unless the PFS priority string
829 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
830 <https://reproducible-builds.org/specs/source-date-epoch/>.
832 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
833 the check for any unsuccessful recipients did not notice the limit, and
834 erroneously found still-pending ones.
836 JH/08 Pipeline CHUNKING command and data together, on kernels that support
837 MSG_MORE. Only in-clear (not on TLS connections).
839 JH/09 Avoid using a temporary file during transport using dkim. Unless a
840 transport-filter is involved we can buffer the headers in memory for
841 creating the signature, and read the spool data file once for the
842 signature and again for transmission.
844 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
845 4.77 as the kernel support then wasn't solid, having issues in 64bit
846 mode. Now, it's been long enough. Add support for FreeBSD also.
848 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
849 case where the routing stage had gathered several addresses to send to
850 a host before calling the transport for the first, we previously failed
851 to close down TLS in the old transport process before passing the TCP
852 connection to the new process. The new one sent a STARTTLS command
853 which naturally failed, giving a failed delivery and bloating the retry
854 database. Investigation and fix prototype from Wolfgang Breyha.
856 JH/12 Fix check on SMTP command input synchronisation. Previously there were
857 false-negatives in the check that the sender had not preempted a response
858 or prompt from Exim (running as a server), due to that code's lack of
859 awareness of the SMTP input buffering.
861 PP/04 Add commandline_checks_require_admin option.
862 Exim drops privileges sanely, various checks such as -be aren't a
863 security problem, as long as you trust local users with access to their
864 own account. When invoked by services which pass untrusted data to
865 Exim, this might be an issue. Set this option in main configuration
866 AND make fixes to the calling application, such as using `--` to stop
869 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
870 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
871 responses to those, into a single TLS record each way (this usually means
872 a single packet). As a side issue, smtp_enforce_sync now works on TLS
875 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
876 affects you only if you're dancing at the edge of the param size limits.
877 If you are, and this message makes sense to you, then: raise the
878 configured limit or use OpenSSL 1.1. Nothing we can do for older
881 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
882 cmdline element to get the documented default one. Previously it was
885 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
886 are unsuable for non-message-specific reasons - eg. connection timeout,
887 banner-time rejection.
889 JH/16 Fix logging of delivery remote port, when specified by router, under
892 PP/06 Repair manualroute's ability to take options in any order, even if one
893 is the name of a transport.
896 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
898 JH/17 Change the list-building routines interface to use the expanding-string
899 triplet model, for better allocation and copying behaviour.
901 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
902 Previously it was constructed the first time a possibly-matching string
903 was met in the configuration file input during startup; now it is done
906 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
907 compatible one, to avoid the (poorly documented) possibility of a config
908 file in the working directory redirecting the DB files, possibly correpting
909 some existing file. CVE-2017-10140 assigned for BDB.
911 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
912 cache-hot. Previously, although the result was properly cached, the
913 initial verify call returned a defer.
915 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
916 the main verify for receipient in uncached-mode.
918 JH/22 Retire historical build files to an "unsupported" subdir. These are
919 defined as "ones for which we have no current evidence of testing".
921 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
922 if present. Previously it was ignored.
924 JH/24 Start using specified-initialisers in C structure init coding. This is
925 a C99 feature (it's 2017, so now considered safe).
927 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
928 if was a fixed-sized field and bitmask ops via macros; it is now more
931 PP/07 GitHub PR 56: Apply MariaDB build fix.
932 Patch provided by Jaroslav Škarvada.
934 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
935 during Coverity cleanups [4.87 JH/47]
936 Diagnosis and fix provided by Michael Fischer v. Mollard.
938 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
939 the right size to place the terminating semicolon on its own folded
940 line, the header hash was calculated to an incorrect value thanks to
941 the (relaxed) space the fold became.
943 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
944 and confused the parent.
946 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
947 which could crash as a result. This could lead to undeliverable messages.
949 JH/28 Logging: "next input sent too soon" now shows where input was truncated
952 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
953 matters on fast-turnover and PID-randomising systems, which were getting
954 out-of-order delivery.
956 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
957 a possibly-overlapping copy. The symptom was that "Remote host closed
958 connection in response to HELO" was logged instead of the actual 4xx
961 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
962 Previously only that bufferd was discarded, resulting in SYMTP command
965 JH/32 DKIM: when a message has multiple signatures matching an identity given
966 in dkim_verify_signers, run the dkim acl once for each. Previously only
967 one run was done. Bug 2189.
969 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
970 "panic the current process" to "deliberately defer". The panic log is
971 still written with the problem list name; the mail and reject logs now
972 get a temp-reject line for the message that was being handled, saying
973 something like "domains check lookup or other defer". The SMTP 451
974 message is still "Temporary local problem".
976 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
977 A crafted sequence of BDAT commands could result in in-use memory beeing
978 freed. CVE-2017-16943.
980 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
981 from SMTP input. Previously it was always done; now only done for DATA
982 and not BDAT commands. CVE-2017-16944.
984 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
985 to the message (such as an overlong header line). Previously this was
986 not done and we did not exit BDAT mode. Followon from the previous item
987 though a different problem.
993 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
994 than -2003 did; needs libidn2 in addition to libidn.
996 JH/02 The path option on a pipe transport is now expanded before use.
998 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
999 Patch provided by "Björn", documentation fix added too.
1001 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1002 missing a wire-to-host endian conversion.
1004 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1005 close after a BDAT command line could be taken as a following command,
1006 giving a synch failure. Fix by only checking for synch immediately
1007 before acknowledging the chunk.
1009 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1010 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1011 macro. Patches provided by Josh Soref.
1013 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1014 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1015 by IANA as a service extension. However, John Klensin suggests that we
1018 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1019 the dkim code may be unix-mode line endings rather than smtp wire-format
1020 CRLF, so prepend a CR to any bare LF.
1022 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1023 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1025 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1026 the first were themselves being wrongly included in the feed into dkim
1027 processing; with most chunk sizes in use this resulted in an incorrect
1028 body hash calculated value.
1030 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1031 DKIM signature block, for verification. Although advised against by
1032 standards it is specifically not ruled illegal.
1034 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1036 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1037 missing a body hash (the bh= tag).
1039 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1040 It seems that HAProxy sends the Proxy Protocol information in clear and
1041 only then does a TLS startup, so do the same.
1043 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1044 TCP connections (such as for Spamd) unless the daemon successfully set
1045 Fast Open mode on its listening sockets. This fixes breakage seen on
1046 too-old kernels or those not configured for Fast Open, at the cost of
1047 requiring both directions being enabled for TFO, and TFO never being used
1048 by non-daemon-related Exim processes.
1050 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1051 endings, at least on the first header line. Try to canonify any that get
1052 past that check, despite the cost.
1054 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1055 now limited to an arbitrary five deep, while parsing addresses with the
1056 strip_excess_angle_brackets option enabled.
1058 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1059 instead leave the unprompted TLS handshake in socket buffer for the
1060 TLS library to consume.
1062 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1064 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1066 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1067 region in one operation, for speed, and this leaves assigned pointers
1068 dangling. Add checks run only under the testsuite which checks all
1069 variables at a store-reset and panics on a dangling pointer; add code
1070 explicitly nulling out all the variables discovered. Fixes one known
1071 bug: a transport crash, where a dangling pointer for $sending_ip_address
1072 originally assigned in a verify callout, is re-used.
1074 PP/06 Drop '.' from @INC in various Perl scripts.
1076 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1078 PP/08 Reduce a number of compilation warnings under clang; building with
1079 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1080 should be warning-free.
1082 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1084 HS/01 Fix portability problems introduced by PP/08 for platforms where
1085 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1087 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1088 chunk. This allows us to accept broken chunked messages. We need a more
1089 general solution here.
1091 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1092 already-broken messages in the queue.
1094 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1096 JH/19 Fix reference counting bug in routing-generated-address tracking.
1102 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1103 supports it and a size is available (ie. the sending peer gave us one).
1105 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1106 years of being deprecated). The replacements are the ACLs
1107 acl_smtp_mime and acl_not_smtp_mime.
1109 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1110 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1111 or even an in-clear connection were permitted. Now, if the host lookup
1112 was dnssec and dane was requested then the host is only used if the
1113 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1114 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1115 if one fails this test.
1116 This means that a poorly-configured remote DNS will make it incommunicado;
1117 but it protects against a DNS-interception attack on it.
1119 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1120 non-noisy when a race steals the message being considered.
1122 JH/05 If main configuration option tls_certificate is unset, generate a
1123 self-signed certificate for inbound TLS connections.
1125 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1126 in rewrites and routers.
1128 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1129 and logged a warning sing 4.83; now they are a configuration file error.
1131 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1132 (lacking @domain). Apply the same qualification processing as RCPT.
1134 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1136 JH/10 Support ${sha256:} applied to a string (as well as the previous
1139 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1140 a cutthrough deliver is pending, as we always want to make a connection.
1141 This also avoids re-routing the message when later placing the cutthrough
1142 connection after a verify cache hit.
1143 Do not update it with the verify result either.
1145 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1146 when routing results in more than one destination address.
1148 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1149 signing (which inhibits the cutthrough capability). Previously only
1150 the presence of an option was tested; now an expansion evaluating as
1151 empty is permissible (obviously it should depend only on data available
1152 when the cutthrough connection is made).
1154 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1155 the relevant preceding SMTP command did not note the pipelining mode.
1157 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1158 Previously they were not counted.
1160 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1161 as one having no matching records. Previously we deferred the message
1162 that needed the lookup.
1164 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1165 distinguished as "(=".
1167 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1168 for missing MX records. Previously it only worked for missing A records.
1170 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1172 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1173 after the data-go-ahead and data-ack. Patch from Jason Betts.
1175 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1176 even for a "none" policy. Patch from Tony Meyer.
1178 JH/22 Fix continued use of a connection for further deliveries. If a port was
1179 specified by a router, it must also match for the delivery to be
1182 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1183 When one of the recipients of a message was unsuitable for the connection
1184 (has no matching addresses), we lost track of needing to mark it
1185 deferred. As a result mail would be lost.
1187 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1189 JH/25 Decoding ACL controls is now done using a binary search; the source code
1190 takes up less space and should be simpler to maintain. Merge the ACL
1191 condition decode tables also, with similar effect.
1193 JH/26 Fix problem with one_time used on a redirect router which returned the
1194 parent address unchanged. A retry would see the parent address marked as
1195 delivered, so not attempt the (identical) child. As a result mail would
1198 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1199 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1200 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1203 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1204 Main config options check_{log,spool}_{inodes,space} are now
1205 100 inodes, 10MB unless set otherwise in the configuration.
1207 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1208 Previously it only applied to the main-section connection policy
1211 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1213 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1214 by me. Added RFC7919 DH primes as an alternative.
1216 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1217 are not in the system include path.
1219 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1220 GnuTLS, when a session startup failed (eg because the client disconnected)
1221 Exim did stdio operations after fclose. This was exposed by a recent
1222 change which nulled out the file handle after the fclose.
1224 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1225 signed directly by the cert-signing cert, rather than an intermediate
1226 OCSP-signing cert. This is the model used by LetsEncrypt.
1228 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1230 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1231 an incoming connection.
1233 HS/02 Bug 1802: Do not half-close the connection after sending a request
1236 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1237 fallback to "prime256v1".
1239 JH/34 SECURITY: Use proper copy of DATA command in error message.
1240 Could leak key material. Remotely exploitable. CVE-2016-9963.
1246 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1247 and 3.4.4 - once the server is enabled to respond to an OCSP request
1248 it does even when not requested, resulting in a stapling non-aware
1249 client dropping the TLS connection.
1251 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1252 support variable-length bit vectors. No functional change.
1254 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1255 The I= interface field on outgoing lines is now after the H= remote
1256 host field, same as incoming lines. There is a separate
1257 outgoing_interface log selector which allows you to disable the
1260 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1261 If not running log_selector +smtp_connection the mainlog would be held
1262 open indefinitely after a "too many connections" event, including to a
1263 deleted file after a log rotate. Leave the per net connection logging
1264 leaving it open for efficiency as that will be quickly detected by the
1265 check on the next write.
1267 HS/01 Bug 1671: Fix post transport crash.
1268 Processing the wait-<transport> messages could crash the delivery
1269 process if the message IDs didn't exist for some reason. When
1270 using 'split_spool_directory=yes' the construction of the spool
1271 file name failed already, exposing the same netto behaviour.
1273 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1274 mime_regex ACL conditions.
1276 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1277 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1278 to HELO, local diagnostic string.
1280 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1281 log line to debug. Even when configured with a tls authenticator many
1282 client connections are expected to not authenticate in this way, so
1283 an authenticate fail is not an error.
1285 HS/02 Add the Exim version string to the process info. This way exiwhat
1286 gives some more detail about the running daemon.
1288 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1289 matter for fast-change records such as DNSBLs.
1291 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1292 retry record, even if constant. There may be multiple transports with
1293 different interface settings and the retry behaviour needs to be kept
1296 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1298 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1300 JH/10 Bug 840: fix log_defer_output option of pipe transport
1302 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1303 hosts_randomize. This matters a lot when combined with mua_wrapper.
1305 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1306 ${quote_pgsql:<string>} operator.
1308 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1310 JH/14 Change header file name for freeradius-client. Relevant if compiling
1311 with Radius support; from the Gentoo tree and checked under Fedora.
1313 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1315 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1316 option transport_filter, meaning no filtering.
1318 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1320 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1321 defaults to "*" (all hosts). The variable is now available when not built
1322 with TLS, default unset, mainly to enable keeping the testsuite sane.
1323 If a server certificate is not supplied (via tls_certificate) an error is
1324 logged, and clients will find TLS connections fail on startup. Presumably
1325 they will retry in-clear.
1326 Packagers of Exim are strongly encouraged to create a server certificate
1327 at installation time.
1329 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1330 with the $config_file variable.
1332 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1333 in transport context, after the attempt, and per-recipient. The latter type
1334 is per host attempted. The event data is the error message, and the errno
1335 information encodes the lookup type (A vs. MX) used for the (first) host,
1336 and the trailing two digits of the smtp 4xx response.
1338 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1339 to write to mainlog (or rejectlog, paniclog) in the window between file
1340 creation and permissions/ownership being changed. Particularly affects
1341 installations where exicyclog is run as root, rather than exim user;
1342 result is that the running daemon panics and dies.
1344 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1346 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1347 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1348 "pri" and "weight". Note that the previous implicit priority given by the
1349 list order is no longer honoured.
1351 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1352 for DKIM processing.
1354 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1355 by defining SUPPORT_SOCKS.
1357 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1358 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1359 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1360 variables are renamed to proxy_{local,external}_{address,port}.
1362 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1363 for a build by defining SUPPORT_I18N
1365 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1366 of the query string, and make ${quote_redis:} do that quoting.
1368 JH/29 Move Events support from Experimental to mainline, enabled by default
1369 and removable for a build by defining DISABLE_EVENT.
1371 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1373 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1374 cached by the daemon.
1376 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1377 by defining LOOKUP_REDIS. The libhiredis library is required.
1379 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1380 keys are given for lookup.
1382 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1383 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1384 only supported when built with TLS support. The PolarSSL SHA routines
1385 are still used when the TLS library is too old for convenient support.
1387 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1388 openssl_options), for security. OpenSSL forces this from version 1.1.0
1389 server-side so match that on older versions.
1391 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1392 allocation for $value could be released as the expansion processing
1393 concluded, but leaving the global pointer active for it.
1395 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1396 and to use the domains and local_parts ACL conditions.
1398 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1399 incorrectly not doubled on cutthrough transmission, hence seen as a
1400 body-termination at the receiving system - resulting in truncated mails.
1401 Commonly the sender saw a TCP-level error, and retransmitted the message
1402 via the normal store-and-forward channel. This could result in duplicates
1403 received - but deduplicating mailstores were liable to retain only the
1404 initial truncated version.
1406 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1408 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1410 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1411 we're in there, support oversigning also; bug 1309.
1413 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1415 HS/04 Add support for keep_environment and add_environment options.
1417 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1418 either intentional arithmetic overflow during PRNG, or testing config-
1421 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1422 delivery resulted in actual delivery. Cancel cutthrough before DATA
1425 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1426 rejects a recipient: pass the reject to the originator.
1428 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1429 Many were false-positives and ignorable, but it's worth fixing the
1432 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1433 for the new environment-manipulation done at startup. Move the routines
1434 from being local to tls.c to being global via the os.c file.
1436 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1437 an extract embedded as result-arg for a map, the first arg for extract
1438 is unavailable so we cannot tell if this is a numbered or keyed
1439 extraction. Accept either.
1445 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1448 JH/02 The smtp transport option "multi_domain" is now expanded.
1450 JH/03 The smtp transport now requests PRDR by default, if the server offers
1453 JH/04 Certificate name checking on server certificates, when exim is a client,
1454 is now done by default. The transport option tls_verify_cert_hostnames
1455 can be used to disable this per-host. The build option
1456 EXPERIMENTAL_CERTNAMES is withdrawn.
1458 JH/05 The value of the tls_verify_certificates smtp transport and main options
1459 default to the word "system" to access the system default CA bundle.
1460 For GnuTLS, only version 3.0.20 or later.
1462 JH/06 Verification of the server certificate for a TLS connection is now tried
1463 (but not required) by default. The verification status is now logged by
1464 default, for both outbound TLS and client-certificate supplying inbound
1467 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1470 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1471 Status Notification (bounce) messages are now MIME format per RFC 3464.
1472 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1473 under the control of the dsn_advertise_hosts option, and routers may
1474 have a dsn_lasthop option.
1476 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1477 default, modifiable by a malware= option. The list separator for
1478 the options can now be changed in the usual way. Bug 68.
1480 JH/10 The smtp_receive_timeout main option is now expanded before use.
1482 JH/11 The incoming_interface log option now also enables logging of the
1483 local interface on delivery outgoing connections.
1485 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1486 if the interface and destination host and port all match.
1488 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1491 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1492 Patch from Andrew Lewis.
1494 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1495 now supports optional time-restrictions, weighting, and priority
1496 modifiers per server. Patch originally by <rommer@active.by>.
1498 JH/16 The spamd_address main option now supports a mixed list of local
1499 and remote servers. Remote servers can be IPv6 addresses, and
1500 specify a port-range.
1502 JH/17 Bug 68: The spamd_address main option now supports an optional
1503 timeout value per server.
1505 JH/18 Bug 1581: Router and transport options headers_add/remove can
1506 now have the list separator specified.
1508 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1511 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1514 JH/21 Support for the A6 type of dns record is withdrawn.
1516 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1517 rather than the verbs used.
1519 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1520 from 255 to 1024 chars.
1522 JH/24 Verification callouts now attempt to use TLS by default.
1524 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1525 are generic router options now. The defaults didn't change.
1527 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1528 Original patch from Alexander Shikoff, worked over by JH.
1530 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1531 files if ZCAT_COMMAND is not executable.
1533 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1535 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1537 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1538 Normally benign, it bites when the pair was led to by a CNAME;
1539 modern usage is to not canonicalize the domain to a CNAME target
1540 (and we were inconsistent anyway for A-only vs AAAA+A).
1542 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1544 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1545 when evaluating $sender_host_dnssec.
1547 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1548 $sender_helo_dnssec variable.
1550 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1552 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1554 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1556 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1557 documented as working, but never had. Support all but $spam_report.
1559 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1560 added for tls authenticator.
1562 HS/03 Add perl_taintmode main config option
1568 TL/01 When running the test suite, the README says that variables such as
1569 no_msglog_check are global and can be placed anywhere in a specific
1570 test's script, however it was observed that placement needed to be near
1571 the beginning for it to behave that way. Changed the runtest perl
1572 script to read through the entire script once to detect and set these
1573 variables, reset to the beginning of the script, and then run through
1574 the script parsing/test process like normal.
1576 TL/02 The BSD's have an arc4random API. One of the functions to induce
1577 adding randomness was arc4random_stir(), but it has been removed in
1578 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1579 function when detected.
1581 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1582 cause callback expansion.
1584 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1585 syntax errors in an expansion can be treated as a string instead of
1586 logging or causing an error, due to the internal use of bool_lax
1587 instead of bool when processing it.
1589 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1590 server certificates when making smtp deliveries.
1592 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1594 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1596 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1598 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1599 Merged patch from Sebastian Wiedenroth.
1601 JH/05 Fix results-pipe from transport process. Several recipients, combined
1602 with certificate use, exposed issues where response data items split
1603 over buffer boundaries were not parsed properly. This eventually
1604 resulted in duplicates being sent. This issue only became common enough
1605 to notice due to the introduction of connection certificate information,
1606 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1608 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1609 size buffer was used, resulting in syntax errors when an expansion
1612 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1613 version 3.3.6 or later.
1615 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1616 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1617 both become "event_action", the variables become $event_name, $event_data
1618 and $event_defer_errno. There is a new variable $verify_mode, usable in
1619 routers, transports and related events. The tls:cert event is now also
1620 raised for inbound connections, if the main configuration event_action
1623 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1624 early OCSP support, but no stapling (appears to be less than 1.0.0).
1626 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1627 server certificate names available under the smtp transport option
1628 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1631 JH/10 Time-related extraction expansions from certificates now use the main
1632 option "timezone" setting for output formatting, and are consistent
1633 between OpenSSL and GnuTLS compilations. Bug 1541.
1635 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1636 encoded parameter in the incoming message. Bug 1558.
1638 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1639 include certificate info, eximon was claiming there were spoolfile
1642 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1644 JH/14 Log delivery-related information more consistently, using the sequence
1645 "H=<name> [<ip>]" wherever possible.
1647 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1648 are problematic for Debian distribution, omit them from the release
1651 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1653 JH/16 Fix string representation of time values on 64bit time_t architectures.
1656 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1657 output list separator was used.
1662 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1663 checkers that were complaining about end of non-void function with no
1666 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1667 This was a regression introduced in 4.83 by another bugfix.
1669 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1671 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1672 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1678 TF/01 Correctly close the server side of TLS when forking for delivery.
1680 When a message was received over SMTP with TLS, Exim failed to clear up
1681 the incoming connection properly after forking off the child process to
1682 deliver the message. In some situations the subsequent outgoing
1683 delivery connection happened to have the same fd number as the incoming
1684 connection previously had. Exim would try to use TLS and fail, logging
1685 a "Bad file descriptor" error.
1687 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1688 utilities have not been installed.
1690 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1691 temporary space as the ACL may create new global variables.
1693 TL/01 LDAP support uses per connection or global context settings, depending
1694 upon the detected version of the libraries at build time.
1696 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1697 to extract and use the src ip:port in logging and expansions as if it
1698 were a direct connection from the outside internet. PPv2 support was
1699 updated based on HAProxy spec change in May 2014.
1701 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1703 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1704 Properly escape header and check for NULL return.
1706 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1709 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1711 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1712 characters in header names, implemented as a verify condition.
1713 Contributed by Michael Fischer v. Mollard.
1715 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1716 results permerror and temperror. Previous values are deprecated but
1717 still accepted. In a future release, err_perm and err_temp will be
1718 completely removed, which will be a backward incompatibility if the
1719 ACL tests for either of these two old results. Patch contributed by
1720 user bes-internal on the mailing list.
1722 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1724 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1725 selectors, in both main and reject logs.
1727 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1730 JH/07 Add malware type "sock" for talking to simple daemon.
1732 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1734 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1735 routers/transports under cutthrough routing.
1737 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1738 numbers. Touch up "bool" conditional to keep the same definition.
1740 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1742 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1744 JH/12 Expand items in router/transport headers_add or headers_remove lists
1745 individually rather than the list as a whole. Bug 1452.
1747 Required for reasonable handling of multiple headers_ options when
1748 they may be empty; requires that headers_remove items with embedded
1749 colons must have them doubled (or the list-separator changed).
1751 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1752 view the policy declared in the DMARC record. Currently, $dmarc_status
1753 is a combined value of both the record presence and the result of the
1756 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1758 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1759 dnslookup router and the smtp transport (applying to the forward
1762 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1763 of ldap servers used for a specific lookup. Patch provided by Heiko
1766 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1767 New variable $lookup_dnssec_authenticated for observability.
1769 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1770 Patch submitted by Lars Timman.
1772 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1774 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1775 Requires trusted mode and valid format message id, aborts otherwise.
1776 Patch contributed by Heiko Schlichting.
1778 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1779 certextract with support for various fields. Bug 1358.
1781 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1782 is requested by default, modifiable by smtp transport option
1785 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1786 operate on certificate variables to give certificate fingerprints
1787 Also new ${sha256:cert_variable}.
1789 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1791 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1792 Christian Aistleitner.
1794 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1796 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1797 file. Patch from Wolfgang Breyha.
1799 JH/25 Expand the coverage of the delivery $host and $host_address to
1800 client authenticators run in verify callout. Bug 1476.
1802 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1803 align with daemon_smtp_ports. Bug 72.
1805 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1806 support and error reporting did not work properly.
1808 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1809 and is readable. Patch from Andrew Colin Kissa.
1811 TL/14 Enhance documentation of ${run expansion and how it parses the
1812 commandline after expansion, particularly in the case when an
1813 unquoted variable expansion results in an empty value.
1815 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1817 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1818 by renaming away from it. Addresses GH issue 15, reported by
1821 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1823 TL/15 SECURITY: prevent double expansion in math comparison functions
1824 (can expand unsanitized data). Not remotely exploitable.
1831 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1833 PP/02 Make -n do something, by making it not do something.
1834 When combined with -bP, the name of an option is not output.
1836 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1839 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1840 $sender_host_name and config options to manage this, and basic check
1843 PP/05 DSCP support for outbound connections and control modifier for inbound.
1845 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1846 (Only plugin which currently uses this is kerberos4, which nobody should
1847 be using, but we should make it available and other future plugins might
1848 conceivably use it, even though it would break NAT; stuff *should* be
1849 using channel bindings instead).
1851 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1852 name; added for Sendmail compatibility; requires admin caller.
1853 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1854 just ignore it); requires trusted caller.
1855 Also parse but ignore: -Ac -Am -X<logfile>
1858 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1860 TL/02 Add +smtp_confirmation as a default logging option.
1862 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1863 Patch by Magnus Holmgren from 2007-02-20.
1865 TL/04 Bugzilla 1281 - Spec typo.
1866 Bugzilla 1283 - Spec typo.
1867 Bugzilla 1290 - Spec grammar fixes.
1869 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1871 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1873 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1874 mailing list by by Dmitry Isaikin.
1876 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1878 JH/02 Support "G" suffix to numbers in ${if comparisons.
1880 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1882 NM/01 Bugzilla 1197 - Spec typo
1883 Bugzilla 1196 - Spec examples corrections
1885 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1887 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1888 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1891 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1892 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1894 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1895 "acl {{name}{arg}...}", and optional args on acl condition
1898 JH/05 Permit multiple router/transport headers_add/remove lines.
1900 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1902 JH/07 Avoid using a waiting database for a single-message-only transport.
1903 Performance patch from Paul Fisher. Bugzilla 1262.
1905 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1908 JH/09 Add $headers_added variable, with content from use of ACL modifier
1909 add_header (but not yet added to the message). Bugzilla 199.
1911 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1912 Pulled from Bugzilla 817 by Wolfgang Breyha.
1914 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1916 (nb: this is the same fix as in Exim 4.80.1)
1918 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1921 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1924 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1926 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1927 advertise SMTP AUTH mechanism to us, instead of a generic
1928 protocol violation error. Also, make Exim more robust to bad
1929 data from the Dovecot auth socket.
1931 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1933 When a queue runner is handling a message, Exim first routes the
1934 recipient addresses, during which it prunes them based on the retry
1935 hints database. After that it attempts to deliver the message to
1936 any remaining recipients. It then updates the hints database using
1939 So if a recipient address works intermittently, it can get repeatedly
1940 deferred at routing time. The retry hints record remains fresh so the
1941 address never reaches the final cutoff time.
1943 This is a fairly common occurrence when a user is bumping up against
1944 their storage quota. Exim had some logic in its local delivery code
1945 to deal with this. However it did not apply to per-recipient defers
1946 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1948 This change adds a proper retry rule check during routing so that the
1949 final cutoff time is checked against the message's age. We only do
1950 this check if there is an address retry record and there is not a
1951 domain retry record; this implies that previous attempts to handle
1952 the address had the retry_use_local_parts option turned on. We use
1953 this as an approximation for the destination being like a local
1954 delivery, as in LMTP.
1956 I suspect this new check makes the old local delivery cutoff check
1957 redundant, but I have not verified this so I left the code in place.
1959 TF/02 Correct gecos expansion when From: is a prefix of the username.
1961 Test 0254 submits a message to Exim with the header
1965 When I ran the test suite under the user fanf2, Exim expanded
1966 the header to contain my full name, whereas it should have added
1967 a Resent-Sender: header. It erroneously treats any prefix of the
1968 username as equal to the username.
1970 This change corrects that bug.
1972 GF/01 DCC debug and logging tidyup
1973 Error conditions log to paniclog rather than rejectlog.
1974 Debug lines prefixed by "DCC: " to remove any ambiguity.
1976 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1978 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1979 Bug spotted by Jeremy Harris; was flawed since initial commit.
1980 Would have resulted in OCSP responses post-SNI triggering an Exim
1981 NULL dereference and crash.
1983 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1985 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1986 Bug detection, analysis and fix by Samuel Thibault.
1987 Bugzilla 1331, Debian bug #698092.
1989 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1991 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1992 Server implementation by Todd Lyons, client by JH.
1993 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1994 config variable "prdr_enable" controls whether the server
1995 advertises the facility. If the client requests PRDR a new
1996 acl_data_smtp_prdr ACL is called once for each recipient, after
1997 the body content is received and before the acl_smtp_data ACL.
1998 The client is controlled by both of: a hosts_try_prdr option
1999 on the smtp transport, and the server advertisement.
2000 Default client logging of deliveries and rejections involving
2001 PRDR are flagged with the string "PRDR".
2003 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2004 fclose(). Diagnosis by Todd Lyons.
2006 PP/17 Update configure.default to handle IPv6 localhost better.
2007 Patch by Alain Williams (plus minor tweaks).
2010 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2011 This is now consistent with GnuTLS, and is now documented: the
2012 previous undocumented portable approach to treating the option as
2013 unset was to force an expansion failure. That still works, and
2014 an empty string is now equivalent.
2016 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2017 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2018 not performing validation itself.
2020 PP/20 Added force_command boolean option to pipe transport.
2021 Patch from Nick Koston, of cPanel Inc.
2023 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2026 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2028 PP/21 Fix eximon continuous updating with timestamped log-files.
2029 Broken in a format-string cleanup in 4.80, missed when I repaired the
2030 other false fix of the same issue.
2031 Report and fix from Heiko Schlichting.
2034 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2035 Report from Prashanth Katuri.
2037 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2038 It's SecureTransport, so affects any MacOS clients which use the
2039 system-integrated TLS libraries, including email clients.
2041 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2042 using a MIME ACL for non-SMTP local injection.
2043 Report and assistance in diagnosis by Warren Baker.
2045 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2047 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2049 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2050 last id that failed so it may be referenced in subsequent ACL's.
2052 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2055 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2056 ldap library initialization, allowing self-signed CA's to be
2057 used. Also properly sets require_cert option later in code by
2058 using NULL (global ldap config) instead of ldap handle (per
2059 session). Bug diagnosis and testing by alxgomz.
2061 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2062 the src/util/ subdirectory.
2064 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2065 renamed to Transport Post Delivery Action by Jeremy Harris, as
2068 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2069 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2070 redis_servers = needs to be configured which will be used by the redis
2071 lookup. Patch from Warren Baker, of The Packet Hub.
2073 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2075 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2076 hostname or reverse DNS when processing a host list. Used suggestions
2077 from multiple comments on this bug.
2079 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2081 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2082 Missed a few lines, added it to make the runtest require no keyboard
2085 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2086 contains upper case chars. Make router use caseful_local_part.
2088 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2089 support when GnuTLS has been built with p11-kit.
2095 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2097 This, or similar/improved, will also be change PP/11 of 4.82.
2103 PP/01 Handle short writes when writing local log-files.
2104 In practice, only affects FreeBSD (8 onwards).
2105 Bugzilla 1053, with thanks to Dmitry Isaikin.
2107 NM/01 Bugzilla 949 - Documentation tweak
2109 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2112 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2114 PP/02 Implemented gsasl authenticator.
2116 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2118 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2119 `pkg-config foo` for cflags/libs.
2121 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2122 with rest of GSASL and with heimdal_gssapi.
2124 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2125 `pkg-config foo` for cflags/libs for the TLS implementation.
2127 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2128 properties get this fed in as external SSF. A number of robustness
2129 and debugging improvements to the cyrus_sasl authenticator.
2131 PP/08 cyrus_sasl server now expands the server_realm option.
2133 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2134 Patch by Jeremy Harris.
2136 PP/10 Added dbmjz lookup type.
2138 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2140 PP/12 MAIL args handles TAB as well as SP, for better interop with
2141 non-compliant senders.
2142 Analysis and variant patch by Todd Lyons.
2144 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2145 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2146 Patch from Dirk Mueller <dmueller@suse.com>
2148 PP/13 tls_peerdn now print-escaped for spool files.
2149 Observed some $tls_peerdn in wild which contained \n, which resulted
2150 in spool file corruption.
2152 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2153 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2154 or write after TLS renegotiation, which otherwise led to messages
2157 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2158 as a tracking header (ie: a signed header comes before the signature).
2159 Patch from Wolfgang Breyha.
2161 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2162 comma-sep list; embedded commas doubled.
2164 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2166 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2168 Report and patch from Dmitry Banschikov.
2170 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2171 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2172 failures. We appear to now support TLS1.1+ with Exim.
2174 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2175 lets Exim select keys and certificates based upon TLS SNI from client.
2176 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2177 before an outbound SMTP session. New log_selector, +tls_sni.
2179 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2180 NULL dereference. Report and patch from Alun Jones.
2182 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2183 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2184 Not seeing resolver debug output on NetBSD, but suspect this is a
2185 resolver implementation change.
2187 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2188 Left warnings. Added "eximon gdb" invocation mode.
2190 PP/21 Defaulting "accept_8bitmime" to true, not false.
2192 PP/22 Added -bw for inetd wait mode support.
2194 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2195 locate the relevant includes and libraries. Made this the default.
2197 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2198 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2200 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2201 This may cause build issues on older platforms.
2203 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2204 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2205 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2206 Added SNI support via GnuTLS too.
2207 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2209 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2211 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2212 Applied second patch from Janne, implementing suggestion to default
2213 multiple-strings-in-record handling to match SPF spec.
2215 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2217 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2218 read-only, out of scope).
2219 Patch from Wolfgang Breyha, report from Stuart Northfield.
2221 PP/29 Fix three issues highlighted by clang analyser static analysis.
2222 Only crash-plausible issue would require the Cambridge-specific
2223 iplookup router and a misconfiguration.
2224 Report from Marcin Mirosław.
2226 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2228 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2229 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2230 As part of this, removing so much warning spew let me fix some minor
2231 real issues in debug logging.
2233 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2234 assignment on my part. Fixed.
2236 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2237 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2238 Janne Snabb (who went above and beyond: thank you).
2240 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2241 string otherwise requires a connection and a bunch more work and it's
2242 relatively easy to get wrong. Should also expose TLS library linkage
2245 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2246 64-bit ${eval} (JH/03).
2248 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2249 GNU libc to support some of the 64-bit stuff, should not lead to
2250 conflicts. Defined before os.h is pulled in, so if a given platform
2251 needs to override this, it can.
2253 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2254 protection layer was required, which is not implemented.
2255 Bugzilla 1254, patch from Wolfgang Breyha.
2257 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2258 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2259 tls_dhparam take prime identifiers. Also unbreak combination of
2260 OpenSSL+DH_params+TLSSNI.
2262 PP/39 Disable SSLv2 by default in OpenSSL support.
2268 PP/01 Solaris build fix for Oracle's LDAP libraries.
2269 Bugzilla 1109, patch from Stephen Usher.
2271 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2273 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2276 TF/02 Fix a couple more cases where we did not log the error message
2277 when unlink() failed. See also change 4.74-TF/03.
2279 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2280 lock up or crash if it happened to be inside a call to libc when it
2281 got a SIGUSR1 from exiwhat.
2283 The SIGUSR1 handler appends the current process status to the process
2284 log which is later printed by exiwhat. It used to use the general
2285 purpose logging code to do this, but several functions it calls are
2286 not safe for signals.
2288 The new output code in the SIGUSR1 handler is specific to the process
2289 log, and simple enough that it's easy to inspect for signal safety.
2290 Removing some special cases also simplifies the general logging code.
2291 Removing the spurious timestamps from the process log simplifies
2294 TF/04 Improved ratelimit ACL condition.
2296 The /noupdate option has been deprecated in favour of /readonly which
2297 has clearer semantics. The /leaky, /strict, and /readonly update modes
2298 are mutually exclusive. The update mode is no longer included in the
2299 database key; it just determines when the database is updated. (This
2300 means that when you upgrade Exim will forget old rate measurements.)
2302 Exim now checks that the per_* options are used with an update mode that
2303 makes sense for the current ACL. For example, when Exim is processing a
2304 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2305 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2306 must specify per_mail/readonly. If you omit the update mode it defaults to
2307 /leaky where that makes sense (as before) or /readonly where required.
2309 The /noupdate option is now undocumented but still supported for
2310 backwards compatibility. It is equivalent to /readonly except that in
2311 ACLs where /readonly is required you may specify /leaky/noupdate or
2312 /strict/noupdate which are treated the same as /readonly.
2314 A useful new feature is the /count= option. This is a generalization
2315 of the per_byte option, so that you can measure the throughput of other
2316 aggregate values. For example, the per_byte option is now equivalent
2317 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2319 The per_rcpt option has been generalized using the /count= mechanism
2320 (though it's more complicated than the per_byte equivalence). When it is
2321 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2322 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2323 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2324 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2325 Note that using per_rcpt with a non-readonly update mode in more than
2326 one ACL will cause the recipients to be double-counted. (The per_mail
2327 and per_byte options don't have this problem.)
2329 The handling of very low rates has changed slightly. If the computed rate
2330 is less than the event's count (usually one) then this event is the first
2331 after a long gap. In this case the rate is set to the same as this event's
2332 count, so that the first message of a spam run is counted properly.
2334 The major new feature is a mechanism for counting the rate of unique
2335 events. The new per_addr option counts the number of different
2336 recipients that someone has sent messages to in the last time period. It
2337 behaves like per_rcpt if all the recipient addresses are different, but
2338 duplicate recipient addresses do not increase the measured rate. Like
2339 the /count= option this is a general mechanism, so the per_addr option
2340 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2341 example, measure the rate that a client uses different sender addresses
2342 with the options per_mail/unique=$sender_address. There are further
2343 details in the main documentation.
2345 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2347 TF/06 Removed a few PCRE remnants.
2349 TF/07 Automatically extract Exim's version number from tags in the git
2350 repository when doing development or release builds.
2352 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2353 Bugzilla 879. Patch from Paul Fisher.
2355 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2356 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2359 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2361 PP/05 Variable $av_failed, true if the AV scanner deferred.
2362 Bugzilla 1078. Patch from John Horne.
2364 PP/06 Stop make process more reliably on build failure.
2365 Bugzilla 1087. Patch from Heiko Schlittermann.
2367 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2368 Bugzilla 1089. Patch from Heiko Schlittermann.
2370 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2371 Bugzilla 1131. Patch from Holger Weiß.
2373 PP/09 Handle IPv6 addresses with SPF.
2374 Bugzilla 860. Patch from Wolfgang Breyha.
2376 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2378 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2381 PP/11 match_* no longer expand right-hand-side by default.
2382 New compile-time build option, EXPAND_LISTMATCH_RHS.
2383 New expansion conditions, "inlist", "inlisti".
2385 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2387 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2389 PP/14 fix log_write() format string regression from TF/03.
2390 Bugzilla 1152. Patch from Dmitry Isaikin.
2396 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2398 PP/02 Harmonised TLS library version reporting; only show if debugging.
2399 Layout now matches that introduced for other libraries in 4.74 PP/03.
2401 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2403 PP/04 New "dns_use_edns0" global option.
2405 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2408 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2409 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2411 TK/01 Updated PolarSSL code to 0.14.2.
2412 Bugzilla 1097. Patch from Andreas Metzler.
2414 PP/07 Catch divide-by-zero in ${eval:...}.
2415 Fixes bugzilla 1102.
2417 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2420 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2421 format-string attack -- SECURITY: remote arbitrary code execution.
2423 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2424 time unintentionally subject to list matching rules, letting the header
2425 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2426 arbitrary string expansion). This allowed for information disclosure.
2428 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2429 INT_MIN/-1 -- value coerced to INT_MAX.
2435 NM/01 Workaround for PCRE version dependency in version reporting
2438 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2439 This fixes portability to compilers other than gcc, notably
2440 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2442 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2443 makefiles for portability to HP-UX and POSIX correctness.
2445 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2446 Also via indented variable definition in the Makefile.
2447 (Debugging by Oliver Heesakkers).
2449 PP/02 Restore caching of spamd results with expanded spamd_address.
2450 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2452 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2453 Improves build reliability. Fix from: Frank Elsner
2455 NM/02 Fix wide character breakage in the rfc2047 coding
2456 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2458 NM/03 Allow underscore in dnslist lookups
2459 Fixes bug 1026. Patch from Graeme Fowler
2461 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2462 Code patches from Adam Ciarcinski of NetBSD.
2464 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2467 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2468 is logged, to avoid truncation. Patch from John Horne.
2470 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2471 Patch from Jakob Hirsch.
2473 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2474 SQL string expansion failure details.
2475 Patch from Andrey Oktyabrski.
2477 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2478 Patch from Simon Arlott.
2480 PP/09 New lookups functionality failed to compile on old gcc which rejects
2481 extern declarations in function scope.
2482 Patch from Oliver Fleischmann
2484 PP/10 Use sig_atomic_t for flags set from signal handlers.
2485 Check getgroups() return and improve debugging.
2486 Fixed developed for diagnosis in bug 927 (which turned out to be
2489 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2490 Patch from Mark Zealey.
2492 PP/12 Bugzilla 1056: Improved spamd server selection.
2493 Patch from Mark Zealey.
2495 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2496 Based on patch from Heiko Schlittermann.
2498 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2499 Patch from Uwe Doering, sign-off by Michael Haardt.
2501 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2502 variable declaration deep within a block. Bug and patch from
2505 PP/15 lookups-Makefile IRIX compatibility coercion.
2507 PP/16 Make DISABLE_DKIM build knob functional.
2509 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2510 Patch by Simon Arlott
2512 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2513 variable argument macros. Our copy now differs from upstream.
2519 TF/01 Failure to get a lock on a hints database can have serious
2520 consequences so log it to the panic log.
2522 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2523 controlled using the smtp_confirmation log selector.
2525 TF/03 Include the error message when we fail to unlink a spool file.
2527 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2528 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2529 for maintaining out-of-tree patches for some time.
2531 PP/01 Bugzilla 139: Documentation and portability issues.
2532 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2533 Handle per-OS dynamic-module compilation flags.
2535 PP/02 Let /dev/null have normal permissions.
2536 The 4.73 fixes were a little too stringent and complained about the
2537 permissions on /dev/null. Exempt it from some checks.
2538 Reported by Andreas M. Kirchwitz.
2540 PP/03 Report version information for many libraries, including
2541 Exim version information for dynamically loaded libraries. Created
2542 version.h, now support a version extension string for distributors
2543 who patch heavily. Dynamic module ABI change.
2545 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2546 privilege escalation vulnerability whereby the Exim run-time user
2547 can cause root to append content of the attacker's choosing to
2550 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2553 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2554 If dropping privileges for untrusted macros, we disabled normal logging
2555 on the basis that it would fail; for the Exim run-time user, this is not
2556 the case, and it resulted in successful deliveries going unlogged.
2557 Fixed. Reported by Andreas Metzler.
2563 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2564 only prepend for the Resent-* case. Fixes regression introduced in
2565 Exim 4.70 by NM/22 for Bugzilla 607.
2567 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2568 increasing numbers of administrators be bitten by this.
2570 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2572 PP/03 Bugzilla 994: added openssl_options main configuration option.
2574 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2576 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2578 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2580 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2581 without a peer certificate, leading to a segfault because of an
2582 assumption that peers always have certificates. Be a little more
2583 paranoid. Problem reported by Martin Tscholak.
2585 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2586 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2587 NB: ClamAV planning to remove STREAM in "middle of 2010".
2588 CL also introduces -bmalware, various -d+acl logging additions and
2589 more caution in buffer sizes.
2591 PP/09 Implemented reverse_ip expansion operator.
2593 PP/10 Bugzilla 937: provide a "debug" ACL control.
2595 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2597 PP/12 Bugzilla 973: Implement --version.
2599 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2601 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2603 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2605 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2606 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2607 ignore trailing whitespace.
2609 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2611 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2614 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2615 Notification from Dr Andrew Aitchison.
2617 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2618 ExtendedDetectionInfo response format.
2619 Notification from John Horne.
2621 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2624 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2625 XSL and documented dependency on system catalogs, with examples of how
2628 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2631 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2632 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2633 configuration file which is writeable by the Exim user or group.
2635 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2636 of configuration files to cover files specified with the -C option if
2637 they are going to be used with root privileges, not just the default
2640 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2641 option (effectively making it always true).
2643 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2644 files to be used while preserving root privileges.
2646 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2647 that rogue child processes cannot use them.
2649 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2650 run-time user, instead of root.
2652 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2653 Exim run-time user without dropping privileges.
2655 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2656 result string, instead of calling string_vformat() twice with the same
2659 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2660 for other users. Others should always drop root privileges if they use
2661 -C on the command line, even for a whitelisted configure file.
2663 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2665 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2671 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2672 $data_path, and $header_path variables; fixed documentation bugs and
2675 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2676 exipick to access non-standard spools, including the "frozen" queue
2679 NM/01 Bugzilla 965: Support mysql stored procedures.
2680 Patch from Alain Williams
2682 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2684 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2685 Patch from Andreas Metzler
2687 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2688 Patch from Kirill Miazine
2690 NM/05 Bugzilla 671: Added umask to procmail example.
2692 JJ/03 installed exipick 20100323.0, fixing doc bug
2694 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2695 directory. Notification and patch from Dan Rosenberg.
2697 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2699 TK/02 Improve log output when DKIM signing operation fails.
2701 MH/01 Treat the transport option dkim_domain as a colon separated
2702 list, not as a single string, and sign the message with each element,
2703 omitting multiple occurences of the same signer.
2705 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2706 Bugzilla 985, 986. Patch by Simon Arlott
2708 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2709 Patch by Simon Arlott
2711 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2712 MBX locking. Notification from Dan Rosenberg.
2718 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2720 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2722 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2724 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2726 NM/04 Bugzilla 915: Flag broken perl installation during build.
2732 TK/01 Added patch by Johannes Berg that expands the main option
2733 "spamd_address" if it starts with a dollar sign.
2735 TK/02 Write list of recipients to X-Envelope-Sender header when building
2736 the mbox-format spool file for content scanning (suggested by Jakob
2739 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2740 (http://www.dcc-servers.net/) support via dccifd. Activated by
2741 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2743 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2744 by Mark Daniel Reidel <mr@df.eu>.
2746 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2747 When building exim an external PCRE library is now needed -
2748 PCRE is a system library on the majority of modern systems.
2749 See entry on PCRE_LIBS in EDITME file.
2751 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2752 conversation. Added nologin parameter to request.
2753 Patch contributed by Kirill Miazine.
2755 TF/01 Do not log submission mode rewrites if they do not change the address.
2757 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2759 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2760 log files in place. Contributed by Roberto Lima.
2762 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2764 TF/03 Bugzilla 615: When checking the local_parts router precondition
2765 after a local_part_suffix or local_part_prefix option, Exim now
2766 does not use the address's named list lookup cache, since this
2767 contains cached lookups for the whole local part.
2769 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2770 Robert Millan. Documentation is in experimental-spec.txt.
2772 TF/04 Bugzilla 668: Fix parallel build (make -j).
2774 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2776 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2777 Patch provided by Jan Srzednicki.
2779 TF/05 Leading white space used to be stripped from $spam_report which
2780 wrecked the formatting. Now it is preserved.
2782 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2783 that they are available at delivery time.
2785 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2787 TF/08 TLS error reporting now respects the incoming_interface and
2788 incoming_port log selectors.
2790 TF/09 Produce a more useful error message if an SMTP transport's hosts
2791 setting expands to an empty string.
2793 NM/06 Bugzilla 744: EXPN did not work under TLS.
2794 Patch provided by Phil Pennock.
2796 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2797 Patch provided by Richard Godbee.
2799 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2800 acl_smtp_notquit, added index entry.
2802 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2803 Patch provided by Eugene Bujak.
2805 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2806 accept(). Patch provided by Maxim Dounin.
2808 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2809 Patch provided by Phil Pennock.
2811 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2813 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2814 Patch provided by Brad "anomie" Jorsch.
2816 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2817 Patch provided by Dean Brooks.
2819 TK/05 Add native DKIM support (does not depend on external libraries).
2821 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2822 Patch provided by Graeme Fowler.
2824 NM/16 Bugzilla 851: Documentation example syntax fix.
2826 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2828 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2831 NM/19 Bugzilla 745: TLS version reporting.
2832 Patch provided by Phil Pennock.
2834 NM/20 Bugzilla 167: bool: condition support.
2835 Patch provided by Phil Pennock.
2837 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2838 clients. Patch provided by Phil Pennock.
2840 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2841 Patch provided by Brad "anomie" Jorsch.
2843 NM/23 Bugzilla 687: Fix misparses in eximstats.
2844 Patch provided by Heiko Schlittermann.
2846 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2847 Patch provided by Heiko Schlittermann.
2849 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2850 plus update to original patch.
2852 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2854 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2855 Patch provided by David Brownlee.
2857 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2859 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2861 NM/30 Bugzilla 888: TLS documentation bugfixes.
2863 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2865 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2866 Unlike the original bugzilla I have changed all shell scripts in src tree.
2868 NM/33 Bugzilla 898: Transport filter timeout fix.
2869 Patch by Todd Rinaldo.
2871 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2872 Patch by Serge Demonchaux.
2874 NM/35 Bugzilla 39: Base64 decode bug fixes.
2875 Patch by Jakob Hirsch.
2877 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2879 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2881 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2883 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2889 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2890 ALT-N's libdkim that I have put here:
2891 http://duncanthrax.net/exim-experimental/
2893 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2894 were called 'true' and it seems that C99 defines that as a reserved
2895 keyword to be used with 'bool' variable types. That means you could
2896 not include C99-style headers which use bools without triggering
2897 build errors in sieve.c.
2899 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2900 as mailq or other aliases. Changed the --help handling significantly
2901 to do whats expected. exim_usage() emits usage/help information.
2903 SC/01 Added the -bylocaldomain option to eximstats.
2905 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2907 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2909 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2915 PH/01 Another patch from the Sieve maintainer.
2917 PH/02 When an IPv6 address is converted to a string for single-key lookup
2918 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2919 dots are used instead of colons so that keys in lsearch files need not
2920 contain colons. This was done some time before quoting was made available
2921 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2922 (notated using the quote facility) so as to distinguish them from IPv4
2923 keys. This meant that lookups for IP addresses in host lists did not work
2924 for iplsearch lookups.
2926 This has been fixed by arranging for IPv6 addresses to be expressed with
2927 colons if the lookup type is iplsearch. This is not incompatible, because
2928 previously such lookups could never work.
2930 The situation is now rather anomalous, since one *can* have colons in
2931 ordinary lsearch keys. However, making the change in all cases is
2932 incompatible and would probably break a number of configurations.
2934 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2937 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2938 conversion specification without a maximum field width, thereby enabling
2939 a rogue spamd server to cause a buffer overflow. While nobody in their
2940 right mind would setup Exim to query an untrusted spamd server, an
2941 attacker that gains access to a server running spamd could potentially
2942 exploit this vulnerability to run arbitrary code as the Exim user.
2944 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2945 $primary_hostname instead of what libspf2 thinks the hosts name is.
2947 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2948 a directory entry by the name of the lookup key. Previously, if a
2949 symlink pointed to a non-existing file or a file in a directory that
2950 Exim lacked permissions to read, a lookup for a key matching that
2951 symlink would fail. Now it is enough that a matching directory entry
2952 exists, symlink or not. (Bugzilla 503.)
2954 PH/03 The body_linecount and body_zerocount variables are now exported in the
2957 PH/04 Added the $dnslist_matched variable.
2959 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2960 This means they are set thereafter only if the connection becomes
2963 PH/06 Added the client_condition to authenticators so that some can be skipped
2964 by clients under certain conditions.
2966 PH/07 The error message for a badly-placed control=no_multiline_responses left
2967 "_responses" off the end of the name.
2969 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2971 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2972 (without spaces) instead of just copying the configuration text.
2974 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2976 PH/11 Added $max_received_linelength.
2978 PH/12 Added +ignore_defer and +include_defer to host lists.
2980 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2981 way in which PCRE > 7.0 is built.
2983 PH/14 Implemented queue_only_load_latch.
2985 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2986 MAIL command. The effect was to mangle the value on 64-bit systems.
2988 PH/16 Another patch from the Sieve maintainer.
2990 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2992 PH/18 If a system quota error occurred while trying to create the file for
2993 a maildir delivery, the message "Mailbox is full" was not appended to the
2994 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2995 only to a quota excession during the actual writing of the file.
2997 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2998 characters?) which causes problems in log lines. The DN values are now
2999 passed through string_printing() before being added to log lines.
3001 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3002 and InterBase are left for another time.)
3004 PH/21 Added message_body_newlines option.
3006 PH/22 Guard against possible overflow in moan_check_errorcopy().
3008 PH/23 POSIX allows open() to be a macro; guard against that.
3010 PH/24 If the recipient of an error message contained an @ in the local part
3011 (suitably quoted, of course), incorrect values were put in $domain and
3012 $local_part during the evaluation of errors_copy.
3018 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3019 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3022 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3023 issue a MAIL command.
3025 PH/02 In an ACL statement such as
3027 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3029 if a client was not listed at all, or was listed with a value other than
3030 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3031 the condition was not true (as it should be), so access was not denied.
3032 The bug was that the ! inversion was incorrectly passed on to the second
3033 item. This has been fixed.
3035 PH/03 Added additional dnslists conditions == and =& which are different from
3036 = and & when the dns lookup returns more than one IP address.
3038 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3039 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3041 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3042 FSYNC, which compiles an option called disable_fsync that allows for
3043 bypassing fsync(). The documentation is heavily laced with warnings.
3045 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3047 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3048 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3049 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3050 including adding "make clean"; (3) Added -fPIC when compiling the test
3051 dynamically loaded module, to get rid of a warning.
3053 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3054 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3055 = 0s. The bug is otherwise harmless.
3057 PH/07 There was a bug in the dovecot authenticator such that the value of
3058 $auth1 could be overwritten, and so not correctly preserved, after a
3059 successful authentication. This usually meant that the value preserved by
3060 the server_setid option was incorrect.
3062 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3064 PH/09 Installed PCRE release 7.0.
3066 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3067 run for batched SMTP input. It is now run at the start of every message
3068 in the batch. While fixing this I discovered that the process information
3069 (output by running exiwhat) was not always getting set for -bs and -bS
3070 input. This is fixed, and it now also says "batched" for BSMTP.
3072 PH/11 Added control=no_pipelining.
3074 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3075 patch, slightly modified), and move the expansion of helo_data till after
3076 the connection is made in the smtp transport (so it can use these
3079 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3081 PH/14 Added log_selector = +pid.
3083 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3085 PH/16 Add ${if forany and ${if forall.
3087 PH/17 Added dsn_from option to vary the From: line in DSNs.
3089 PH/18 Flush SMTP output before performing a callout, unless control =
3090 no_callout_flush is set.
3092 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3093 was true (the default) a successful delivery failed to delete the retry
3094 item, thus causing premature timeout of the address. The bug is now
3097 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3099 PH/21 Long custom messages for fakedefer and fakereject are now split up
3100 into multiline responses in the same way that messages for "deny" and
3101 other ACL rejections are.
3103 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3104 with slight modification.
3106 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3107 draft, changing the syntax and factoring some duplicate code".
3109 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3110 for deliveries of the second and subsequent messages over the same SMTP
3113 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3114 ${reduce, with only minor "tidies".
3116 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3118 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3119 expansion side effects.
3121 PH/27 When a message times out after an over-quota error from an Exim-imposed
3122 quota, the bounce message says "mailbox is full". This message was not
3123 being given when it was a system quota that was exceeded. It now should
3126 MH/03 Made $recipients available in local_scan(). local_scan() already has
3127 better access to the recipient list through recipients_list[], but
3128 $recipients can be useful in postmaster-provided expansion strings.
3130 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3131 in the case of a MAIL command with additional options following the
3132 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3133 were accidentally chopped off.
3135 PH/29 SMTP synchronization checks are implemented when a command is read -
3136 there is a check that no more input is waiting when there shouldn't be
3137 any. However, for some commands, a delay in an ACL can mean that it is
3138 some time before the response is written. In this time, more input might
3139 arrive, invalidly. So now there are extra checks after an ACL has run for
3140 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3141 pipelining has not been advertised.
3143 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3145 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3146 returned in the SMTP response when smtp_return_error_details was false.
3147 This has been fixed.
3149 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3150 instead of the C I/O that was originally supplied, because problems were
3151 reported on Solaris.
3153 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3154 Exim which did not show up earlier: it was assuming that a call to
3155 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3156 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3157 was a macro that became an assignment, so it seemed to work. This has
3158 changed to a proper function call with a void return, hence the compile
3159 error. Exim's code has been fixed.
3161 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3164 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3166 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3168 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3169 because it turns out to be too big or there is a timeout) there is no
3170 "Completed" line in the log. When some messages of this type were
3171 selected by exigrep, they were listed as "not completed". Others were
3172 picked up by some special patterns. I have improved the selection
3173 criteria to be more general.
3175 PH/38 The host_find_failed option in the manualroute router can now be set
3176 to "ignore", to completely ignore a host whose IP address cannot be
3177 found. If all hosts are ignored, the behaviour is controlled by the new
3178 host_all_ignored option.
3180 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3181 homing or because of multiple MX records with /mx) generated more than
3182 one IP address, and the following item turned out to be the local host,
3183 all the secondary addresses of the first item were incorrectly removed
3184 from the list, along with the local host and any following hosts (which
3185 is what is supposed to happen).
3187 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3188 whoever called Exim into the -H file. In the case of the daemon it was
3189 behaving confusingly. When first started, it used values for whoever
3190 started the daemon, but after a SIGHUP it used the Exim user (because it
3191 calls itself on a restart). I have changed the code so that it now always
3194 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3195 message are rejected with the same error (e.g. no authentication or bad
3196 sender address), and a DATA command is nevertheless sent (as can happen
3197 with PIPELINING or a stupid MUA), the error message that was given to the
3198 RCPT commands is included in the rejection of the DATA command. This is
3199 intended to be helpful for MUAs that show only the final error to their
3202 PH/42 Another patch from the Sieve maintainer.
3204 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3205 Eximstats - Fixed some broken HTML links and added missing column headers
3207 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3208 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3210 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3212 PH/43 Yet another patch from the Sieve maintainer.
3214 PH/44 I found a way to check for a TCP/IP connection going away before sending
3215 the response to the final '.' that terminates a message, but only in the
3216 case where the client has not sent further data following the '.'
3217 (unfortunately, this is allowed). However, in many cases there won't be
3218 any further data because there won't be any more messages to send. A call
3219 to select() can be used: if it shows that the input is "ready", there is
3220 either input waiting, or the socket has been closed. An attempt to read
3221 the next input character can distinguish the two cases. Previously, Exim
3222 would have sent an OK response which the client would never have see.
3223 This could lead to message repetition. This fix should cure that, at
3224 least in a lot of common cases.
3226 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3227 advertised in response to EHLO.
3233 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3234 fixed by 4.65/MH/01 (is this a record?) are fixed:
3236 (i) An empty string was always treated as zero by the numeric comparison
3237 operators. This behaviour has been restored.
3239 (ii) It is documented that the numeric comparison operators always treat
3240 their arguments as decimal numbers. This was broken in that numbers
3241 starting with 0 were being interpreted as octal.
3243 While fixing these problems I realized that there was another issue that
3244 hadn't been noticed. Values of message_size_limit (both the global option
3245 and the transport option) were treated as octal if they started with 0.
3246 The documentation was vague. These values are now always treated as
3247 decimal, and I will make that clear in the documentation.
3253 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3254 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3257 MH/01 Don't check that the operands of numeric comparison operators are
3258 integers when their expansion is in "skipping" mode (fixes bug
3259 introduced by 4.64-PH/07).
3261 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3262 child addresses, Exim now panics and dies. Previously, because the count
3263 is held in a short int, deliveries were likely to be lost. As such a
3264 large number of recipients for a single message is ridiculous
3265 (performance will be very, very poor), I have chosen to impose a limit
3266 rather than extend the field.
3272 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3273 leftover -K file (the existence of which was triggered by #402).
3274 While we were at it, introduced process PID as part of the -K
3275 filename. This should rule out race conditions when creating
3278 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3279 processing considerably. Previous code took too long for large mails,
3280 triggering a timeout which in turn triggers #401.
3282 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3283 in the DK code in transports.c. sendfile() is not really portable,
3284 hence the _LINUX specificness.
3286 TF/01 In the add_headers option to the mail command in an Exim filter,
3287 there was a bug that Exim would claim a syntax error in any
3288 header after the first one which had an odd number of characters
3291 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3292 callout verification, Exim cached a "reject" for the entire domain. This
3293 is correct for most verifications, but it is not correct for a recipient
3294 verification with use_sender or use_postmaster set, because in that case
3295 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3296 case of MAIL FROM:<> rejection from other early rejections (e.g.
3297 rejection of HELO). When verifying a recipient using a non-null MAIL
3298 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3299 Whatever the result of the callout, the value of the domain cache is
3300 left unchanged (for any other kind of callout, getting as far as trying
3301 RCPT means that the domain itself is ok).
3303 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3306 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3307 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3308 session, the daemon ignores ECONNECT errors and logs others; it now
3309 ignores EPIPE as well.
3311 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3312 (quoted-printable decoding).
3314 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3315 later the small subsequent patch to fix an introduced bug.
3317 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3319 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3321 PH/08 An error is now given if message_size_limit is specified negative.
3323 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3324 to be given (somewhat) arbitrary names.
3326 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3329 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3330 miscellaneous code fixes
3332 PH/10 Added the log_reject_target ACL modifier to specify where to log
3335 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3336 hostname. This is wrong, because it relates to the incoming message (and
3337 probably the interface on which it is arriving) and not to the outgoing
3338 callout (which could be using a different interface). This has been
3339 changed to use the value of the helo_data option from the smtp transport
3340 instead - this is what is used when a message is actually being sent. If
3341 there is no remote transport (possible with a router that sets up host
3342 addresses), $smtp_active_hostname is used.
3344 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3345 tweaks were necessary in order to get it to work (see also 21 below):
3346 (a) The code assumed that strncpy() returns a negative number on buffer
3347 overflow, which isn't the case. Replaced with Exim's string_format()
3349 (b) There were several signed/unsigned issues. I just did the minimum
3350 hacking in of casts. There is scope for a larger refactoring.
3351 (c) The code used strcasecmp() which is not a standard C function.
3352 Replaced with Exim's strcmpic() function.
3353 (d) The code set only $1; it now sets $auth1 as well.
3354 (e) A simple test gave the error "authentication client didn't specify
3355 service in request". It would seem that Dovecot has changed its
3356 interface. Fortunately there's a specification; I followed it and
3357 changed what the client sends and it appears to be working now.
3359 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3362 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3363 address A is aliased to B and C, where B exists and C does not. Without
3364 -v the output is "A verified" because verification stops after a
3365 successful redirection if more than one address is generated. However,
3366 with -v the child addresses are also verified. Exim was outputting "A
3367 failed to verify" and then showing the successful verification for C,
3368 with its parentage. It now outputs "B failed to verify", showing B's
3369 parentage before showing the successful verification of C.
3371 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3372 look up a TXT record in a specific list after matching in a combined
3375 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3376 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3377 they consult the DNS. I had assumed they would set it the way they
3378 wanted; and indeed my experiments on Linux seem to show that in some
3379 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3380 To be on the safe side, however, I have now made the interface to
3381 host_find_byname() similar to host_find_bydns(), with an argument
3382 containing the DNS resolver options. The host_find_byname() function now
3383 sets these options at its start, just as host_find_bydns() does. The smtp
3384 transport options dns_qualify_single and dns_search_parents are passed to
3385 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3386 of host_find_byname() use the default settings of RES_DEFNAMES
3387 (qualify_single) but not RES_DNSRCH (search_parents).
3389 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3390 spool_read_header() do less string testing, by means of a preliminary
3391 switch on the second character of optional "-foo" lines. (This is
3392 overdue, caused by the large number of possibilities that now exist.
3393 Originally there were few.) While I was there, I also converted the
3394 str(n)cmp tests so they don't re-test the leading "-" and the first
3395 character, in the hope this might squeeze out yet more improvement.
3397 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3398 flag allowing group syntax was set by the header_syntax check but not
3399 turned off, possible causing trouble later; (2) The flag was not being
3400 set at all for the header_verify test, causing "group"-style headers to
3401 be rejected. I have now set it in this case, and also caused header_
3402 verify to ignore an empty address taken from a group. While doing this, I
3403 came across some other cases where the code for allowing group syntax
3404 while scanning a header line wasn't quite right (mostly, not resetting
3405 the flag correctly in the right place). These bugs could have caused
3406 trouble for malformed header lines. I hope it is now all correct.
3408 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3409 with the "reply" argument non-NULL. The code, however (which originally
3410 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3411 but it didn't always do it. This confused somebody who was copying the
3412 code for some other use. I have removed all the tests.
3414 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3415 feature that was used to support insecure browsers during the U.S. crypto
3416 embargo. It requires special client support, and Exim is probably the
3417 only MTA that supported it -- and would never use it because real RSA is
3418 always available. This code has been removed, because it had the bad
3419 effect of slowing Exim down by computing (never used) parameters for the
3420 RSA_EXPORT functionality.
3422 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3423 authenticator to fail if there's a tab character in the incoming data
3424 (there should never be unless someone is messing about, as it's supposed
3425 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3426 if the connection is using TLS or if the remote IP is the same as the
3427 local IP, and the "valid-client-cert option" if a client certificate has
3430 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3431 authenticators. This can be used for authorization after authentication
3432 succeeds. (In the case of plaintext, it servers for both authentication
3435 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3436 if any retry times were supplied.
3438 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3439 connection, where there is no client IP address to check. In this
3440 situation, the verify now always succeeds.
3442 PH/25 Applied John Jetmore's -Mset patch.
3444 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3446 PH/27 In a string expansion for a processed (not raw) header when multiple
3447 headers of the same name were present, leading whitespace was being
3448 removed from all of them, but trailing whitespace was being removed only
3449 from the last one. Now trailing whitespace is removed from each header
3450 before concatenation. Completely empty headers in a concatenation (as
3451 before) are ignored.
3453 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3454 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3456 PH/29 [Removed. This was a change that I later backed out, and forgot to
3457 correct the ChangeLog entry (that I had efficiently created) before
3458 committing the later change.]
3460 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3461 address errors (4xx response to RCPT) over the same connection as other
3462 messages routed to the same hosts. Such deliveries are always "forced",
3463 so retry times are not inspected. This resulted in far too many retries
3464 for the affected addresses. The effect occurred only when there were more
3465 hosts than the hosts_max_try setting in the smtp transport when it had
3466 the 4xx errors. Those hosts that it had tried were not added to the list
3467 of hosts for which the message was waiting, so if all were tried, there
3468 was no problem. Two fixes have been applied:
3470 (i) If there are any address or message errors in an SMTP delivery, none
3471 of the hosts (tried or untried) are now added to the list of hosts
3472 for which the message is waiting, so the message should not be a
3473 candidate for sending over the same connection that was used for a
3474 successful delivery of some other message. This seems entirely
3475 reasonable: after all the message is NOT "waiting for some host".
3476 This is so "obvious" that I'm not sure why it wasn't done
3477 previously. Hope I haven't missed anything, but it can't do any
3478 harm, as the worst effect is to miss an optimization.
3480 (ii) If, despite (i), such a delivery is accidentally attempted, the
3481 routing retry time is respected, so at least it doesn't keep
3482 hammering the server.
3484 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3485 in ${readsocket because some servers need this prod.
3487 PH/32 Added some extra debug output when updating a wait-xxx database.
3489 PH/33 The hint "could be header name not terminated by colon", which has been
3490 given for certain expansion errors for a long time, was not being given
3491 for the ${if def:h_colon_omitted{... case.
3493 PH/34 The spec says: "With one important exception, whenever a domain list is
3494 being scanned, $domain contains the subject domain." There was at least
3495 one case where this was not true.
3497 PH/35 The error "getsockname() failed: connection reset by peer" was being
3498 written to the panic log as well as the main log, but it isn't really
3499 panic-worthy as it just means the connection died rather early on. I have
3500 removed the panic log writing for the ECONNRESET error when getsockname()
3503 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3504 runs only) independently of the message's sender address. This meant
3505 that, if the 4xx error was in fact related to the sender, a different
3506 message to the same recipient with a different sender could confuse
3507 things. In particular, this can happen when sending to a greylisting
3508 server, but other circumstances could also provoke similar problems.
3509 I have changed the default so that the retry time for these errors is now
3510 based a combination of the sender and recipient addresses. This change
3511 can be overridden by setting address_retry_include_sender=false in the
3514 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3515 remote server are returned as part of bounce messages. This was not
3516 happening for LMTP over a pipe (the lmtp transport), but now it is the
3517 same for both kinds of LMTP.
3519 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3520 in header lines that were in fact CNAMEs. This is no longer the case.
3522 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3523 and queue runs started by the daemon processed all messages. This has
3524 been fixed so that -R and -S can now usefully be given with -q<time>.
3526 PH/40 Import PCRE release 6.7 (fixes some bugs).
3528 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3530 PH/42 Give an error if -q is specified more than once.
3532 PH/43 Renamed the variables $interface_address and $interface_port as
3533 $received_ip_address and $received_port, to make it clear that these
3534 values apply to message reception, and not to the outgoing interface when
3535 a message is delivered. (The old names remain recognized, of course.)
3537 PH/44 There was no timeout on the connect() call when using a Unix domain
3538 socket in the ${readsocket expansion. There now is.
3540 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3541 be meaningful with "accept".
3543 SC/01 Eximstats V1.43
3544 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3546 SC/02 Eximstats V1.44
3547 Use a glob alias rather than an array ref in the generated
3548 parser. This improves both readability and performance.
3550 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3551 Collect SpamAssassin and rejection statistics.
3552 Don't display local sender or destination tables unless
3553 there is data to show.
3554 Added average volumes into the top table text output.
3556 SC/04 Eximstats V1.46
3557 Collect data on the number of addresses (recipients)
3558 as well as the number of messages.
3560 SC/05 Eximstats V1.47
3561 Added 'Message too big' to the list of mail rejection
3562 reasons (thanks to Marco Gaiarin).
3564 SC/06 Eximstats V1.48
3565 Mainlog lines which have GMT offsets and are too short to
3566 have a flag are now skipped.
3568 SC/07 Eximstats V1.49 (Alain Williams)
3569 Added the -emptyok flag.
3571 SC/08 Eximstats V1.50
3572 Fixes for obtaining the IP address from reject messages.
3574 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3575 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3576 whitespace changes from 4.64-PH/27
3578 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3581 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3582 are found, allow negative numbers in numeric criteria)
3584 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3586 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3587 to match changes made in 4.64-PH/43
3589 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3591 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3592 "domain" is actually an IP address, and force a failure. This locks out
3593 those revolvers/nameservers that support "A-for-A" lookups, in
3594 contravention of the specifications.
3596 PH/48 When a host name was looked up from an IP address, and the subsequent
3597 forward lookup of the name timed out, the host name was left in
3598 $sender_host_name, contrary to the specification.
3600 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3601 restricted to single-key lookups, Exim was not diagnosing an error if
3602 * or *@ was used with a query-style lookup.
3604 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3606 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3607 long ago, but noone interested enough thought of it. Let's just say that
3608 the "1.1" means that there are some new functions that weren't there at
3609 some point in the past.
3611 PH/51 Error processing for expansion failure of helo_data from an smtp
3612 transport during callout processing was broken.
3614 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3615 tested/used via the -bh/-bhc/-bs options.
3617 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3618 bug, fixed in subsequent PCRE releases).
3620 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3621 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3623 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3629 SC/01 Use a glob alias rather than an array ref in eximstats generated
3630 parser. This improves both readability and performance.
3632 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3633 Don't display local sender or destination tables in eximstats unless
3634 there is data to show.
3635 Added average volumes into the eximstats top table text output.
3637 SC/03 Collect data on the number of addresses (recipients) as well
3638 as the number of messages in eximstats.
3640 TF/01 Correct an error in the documentation for the redirect router. Exim
3641 does (usually) call initgroups() when daemonizing.
3643 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3644 with consistent privilege compared to when running as a daemon.
3646 TF/03 Note in the spec that $authenticated_id is not set for local
3647 submissions from trusted users.
3649 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3650 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3652 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3653 by adding some example configuration directives to the default
3654 configuration file. A little bit of work is required to uncomment the
3655 directives and define how usernames and passwords are checked, but
3656 there is now a framework to start from.
3658 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3659 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3660 without this. I don't know how relevant this is to other LDAP libraries.
3662 PH/02 Add the verb name to the "unknown ACL verb" error.
3664 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3666 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3668 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3669 directory not expanded when it should be if an expanded home directory
3670 was set for the address (which is overridden by the transport).
3672 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3675 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3676 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3677 because it is too late at that time, and has no effect.
3679 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3680 security issue with \' (bugzilla #107). I could not use the
3681 PQescapeStringConn() function, because it needs a PGconn value as one of
3684 PH/08 When testing addresses using -bt, indicate those final addresses that
3685 are duplicates that would not cause an additional delivery. At least one
3686 person was confused, thinking that -bt output corresponded to deliveries.
3687 (Suppressing duplicates isn't a good idea as you lose the information
3688 about possibly different redirections that led to the duplicates.)
3690 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3691 systems where poll() doesn't work, in particular OS X.
3693 PH/10 Added more information to debugging output for retry time not reached.
3695 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3696 operations in malware.c.
3698 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3701 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3702 syslog_duplication set false, log lines that would normally be written
3703 both the the main log and to the reject log were not written to syslog at
3706 PH/14 In the default configuration, change the use of "message" in ACL warn
3707 statements to "add_header".
3709 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3710 not followed by a command (e.g. "seen endif").
3712 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3713 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3716 PH/17 Added extra conditions to the default value of delay_warning_condition
3720 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3721 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3722 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3725 The Auto-Submitted: and various List- headers are standardised, whereas I
3726 don't think Precedence: ever was.
3728 PH/18 Refactored debugging code in route_finduser() to show more information,
3729 in particular, the error code if getpwnam() issues one.
3731 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3732 This is apparently needed in addition to the PH/07 change above to avoid
3733 any possible encoding problems.
3735 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3736 but not after initializing Perl.
3738 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3739 output them only if debugging. By default they are written stderr,
3740 apparently, which is not desirable.
3742 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3745 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3748 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3750 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3751 authenticated or an ident call has been made. Suppress the default
3752 values for $authenticated_id and $authenticated_sender (but permit -oMai
3753 and -oMas) when testing with -bh.
3755 PH/24 Re-jigged the order of the tests in the default configuration so that the
3756 tests for valid domains and recipients precede the DNS black list and CSA
3757 tests, on the grounds that those ones are more expensive.
3759 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3760 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3761 This bug exists in every version of Exim that I still have, right back to
3764 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3765 However, an attempt to turn on case-sensitivity in a regex key by
3766 including (?-i) didn't work because the subject string was already
3767 lowercased, and the effects were non-intuitive. It turns out that a
3768 one-line patch can be used to allow (?-i) to work as expected.
3774 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3775 other effects) broke the use of negated acl sub-conditions.
3777 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3780 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3781 "Deny" causes Exim to reject the incoming connection with a 554 error.
3782 Unfortunately, if there is a major crisis, such as a disk failure,
3783 tcp-wrappers gives "deny", whereas what one would like would be some
3784 kind of temporary error. A kludge has been added to help with this.
3785 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3786 554 error is used if errno is still zero or contains ENOENT (which occurs
3787 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3790 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3792 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3793 errors. Otherwise a message that provokes a temporary error (when other
3794 messages do not) can cause a whole host to time out.
3796 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3797 addresses were routed directly to files or pipes from a redirect router.
3798 File deliveries just didn't batch; pipe deliveries might have suffered
3801 PH/06 A failure to get a lock for a hints database would erroneously always say
3802 "Failed to get write lock", even when it was really a read lock.
3804 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3805 of 0600. This has been changed to use the value of the lockfile_mode
3806 option (which defaults to 0600).
3808 PH/08 Applied small patch from the Sieve maintainer.
3810 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3811 folder from quota calculations, a direct delivery into this folder messed
3812 up the contents of the maildirsize file. This was because the regex was
3813 used only to exclude .Trash (or whatever) when the size of the mailbox
3814 was calculated. There was no check that a delivery was happening into an
3815 excluded directory. This bug has been fixed by ignoring all quota
3816 processing for deliveries into excluded directories.
3818 PH/10 Added the maildirfolder_create_regex option to appendfile.
3824 PH/01 The code for finding all the local interface addresses on a FreeBSD
3825 system running IPv6 was broken. This may well have applied to all BSD
3826 systems, as well as to others that have similar system calls. The broken
3827 code found IPv4 interfaces correctly, but gave incorrect values for the
3828 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3829 that it would not match correctly against @[] and not recognize the IPv6
3832 PH/02 The ipliteral router was not recognizing addresses of the form user@
3833 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3835 PH/03 Added disable_ipv6.
3837 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3838 decoded form, because it is most often used to construct To: headers
3839 lines in autoreplies, and the decoded form may well be syntactically
3840 invalid. However, $reply_address has leading white space removed, and all
3841 newlines turned into spaces so that the autoreply transport does not
3844 PH/05 If group was specified without a user on a router, and no group or user
3845 was specified on a transport, the group from the router was ignored.
3847 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3848 for visible compile-time settings that can be used to change these
3849 numbers, for those that want even more. Backwards compatibility with old
3850 spool files has been maintained. However, going back to a previous Exim
3851 release will lost any variables that are in spool files.
3853 PH/07 Two small changes when running in the test harness: increase delay when
3854 passing a TCP/IP connection to a new process, in case the original
3855 process has to generate a bounce, and remove special handling of
3856 127.0.0.2 (sic), which is no longer necessary.
3858 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3859 be the same on different OS.
3861 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3864 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3865 whether --show-vars was specified or not
3867 JJ/02 exipick: Added support for new ACL variable spool format introduced
3870 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3871 syntactically invalid From: or Reply-to: line, and a filter used this to
3872 generate an autoreply, and therefore failed to obtain an address for the
3873 autoreply, Exim could try to deliver to a non-existent relative file
3874 name, causing unrelated and misleading errors. What now happens is that
3875 it logs this as a hard delivery error, but does not attempt to create a
3878 PH/11 The exinext utility has a -C option for testing purposes, but although
3879 the given file was scanned by exinext itself; it wasn't being passed on
3880 when Exim was called.
3882 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3883 an end-of-file indication when reading a command response.
3885 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3886 compiled. In many other places in Exim, IPv6 addresses are always
3887 recognized, so I have changed this. It also means that IPv4 domain
3888 literals of the form [IPV4:n.n.n.n] are now always recognized.
3890 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3891 used if the router is not running as root, for example, when verifying at
3892 ACL time, or when using -bh. The debugging output from this situation was
3893 non-existent - all you got was a failure to exec. I have made two
3896 (a) Failures to set uid/gid, the current directory, or a process leader
3897 in a subprocess such as that created by queryprogram now generate
3898 suitable debugging output when -d is set.
3900 (b) The queryprogram router detects when it is not running as root,
3901 outputs suitable debugging information if -d is set, and then runs
3902 the subprocess without attempting to change uid/gid.
3904 PH/15 Minor change to Makefile for building test_host (undocumented testing
3907 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3908 additional section of a DNS packet that returns MX or SRV records.
3909 Instead, it always explicitly searches for A/AAAA records. This avoids
3910 major problems that occur when a DNS server includes only records of one
3911 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3912 fixed another bug: if SRV records were looked up and the corresponding
3913 address records were *not* found in the additional section, the port
3914 values from the SRV records were lost.
3916 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3917 using the correct key (the original address) when searching the retry
3918 rules in order to find which one to use for generating the retry hint.
3920 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3921 adding the default one. Similarly, if it contains a Reply-To: header, the
3922 errors_reply_to option, if set, is not used.
3924 PH/19 When calculating a retry time, Exim used to measure the "time since
3925 failure" by looking at the "first failed" field in the retry record. Now
3926 it does not use this if it is later than than the arrival time of the
3927 message. Instead it uses the arrival time. This makes for better
3928 behaviour in cases where some deliveries succeed, thus re-setting the
3929 "first failed" field. An example is a quota failure for a huge message
3930 when small messages continue to be delivered. Without this change, the
3931 "time since failure" will always be short, possible causing more frequent
3932 delivery attempts for the huge message than are intended.
3933 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3935 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3936 $1, $2, $3) because the numerical variables can be reset during some
3937 expansion items (e.g. "match"), thereby losing the authentication data.
3939 PH/21 Make -bV show the size of off_t variables so that the test suite can
3940 decide whether to run tests for quotas > 2G.
3942 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3943 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3944 filecount value is greater than 2G or if a quota value is greater than 2G
3945 on a system where the size of off_t is not greater than 4, a panic error
3948 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3949 never match. The debug and -bh output now contains an explicit error
3950 message indicating a malformed IPv4 address or mask.
3952 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3953 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3954 PH/23 above applies.
3956 PH/25 Do not write to syslog when running in the test harness. The only
3957 occasion when this arises is a failure to open the main or panic logs
3958 (for which there is an explicit test).
3960 PH/26 Added the /no_tell option to "control=freeze".
3962 PH/27 If a host name lookup failed very early in a connection, for example, if
3963 the IP address matched host_lookup and the reverse lookup yielded a name
3964 that did not have a forward lookup, an error message of the form "no IP
3965 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3966 could be logged. Now it outputs the IP address instead of "NULL".
3968 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3969 allows the sender and the authenticated sender to be set when
3970 submitting a message from within Exim. Since child_open_exim() is
3971 documented for local_scan(), the new function should be too.
3973 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3974 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3975 results in an empty string is now treated as unset.
3977 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3979 PH/31 Added sender_verify_fail logging option.
3981 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3982 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3983 tidied the source and removed it altogether.
3985 PH/33 When a queue run was abandoned because the load average was too high, a
3986 log line was always written; now it is written only if the queue_run log
3987 selector is set. In addition, the log line for abandonment now contains
3988 information about the queue run such as the pid. This is always present
3989 in "start" and "stop" lines but was omitted from the "abandon" line.
3991 PH/34 Omit spaces between a header name and the colon in the error message that
3992 is given when verify = headers_syntax fails (if there are lots of them,
3993 the message gets confusing).
3995 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3996 names, as there are now some PTR records that contain slashes. This check
3997 is only to protect against broken name servers that fall over on strange
3998 characters, so the fact that it applies to all lookups doesn't matter.
4000 PH/36 Now that the new test suite is complete, we can remove some of the
4001 special code in Exim that was needed for the old test suite. For example,
4002 sorting DNS records because real resolvers return them in an arbitrary
4003 order. The new test suite's fake resolver always returns records in the
4006 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4007 bounces) except when queue_only is set, to avoid logging races between
4008 the different processes.
4010 PH/38 Panic-die if .include specifies a non-absolute path.
4012 PH/39 A tweak to the "H" retry rule from its user.
4014 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4015 a label. They prevented compilation on older perls.
4017 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4018 a warning to be raised on newish perls.
4020 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4021 on queue. Changes to match documented behaviour of showing count of
4022 messages matching specified criteria.
4024 PH/40 Changed the default ident timeout from 30s to 5s.
4026 PH/41 Added support for the use of login_cap features, on those BSD systems
4027 that have them, for controlling the resources used by pipe deliveries.
4029 PH/42 The content-scanning code uses fopen() to create files in which to put
4030 message data. Previously it was not paying any attention to the mode of
4031 the files. Exim runs with umask(0) because the rest of the code creates
4032 files with open(), and sets the required mode explicitly. Thus, these
4033 files were ending up world-writeable. This was not a big issue, because,
4034 being within the spool directory, they were not world-accessible. I have
4035 created a function called modefopen, which takes an additional mode
4036 argument. It sets umask(777), creates the file, chmods it to the required
4037 mode, then resets the umask. All the relevant calls to fopen() in the
4038 content scanning code have been changed to use this function.
4040 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4041 to 24 hours. This avoids potential overflow problems when processing G
4042 and H retry rules. I suspect nobody ever tinkers with this value.
4044 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4046 PH/45 When the plaintext authenticator is running as a client, the server's
4047 challenges are checked to ensure they are valid base64 strings. By
4048 default, the authentication attempt is cancelled if an invalid string is
4049 received. Setting client_ignore_invalid_base64 true ignores these errors.
4050 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4051 they are received. Thus, the responses can be made to depend on the
4052 challenges. If an invalid string is ignored, an empty string is placed in
4055 PH/46 Messages that are created by the autoreply transport now contains a
4056 References: header, in accordance with RFCs 2822 and 3834.
4058 PH/47 Added authenticated_sender_force to the smtp transport.
4060 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4062 PH/49 Installed latest patch from the Sieve maintainer.
4064 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4065 was also set, the appendfile transport was unnecessarily scanning a
4066 directory of message files (e.g. for maildir delivery) to find the count
4067 of files (along with the size), even though it did not need this
4068 information. It now does the scan only if it needs to find either the
4069 size of the count of files.
4071 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4073 PH/52 Two bugs concerned with error handling when the smtp transport is
4076 (i) Exim was not creating retry information for temporary errors given
4077 for individual recipients after the DATA command when the smtp transport
4078 was used in LMTP mode. This meant that they could be retried too
4079 frequently, and not timed out correctly.
4081 (ii) Exim was setting the flag that allows error details to be returned
4082 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4083 recipients that were returned after the DATA command.
4085 PH/53 This is related to PH/52, but is more general: for any failing address,
4086 when detailed error information was permitted to be returned to the
4087 sender, but the error was temporary, then after the final timeout, only
4088 "retry timeout exceeded" was returned. Now it returns the full error as
4089 well as "retry timeout exceeded".
4091 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4092 do this, and (what is worse) MTAs that accept it.
4094 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4095 will now be deprecated.
4097 PH/56 New os.c-cygwin from the Cygwin maintainer.
4099 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4100 formats (previously only available in exim formats via -bpr, -bpru,
4101 and -bpra. Now also available in native and exiqgrep formats)
4103 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4104 with very large, slow to parse queues
4106 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4108 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4110 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4111 responses to authentication challenges, though it was showing the
4112 challenges; (ii) I've removed the CR characters from the debug output for
4115 PH/58 Allow for the insertion of a newline as well as a space when a string
4116 is turned into more than one encoded-word during RFC 2047 encoding. The
4117 Sieve code now uses this.
4119 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4120 data_4xx, lost_connection, tls_required.
4122 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4123 message was being sent as an SMTP response.
4125 PH/61 Add -l and -k options to exicyclog.
4127 PH/62 When verifying, if an address was redirected to one new address, so that
4128 verification continued, and the new address failed or deferred after
4129 having set something in $address_data, the value of $address_data was not
4130 passed back to the ACL. This was different to the case when no
4131 redirection occurred. The value is now passed back in both cases.
4133 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4134 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4135 use login_cap.h, so on its own it isn't the distinguishing feature. The
4136 new name refers directly to the setclassresources() function.
4138 PH/65 Added configuration files for NetBSD3.
4140 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4142 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4143 is preferred over IPv4.
4145 PH/68 The bounce_return_message and bounce_return_body options were not being
4146 honoured for bounces generated during the reception of non-SMTP messages.
4147 In particular, this applied to messages rejected by the ACL. This bug has
4148 been fixed. However, if bounce_return_message is true and bounce_return_
4149 body is false, the headers that are returned for a non-SMTP message
4150 include only those that have been read before the error was detected.
4151 (In the case of an ACL rejection, they have all been read.)
4153 PH/69 The HTML version of the specification is now built in a directory called
4154 spec_html instead of spec.html, because the latter looks like a path with
4155 a MIME-type, and this confuses some software.
4157 PH/70 Catch two compiler warnings in sieve.c.
4159 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4160 function verify_get_ident() calls ip_connect() to connect a socket, but
4161 if the "connect()" function timed out, ip_connect() used to close the
4162 socket. However, verify_get_ident() also closes the socket later, and in
4163 between Exim writes to the log, which may get opened at this point. When
4164 the socket was closed in ip_connect(), the log could get the same file
4165 descriptor number as the socket. This naturally causes chaos. The fix is
4166 not to close the socket in ip_connect(); the socket should be closed by
4167 the function that creates it. There was only one place in the code where
4168 this was missing, in the iplookup router, which I don't think anybody now
4169 uses, but I've fixed it anyway.
4171 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4172 well as to direct DNS lookups. Otherwise the handling of names in host
4173 lists is inconsistent and therefore confusing.
4179 PH/01 Two changes to the default runtime configuration:
4181 (1) Move the checks for relay_from_hosts and authenticated clients from
4182 after to before the (commented out) DNS black list checks.
4184 (2) Add control=submission to the relay_from_hosts and authenticated
4185 clients checks, on the grounds that messages accepted by these
4186 statements are most likely to be submissions.
4188 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4190 (1) Generate an error if the third argument for the ${prvs expansion is
4193 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4196 (3) Reset the variables that are obtained from the first argument of
4197 ${prvscheck and used in the second argument before leaving the code,
4198 because their memory is reclaimed, so using them afterwards may do
4201 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4202 one (it's much easier than Tom thought :-).
4204 (5) Because of (4), we can now allow for the use of $prvscheck_result
4205 inside the third argument.
4207 PH/03 For some reason, the default setting of PATH when running a command from
4208 a pipe transport was just "/usr/bin". I have changed it to
4211 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4212 anything to be listed in the output from -bV.
4214 PH/05 When a filter generated an autoreply, the entire To: header line was
4215 quoted in the delivery log line, like this:
4217 => >A.N.Other <ano@some.domain> <original@ddress> ...
4219 This has been changed so that it extracts the operative address. There
4220 may be more than one such address. If so, they are comma-separated, like
4223 => >ano@some.domain,ona@other.domain <original@ddress> ...
4225 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4226 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4227 not being looked up in the rDNS to get a host name, Exim was showing the
4228 IP address twice in Received: lines, even though the IP addresses were
4229 identical. For example:
4231 Received: from [1.2.3.4] (helo=[1.2.3.4])
4233 However, if the real host name was known, it was omitting the HELO data
4234 if it matched the actual IP address. This has been tidied up so that it
4235 doesn't show the same IP address twice.
4237 PH/07 When both +timestamp and +memory debugging was on, the value given by
4238 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4239 called by the debug printing, thereby overwriting the timestamp buffer.
4240 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4242 PH/08 When the original message was included in an autoreply transport, it
4243 always said "this is a copy of the message, including all the headers",
4244 even if body_only or headers_only was set. It now gives an appropriate
4247 PH/09 Applied a patch from the Sieve maintainer which:
4249 o fixes some comments
4250 o adds the (disabled) notify extension core
4251 o adds some debug output for the result of if/elsif tests
4252 o points to the current vacation draft in the documentation
4253 and documents the missing references header update
4257 o fixes a bug in processing the envelope test (when testing
4258 multiple envelope elements, the last element determined the
4261 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4262 Electronic Mail") by including:
4264 Auto-submitted: auto-generated
4266 in the messages that it generates (bounce messages and others, such as
4267 warnings). In the case of bounce messages for non-SMTP messages, there was
4268 also a typo: it was using "Auto_submitted" (underscore instead of
4269 hyphen). Since every message generated by Exim is necessarily in response
4270 to another message, thes have all been changed to:
4272 Auto-Submitted: auto-replied
4274 in accordance with these statements in the RFC:
4276 The auto-replied keyword:
4278 - SHOULD be used on messages sent in direct response to another
4279 message by an automatic process,
4281 - MUST NOT be used on manually-generated messages,
4283 - MAY be used on Delivery Status Notifications (DSNs) and Message
4284 Disposition Notifications (MDNs),
4286 - MUST NOT be used on messages generated by automatic or periodic
4287 processes, except for messages which are automatic responses to
4290 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4291 to the default Received: header definition.
4293 PH/12 Added log selector acl_warn_skipped (default on).
4295 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4296 variables because (a) they are in the wrong storage pool and (b) even if
4297 they were copied, it wouldn't work properly because of the caching.
4299 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4300 checking when decoding. Apparently there are clients that generate
4301 overlong encoded strings. Why am I not surprised?
4303 PH/15 If the first argument of "${if match_address" was not empty, but did not
4304 contain an "@" character, Exim crashed. Now it writes a panic log message
4305 and treats the condition as false.
4307 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4309 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4310 extension "envelope-auth". The code is finished and in agreement with
4311 other implementations, but there is no documentation so far and in fact,
4312 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4313 not changing the active code.
4315 Print executed "if" and "elsif" statements when debugging is used. This
4316 helps a great deal to understand what a filter does.
4318 Document more things not specified clearly in RFC3028. I had all this
4319 sorted out, when out of a sudden new issues came to my mind. Oops."
4321 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4324 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4325 canonical form (as documented). However, after a host name lookup from
4326 the IP address, check_host() was doing a simple string comparison with
4327 addresses acquired from the DNS when checking that the found name did
4328 have the original IP as one of its addresses. Since any found IPv6
4329 addresses are likely to be in abbreviated form, the comparison could
4330 fail. Luckily, there already exists a function for doing the comparison
4331 by converting both addresses to binary, so now that is used instead of
4332 the text comparison.
4334 PH/20 There was another similar case to PH/19, when a complete host name was
4335 given in a host list; looking up its IP address could give an abbreviated
4336 form, whereas the current host's name might or might not be abbreviated.
4337 The same fix has been applied.
4343 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4344 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4347 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4348 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4350 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4352 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4353 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4354 Also, for "Auto-Submitted", treat anything other than "no" as
4355 non-personal, in accordance with RFC 3834. (Previously it treated
4356 anything starting "auto-" as non-personal.)
4358 TF/01 The control=submission/name=... option had a problem with syntax
4359 errors if the name included a slash character. The /name= option
4360 now slurps the rest of the string, so it can include any characters
4361 but it must come last in the list of options (after /sender_retain
4364 PH/05 Some modifications to the interface to the fake nameserver for the new
4372 TK/01 Added the "success_on_redirect" address verification option. See
4373 NewStuff for rationale and an example.
4375 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4377 PH/02 Patch to exigrep to allow it to work on syslog lines.
4379 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4380 fread() to skip over the body file's header line, because in Cygwin the
4381 header line is locked and is inaccessible.
4383 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4384 co-exist for some time) to make it clear that it is the Exim ID that is
4385 referenced, not the Message-ID: header line.
4387 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4388 string_format() function, because snprintf() does not exist on all
4391 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4392 ${stat: expansion item.
4394 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4395 protocol synchronization error", to keep the pedants happy.
4397 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4398 well as for IRIX systems, when gcc is being used. See the host.c source
4401 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4403 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4406 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4407 transport and to the smtp transport in LMTP mode.
4409 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4411 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4412 run a filter in a subprocess. This could lead to confusion in subsequent
4413 lookups in the parent process. There should also be a search_tidyup() at
4414 the end of the subprocess.
4416 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4417 only if the host matched helo_try_verify_hosts, which caused the
4418 verification to occur when the EHLO/HELO command was issued. The ACL just
4419 tested the remembered result. Now, if a previous verification attempt has
4420 not happened, "verify = helo" does it there and then.
4422 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4424 TK/03 Fix log output including CR from clamd.
4426 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4427 exist provoked a memory error which could cause a segfault.
4429 PH/15 Installed PCRE 6.2
4431 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4433 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4434 of the problem. Specifically, suggested +O2 rather than +O1 for the
4437 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4439 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4440 record, the port was not correctly logged when the outgoing_port log
4441 selector was set (it logged the transort's default port).
4443 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4444 fallback_hosts, and "hosts" in the smtp transport.
4446 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4447 host errors such as "Connection refused".
4449 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4450 authentication with radiusclient 0.4.9:
4452 - Error returned from rc_read_config was caught wrongly
4453 - Username/password not passed on to radius server due to wrong length.
4455 The presumption is that some radiusclient API changes for 4.51/PH/17
4456 were not taken care of correctly. The code is still untested by me (my
4457 Linux distribution still has 0.3.2 of radiusclient), but it was
4458 contributed by a Radius user.
4460 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4461 expanding the "port" option of the smtp transport.
4463 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4464 while reading a MIME header. Thanks to Tom Hughes for a patch.
4466 PH/24 Include config.h inside local_scan.h so that configuration settings are
4469 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4470 that in an ACL for RCPT (for example), you can examine exactly what was
4473 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4474 commands, but it was not correctly comparing the address with the actual
4475 client host address. Thus, it would show the EHLO address in Received:
4476 header lines when this was not necessary.
4478 PH/27 Added the % operator to ${eval:}.
4480 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4481 it should be ignoring failures (because with -C, for example, it has lost
4482 privilege). It wasn't ignoring creation failures other than "already
4485 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4488 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4489 because an input line was too long, either on its own, or by virtue of
4490 too many continuations, the temporary file was not being removed, and the
4491 return code was incorrect.
4493 PH/31 Missing "BOOL" in function definition in filtertest.c.
4495 PH/32 Applied Sieve patches from the maintainer.
4497 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4499 PH/33 Added "verify = not_blind".
4501 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4502 Local/Makefile (with some defaults set). These are used in built scripts
4503 such as exicyclog, but they have never been used in the exim_install
4504 script (though there are many overriding facilities there). I have
4505 arranged that the exim_install script now takes note of these two
4508 PH/35 Installed configuration files for Dragonfly.
4510 PH/36 When a locally submitted message by a trusted user did not contain a
4511 From: header, and the sender address was obtained from -f or from an SMTP
4512 MAIL command, and the trusted user did not use -F to supply a sender
4513 name, $originator_name was incorrectly used when constructing a From:
4514 header. Furthermore, $originator_name was used for submission mode
4515 messages from external hosts without From: headers in a similar way,
4516 which is clearly wrong.
4518 PH/37 Added control=suppress_local_fixups.
4520 PH/38 When log_selector = +received_sender was set, and the addition of the
4521 sender made the log line's construction buffer exactly full, or one byte
4522 less than full, an overflow happened when the terminating "\n" was
4525 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4526 when the result of a list match is failure because a DNS lookup failed.
4528 PH/40 RM_COMMAND is now used in the building process.
4530 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4531 the "build-* directories that it finds.
4533 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4534 address in a domain literal was a prefix of an interface address.
4536 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4537 when verifying a sender address, unless rewrite_headers is false.
4539 PH/44 Wrote a long comment about why errors_to addresses are verified as
4540 recipients, not senders.
4542 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4543 the ratelimit ACL was added.
4545 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4547 PH/46 Added extra information about PostgreSQL errors to the error string.
4549 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4550 avoiding the need to install special zones in a real server. This is
4551 backwards compatible; if it can't find the fake resolver, it drops back.
4552 Thus, both old and new test suites can be run.
4554 TF/02 Added util/ratelimit.pl
4556 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4557 clock is set back in time.
4559 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4560 Candler <B.Candler@pobox.com>.
4562 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4563 OK for addresses that are the result of redirections.
4565 PH/48 A number of further additions for the benefit of the new test suite,
4566 including a fake gethostbyname() that interfaces to the fake DNS resolver
4569 TF/06 The fix for widen_domains has also been applied to qualify_single and
4570 search_parents which are the other dnslookup options that can cause
4573 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4576 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4578 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4579 service" if it encountered an MX record with an empty target hostname.
4580 The message is now "an MX or SRV record indicated no SMTP service".
4582 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4583 if the DNS of the sending site is misconfigured. This is quite a
4584 common situation. This change restores the behaviour of treating a
4585 helo verification defer as a failure.
4587 PH/51 If self=fail was set on a router, the bounce message did not include the
4588 actual error message.
4594 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4596 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4597 command itself ended in error, the underlying message about the transport
4598 filter timeout was being overwritten with the pipe command error. Now the
4599 underlying error message should be appended to the second error message.
4601 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4603 PH/02 Reduce the amount of output that "make" produces by default. Full output
4604 can still be requested.
4606 PH/03 The warning log line about a condition test deferring for a "warn" verb
4607 was being output only once per connection, rather than after each
4608 occurrence (because it was using the same function as for successful
4609 "warn" verbs). This seems wrong, so I have changed it.
4611 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4612 it should not have, which might have caused a crash in the right
4613 circumstances, but probably never did.
4615 PH/04 Installed a modified version of Tony Finch's patch to make submission
4616 mode fix the return path as well as the Sender: header line, and to
4617 add a /name= option so that you can make the user's friendly name appear
4620 TF/03 Added the control = fakedefer ACL modifier.
4622 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4623 Mark Lowes for thorough testing.
4625 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4627 TK/03 Merged latest SRS patch from Miles Wilton.
4629 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4630 with the definition in sysexits.h (which is #included earlier).
4631 Fortunately, Exim does not actually use EX_OK. The code used to try to
4632 preserve the sysexits.h value, by assuming that macro definitions were
4633 scanned for macro replacements. I have been disabused of this notion,
4634 so now the code just undefines EX_OK before #including unistd.h.
4636 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4637 in the smtp transport. When a block could not be written in a single
4638 write() function, the timeout was being re-applied to each part-write.
4639 This seems wrong - if the receiver was accepting one byte at a time it
4640 would take for ever. The timeout is now adjusted when this happens. It
4641 doesn't have to be particularly precise.
4643 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4644 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4646 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4647 without a domain if the check to <postmaster@domain> fails.
4649 SC/01 Eximstats: added -xls and the ability to specify output files
4650 (patch written by Frank Heydlauf).
4652 SC/02 Eximstats: use FileHandles for outputting results.
4654 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4656 SC/04 Eximstats: fixed display of large numbers with -nvr option
4658 SC/05 Eximstats: fixed merging of reports with empty tables.
4660 SC/06 Eximstats: added the -include_original_destination flag
4662 SC/07 Eximstats: removed tabs and trailing whitespace.
4664 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4666 TK/06 MBOX spool code: Add real "From " MBOX separator line
4667 so the .eml file is really in mbox format (even though
4668 most programs do not really care). Patch from Alex Miller.
4670 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4671 The latter is generated from $received_to and is only set if the
4672 message has one envelope recipient. SA can use these headers,
4673 obviously out-of-the-box. Patch from Alex Miller.
4675 PH/08 The ${def test on a variable was returning false if the variable's
4676 value was "0", contrary to what the specification has always said!
4677 The result should be true unless the variable is empty.
4679 PH/09 The syntax error of a character other than { following "${if
4680 def:variable_name" (after optional whitespace) was not being diagnosed.
4681 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4682 accidental colon was present, for example, could give incorrect results.
4684 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4685 result is used (not including appendfile, where other changes are about
4688 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4689 This involved changing a lot of size variables from int to off_t. It
4690 should work with maildirs and everything.
4692 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4693 spamd dying while we are connected to it.
4695 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4698 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4699 function for BDB 4.3.
4701 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4703 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4704 the sources for PCRE have been re-arranged and more files are now
4707 PH/15 The code I had for printing potentially long long variables in PH/11
4708 above was not the best (it lost precision). The length of off_t variables
4709 is now inspected at build time, and an appropriate printing format (%ld
4710 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4711 to be "long long int" or "long int". This is needed for the internal
4712 formatting function string_vformat().
4714 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4715 the configuration file to be ":syslog", then the script "guesses" where
4716 the logs files are, rather than using the compiled in default. In our
4717 case the guess is not the same as the compiled default, so the script
4718 suddenly stopped working when I started to use syslog. The patch checks
4719 to see if log_file_path is "". If so, it attempts to read it from exim
4720 with no configuration file to get the compiled in version, before it
4721 falls back to the previous guessing code."
4723 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4724 implementing BATV in an Exim configuration. See NewStuff for the gory
4727 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4728 Makefile that are specific to HP-UX.
4730 PH/18 If the "use_postmaster" option was set for a recipient callout together
4731 with the "random" option, the postmaster address was used as the MAIL
4732 FROM address for the random test, but not for the subsequent recipient
4733 test. It is now used for both.
4735 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4736 patch removes a few documentation additions to RFC 3028, because the
4737 latest draft now contains them. It adds the new en;ascii-case comparator
4738 and a new error check for 8bit text in MIME parts. Comparator and
4739 require names are now matched exactly. I enabled the subaddress
4740 extension, but it is not well tested yet (read: it works for me)."
4742 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4743 rework some of the code of TK/09 above to avoid the hardwired use of
4744 "%lld" and "long long". Replaced the call to snprintf() with a call to
4747 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4748 records point to non-existent hosts", "retry timeout exceeded", and
4749 "retry time not reached for any host after a long failure period".
4751 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4752 experimental DomainKeys support:
4754 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4755 (2) On an error such as an illegally used "control", the wrong name for
4756 the control was given.
4758 These problems did NOT occur unless DomainKeys support was compiled.
4760 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4762 PH/24 Added ${if match_ip condition.
4764 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4765 databases so that it will be absolutely obvious if a crash occurs in the
4766 DB library. This is a regular occurrence (often caused by mis-matched
4769 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4770 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4771 picked up on a user's system that detects such things. There doesn't seem
4772 to be a gcc warning option for this - only an attribute that has to be
4773 put on the function's prototype. It seems that in Fedora Core 4 they have
4774 set this on a number of new functions. No doubt there will be more in due
4777 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4778 specify a transport. However, if an address that was verified by such a
4779 router was the subject of a callout, Exim crashed because it tried to
4780 read the rcpt_include_affixes from the non-existent transport. Now it
4781 just assumes that the setting of that option is false. This bug was
4782 introduced by 4.51/PH/31.
4784 PH/28 Changed -d+all to exclude +memory, because that information is very
4785 rarely of interest, but it makes the output a lot bigger. People tend to
4786 do -d+all out of habit.
4788 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4789 code in os-type was giving problems when libc.so lives in lib64, like on
4792 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4793 aren't the modern standard, and it seems that some systems' include files
4794 don't always have them. Exim was already checking for some of the newer
4795 ones like T_AAAA, and defining it itself. I've added checks for all the
4796 record types that Exim uses.
4798 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4799 not automatically generating a new one, as it is supposed to. This
4800 prevented TLS from working. If the file did exist, but contained invalid
4801 data, a new version was generated, as expected. It was only the case of a
4802 non-existent file that was broken.
4804 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4805 with a change in libdomainkeys > 0.64.
4807 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4808 from DNS. If the selector record carries the flag, it now has
4809 precedence over the domain-wide flag.
4811 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4813 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4814 the use of a port name that isn't defined in /etc/services) occurred, the
4815 message was deferred as in a normal delivery, and thus remained on the
4816 spool, instead of being failed because of the mua_wrapper setting. This
4817 is now fixed, and I tidied up some of the mua_wrapper messages at the
4820 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4821 the messages in a hash of arrays rather than using individual hashes.
4822 This is a bit cleaner and results in dramatic memory savings, albeit
4823 at a slight CPU cost.
4825 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4826 as requested by Marc Sherman.
4828 SC/10 Eximstats: added histograms for user specified patterns as requested
4831 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4833 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4834 fopen() in the content-scanning modules that did not already have it.
4840 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4841 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4843 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4845 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4847 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4848 to test Sieve filters that use "vacation".
4850 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4851 that changes the way the GnuTLS parameters are stored in the cache file.
4852 The new format can be generated externally. For backward compatibility,
4853 if the data in the cache doesn't make sense, Exim assumes it has read an
4854 old-format file, and it generates new data and writes a new file. This
4855 means that you can't go back to an older release without removing the
4858 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4859 work if there are any delivery delays because "one_time" forces the
4860 parent to be marked "delivered", so its unseen clone is never tried
4861 again. For this reason, Exim now forbids the simultaneous setting of
4864 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4865 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4866 entry if you want to know the details. The fix, however, neglected to
4867 consider the case when local delivery batching is involved. The test for
4868 "previously delivered" was not happening when checking to see if an
4869 address could be batched with a previous (undelivered) one; under
4870 certain circumstances this could lead to multiple deliveries to the same
4873 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4874 in its include files, and this causes problems building Exim.
4876 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4877 header_syntax) but Exim was just ignoring anything given after a slash.
4878 In particular, this caused confusion with an attempt to use "verify =
4879 reverse_host_lookup/defer_ok". An error is now given when options are
4880 supplied for verify items that do not have them. (Maybe reverse_host_
4881 lookup should have a defer_ok option, but that's a different point.)
4883 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4884 defined by RFC 821) to 2048, because there were problems with some AUTH
4885 commands, and RFC 1869 says the size should be increased for extended
4886 SMTP commands that take arguments.
4888 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4891 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4892 "unknown" error; now it says that the functionality isn't in the binary.
4894 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4895 an address' error message when a string expansion fails (syntax or
4896 whatever). Otherwise the password may appear in the log. Following change
4897 PH/42 below, there is no longer a chance of it appearing in a bounce
4900 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4902 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4903 of its addresses was ever tried. (Bugzilla bug #2.)
4905 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4906 the result incorrectly in the debug output. (It correctly added a newline
4907 to what was transported.)
4909 TF/01 Added $received_time.
4911 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4912 commented out examples of how to interface to a virus scanner and to
4913 SpamAssassin. Also added commented examples of av_scanner and
4914 spamd_address settings.
4916 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4917 and controls are allowed in which ACLs. There were a couple of minor
4918 errors. Some of the entries in the conditions table (which is a table of
4919 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4920 negation of where the condition IS allowed.
4922 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4924 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4925 header file does not have a version number, so I've had to invent a new
4926 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4927 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4928 radiusclient), but it was contributed by a Radius user.
4930 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4931 files or directories, for OpenSSL.
4933 PH/19 When an Exim process that is running as root has to create an Exim log
4934 file, it does so in a subprocess that runs as exim:exim so as to get the
4935 ownership right at creation (otherwise, other Exim processes might see
4936 the file with the wrong ownership). There was no test for failure of this
4937 fork() call, which would lead to the process getting stuck as it waited
4938 for a non-existent subprocess. Forks do occasionally fail when resources
4939 run out. I reviewed all the other calls to fork(); they all seem to check
4942 PH/20 When checking for unexpected SMTP input at connect time (before writing
4943 the banner), Exim was not dealing correctly with a non-positive return
4944 from the read() function. If the client had disconnected by this time,
4945 the result was a log entry for a synchronization error with an empty
4946 string after "input=" when read() returned zero. If read() returned -1
4947 (an event I could not check), uninitialized data bytes were printed.
4948 There were reports of junk text (parts of files, etc) appearing after
4951 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4953 PH/22 Added support for macro redefinition, and (re)definition in between
4954 driver and ACL definitions.
4956 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4957 forgetting to use the resulting value; it was using the unexpanded value.
4959 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4960 hadn't been configured. The fix is from Juergen Kreileder, who
4961 understands it better than I do:
4963 "Here's what I see happening with three configured cyrus_sasl
4964 authenticators configured (plain, login, cram-md5):
4966 On startup auth_cyrus_sasl_init() gets called for each of these.
4967 This means three calls to sasl_listmech() without a specified mech_list.
4968 => SASL tests which mechs of all available mechs actually work
4969 => three warnings about OTP not working
4970 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4972 With the patch, sasl_listmech() also gets called three times. But now
4973 SASL's mech_list option is set to the server_mech specified in the the
4974 authenticator. Or in other words, the answer from sasl_listmech()
4975 gets limited to just the mech you're testing for (which is different
4977 => the return list contains just 'plain' or 'login', 'cram-md5' or
4978 nothing depending on the value of ob->server_mech.
4980 I've just tested the patch: Authentication still works fine,
4981 unavailable mechs specified in the exim configuration are still
4982 caught, and the auth.log warnings about OTP are gone."
4984 PH/25 When debugging is enabled, the contents of the command line are added
4985 to the debugging output, even when log_selector=+arguments is not
4988 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4989 answer is "GNU", and only if the return is "GNU/something" is the answer
4992 PH/27 $acl_verify_message is now set immediately after the failure of a
4993 verification in an ACL, and so is available in subsequent modifiers. In
4994 particular, the message can be preserved by coding like this:
4996 warn !verify = sender
4997 set acl_m0 = $acl_verify_message
4999 Previously, $acl_verify_message was set only while expanding "message"
5000 and "log_message" when a very denied access.
5002 PH/28 Modified OS/os.c-Linux with
5004 -#ifndef OS_LOAD_AVERAGE
5005 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5007 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5008 nomenclature these days.)
5010 PH/29 Installed patch from the Sieve maintainer that adds the options
5011 sieve_useraddress and sieve_subaddress to the redirect router.
5013 PH/30 In these circumstances:
5014 . Two addresses routed to the same list of hosts;
5015 . First host does not offer TLS;
5016 . First host accepts first address;
5017 . First host gives temporary error to second address;
5018 . Second host offers TLS and a TLS session is established;
5019 . Second host accepts second address.
5020 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5021 and peerdn, if requested) that were in fact used only for the second
5024 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5025 attention to any local part prefix or suffix that was matched by the
5026 router that accepted the address. It now behaves in the same way as it
5027 does for delivery: the affixes are removed from the local part unless
5028 rcpt_include_affixes is set on the transport.
5030 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5031 timeout during the DATA phase of an incoming message.
5033 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5034 applied a patch sanctioned by the Sieve maintainer.
5036 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5037 the uid or gid is negative. A case of a negative gid caused this to be
5038 noticed. The fix allows for either to be negative.
5040 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5041 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5044 PH/36 The change PH/12 above was broken. Fixed it.
5046 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5047 the grounds that routing the same address twice would always produce the
5048 same answer. This might have been true once, but it is certainly no
5049 longer true now. Routing a child address may depend on the previous
5050 routing that produced that child. Some complicated redirection strategies
5051 went wrong when messages had multiple recipients, and made Exim's
5052 behaviour dependent on the order in which the addresses were given.
5054 I have moved the duplicate checking until after the routing is complete.
5055 Exim scans the addresses that are assigned to local and remote
5056 transports, and removes any duplicates. This means that more work will be
5057 done, as duplicates will always all be routed, but duplicates are
5058 presumably rare, so I don't expect this is of any significance.
5060 For deliveries to pipes, files, and autoreplies, the duplicate checking
5061 still happens during the routing process, since they are not going to be
5064 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5065 It corrects a timeout issue with spamd. This is Ian's comment: "The
5066 background is that sometimes spamd either never reads data from a
5067 connection it has accepted, or it never writes response data. The exiscan
5068 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5069 blindly assumes that writes won't block so it may never time out."
5071 PH/39 Allow G after quota size as well as K and M.
5073 PH/40 The value set for $authenticated_id in an authenticator may not contain
5074 binary zeroes or newlines because the value is written to log lines and
5075 to spool files. There was no check on this. Now the value is run through
5076 the string_printing() function so that such characters are converted to
5077 printable escape sequences.
5079 PH/41 $message_linecount is a new variable that contains the total number of
5080 lines in the message. Compare $body_linecount, which is the count for the
5083 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5084 bounce and delay warning messages, except in certain special cases, which
5087 (a) An SMTP error message from a remote host;
5088 (b) A message specified in a :fail: redirection;
5089 (c) A message specified in a "fail" command in a system filter;
5090 (d) A message specified in a FAIL return from the queryprogram router;
5091 (e) A message specified by the cannot_route_message router option.
5093 In these cases only, Exim does include the error details in bounce and
5094 warning messages. There are also a few cases where bland messages such
5095 as "unrouteable address" or "local delivery error" are given.
5097 PH/43 $value is now also set for the "else" part of a ${run expansion.
5099 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5100 being worked on, but at least Exim now implements the latest version to
5103 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5104 process to complete was treated as a delivery failure, a timeout while
5105 writing the message to the pipe was logged, but erroneously treated as a
5106 successful delivery. Such timeouts include transport filter timeouts. For
5107 consistency with the overall process timeout, these timeouts are now
5108 treated as errors, giving rise to delivery failures by default. However,
5109 there is now a new Boolean option for the pipe transport called
5110 timeout_defer, which, if set TRUE, converts the failures into defers for
5111 both kinds of timeout. A transport filter timeout is now identified in
5114 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5115 systems where "make" and "gmake" are different, calling "gmake" at top
5116 level broke things. I've arranged for the value of $(MAKE) to be passed
5117 from the Makefile to this script so that it can call the same version of
5121 A note about Exim versions 4.44 and 4.50
5122 ----------------------------------------
5124 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5125 changes of various kinds. As a consequence, a big documentation update was
5126 needed. This delayed the release for rather longer than seemed good, especially
5127 in the light of a couple of (minor) security issues. Therefore, the changes
5128 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5129 release. So 4.44 and 4.50 are in effect two different branches that both start
5132 I have left the 4.50 change log unchanged; it contains all the changes since
5133 4.43. The change log for 4.44 is below; many of its items are identical to
5134 those for 4.50. This seems to be the most sensible way to preserve the
5135 historical information.
5141 1. Minor wording change to the doc/README.SIEVE file.
5143 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5144 computation of the current number of files was incorrect.
5146 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5147 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5150 4. Give more explanation in the error message when the command for a transport
5151 filter fails to execute.
5153 5. There are several places where Exim runs a non-Exim command in a
5154 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5155 was being done only for the command run by the queryprogram router. It is
5156 now done for all such subprocesses. The other cases are: ${run, transport
5157 filters, and the commands run by the lmtp and pipe transports.
5159 6. Added CONFIGURE_GROUP build-time option.
5161 7. Some older OS have a limit of 256 on the maximum number of file
5162 descriptors. Exim was using setrlimit() to set 1000 as a large value
5163 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5164 systems. I've change it so that if it can't get 1000, it tries for 256.
5166 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5167 was an oversight, and furthermore, ever since the addition of extra
5168 controls (e.g. 4.43/32), the checks on when to allow different forms of
5169 "control" were broken. There should now be diagnostics for all cases when a
5170 control that does not make sense is encountered.
5172 9. Added the /retain_sender option to "control=submission".
5174 10. $recipients is now available in the predata ACL (oversight).
5176 11. Tidy the search cache before the fork to do a delivery from a message
5177 received from the command line. Otherwise the child will trigger a lookup
5178 failure and thereby defer the delivery if it tries to use (for example) a
5179 cached ldap connection that the parent has called unbind on.
5181 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5182 of $address_data from the recipient verification was clobbered by the
5183 sender verification.
5185 13. The value of address_data from a sender verification is now available in
5186 $sender_address_data in subsequent conditions in the ACL statement.
5188 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5190 15. Added a new option "connect=<time>" to callout options, to set a different
5193 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5194 was its contents. (It was OK if the option was not defined at all.)
5196 17. A "Completed" log line is now written for messages that are removed from
5197 the spool by the -Mrm option.
5199 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5200 information about exactly what failed.
5202 19. Added -dd to debug only the daemon process.
5204 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5205 handles timeouts, both on the server side and network timeouts. Renamed the
5206 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5208 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5209 It is now set to "smtps".
5211 22. $host_address is now set to the target address during the checking of
5212 ignore_target_hosts.
5214 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5215 being passed; this would have caused $sender_host_name to have been used if
5216 matching the list had actually called for a host name (not very likely,
5217 since this list is usually IP addresses). A host name is now passed as
5220 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5221 code that specifies a non-restarting handler (typically sigaction() in
5222 modern systems) in an attempt to fix a rare and obscure crash bug.
5224 25. Narrowed the window for a race in the daemon that could cause it to ignore
5225 SIGCHLD signals. This is not a major problem, because they are used only to
5226 wake it up if nothing else does.
5228 26. A malformed maildirsize file could cause Exim to calculate negative values
5229 for the mailbox size or file count. Odd effects could occur as a result.
5230 The maildirsize information is now recalculated if the size or filecount
5233 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5234 support for a long time. Removed HAVE_SYS_VFS_H.
5236 28. Installed the latest version of exipick from John Jetmore.
5238 29. In an address list, if the pattern was not a regular expression, an empty
5239 subject address (from a bounce message) matched only if the pattern was an
5240 empty string. Non-empty patterns were not even tested. This was the wrong
5241 because it is perfectly reasonable to use an empty address as part of a
5242 database query. An empty address is now tested by patterns that are
5243 lookups. However, all the other forms of pattern expect the subject to
5244 contain a local part and a domain, and therefore, for them, an empty
5245 address still always fails if the pattern is not itself empty.
5247 30. Exim went into a mad DNS loop when attempting to do a callout where the
5248 host was specified on an smtp transport, and looking it up yielded more
5249 than one IP address.
5251 31. Re-factored the code for checking spool and log partition space into a
5252 function that finds that data and another that does the check. The former
5253 is then used to implement four new variables: $spool_space, $log_space,
5254 $spool_inodes, and $log_inodes.
5256 32. The RFC2047 encoding function was originally intended for short strings
5257 such as real names; it was not keeping to the 75-character limit for
5258 encoded words that the RFC imposes. It now respects the limit, and
5259 generates multiple encoded words if necessary. To be on the safe side, I
5260 have increased the buffer size for the ${rfc2047: expansion operator from
5263 33. It is now permitted to omit both strings after an "if" condition; if the
5264 condition is true, the result is "true". As before, when the second string
5265 is omitted, a false condition yields an empty string. This makes it less
5266 cumbersome to write custom ACL and router conditions.
5268 34. Failure to deliver a bounce message always caused it to be frozen, even if
5269 there was an errors_to setting on the router. The errors_to setting is now
5272 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5273 canonical form (fully expanded) before being placed in
5274 $sender_host_address.
5276 36. The table in the code that translates DNS record types into text (T_A to
5277 "A" for instance) was missing entries for NS and CNAME. It is just possible
5278 that this could have caused confusion if both these types were looked up
5279 for the same domain, because the text type is used as part of Exim's
5280 per-process caching. But the chance of anyone hitting this buglet seems
5283 37. The dnsdb lookup has been extended in a number of ways.
5285 (1) There is a new type, "zns", which walks up the domain tree until it
5286 finds some nameserver records. It should be used with care.
5288 (2) There is a new type, "mxh", which is like "mx" except that it returns
5289 just the host names, not the priorities.
5291 (3) It is now possible to give a list of domains (or IP addresses) to be
5292 looked up. The behaviour when one of the lookups defers can be
5293 controlled by a keyword.
5295 (4) It is now possible to specify the separator character for use when
5296 multiple records are returned.
5298 38. The dnslists ACL condition has been extended: it is now possible to supply
5299 a list of IP addresses and/or domains to be looked up in a particular DNS
5302 39. Added log_selector=+queue_time_overall.
5304 40. When running the queue in the test harness, wait just a tad after forking a
5305 delivery process, to get repeatability of debugging output.
5307 41. Include certificate and key file names in error message when GnuTLS fails
5308 to set them up, because the GnuTLS error message doesn't include the name
5309 of the failing file when there is a problem reading it.
5311 42. Allow both -bf and -bF in the same test run.
5313 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5315 44. The "Exiscan patch" is now merged into the mainline Exim source.
5317 45. Sometimes the final signoff response after QUIT could fail to get
5318 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5319 < 0 before doing a fflush(). This bug looks as though it goes back to the
5320 introduction of TLS in release 3.20, but "sometimes" must have been rare
5321 because the tests only now provoked it.
5323 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5324 (this can affect the format of dates).
5326 47. exim_tidydb, when checking for the continued existence of a message for
5327 which it has found a message-specific retry record, was not finding
5328 messages that were in split spool directories. Consequently, it was
5329 deleting retry records that should have stayed in existence.
5331 48. Steve fixed some bugs in eximstats.
5333 49. The SPA authentication driver was not abandoning authentication and moving
5334 on to the next authenticator when an expansion was forced to fail,
5335 contradicting the general specification for all authenticators. Instead it
5336 was generating a temporary error. It now behaves as specified.
5338 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5339 (the order specifies the preference for clients). The order is now AES256,
5340 AES128, 3DES, ARCFOUR128.
5342 51. Small patch to Sieve code - explicitly set From: when generating an
5345 52. Exim crashed if a remote delivery caused a very long error message to be
5346 recorded - for instance if somebody sent an entire SpamAssassin report back
5347 as a large number of 550 error lines. This bug was coincidentally fixed by
5348 increasing the size of one of Exim's internal buffers (big_buffer) that
5349 happened as part of the Exiscan merge. However, to be on the safe side, I
5350 have made the code more robust (and fixed the comments that describe what
5353 53. Now that there can be additional text after "Completed" in log lines (if
5354 the queue_time_overall log selector is set), a one-byte patch to exigrep
5355 was needed to allow it to recognize "Completed" as not the last thing in
5358 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5359 patch that reportedly fixes this has been added. I am not expert enough to
5360 create a test for it. This is what the patch creator wrote:
5362 "I found a little strange behaviour of ldap code when working with
5363 Windows 2003 AD Domain, where users was placed in more than one
5364 Organization Units. When I tried to give exim partial DN, the exit code
5365 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5366 But simultaneously result of request was absolutely normal ldap result,
5367 so I produce this patch..."
5369 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5370 REFERENCE, so I have modified the code to exclude the patch when that macro
5373 55. Some experimental protocols are using DNS PTR records for new purposes. The
5374 keys for these records are domain names, not reversed IP addresses. The
5375 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5376 leaves it alone. Component reversal etc. now happens only for IP addresses.
5379 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5381 57. Double the size of the debug message buffer (to 2048) so that more of very
5382 long debug lines gets shown.
5384 58. The exicyclog utility now does better if the number of log files to keep
5385 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5387 59. Two changes related to the smtp_active_hostname option:
5389 (1) $smtp_active_hostname is now available as a variable.
5390 (2) The default for smtp_banner uses $smtp_active_hostname instead
5391 of $primary_hostname.
5393 60. The host_aton() function is supposed to be passed a string that is known
5394 to be a valid IP address. However, in the case of IPv6 addresses, it was
5395 not checking this. This is a hostage to fortune. Exim now panics and dies
5396 if the condition is not met. A case was found where this could be provoked
5397 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5398 components; fortuitously, this particular loophole had already been fixed
5399 by change 4.50/55 above.
5401 If there are any other similar loopholes, the new check in host_aton()
5402 itself should stop them being exploited. The report I received stated that
5403 data on the command line could provoke the exploit when Exim was running as
5404 exim, but did not say which command line option was involved. All I could
5405 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5406 running as the user.
5409 61. There was a buffer overflow vulnerability in the SPA authentication code
5410 (which came originally from the Samba project). I have added a test to the
5411 spa_base64_to_bits() function which I hope fixes it.
5414 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5415 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5417 63. The daemon start-up calls getloadavg() while still root for those OS that
5418 need the first call to be done as root, but it missed one case: when
5419 deliver_queue_load_max is set with deliver_drop_privilege. This is
5420 necessary for the benefit of the queue runner, because there is no re-exec
5421 when deliver_drop_privilege is set.
5423 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5424 This has been fixed.
5426 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5427 was in use, was not putting the data itself into the right store pool;
5428 consequently, it could be overwritten for a subsequent message in the same
5429 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5432 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5434 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5435 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5436 the function were treating the return as a boolean value, which happened to
5437 work because 0=false and not-0=true, but is not correct code.
5439 68. The host_aton() function was not handling scoped IPv6 addresses (those
5440 with, for example, "%eth0" on the end) correctly.
5442 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5443 negated items (that is, ~something) in unsigned ints. Some compilers
5444 apparently mutter when there is no cast.
5446 70. If an address verification called from an ACL failed, and did not produce a
5447 user-specific message (i.e. there was only a "system" message), nothing was
5448 put in $acl_verify_message. In this situation, it now puts the system
5451 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5452 session; change 4.31/43 added the unwanted input to the log line - except
5453 that it did not do this in the start of session case. It now does.
5455 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5456 This is wrong and can cause the other end to generate a synchronization
5457 error if it is another Exim or anything else that does the synchronization
5458 check. A QUIT command is no longer sent after a timeout.
5460 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5461 during host lookups.
5463 74. The defer_ok option of callout verification was not working if it was used
5464 when verifying addresses in header lines, that is, for this case:
5466 verify = header_sender/callout=defer_ok
5468 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5469 those file descriptors could be used for SMTP connections. If anything
5470 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5471 could be sent to the SMTP client, causing chaos. The daemon now opens
5472 stdin, stdout, and stderr to /dev/null when it puts itself into the
5475 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5476 log by default. The user can override this with suitable Perl magic.
5478 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5479 the log message when discard triggers, was not working for the DATA ACL or
5480 for the non-SMTP ACL.
5482 78. Error message wording change in sieve.c.
5484 79. If smtp_accept_max_per_host was set, the number of connections could be
5485 restricted to fewer than expected, because the daemon was trying to set up
5486 a new connection before checking whether the processes handling previous
5487 connections had finished. The check for completed processes is now done
5488 earlier. On busy systems, this bug wouldn't be noticed because something
5489 else would have woken the daemon, and it would have reaped the completed
5492 80. If a message was submitted locally by a user whose login name contained one
5493 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5494 It caused a spool format error. I have fixed the spool reading code. A
5495 related problem was that the "from" clause in the Received: line became
5496 illegal because of the space(s). It is now covered by ${quote_local_part.
5498 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5501 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5503 83. Patch from Sieve maintainer to fix unterminated string problem in
5504 "vacation" handling.
5506 84. Some minor changes to the Linux configuration files to help with other
5507 OS variants using glibc.
5509 85. One more patch for Sieve to update vacation handling to latest spec.
5512 ----------------------------------------------------
5513 See the note above about the 4.44 and 4.50 releases.
5514 ----------------------------------------------------
5520 1. Change 4.43/35 introduced a bug that caused file counts to be
5521 incorrectly computed when quota_filecount was set in an appendfile
5524 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5525 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5528 3. Give more explanation in the error message when the command for a transport
5529 filter fails to execute.
5531 4. There are several places where Exim runs a non-Exim command in a
5532 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5533 was being done only for the command run by the queryprogram router. It is
5534 now done for all such subprocesses. The other cases are: ${run, transport
5535 filters, and the commands run by the lmtp and pipe transports.
5537 5. Some older OS have a limit of 256 on the maximum number of file
5538 descriptors. Exim was using setrlimit() to set 1000 as a large value
5539 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5540 systems. I've change it so that if it can't get 1000, it tries for 256.
5542 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5543 was an oversight, and furthermore, ever since the addition of extra
5544 controls (e.g. 4.43/32), the checks on when to allow different forms of
5545 "control" were broken. There should now be diagnostics for all cases when a
5546 control that does not make sense is encountered.
5548 7. $recipients is now available in the predata ACL (oversight).
5550 8. Tidy the search cache before the fork to do a delivery from a message
5551 received from the command line. Otherwise the child will trigger a lookup
5552 failure and thereby defer the delivery if it tries to use (for example) a
5553 cached ldap connection that the parent has called unbind on.
5555 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5556 of $address_data from the recipient verification was clobbered by the
5557 sender verification.
5559 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5560 was its contents. (It was OK if the option was not defined at all.)
5562 11. A "Completed" log line is now written for messages that are removed from
5563 the spool by the -Mrm option.
5565 12. $host_address is now set to the target address during the checking of
5566 ignore_target_hosts.
5568 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5569 being passed; this would have caused $sender_host_name to have been used if
5570 matching the list had actually called for a host name (not very likely,
5571 since this list is usually IP addresses). A host name is now passed as
5574 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5575 code that specifies a non-restarting handler (typically sigaction() in
5576 modern systems) in an attempt to fix a rare and obscure crash bug.
5578 15. Narrowed the window for a race in the daemon that could cause it to ignore
5579 SIGCHLD signals. This is not a major problem, because they are used only to
5580 wake it up if nothing else does.
5582 16. A malformed maildirsize file could cause Exim to calculate negative values
5583 for the mailbox size or file count. Odd effects could occur as a result.
5584 The maildirsize information is now recalculated if the size or filecount
5587 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5588 support for a long time. Removed HAVE_SYS_VFS_H.
5590 18. Updated exipick to current release from John Jetmore.
5592 19. Allow an empty sender to be matched against a lookup in an address list.
5593 Previously the only cases considered were a regular expression, or an
5596 20. Exim went into a mad DNS lookup loop when doing a callout where the
5597 host was specified on the transport, if the DNS lookup yielded more than
5600 21. The RFC2047 encoding function was originally intended for short strings
5601 such as real names; it was not keeping to the 75-character limit for
5602 encoded words that the RFC imposes. It now respects the limit, and
5603 generates multiple encoded words if necessary. To be on the safe side, I
5604 have increased the buffer size for the ${rfc2047: expansion operator from
5607 22. Failure to deliver a bounce message always caused it to be frozen, even if
5608 there was an errors_to setting on the router. The errors_to setting is now
5611 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5612 canonical form (fully expanded) before being placed in
5613 $sender_host_address.
5615 24. Updated eximstats to version 1.33
5617 25. Include certificate and key file names in error message when GnuTLS fails
5618 to set them up, because the GnuTLS error message doesn't include the name
5619 of the failing file when there is a problem reading it.
5621 26. Expand error message when OpenSSL has problems setting up cert/key files.
5624 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5625 (this can affect the format of dates).
5627 28. exim_tidydb, when checking for the continued existence of a message for
5628 which it has found a message-specific retry record, was not finding
5629 messages that were in split spool directories. Consequently, it was
5630 deleting retry records that should have stayed in existence.
5632 29. eximstats updated to version 1.35
5633 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5634 1.35 - bugfix such that pie charts by volume are generated correctly
5636 30. The SPA authentication driver was not abandoning authentication and moving
5637 on to the next authenticator when an expansion was forced to fail,
5638 contradicting the general specification for all authenticators. Instead it
5639 was generating a temporary error. It now behaves as specified.
5641 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5642 (the order specifies the preference for clients). The order is now AES256,
5643 AES128, 3DES, ARCFOUR128.
5645 31. Small patch to Sieve code - explicitly set From: when generating an
5648 32. Exim crashed if a remote delivery caused a very long error message to be
5649 recorded - for instance if somebody sent an entire SpamAssassin report back
5650 as a large number of 550 error lines. This bug was coincidentally fixed by
5651 increasing the size of one of Exim's internal buffers (big_buffer) that
5652 happened as part of the Exiscan merge. However, to be on the safe side, I
5653 have made the code more robust (and fixed the comments that describe what
5656 33. Some experimental protocols are using DNS PTR records for new purposes. The
5657 keys for these records are domain names, not reversed IP addresses. The
5658 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5659 leaves it alone. Component reversal etc. now happens only for IP addresses.
5662 34. The host_aton() function is supposed to be passed a string that is known
5663 to be a valid IP address. However, in the case of IPv6 addresses, it was
5664 not checking this. This is a hostage to fortune. Exim now panics and dies
5665 if the condition is not met. A case was found where this could be provoked
5666 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5667 components; fortuitously, this particular loophole had already been fixed
5668 by change 4.50/55 or 4.44/33 above.
5670 If there are any other similar loopholes, the new check in host_aton()
5671 itself should stop them being exploited. The report I received stated that
5672 data on the command line could provoke the exploit when Exim was running as
5673 exim, but did not say which command line option was involved. All I could
5674 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5675 running as the user.
5678 35. There was a buffer overflow vulnerability in the SPA authentication code
5679 (which came originally from the Samba project). I have added a test to the
5680 spa_base64_to_bits() function which I hope fixes it.
5683 36. The daemon start-up calls getloadavg() while still root for those OS that
5684 need the first call to be done as root, but it missed one case: when
5685 deliver_queue_load_max is set with deliver_drop_privilege. This is
5686 necessary for the benefit of the queue runner, because there is no re-exec
5687 when deliver_drop_privilege is set.
5689 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5690 was in use, was not putting the data itself into the right store pool;
5691 consequently, it could be overwritten for a subsequent message in the same
5692 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5695 38. Sometimes the final signoff response after QUIT could fail to get
5696 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5697 < 0 before doing a fflush(). This bug looks as though it goes back to the
5698 introduction of TLS in release 3.20, but "sometimes" must have been rare
5699 because the tests only now provoked it.
5705 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5706 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5707 Now it returns an integer. A number of calls were still expecting a T/F
5708 return. Fortuitously, in all cases, the tests worked in OK situations,
5709 which is the norm. However, things would have gone wrong on any write
5710 failures on the smtp file descriptor. This function is used when sending
5711 messages over SMTP and also when doing verify callouts.
5713 2. When Exim is called to do synchronous delivery of a locally submitted
5714 message (the -odf or -odi options), it no longer closes stderr before doing
5717 3. Implemented the mua_wrapper option.
5719 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5721 5. Implemented the functions header_remove(), header_testname(),
5722 header_add_at_position(), and receive_remove_recipient(), and exported them
5725 6. If an ACL "warn" statement specified the addition of headers, Exim already
5726 inserted X-ACL-Warn: at the start if there was no header name. However, it
5727 was not making this test for the second and subsequent header lines if
5728 there were newlines in the string. This meant that an invalid header could
5729 be inserted if Exim was badly configured.
5731 7. Allow an ACL "warn" statement to add header lines at the start or after all
5732 the Received: headers, as well as at the end.
5734 8. Added the rcpt_4xx retry error code.
5736 9. Added postmaster_mailfrom=xxx to callout verification option.
5738 10. Added mailfrom=xxxx to the callout verification option, for verify=
5741 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5742 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5744 12. Inserted some casts to stop certain compilers warning when using pointer
5745 differences as field lengths or precisions in printf-type calls (mostly
5746 affecting debugging statements).
5748 13. Added optional readline() support for -be (dynamically loaded).
5750 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5751 same clock tick as a message's arrival, so that its received time was the
5752 same as the "first fail" time on the retry record, and that message
5753 remained on the queue past the ultimate address timeout, every queue runner
5754 would try a delivery (because it was past the ultimate address timeout) but
5755 after another failure, the ultimate address timeout, which should have then
5756 bounced the address, did not kick in. This was a "< instead of <=" error;
5757 in most cases the first failure would have been in the next clock tick
5758 after the received time, and all would be well.
5760 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5761 being recognized when the domain list was tested by the match_domain
5762 condition in an expansion string.
5764 16. Added the ${str2b64: operator.
5766 17. Exim was always calling setrlimit() to set a large limit for the number of
5767 processes, without checking whether the existing limit was already
5768 adequate. (It did check for the limit on file descriptors.) Furthermore,
5769 errors from getrlimit() and setrlimit() were being ignored. Now they are
5770 logged to the main and panic logs, but Exim does carry on, to try to do its
5771 job under whatever limits there are.
5773 18. Imported PCRE 5.0.
5775 19. Trivial typo in log message " temporarily refused connection" (the leading
5778 20. If the log selector return_path_on_delivery was set and an address was
5779 redirected to /dev/null, the delivery process crashed because it assumed
5780 that a return path would always be set for a "successful" delivery. In this
5781 case, the whole delivery is bypassed as an optimization, and therefore no
5784 21. Internal re-arrangement: the function for sending a challenge and reading
5785 a response while authentication was assuming a zero-terminated challenge
5786 string. It's now changed to take a pointer and a length, to allow for
5787 binary data in such strings.
5789 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5791 23. Exim was not respecting finduser_retries when seeking the login of the
5792 uid under which it was called; it was always trying 10 times. (The default
5793 setting of finduser_retries is zero.) Also, it was sleeping after the final
5794 failure, which is pointless.
5796 24. Implemented tls_on_connect_ports.
5798 25. Implemented acl_smtp_predata.
5800 26. If the domain in control=submission is set empty, Exim assumes that the
5801 authenticated id is a complete email address when it generates From: or
5802 Sender: header lines.
5804 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5805 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5806 chown and chgrp in /bin and hostname in /usr/bin.
5808 28. Exim was keeping the "process log" file open after each use, just as it
5809 does for the main log. This opens the possibility of it remaining open for
5810 long periods when the USR1 signal hits a daemon. Occasional processlog
5811 errors were reported, that could have been caused by this. Anyway, it seems
5812 much more sensible not to leave this file open at all, so that is what now
5815 29. The long-running daemon process does not normally write to the log once it
5816 has entered its main loop, and it closes the log before doing so. This is
5817 so that log files can straightforwardly be renamed and moved. However,
5818 there are a couple of unusual error situations where the daemon does write
5819 log entries, and I had neglected to close the log afterwards.
5821 30. The text of an SMTP error response that was received during a remote
5822 delivery was being truncated at 512 bytes. This is too short for some of
5823 the long messages that one sometimes sees. I've increased the limit to
5826 31. It is now possible to make retry rules that apply only when a message has a
5827 specific sender, in particular, an empty sender.
5829 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5830 it possible to be selective about when SMTP synchronization is enforced.
5832 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5834 32. Implemented hosts_connection_nolog.
5836 33. Added an ACL for QUIT.
5838 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5841 35. Added mailbox_size and mailbox_filecount to appendfile.
5843 36. Added control = no_multiline_responses to ACLs.
5845 37. There was a bug in the logic of the code that waits for the clock to tick
5846 in the case where the clock went backwards by a substantial amount such
5847 that the microsecond fraction of "now" was more than the microsecond
5848 fraction of "then" (but the whole seconds number was less).
5850 38. Added support for the libradius Radius client library this is found on
5851 FreeBSD (previously only the radiusclient library was supported).
5857 1. When certain lookups returned multiple values in the form name=value, the
5858 quoting of the values was not always being done properly. Specifically:
5859 (a) If the value started with a double quote, but contained no whitespace,
5861 (b) If the value contained whitespace other than a space character (i.e.
5862 tabs or newlines or carriage returns) it was not quoted.
5863 This fix has been applied to the mysql and pgsql lookups by writing a
5864 separate quoting function and calling it from the lookup code. The fix
5865 should probably also be applied to nisplus, ibase and oracle lookups, but
5866 since I cannot test any of those, I have not disturbed their existing code.
5868 2. A hit in the callout cache for a specific address caused a log line with no
5869 reason for rejecting RCPT. Now it says "Previous (cached) callout
5870 verification failure".
5872 3. There was an off-by-one bug in the queryprogram router. An over-long
5873 return line was truncated at 256 instead of 255 characters, thereby
5874 overflowing its buffer with the terminating zero. As well as fixing this, I
5875 have increased the buffer size to 1024 (and made a note to document this).
5877 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5878 when Exim is waiting for an SMTP response from a remote server, Exim
5879 restarts its select() call on the socket, thereby resetting its timeout.
5880 This is not a problem when such interrupts are rare. Somebody set up a cron
5881 job to run exiwhat every 2 minutes, which is less than the normal select()
5882 timeout (5 or 10 minutes). This meant that the select() timeout never
5883 kicked in because it was always reset. I have fixed this by comparing the
5884 time when an interrupt arrives with the time at the start of the first call
5885 to select(). If more time than the timeout has elapsed, the interrupt is
5886 treated as a timeout.
5888 5. Some internal re-factoring in preparation for the addition of Sieve
5889 extensions (by MH). In particular, the "personal" test is moved to a
5890 separate function, and given an option for scanning Cc: and Bcc: (which is
5891 not set for Exim filters).
5893 6. When Exim created an email address using the login of the caller as the
5894 local part (e.g. when creating a From: or Sender: header line), it was not
5895 quoting the local part when it contained special characters such as @.
5897 7. Installed new OpenBSD configuration files.
5899 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5900 try to make them clearer.
5902 9. Callout options, other than the timeout value, were being ignored when
5903 verifying sender addresses in header lines. For example, when using
5905 verify = header_sender/callout=no_cache
5907 the cache was (incorrectly) being used.
5909 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5910 only the Cygwin environment.
5912 11. When return_path_on_delivery was set as a log selector, if different remote
5913 addresses in the same message used different return paths and parallel
5914 remote delivery occurred, the wrong values would sometimes be logged.
5915 (Whenever a remote delivery process finished, the return path value from
5916 the most recently started remote delivery process was logged.)
5918 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5919 lines when AUTH and/or TLS are in use. This is the "received protocol"
5920 field. Exim used to use "asmtp" for authenticated SMTP, without any
5921 indication (in the protocol name) for TLS use. Now it follows the RFC and
5922 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5923 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5924 names appear in log lines as well as in Received: header lines.
5926 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5927 extensions, and comparison tests, and to fix some bugs.
5929 14. Changes to the "personal" filter test:
5931 (1) The test was buggy in that it was just doing the equivalent of
5932 "contains" tests on header lines. For example, if a user's address was
5933 anne@some.where, the "personal" test would incorrectly be true for
5935 To: susanne@some.where
5937 This test is now done by extracting each address from the header in turn,
5938 and checking the entire address. Other tests that are part of "personal"
5939 are now done using regular expressions (for example, to check local parts
5940 of addresses in From: header lines).
5942 (2) The list of non-personal local parts in From: addresses has been
5943 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5944 taken from the Sieve specification recommendations.
5946 (3) If the message contains any header line starting with "List-" it is
5947 treated as non-personal.
5949 (4) The test for "circular" in the Subject: header line has been removed
5950 because it now seems ill-conceived.
5952 15. Minor typos in src/EDITME comments corrected.
5954 16. Installed latest exipick from John Jetmore.
5956 17. If headers_add on a router specified a text string that was too long for
5957 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5958 of string_sprintf() is now avoided.
5960 18. $message_body_size was not set (it was always zero) when running the DATA
5961 ACL and the local_scan() function.
5963 19. For the "mail" command in an Exim filter, no default was being set for
5964 the once_repeat time, causing a random time value to be used if "once" was
5965 specified. (If the value happened to be <= 0, no repeat happened.) The
5966 default is now 0s, meaning "never repeat". The "vacation" command was OK
5967 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5968 (I found it when inspecting the code).
5970 20. There is now an overall timeout for performing a callout verification. It
5971 defaults to 4 times the callout timeout, which applies to individual SMTP
5972 commands during the callout. The overall timeout applies when there is more
5973 than one host that can be tried. The timeout is checked before trying the
5974 next host. This prevents very long delays if there are a large number of
5975 hosts and all are timing out (e.g. when the network connections are timing
5976 out). The value of the overall timeout can be changed by specifying an
5977 additional sub-option for "callout", called "maxwait". For example:
5979 verify = sender/callout=5s,maxwait=20s
5981 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5982 to the end before writing, but this should make it even safer).
5984 22. Exim was forgetting that it had advertised PIPELINING for the second and
5985 subsequent messages on an SMTP connection. It was also not resetting its
5986 memory on STARTTLS and an internal HELO.
5988 23. When Exim logs an SMTP synchronization error within a session, it now
5989 records whether PIPELINING has been advertised or not.
5991 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5992 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5993 rather than long int.
5995 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5997 26. Added the never_mail option to autoreply.
6003 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6004 crash if the getsockname() call failed; this can happen if a connection is
6005 closed very soon after it is established. The problem was simply in the
6006 order in which certain operations were done, causing Exim to try to write
6007 to the SMTP stream before it had set up the file descriptor. The bug has
6008 been fixed by making things happen in the correct order.
6014 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6015 before the connection was closed, thus losing the rejection response.
6017 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6018 some early Solaris releases, but causes trouble in current releases where
6019 socklen_t is defined.
6021 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6024 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6027 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6028 that releasing the top of it at the end releases what was used for sub-
6029 expansions (unless the block got too big). However, discard this block if
6030 the first thing is a variable or header, so that we can use its block when
6031 it is dynamic (useful for very large $message_headers, for example).
6033 6. Lookups now cache *every* query, not just the most recent. A new, separate
6034 store pool is used for this. It can be recovered when all lookup caches are
6035 flushed. Lookups now release memory at the end of their result strings.
6036 This has involved some general refactoring of the lookup sources.
6038 7. Some code has been added to the store_xxx() functions to reduce the amount
6039 of flapping under certain conditions.
6041 8. log_incoming_interface used to affect only the <= reception log lines. Now
6042 it causes the local interface and port to be added to several more SMTP log
6043 lines, for example "SMTP connection from", and rejection lines.
6045 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6047 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6049 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6050 could be overwritten at the end of the current message (or the start of a
6051 new message if it was set in a HELO ACL). The value is now preserved for
6052 the duration of the SMTP connection.
6054 12. If a transport had a headers_rewrite setting, and a matching header line
6055 contained an unqualified address, that address was qualified, even if it
6056 did not match any rewriting rules. The underlying bug was that the values
6057 of the flags that permit the existence of unqualified sender and recipient
6058 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6059 non-local messages, and by -bnq for local messages) were not being
6060 preserved with the message after it was received.
6062 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6063 "next input=" as part of the text comprising the host identity instead of
6064 the correct text. The code was using the same buffer for two different
6065 strings. However, depending on which order the printing function evaluated
6066 its arguments, the bug did not always show up. Under Linux, for example, my
6067 test suite worked just fine.
6069 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6070 doesn't work with some older versions of Perl. It has been changed to "my",
6071 which in any case is probably the better facility to use.
6073 15. A really picky compiler found some instances of statements for creating
6074 error messages that either had too many or two few arguments for the format
6077 16. The size of the buffer for calls to the DNS resolver has been increased
6078 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6079 for addresses that have a lot of PTR records. This alleviates a problem; it
6080 does not fully solve it.
6082 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6083 buffer now truncates the list and logs the incident, which is the same
6084 action as happens when Exim is looking up a host name and its aliases.
6085 Previously in this situation something unpredictable would happen;
6086 sometimes it was "internal error: store_reset failed".
6088 18. If a server dropped the connection unexpectedly when an Exim client was
6089 using GnuTLS and trying to read a response, the client delivery process
6090 crashed while trying to generate an error log message.
6092 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6093 string, for example:
6095 warn message = H1: something\nH2: something
6097 the text was added as a single header line from Exim's point of view
6098 though it ended up OK in the delivered message. However, searching for the
6099 second and subsequent header lines using $h_h2: did not work. This has been
6100 fixed. Similarly, if a system filter added multiple headers in this way,
6101 the routers could not see them.
6103 20. Expanded the error message when iplsearch is called with an invalid key to
6104 suggest using net-iplsearch in a host list.
6106 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6107 ACLs are no longer actually imposed (and a message to that effect is
6110 22. If a "gecos" field in a passwd entry contained escaped characters, in
6111 particular, if it contained a \" sequence, Exim got it wrong when building
6112 a From: or a Sender: header from that name. A second bug also caused
6113 incorrect handling when an unquoted " was present following a character
6114 that needed quoting.
6116 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6117 was not being matched caselessly.
6119 24. Arranged for all hyphens in the exim.8 source to be escaped with
6122 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6123 properly. Recipient callout cache records were still being keyed to include
6124 the sender, even when use_sender was set false. This led to far more
6125 callouts that were necessary. The sender is no longer included in the key
6126 when use_sender is false.
6128 26. Added "control = submission" modifier to ACLs.
6130 27. Added the ${base62d: operator to decode base 62 numbers.
6132 28. dnsdb lookups can now access SRV records.
6134 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6135 the configuration file.
6137 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6138 (-v) mode. This makes the output for a verbose queue run more intelligible.
6140 31. Added a use_postmaster feature to recipient callouts.
6142 32. Added the $body_zerocount variable, containing the number of binary zero
6143 bytes in the message body.
6145 33. The time of last modification of the "new" subdirectory is now used as the
6146 "mailbox time last read" when there is a quota error for a maildir
6149 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6151 35. Added +ignore_unknown as a special item in host lists.
6153 36. Code for decoding IPv6 addresses in host lists is now included, even if
6154 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6155 address was recognized as an IP address, but was then not correctly decoded
6156 into binary, causing unexpected and incorrect effects when compared with
6163 1. Very minor rewording of debugging text in manualroute to say "list of
6164 hosts" instead of "hostlist".
6166 2. If verify=header_syntax was set, and a header line with an unqualified
6167 address (no domain) and a large number of spaces between the end of the
6168 name and the colon was received, the reception process suffered a buffer
6169 overflow, and (when I tested it) crashed. This was caused by some obsolete
6170 code that should have been removed. The fix is to remove it!
6172 3. When running in the test harness, delay a bit after writing a bounce
6173 message to get a bit more predictability in the log output.
6175 4. Added a call to search_tidyup() just before forking a reception process. In
6176 theory, someone could use a lookup in the expansion of smtp_accept_max_
6177 per_host which, without the tidyup, could leave open a database connection.
6179 5. Added the variables $recipient_data and $sender_data which get set from a
6180 lookup success in an ACL "recipients" or "senders" condition, or a router
6181 "senders" option, similar to $domain_data and $local_part_data.
6183 6. Moved the writing of debug_print from before to after the "senders" test
6186 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6187 problems for message scanning, either using a data ACL, or using
6188 local_scan() because the Received: header was not generated till after they
6189 were called (in order to set the time as the time of reception completion).
6190 I have revised the way this works. The header is now generated after the
6191 body is received, but before the ACL or local_scan() are called. After they
6192 are run, the timestamp in the header is updated.
6198 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6199 before starting a queue runner without re-exec. This happened only when
6200 deliver_drop_privilege was set or when the Exim user was set to root. The
6201 effect of the bug was that timeouts during subsequent deliveries caused
6202 crashes instead of being properly handled. The handler is now left at its
6203 default (and expected) setting.
6205 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6206 message, again when deliver_drop_privilege is set or Exim is run as root.
6207 The bug described in (1) was not present in this case, but the tidying up
6208 of the other signals was missing. I have made the two cases consistent.
6210 3. The ignore_target_hosts setting on a manualroute router was being ignored
6211 for hosts that were looked up using the /MX notation.
6213 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6216 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6217 operated on the sender address. After changing the $sender_address to <>
6218 for the sender address verify, Exim was re-instated it as the original
6219 (before rewriting) address, but remembering that it had rewritten it, so it
6220 wasn't rewriting it again. This bug also had the effect of breaking the
6221 sender address verification caching when the sender address was rewritten.
6223 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6224 This has been changed so that if the ip literal address matches
6225 ignore_target_hosts, the router declines.
6227 7. Added expansion conditions match_domain, match_address, and match_local_
6228 part (NOT match_host).
6230 8. The placeholder for the Received: header didn't have a length field set.
6232 9. Added code to Exim itself and to exim_lock to test for a specific race
6233 condition that could lead to file corruption when using MBX delivery. The
6234 issue is with the lockfile that is created in /tmp. If this file is removed
6235 after a process has opened it but before that process has acquired a lock,
6236 there is the potential for a second process to recreate the file and also
6237 acquire a lock. This could lead to two Exim processes writing to the file
6238 at the same time. The added code performs the same test as UW imapd; it
6239 checks after acquiring the lock that its file descriptor still refers to
6240 the same named file.
6242 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6243 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6246 11. Added the smtp_active_hostname option. If used, this will typically be made
6247 to depend on the incoming interface address. Because $interface_address is
6248 not set up until the daemon has forked a reception process, error responses
6249 that can happen earlier (such as "too many connections") no longer contain
6252 12. If an expansion in a condition on a "warn" statement fails because a lookup
6253 defers, the "warn" statement is abandoned, and the next ACL statement is
6254 processed. Previously this caused the whole ACL to be aborted.
6256 13. Added the iplsearch lookup type.
6258 14. Added ident_timeout as a log selector.
6260 15. Added tls_certificate_verified as a log selector.
6262 16. Added a global option tls_require_ciphers (compare the smtp transport
6263 option of the same name). This controls incoming TLS connections.
6265 17. I finally figured out how to make tls_require_ciphers do a similar thing
6266 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6267 before starting the TLS session.
6269 18. Tabs are now shown as \t in -bP output.
6271 19. If the log selector return_path_on_delivery was set, Exim crashed when
6272 bouncing a message because it had too many Received: header lines.
6274 20. If two routers both had headers_remove settings, and the first one included
6275 a superfluous trailing colon, the final name in the first list and the
6276 first name in the second list were incorrectly joined into one item (with a
6277 colon in the middle).
6283 1. Added -C and -D options to the exinext utility, mainly to make it easier
6284 to include in the automated testing, but these could be helpful when
6285 multiple configurations are in use.
6287 2. The exinext utility was not formatting the output nicely when there was
6288 an alternate port involved in the retry record key, nor when there was a
6289 message id as well (for retries that were specific to a specific message
6290 and a specific host). It was also confused by IPv6 addresses, because of
6291 the additional colons they contain. I have fixed the IPv4 problem, and
6292 patched it up to do a reasonable job for IPv6.
6294 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6295 delivery, the log line now contains "pipelined" if PIPELINING was used.
6297 4. An SMTP transport process used to panic and die if the bind() call to set
6298 an explicit outgoing interface failed. This has been changed; it is now
6299 treated in the same way as a connect() failure.
6301 5. A reference to $sender_host_name in the part of a conditional expansion
6302 that was being skipped was still causing a DNS lookup. This no longer
6305 6. The def: expansion condition was not recognizing references to header lines
6306 that used bh_ and bheader_.
6308 7. Added the _cache feature to named lists.
6310 8. The code for checking quota_filecount in the appendfile transport was
6311 allowing one more file than it should have been.
6313 9. For compatibility with Sendmail, the command line option
6321 and sets the incoming protocol and host name (for trusted callers). The
6322 host name and its colon can be omitted when only the protocol is to be set.
6323 Note the Exim already has two private options, -pd and -ps, that refer to
6324 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6325 "s", but I don't think that's a major issue.
6327 10. A number of refactoring changes to the code, none of which should affect
6330 (a) The number of logging options was getting close to filling up the
6331 32-bit word that was used as a bit map. I have split them into two classes:
6332 those that are passed in the argument to log_write(), and those that are
6333 only ever tested independently outside of that function. These are now in
6334 separate 32-bit words, so there is plenty of room for expansion again.
6335 There is no change in the user interface or the logging behaviour.
6337 (b) When building, for example, log lines, the code previously used a
6338 macro that called string_cat() twice, in order to add two strings. This is
6339 not really sufficiently general. Furthermore, there was one instance where
6340 it was actually wrong because one of the argument was used twice, and in
6341 one call a function was used. (As it happened, calling the function twice
6342 did not affect the overall behaviour.) The macro has been replaced by a
6343 function that can join an arbitrary number of extra strings onto a growing
6346 (c) The code for expansion conditions now uses a table and a binary chop
6347 instead of a serial search (which was left over from when there were very
6348 few conditions). Also, it now recognizes conditions like "pam" even when
6349 the relevant support is not compiled in: a suitably worded error message is
6350 given if an attempt is made to use such a condition.
6352 11. Added ${time_interval:xxxxx}.
6354 12. A bug was causing one of the ddress fields not to be passed back correctly
6355 from remote delivery subprocesses. The field in question was not being
6356 subsequently used, so this caused to problems in practice.
6358 13. Added new log selectors queue_time and deliver_time.
6360 14. Might have fixed a bug in maildirsizefile handling that threw up
6361 "unexpected character" debug warnings, and recalculated the data
6362 unnecessarily. In any case, I expanded the warning message to give more
6365 15. Added the message "Restricted characters in address" to the statements in
6366 the default ACL that block characters like @ and % in local parts.
6368 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6369 Three changes have been made:
6371 (a) There was a serious bug; a negative response to MAIL caused the whole
6372 recipient domain to be cached as invalid, thereby blocking all messages
6373 to all local parts at the same domain, from all senders. This bug has
6374 been fixed. The domain is no longer cached after a negative response to
6375 MAIL if the sender used is not empty.
6377 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6380 (c) A new callout option, "use_sender" has been added for people who want
6381 the modified behaviour.
6387 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6390 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6391 indeed breaks things for older releases.
6393 3. Added additional logging to the case where there is a problem reading data
6394 from a filter that is running in a subprocess using a pipe, in order to
6395 try to track down a specific problem.
6397 4. Testing facility fudge: when running in the test harness and attempting
6398 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6399 getting "No route to host". Convert this to a timeout.
6401 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6404 6. Some OS don't have socklen_t but use size_t instead. This affects the
6405 fifth argument of getsockopt() amongst other things. This is now
6406 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6407 can be set for individual OS. I have set it for SunOS5, OSF1, and
6408 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6409 some earlier ones do not.
6411 7. Change 4.30/15 was not doing the test caselessly.
6413 8. The standard form for an IPv6 address literal was being rejected by address
6414 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6415 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6416 this, as well as the form without the "IPv6" on the front (but only when
6417 address literals are enabled, of course).
6419 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6421 10. Exim crashed if a message with an empty sender address specified by -f
6422 encountered a router with an errors_to setting. This could be provoked only
6423 by a command such as
6427 where an empty string was supplied; "<>" did not hit this bug.
6429 11. Installed PCRE release 4.5.
6431 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6432 remained set. It is now erased.
6434 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6435 times from message ids (which are base 36 rather than the normal 62).
6437 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6438 were being counted as actual protocol errors, and logged if the log
6439 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6440 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6441 and DATA following a set of rejected RCPTs do not count as protocol errors.
6442 In other words, Exim assumes they were pipelined, though this may not
6443 actually be the case. Of course, in all cases the client gets an
6444 appropriate error code.
6446 15. If a lookup fails in an ACL condition, a message about the failure may
6447 be available; it is used if testing the ACL cannot continue, because most
6448 such messages specify what the cause of the deferral is. However, some
6449 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6450 that caused an old message to be retained and used if a later statement
6451 caused a defer, replacing the real cause of the deferral.
6453 16. If an IP address had so many PTR records that the DNS lookup buffer
6454 was not large enough to hold them, Exim could crash while trying to process
6455 the truncated data. It now detects and logs this case.
6457 17. Further to 4.21/58, another change has been made: if (and only if) the
6458 first line of a message (the first header line) ends with CRLF, a bare LF
6459 in a subsequent header line has a space inserted after it, so as not to
6460 terminate the header.
6462 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6463 unnecessarily, used atoi() instead of strtol(), and didn't check the
6464 termination when getting file sizes from file names by regex.
6466 19. Completely re-implemented the support for maildirsize files, in the light
6467 of a number of problems with the previous contributed implementation
6468 (4.30/29). In particular:
6470 . If the quota is zero, the maildirsize file is maintained, but no quota is
6473 . If the maildir directory does not exist, it is created before any attempt
6474 to write a maildirsize file.
6476 . The quota value in the file is just a cache; if the quota is changed in
6477 the transport, the new value overrides.
6479 . A regular expression is available for excluding directories from the
6482 20. The autoreply transport checks the characters in options that define the
6483 message's headers; it allows continued headers, but it was checking with
6484 isspace() after an embedded newline instead of explicitly looking for a
6487 21. If all the "regular" hosts to which an address was routed had passed their
6488 expiry times, and had not reached their retry times, the address was
6489 bounced, even if fallback hosts were defined. Now Exim should go on to try
6492 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6493 equivalent code in the SMTP transport. Some hosts send humungous responses
6494 to HELO/EHLO, more than 1024 it seems.
6496 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6497 gives compiler warnings in some environments. I've now done it "properly",
6500 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6501 (because of problems with the built-in one) was declared to return uschar *
6502 instead of char *, causing compiler failure.
6504 25. Fixed a file descriptor leak when processing alias/forward files.
6506 26. Fixed a minor format string issue in dbfn.c.
6508 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6510 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6511 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6514 29. When a sender address is verified, it is cached, to save repeating the test
6515 when there is more than one recipient in a message. However, when the
6516 verification involves a callout, it is possible for different callout
6517 options to be set for different recipients. It is too complicated to keep
6518 track of this in the cache, so now Exim always runs a verification when a
6519 callout is required, relying on the callout cache for the optimization.
6520 The overhead is duplication of the address routing, but this should not be
6523 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6524 to be verified with callout=postmaster, and the main callout worked but the
6525 postmaster check failed, the verification correctly failed. However, if a
6526 subsequent RCPT command asked for sender verification *without* the
6527 postmaster check, incorrect caching caused this verification also to fail,
6530 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6531 it was not caching the DNS options (qualify_single, search_parents) that
6532 were used when the lookup failed. A subsequent lookup with different
6533 options therefore always gave the same answer, though there were cases
6534 where it should not have. (Example: a "domains = !$mx_any" option on a
6535 dnslookup router: the "domains" option is always processed without any
6536 widening, but the router might have qualify_single set.) Now Exim uses the
6537 cached value only when the same options are set.
6539 32. Added John Jetmore's "exipick" utility to the distribution.
6541 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6542 than a timeout (e.g. a certificate is required, and is not provided), an
6543 Exim server now closes the connection immediately. Previously it waited for
6544 the client to close - but if the client is SSL, it seems that they each
6545 wait for each other, leading to a delay before one of them times out.
6547 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6548 maintained 0.8.x compatibility because I don't think many are using it, and
6549 it is clearly obsolete.
6551 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6554 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6555 client certificate was expired. A simple patch fixes this, though I don't
6556 understand the full logic of why the verify callback is called multiple
6559 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6560 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6561 which causes problems with some clients (such as the Certicom SSL Plus
6562 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6563 disables the coutermeasure allowing Eudora to connect."
6565 38. Exim was not checking that a write() to a log file succeeded. This could
6566 lead to Bad Things if a log got too big, in particular if it hit a file
6567 size limit. Exim now panics and dies if it cannot write to a log file, just
6568 as it does if it cannot open a log file.
6570 39. Modified OS/Makefile-Linux so that it now contains
6572 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6574 The two -D definitions ensure that Exim is compiled with large file
6575 support, which makes it possible to handle log files that are bigger than
6578 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6579 instance) a domain was checked against a named list that involved a lookup,
6580 causing $domain_data to be set, then another domain was checked against the
6581 same list, then the first domain was re-checked, the value of $domain_data
6582 after the final check could be wrong. In particular, if the second check
6583 failed, it could be set empty. This bug probably also applied to
6586 41. The strip_trailing_dot option was not being applied to the address given
6587 with the -f command-line option.
6589 42. The code for reading a message's header from the spool was incrementing
6590 $received_count, but never initializing it. This meant that the value was
6591 incorrect (doubled) while delivering a message in the same process in which
6592 it was received. In the most common configuration of Exim, this never
6593 happens - a fresh exec is done - but it can happen when
6594 deliver_drop_privilege is set.
6596 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6597 it now includes up to 150 characters of the unexpected data in the log
6600 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6601 and building data strings. The size of both of these buffers was 10 000
6602 bytes - far larger than anybody would *ever* want, thought I. Needless to
6603 say, somebody hit the limit. I have increased the maximum line length to
6604 20 000 and the maximum data length of concatenated lines to 100 000. I have
6605 also fixed two bugs, because there was no checking on these buffers. Tsk,
6606 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6607 buffer is too small.
6609 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6610 lsearch lookups. Now it does.
6612 46. When parsing a route_list item in a manualroute router, a fixed-length
6613 buffer was used for the list of hosts. I made this 1024 bytes long,
6614 thinking that nobody would ever have a list of hosts that long. Wrong.
6615 Somebody had a whole pile of complicated expansion conditions, and the
6616 string was silently truncated, leading to an expansion error. It turns out
6617 that it is easier to change to an unlimited length (owing to other changes
6618 that have happened since this code was originally written) than to build
6619 structure for giving a limitation error. The length of the item that
6620 expands into the list of hosts is now unlimited.
6622 47. The lsearch lookup could not handle data where the length of text line was
6623 more than 4095 characters. Such lines were truncated, leading to shortened
6624 data being returned. It should now handle lines of any length.
6626 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6627 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6630 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6631 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6632 no longer happens in comment lines. A list of replacements is now placed
6633 at the head of all of the source files, except those whose only change is
6634 to replace PERL_COMMAND in the very first #! line.
6636 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6637 messages on the queue, with a bottom-up merge sort, using code contributed
6638 by Michael Haardt. This should make operations like -bp somewhat faster on
6639 large queues. It won't affect queue runners, except when queue_run_in_order
6642 51. Installed eximstats 1.31 in the distribution.
6644 52. Added support for SRV lookups to the dnslookup router.
6646 53. If an ACL referred to $message_body or $message_body_end, the value was not
6647 reset for any messages that followed in the same SMTP session.
6649 54. The store-handling optimization for building very long strings was not
6650 differentiating between the different store pools. I don't think this
6651 actually made any difference in practice, but I've tidied it.
6653 55. While running the routers to verify a sender address, $sender_address
6654 was still set to the sender address. This is wrong, because when routing to
6655 send a bounce to the sender, it would be empty. Therefore, I have changed
6656 it so that, while verifying a sender address, $sender_address is set to <>.
6657 (There is no change to what happens when verifying a recipient address.)
6659 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6660 target A or AAAA records (if not already returned) without resetting the
6661 qualify_single or search_parents options of the DNS resolver. These are
6662 inappropriate in this case because the targets of MX and SRV records must
6663 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6664 target that, when qualified, matched something in the local domain. These
6665 two options are now turned off when doing these lookups.
6667 57. It seems that at least some releases of Reiserfs (which does not have the
6668 concept of a fixed number of inodes) returns zero and not -1 for the
6669 number of available inodes. This interacted badly with check_spool_inodes,
6670 which assumed that -1 was the "no such thing" setting. What I have done is
6671 to check that the total number of inodes is greater than zero before doing
6672 the test of how many are available.
6674 58. When a "warn" ACL statement has a log_message modifier, the message is
6675 remembered, and not repeated. This is to avoid a lot of repetition when a
6676 message has many recipients that cause the same warning to be written.
6677 However, Exim was preserving the list of already written lines for an
6678 entire SMTP session, which doesn't seem right. The memory is now reset if a
6679 new message is started.
6681 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6682 debugging output unless log_rewrite was also set.
6684 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6685 of (int)(handle) when we know that handle contains (void *)(-1).
6687 61. The Exim daemon panic-logs an error return when it closes the incoming
6688 connection. However "connection reset by peer" seems to be common, and
6689 isn't really an error worthy of noting specially, so that particular error
6692 62. When Exim is trying to find all the local interfaces, it used to panic and
6693 die if the ioctl to get the interface flags failed. However, it seems that
6694 on at least one OS (Solaris 9) it is possible to have an interface that is
6695 included in the list of interfaces, but for which you get a failure error
6696 for this call. This happens when the interface is not "plumbed" into a
6697 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6698 failure of the "get flags" call assumes that the interface is down.
6700 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6701 makes life easier for people who are doing arithmetic on fields extracted
6702 from dates, where you often get leading zeros that should not be
6703 interpreted as octal.
6705 64. Added qualify_domain to the redirect router, to override the global
6708 65. If a pathologically long header line contained very many addresses (the
6709 report of this problem mentioned 10 000) and each of them was rewritten,
6710 Exim could use up a very large amount of memory. (It kept on making new
6711 copies of the header line as it rewrote, and never released the old ones.)
6712 At the expense of a bit more processing, the header rewriting function has
6713 been changed so that it no longer eats memory in this way.
6715 66. The generation of the Received: header has been moved from the time that a
6716 message starts to be received, to the time that it finishes. The timestamp
6717 in the Received: header should now be very close to that of the <= log
6718 line. There are two side-effects of this change:
6720 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6721 logged header lines no longer include the local Received: line, because
6722 it has not yet been created. The same applies to a copy of the message
6723 that is returned to a non-SMTP sender when a message is rejected.
6725 (b) When a filter file is tested using -bf, no additional Received: header
6726 is added to the test message. After some thought, I decided that this
6729 This change does not affect the value of $received_for. It is still set
6730 after address rewriting, but before local_scan() is called.
6732 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6734 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6735 gave an unhelpful panic error message, and a defer error. I have managed to
6736 change this behaviour so that it now rejects any supplied certificate,
6737 which seems right, as the list of acceptable certificates is empty.
6739 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6740 gave an unhelpful defer error. I have not managed to make this reject any
6741 supplied certificates, but the error message it gives is "no certificate
6742 supplied", which is not helpful.
6744 70. exigrep's output now also includes lines that are not associated with any
6745 message, but which match the given pattern. Implemented by a patch from
6746 Martin Sluka, which also tidied up the Perl a bit.
6748 71. Recipient callout verification, like sender verification, was using <> in
6749 the MAIL FROM command. This isn't really the right thing, since the actual
6750 sender may affect whether the remote host accepts the recipient or not. I
6751 have changed it to use the actual sender in the callout; this means that
6752 the cache record is now keyed on a recipient/sender pair, not just the
6753 recipient address. There doesn't seem to be a real danger of callout loops,
6754 since a callout by the remote host to check the sender would use <>.
6755 [SEE ABOVE: changed after hitting problems.]
6757 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6758 temporary errors. However, in the case of such a code being given after
6759 the end of a data transmission (i.e. after ".") Exim was failing to write
6760 a retry record for the message. (Yes, there was some broken host that was
6761 actually sending 8xx at this point.)
6763 73. An unknown lookup type in a host list could cause Exim to panic-die when
6764 the list was checked. (An example that provoked this was putting <; in the
6765 middle of a list instead of at the start.) If this happened during a DATA
6766 ACL check, a -D file could be left lying around. This kind of configuration
6767 error no longer causes Exim to die; instead it causes a defer error. The
6768 incident is still logged to the main and panic logs.
6770 74. Buglet left over from Exim 3 conversion. The message "too many messages
6771 in one connection" was written to the rejectlog but not the mainlog, except
6772 when address rewriting (yes!) was being logged.
6774 75. Added write_rejectlog option.
6776 76. When a system filter was run not as root (that is, when system_filter_user
6777 was set), the values of the $n variables were not being returned to the
6778 main process; thus, they were not subsequently available in the $sn
6781 77. Added +return_path_on_delivery log selector.
6783 78. A connection timeout was being treated differently from recipients deferred
6784 when testing hosts_max_try with a message that was older than the host's
6785 retry timeout. (The host should not be counted, thus allowing all hosts to
6786 be tried at least once before bouncing.) This may have been the cause of an
6787 occasionally reported bug whereby a message would remain on the queue
6788 longer than the retry timeout, but would be bounced if a delivery was
6789 forced. I say "may" because I never totally pinned down the problem;
6790 setting up timeout/retry tests is difficult. See also the next item.
6792 79. The ultimate address timeout was not being applied to errors that involved
6793 a combination of host plus message (for example, a timeout on a MAIL
6794 command). When an address resolved to a number of possible hosts, and they
6795 were not all tried for each delivery (e.g. because of hosts_max_try), a
6796 message could remain on the queue longer than the retry timeout.
6798 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6801 81. Fixed an obscure SMTP outgoing bug which required at least the following
6802 conditions: (a) there was another message waiting for the same server;
6803 (b) the server returned 5xx to all RCPT commands in the first message so
6804 that the message was not completed; (c) the server dropped the connection
6805 or gave a negative response to the RSET that Exim sends to abort the
6806 transaction. The observed case was a dropped connection after DATA that had
6807 been sent in pipelining mode. That is, the server had advertised PIPELINING
6808 but was not implementing it correctly. The effect of the bug was incorrect
6809 behaviour, such as trying another host, and this could lead to a crash.
6815 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6816 and daemon.c were passed as pointers to ints; they should have been
6817 pointers to socklen_t variables (which are typically unsigned ints).
6819 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6822 3. Fixed a really odd bug that affected only the testing scheme; patching a
6823 certain fixed string in the binary changed the value of another string that
6824 happened to be identical to the end of the original first string.
6826 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6827 name", it returns that address as the IP address. On some operating
6828 systems (e.g. Solaris), it also passes back the IP address string as the
6829 "host name". However, on others (e.g. Linux), it passes back an empty
6830 string. Exim wasn't checking for this, and was changing the host name to an
6831 empty string, assuming it had been canonicalized.
6833 5. Although rare, it is permitted to have more than one PTR record for a given
6834 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6835 all the names associated with an address, because they do in Solaris.
6836 However, it seems that they do not in Linux for data that comes from the
6837 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6838 I found this out when I moved to a new Linux workstation and tried to run
6839 the Exim test suite.
6841 To get round this problem I have changed the code so that it now does its
6842 own call to the DNS to look up PTR records when searching for a host name.
6843 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6844 addresses that are only in /etc/hosts are still found.
6846 This behaviour is, however, controlled by an option called host_lookup_
6847 order, which defaults to "bydns:byaddr". If people want to use the other
6848 order, or indeed, just use one or the other means of lookup, they can
6849 specify it in this variable.
6851 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6852 some operating systems, this comes back from gethostbyaddr() as an empty
6853 string, and this is what Exim used to test for. However, it seems that in
6854 other systems, "." is yielded. Exim now tests for this case too.
6856 7. The values of check_spool_space and check_log_space are now held internally
6857 as a number of kilobytes instead of an absolute number of bytes. If a
6858 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6859 kilobyte. This means that much larger values can be stored.
6861 8. Exim monitor: an attempt to get the action menu when not actually pointing
6862 at a message produces an empty menu entitled "No message selected". This
6863 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6864 no entries in it ("Shell widget menu has zero width and/or height"). So I
6865 have added a single, blank menu entry in this case.
6867 9. Added ${quote_local_part.
6869 10. MIME decoding is now applied to the contents of Subject: header lines when
6872 11. Now that a reference to $sender_host_address automatically causes a reverse
6873 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6874 host lookup before query-style lookups in lists that might use this
6875 variable. This has therefore been abolished, and the "net-" prefix is no
6876 longer necessary for query-style lookups.
6878 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6879 have been a typo for LFLAGS, so it has been changed.
6881 13. The install script calls Exim with "-C /dev/null" in order to find the
6882 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6883 to be output. However, since Exim outputs its version number before the
6884 error, it didn't break the script. It just looked ugly. I fixed this by
6885 always allowing "-C /dev/null" if the caller is root.
6887 14. Ignore overlarge ACL variable number when reading spool file - insurance
6888 against a later release with more variables having written the file.
6890 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6891 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6892 this, as well as the form without the "IPv6" on the front.
6894 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6895 OS/Makefile-Darwin file.
6897 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6898 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6900 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6901 variable. However, some OS use ints for inodes. I've added cast to long int
6902 to get rid of the compiler warning.
6904 19. I had forgotten to lock out "/../" in configuration file names when
6905 ALT_CONFIG_PREFIX was set.
6907 20. Routers used for verification do not need to specify transports. However,
6908 if such a router generated a host list, and callout was configured, Exim
6909 crashed, because it could not find a port number from the (non-existent)
6910 transport. It now assumes port 25 in this circumstance.
6912 21. Added the -t option to exigrep.
6914 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6915 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6916 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6917 them is set, LOOKUP_LSEARCH is forced.
6919 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6921 24. Added sender and host information to the "rejected by local_scan()" log
6922 line; previously there was no indication of these.
6924 25. Added .include_if_exists.
6926 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6927 Linux. It turns out that not all file systems support this. Apparently some
6928 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6929 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6930 not supported on the file descriptor, is now ignored when Exim is trying to
6931 sync a directory. This applies only to Linux.
6933 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6935 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6936 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6939 29. Added support for maildirsize files from supplied patch (modified a bit).
6941 30. The use of :fail: followed by an empty string could lead Exim to respond to
6942 sender verification failures with (e.g.):
6944 550 Verification failed for <xxx>
6945 550 Sender verify failed
6947 where the first response line was missing the '-' that indicates it is not
6948 the final line of the response.
6950 31. The loop for finding the name of the user that called Exim had a hardwired
6951 limit of 10; it now uses the value of finduser_retries, which is used for
6952 all other user lookups.
6954 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6957 33. Exim was neglecting to zero errno before one call of strtol() when
6958 expanding a string and expecting an integer value. On some systems this
6959 resulted in spurious "integer overflow" errors. Also, it was casting the
6960 result into an int without checking.
6962 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6963 did not work. The code looks as if it has *never* worked, though it appears
6964 to have been documented since at least release 1.62. I have made it work.
6966 35. The "timeout_DNS" error in retry rules, also documented since at least
6967 1.62, also never worked. As it isn't clear exactly what this means, and
6968 clearly it isn't a major issue, I have abolished the feature by treating it
6969 as "timeout", and writing a warning to the main and panic logs.
6971 36. The display of retry rules for -brt wasn't always showing the error code
6974 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6975 timeout_connect_A, timeout_connect_MX.
6977 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6978 to the empty sender.
6980 39. The daemon was not analysing the content of -oX till after it had closed
6981 stderr and disconnected from the controlling terminal. This meant that any
6982 syntax errors were only noted on the panic log, and the return code from
6983 the command was 0. By re-arranging the code a little, I've made the
6984 decoding happen first, so such errors now appear on stderr, and the return
6985 code is 1. However, the actual setting up of the sockets still happens in
6986 the disconnected process, so errors there are still only recorded on the
6989 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6990 connections (as happens on some IP stacks) was logged at start up time as
6991 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6992 it from "IPv6 and IPv4", which means that two separate sockets are being
6995 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6996 says whether AF_INET or AF_INET6 was passed as an argument.
6998 42. Exiwhat output was messed up when time zones were included in log
7001 43. Exiwhat now gives more information about the daemon's listening ports,
7002 and whether -tls-on-connect was used.
7004 44. The "port" option of the smtp transport is now expanded.
7006 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7007 silently ignored. Now an error message is written to the main and panic
7010 46. There's a new ACL modifier called "logwrite" which writes to a log file
7011 as soon as it is encountered.
7013 47. Added $local_user_uid and $local_user_gid at routing time.
7015 48. Exim crashed when trying to verify a sender address that was being
7018 49. Exim was recognizing only a space character after ".include". It now also
7019 recognizes a tab character.
7021 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7022 extracting the relevant information from the specification. The man page no
7023 longer contains scrambled data for the -d option, and I've added a section
7024 at the front about calling Exim under different names.
7026 51. Added "extra_headers" argument to the "mail" command in filter files.
7028 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7031 53. Installed eximstats 1.29.
7033 54. Added transport_filter_timeout as a generic transport option.
7035 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7036 Cc: header lines. This was required by RFC 822, but it not required by RFC
7039 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7040 incoming messages that did not have them. Now it does so only if the
7041 message originates locally, that is, if there is no associated remote host
7042 address. When Resent- header lines are present, this applies to the Resent-
7043 lines rather than the non-Resent- lines.
7045 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7046 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7048 58. Messages for configuration errors now include the name of the main
7049 configuration file - useful now that there may be more than one file in a
7050 list (.included file names were always shown).
7052 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7053 for those rare installations that do not start the daemon as root or run it
7054 setuid root. I've cut out the call to initgroups() if the daemon is not
7057 60. The Exim user and group can now be bound into the binary as text strings
7058 that are looked up at the start of Exim's processing.
7060 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7062 62. Added $mailstore_basename variable.
7064 63. Installed patch to sieve.c from Michael Haardt.
7066 64. When Exim failed to open the panic log after failing to open the main log,
7067 the original message it was trying to log was written to stderr and debug
7068 output, but if they were not available (the usual case in production), it
7069 was lost. Now it is written to syslog before the two lines that record the
7070 failures to open the logs.
7072 65. Users' Exim filters run in subprocesses under the user's uid. It is
7073 possible for a "deliver" command or an alias in a "personal" command to
7074 provoke an address rewrite. If logging of address rewriting is configured,
7075 this fails because the process is not running as root or exim. There may be
7076 a better way of dealing with this, but for the moment (because 4.30 needs
7077 to be released), I have disabled address rewrite logging when running a
7078 filter in a non-root, non-exim process.
7084 1. The buildconfig auxiliary program wasn't quoting the value set for
7085 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7086 not defined. This bug was masked in 4.22 by the effect that was fixed in
7089 2. Some messages that were rejected after a message id was allocated were
7090 shown as "incomplete" by exigrep. It no longer does this for messages that
7091 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7093 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7094 have allow_domain_literals set, the ID did not get logged in the <= line.
7095 Domain literals are now always recognized in Message-ID: header lines.
7097 4. The first argument for a ${extract expansion item is the key name or field
7098 number. Leading and trailing spaces in this item were not being ignored,
7099 causing some misleading effects.
7101 5. When deliver_drop_privilege was set, single queue runner processes started
7102 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7103 same command in the process it spins off) were not dropping privilege.
7105 6. When the daemon running as "exim" started a queue runner, it always
7106 re-executed Exim in the spun-off process. This is a waste of effort when
7107 deliver_drop_privilege is set. The new process now just calls the
7108 queue-runner function directly.
7114 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7117 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7118 ignored. Though the use of -oP was forcing the writing of a pid file, it
7119 was always written to the default place.
7121 3. If the message "no IP address found for host xxxx" is generated during
7122 incoming verification, it is now followed by identification of the incoming
7123 connection (so you can more easily find what provoked it).
7125 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7127 5. Added some features to "harden" Exim a bit more against certain attacks:
7129 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7130 be put in Local/Makefile. This is like the never_users runtime option,
7131 but it cannot be overridden. The default setting is "root".
7133 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7134 prefix string with which any file named in a -C command line option
7137 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7138 is retained for -C and -D only if the caller of Exim is root. Without
7139 it, the exim user may also use -C and -D and retain privilege.
7141 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7142 command line option is disabled.
7144 6. Macro names set by the -D option must start with an upper case letter, just
7145 like macro names defined in the configuration file.
7147 7. Added "dereference=" facility to LDAP.
7149 8. Two instances of the typo "uknown" in the source files are fixed.
7151 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7152 the Configure-Makefile script screwed up while processing it.
7154 10. Incorporated PCRE 4.4.
7156 11. The SMTP synchronization check was not operating right at the start of an
7157 SMTP session. For example, it could not catch a HELO sent before the client
7158 waited for the greeting. There is now a check for outstanding input at the
7159 point when the greeting is written. Because of the duplex, asynchronous
7160 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7161 way, but not yet received, when the check is performed.
7163 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7164 on TCP/IP sockets, because this apparently causes some broken clients to
7167 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7168 unchanged) from the Cygwin maintainer.
7170 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7171 was supported instead of testing for mbx. Effectively a typo.
7173 15. The spa authenticator server code was not checking that the input it
7174 received was valid base64.
7176 16. The debug output line for the "set" modifier in ACLs was not showing the
7177 name of the variable that was being set.
7179 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7181 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7182 lookup on its own. Something else was needed to trigger the lookup. For
7183 example, a match in host_lookup or the need for a host name in a host list.
7184 Now, if $sender_host_name is referenced and the host name has not yet been
7185 looked up, a lookup is performed. If the lookup fails, the variable remains
7186 empty, and $host_lookup_failed is set to "1".
7188 19. Added "eqi" as a case-independent comparison operator.
7190 20. The saslauthd authentication condition could segfault if neither service
7191 nor realm was specified.
7193 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7194 error message that was logged was misleading, and incoming SMTP
7195 connections were dropped. The message is now more accurate, and temporary
7196 errors are given to SMTP connections.
7198 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7199 (except RSET) with a 421 error, until QUIT is received. However, it was
7200 failing to send a response to QUIT.
7202 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7203 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7204 ACL is now run after the helo_try_verify_hosts code.
7206 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7207 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7208 case-independent, so other case variants are also recognized). Apparently
7209 some systems use these upper case variants.
7211 25. If more than two messages were waiting for the same host, and a transport
7212 filter was specified for the transport, Exim sent two messages over the
7213 same TCP/IP connection, and then failed with "socket operation on non-
7214 socket" when it tried to send the third.
7216 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7218 27. The extern definition of crypt16() in expand.c was not being excluded when
7219 the OS had its own crypt16() function.
7221 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7222 as a preferred synonym for return_size_limit, both as an option and as an
7225 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7227 30. Changed the default configuration ACL to relax the local part checking rule
7228 for addresses that are not in any local domains. For these addresses,
7229 slashes and pipe symbols are allowed within local parts, but the sequence
7230 /../ is explicitly forbidden.
7232 31. SPA server authentication was not clearing the challenge buffer before
7235 32. log_message in a "warn" ACL statement was writing to the reject log as
7236 well as to the main log, which contradicts the documentation and doesn't
7237 seem right (because no rejection is happening). So I have stopped it.
7239 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7240 However, I am unable to do any testing of this.
7242 34. Fixed an infelicity in the appendfile transport. When checking directories
7243 for a mailbox, to see if any needed to be created, it was accidentally
7244 using path names with one or more superfluous leading slashes; tracing
7245 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7247 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7248 contents are added to the log line that is written for every discarded
7249 recipient. (Previously a log_message setting was ignored.)
7251 36. The ${quote: operator now quotes the string if it is empty.
7253 37. The install script runs exim in order to find its version number. If for
7254 some reason other than non-existence or emptiness, which it checks, it
7255 could not run './exim', it was installing it with an empty version number,
7256 i.e. as "exim-". This error state is now caught, and the installation is
7259 38. An argument was missing from the function that creates an error message
7260 when Exim fails to connect to the socket for saslauthd authentication.
7261 This could cause Exim to crash, or give a corrupted message.
7263 39. Added isip, isip4, and isip6 to ${if conditions.
7265 40. The ACL variables $acl_xx are now saved with the message, and can be
7266 accessed later in routers, transports, and filters.
7268 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7269 strings in the file are not string-expanded.
7271 42. If a MAIL command specified a SIZE value that was too large to fit into an
7272 int variable, the check against message_size_limit failed. Such values are
7273 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7274 day this will have to be increased, but I don't think I want to be around
7275 when emails are that large.
7282 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7283 iconv() is not standard in FreeBSD.
7285 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7286 IPv6 enabled. The observed symptom was a segmentation fault on return from
7287 the function os_common_find_running_interfaces() in src/os.c.
7289 3. In the check_special_case() function in daemon.c I had used "errno" as an
7290 argument name, which causes warnings on some systems. This was basically a
7291 typo, since it was named "eno" in the comments!
7293 4. The code that waits for the clock to tick (at a resolution of some fraction
7294 of a second) so as to ensure message-id uniqueness was always waiting for
7295 at least one whole tick, when it could have waited for less. [This is
7296 almost certainly not relevant at current processor speeds, where it is
7297 unlikely to ever wait at all. But we try to future-proof.]
7299 5. The function that sleeps for a time interval that includes fractions of a
7300 second contained a race. It did not block SIGALRM between setting the
7301 timer, and suspending (a couple of lines later). If the interval was short
7302 and the sigsuspend() was delayed until after it had expired, the suspension
7303 never ended. On busy systems this could lead to processes getting stuck for
7306 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7307 process, before it forks any delivery processes. The open lookup caching
7308 mechanism meant that the open file or database connection was passed into
7309 the delivery process. The problem was that delivery processes always tidy
7310 up cached lookup data. This could cause a problem for the next delivery
7311 process started by the queue runner, because the external queue runner
7312 process does not know about the closure. So the next delivery process
7313 still has data in the lookup cache. In the case of a file lookup, there was
7314 no problem because closing a file descriptor in a subprocess doesn't affect
7315 the parent. However, if the lookup was caching a connection to a database,
7316 the connection was closed, and the second delivery process was likely to
7317 see errors such as "PGSQL: query failed: server closed the connection
7318 unexpectedly". The problem has been fixed by closing all cached lookups
7319 in a queue runner before running a delivery process.
7321 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7322 seem to have the "const" qualifier which it has on other OS. I've
7325 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7326 *of the same type* (client or server) with the same public name that an
7327 error should be diagnosed.
7329 9. When Exim looked up a host name for an IP address, but failed to find the
7330 original IP address when looking up the host name (a safety check), it
7331 output the message "<ip address> does not match any IP for NULL", which was
7332 confusing, to say the least. The bug was that the host name should have
7333 appeared instead of "NULL".
7335 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7336 user that is built into the binary, and the -C or -D options is used, root
7337 privilege is dropped before the configuration file is read. In addition,
7338 logging is switched to stderr instead of the normal log files. If the
7339 configuration then re-defines the Exim user, the unprivileged environment
7340 is probably not what is expected, so Exim logs a panic warning message (but
7343 However, if deliver_drop_privilege is set, the unprivileged state may well
7344 be exactly what is intended, so the warning has been cut out in that case,
7345 and Exim is allowed to try to write to its normal log files.
7351 1. smtp_return_error_details was not giving details for temporary sender
7352 or receiver verification errors.
7354 2. Diagnose a configuration error if two authenticators have the same public
7357 3. Exim used not to create the message log file for a message until the first
7358 delivery attempt. This could be confusing when incoming messages were held
7359 for policy or load reasons. The message log file is now created at the time
7360 the message is received, and an initial "Received" line is written to it.
7362 4. The automatically generated man page for command line options had a minor
7363 bug that caused no ill effects; however, a more serious problem was that
7364 the procedure for building the man page automatically didn't always
7365 operate. Consequently, release 4.20 contains an out-of-date version. This
7366 shouldn't happen again.
7368 5. When building Exim with embedded Perl support, the script that builds the
7369 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7370 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7372 6. The freeze_tell option was not being used for messages that were frozen on
7373 arrival, either by an ACL or by local_scan().
7375 7. Added the smtp_incomplete_transaction log selector.
7377 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7378 was accepting AUTH without a new EHLO.
7380 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7381 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7384 10. Logging of TCP/IP connections (when configured) now happens in the main
7385 daemon process instead of the child process, so that the TCP/IP connection
7386 count is more accurate (but it can never be perfect).
7388 11. The use of "drop" in a nested ACL was not being handled correctly in the
7389 outer ACL. Now, if condition failure induced by the nested "drop" causes
7390 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7391 or "require"), the connection is dropped.
7393 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7394 that yield "discard" can now be used with an "accept" or a "discard" verb,
7395 but an error is generated for any others (because I can't see a useful way
7396 to define what should happen).
7398 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7399 are now processed in the same way as lines in the Exim configuration file.
7400 In particular, continuation lines are supported.
7402 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7404 15. Added -ti meaning -t -i.
7406 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7407 domains, and warn by logging if others are found.
7409 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7410 returned from GIFCONF when Exim is trying to find the list of interfaces on
7411 a host. The code in os.c has been modified to copy each ifreq to an aligned
7412 structure in all cases.
7414 Also, in some cases, the returned ifreq's were being copied to a 'struct
7415 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7416 means the last couple of bytes of an IPv6 address could be chopped if the
7417 ifreq contained only a normal sockaddr (14 bytes storage).
7419 18. Named domain lists were not supported in the hosts_treat_as_local option.
7420 An entry such as +xxxx was not recognized, and was treated as a literal
7423 19. Ensure that header lines added by a DATA ACL are included in the reject log
7424 if the ACL subsequently rejects the message.
7426 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7427 MD5 (which is deprecated).
7429 21. When testing a filter file using -bf, Exim was writing a message when it
7430 took the sender from a "From " line in the message, but it was not doing so
7431 when it took $return_path from a Return-Path: header line. It now does.
7433 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7434 with a valid header line field name (a series of printing characters
7435 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7437 23. Changed "disc" in the source to "disk" to conform to the documentation and
7438 the book and for uniformity.
7440 24. Ignore Sendmail's -Ooption=value command line item.
7442 25. When execve() failed while trying to run a command in a pipe transport,
7443 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7444 could be confused with a return value of 69 from the command itself. This
7445 has been changed to 127, the value the shell returns if it is asked to run
7446 a non-existent command. The wording for the related log line suggests a
7447 non-existent command as the problem.
7449 26. If received_header_text expands to an empty string, do not add a Received:
7450 header line to the message. (Well, it adds a token one on the spool, but
7451 marks it "old" so that it doesn't get used or transmitted.)
7453 27. Installed eximstats 1.28 (addition of -nt option).
7455 28. There was no check for failure on the call to getsockname() in the daemon
7456 code. This can fail if there is a shortage of resources on the system, with
7457 ENOMEM, for example. A temporary error is now given on failure.
7459 29. Contrary to the C standard, it seems that in some environments, the
7460 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7461 program. Exim now does this explicitly; it affects the formatting of
7462 timestamps using strftime().
7464 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7465 complaints. I've now initialized all the variables, to avoid this.
7467 32. Header lines added by a system filter were not being "seen" during
7468 transport-time rewrites.
7470 33. The info_callback() function passed to OpenSSL is set up with type void
7471 (*)(SSL *, int, int), as described somewhere. However, when calling the
7472 function (actually a macro) that sets it up, the type void(*)() is
7473 expected. I've put in a cast to prevent warnings from picky compilers.
7475 34. If a DNS black list lookup found a CNAME record, but there were no A
7476 records associated with the domain it pointed at, Exim crashed.
7478 35. If a DNS black list lookup returned more than one A record, Exim ignored
7479 all but the first. It now scans all returned addresses if a particular IP
7480 value is being sought. In this situation, the contents of the
7481 $dnslist_value variable are a list of all the addresses, separated by a
7484 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7485 to accept a host name and all its aliases if the forward lookup for any of
7486 them yielded the IP address of the incoming connection. Now it accepts only
7487 those names whose forward lookup yields the correct IP address. Any other
7488 names are discarded. This closes a loophole whereby a rogue DNS
7489 administrator could create reverse DNS records to break through a
7490 wildcarded host restriction in an ACL.
7492 37. If a user filter or a system filter that ran in a subprocess used any of
7493 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7494 the wrong values were passed to the pipe command ($thisaddress had the
7495 value of $0, $0 had the value of $1, etc). This bug was introduced by
7496 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7498 38. Improved the line breaking for long SMTP error messages from ACLs.
7499 Previously, if there was no break point between 40 and 75 characters, Exim
7500 left the rest of the message alone. Two changes have been made: (a) I've
7501 reduced the minimum length to 35 characters; (b) if it can't find a break
7502 point between 35 and 75 characters, it looks ahead and uses the first one
7503 that it finds. This may give the occasional overlong line, but at least the
7504 remaining text gets split now.
7506 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7507 file descriptors might be low, and that setting 1000 would always raise it.
7508 It turns out that in some environments, the limit is already over 1000 and
7509 that lowering it causes trouble. So now Exim takes care not to decrease it.
7511 40. When delivering a message, the value of $return_path is set to $sender_
7512 address at the start of routing (routers may change the value). By an
7513 oversight, this default was not being set up when an address was tested by
7514 -bt or -bv, which affected the outcome if any router or filter referred to
7517 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7518 log only when "message" or "log_message" are set. However, if one of the
7519 conditions was an address verification, or a call to a nested ACL, the
7520 messages generated by the underlying test were being passed through. This
7521 no longer happens. The underlying message is available in $acl_verify_
7522 message for both "message" and "log_message" expansions, so it can be
7523 passed through if needed.
7525 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7526 new expansion $bh_ to give the encoded byte string without charset
7527 translation. Translation happens only if iconv() is available; HAVE_ICONV
7528 indicates this at build time. HEADERS_CHARSET gives the charset to
7529 translate to; headers_charset can change it in the configuration, and
7530 "headers charset" can change it in an individual filter file.
7532 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7533 that creates RFC 2047 encoded "words" labels them as that charset instead
7534 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7535 expansion operator; (ii) when Exim creates a From: line for a local
7536 message; (iii) when a header line is rewritten to include a "phrase" part.
7538 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7539 buggy, causing it to skip the first lines of messages whose message ID
7540 ended in 'D'. This would not have bitten before Exim release 4.14, because
7541 message IDs were unlikely to end in 'D' before then. The effect was to have
7542 incorrect size information for certain domains.
7544 45. #include "config.h" was missing at the start of the crypt16.c module. This
7545 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7548 46. If there was a timeout during a "random" callout check, Exim treated it as
7549 a failure of the random address, and carried on sending RSET and the real
7550 address. If the delay was just some slowness somewhere, the response to the
7551 original RCPT would be taken as a response to RSET and so on, causing
7552 mayhem of various kinds.
7554 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7555 when I implemented it. It didn't allow for the fact that some option values
7556 may legitimately be negative (e.g. size_addition), and it didn't even do
7557 the right test for positive values.
7559 48. Domain names in DNS records are case-independent. Exim always looks them up
7560 in lower case. Some resolvers return domain names in exactly the case they
7561 appear in the zone file, that is, they may contain uppercase letters. Not
7562 all resolvers do this - some return always lower case. Exim was treating a
7563 change of case by a resolver as a change of domain, similar to a widening
7564 of a domain abbreviation. This triggered its re-routing code and so it was
7565 trying to route what was effectively the same domain again. This normally
7566 caused routing to fail (because the router wouldn't handle the domain
7567 twice). Now Exim checks for this case specially, and just changes the
7568 casing of the domain that it ultimately uses when it transmits the message
7571 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7574 50. If a filter generated a file delivery with a non-absolute name (possible if
7575 no home directory exists for the router), the forbid_file option was not
7578 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7579 the existing equality matching.
7581 52. Exim was using ints instead of ino_t variables in some places where it was
7582 dealing with inode numbers.
7584 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7585 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7586 called TMPDIR, and if it finds it is different, it changes its value.
7588 54. The smtp_printf() function is now made available to local_scan() so
7589 additional output lines can be written before returning. There is also an
7590 smtp_fflush() function to enable the detection of a dropped connection.
7591 The variables smtp_input and smtp_batched_input are exported to
7594 55. Changed the default runtime configuration: the message "Unknown user"
7595 has been removed from the ACL, and instead placed on the localuser router,
7596 using the cannot_route_message feature. This means that any verification
7597 failures that generate their own messages won't get overridden. Similarly,
7598 the "Unrouteable address" message that was in the ACL for unverifiable
7599 relay addresses has also been removed.
7601 56. Added hosts_avoid_esmtp to the smtp transport.
7603 57. The exicyclog script was not checking for the esoteric option
7604 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7605 will work only if exicyclog is run under the appropriate euid.
7607 58. Following a discussion on the list, the rules by which Exim recognises line
7608 endings on incoming messages have been changed. The -dropcr and drop_cr
7609 options are now no-ops, retained only for backwards compatibility. The
7610 following line terminators are recognized: LF CRLF CR. However, special
7611 processing applies to CR:
7613 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7614 nor a local message in the state where . is a terminator.
7616 (ii) If a bare CR is encountered in a header line, an extra space is added
7617 after the line terminator so as not to end the header. The reasoning
7618 behind this is that bare CRs in header lines are most likely either
7619 to be mistakes, or people trying to play silly games.
7621 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7622 was being incorrectly given as 18 bytes larger than it should have been.
7623 This is a VOB (very old bug).
7625 60. This may never have affected anything current, but just in case it has:
7626 When the local host is found other than at the start of a list of hosts,
7627 the local host, those with the same MX, and any that follow, are discarded.
7628 When the list in question was part of a longer list of hosts, the following
7629 hosts (not currently being processed) were also being discarded. This no
7630 longer happens. I'm not sure if this situation could ever has previously
7633 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7636 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7637 header. This is something that is recommended in a new Internet Draft, and
7638 is something that is documented as being done by Sendmail. There are two
7639 possible values. For messages generated by the autoreply transport, Exim
7642 Auto-Submitted: auto-replied
7644 whereas for all other generated messages (e.g. bounces) it adds
7646 Auto-Submitted: auto-generated
7648 63. The "personal" condition in filters now includes a test for the
7649 Auto-Submitted: header. If it contains the string "auto-" the message it
7650 not considered personal.
7652 64. Added rcpt_include_affixes as a generic transport option.
7654 65. Added queue_only_override (default true).
7656 66. Added the syslog_duplication option.
7658 67. If what should have been the first header line of a message consisted of
7659 a space followed by a colon, Exim was mis-interpreting it as a header line.
7660 It isn't of course - it is syntactically invalid and should therefore be
7661 treated as the start of the message body. The misbehaviour could have
7662 caused a number of strange effects, including loss of data in subsequent
7663 header lines, and spool format errors.
7665 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7666 client host had authenticated. This control can now be exercised by an ACL
7667 for more flexibility.
7669 69. By default, callouts do not happen when testing with -bh. There is now a
7670 variant, -bhc, which does actually run the callout code, including
7671 consulting and updating the callout cache.
7673 70. Added support for saslauthd authentication, courtesy of Alexander
7676 71. If statvfs() failed on the spool or log directories while checking their
7677 size for availability, Exim confusingly gave the error "space shortage".
7678 Furthermore, in debugging mode it crashed with a floating point exception.
7679 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7680 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7681 serious problem, Exim now writes to the main and panic logs when this
7682 happens, with details of the failure. It then refuses to accept the
7683 incoming message, giving the message "spool directory problem" or "log
7684 directory problem" with a 421 code for SMTP messages.
7686 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7687 0, 1, and 2 exist, because some OS complain for execs without them (see
7688 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7689 descriptors. However, the code omitted to check that the open succeeded,
7690 causing mysterious errors if for some reason the permissions on /dev/null
7691 got screwed. Now Exim writes a message to the main and panic logs, and
7692 bombs out if it can't open /dev/null.
7694 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7695 interact so that it is all more flexible. It is supposed to remain
7696 backwards compatible. Also added extra_local_interfaces.
7698 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7699 to bomb out with an assertion failure - to the client this appears as a
7700 connection drop. This problem occurs in the part of the code that was taken
7701 from the Samba project. Fortunately, the assertion is in a very simple
7702 function, so I have fixed this by reproducing the function inline in the
7703 one place where it is called, and arranging for authentication to fail
7704 instead of killing the process with assert().
7706 75. The SPA client code was not working when the server requested OEM rather
7707 than Unicode encoding.
7709 76. Added code to make require_files with a specific uid setting more usable in
7710 the case where statting the file as root fails - usually a non-root-mounted
7711 NFS file system. When this happens and the failure is EACCES, Exim now
7712 forks a subprocess and does the per-uid checking as the relevant uid.
7714 77. Added process_log_path.
7716 78. If log_file_path was not explicitly set, a setting of check_log_space or
7717 check_log_inodes was ignored.
7719 79. If a space check for the spool or log partitions fails, the incident is now
7720 logged. Of course, in the latter case the data may get lost...
7722 80. Added the %p formatting code to string_format() so that it can be used to
7723 print addresses in debug_print(). Adjusted all the address printing in the
7724 debugging in store.c to use %p rather than %d.
7726 81. There was a concern that a line of code in smtp_in.c could overflow a
7727 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7728 initially expressed, the concern was not well-founded, because trailing
7729 spaces are removed early. However, if the trailing spaces were followed by
7730 a NULL, they did not get removed, so the overflow was possible. Two fixes
7733 (a) I re-wrote the offending code in a cleaner fashion.
7734 (b) If an incoming SMTP command contains a NULL character, it is rejected
7737 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7738 runs initgroups(), so that if the Exim user is in any additional groups,
7739 they will be used during message reception.
7745 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob