1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
8 Since Exim version 4.90
11 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
12 Previously only the last row was returned.
14 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
15 we assumed that tags in the header were well-formed, and parsed the
16 element content after inspecting only the first char of the tag.
17 Assumptions at that stage could crash the receive process on malformed
20 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
21 While running the DKIM ACL we operate on the Permanent memory pool so that
22 variables created with "set" persist to the DATA ACL. Also (at any time)
23 DNS lookups that fail create cache records using the Permanent pool. But
24 expansions release any allocations made on the current pool - so a dnsdb
25 lookup expansion done in the DKIM ACL releases the memory used for the
26 DNS negative-cache, and bad things result. Solution is to switch to the
27 Main pool for expansions.
28 While we're in that code, add checks on the DNS cache during store_reset,
29 active in the testsuite.
30 Problem spotted, and debugging aided, by Wolfgang Breyha.
36 JH/01 Rework error string handling in TLS interface so that the caller in
37 more cases is responsible for logging. This permits library-sourced
38 string to be attached to addresses during delivery, and collapses
39 pairs of long lines into single ones.
41 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
42 during configuration. Wildcards are allowed and expanded.
44 JH/02 Rework error string handling in DKIM to pass more info back to callers.
45 This permits better logging.
47 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
48 do not close it down and have the child transport start it up again on
49 the passed-on TCP connection. Instead, proxy the child (and any
50 subsequent ones) for TLS via a unix-domain socket channel. Logging is
51 affected: the continued delivery log lines do not have any DNSSEC, TLS
52 Certificate or OCSP information. TLS cipher information is still logged.
54 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
55 identical IP addresses on different listening ports. Will also affect
58 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
59 add noisy ifdef guards to special-case this sillyness.
60 Patch from Bernd Kuhls.
62 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
63 than 255 are no longer allowed.
65 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
66 Disable the session-cache too, which might reduce our load. Since we
67 currrectly use a new context for every connection, both as server and
68 client, there is no benefit for these.
69 GnuTLS appears to not support tickets server-side by default (we don't
70 call gnutls_session_ticket_enable_server()) but client side is enabled
71 by default on recent versions (3.1.3 +) unless the PFS priority string
74 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
75 <https://reproducible-builds.org/specs/source-date-epoch/>.
77 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
78 the check for any unsuccessful recipients did not notice the limit, and
79 erroneously found still-pending ones.
81 JH/08 Pipeline CHUNKING command and data together, on kernels that support
82 MSG_MORE. Only in-clear (not on TLS connections).
84 JH/09 Avoid using a temporary file during transport using dkim. Unless a
85 transport-filter is involved we can buffer the headers in memory for
86 creating the signature, and read the spool data file once for the
87 signature and again for transmission.
89 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
90 4.77 as the kernel support then wasn't solid, having issues in 64bit
91 mode. Now, it's been long enough. Add support for FreeBSD also.
93 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
94 case where the routing stage had gathered several addresses to send to
95 a host before calling the transport for the first, we previously failed
96 to close down TLS in the old transport process before passing the TCP
97 connection to the new process. The new one sent a STARTTLS command
98 which naturally failed, giving a failed delivery and bloating the retry
99 database. Investigation and fix prototype from Wolfgang Breyha.
101 JH/12 Fix check on SMTP command input synchronisation. Previously there were
102 false-negatives in the check that the sender had not preempted a response
103 or prompt from Exim (running as a server), due to that code's lack of
104 awareness of the SMTP input buffering.
106 PP/04 Add commandline_checks_require_admin option.
107 Exim drops privileges sanely, various checks such as -be aren't a
108 security problem, as long as you trust local users with access to their
109 own account. When invoked by services which pass untrusted data to
110 Exim, this might be an issue. Set this option in main configuration
111 AND make fixes to the calling application, such as using `--` to stop
114 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
115 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
116 responses to those, into a single TLS record each way (this usually means
117 a single packet). As a side issue, smtp_enforce_sync now works on TLS
120 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
121 affects you only if you're dancing at the edge of the param size limits.
122 If you are, and this message makes sense to you, then: raise the
123 configured limit or use OpenSSL 1.1. Nothing we can do for older
126 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
127 cmdline element to get the documented default one. Previously it was
130 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
131 are unsuable for non-message-specific reasons - eg. connection timeout,
132 banner-time rejection.
134 JH/16 Fix logging of delivery remote port, when specified by router, under
137 PP/06 Repair manualroute's ability to take options in any order, even if one
138 is the name of a transport.
141 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
143 JH/17 Change the list-building routines interface to use the expanding-string
144 triplet model, for better allocation and copying behaviour.
146 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
147 Previously it was constructed the first time a possibly-matching string
148 was met in the configuration file input during startup; now it is done
151 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
152 compatible one, to avoid the (poorly documented) possibility of a config
153 file in the working directory redirecting the DB files, possibly correpting
154 some existing file. CVE-2017-10140 assigned for BDB.
156 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
157 cache-hot. Previously, although the result was properly cached, the
158 initial verify call returned a defer.
160 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
161 the main verify for receipient in uncached-mode.
163 JH/22 Retire historical build files to an "unsupported" subdir. These are
164 defined as "ones for which we have no current evidence of testing".
166 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
167 if present. Previously it was ignored.
169 JH/24 Start using specified-initialisers in C structure init coding. This is
170 a C99 feature (it's 2017, so now considered safe).
172 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
173 if was a fixed-sized field and bitmask ops via macros; it is now more
176 PP/07 GitHub PR 56: Apply MariaDB build fix.
177 Patch provided by Jaroslav Škarvada.
179 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
180 during Coverity cleanups [4.87 JH/47]
181 Diagnosis and fix provided by Michael Fischer v. Mollard.
183 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
184 the right size to place the terminating semicolon on its own folded
185 line, the header hash was calculated to an incorrect value thanks to
186 the (relaxed) space the fold became.
188 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
189 and confused the parent.
191 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
192 which could crash as a result. This could lead to undeliverable messages.
194 JH/28 Logging: "next input sent too soon" now shows where input was truncated
197 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
198 matters on fast-turnover and PID-randomising systems, which were getting
199 out-of-order delivery.
201 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
202 a possibly-overlapping copy. The symptom was that "Remote host closed
203 connection in response to HELO" was logged instead of the actual 4xx
206 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
207 Previously only that bufferd was discarded, resulting in SYMTP command
210 JH/32 DKIM: when a message has multiple signatures matching an identity given
211 in dkim_verify_signers, run the dkim acl once for each. Previously only
212 one run was done. Bug 2189.
214 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
215 "panic the current process" to "deliberately defer". The panic log is
216 still written with the problem list name; the mail and reject logs now
217 get a temp-reject line for the message that was being handled, saying
218 something like "domains check lookup or other defer". The SMTP 451
219 message is still "Temporary local problem".
221 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
222 A crafted sequence of BDAT commands could result in in-use memory beeing
223 freed. CVE-2017-16943.
225 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
226 from SMTP input. Previously it was always done; now only done for DATA
227 and not BDAT commands. CVE-2017-16944.
229 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
230 to the message (such as an overlong header line). Previously this was
231 not done and we did not exit BDAT mode. Followon from the previous item
232 though a different problem.
238 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
239 than -2003 did; needs libidn2 in addition to libidn.
241 JH/02 The path option on a pipe transport is now expanded before use.
243 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
244 Patch provided by "Björn", documentation fix added too.
246 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
247 missing a wire-to-host endian conversion.
249 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
250 close after a BDAT command line could be taken as a following command,
251 giving a synch failure. Fix by only checking for synch immediately
252 before acknowledging the chunk.
254 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
255 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
256 macro. Patches provided by Josh Soref.
258 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
259 Previously we did not; the RFC seems ambiguous and VRFY is not listed
260 by IANA as a service extension. However, John Klensin suggests that we
263 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
264 the dkim code may be unix-mode line endings rather than smtp wire-format
265 CRLF, so prepend a CR to any bare LF.
267 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
268 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
270 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
271 the first were themselves being wrongly included in the feed into dkim
272 processing; with most chunk sizes in use this resulted in an incorrect
273 body hash calculated value.
275 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
276 DKIM signature block, for verification. Although advised against by
277 standards it is specifically not ruled illegal.
279 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
281 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
282 missing a body hash (the bh= tag).
284 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
285 It seems that HAProxy sends the Proxy Protocol information in clear and
286 only then does a TLS startup, so do the same.
288 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
289 TCP connections (such as for Spamd) unless the daemon successfully set
290 Fast Open mode on its listening sockets. This fixes breakage seen on
291 too-old kernels or those not configured for Fast Open, at the cost of
292 requiring both directions being enabled for TFO, and TFO never being used
293 by non-daemon-related Exim processes.
295 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
296 endings, at least on the first header line. Try to canonify any that get
297 past that check, despite the cost.
299 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
300 now limited to an arbitrary five deep, while parsing addresses with the
301 strip_excess_angle_brackets option enabled.
303 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
304 instead leave the unprompted TLS handshake in socket buffer for the
305 TLS library to consume.
307 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
309 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
311 JH/16 Drop variables when they go out of scope. Memory management drops a whole
312 region in one operation, for speed, and this leaves assigned pointers
313 dangling. Add checks run only under the testsuite which checks all
314 variables at a store-reset and panics on a dangling pointer; add code
315 explicitly nulling out all the variables discovered. Fixes one known
316 bug: a transport crash, where a dangling pointer for $sending_ip_address
317 originally assigned in a verify callout, is re-used.
319 PP/06 Drop '.' from @INC in various Perl scripts.
321 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
323 PP/08 Reduce a number of compilation warnings under clang; building with
324 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
325 should be warning-free.
327 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
329 HS/01 Fix portability problems introduced by PP/08 for platforms where
330 realloc(NULL) is not equivalent to malloc() [SunOS et al].
332 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
333 chunk. This allows us to accept broken chunked messages. We need a more
334 general solution here.
336 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
337 already-broken messages in the queue.
339 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
341 JH/19 Fix reference counting bug in routing-generated-address tracking.
347 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
348 supports it and a size is available (ie. the sending peer gave us one).
350 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
351 years of being deprecated). The replacements are the ACLs
352 acl_smtp_mime and acl_not_smtp_mime.
354 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
355 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
356 or even an in-clear connection were permitted. Now, if the host lookup
357 was dnssec and dane was requested then the host is only used if the
358 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
359 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
360 if one fails this test.
361 This means that a poorly-configured remote DNS will make it incommunicado;
362 but it protects against a DNS-interception attack on it.
364 JH/04 Bug 1810: make continued-use of an open smtp transport connection
365 non-noisy when a race steals the message being considered.
367 JH/05 If main configuration option tls_certificate is unset, generate a
368 self-signed certificate for inbound TLS connections.
370 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
371 in rewrites and routers.
373 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
374 and logged a warning sing 4.83; now they are a configuration file error.
376 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
377 (lacking @domain). Apply the same qualification processing as RCPT.
379 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
381 JH/10 Support ${sha256:} applied to a string (as well as the previous
384 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
385 a cutthrough deliver is pending, as we always want to make a connection.
386 This also avoids re-routing the message when later placing the cutthrough
387 connection after a verify cache hit.
388 Do not update it with the verify result either.
390 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
391 when routing results in more than one destination address.
393 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
394 signing (which inhibits the cutthrough capability). Previously only
395 the presence of an option was tested; now an expansion evaluating as
396 empty is permissible (obviously it should depend only on data available
397 when the cutthrough connection is made).
399 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
400 the relevant preceding SMTP command did not note the pipelining mode.
402 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
403 Previously they were not counted.
405 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
406 as one having no matching records. Previously we deferred the message
407 that needed the lookup.
409 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
410 distinguished as "(=".
412 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
413 for missing MX records. Previously it only worked for missing A records.
415 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
417 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
418 after the data-go-ahead and data-ack. Patch from Jason Betts.
420 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
421 even for a "none" policy. Patch from Tony Meyer.
423 JH/22 Fix continued use of a connection for further deliveries. If a port was
424 specified by a router, it must also match for the delivery to be
427 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
428 When one of the recipients of a message was unsuitable for the connection
429 (has no matching addresses), we lost track of needing to mark it
430 deferred. As a result mail would be lost.
432 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
434 JH/25 Decoding ACL controls is now done using a binary search; the source code
435 takes up less space and should be simpler to maintain. Merge the ACL
436 condition decode tables also, with similar effect.
438 JH/26 Fix problem with one_time used on a redirect router which returned the
439 parent address unchanged. A retry would see the parent address marked as
440 delivered, so not attempt the (identical) child. As a result mail would
443 JH/27 Fix a possible security hole, wherein a process operating with the Exim
444 UID can gain a root shell. Credit to http://www.halfdog.net/ for
445 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
448 JH/28 Enable {spool,log} filesystem space and inode checks as default.
449 Main config options check_{log,spool}_{inodes,space} are now
450 100 inodes, 10MB unless set otherwise in the configuration.
452 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
453 Previously it only applied to the main-section connection policy
456 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
458 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
459 by me. Added RFC7919 DH primes as an alternative.
461 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
462 are not in the system include path.
464 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
465 GnuTLS, when a session startup failed (eg because the client disconnected)
466 Exim did stdio operations after fclose. This was exposed by a recent
467 change which nulled out the file handle after the fclose.
469 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
470 signed directly by the cert-signing cert, rather than an intermediate
471 OCSP-signing cert. This is the model used by LetsEncrypt.
473 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
475 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
476 an incoming connection.
478 HS/02 Bug 1802: Do not half-close the connection after sending a request
481 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
482 fallback to "prime256v1".
484 JH/34 SECURITY: Use proper copy of DATA command in error message.
485 Could leak key material. Remotely exploitable. CVE-2016-9963.
491 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
492 and 3.4.4 - once the server is enabled to respond to an OCSP request
493 it does even when not requested, resulting in a stapling non-aware
494 client dropping the TLS connection.
496 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
497 support variable-length bit vectors. No functional change.
499 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
500 The I= interface field on outgoing lines is now after the H= remote
501 host field, same as incoming lines. There is a separate
502 outgoing_interface log selector which allows you to disable the
505 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
506 If not running log_selector +smtp_connection the mainlog would be held
507 open indefinitely after a "too many connections" event, including to a
508 deleted file after a log rotate. Leave the per net connection logging
509 leaving it open for efficiency as that will be quickly detected by the
510 check on the next write.
512 HS/01 Bug 1671: Fix post transport crash.
513 Processing the wait-<transport> messages could crash the delivery
514 process if the message IDs didn't exist for some reason. When
515 using 'split_spool_directory=yes' the construction of the spool
516 file name failed already, exposing the same netto behaviour.
518 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
519 mime_regex ACL conditions.
521 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
522 to DSN fail messages (bounces): remote IP, remote greeting, remote response
523 to HELO, local diagnostic string.
525 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
526 log line to debug. Even when configured with a tls authenticator many
527 client connections are expected to not authenticate in this way, so
528 an authenticate fail is not an error.
530 HS/02 Add the Exim version string to the process info. This way exiwhat
531 gives some more detail about the running daemon.
533 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
534 matter for fast-change records such as DNSBLs.
536 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
537 retry record, even if constant. There may be multiple transports with
538 different interface settings and the retry behaviour needs to be kept
541 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
543 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
545 JH/10 Bug 840: fix log_defer_output option of pipe transport
547 JH/11 Bug 830: use same host for all RCPTS of a message, even under
548 hosts_randomize. This matters a lot when combined with mua_wrapper.
550 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
551 ${quote_pgsql:<string>} operator.
553 JH/13 Bug 1708: avoid misaligned access in cached lookup.
555 JH/14 Change header file name for freeradius-client. Relevant if compiling
556 with Radius support; from the Gentoo tree and checked under Fedora.
558 JH/15 Bug 1712: Introduce $prdr_requested flag variable
560 JH/16 Bug 1714: Permit an empty string as expansion result for transport
561 option transport_filter, meaning no filtering.
563 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
565 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
566 defaults to "*" (all hosts). The variable is now available when not built
567 with TLS, default unset, mainly to enable keeping the testsuite sane.
568 If a server certificate is not supplied (via tls_certificate) an error is
569 logged, and clients will find TLS connections fail on startup. Presumably
570 they will retry in-clear.
571 Packagers of Exim are strongly encouraged to create a server certificate
572 at installation time.
574 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
575 with the $config_file variable.
577 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
578 in transport context, after the attempt, and per-recipient. The latter type
579 is per host attempted. The event data is the error message, and the errno
580 information encodes the lookup type (A vs. MX) used for the (first) host,
581 and the trailing two digits of the smtp 4xx response.
583 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
584 to write to mainlog (or rejectlog, paniclog) in the window between file
585 creation and permissions/ownership being changed. Particularly affects
586 installations where exicyclog is run as root, rather than exim user;
587 result is that the running daemon panics and dies.
589 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
591 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
592 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
593 "pri" and "weight". Note that the previous implicit priority given by the
594 list order is no longer honoured.
596 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
599 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
600 by defining SUPPORT_SOCKS.
602 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
603 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
604 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
605 variables are renamed to proxy_{local,external}_{address,port}.
607 JH/27 Move Internationalisation support from Experimental to mainline, enabled
608 for a build by defining SUPPORT_I18N
610 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
611 of the query string, and make ${quote_redis:} do that quoting.
613 JH/29 Move Events support from Experimental to mainline, enabled by default
614 and removable for a build by defining DISABLE_EVENT.
616 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
618 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
619 cached by the daemon.
621 JH/32 Move Redis support from Experimental to mainline, enabled for a build
622 by defining LOOKUP_REDIS. The libhiredis library is required.
624 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
625 keys are given for lookup.
627 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
628 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
629 only supported when built with TLS support. The PolarSSL SHA routines
630 are still used when the TLS library is too old for convenient support.
632 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
633 openssl_options), for security. OpenSSL forces this from version 1.1.0
634 server-side so match that on older versions.
636 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
637 allocation for $value could be released as the expansion processing
638 concluded, but leaving the global pointer active for it.
640 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
641 and to use the domains and local_parts ACL conditions.
643 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
644 incorrectly not doubled on cutthrough transmission, hence seen as a
645 body-termination at the receiving system - resulting in truncated mails.
646 Commonly the sender saw a TCP-level error, and retransmitted the message
647 via the normal store-and-forward channel. This could result in duplicates
648 received - but deduplicating mailstores were liable to retain only the
649 initial truncated version.
651 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
653 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
655 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
656 we're in there, support oversigning also; bug 1309.
658 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
660 HS/04 Add support for keep_environment and add_environment options.
662 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
663 either intentional arithmetic overflow during PRNG, or testing config-
666 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
667 delivery resulted in actual delivery. Cancel cutthrough before DATA
670 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
671 rejects a recipient: pass the reject to the originator.
673 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
674 Many were false-positives and ignorable, but it's worth fixing the
677 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
678 for the new environment-manipulation done at startup. Move the routines
679 from being local to tls.c to being global via the os.c file.
681 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
682 an extract embedded as result-arg for a map, the first arg for extract
683 is unavailable so we cannot tell if this is a numbered or keyed
684 extraction. Accept either.
690 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
693 JH/02 The smtp transport option "multi_domain" is now expanded.
695 JH/03 The smtp transport now requests PRDR by default, if the server offers
698 JH/04 Certificate name checking on server certificates, when exim is a client,
699 is now done by default. The transport option tls_verify_cert_hostnames
700 can be used to disable this per-host. The build option
701 EXPERIMENTAL_CERTNAMES is withdrawn.
703 JH/05 The value of the tls_verify_certificates smtp transport and main options
704 default to the word "system" to access the system default CA bundle.
705 For GnuTLS, only version 3.0.20 or later.
707 JH/06 Verification of the server certificate for a TLS connection is now tried
708 (but not required) by default. The verification status is now logged by
709 default, for both outbound TLS and client-certificate supplying inbound
712 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
715 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
716 Status Notification (bounce) messages are now MIME format per RFC 3464.
717 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
718 under the control of the dsn_advertise_hosts option, and routers may
719 have a dsn_lasthop option.
721 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
722 default, modifiable by a malware= option. The list separator for
723 the options can now be changed in the usual way. Bug 68.
725 JH/10 The smtp_receive_timeout main option is now expanded before use.
727 JH/11 The incoming_interface log option now also enables logging of the
728 local interface on delivery outgoing connections.
730 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
731 if the interface and destination host and port all match.
733 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
736 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
737 Patch from Andrew Lewis.
739 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
740 now supports optional time-restrictions, weighting, and priority
741 modifiers per server. Patch originally by <rommer@active.by>.
743 JH/16 The spamd_address main option now supports a mixed list of local
744 and remote servers. Remote servers can be IPv6 addresses, and
745 specify a port-range.
747 JH/17 Bug 68: The spamd_address main option now supports an optional
748 timeout value per server.
750 JH/18 Bug 1581: Router and transport options headers_add/remove can
751 now have the list separator specified.
753 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
756 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
759 JH/21 Support for the A6 type of dns record is withdrawn.
761 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
762 rather than the verbs used.
764 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
765 from 255 to 1024 chars.
767 JH/24 Verification callouts now attempt to use TLS by default.
769 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
770 are generic router options now. The defaults didn't change.
772 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
773 Original patch from Alexander Shikoff, worked over by JH.
775 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
776 files if ZCAT_COMMAND is not executable.
778 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
780 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
782 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
783 Normally benign, it bites when the pair was led to by a CNAME;
784 modern usage is to not canonicalize the domain to a CNAME target
785 (and we were inconsistent anyway for A-only vs AAAA+A).
787 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
789 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
790 when evaluating $sender_host_dnssec.
792 JH/31 Check the HELO verification lookup for DNSSEC, adding new
793 $sender_helo_dnssec variable.
795 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
797 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
799 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
801 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
802 documented as working, but never had. Support all but $spam_report.
804 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
805 added for tls authenticator.
807 HS/03 Add perl_taintmode main config option
813 TL/01 When running the test suite, the README says that variables such as
814 no_msglog_check are global and can be placed anywhere in a specific
815 test's script, however it was observed that placement needed to be near
816 the beginning for it to behave that way. Changed the runtest perl
817 script to read through the entire script once to detect and set these
818 variables, reset to the beginning of the script, and then run through
819 the script parsing/test process like normal.
821 TL/02 The BSD's have an arc4random API. One of the functions to induce
822 adding randomness was arc4random_stir(), but it has been removed in
823 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
824 function when detected.
826 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
827 cause callback expansion.
829 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
830 syntax errors in an expansion can be treated as a string instead of
831 logging or causing an error, due to the internal use of bool_lax
832 instead of bool when processing it.
834 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
835 server certificates when making smtp deliveries.
837 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
839 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
841 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
843 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
844 Merged patch from Sebastian Wiedenroth.
846 JH/05 Fix results-pipe from transport process. Several recipients, combined
847 with certificate use, exposed issues where response data items split
848 over buffer boundaries were not parsed properly. This eventually
849 resulted in duplicates being sent. This issue only became common enough
850 to notice due to the introduction of connection certificate information,
851 the item size being so much larger. Found and fixed by Wolfgang Breyha.
853 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
854 size buffer was used, resulting in syntax errors when an expansion
857 JH/07 Add support for directories of certificates when compiled with a GnuTLS
858 version 3.3.6 or later.
860 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
861 is EXPERIMENTAL_EVENT, the main-configuration and transport options
862 both become "event_action", the variables become $event_name, $event_data
863 and $event_defer_errno. There is a new variable $verify_mode, usable in
864 routers, transports and related events. The tls:cert event is now also
865 raised for inbound connections, if the main configuration event_action
868 TL/06 In test suite, disable OCSP for old versions of openssl which contained
869 early OCSP support, but no stapling (appears to be less than 1.0.0).
871 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
872 server certificate names available under the smtp transport option
873 "tls_verify_cert_hostname" now do not permit multi-component wildcard
876 JH/10 Time-related extraction expansions from certificates now use the main
877 option "timezone" setting for output formatting, and are consistent
878 between OpenSSL and GnuTLS compilations. Bug 1541.
880 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
881 encoded parameter in the incoming message. Bug 1558.
883 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
884 include certificate info, eximon was claiming there were spoolfile
887 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
889 JH/14 Log delivery-related information more consistently, using the sequence
890 "H=<name> [<ip>]" wherever possible.
892 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
893 are problematic for Debian distribution, omit them from the release
896 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
898 JH/16 Fix string representation of time values on 64bit time_t architectures.
901 JH/17 Fix a null-indirection in certextract expansions when a nondefault
902 output list separator was used.
907 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
908 checkers that were complaining about end of non-void function with no
911 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
912 This was a regression introduced in 4.83 by another bugfix.
914 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
916 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
917 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
923 TF/01 Correctly close the server side of TLS when forking for delivery.
925 When a message was received over SMTP with TLS, Exim failed to clear up
926 the incoming connection properly after forking off the child process to
927 deliver the message. In some situations the subsequent outgoing
928 delivery connection happened to have the same fd number as the incoming
929 connection previously had. Exim would try to use TLS and fail, logging
930 a "Bad file descriptor" error.
932 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
933 utilities have not been installed.
935 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
936 temporary space as the ACL may create new global variables.
938 TL/01 LDAP support uses per connection or global context settings, depending
939 upon the detected version of the libraries at build time.
941 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
942 to extract and use the src ip:port in logging and expansions as if it
943 were a direct connection from the outside internet. PPv2 support was
944 updated based on HAProxy spec change in May 2014.
946 JH/02 Add ${listextract {number}{list}{success}{fail}}.
948 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
949 Properly escape header and check for NULL return.
951 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
954 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
956 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
957 characters in header names, implemented as a verify condition.
958 Contributed by Michael Fischer v. Mollard.
960 TL/05 Rename SPF condition results err_perm and err_temp to standardized
961 results permerror and temperror. Previous values are deprecated but
962 still accepted. In a future release, err_perm and err_temp will be
963 completely removed, which will be a backward incompatibility if the
964 ACL tests for either of these two old results. Patch contributed by
965 user bes-internal on the mailing list.
967 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
969 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
970 selectors, in both main and reject logs.
972 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
975 JH/07 Add malware type "sock" for talking to simple daemon.
977 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
979 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
980 routers/transports under cutthrough routing.
982 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
983 numbers. Touch up "bool" conditional to keep the same definition.
985 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
987 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
989 JH/12 Expand items in router/transport headers_add or headers_remove lists
990 individually rather than the list as a whole. Bug 1452.
992 Required for reasonable handling of multiple headers_ options when
993 they may be empty; requires that headers_remove items with embedded
994 colons must have them doubled (or the list-separator changed).
996 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
997 view the policy declared in the DMARC record. Currently, $dmarc_status
998 is a combined value of both the record presence and the result of the
1001 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1003 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1004 dnslookup router and the smtp transport (applying to the forward
1007 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1008 of ldap servers used for a specific lookup. Patch provided by Heiko
1011 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1012 New variable $lookup_dnssec_authenticated for observability.
1014 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1015 Patch submitted by Lars Timman.
1017 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1019 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1020 Requires trusted mode and valid format message id, aborts otherwise.
1021 Patch contributed by Heiko Schlichting.
1023 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1024 certextract with support for various fields. Bug 1358.
1026 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1027 is requested by default, modifiable by smtp transport option
1030 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1031 operate on certificate variables to give certificate fingerprints
1032 Also new ${sha256:cert_variable}.
1034 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1036 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1037 Christian Aistleitner.
1039 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1041 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1042 file. Patch from Wolfgang Breyha.
1044 JH/25 Expand the coverage of the delivery $host and $host_address to
1045 client authenticators run in verify callout. Bug 1476.
1047 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1048 align with daemon_smtp_ports. Bug 72.
1050 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1051 support and error reporting did not work properly.
1053 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1054 and is readable. Patch from Andrew Colin Kissa.
1056 TL/14 Enhance documentation of ${run expansion and how it parses the
1057 commandline after expansion, particularly in the case when an
1058 unquoted variable expansion results in an empty value.
1060 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1062 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1063 by renaming away from it. Addresses GH issue 15, reported by
1066 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1068 TL/15 SECURITY: prevent double expansion in math comparison functions
1069 (can expand unsanitized data). Not remotely exploitable.
1076 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1078 PP/02 Make -n do something, by making it not do something.
1079 When combined with -bP, the name of an option is not output.
1081 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1084 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1085 $sender_host_name and config options to manage this, and basic check
1088 PP/05 DSCP support for outbound connections and control modifier for inbound.
1090 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1091 (Only plugin which currently uses this is kerberos4, which nobody should
1092 be using, but we should make it available and other future plugins might
1093 conceivably use it, even though it would break NAT; stuff *should* be
1094 using channel bindings instead).
1096 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1097 name; added for Sendmail compatibility; requires admin caller.
1098 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1099 just ignore it); requires trusted caller.
1100 Also parse but ignore: -Ac -Am -X<logfile>
1103 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1105 TL/02 Add +smtp_confirmation as a default logging option.
1107 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1108 Patch by Magnus Holmgren from 2007-02-20.
1110 TL/04 Bugzilla 1281 - Spec typo.
1111 Bugzilla 1283 - Spec typo.
1112 Bugzilla 1290 - Spec grammar fixes.
1114 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1116 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1118 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1119 mailing list by by Dmitry Isaikin.
1121 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1123 JH/02 Support "G" suffix to numbers in ${if comparisons.
1125 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1127 NM/01 Bugzilla 1197 - Spec typo
1128 Bugzilla 1196 - Spec examples corrections
1130 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1132 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1133 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1136 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1137 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1139 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1140 "acl {{name}{arg}...}", and optional args on acl condition
1143 JH/05 Permit multiple router/transport headers_add/remove lines.
1145 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1147 JH/07 Avoid using a waiting database for a single-message-only transport.
1148 Performance patch from Paul Fisher. Bugzilla 1262.
1150 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1153 JH/09 Add $headers_added variable, with content from use of ACL modifier
1154 add_header (but not yet added to the message). Bugzilla 199.
1156 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1157 Pulled from Bugzilla 817 by Wolfgang Breyha.
1159 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1161 (nb: this is the same fix as in Exim 4.80.1)
1163 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1166 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1169 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1171 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1172 advertise SMTP AUTH mechanism to us, instead of a generic
1173 protocol violation error. Also, make Exim more robust to bad
1174 data from the Dovecot auth socket.
1176 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1178 When a queue runner is handling a message, Exim first routes the
1179 recipient addresses, during which it prunes them based on the retry
1180 hints database. After that it attempts to deliver the message to
1181 any remaining recipients. It then updates the hints database using
1184 So if a recipient address works intermittently, it can get repeatedly
1185 deferred at routing time. The retry hints record remains fresh so the
1186 address never reaches the final cutoff time.
1188 This is a fairly common occurrence when a user is bumping up against
1189 their storage quota. Exim had some logic in its local delivery code
1190 to deal with this. However it did not apply to per-recipient defers
1191 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1193 This change adds a proper retry rule check during routing so that the
1194 final cutoff time is checked against the message's age. We only do
1195 this check if there is an address retry record and there is not a
1196 domain retry record; this implies that previous attempts to handle
1197 the address had the retry_use_local_parts option turned on. We use
1198 this as an approximation for the destination being like a local
1199 delivery, as in LMTP.
1201 I suspect this new check makes the old local delivery cutoff check
1202 redundant, but I have not verified this so I left the code in place.
1204 TF/02 Correct gecos expansion when From: is a prefix of the username.
1206 Test 0254 submits a message to Exim with the header
1210 When I ran the test suite under the user fanf2, Exim expanded
1211 the header to contain my full name, whereas it should have added
1212 a Resent-Sender: header. It erroneously treats any prefix of the
1213 username as equal to the username.
1215 This change corrects that bug.
1217 GF/01 DCC debug and logging tidyup
1218 Error conditions log to paniclog rather than rejectlog.
1219 Debug lines prefixed by "DCC: " to remove any ambiguity.
1221 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1223 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1224 Bug spotted by Jeremy Harris; was flawed since initial commit.
1225 Would have resulted in OCSP responses post-SNI triggering an Exim
1226 NULL dereference and crash.
1228 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1230 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1231 Bug detection, analysis and fix by Samuel Thibault.
1232 Bugzilla 1331, Debian bug #698092.
1234 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1236 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1237 Server implementation by Todd Lyons, client by JH.
1238 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1239 config variable "prdr_enable" controls whether the server
1240 advertises the facility. If the client requests PRDR a new
1241 acl_data_smtp_prdr ACL is called once for each recipient, after
1242 the body content is received and before the acl_smtp_data ACL.
1243 The client is controlled by both of: a hosts_try_prdr option
1244 on the smtp transport, and the server advertisement.
1245 Default client logging of deliveries and rejections involving
1246 PRDR are flagged with the string "PRDR".
1248 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1249 fclose(). Diagnosis by Todd Lyons.
1251 PP/17 Update configure.default to handle IPv6 localhost better.
1252 Patch by Alain Williams (plus minor tweaks).
1255 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1256 This is now consistent with GnuTLS, and is now documented: the
1257 previous undocumented portable approach to treating the option as
1258 unset was to force an expansion failure. That still works, and
1259 an empty string is now equivalent.
1261 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1262 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1263 not performing validation itself.
1265 PP/20 Added force_command boolean option to pipe transport.
1266 Patch from Nick Koston, of cPanel Inc.
1268 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1271 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1273 PP/21 Fix eximon continuous updating with timestamped log-files.
1274 Broken in a format-string cleanup in 4.80, missed when I repaired the
1275 other false fix of the same issue.
1276 Report and fix from Heiko Schlichting.
1279 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1280 Report from Prashanth Katuri.
1282 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1283 It's SecureTransport, so affects any MacOS clients which use the
1284 system-integrated TLS libraries, including email clients.
1286 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1287 using a MIME ACL for non-SMTP local injection.
1288 Report and assistance in diagnosis by Warren Baker.
1290 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1292 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1294 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1295 last id that failed so it may be referenced in subsequent ACL's.
1297 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1300 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1301 ldap library initialization, allowing self-signed CA's to be
1302 used. Also properly sets require_cert option later in code by
1303 using NULL (global ldap config) instead of ldap handle (per
1304 session). Bug diagnosis and testing by alxgomz.
1306 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1307 the src/util/ subdirectory.
1309 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1310 renamed to Transport Post Delivery Action by Jeremy Harris, as
1313 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1314 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1315 redis_servers = needs to be configured which will be used by the redis
1316 lookup. Patch from Warren Baker, of The Packet Hub.
1318 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1320 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1321 hostname or reverse DNS when processing a host list. Used suggestions
1322 from multiple comments on this bug.
1324 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1326 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1327 Missed a few lines, added it to make the runtest require no keyboard
1330 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1331 contains upper case chars. Make router use caseful_local_part.
1333 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1334 support when GnuTLS has been built with p11-kit.
1340 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1342 This, or similar/improved, will also be change PP/11 of 4.82.
1348 PP/01 Handle short writes when writing local log-files.
1349 In practice, only affects FreeBSD (8 onwards).
1350 Bugzilla 1053, with thanks to Dmitry Isaikin.
1352 NM/01 Bugzilla 949 - Documentation tweak
1354 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1357 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1359 PP/02 Implemented gsasl authenticator.
1361 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1363 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1364 `pkg-config foo` for cflags/libs.
1366 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1367 with rest of GSASL and with heimdal_gssapi.
1369 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1370 `pkg-config foo` for cflags/libs for the TLS implementation.
1372 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1373 properties get this fed in as external SSF. A number of robustness
1374 and debugging improvements to the cyrus_sasl authenticator.
1376 PP/08 cyrus_sasl server now expands the server_realm option.
1378 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1379 Patch by Jeremy Harris.
1381 PP/10 Added dbmjz lookup type.
1383 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1385 PP/12 MAIL args handles TAB as well as SP, for better interop with
1386 non-compliant senders.
1387 Analysis and variant patch by Todd Lyons.
1389 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1390 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1391 Patch from Dirk Mueller <dmueller@suse.com>
1393 PP/13 tls_peerdn now print-escaped for spool files.
1394 Observed some $tls_peerdn in wild which contained \n, which resulted
1395 in spool file corruption.
1397 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1398 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1399 or write after TLS renegotiation, which otherwise led to messages
1402 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1403 as a tracking header (ie: a signed header comes before the signature).
1404 Patch from Wolfgang Breyha.
1406 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1407 comma-sep list; embedded commas doubled.
1409 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1411 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1413 Report and patch from Dmitry Banschikov.
1415 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1416 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1417 failures. We appear to now support TLS1.1+ with Exim.
1419 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1420 lets Exim select keys and certificates based upon TLS SNI from client.
1421 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1422 before an outbound SMTP session. New log_selector, +tls_sni.
1424 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1425 NULL dereference. Report and patch from Alun Jones.
1427 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1428 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1429 Not seeing resolver debug output on NetBSD, but suspect this is a
1430 resolver implementation change.
1432 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1433 Left warnings. Added "eximon gdb" invocation mode.
1435 PP/21 Defaulting "accept_8bitmime" to true, not false.
1437 PP/22 Added -bw for inetd wait mode support.
1439 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1440 locate the relevant includes and libraries. Made this the default.
1442 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1443 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1445 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1446 This may cause build issues on older platforms.
1448 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1449 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1450 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1451 Added SNI support via GnuTLS too.
1452 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1454 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1456 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1457 Applied second patch from Janne, implementing suggestion to default
1458 multiple-strings-in-record handling to match SPF spec.
1460 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1462 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1463 read-only, out of scope).
1464 Patch from Wolfgang Breyha, report from Stuart Northfield.
1466 PP/29 Fix three issues highlighted by clang analyser static analysis.
1467 Only crash-plausible issue would require the Cambridge-specific
1468 iplookup router and a misconfiguration.
1469 Report from Marcin Mirosław.
1471 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1473 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1474 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1475 As part of this, removing so much warning spew let me fix some minor
1476 real issues in debug logging.
1478 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1479 assignment on my part. Fixed.
1481 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1482 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1483 Janne Snabb (who went above and beyond: thank you).
1485 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1486 string otherwise requires a connection and a bunch more work and it's
1487 relatively easy to get wrong. Should also expose TLS library linkage
1490 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1491 64-bit ${eval} (JH/03).
1493 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1494 GNU libc to support some of the 64-bit stuff, should not lead to
1495 conflicts. Defined before os.h is pulled in, so if a given platform
1496 needs to override this, it can.
1498 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1499 protection layer was required, which is not implemented.
1500 Bugzilla 1254, patch from Wolfgang Breyha.
1502 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1503 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1504 tls_dhparam take prime identifiers. Also unbreak combination of
1505 OpenSSL+DH_params+TLSSNI.
1507 PP/39 Disable SSLv2 by default in OpenSSL support.
1513 PP/01 Solaris build fix for Oracle's LDAP libraries.
1514 Bugzilla 1109, patch from Stephen Usher.
1516 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1518 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1521 TF/02 Fix a couple more cases where we did not log the error message
1522 when unlink() failed. See also change 4.74-TF/03.
1524 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1525 lock up or crash if it happened to be inside a call to libc when it
1526 got a SIGUSR1 from exiwhat.
1528 The SIGUSR1 handler appends the current process status to the process
1529 log which is later printed by exiwhat. It used to use the general
1530 purpose logging code to do this, but several functions it calls are
1531 not safe for signals.
1533 The new output code in the SIGUSR1 handler is specific to the process
1534 log, and simple enough that it's easy to inspect for signal safety.
1535 Removing some special cases also simplifies the general logging code.
1536 Removing the spurious timestamps from the process log simplifies
1539 TF/04 Improved ratelimit ACL condition.
1541 The /noupdate option has been deprecated in favour of /readonly which
1542 has clearer semantics. The /leaky, /strict, and /readonly update modes
1543 are mutually exclusive. The update mode is no longer included in the
1544 database key; it just determines when the database is updated. (This
1545 means that when you upgrade Exim will forget old rate measurements.)
1547 Exim now checks that the per_* options are used with an update mode that
1548 makes sense for the current ACL. For example, when Exim is processing a
1549 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1550 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1551 must specify per_mail/readonly. If you omit the update mode it defaults to
1552 /leaky where that makes sense (as before) or /readonly where required.
1554 The /noupdate option is now undocumented but still supported for
1555 backwards compatibility. It is equivalent to /readonly except that in
1556 ACLs where /readonly is required you may specify /leaky/noupdate or
1557 /strict/noupdate which are treated the same as /readonly.
1559 A useful new feature is the /count= option. This is a generalization
1560 of the per_byte option, so that you can measure the throughput of other
1561 aggregate values. For example, the per_byte option is now equivalent
1562 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1564 The per_rcpt option has been generalized using the /count= mechanism
1565 (though it's more complicated than the per_byte equivalence). When it is
1566 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1567 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1568 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1569 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1570 Note that using per_rcpt with a non-readonly update mode in more than
1571 one ACL will cause the recipients to be double-counted. (The per_mail
1572 and per_byte options don't have this problem.)
1574 The handling of very low rates has changed slightly. If the computed rate
1575 is less than the event's count (usually one) then this event is the first
1576 after a long gap. In this case the rate is set to the same as this event's
1577 count, so that the first message of a spam run is counted properly.
1579 The major new feature is a mechanism for counting the rate of unique
1580 events. The new per_addr option counts the number of different
1581 recipients that someone has sent messages to in the last time period. It
1582 behaves like per_rcpt if all the recipient addresses are different, but
1583 duplicate recipient addresses do not increase the measured rate. Like
1584 the /count= option this is a general mechanism, so the per_addr option
1585 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1586 example, measure the rate that a client uses different sender addresses
1587 with the options per_mail/unique=$sender_address. There are further
1588 details in the main documentation.
1590 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1592 TF/06 Removed a few PCRE remnants.
1594 TF/07 Automatically extract Exim's version number from tags in the git
1595 repository when doing development or release builds.
1597 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1598 Bugzilla 879. Patch from Paul Fisher.
1600 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1601 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1604 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1606 PP/05 Variable $av_failed, true if the AV scanner deferred.
1607 Bugzilla 1078. Patch from John Horne.
1609 PP/06 Stop make process more reliably on build failure.
1610 Bugzilla 1087. Patch from Heiko Schlittermann.
1612 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1613 Bugzilla 1089. Patch from Heiko Schlittermann.
1615 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1616 Bugzilla 1131. Patch from Holger Weiß.
1618 PP/09 Handle IPv6 addresses with SPF.
1619 Bugzilla 860. Patch from Wolfgang Breyha.
1621 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1623 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1626 PP/11 match_* no longer expand right-hand-side by default.
1627 New compile-time build option, EXPAND_LISTMATCH_RHS.
1628 New expansion conditions, "inlist", "inlisti".
1630 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1632 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1634 PP/14 fix log_write() format string regression from TF/03.
1635 Bugzilla 1152. Patch from Dmitry Isaikin.
1641 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1643 PP/02 Harmonised TLS library version reporting; only show if debugging.
1644 Layout now matches that introduced for other libraries in 4.74 PP/03.
1646 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1648 PP/04 New "dns_use_edns0" global option.
1650 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1653 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1654 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1656 TK/01 Updated PolarSSL code to 0.14.2.
1657 Bugzilla 1097. Patch from Andreas Metzler.
1659 PP/07 Catch divide-by-zero in ${eval:...}.
1660 Fixes bugzilla 1102.
1662 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1665 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1666 format-string attack -- SECURITY: remote arbitrary code execution.
1668 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1669 time unintentionally subject to list matching rules, letting the header
1670 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1671 arbitrary string expansion). This allowed for information disclosure.
1673 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1674 INT_MIN/-1 -- value coerced to INT_MAX.
1680 NM/01 Workaround for PCRE version dependency in version reporting
1683 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1684 This fixes portability to compilers other than gcc, notably
1685 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1687 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1688 makefiles for portability to HP-UX and POSIX correctness.
1690 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1691 Also via indented variable definition in the Makefile.
1692 (Debugging by Oliver Heesakkers).
1694 PP/02 Restore caching of spamd results with expanded spamd_address.
1695 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1697 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1698 Improves build reliability. Fix from: Frank Elsner
1700 NM/02 Fix wide character breakage in the rfc2047 coding
1701 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1703 NM/03 Allow underscore in dnslist lookups
1704 Fixes bug 1026. Patch from Graeme Fowler
1706 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1707 Code patches from Adam Ciarcinski of NetBSD.
1709 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1712 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1713 is logged, to avoid truncation. Patch from John Horne.
1715 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1716 Patch from Jakob Hirsch.
1718 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1719 SQL string expansion failure details.
1720 Patch from Andrey Oktyabrski.
1722 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1723 Patch from Simon Arlott.
1725 PP/09 New lookups functionality failed to compile on old gcc which rejects
1726 extern declarations in function scope.
1727 Patch from Oliver Fleischmann
1729 PP/10 Use sig_atomic_t for flags set from signal handlers.
1730 Check getgroups() return and improve debugging.
1731 Fixed developed for diagnosis in bug 927 (which turned out to be
1734 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1735 Patch from Mark Zealey.
1737 PP/12 Bugzilla 1056: Improved spamd server selection.
1738 Patch from Mark Zealey.
1740 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1741 Based on patch from Heiko Schlittermann.
1743 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1744 Patch from Uwe Doering, sign-off by Michael Haardt.
1746 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1747 variable declaration deep within a block. Bug and patch from
1750 PP/15 lookups-Makefile IRIX compatibility coercion.
1752 PP/16 Make DISABLE_DKIM build knob functional.
1754 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1755 Patch by Simon Arlott
1757 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1758 variable argument macros. Our copy now differs from upstream.
1764 TF/01 Failure to get a lock on a hints database can have serious
1765 consequences so log it to the panic log.
1767 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1768 controlled using the smtp_confirmation log selector.
1770 TF/03 Include the error message when we fail to unlink a spool file.
1772 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1773 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1774 for maintaining out-of-tree patches for some time.
1776 PP/01 Bugzilla 139: Documentation and portability issues.
1777 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1778 Handle per-OS dynamic-module compilation flags.
1780 PP/02 Let /dev/null have normal permissions.
1781 The 4.73 fixes were a little too stringent and complained about the
1782 permissions on /dev/null. Exempt it from some checks.
1783 Reported by Andreas M. Kirchwitz.
1785 PP/03 Report version information for many libraries, including
1786 Exim version information for dynamically loaded libraries. Created
1787 version.h, now support a version extension string for distributors
1788 who patch heavily. Dynamic module ABI change.
1790 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1791 privilege escalation vulnerability whereby the Exim run-time user
1792 can cause root to append content of the attacker's choosing to
1795 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1798 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1799 If dropping privileges for untrusted macros, we disabled normal logging
1800 on the basis that it would fail; for the Exim run-time user, this is not
1801 the case, and it resulted in successful deliveries going unlogged.
1802 Fixed. Reported by Andreas Metzler.
1808 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1809 only prepend for the Resent-* case. Fixes regression introduced in
1810 Exim 4.70 by NM/22 for Bugzilla 607.
1812 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1813 increasing numbers of administrators be bitten by this.
1815 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1817 PP/03 Bugzilla 994: added openssl_options main configuration option.
1819 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1821 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1823 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1825 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1826 without a peer certificate, leading to a segfault because of an
1827 assumption that peers always have certificates. Be a little more
1828 paranoid. Problem reported by Martin Tscholak.
1830 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1831 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1832 NB: ClamAV planning to remove STREAM in "middle of 2010".
1833 CL also introduces -bmalware, various -d+acl logging additions and
1834 more caution in buffer sizes.
1836 PP/09 Implemented reverse_ip expansion operator.
1838 PP/10 Bugzilla 937: provide a "debug" ACL control.
1840 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1842 PP/12 Bugzilla 973: Implement --version.
1844 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1846 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1848 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1850 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1851 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1852 ignore trailing whitespace.
1854 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1856 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1859 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1860 Notification from Dr Andrew Aitchison.
1862 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1863 ExtendedDetectionInfo response format.
1864 Notification from John Horne.
1866 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1869 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1870 XSL and documented dependency on system catalogs, with examples of how
1873 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1876 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1877 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1878 configuration file which is writeable by the Exim user or group.
1880 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1881 of configuration files to cover files specified with the -C option if
1882 they are going to be used with root privileges, not just the default
1885 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1886 option (effectively making it always true).
1888 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1889 files to be used while preserving root privileges.
1891 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1892 that rogue child processes cannot use them.
1894 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1895 run-time user, instead of root.
1897 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1898 Exim run-time user without dropping privileges.
1900 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1901 result string, instead of calling string_vformat() twice with the same
1904 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1905 for other users. Others should always drop root privileges if they use
1906 -C on the command line, even for a whitelisted configure file.
1908 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1910 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1916 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1917 $data_path, and $header_path variables; fixed documentation bugs and
1920 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1921 exipick to access non-standard spools, including the "frozen" queue
1924 NM/01 Bugzilla 965: Support mysql stored procedures.
1925 Patch from Alain Williams
1927 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1929 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1930 Patch from Andreas Metzler
1932 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1933 Patch from Kirill Miazine
1935 NM/05 Bugzilla 671: Added umask to procmail example.
1937 JJ/03 installed exipick 20100323.0, fixing doc bug
1939 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1940 directory. Notification and patch from Dan Rosenberg.
1942 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1944 TK/02 Improve log output when DKIM signing operation fails.
1946 MH/01 Treat the transport option dkim_domain as a colon separated
1947 list, not as a single string, and sign the message with each element,
1948 omitting multiple occurences of the same signer.
1950 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1951 Bugzilla 985, 986. Patch by Simon Arlott
1953 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1954 Patch by Simon Arlott
1956 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1957 MBX locking. Notification from Dan Rosenberg.
1963 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1965 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1967 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1969 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1971 NM/04 Bugzilla 915: Flag broken perl installation during build.
1977 TK/01 Added patch by Johannes Berg that expands the main option
1978 "spamd_address" if it starts with a dollar sign.
1980 TK/02 Write list of recipients to X-Envelope-Sender header when building
1981 the mbox-format spool file for content scanning (suggested by Jakob
1984 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1985 (http://www.dcc-servers.net/) support via dccifd. Activated by
1986 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1988 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1989 by Mark Daniel Reidel <mr@df.eu>.
1991 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1992 When building exim an external PCRE library is now needed -
1993 PCRE is a system library on the majority of modern systems.
1994 See entry on PCRE_LIBS in EDITME file.
1996 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1997 conversation. Added nologin parameter to request.
1998 Patch contributed by Kirill Miazine.
2000 TF/01 Do not log submission mode rewrites if they do not change the address.
2002 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2004 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2005 log files in place. Contributed by Roberto Lima.
2007 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2009 TF/03 Bugzilla 615: When checking the local_parts router precondition
2010 after a local_part_suffix or local_part_prefix option, Exim now
2011 does not use the address's named list lookup cache, since this
2012 contains cached lookups for the whole local part.
2014 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2015 Robert Millan. Documentation is in experimental-spec.txt.
2017 TF/04 Bugzilla 668: Fix parallel build (make -j).
2019 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2021 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2022 Patch provided by Jan Srzednicki.
2024 TF/05 Leading white space used to be stripped from $spam_report which
2025 wrecked the formatting. Now it is preserved.
2027 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2028 that they are available at delivery time.
2030 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2032 TF/08 TLS error reporting now respects the incoming_interface and
2033 incoming_port log selectors.
2035 TF/09 Produce a more useful error message if an SMTP transport's hosts
2036 setting expands to an empty string.
2038 NM/06 Bugzilla 744: EXPN did not work under TLS.
2039 Patch provided by Phil Pennock.
2041 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2042 Patch provided by Richard Godbee.
2044 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2045 acl_smtp_notquit, added index entry.
2047 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2048 Patch provided by Eugene Bujak.
2050 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2051 accept(). Patch provided by Maxim Dounin.
2053 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2054 Patch provided by Phil Pennock.
2056 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2058 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2059 Patch provided by Brad "anomie" Jorsch.
2061 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2062 Patch provided by Dean Brooks.
2064 TK/05 Add native DKIM support (does not depend on external libraries).
2066 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2067 Patch provided by Graeme Fowler.
2069 NM/16 Bugzilla 851: Documentation example syntax fix.
2071 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2073 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2076 NM/19 Bugzilla 745: TLS version reporting.
2077 Patch provided by Phil Pennock.
2079 NM/20 Bugzilla 167: bool: condition support.
2080 Patch provided by Phil Pennock.
2082 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2083 clients. Patch provided by Phil Pennock.
2085 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2086 Patch provided by Brad "anomie" Jorsch.
2088 NM/23 Bugzilla 687: Fix misparses in eximstats.
2089 Patch provided by Heiko Schlittermann.
2091 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2092 Patch provided by Heiko Schlittermann.
2094 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2095 plus update to original patch.
2097 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2099 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2100 Patch provided by David Brownlee.
2102 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2104 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2106 NM/30 Bugzilla 888: TLS documentation bugfixes.
2108 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2110 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2111 Unlike the original bugzilla I have changed all shell scripts in src tree.
2113 NM/33 Bugzilla 898: Transport filter timeout fix.
2114 Patch by Todd Rinaldo.
2116 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2117 Patch by Serge Demonchaux.
2119 NM/35 Bugzilla 39: Base64 decode bug fixes.
2120 Patch by Jakob Hirsch.
2122 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2124 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2126 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2128 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2134 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2135 ALT-N's libdkim that I have put here:
2136 http://duncanthrax.net/exim-experimental/
2138 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2139 were called 'true' and it seems that C99 defines that as a reserved
2140 keyword to be used with 'bool' variable types. That means you could
2141 not include C99-style headers which use bools without triggering
2142 build errors in sieve.c.
2144 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2145 as mailq or other aliases. Changed the --help handling significantly
2146 to do whats expected. exim_usage() emits usage/help information.
2148 SC/01 Added the -bylocaldomain option to eximstats.
2150 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2152 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2154 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2160 PH/01 Another patch from the Sieve maintainer.
2162 PH/02 When an IPv6 address is converted to a string for single-key lookup
2163 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2164 dots are used instead of colons so that keys in lsearch files need not
2165 contain colons. This was done some time before quoting was made available
2166 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2167 (notated using the quote facility) so as to distinguish them from IPv4
2168 keys. This meant that lookups for IP addresses in host lists did not work
2169 for iplsearch lookups.
2171 This has been fixed by arranging for IPv6 addresses to be expressed with
2172 colons if the lookup type is iplsearch. This is not incompatible, because
2173 previously such lookups could never work.
2175 The situation is now rather anomalous, since one *can* have colons in
2176 ordinary lsearch keys. However, making the change in all cases is
2177 incompatible and would probably break a number of configurations.
2179 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2182 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2183 conversion specification without a maximum field width, thereby enabling
2184 a rogue spamd server to cause a buffer overflow. While nobody in their
2185 right mind would setup Exim to query an untrusted spamd server, an
2186 attacker that gains access to a server running spamd could potentially
2187 exploit this vulnerability to run arbitrary code as the Exim user.
2189 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2190 $primary_hostname instead of what libspf2 thinks the hosts name is.
2192 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2193 a directory entry by the name of the lookup key. Previously, if a
2194 symlink pointed to a non-existing file or a file in a directory that
2195 Exim lacked permissions to read, a lookup for a key matching that
2196 symlink would fail. Now it is enough that a matching directory entry
2197 exists, symlink or not. (Bugzilla 503.)
2199 PH/03 The body_linecount and body_zerocount variables are now exported in the
2202 PH/04 Added the $dnslist_matched variable.
2204 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2205 This means they are set thereafter only if the connection becomes
2208 PH/06 Added the client_condition to authenticators so that some can be skipped
2209 by clients under certain conditions.
2211 PH/07 The error message for a badly-placed control=no_multiline_responses left
2212 "_responses" off the end of the name.
2214 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2216 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2217 (without spaces) instead of just copying the configuration text.
2219 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2221 PH/11 Added $max_received_linelength.
2223 PH/12 Added +ignore_defer and +include_defer to host lists.
2225 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2226 way in which PCRE > 7.0 is built.
2228 PH/14 Implemented queue_only_load_latch.
2230 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2231 MAIL command. The effect was to mangle the value on 64-bit systems.
2233 PH/16 Another patch from the Sieve maintainer.
2235 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2237 PH/18 If a system quota error occurred while trying to create the file for
2238 a maildir delivery, the message "Mailbox is full" was not appended to the
2239 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2240 only to a quota excession during the actual writing of the file.
2242 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2243 characters?) which causes problems in log lines. The DN values are now
2244 passed through string_printing() before being added to log lines.
2246 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2247 and InterBase are left for another time.)
2249 PH/21 Added message_body_newlines option.
2251 PH/22 Guard against possible overflow in moan_check_errorcopy().
2253 PH/23 POSIX allows open() to be a macro; guard against that.
2255 PH/24 If the recipient of an error message contained an @ in the local part
2256 (suitably quoted, of course), incorrect values were put in $domain and
2257 $local_part during the evaluation of errors_copy.
2263 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2264 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2267 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2268 issue a MAIL command.
2270 PH/02 In an ACL statement such as
2272 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2274 if a client was not listed at all, or was listed with a value other than
2275 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2276 the condition was not true (as it should be), so access was not denied.
2277 The bug was that the ! inversion was incorrectly passed on to the second
2278 item. This has been fixed.
2280 PH/03 Added additional dnslists conditions == and =& which are different from
2281 = and & when the dns lookup returns more than one IP address.
2283 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2284 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2286 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2287 FSYNC, which compiles an option called disable_fsync that allows for
2288 bypassing fsync(). The documentation is heavily laced with warnings.
2290 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2292 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2293 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2294 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2295 including adding "make clean"; (3) Added -fPIC when compiling the test
2296 dynamically loaded module, to get rid of a warning.
2298 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2299 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2300 = 0s. The bug is otherwise harmless.
2302 PH/07 There was a bug in the dovecot authenticator such that the value of
2303 $auth1 could be overwritten, and so not correctly preserved, after a
2304 successful authentication. This usually meant that the value preserved by
2305 the server_setid option was incorrect.
2307 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2309 PH/09 Installed PCRE release 7.0.
2311 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2312 run for batched SMTP input. It is now run at the start of every message
2313 in the batch. While fixing this I discovered that the process information
2314 (output by running exiwhat) was not always getting set for -bs and -bS
2315 input. This is fixed, and it now also says "batched" for BSMTP.
2317 PH/11 Added control=no_pipelining.
2319 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2320 patch, slightly modified), and move the expansion of helo_data till after
2321 the connection is made in the smtp transport (so it can use these
2324 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2326 PH/14 Added log_selector = +pid.
2328 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2330 PH/16 Add ${if forany and ${if forall.
2332 PH/17 Added dsn_from option to vary the From: line in DSNs.
2334 PH/18 Flush SMTP output before performing a callout, unless control =
2335 no_callout_flush is set.
2337 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2338 was true (the default) a successful delivery failed to delete the retry
2339 item, thus causing premature timeout of the address. The bug is now
2342 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2344 PH/21 Long custom messages for fakedefer and fakereject are now split up
2345 into multiline responses in the same way that messages for "deny" and
2346 other ACL rejections are.
2348 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2349 with slight modification.
2351 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2352 draft, changing the syntax and factoring some duplicate code".
2354 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2355 for deliveries of the second and subsequent messages over the same SMTP
2358 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2359 ${reduce, with only minor "tidies".
2361 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2363 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2364 expansion side effects.
2366 PH/27 When a message times out after an over-quota error from an Exim-imposed
2367 quota, the bounce message says "mailbox is full". This message was not
2368 being given when it was a system quota that was exceeded. It now should
2371 MH/03 Made $recipients available in local_scan(). local_scan() already has
2372 better access to the recipient list through recipients_list[], but
2373 $recipients can be useful in postmaster-provided expansion strings.
2375 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2376 in the case of a MAIL command with additional options following the
2377 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2378 were accidentally chopped off.
2380 PH/29 SMTP synchronization checks are implemented when a command is read -
2381 there is a check that no more input is waiting when there shouldn't be
2382 any. However, for some commands, a delay in an ACL can mean that it is
2383 some time before the response is written. In this time, more input might
2384 arrive, invalidly. So now there are extra checks after an ACL has run for
2385 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2386 pipelining has not been advertised.
2388 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2390 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2391 returned in the SMTP response when smtp_return_error_details was false.
2392 This has been fixed.
2394 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2395 instead of the C I/O that was originally supplied, because problems were
2396 reported on Solaris.
2398 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2399 Exim which did not show up earlier: it was assuming that a call to
2400 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2401 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2402 was a macro that became an assignment, so it seemed to work. This has
2403 changed to a proper function call with a void return, hence the compile
2404 error. Exim's code has been fixed.
2406 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2409 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2411 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2413 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2414 because it turns out to be too big or there is a timeout) there is no
2415 "Completed" line in the log. When some messages of this type were
2416 selected by exigrep, they were listed as "not completed". Others were
2417 picked up by some special patterns. I have improved the selection
2418 criteria to be more general.
2420 PH/38 The host_find_failed option in the manualroute router can now be set
2421 to "ignore", to completely ignore a host whose IP address cannot be
2422 found. If all hosts are ignored, the behaviour is controlled by the new
2423 host_all_ignored option.
2425 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2426 homing or because of multiple MX records with /mx) generated more than
2427 one IP address, and the following item turned out to be the local host,
2428 all the secondary addresses of the first item were incorrectly removed
2429 from the list, along with the local host and any following hosts (which
2430 is what is supposed to happen).
2432 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2433 whoever called Exim into the -H file. In the case of the daemon it was
2434 behaving confusingly. When first started, it used values for whoever
2435 started the daemon, but after a SIGHUP it used the Exim user (because it
2436 calls itself on a restart). I have changed the code so that it now always
2439 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2440 message are rejected with the same error (e.g. no authentication or bad
2441 sender address), and a DATA command is nevertheless sent (as can happen
2442 with PIPELINING or a stupid MUA), the error message that was given to the
2443 RCPT commands is included in the rejection of the DATA command. This is
2444 intended to be helpful for MUAs that show only the final error to their
2447 PH/42 Another patch from the Sieve maintainer.
2449 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2450 Eximstats - Fixed some broken HTML links and added missing column headers
2452 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2453 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2455 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2457 PH/43 Yet another patch from the Sieve maintainer.
2459 PH/44 I found a way to check for a TCP/IP connection going away before sending
2460 the response to the final '.' that terminates a message, but only in the
2461 case where the client has not sent further data following the '.'
2462 (unfortunately, this is allowed). However, in many cases there won't be
2463 any further data because there won't be any more messages to send. A call
2464 to select() can be used: if it shows that the input is "ready", there is
2465 either input waiting, or the socket has been closed. An attempt to read
2466 the next input character can distinguish the two cases. Previously, Exim
2467 would have sent an OK response which the client would never have see.
2468 This could lead to message repetition. This fix should cure that, at
2469 least in a lot of common cases.
2471 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2472 advertised in response to EHLO.
2478 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2479 fixed by 4.65/MH/01 (is this a record?) are fixed:
2481 (i) An empty string was always treated as zero by the numeric comparison
2482 operators. This behaviour has been restored.
2484 (ii) It is documented that the numeric comparison operators always treat
2485 their arguments as decimal numbers. This was broken in that numbers
2486 starting with 0 were being interpreted as octal.
2488 While fixing these problems I realized that there was another issue that
2489 hadn't been noticed. Values of message_size_limit (both the global option
2490 and the transport option) were treated as octal if they started with 0.
2491 The documentation was vague. These values are now always treated as
2492 decimal, and I will make that clear in the documentation.
2498 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2499 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2502 MH/01 Don't check that the operands of numeric comparison operators are
2503 integers when their expansion is in "skipping" mode (fixes bug
2504 introduced by 4.64-PH/07).
2506 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2507 child addresses, Exim now panics and dies. Previously, because the count
2508 is held in a short int, deliveries were likely to be lost. As such a
2509 large number of recipients for a single message is ridiculous
2510 (performance will be very, very poor), I have chosen to impose a limit
2511 rather than extend the field.
2517 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2518 leftover -K file (the existence of which was triggered by #402).
2519 While we were at it, introduced process PID as part of the -K
2520 filename. This should rule out race conditions when creating
2523 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2524 processing considerably. Previous code took too long for large mails,
2525 triggering a timeout which in turn triggers #401.
2527 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2528 in the DK code in transports.c. sendfile() is not really portable,
2529 hence the _LINUX specificness.
2531 TF/01 In the add_headers option to the mail command in an Exim filter,
2532 there was a bug that Exim would claim a syntax error in any
2533 header after the first one which had an odd number of characters
2536 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2537 callout verification, Exim cached a "reject" for the entire domain. This
2538 is correct for most verifications, but it is not correct for a recipient
2539 verification with use_sender or use_postmaster set, because in that case
2540 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2541 case of MAIL FROM:<> rejection from other early rejections (e.g.
2542 rejection of HELO). When verifying a recipient using a non-null MAIL
2543 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2544 Whatever the result of the callout, the value of the domain cache is
2545 left unchanged (for any other kind of callout, getting as far as trying
2546 RCPT means that the domain itself is ok).
2548 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2551 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2552 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2553 session, the daemon ignores ECONNECT errors and logs others; it now
2554 ignores EPIPE as well.
2556 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2557 (quoted-printable decoding).
2559 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2560 later the small subsequent patch to fix an introduced bug.
2562 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2564 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2566 PH/08 An error is now given if message_size_limit is specified negative.
2568 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2569 to be given (somewhat) arbitrary names.
2571 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2574 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2575 miscellaneous code fixes
2577 PH/10 Added the log_reject_target ACL modifier to specify where to log
2580 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2581 hostname. This is wrong, because it relates to the incoming message (and
2582 probably the interface on which it is arriving) and not to the outgoing
2583 callout (which could be using a different interface). This has been
2584 changed to use the value of the helo_data option from the smtp transport
2585 instead - this is what is used when a message is actually being sent. If
2586 there is no remote transport (possible with a router that sets up host
2587 addresses), $smtp_active_hostname is used.
2589 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2590 tweaks were necessary in order to get it to work (see also 21 below):
2591 (a) The code assumed that strncpy() returns a negative number on buffer
2592 overflow, which isn't the case. Replaced with Exim's string_format()
2594 (b) There were several signed/unsigned issues. I just did the minimum
2595 hacking in of casts. There is scope for a larger refactoring.
2596 (c) The code used strcasecmp() which is not a standard C function.
2597 Replaced with Exim's strcmpic() function.
2598 (d) The code set only $1; it now sets $auth1 as well.
2599 (e) A simple test gave the error "authentication client didn't specify
2600 service in request". It would seem that Dovecot has changed its
2601 interface. Fortunately there's a specification; I followed it and
2602 changed what the client sends and it appears to be working now.
2604 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2607 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2608 address A is aliased to B and C, where B exists and C does not. Without
2609 -v the output is "A verified" because verification stops after a
2610 successful redirection if more than one address is generated. However,
2611 with -v the child addresses are also verified. Exim was outputting "A
2612 failed to verify" and then showing the successful verification for C,
2613 with its parentage. It now outputs "B failed to verify", showing B's
2614 parentage before showing the successful verification of C.
2616 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2617 look up a TXT record in a specific list after matching in a combined
2620 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2621 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2622 they consult the DNS. I had assumed they would set it the way they
2623 wanted; and indeed my experiments on Linux seem to show that in some
2624 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2625 To be on the safe side, however, I have now made the interface to
2626 host_find_byname() similar to host_find_bydns(), with an argument
2627 containing the DNS resolver options. The host_find_byname() function now
2628 sets these options at its start, just as host_find_bydns() does. The smtp
2629 transport options dns_qualify_single and dns_search_parents are passed to
2630 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2631 of host_find_byname() use the default settings of RES_DEFNAMES
2632 (qualify_single) but not RES_DNSRCH (search_parents).
2634 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2635 spool_read_header() do less string testing, by means of a preliminary
2636 switch on the second character of optional "-foo" lines. (This is
2637 overdue, caused by the large number of possibilities that now exist.
2638 Originally there were few.) While I was there, I also converted the
2639 str(n)cmp tests so they don't re-test the leading "-" and the first
2640 character, in the hope this might squeeze out yet more improvement.
2642 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2643 flag allowing group syntax was set by the header_syntax check but not
2644 turned off, possible causing trouble later; (2) The flag was not being
2645 set at all for the header_verify test, causing "group"-style headers to
2646 be rejected. I have now set it in this case, and also caused header_
2647 verify to ignore an empty address taken from a group. While doing this, I
2648 came across some other cases where the code for allowing group syntax
2649 while scanning a header line wasn't quite right (mostly, not resetting
2650 the flag correctly in the right place). These bugs could have caused
2651 trouble for malformed header lines. I hope it is now all correct.
2653 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2654 with the "reply" argument non-NULL. The code, however (which originally
2655 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2656 but it didn't always do it. This confused somebody who was copying the
2657 code for some other use. I have removed all the tests.
2659 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2660 feature that was used to support insecure browsers during the U.S. crypto
2661 embargo. It requires special client support, and Exim is probably the
2662 only MTA that supported it -- and would never use it because real RSA is
2663 always available. This code has been removed, because it had the bad
2664 effect of slowing Exim down by computing (never used) parameters for the
2665 RSA_EXPORT functionality.
2667 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2668 authenticator to fail if there's a tab character in the incoming data
2669 (there should never be unless someone is messing about, as it's supposed
2670 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2671 if the connection is using TLS or if the remote IP is the same as the
2672 local IP, and the "valid-client-cert option" if a client certificate has
2675 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2676 authenticators. This can be used for authorization after authentication
2677 succeeds. (In the case of plaintext, it servers for both authentication
2680 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2681 if any retry times were supplied.
2683 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2684 connection, where there is no client IP address to check. In this
2685 situation, the verify now always succeeds.
2687 PH/25 Applied John Jetmore's -Mset patch.
2689 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2691 PH/27 In a string expansion for a processed (not raw) header when multiple
2692 headers of the same name were present, leading whitespace was being
2693 removed from all of them, but trailing whitespace was being removed only
2694 from the last one. Now trailing whitespace is removed from each header
2695 before concatenation. Completely empty headers in a concatenation (as
2696 before) are ignored.
2698 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2699 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2701 PH/29 [Removed. This was a change that I later backed out, and forgot to
2702 correct the ChangeLog entry (that I had efficiently created) before
2703 committing the later change.]
2705 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2706 address errors (4xx response to RCPT) over the same connection as other
2707 messages routed to the same hosts. Such deliveries are always "forced",
2708 so retry times are not inspected. This resulted in far too many retries
2709 for the affected addresses. The effect occurred only when there were more
2710 hosts than the hosts_max_try setting in the smtp transport when it had
2711 the 4xx errors. Those hosts that it had tried were not added to the list
2712 of hosts for which the message was waiting, so if all were tried, there
2713 was no problem. Two fixes have been applied:
2715 (i) If there are any address or message errors in an SMTP delivery, none
2716 of the hosts (tried or untried) are now added to the list of hosts
2717 for which the message is waiting, so the message should not be a
2718 candidate for sending over the same connection that was used for a
2719 successful delivery of some other message. This seems entirely
2720 reasonable: after all the message is NOT "waiting for some host".
2721 This is so "obvious" that I'm not sure why it wasn't done
2722 previously. Hope I haven't missed anything, but it can't do any
2723 harm, as the worst effect is to miss an optimization.
2725 (ii) If, despite (i), such a delivery is accidentally attempted, the
2726 routing retry time is respected, so at least it doesn't keep
2727 hammering the server.
2729 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2730 in ${readsocket because some servers need this prod.
2732 PH/32 Added some extra debug output when updating a wait-xxx database.
2734 PH/33 The hint "could be header name not terminated by colon", which has been
2735 given for certain expansion errors for a long time, was not being given
2736 for the ${if def:h_colon_omitted{... case.
2738 PH/34 The spec says: "With one important exception, whenever a domain list is
2739 being scanned, $domain contains the subject domain." There was at least
2740 one case where this was not true.
2742 PH/35 The error "getsockname() failed: connection reset by peer" was being
2743 written to the panic log as well as the main log, but it isn't really
2744 panic-worthy as it just means the connection died rather early on. I have
2745 removed the panic log writing for the ECONNRESET error when getsockname()
2748 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2749 runs only) independently of the message's sender address. This meant
2750 that, if the 4xx error was in fact related to the sender, a different
2751 message to the same recipient with a different sender could confuse
2752 things. In particular, this can happen when sending to a greylisting
2753 server, but other circumstances could also provoke similar problems.
2754 I have changed the default so that the retry time for these errors is now
2755 based a combination of the sender and recipient addresses. This change
2756 can be overridden by setting address_retry_include_sender=false in the
2759 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2760 remote server are returned as part of bounce messages. This was not
2761 happening for LMTP over a pipe (the lmtp transport), but now it is the
2762 same for both kinds of LMTP.
2764 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2765 in header lines that were in fact CNAMEs. This is no longer the case.
2767 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2768 and queue runs started by the daemon processed all messages. This has
2769 been fixed so that -R and -S can now usefully be given with -q<time>.
2771 PH/40 Import PCRE release 6.7 (fixes some bugs).
2773 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2775 PH/42 Give an error if -q is specified more than once.
2777 PH/43 Renamed the variables $interface_address and $interface_port as
2778 $received_ip_address and $received_port, to make it clear that these
2779 values apply to message reception, and not to the outgoing interface when
2780 a message is delivered. (The old names remain recognized, of course.)
2782 PH/44 There was no timeout on the connect() call when using a Unix domain
2783 socket in the ${readsocket expansion. There now is.
2785 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2786 be meaningful with "accept".
2788 SC/01 Eximstats V1.43
2789 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2791 SC/02 Eximstats V1.44
2792 Use a glob alias rather than an array ref in the generated
2793 parser. This improves both readability and performance.
2795 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2796 Collect SpamAssassin and rejection statistics.
2797 Don't display local sender or destination tables unless
2798 there is data to show.
2799 Added average volumes into the top table text output.
2801 SC/04 Eximstats V1.46
2802 Collect data on the number of addresses (recipients)
2803 as well as the number of messages.
2805 SC/05 Eximstats V1.47
2806 Added 'Message too big' to the list of mail rejection
2807 reasons (thanks to Marco Gaiarin).
2809 SC/06 Eximstats V1.48
2810 Mainlog lines which have GMT offsets and are too short to
2811 have a flag are now skipped.
2813 SC/07 Eximstats V1.49 (Alain Williams)
2814 Added the -emptyok flag.
2816 SC/08 Eximstats V1.50
2817 Fixes for obtaining the IP address from reject messages.
2819 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2820 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2821 whitespace changes from 4.64-PH/27
2823 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2826 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2827 are found, allow negative numbers in numeric criteria)
2829 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2831 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2832 to match changes made in 4.64-PH/43
2834 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2836 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2837 "domain" is actually an IP address, and force a failure. This locks out
2838 those revolvers/nameservers that support "A-for-A" lookups, in
2839 contravention of the specifications.
2841 PH/48 When a host name was looked up from an IP address, and the subsequent
2842 forward lookup of the name timed out, the host name was left in
2843 $sender_host_name, contrary to the specification.
2845 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2846 restricted to single-key lookups, Exim was not diagnosing an error if
2847 * or *@ was used with a query-style lookup.
2849 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2851 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2852 long ago, but noone interested enough thought of it. Let's just say that
2853 the "1.1" means that there are some new functions that weren't there at
2854 some point in the past.
2856 PH/51 Error processing for expansion failure of helo_data from an smtp
2857 transport during callout processing was broken.
2859 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2860 tested/used via the -bh/-bhc/-bs options.
2862 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2863 bug, fixed in subsequent PCRE releases).
2865 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2866 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2868 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2874 SC/01 Use a glob alias rather than an array ref in eximstats generated
2875 parser. This improves both readability and performance.
2877 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2878 Don't display local sender or destination tables in eximstats unless
2879 there is data to show.
2880 Added average volumes into the eximstats top table text output.
2882 SC/03 Collect data on the number of addresses (recipients) as well
2883 as the number of messages in eximstats.
2885 TF/01 Correct an error in the documentation for the redirect router. Exim
2886 does (usually) call initgroups() when daemonizing.
2888 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2889 with consistent privilege compared to when running as a daemon.
2891 TF/03 Note in the spec that $authenticated_id is not set for local
2892 submissions from trusted users.
2894 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2895 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2897 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2898 by adding some example configuration directives to the default
2899 configuration file. A little bit of work is required to uncomment the
2900 directives and define how usernames and passwords are checked, but
2901 there is now a framework to start from.
2903 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2904 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2905 without this. I don't know how relevant this is to other LDAP libraries.
2907 PH/02 Add the verb name to the "unknown ACL verb" error.
2909 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2911 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2913 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2914 directory not expanded when it should be if an expanded home directory
2915 was set for the address (which is overridden by the transport).
2917 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2920 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2921 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2922 because it is too late at that time, and has no effect.
2924 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2925 security issue with \' (bugzilla #107). I could not use the
2926 PQescapeStringConn() function, because it needs a PGconn value as one of
2929 PH/08 When testing addresses using -bt, indicate those final addresses that
2930 are duplicates that would not cause an additional delivery. At least one
2931 person was confused, thinking that -bt output corresponded to deliveries.
2932 (Suppressing duplicates isn't a good idea as you lose the information
2933 about possibly different redirections that led to the duplicates.)
2935 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2936 systems where poll() doesn't work, in particular OS X.
2938 PH/10 Added more information to debugging output for retry time not reached.
2940 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2941 operations in malware.c.
2943 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2946 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2947 syslog_duplication set false, log lines that would normally be written
2948 both the the main log and to the reject log were not written to syslog at
2951 PH/14 In the default configuration, change the use of "message" in ACL warn
2952 statements to "add_header".
2954 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2955 not followed by a command (e.g. "seen endif").
2957 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2958 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2961 PH/17 Added extra conditions to the default value of delay_warning_condition
2965 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2966 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2967 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2970 The Auto-Submitted: and various List- headers are standardised, whereas I
2971 don't think Precedence: ever was.
2973 PH/18 Refactored debugging code in route_finduser() to show more information,
2974 in particular, the error code if getpwnam() issues one.
2976 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2977 This is apparently needed in addition to the PH/07 change above to avoid
2978 any possible encoding problems.
2980 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2981 but not after initializing Perl.
2983 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2984 output them only if debugging. By default they are written stderr,
2985 apparently, which is not desirable.
2987 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2990 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2993 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2995 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2996 authenticated or an ident call has been made. Suppress the default
2997 values for $authenticated_id and $authenticated_sender (but permit -oMai
2998 and -oMas) when testing with -bh.
3000 PH/24 Re-jigged the order of the tests in the default configuration so that the
3001 tests for valid domains and recipients precede the DNS black list and CSA
3002 tests, on the grounds that those ones are more expensive.
3004 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3005 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3006 This bug exists in every version of Exim that I still have, right back to
3009 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3010 However, an attempt to turn on case-sensitivity in a regex key by
3011 including (?-i) didn't work because the subject string was already
3012 lowercased, and the effects were non-intuitive. It turns out that a
3013 one-line patch can be used to allow (?-i) to work as expected.
3019 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3020 other effects) broke the use of negated acl sub-conditions.
3022 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3025 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3026 "Deny" causes Exim to reject the incoming connection with a 554 error.
3027 Unfortunately, if there is a major crisis, such as a disk failure,
3028 tcp-wrappers gives "deny", whereas what one would like would be some
3029 kind of temporary error. A kludge has been added to help with this.
3030 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3031 554 error is used if errno is still zero or contains ENOENT (which occurs
3032 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3035 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3037 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3038 errors. Otherwise a message that provokes a temporary error (when other
3039 messages do not) can cause a whole host to time out.
3041 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3042 addresses were routed directly to files or pipes from a redirect router.
3043 File deliveries just didn't batch; pipe deliveries might have suffered
3046 PH/06 A failure to get a lock for a hints database would erroneously always say
3047 "Failed to get write lock", even when it was really a read lock.
3049 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3050 of 0600. This has been changed to use the value of the lockfile_mode
3051 option (which defaults to 0600).
3053 PH/08 Applied small patch from the Sieve maintainer.
3055 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3056 folder from quota calculations, a direct delivery into this folder messed
3057 up the contents of the maildirsize file. This was because the regex was
3058 used only to exclude .Trash (or whatever) when the size of the mailbox
3059 was calculated. There was no check that a delivery was happening into an
3060 excluded directory. This bug has been fixed by ignoring all quota
3061 processing for deliveries into excluded directories.
3063 PH/10 Added the maildirfolder_create_regex option to appendfile.
3069 PH/01 The code for finding all the local interface addresses on a FreeBSD
3070 system running IPv6 was broken. This may well have applied to all BSD
3071 systems, as well as to others that have similar system calls. The broken
3072 code found IPv4 interfaces correctly, but gave incorrect values for the
3073 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3074 that it would not match correctly against @[] and not recognize the IPv6
3077 PH/02 The ipliteral router was not recognizing addresses of the form user@
3078 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3080 PH/03 Added disable_ipv6.
3082 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3083 decoded form, because it is most often used to construct To: headers
3084 lines in autoreplies, and the decoded form may well be syntactically
3085 invalid. However, $reply_address has leading white space removed, and all
3086 newlines turned into spaces so that the autoreply transport does not
3089 PH/05 If group was specified without a user on a router, and no group or user
3090 was specified on a transport, the group from the router was ignored.
3092 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3093 for visible compile-time settings that can be used to change these
3094 numbers, for those that want even more. Backwards compatibility with old
3095 spool files has been maintained. However, going back to a previous Exim
3096 release will lost any variables that are in spool files.
3098 PH/07 Two small changes when running in the test harness: increase delay when
3099 passing a TCP/IP connection to a new process, in case the original
3100 process has to generate a bounce, and remove special handling of
3101 127.0.0.2 (sic), which is no longer necessary.
3103 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3104 be the same on different OS.
3106 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3109 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3110 whether --show-vars was specified or not
3112 JJ/02 exipick: Added support for new ACL variable spool format introduced
3115 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3116 syntactically invalid From: or Reply-to: line, and a filter used this to
3117 generate an autoreply, and therefore failed to obtain an address for the
3118 autoreply, Exim could try to deliver to a non-existent relative file
3119 name, causing unrelated and misleading errors. What now happens is that
3120 it logs this as a hard delivery error, but does not attempt to create a
3123 PH/11 The exinext utility has a -C option for testing purposes, but although
3124 the given file was scanned by exinext itself; it wasn't being passed on
3125 when Exim was called.
3127 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3128 an end-of-file indication when reading a command response.
3130 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3131 compiled. In many other places in Exim, IPv6 addresses are always
3132 recognized, so I have changed this. It also means that IPv4 domain
3133 literals of the form [IPV4:n.n.n.n] are now always recognized.
3135 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3136 used if the router is not running as root, for example, when verifying at
3137 ACL time, or when using -bh. The debugging output from this situation was
3138 non-existent - all you got was a failure to exec. I have made two
3141 (a) Failures to set uid/gid, the current directory, or a process leader
3142 in a subprocess such as that created by queryprogram now generate
3143 suitable debugging output when -d is set.
3145 (b) The queryprogram router detects when it is not running as root,
3146 outputs suitable debugging information if -d is set, and then runs
3147 the subprocess without attempting to change uid/gid.
3149 PH/15 Minor change to Makefile for building test_host (undocumented testing
3152 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3153 additional section of a DNS packet that returns MX or SRV records.
3154 Instead, it always explicitly searches for A/AAAA records. This avoids
3155 major problems that occur when a DNS server includes only records of one
3156 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3157 fixed another bug: if SRV records were looked up and the corresponding
3158 address records were *not* found in the additional section, the port
3159 values from the SRV records were lost.
3161 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3162 using the correct key (the original address) when searching the retry
3163 rules in order to find which one to use for generating the retry hint.
3165 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3166 adding the default one. Similarly, if it contains a Reply-To: header, the
3167 errors_reply_to option, if set, is not used.
3169 PH/19 When calculating a retry time, Exim used to measure the "time since
3170 failure" by looking at the "first failed" field in the retry record. Now
3171 it does not use this if it is later than than the arrival time of the
3172 message. Instead it uses the arrival time. This makes for better
3173 behaviour in cases where some deliveries succeed, thus re-setting the
3174 "first failed" field. An example is a quota failure for a huge message
3175 when small messages continue to be delivered. Without this change, the
3176 "time since failure" will always be short, possible causing more frequent
3177 delivery attempts for the huge message than are intended.
3178 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3180 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3181 $1, $2, $3) because the numerical variables can be reset during some
3182 expansion items (e.g. "match"), thereby losing the authentication data.
3184 PH/21 Make -bV show the size of off_t variables so that the test suite can
3185 decide whether to run tests for quotas > 2G.
3187 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3188 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3189 filecount value is greater than 2G or if a quota value is greater than 2G
3190 on a system where the size of off_t is not greater than 4, a panic error
3193 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3194 never match. The debug and -bh output now contains an explicit error
3195 message indicating a malformed IPv4 address or mask.
3197 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3198 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3199 PH/23 above applies.
3201 PH/25 Do not write to syslog when running in the test harness. The only
3202 occasion when this arises is a failure to open the main or panic logs
3203 (for which there is an explicit test).
3205 PH/26 Added the /no_tell option to "control=freeze".
3207 PH/27 If a host name lookup failed very early in a connection, for example, if
3208 the IP address matched host_lookup and the reverse lookup yielded a name
3209 that did not have a forward lookup, an error message of the form "no IP
3210 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3211 could be logged. Now it outputs the IP address instead of "NULL".
3213 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3214 allows the sender and the authenticated sender to be set when
3215 submitting a message from within Exim. Since child_open_exim() is
3216 documented for local_scan(), the new function should be too.
3218 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3219 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3220 results in an empty string is now treated as unset.
3222 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3224 PH/31 Added sender_verify_fail logging option.
3226 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3227 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3228 tidied the source and removed it altogether.
3230 PH/33 When a queue run was abandoned because the load average was too high, a
3231 log line was always written; now it is written only if the queue_run log
3232 selector is set. In addition, the log line for abandonment now contains
3233 information about the queue run such as the pid. This is always present
3234 in "start" and "stop" lines but was omitted from the "abandon" line.
3236 PH/34 Omit spaces between a header name and the colon in the error message that
3237 is given when verify = headers_syntax fails (if there are lots of them,
3238 the message gets confusing).
3240 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3241 names, as there are now some PTR records that contain slashes. This check
3242 is only to protect against broken name servers that fall over on strange
3243 characters, so the fact that it applies to all lookups doesn't matter.
3245 PH/36 Now that the new test suite is complete, we can remove some of the
3246 special code in Exim that was needed for the old test suite. For example,
3247 sorting DNS records because real resolvers return them in an arbitrary
3248 order. The new test suite's fake resolver always returns records in the
3251 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3252 bounces) except when queue_only is set, to avoid logging races between
3253 the different processes.
3255 PH/38 Panic-die if .include specifies a non-absolute path.
3257 PH/39 A tweak to the "H" retry rule from its user.
3259 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3260 a label. They prevented compilation on older perls.
3262 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3263 a warning to be raised on newish perls.
3265 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3266 on queue. Changes to match documented behaviour of showing count of
3267 messages matching specified criteria.
3269 PH/40 Changed the default ident timeout from 30s to 5s.
3271 PH/41 Added support for the use of login_cap features, on those BSD systems
3272 that have them, for controlling the resources used by pipe deliveries.
3274 PH/42 The content-scanning code uses fopen() to create files in which to put
3275 message data. Previously it was not paying any attention to the mode of
3276 the files. Exim runs with umask(0) because the rest of the code creates
3277 files with open(), and sets the required mode explicitly. Thus, these
3278 files were ending up world-writeable. This was not a big issue, because,
3279 being within the spool directory, they were not world-accessible. I have
3280 created a function called modefopen, which takes an additional mode
3281 argument. It sets umask(777), creates the file, chmods it to the required
3282 mode, then resets the umask. All the relevant calls to fopen() in the
3283 content scanning code have been changed to use this function.
3285 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3286 to 24 hours. This avoids potential overflow problems when processing G
3287 and H retry rules. I suspect nobody ever tinkers with this value.
3289 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3291 PH/45 When the plaintext authenticator is running as a client, the server's
3292 challenges are checked to ensure they are valid base64 strings. By
3293 default, the authentication attempt is cancelled if an invalid string is
3294 received. Setting client_ignore_invalid_base64 true ignores these errors.
3295 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3296 they are received. Thus, the responses can be made to depend on the
3297 challenges. If an invalid string is ignored, an empty string is placed in
3300 PH/46 Messages that are created by the autoreply transport now contains a
3301 References: header, in accordance with RFCs 2822 and 3834.
3303 PH/47 Added authenticated_sender_force to the smtp transport.
3305 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3307 PH/49 Installed latest patch from the Sieve maintainer.
3309 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3310 was also set, the appendfile transport was unnecessarily scanning a
3311 directory of message files (e.g. for maildir delivery) to find the count
3312 of files (along with the size), even though it did not need this
3313 information. It now does the scan only if it needs to find either the
3314 size of the count of files.
3316 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3318 PH/52 Two bugs concerned with error handling when the smtp transport is
3321 (i) Exim was not creating retry information for temporary errors given
3322 for individual recipients after the DATA command when the smtp transport
3323 was used in LMTP mode. This meant that they could be retried too
3324 frequently, and not timed out correctly.
3326 (ii) Exim was setting the flag that allows error details to be returned
3327 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3328 recipients that were returned after the DATA command.
3330 PH/53 This is related to PH/52, but is more general: for any failing address,
3331 when detailed error information was permitted to be returned to the
3332 sender, but the error was temporary, then after the final timeout, only
3333 "retry timeout exceeded" was returned. Now it returns the full error as
3334 well as "retry timeout exceeded".
3336 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3337 do this, and (what is worse) MTAs that accept it.
3339 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3340 will now be deprecated.
3342 PH/56 New os.c-cygwin from the Cygwin maintainer.
3344 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3345 formats (previously only available in exim formats via -bpr, -bpru,
3346 and -bpra. Now also available in native and exiqgrep formats)
3348 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3349 with very large, slow to parse queues
3351 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3353 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3355 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3356 responses to authentication challenges, though it was showing the
3357 challenges; (ii) I've removed the CR characters from the debug output for
3360 PH/58 Allow for the insertion of a newline as well as a space when a string
3361 is turned into more than one encoded-word during RFC 2047 encoding. The
3362 Sieve code now uses this.
3364 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3365 data_4xx, lost_connection, tls_required.
3367 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3368 message was being sent as an SMTP response.
3370 PH/61 Add -l and -k options to exicyclog.
3372 PH/62 When verifying, if an address was redirected to one new address, so that
3373 verification continued, and the new address failed or deferred after
3374 having set something in $address_data, the value of $address_data was not
3375 passed back to the ACL. This was different to the case when no
3376 redirection occurred. The value is now passed back in both cases.
3378 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3379 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3380 use login_cap.h, so on its own it isn't the distinguishing feature. The
3381 new name refers directly to the setclassresources() function.
3383 PH/65 Added configuration files for NetBSD3.
3385 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3387 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3388 is preferred over IPv4.
3390 PH/68 The bounce_return_message and bounce_return_body options were not being
3391 honoured for bounces generated during the reception of non-SMTP messages.
3392 In particular, this applied to messages rejected by the ACL. This bug has
3393 been fixed. However, if bounce_return_message is true and bounce_return_
3394 body is false, the headers that are returned for a non-SMTP message
3395 include only those that have been read before the error was detected.
3396 (In the case of an ACL rejection, they have all been read.)
3398 PH/69 The HTML version of the specification is now built in a directory called
3399 spec_html instead of spec.html, because the latter looks like a path with
3400 a MIME-type, and this confuses some software.
3402 PH/70 Catch two compiler warnings in sieve.c.
3404 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3405 function verify_get_ident() calls ip_connect() to connect a socket, but
3406 if the "connect()" function timed out, ip_connect() used to close the
3407 socket. However, verify_get_ident() also closes the socket later, and in
3408 between Exim writes to the log, which may get opened at this point. When
3409 the socket was closed in ip_connect(), the log could get the same file
3410 descriptor number as the socket. This naturally causes chaos. The fix is
3411 not to close the socket in ip_connect(); the socket should be closed by
3412 the function that creates it. There was only one place in the code where
3413 this was missing, in the iplookup router, which I don't think anybody now
3414 uses, but I've fixed it anyway.
3416 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3417 well as to direct DNS lookups. Otherwise the handling of names in host
3418 lists is inconsistent and therefore confusing.
3424 PH/01 Two changes to the default runtime configuration:
3426 (1) Move the checks for relay_from_hosts and authenticated clients from
3427 after to before the (commented out) DNS black list checks.
3429 (2) Add control=submission to the relay_from_hosts and authenticated
3430 clients checks, on the grounds that messages accepted by these
3431 statements are most likely to be submissions.
3433 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3435 (1) Generate an error if the third argument for the ${prvs expansion is
3438 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3441 (3) Reset the variables that are obtained from the first argument of
3442 ${prvscheck and used in the second argument before leaving the code,
3443 because their memory is reclaimed, so using them afterwards may do
3446 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3447 one (it's much easier than Tom thought :-).
3449 (5) Because of (4), we can now allow for the use of $prvscheck_result
3450 inside the third argument.
3452 PH/03 For some reason, the default setting of PATH when running a command from
3453 a pipe transport was just "/usr/bin". I have changed it to
3456 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3457 anything to be listed in the output from -bV.
3459 PH/05 When a filter generated an autoreply, the entire To: header line was
3460 quoted in the delivery log line, like this:
3462 => >A.N.Other <ano@some.domain> <original@ddress> ...
3464 This has been changed so that it extracts the operative address. There
3465 may be more than one such address. If so, they are comma-separated, like
3468 => >ano@some.domain,ona@other.domain <original@ddress> ...
3470 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3471 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3472 not being looked up in the rDNS to get a host name, Exim was showing the
3473 IP address twice in Received: lines, even though the IP addresses were
3474 identical. For example:
3476 Received: from [1.2.3.4] (helo=[1.2.3.4])
3478 However, if the real host name was known, it was omitting the HELO data
3479 if it matched the actual IP address. This has been tidied up so that it
3480 doesn't show the same IP address twice.
3482 PH/07 When both +timestamp and +memory debugging was on, the value given by
3483 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3484 called by the debug printing, thereby overwriting the timestamp buffer.
3485 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3487 PH/08 When the original message was included in an autoreply transport, it
3488 always said "this is a copy of the message, including all the headers",
3489 even if body_only or headers_only was set. It now gives an appropriate
3492 PH/09 Applied a patch from the Sieve maintainer which:
3494 o fixes some comments
3495 o adds the (disabled) notify extension core
3496 o adds some debug output for the result of if/elsif tests
3497 o points to the current vacation draft in the documentation
3498 and documents the missing references header update
3502 o fixes a bug in processing the envelope test (when testing
3503 multiple envelope elements, the last element determined the
3506 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3507 Electronic Mail") by including:
3509 Auto-submitted: auto-generated
3511 in the messages that it generates (bounce messages and others, such as
3512 warnings). In the case of bounce messages for non-SMTP messages, there was
3513 also a typo: it was using "Auto_submitted" (underscore instead of
3514 hyphen). Since every message generated by Exim is necessarily in response
3515 to another message, thes have all been changed to:
3517 Auto-Submitted: auto-replied
3519 in accordance with these statements in the RFC:
3521 The auto-replied keyword:
3523 - SHOULD be used on messages sent in direct response to another
3524 message by an automatic process,
3526 - MUST NOT be used on manually-generated messages,
3528 - MAY be used on Delivery Status Notifications (DSNs) and Message
3529 Disposition Notifications (MDNs),
3531 - MUST NOT be used on messages generated by automatic or periodic
3532 processes, except for messages which are automatic responses to
3535 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3536 to the default Received: header definition.
3538 PH/12 Added log selector acl_warn_skipped (default on).
3540 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3541 variables because (a) they are in the wrong storage pool and (b) even if
3542 they were copied, it wouldn't work properly because of the caching.
3544 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3545 checking when decoding. Apparently there are clients that generate
3546 overlong encoded strings. Why am I not surprised?
3548 PH/15 If the first argument of "${if match_address" was not empty, but did not
3549 contain an "@" character, Exim crashed. Now it writes a panic log message
3550 and treats the condition as false.
3552 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3554 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3555 extension "envelope-auth". The code is finished and in agreement with
3556 other implementations, but there is no documentation so far and in fact,
3557 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3558 not changing the active code.
3560 Print executed "if" and "elsif" statements when debugging is used. This
3561 helps a great deal to understand what a filter does.
3563 Document more things not specified clearly in RFC3028. I had all this
3564 sorted out, when out of a sudden new issues came to my mind. Oops."
3566 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3569 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3570 canonical form (as documented). However, after a host name lookup from
3571 the IP address, check_host() was doing a simple string comparison with
3572 addresses acquired from the DNS when checking that the found name did
3573 have the original IP as one of its addresses. Since any found IPv6
3574 addresses are likely to be in abbreviated form, the comparison could
3575 fail. Luckily, there already exists a function for doing the comparison
3576 by converting both addresses to binary, so now that is used instead of
3577 the text comparison.
3579 PH/20 There was another similar case to PH/19, when a complete host name was
3580 given in a host list; looking up its IP address could give an abbreviated
3581 form, whereas the current host's name might or might not be abbreviated.
3582 The same fix has been applied.
3588 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3589 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3592 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3593 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3595 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3597 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3598 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3599 Also, for "Auto-Submitted", treat anything other than "no" as
3600 non-personal, in accordance with RFC 3834. (Previously it treated
3601 anything starting "auto-" as non-personal.)
3603 TF/01 The control=submission/name=... option had a problem with syntax
3604 errors if the name included a slash character. The /name= option
3605 now slurps the rest of the string, so it can include any characters
3606 but it must come last in the list of options (after /sender_retain
3609 PH/05 Some modifications to the interface to the fake nameserver for the new
3617 TK/01 Added the "success_on_redirect" address verification option. See
3618 NewStuff for rationale and an example.
3620 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3622 PH/02 Patch to exigrep to allow it to work on syslog lines.
3624 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3625 fread() to skip over the body file's header line, because in Cygwin the
3626 header line is locked and is inaccessible.
3628 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3629 co-exist for some time) to make it clear that it is the Exim ID that is
3630 referenced, not the Message-ID: header line.
3632 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3633 string_format() function, because snprintf() does not exist on all
3636 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3637 ${stat: expansion item.
3639 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3640 protocol synchronization error", to keep the pedants happy.
3642 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3643 well as for IRIX systems, when gcc is being used. See the host.c source
3646 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3648 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3651 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3652 transport and to the smtp transport in LMTP mode.
3654 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3656 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3657 run a filter in a subprocess. This could lead to confusion in subsequent
3658 lookups in the parent process. There should also be a search_tidyup() at
3659 the end of the subprocess.
3661 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3662 only if the host matched helo_try_verify_hosts, which caused the
3663 verification to occur when the EHLO/HELO command was issued. The ACL just
3664 tested the remembered result. Now, if a previous verification attempt has
3665 not happened, "verify = helo" does it there and then.
3667 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3669 TK/03 Fix log output including CR from clamd.
3671 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3672 exist provoked a memory error which could cause a segfault.
3674 PH/15 Installed PCRE 6.2
3676 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3678 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3679 of the problem. Specifically, suggested +O2 rather than +O1 for the
3682 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3684 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3685 record, the port was not correctly logged when the outgoing_port log
3686 selector was set (it logged the transort's default port).
3688 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3689 fallback_hosts, and "hosts" in the smtp transport.
3691 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3692 host errors such as "Connection refused".
3694 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3695 authentication with radiusclient 0.4.9:
3697 - Error returned from rc_read_config was caught wrongly
3698 - Username/password not passed on to radius server due to wrong length.
3700 The presumption is that some radiusclient API changes for 4.51/PH/17
3701 were not taken care of correctly. The code is still untested by me (my
3702 Linux distribution still has 0.3.2 of radiusclient), but it was
3703 contributed by a Radius user.
3705 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3706 expanding the "port" option of the smtp transport.
3708 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3709 while reading a MIME header. Thanks to Tom Hughes for a patch.
3711 PH/24 Include config.h inside local_scan.h so that configuration settings are
3714 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3715 that in an ACL for RCPT (for example), you can examine exactly what was
3718 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3719 commands, but it was not correctly comparing the address with the actual
3720 client host address. Thus, it would show the EHLO address in Received:
3721 header lines when this was not necessary.
3723 PH/27 Added the % operator to ${eval:}.
3725 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3726 it should be ignoring failures (because with -C, for example, it has lost
3727 privilege). It wasn't ignoring creation failures other than "already
3730 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3733 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3734 because an input line was too long, either on its own, or by virtue of
3735 too many continuations, the temporary file was not being removed, and the
3736 return code was incorrect.
3738 PH/31 Missing "BOOL" in function definition in filtertest.c.
3740 PH/32 Applied Sieve patches from the maintainer.
3742 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3744 PH/33 Added "verify = not_blind".
3746 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3747 Local/Makefile (with some defaults set). These are used in built scripts
3748 such as exicyclog, but they have never been used in the exim_install
3749 script (though there are many overriding facilities there). I have
3750 arranged that the exim_install script now takes note of these two
3753 PH/35 Installed configuration files for Dragonfly.
3755 PH/36 When a locally submitted message by a trusted user did not contain a
3756 From: header, and the sender address was obtained from -f or from an SMTP
3757 MAIL command, and the trusted user did not use -F to supply a sender
3758 name, $originator_name was incorrectly used when constructing a From:
3759 header. Furthermore, $originator_name was used for submission mode
3760 messages from external hosts without From: headers in a similar way,
3761 which is clearly wrong.
3763 PH/37 Added control=suppress_local_fixups.
3765 PH/38 When log_selector = +received_sender was set, and the addition of the
3766 sender made the log line's construction buffer exactly full, or one byte
3767 less than full, an overflow happened when the terminating "\n" was
3770 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3771 when the result of a list match is failure because a DNS lookup failed.
3773 PH/40 RM_COMMAND is now used in the building process.
3775 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3776 the "build-* directories that it finds.
3778 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3779 address in a domain literal was a prefix of an interface address.
3781 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3782 when verifying a sender address, unless rewrite_headers is false.
3784 PH/44 Wrote a long comment about why errors_to addresses are verified as
3785 recipients, not senders.
3787 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3788 the ratelimit ACL was added.
3790 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3792 PH/46 Added extra information about PostgreSQL errors to the error string.
3794 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3795 avoiding the need to install special zones in a real server. This is
3796 backwards compatible; if it can't find the fake resolver, it drops back.
3797 Thus, both old and new test suites can be run.
3799 TF/02 Added util/ratelimit.pl
3801 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3802 clock is set back in time.
3804 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3805 Candler <B.Candler@pobox.com>.
3807 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3808 OK for addresses that are the result of redirections.
3810 PH/48 A number of further additions for the benefit of the new test suite,
3811 including a fake gethostbyname() that interfaces to the fake DNS resolver
3814 TF/06 The fix for widen_domains has also been applied to qualify_single and
3815 search_parents which are the other dnslookup options that can cause
3818 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3821 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3823 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3824 service" if it encountered an MX record with an empty target hostname.
3825 The message is now "an MX or SRV record indicated no SMTP service".
3827 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3828 if the DNS of the sending site is misconfigured. This is quite a
3829 common situation. This change restores the behaviour of treating a
3830 helo verification defer as a failure.
3832 PH/51 If self=fail was set on a router, the bounce message did not include the
3833 actual error message.
3839 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3841 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3842 command itself ended in error, the underlying message about the transport
3843 filter timeout was being overwritten with the pipe command error. Now the
3844 underlying error message should be appended to the second error message.
3846 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3848 PH/02 Reduce the amount of output that "make" produces by default. Full output
3849 can still be requested.
3851 PH/03 The warning log line about a condition test deferring for a "warn" verb
3852 was being output only once per connection, rather than after each
3853 occurrence (because it was using the same function as for successful
3854 "warn" verbs). This seems wrong, so I have changed it.
3856 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3857 it should not have, which might have caused a crash in the right
3858 circumstances, but probably never did.
3860 PH/04 Installed a modified version of Tony Finch's patch to make submission
3861 mode fix the return path as well as the Sender: header line, and to
3862 add a /name= option so that you can make the user's friendly name appear
3865 TF/03 Added the control = fakedefer ACL modifier.
3867 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3868 Mark Lowes for thorough testing.
3870 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3872 TK/03 Merged latest SRS patch from Miles Wilton.
3874 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3875 with the definition in sysexits.h (which is #included earlier).
3876 Fortunately, Exim does not actually use EX_OK. The code used to try to
3877 preserve the sysexits.h value, by assuming that macro definitions were
3878 scanned for macro replacements. I have been disabused of this notion,
3879 so now the code just undefines EX_OK before #including unistd.h.
3881 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3882 in the smtp transport. When a block could not be written in a single
3883 write() function, the timeout was being re-applied to each part-write.
3884 This seems wrong - if the receiver was accepting one byte at a time it
3885 would take for ever. The timeout is now adjusted when this happens. It
3886 doesn't have to be particularly precise.
3888 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3889 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3891 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3892 without a domain if the check to <postmaster@domain> fails.
3894 SC/01 Eximstats: added -xls and the ability to specify output files
3895 (patch written by Frank Heydlauf).
3897 SC/02 Eximstats: use FileHandles for outputting results.
3899 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3901 SC/04 Eximstats: fixed display of large numbers with -nvr option
3903 SC/05 Eximstats: fixed merging of reports with empty tables.
3905 SC/06 Eximstats: added the -include_original_destination flag
3907 SC/07 Eximstats: removed tabs and trailing whitespace.
3909 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3911 TK/06 MBOX spool code: Add real "From " MBOX separator line
3912 so the .eml file is really in mbox format (even though
3913 most programs do not really care). Patch from Alex Miller.
3915 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3916 The latter is generated from $received_to and is only set if the
3917 message has one envelope recipient. SA can use these headers,
3918 obviously out-of-the-box. Patch from Alex Miller.
3920 PH/08 The ${def test on a variable was returning false if the variable's
3921 value was "0", contrary to what the specification has always said!
3922 The result should be true unless the variable is empty.
3924 PH/09 The syntax error of a character other than { following "${if
3925 def:variable_name" (after optional whitespace) was not being diagnosed.
3926 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3927 accidental colon was present, for example, could give incorrect results.
3929 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3930 result is used (not including appendfile, where other changes are about
3933 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3934 This involved changing a lot of size variables from int to off_t. It
3935 should work with maildirs and everything.
3937 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3938 spamd dying while we are connected to it.
3940 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3943 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3944 function for BDB 4.3.
3946 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3948 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3949 the sources for PCRE have been re-arranged and more files are now
3952 PH/15 The code I had for printing potentially long long variables in PH/11
3953 above was not the best (it lost precision). The length of off_t variables
3954 is now inspected at build time, and an appropriate printing format (%ld
3955 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3956 to be "long long int" or "long int". This is needed for the internal
3957 formatting function string_vformat().
3959 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3960 the configuration file to be ":syslog", then the script "guesses" where
3961 the logs files are, rather than using the compiled in default. In our
3962 case the guess is not the same as the compiled default, so the script
3963 suddenly stopped working when I started to use syslog. The patch checks
3964 to see if log_file_path is "". If so, it attempts to read it from exim
3965 with no configuration file to get the compiled in version, before it
3966 falls back to the previous guessing code."
3968 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3969 implementing BATV in an Exim configuration. See NewStuff for the gory
3972 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3973 Makefile that are specific to HP-UX.
3975 PH/18 If the "use_postmaster" option was set for a recipient callout together
3976 with the "random" option, the postmaster address was used as the MAIL
3977 FROM address for the random test, but not for the subsequent recipient
3978 test. It is now used for both.
3980 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3981 patch removes a few documentation additions to RFC 3028, because the
3982 latest draft now contains them. It adds the new en;ascii-case comparator
3983 and a new error check for 8bit text in MIME parts. Comparator and
3984 require names are now matched exactly. I enabled the subaddress
3985 extension, but it is not well tested yet (read: it works for me)."
3987 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3988 rework some of the code of TK/09 above to avoid the hardwired use of
3989 "%lld" and "long long". Replaced the call to snprintf() with a call to
3992 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3993 records point to non-existent hosts", "retry timeout exceeded", and
3994 "retry time not reached for any host after a long failure period".
3996 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3997 experimental DomainKeys support:
3999 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4000 (2) On an error such as an illegally used "control", the wrong name for
4001 the control was given.
4003 These problems did NOT occur unless DomainKeys support was compiled.
4005 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4007 PH/24 Added ${if match_ip condition.
4009 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4010 databases so that it will be absolutely obvious if a crash occurs in the
4011 DB library. This is a regular occurrence (often caused by mis-matched
4014 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4015 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4016 picked up on a user's system that detects such things. There doesn't seem
4017 to be a gcc warning option for this - only an attribute that has to be
4018 put on the function's prototype. It seems that in Fedora Core 4 they have
4019 set this on a number of new functions. No doubt there will be more in due
4022 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4023 specify a transport. However, if an address that was verified by such a
4024 router was the subject of a callout, Exim crashed because it tried to
4025 read the rcpt_include_affixes from the non-existent transport. Now it
4026 just assumes that the setting of that option is false. This bug was
4027 introduced by 4.51/PH/31.
4029 PH/28 Changed -d+all to exclude +memory, because that information is very
4030 rarely of interest, but it makes the output a lot bigger. People tend to
4031 do -d+all out of habit.
4033 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4034 code in os-type was giving problems when libc.so lives in lib64, like on
4037 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4038 aren't the modern standard, and it seems that some systems' include files
4039 don't always have them. Exim was already checking for some of the newer
4040 ones like T_AAAA, and defining it itself. I've added checks for all the
4041 record types that Exim uses.
4043 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4044 not automatically generating a new one, as it is supposed to. This
4045 prevented TLS from working. If the file did exist, but contained invalid
4046 data, a new version was generated, as expected. It was only the case of a
4047 non-existent file that was broken.
4049 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4050 with a change in libdomainkeys > 0.64.
4052 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4053 from DNS. If the selector record carries the flag, it now has
4054 precedence over the domain-wide flag.
4056 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4058 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4059 the use of a port name that isn't defined in /etc/services) occurred, the
4060 message was deferred as in a normal delivery, and thus remained on the
4061 spool, instead of being failed because of the mua_wrapper setting. This
4062 is now fixed, and I tidied up some of the mua_wrapper messages at the
4065 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4066 the messages in a hash of arrays rather than using individual hashes.
4067 This is a bit cleaner and results in dramatic memory savings, albeit
4068 at a slight CPU cost.
4070 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4071 as requested by Marc Sherman.
4073 SC/10 Eximstats: added histograms for user specified patterns as requested
4076 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4078 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4079 fopen() in the content-scanning modules that did not already have it.
4085 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4086 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4088 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4090 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4092 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4093 to test Sieve filters that use "vacation".
4095 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4096 that changes the way the GnuTLS parameters are stored in the cache file.
4097 The new format can be generated externally. For backward compatibility,
4098 if the data in the cache doesn't make sense, Exim assumes it has read an
4099 old-format file, and it generates new data and writes a new file. This
4100 means that you can't go back to an older release without removing the
4103 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4104 work if there are any delivery delays because "one_time" forces the
4105 parent to be marked "delivered", so its unseen clone is never tried
4106 again. For this reason, Exim now forbids the simultaneous setting of
4109 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4110 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4111 entry if you want to know the details. The fix, however, neglected to
4112 consider the case when local delivery batching is involved. The test for
4113 "previously delivered" was not happening when checking to see if an
4114 address could be batched with a previous (undelivered) one; under
4115 certain circumstances this could lead to multiple deliveries to the same
4118 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4119 in its include files, and this causes problems building Exim.
4121 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4122 header_syntax) but Exim was just ignoring anything given after a slash.
4123 In particular, this caused confusion with an attempt to use "verify =
4124 reverse_host_lookup/defer_ok". An error is now given when options are
4125 supplied for verify items that do not have them. (Maybe reverse_host_
4126 lookup should have a defer_ok option, but that's a different point.)
4128 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4129 defined by RFC 821) to 2048, because there were problems with some AUTH
4130 commands, and RFC 1869 says the size should be increased for extended
4131 SMTP commands that take arguments.
4133 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4136 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4137 "unknown" error; now it says that the functionality isn't in the binary.
4139 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4140 an address' error message when a string expansion fails (syntax or
4141 whatever). Otherwise the password may appear in the log. Following change
4142 PH/42 below, there is no longer a chance of it appearing in a bounce
4145 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4147 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4148 of its addresses was ever tried. (Bugzilla bug #2.)
4150 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4151 the result incorrectly in the debug output. (It correctly added a newline
4152 to what was transported.)
4154 TF/01 Added $received_time.
4156 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4157 commented out examples of how to interface to a virus scanner and to
4158 SpamAssassin. Also added commented examples of av_scanner and
4159 spamd_address settings.
4161 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4162 and controls are allowed in which ACLs. There were a couple of minor
4163 errors. Some of the entries in the conditions table (which is a table of
4164 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4165 negation of where the condition IS allowed.
4167 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4169 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4170 header file does not have a version number, so I've had to invent a new
4171 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4172 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4173 radiusclient), but it was contributed by a Radius user.
4175 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4176 files or directories, for OpenSSL.
4178 PH/19 When an Exim process that is running as root has to create an Exim log
4179 file, it does so in a subprocess that runs as exim:exim so as to get the
4180 ownership right at creation (otherwise, other Exim processes might see
4181 the file with the wrong ownership). There was no test for failure of this
4182 fork() call, which would lead to the process getting stuck as it waited
4183 for a non-existent subprocess. Forks do occasionally fail when resources
4184 run out. I reviewed all the other calls to fork(); they all seem to check
4187 PH/20 When checking for unexpected SMTP input at connect time (before writing
4188 the banner), Exim was not dealing correctly with a non-positive return
4189 from the read() function. If the client had disconnected by this time,
4190 the result was a log entry for a synchronization error with an empty
4191 string after "input=" when read() returned zero. If read() returned -1
4192 (an event I could not check), uninitialized data bytes were printed.
4193 There were reports of junk text (parts of files, etc) appearing after
4196 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4198 PH/22 Added support for macro redefinition, and (re)definition in between
4199 driver and ACL definitions.
4201 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4202 forgetting to use the resulting value; it was using the unexpanded value.
4204 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4205 hadn't been configured. The fix is from Juergen Kreileder, who
4206 understands it better than I do:
4208 "Here's what I see happening with three configured cyrus_sasl
4209 authenticators configured (plain, login, cram-md5):
4211 On startup auth_cyrus_sasl_init() gets called for each of these.
4212 This means three calls to sasl_listmech() without a specified mech_list.
4213 => SASL tests which mechs of all available mechs actually work
4214 => three warnings about OTP not working
4215 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4217 With the patch, sasl_listmech() also gets called three times. But now
4218 SASL's mech_list option is set to the server_mech specified in the the
4219 authenticator. Or in other words, the answer from sasl_listmech()
4220 gets limited to just the mech you're testing for (which is different
4222 => the return list contains just 'plain' or 'login', 'cram-md5' or
4223 nothing depending on the value of ob->server_mech.
4225 I've just tested the patch: Authentication still works fine,
4226 unavailable mechs specified in the exim configuration are still
4227 caught, and the auth.log warnings about OTP are gone."
4229 PH/25 When debugging is enabled, the contents of the command line are added
4230 to the debugging output, even when log_selector=+arguments is not
4233 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4234 answer is "GNU", and only if the return is "GNU/something" is the answer
4237 PH/27 $acl_verify_message is now set immediately after the failure of a
4238 verification in an ACL, and so is available in subsequent modifiers. In
4239 particular, the message can be preserved by coding like this:
4241 warn !verify = sender
4242 set acl_m0 = $acl_verify_message
4244 Previously, $acl_verify_message was set only while expanding "message"
4245 and "log_message" when a very denied access.
4247 PH/28 Modified OS/os.c-Linux with
4249 -#ifndef OS_LOAD_AVERAGE
4250 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4252 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4253 nomenclature these days.)
4255 PH/29 Installed patch from the Sieve maintainer that adds the options
4256 sieve_useraddress and sieve_subaddress to the redirect router.
4258 PH/30 In these circumstances:
4259 . Two addresses routed to the same list of hosts;
4260 . First host does not offer TLS;
4261 . First host accepts first address;
4262 . First host gives temporary error to second address;
4263 . Second host offers TLS and a TLS session is established;
4264 . Second host accepts second address.
4265 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4266 and peerdn, if requested) that were in fact used only for the second
4269 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4270 attention to any local part prefix or suffix that was matched by the
4271 router that accepted the address. It now behaves in the same way as it
4272 does for delivery: the affixes are removed from the local part unless
4273 rcpt_include_affixes is set on the transport.
4275 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4276 timeout during the DATA phase of an incoming message.
4278 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4279 applied a patch sanctioned by the Sieve maintainer.
4281 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4282 the uid or gid is negative. A case of a negative gid caused this to be
4283 noticed. The fix allows for either to be negative.
4285 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4286 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4289 PH/36 The change PH/12 above was broken. Fixed it.
4291 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4292 the grounds that routing the same address twice would always produce the
4293 same answer. This might have been true once, but it is certainly no
4294 longer true now. Routing a child address may depend on the previous
4295 routing that produced that child. Some complicated redirection strategies
4296 went wrong when messages had multiple recipients, and made Exim's
4297 behaviour dependent on the order in which the addresses were given.
4299 I have moved the duplicate checking until after the routing is complete.
4300 Exim scans the addresses that are assigned to local and remote
4301 transports, and removes any duplicates. This means that more work will be
4302 done, as duplicates will always all be routed, but duplicates are
4303 presumably rare, so I don't expect this is of any significance.
4305 For deliveries to pipes, files, and autoreplies, the duplicate checking
4306 still happens during the routing process, since they are not going to be
4309 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4310 It corrects a timeout issue with spamd. This is Ian's comment: "The
4311 background is that sometimes spamd either never reads data from a
4312 connection it has accepted, or it never writes response data. The exiscan
4313 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4314 blindly assumes that writes won't block so it may never time out."
4316 PH/39 Allow G after quota size as well as K and M.
4318 PH/40 The value set for $authenticated_id in an authenticator may not contain
4319 binary zeroes or newlines because the value is written to log lines and
4320 to spool files. There was no check on this. Now the value is run through
4321 the string_printing() function so that such characters are converted to
4322 printable escape sequences.
4324 PH/41 $message_linecount is a new variable that contains the total number of
4325 lines in the message. Compare $body_linecount, which is the count for the
4328 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4329 bounce and delay warning messages, except in certain special cases, which
4332 (a) An SMTP error message from a remote host;
4333 (b) A message specified in a :fail: redirection;
4334 (c) A message specified in a "fail" command in a system filter;
4335 (d) A message specified in a FAIL return from the queryprogram router;
4336 (e) A message specified by the cannot_route_message router option.
4338 In these cases only, Exim does include the error details in bounce and
4339 warning messages. There are also a few cases where bland messages such
4340 as "unrouteable address" or "local delivery error" are given.
4342 PH/43 $value is now also set for the "else" part of a ${run expansion.
4344 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4345 being worked on, but at least Exim now implements the latest version to
4348 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4349 process to complete was treated as a delivery failure, a timeout while
4350 writing the message to the pipe was logged, but erroneously treated as a
4351 successful delivery. Such timeouts include transport filter timeouts. For
4352 consistency with the overall process timeout, these timeouts are now
4353 treated as errors, giving rise to delivery failures by default. However,
4354 there is now a new Boolean option for the pipe transport called
4355 timeout_defer, which, if set TRUE, converts the failures into defers for
4356 both kinds of timeout. A transport filter timeout is now identified in
4359 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4360 systems where "make" and "gmake" are different, calling "gmake" at top
4361 level broke things. I've arranged for the value of $(MAKE) to be passed
4362 from the Makefile to this script so that it can call the same version of
4366 A note about Exim versions 4.44 and 4.50
4367 ----------------------------------------
4369 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4370 changes of various kinds. As a consequence, a big documentation update was
4371 needed. This delayed the release for rather longer than seemed good, especially
4372 in the light of a couple of (minor) security issues. Therefore, the changes
4373 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4374 release. So 4.44 and 4.50 are in effect two different branches that both start
4377 I have left the 4.50 change log unchanged; it contains all the changes since
4378 4.43. The change log for 4.44 is below; many of its items are identical to
4379 those for 4.50. This seems to be the most sensible way to preserve the
4380 historical information.
4386 1. Minor wording change to the doc/README.SIEVE file.
4388 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4389 computation of the current number of files was incorrect.
4391 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4392 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4395 4. Give more explanation in the error message when the command for a transport
4396 filter fails to execute.
4398 5. There are several places where Exim runs a non-Exim command in a
4399 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4400 was being done only for the command run by the queryprogram router. It is
4401 now done for all such subprocesses. The other cases are: ${run, transport
4402 filters, and the commands run by the lmtp and pipe transports.
4404 6. Added CONFIGURE_GROUP build-time option.
4406 7. Some older OS have a limit of 256 on the maximum number of file
4407 descriptors. Exim was using setrlimit() to set 1000 as a large value
4408 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4409 systems. I've change it so that if it can't get 1000, it tries for 256.
4411 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4412 was an oversight, and furthermore, ever since the addition of extra
4413 controls (e.g. 4.43/32), the checks on when to allow different forms of
4414 "control" were broken. There should now be diagnostics for all cases when a
4415 control that does not make sense is encountered.
4417 9. Added the /retain_sender option to "control=submission".
4419 10. $recipients is now available in the predata ACL (oversight).
4421 11. Tidy the search cache before the fork to do a delivery from a message
4422 received from the command line. Otherwise the child will trigger a lookup
4423 failure and thereby defer the delivery if it tries to use (for example) a
4424 cached ldap connection that the parent has called unbind on.
4426 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4427 of $address_data from the recipient verification was clobbered by the
4428 sender verification.
4430 13. The value of address_data from a sender verification is now available in
4431 $sender_address_data in subsequent conditions in the ACL statement.
4433 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4435 15. Added a new option "connect=<time>" to callout options, to set a different
4438 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4439 was its contents. (It was OK if the option was not defined at all.)
4441 17. A "Completed" log line is now written for messages that are removed from
4442 the spool by the -Mrm option.
4444 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4445 information about exactly what failed.
4447 19. Added -dd to debug only the daemon process.
4449 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4450 handles timeouts, both on the server side and network timeouts. Renamed the
4451 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4453 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4454 It is now set to "smtps".
4456 22. $host_address is now set to the target address during the checking of
4457 ignore_target_hosts.
4459 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4460 being passed; this would have caused $sender_host_name to have been used if
4461 matching the list had actually called for a host name (not very likely,
4462 since this list is usually IP addresses). A host name is now passed as
4465 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4466 code that specifies a non-restarting handler (typically sigaction() in
4467 modern systems) in an attempt to fix a rare and obscure crash bug.
4469 25. Narrowed the window for a race in the daemon that could cause it to ignore
4470 SIGCHLD signals. This is not a major problem, because they are used only to
4471 wake it up if nothing else does.
4473 26. A malformed maildirsize file could cause Exim to calculate negative values
4474 for the mailbox size or file count. Odd effects could occur as a result.
4475 The maildirsize information is now recalculated if the size or filecount
4478 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4479 support for a long time. Removed HAVE_SYS_VFS_H.
4481 28. Installed the latest version of exipick from John Jetmore.
4483 29. In an address list, if the pattern was not a regular expression, an empty
4484 subject address (from a bounce message) matched only if the pattern was an
4485 empty string. Non-empty patterns were not even tested. This was the wrong
4486 because it is perfectly reasonable to use an empty address as part of a
4487 database query. An empty address is now tested by patterns that are
4488 lookups. However, all the other forms of pattern expect the subject to
4489 contain a local part and a domain, and therefore, for them, an empty
4490 address still always fails if the pattern is not itself empty.
4492 30. Exim went into a mad DNS loop when attempting to do a callout where the
4493 host was specified on an smtp transport, and looking it up yielded more
4494 than one IP address.
4496 31. Re-factored the code for checking spool and log partition space into a
4497 function that finds that data and another that does the check. The former
4498 is then used to implement four new variables: $spool_space, $log_space,
4499 $spool_inodes, and $log_inodes.
4501 32. The RFC2047 encoding function was originally intended for short strings
4502 such as real names; it was not keeping to the 75-character limit for
4503 encoded words that the RFC imposes. It now respects the limit, and
4504 generates multiple encoded words if necessary. To be on the safe side, I
4505 have increased the buffer size for the ${rfc2047: expansion operator from
4508 33. It is now permitted to omit both strings after an "if" condition; if the
4509 condition is true, the result is "true". As before, when the second string
4510 is omitted, a false condition yields an empty string. This makes it less
4511 cumbersome to write custom ACL and router conditions.
4513 34. Failure to deliver a bounce message always caused it to be frozen, even if
4514 there was an errors_to setting on the router. The errors_to setting is now
4517 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4518 canonical form (fully expanded) before being placed in
4519 $sender_host_address.
4521 36. The table in the code that translates DNS record types into text (T_A to
4522 "A" for instance) was missing entries for NS and CNAME. It is just possible
4523 that this could have caused confusion if both these types were looked up
4524 for the same domain, because the text type is used as part of Exim's
4525 per-process caching. But the chance of anyone hitting this buglet seems
4528 37. The dnsdb lookup has been extended in a number of ways.
4530 (1) There is a new type, "zns", which walks up the domain tree until it
4531 finds some nameserver records. It should be used with care.
4533 (2) There is a new type, "mxh", which is like "mx" except that it returns
4534 just the host names, not the priorities.
4536 (3) It is now possible to give a list of domains (or IP addresses) to be
4537 looked up. The behaviour when one of the lookups defers can be
4538 controlled by a keyword.
4540 (4) It is now possible to specify the separator character for use when
4541 multiple records are returned.
4543 38. The dnslists ACL condition has been extended: it is now possible to supply
4544 a list of IP addresses and/or domains to be looked up in a particular DNS
4547 39. Added log_selector=+queue_time_overall.
4549 40. When running the queue in the test harness, wait just a tad after forking a
4550 delivery process, to get repeatability of debugging output.
4552 41. Include certificate and key file names in error message when GnuTLS fails
4553 to set them up, because the GnuTLS error message doesn't include the name
4554 of the failing file when there is a problem reading it.
4556 42. Allow both -bf and -bF in the same test run.
4558 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4560 44. The "Exiscan patch" is now merged into the mainline Exim source.
4562 45. Sometimes the final signoff response after QUIT could fail to get
4563 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4564 < 0 before doing a fflush(). This bug looks as though it goes back to the
4565 introduction of TLS in release 3.20, but "sometimes" must have been rare
4566 because the tests only now provoked it.
4568 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4569 (this can affect the format of dates).
4571 47. exim_tidydb, when checking for the continued existence of a message for
4572 which it has found a message-specific retry record, was not finding
4573 messages that were in split spool directories. Consequently, it was
4574 deleting retry records that should have stayed in existence.
4576 48. Steve fixed some bugs in eximstats.
4578 49. The SPA authentication driver was not abandoning authentication and moving
4579 on to the next authenticator when an expansion was forced to fail,
4580 contradicting the general specification for all authenticators. Instead it
4581 was generating a temporary error. It now behaves as specified.
4583 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4584 (the order specifies the preference for clients). The order is now AES256,
4585 AES128, 3DES, ARCFOUR128.
4587 51. Small patch to Sieve code - explicitly set From: when generating an
4590 52. Exim crashed if a remote delivery caused a very long error message to be
4591 recorded - for instance if somebody sent an entire SpamAssassin report back
4592 as a large number of 550 error lines. This bug was coincidentally fixed by
4593 increasing the size of one of Exim's internal buffers (big_buffer) that
4594 happened as part of the Exiscan merge. However, to be on the safe side, I
4595 have made the code more robust (and fixed the comments that describe what
4598 53. Now that there can be additional text after "Completed" in log lines (if
4599 the queue_time_overall log selector is set), a one-byte patch to exigrep
4600 was needed to allow it to recognize "Completed" as not the last thing in
4603 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4604 patch that reportedly fixes this has been added. I am not expert enough to
4605 create a test for it. This is what the patch creator wrote:
4607 "I found a little strange behaviour of ldap code when working with
4608 Windows 2003 AD Domain, where users was placed in more than one
4609 Organization Units. When I tried to give exim partial DN, the exit code
4610 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4611 But simultaneously result of request was absolutely normal ldap result,
4612 so I produce this patch..."
4614 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4615 REFERENCE, so I have modified the code to exclude the patch when that macro
4618 55. Some experimental protocols are using DNS PTR records for new purposes. The
4619 keys for these records are domain names, not reversed IP addresses. The
4620 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4621 leaves it alone. Component reversal etc. now happens only for IP addresses.
4624 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4626 57. Double the size of the debug message buffer (to 2048) so that more of very
4627 long debug lines gets shown.
4629 58. The exicyclog utility now does better if the number of log files to keep
4630 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4632 59. Two changes related to the smtp_active_hostname option:
4634 (1) $smtp_active_hostname is now available as a variable.
4635 (2) The default for smtp_banner uses $smtp_active_hostname instead
4636 of $primary_hostname.
4638 60. The host_aton() function is supposed to be passed a string that is known
4639 to be a valid IP address. However, in the case of IPv6 addresses, it was
4640 not checking this. This is a hostage to fortune. Exim now panics and dies
4641 if the condition is not met. A case was found where this could be provoked
4642 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4643 components; fortuitously, this particular loophole had already been fixed
4644 by change 4.50/55 above.
4646 If there are any other similar loopholes, the new check in host_aton()
4647 itself should stop them being exploited. The report I received stated that
4648 data on the command line could provoke the exploit when Exim was running as
4649 exim, but did not say which command line option was involved. All I could
4650 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4651 running as the user.
4654 61. There was a buffer overflow vulnerability in the SPA authentication code
4655 (which came originally from the Samba project). I have added a test to the
4656 spa_base64_to_bits() function which I hope fixes it.
4659 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4660 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4662 63. The daemon start-up calls getloadavg() while still root for those OS that
4663 need the first call to be done as root, but it missed one case: when
4664 deliver_queue_load_max is set with deliver_drop_privilege. This is
4665 necessary for the benefit of the queue runner, because there is no re-exec
4666 when deliver_drop_privilege is set.
4668 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4669 This has been fixed.
4671 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4672 was in use, was not putting the data itself into the right store pool;
4673 consequently, it could be overwritten for a subsequent message in the same
4674 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4677 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4679 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4680 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4681 the function were treating the return as a boolean value, which happened to
4682 work because 0=false and not-0=true, but is not correct code.
4684 68. The host_aton() function was not handling scoped IPv6 addresses (those
4685 with, for example, "%eth0" on the end) correctly.
4687 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4688 negated items (that is, ~something) in unsigned ints. Some compilers
4689 apparently mutter when there is no cast.
4691 70. If an address verification called from an ACL failed, and did not produce a
4692 user-specific message (i.e. there was only a "system" message), nothing was
4693 put in $acl_verify_message. In this situation, it now puts the system
4696 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4697 session; change 4.31/43 added the unwanted input to the log line - except
4698 that it did not do this in the start of session case. It now does.
4700 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4701 This is wrong and can cause the other end to generate a synchronization
4702 error if it is another Exim or anything else that does the synchronization
4703 check. A QUIT command is no longer sent after a timeout.
4705 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4706 during host lookups.
4708 74. The defer_ok option of callout verification was not working if it was used
4709 when verifying addresses in header lines, that is, for this case:
4711 verify = header_sender/callout=defer_ok
4713 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4714 those file descriptors could be used for SMTP connections. If anything
4715 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4716 could be sent to the SMTP client, causing chaos. The daemon now opens
4717 stdin, stdout, and stderr to /dev/null when it puts itself into the
4720 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4721 log by default. The user can override this with suitable Perl magic.
4723 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4724 the log message when discard triggers, was not working for the DATA ACL or
4725 for the non-SMTP ACL.
4727 78. Error message wording change in sieve.c.
4729 79. If smtp_accept_max_per_host was set, the number of connections could be
4730 restricted to fewer than expected, because the daemon was trying to set up
4731 a new connection before checking whether the processes handling previous
4732 connections had finished. The check for completed processes is now done
4733 earlier. On busy systems, this bug wouldn't be noticed because something
4734 else would have woken the daemon, and it would have reaped the completed
4737 80. If a message was submitted locally by a user whose login name contained one
4738 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4739 It caused a spool format error. I have fixed the spool reading code. A
4740 related problem was that the "from" clause in the Received: line became
4741 illegal because of the space(s). It is now covered by ${quote_local_part.
4743 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4746 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4748 83. Patch from Sieve maintainer to fix unterminated string problem in
4749 "vacation" handling.
4751 84. Some minor changes to the Linux configuration files to help with other
4752 OS variants using glibc.
4754 85. One more patch for Sieve to update vacation handling to latest spec.
4757 ----------------------------------------------------
4758 See the note above about the 4.44 and 4.50 releases.
4759 ----------------------------------------------------
4765 1. Change 4.43/35 introduced a bug that caused file counts to be
4766 incorrectly computed when quota_filecount was set in an appendfile
4769 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4770 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4773 3. Give more explanation in the error message when the command for a transport
4774 filter fails to execute.
4776 4. There are several places where Exim runs a non-Exim command in a
4777 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4778 was being done only for the command run by the queryprogram router. It is
4779 now done for all such subprocesses. The other cases are: ${run, transport
4780 filters, and the commands run by the lmtp and pipe transports.
4782 5. Some older OS have a limit of 256 on the maximum number of file
4783 descriptors. Exim was using setrlimit() to set 1000 as a large value
4784 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4785 systems. I've change it so that if it can't get 1000, it tries for 256.
4787 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4788 was an oversight, and furthermore, ever since the addition of extra
4789 controls (e.g. 4.43/32), the checks on when to allow different forms of
4790 "control" were broken. There should now be diagnostics for all cases when a
4791 control that does not make sense is encountered.
4793 7. $recipients is now available in the predata ACL (oversight).
4795 8. Tidy the search cache before the fork to do a delivery from a message
4796 received from the command line. Otherwise the child will trigger a lookup
4797 failure and thereby defer the delivery if it tries to use (for example) a
4798 cached ldap connection that the parent has called unbind on.
4800 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4801 of $address_data from the recipient verification was clobbered by the
4802 sender verification.
4804 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4805 was its contents. (It was OK if the option was not defined at all.)
4807 11. A "Completed" log line is now written for messages that are removed from
4808 the spool by the -Mrm option.
4810 12. $host_address is now set to the target address during the checking of
4811 ignore_target_hosts.
4813 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4814 being passed; this would have caused $sender_host_name to have been used if
4815 matching the list had actually called for a host name (not very likely,
4816 since this list is usually IP addresses). A host name is now passed as
4819 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4820 code that specifies a non-restarting handler (typically sigaction() in
4821 modern systems) in an attempt to fix a rare and obscure crash bug.
4823 15. Narrowed the window for a race in the daemon that could cause it to ignore
4824 SIGCHLD signals. This is not a major problem, because they are used only to
4825 wake it up if nothing else does.
4827 16. A malformed maildirsize file could cause Exim to calculate negative values
4828 for the mailbox size or file count. Odd effects could occur as a result.
4829 The maildirsize information is now recalculated if the size or filecount
4832 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4833 support for a long time. Removed HAVE_SYS_VFS_H.
4835 18. Updated exipick to current release from John Jetmore.
4837 19. Allow an empty sender to be matched against a lookup in an address list.
4838 Previously the only cases considered were a regular expression, or an
4841 20. Exim went into a mad DNS lookup loop when doing a callout where the
4842 host was specified on the transport, if the DNS lookup yielded more than
4845 21. The RFC2047 encoding function was originally intended for short strings
4846 such as real names; it was not keeping to the 75-character limit for
4847 encoded words that the RFC imposes. It now respects the limit, and
4848 generates multiple encoded words if necessary. To be on the safe side, I
4849 have increased the buffer size for the ${rfc2047: expansion operator from
4852 22. Failure to deliver a bounce message always caused it to be frozen, even if
4853 there was an errors_to setting on the router. The errors_to setting is now
4856 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4857 canonical form (fully expanded) before being placed in
4858 $sender_host_address.
4860 24. Updated eximstats to version 1.33
4862 25. Include certificate and key file names in error message when GnuTLS fails
4863 to set them up, because the GnuTLS error message doesn't include the name
4864 of the failing file when there is a problem reading it.
4866 26. Expand error message when OpenSSL has problems setting up cert/key files.
4869 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4870 (this can affect the format of dates).
4872 28. exim_tidydb, when checking for the continued existence of a message for
4873 which it has found a message-specific retry record, was not finding
4874 messages that were in split spool directories. Consequently, it was
4875 deleting retry records that should have stayed in existence.
4877 29. eximstats updated to version 1.35
4878 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4879 1.35 - bugfix such that pie charts by volume are generated correctly
4881 30. The SPA authentication driver was not abandoning authentication and moving
4882 on to the next authenticator when an expansion was forced to fail,
4883 contradicting the general specification for all authenticators. Instead it
4884 was generating a temporary error. It now behaves as specified.
4886 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4887 (the order specifies the preference for clients). The order is now AES256,
4888 AES128, 3DES, ARCFOUR128.
4890 31. Small patch to Sieve code - explicitly set From: when generating an
4893 32. Exim crashed if a remote delivery caused a very long error message to be
4894 recorded - for instance if somebody sent an entire SpamAssassin report back
4895 as a large number of 550 error lines. This bug was coincidentally fixed by
4896 increasing the size of one of Exim's internal buffers (big_buffer) that
4897 happened as part of the Exiscan merge. However, to be on the safe side, I
4898 have made the code more robust (and fixed the comments that describe what
4901 33. Some experimental protocols are using DNS PTR records for new purposes. The
4902 keys for these records are domain names, not reversed IP addresses. The
4903 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4904 leaves it alone. Component reversal etc. now happens only for IP addresses.
4907 34. The host_aton() function is supposed to be passed a string that is known
4908 to be a valid IP address. However, in the case of IPv6 addresses, it was
4909 not checking this. This is a hostage to fortune. Exim now panics and dies
4910 if the condition is not met. A case was found where this could be provoked
4911 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4912 components; fortuitously, this particular loophole had already been fixed
4913 by change 4.50/55 or 4.44/33 above.
4915 If there are any other similar loopholes, the new check in host_aton()
4916 itself should stop them being exploited. The report I received stated that
4917 data on the command line could provoke the exploit when Exim was running as
4918 exim, but did not say which command line option was involved. All I could
4919 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4920 running as the user.
4923 35. There was a buffer overflow vulnerability in the SPA authentication code
4924 (which came originally from the Samba project). I have added a test to the
4925 spa_base64_to_bits() function which I hope fixes it.
4928 36. The daemon start-up calls getloadavg() while still root for those OS that
4929 need the first call to be done as root, but it missed one case: when
4930 deliver_queue_load_max is set with deliver_drop_privilege. This is
4931 necessary for the benefit of the queue runner, because there is no re-exec
4932 when deliver_drop_privilege is set.
4934 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4935 was in use, was not putting the data itself into the right store pool;
4936 consequently, it could be overwritten for a subsequent message in the same
4937 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4940 38. Sometimes the final signoff response after QUIT could fail to get
4941 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4942 < 0 before doing a fflush(). This bug looks as though it goes back to the
4943 introduction of TLS in release 3.20, but "sometimes" must have been rare
4944 because the tests only now provoked it.
4950 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4951 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4952 Now it returns an integer. A number of calls were still expecting a T/F
4953 return. Fortuitously, in all cases, the tests worked in OK situations,
4954 which is the norm. However, things would have gone wrong on any write
4955 failures on the smtp file descriptor. This function is used when sending
4956 messages over SMTP and also when doing verify callouts.
4958 2. When Exim is called to do synchronous delivery of a locally submitted
4959 message (the -odf or -odi options), it no longer closes stderr before doing
4962 3. Implemented the mua_wrapper option.
4964 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4966 5. Implemented the functions header_remove(), header_testname(),
4967 header_add_at_position(), and receive_remove_recipient(), and exported them
4970 6. If an ACL "warn" statement specified the addition of headers, Exim already
4971 inserted X-ACL-Warn: at the start if there was no header name. However, it
4972 was not making this test for the second and subsequent header lines if
4973 there were newlines in the string. This meant that an invalid header could
4974 be inserted if Exim was badly configured.
4976 7. Allow an ACL "warn" statement to add header lines at the start or after all
4977 the Received: headers, as well as at the end.
4979 8. Added the rcpt_4xx retry error code.
4981 9. Added postmaster_mailfrom=xxx to callout verification option.
4983 10. Added mailfrom=xxxx to the callout verification option, for verify=
4986 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4987 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4989 12. Inserted some casts to stop certain compilers warning when using pointer
4990 differences as field lengths or precisions in printf-type calls (mostly
4991 affecting debugging statements).
4993 13. Added optional readline() support for -be (dynamically loaded).
4995 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4996 same clock tick as a message's arrival, so that its received time was the
4997 same as the "first fail" time on the retry record, and that message
4998 remained on the queue past the ultimate address timeout, every queue runner
4999 would try a delivery (because it was past the ultimate address timeout) but
5000 after another failure, the ultimate address timeout, which should have then
5001 bounced the address, did not kick in. This was a "< instead of <=" error;
5002 in most cases the first failure would have been in the next clock tick
5003 after the received time, and all would be well.
5005 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5006 being recognized when the domain list was tested by the match_domain
5007 condition in an expansion string.
5009 16. Added the ${str2b64: operator.
5011 17. Exim was always calling setrlimit() to set a large limit for the number of
5012 processes, without checking whether the existing limit was already
5013 adequate. (It did check for the limit on file descriptors.) Furthermore,
5014 errors from getrlimit() and setrlimit() were being ignored. Now they are
5015 logged to the main and panic logs, but Exim does carry on, to try to do its
5016 job under whatever limits there are.
5018 18. Imported PCRE 5.0.
5020 19. Trivial typo in log message " temporarily refused connection" (the leading
5023 20. If the log selector return_path_on_delivery was set and an address was
5024 redirected to /dev/null, the delivery process crashed because it assumed
5025 that a return path would always be set for a "successful" delivery. In this
5026 case, the whole delivery is bypassed as an optimization, and therefore no
5029 21. Internal re-arrangement: the function for sending a challenge and reading
5030 a response while authentication was assuming a zero-terminated challenge
5031 string. It's now changed to take a pointer and a length, to allow for
5032 binary data in such strings.
5034 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5036 23. Exim was not respecting finduser_retries when seeking the login of the
5037 uid under which it was called; it was always trying 10 times. (The default
5038 setting of finduser_retries is zero.) Also, it was sleeping after the final
5039 failure, which is pointless.
5041 24. Implemented tls_on_connect_ports.
5043 25. Implemented acl_smtp_predata.
5045 26. If the domain in control=submission is set empty, Exim assumes that the
5046 authenticated id is a complete email address when it generates From: or
5047 Sender: header lines.
5049 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5050 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5051 chown and chgrp in /bin and hostname in /usr/bin.
5053 28. Exim was keeping the "process log" file open after each use, just as it
5054 does for the main log. This opens the possibility of it remaining open for
5055 long periods when the USR1 signal hits a daemon. Occasional processlog
5056 errors were reported, that could have been caused by this. Anyway, it seems
5057 much more sensible not to leave this file open at all, so that is what now
5060 29. The long-running daemon process does not normally write to the log once it
5061 has entered its main loop, and it closes the log before doing so. This is
5062 so that log files can straightforwardly be renamed and moved. However,
5063 there are a couple of unusual error situations where the daemon does write
5064 log entries, and I had neglected to close the log afterwards.
5066 30. The text of an SMTP error response that was received during a remote
5067 delivery was being truncated at 512 bytes. This is too short for some of
5068 the long messages that one sometimes sees. I've increased the limit to
5071 31. It is now possible to make retry rules that apply only when a message has a
5072 specific sender, in particular, an empty sender.
5074 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5075 it possible to be selective about when SMTP synchronization is enforced.
5077 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5079 32. Implemented hosts_connection_nolog.
5081 33. Added an ACL for QUIT.
5083 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5086 35. Added mailbox_size and mailbox_filecount to appendfile.
5088 36. Added control = no_multiline_responses to ACLs.
5090 37. There was a bug in the logic of the code that waits for the clock to tick
5091 in the case where the clock went backwards by a substantial amount such
5092 that the microsecond fraction of "now" was more than the microsecond
5093 fraction of "then" (but the whole seconds number was less).
5095 38. Added support for the libradius Radius client library this is found on
5096 FreeBSD (previously only the radiusclient library was supported).
5102 1. When certain lookups returned multiple values in the form name=value, the
5103 quoting of the values was not always being done properly. Specifically:
5104 (a) If the value started with a double quote, but contained no whitespace,
5106 (b) If the value contained whitespace other than a space character (i.e.
5107 tabs or newlines or carriage returns) it was not quoted.
5108 This fix has been applied to the mysql and pgsql lookups by writing a
5109 separate quoting function and calling it from the lookup code. The fix
5110 should probably also be applied to nisplus, ibase and oracle lookups, but
5111 since I cannot test any of those, I have not disturbed their existing code.
5113 2. A hit in the callout cache for a specific address caused a log line with no
5114 reason for rejecting RCPT. Now it says "Previous (cached) callout
5115 verification failure".
5117 3. There was an off-by-one bug in the queryprogram router. An over-long
5118 return line was truncated at 256 instead of 255 characters, thereby
5119 overflowing its buffer with the terminating zero. As well as fixing this, I
5120 have increased the buffer size to 1024 (and made a note to document this).
5122 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5123 when Exim is waiting for an SMTP response from a remote server, Exim
5124 restarts its select() call on the socket, thereby resetting its timeout.
5125 This is not a problem when such interrupts are rare. Somebody set up a cron
5126 job to run exiwhat every 2 minutes, which is less than the normal select()
5127 timeout (5 or 10 minutes). This meant that the select() timeout never
5128 kicked in because it was always reset. I have fixed this by comparing the
5129 time when an interrupt arrives with the time at the start of the first call
5130 to select(). If more time than the timeout has elapsed, the interrupt is
5131 treated as a timeout.
5133 5. Some internal re-factoring in preparation for the addition of Sieve
5134 extensions (by MH). In particular, the "personal" test is moved to a
5135 separate function, and given an option for scanning Cc: and Bcc: (which is
5136 not set for Exim filters).
5138 6. When Exim created an email address using the login of the caller as the
5139 local part (e.g. when creating a From: or Sender: header line), it was not
5140 quoting the local part when it contained special characters such as @.
5142 7. Installed new OpenBSD configuration files.
5144 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5145 try to make them clearer.
5147 9. Callout options, other than the timeout value, were being ignored when
5148 verifying sender addresses in header lines. For example, when using
5150 verify = header_sender/callout=no_cache
5152 the cache was (incorrectly) being used.
5154 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5155 only the Cygwin environment.
5157 11. When return_path_on_delivery was set as a log selector, if different remote
5158 addresses in the same message used different return paths and parallel
5159 remote delivery occurred, the wrong values would sometimes be logged.
5160 (Whenever a remote delivery process finished, the return path value from
5161 the most recently started remote delivery process was logged.)
5163 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5164 lines when AUTH and/or TLS are in use. This is the "received protocol"
5165 field. Exim used to use "asmtp" for authenticated SMTP, without any
5166 indication (in the protocol name) for TLS use. Now it follows the RFC and
5167 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5168 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5169 names appear in log lines as well as in Received: header lines.
5171 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5172 extensions, and comparison tests, and to fix some bugs.
5174 14. Changes to the "personal" filter test:
5176 (1) The test was buggy in that it was just doing the equivalent of
5177 "contains" tests on header lines. For example, if a user's address was
5178 anne@some.where, the "personal" test would incorrectly be true for
5180 To: susanne@some.where
5182 This test is now done by extracting each address from the header in turn,
5183 and checking the entire address. Other tests that are part of "personal"
5184 are now done using regular expressions (for example, to check local parts
5185 of addresses in From: header lines).
5187 (2) The list of non-personal local parts in From: addresses has been
5188 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5189 taken from the Sieve specification recommendations.
5191 (3) If the message contains any header line starting with "List-" it is
5192 treated as non-personal.
5194 (4) The test for "circular" in the Subject: header line has been removed
5195 because it now seems ill-conceived.
5197 15. Minor typos in src/EDITME comments corrected.
5199 16. Installed latest exipick from John Jetmore.
5201 17. If headers_add on a router specified a text string that was too long for
5202 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5203 of string_sprintf() is now avoided.
5205 18. $message_body_size was not set (it was always zero) when running the DATA
5206 ACL and the local_scan() function.
5208 19. For the "mail" command in an Exim filter, no default was being set for
5209 the once_repeat time, causing a random time value to be used if "once" was
5210 specified. (If the value happened to be <= 0, no repeat happened.) The
5211 default is now 0s, meaning "never repeat". The "vacation" command was OK
5212 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5213 (I found it when inspecting the code).
5215 20. There is now an overall timeout for performing a callout verification. It
5216 defaults to 4 times the callout timeout, which applies to individual SMTP
5217 commands during the callout. The overall timeout applies when there is more
5218 than one host that can be tried. The timeout is checked before trying the
5219 next host. This prevents very long delays if there are a large number of
5220 hosts and all are timing out (e.g. when the network connections are timing
5221 out). The value of the overall timeout can be changed by specifying an
5222 additional sub-option for "callout", called "maxwait". For example:
5224 verify = sender/callout=5s,maxwait=20s
5226 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5227 to the end before writing, but this should make it even safer).
5229 22. Exim was forgetting that it had advertised PIPELINING for the second and
5230 subsequent messages on an SMTP connection. It was also not resetting its
5231 memory on STARTTLS and an internal HELO.
5233 23. When Exim logs an SMTP synchronization error within a session, it now
5234 records whether PIPELINING has been advertised or not.
5236 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5237 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5238 rather than long int.
5240 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5242 26. Added the never_mail option to autoreply.
5248 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5249 crash if the getsockname() call failed; this can happen if a connection is
5250 closed very soon after it is established. The problem was simply in the
5251 order in which certain operations were done, causing Exim to try to write
5252 to the SMTP stream before it had set up the file descriptor. The bug has
5253 been fixed by making things happen in the correct order.
5259 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5260 before the connection was closed, thus losing the rejection response.
5262 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5263 some early Solaris releases, but causes trouble in current releases where
5264 socklen_t is defined.
5266 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5269 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5272 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5273 that releasing the top of it at the end releases what was used for sub-
5274 expansions (unless the block got too big). However, discard this block if
5275 the first thing is a variable or header, so that we can use its block when
5276 it is dynamic (useful for very large $message_headers, for example).
5278 6. Lookups now cache *every* query, not just the most recent. A new, separate
5279 store pool is used for this. It can be recovered when all lookup caches are
5280 flushed. Lookups now release memory at the end of their result strings.
5281 This has involved some general refactoring of the lookup sources.
5283 7. Some code has been added to the store_xxx() functions to reduce the amount
5284 of flapping under certain conditions.
5286 8. log_incoming_interface used to affect only the <= reception log lines. Now
5287 it causes the local interface and port to be added to several more SMTP log
5288 lines, for example "SMTP connection from", and rejection lines.
5290 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5292 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5294 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5295 could be overwritten at the end of the current message (or the start of a
5296 new message if it was set in a HELO ACL). The value is now preserved for
5297 the duration of the SMTP connection.
5299 12. If a transport had a headers_rewrite setting, and a matching header line
5300 contained an unqualified address, that address was qualified, even if it
5301 did not match any rewriting rules. The underlying bug was that the values
5302 of the flags that permit the existence of unqualified sender and recipient
5303 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5304 non-local messages, and by -bnq for local messages) were not being
5305 preserved with the message after it was received.
5307 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5308 "next input=" as part of the text comprising the host identity instead of
5309 the correct text. The code was using the same buffer for two different
5310 strings. However, depending on which order the printing function evaluated
5311 its arguments, the bug did not always show up. Under Linux, for example, my
5312 test suite worked just fine.
5314 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5315 doesn't work with some older versions of Perl. It has been changed to "my",
5316 which in any case is probably the better facility to use.
5318 15. A really picky compiler found some instances of statements for creating
5319 error messages that either had too many or two few arguments for the format
5322 16. The size of the buffer for calls to the DNS resolver has been increased
5323 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5324 for addresses that have a lot of PTR records. This alleviates a problem; it
5325 does not fully solve it.
5327 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5328 buffer now truncates the list and logs the incident, which is the same
5329 action as happens when Exim is looking up a host name and its aliases.
5330 Previously in this situation something unpredictable would happen;
5331 sometimes it was "internal error: store_reset failed".
5333 18. If a server dropped the connection unexpectedly when an Exim client was
5334 using GnuTLS and trying to read a response, the client delivery process
5335 crashed while trying to generate an error log message.
5337 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5338 string, for example:
5340 warn message = H1: something\nH2: something
5342 the text was added as a single header line from Exim's point of view
5343 though it ended up OK in the delivered message. However, searching for the
5344 second and subsequent header lines using $h_h2: did not work. This has been
5345 fixed. Similarly, if a system filter added multiple headers in this way,
5346 the routers could not see them.
5348 20. Expanded the error message when iplsearch is called with an invalid key to
5349 suggest using net-iplsearch in a host list.
5351 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5352 ACLs are no longer actually imposed (and a message to that effect is
5355 22. If a "gecos" field in a passwd entry contained escaped characters, in
5356 particular, if it contained a \" sequence, Exim got it wrong when building
5357 a From: or a Sender: header from that name. A second bug also caused
5358 incorrect handling when an unquoted " was present following a character
5359 that needed quoting.
5361 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5362 was not being matched caselessly.
5364 24. Arranged for all hyphens in the exim.8 source to be escaped with
5367 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5368 properly. Recipient callout cache records were still being keyed to include
5369 the sender, even when use_sender was set false. This led to far more
5370 callouts that were necessary. The sender is no longer included in the key
5371 when use_sender is false.
5373 26. Added "control = submission" modifier to ACLs.
5375 27. Added the ${base62d: operator to decode base 62 numbers.
5377 28. dnsdb lookups can now access SRV records.
5379 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5380 the configuration file.
5382 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5383 (-v) mode. This makes the output for a verbose queue run more intelligible.
5385 31. Added a use_postmaster feature to recipient callouts.
5387 32. Added the $body_zerocount variable, containing the number of binary zero
5388 bytes in the message body.
5390 33. The time of last modification of the "new" subdirectory is now used as the
5391 "mailbox time last read" when there is a quota error for a maildir
5394 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5396 35. Added +ignore_unknown as a special item in host lists.
5398 36. Code for decoding IPv6 addresses in host lists is now included, even if
5399 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5400 address was recognized as an IP address, but was then not correctly decoded
5401 into binary, causing unexpected and incorrect effects when compared with
5408 1. Very minor rewording of debugging text in manualroute to say "list of
5409 hosts" instead of "hostlist".
5411 2. If verify=header_syntax was set, and a header line with an unqualified
5412 address (no domain) and a large number of spaces between the end of the
5413 name and the colon was received, the reception process suffered a buffer
5414 overflow, and (when I tested it) crashed. This was caused by some obsolete
5415 code that should have been removed. The fix is to remove it!
5417 3. When running in the test harness, delay a bit after writing a bounce
5418 message to get a bit more predictability in the log output.
5420 4. Added a call to search_tidyup() just before forking a reception process. In
5421 theory, someone could use a lookup in the expansion of smtp_accept_max_
5422 per_host which, without the tidyup, could leave open a database connection.
5424 5. Added the variables $recipient_data and $sender_data which get set from a
5425 lookup success in an ACL "recipients" or "senders" condition, or a router
5426 "senders" option, similar to $domain_data and $local_part_data.
5428 6. Moved the writing of debug_print from before to after the "senders" test
5431 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5432 problems for message scanning, either using a data ACL, or using
5433 local_scan() because the Received: header was not generated till after they
5434 were called (in order to set the time as the time of reception completion).
5435 I have revised the way this works. The header is now generated after the
5436 body is received, but before the ACL or local_scan() are called. After they
5437 are run, the timestamp in the header is updated.
5443 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5444 before starting a queue runner without re-exec. This happened only when
5445 deliver_drop_privilege was set or when the Exim user was set to root. The
5446 effect of the bug was that timeouts during subsequent deliveries caused
5447 crashes instead of being properly handled. The handler is now left at its
5448 default (and expected) setting.
5450 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5451 message, again when deliver_drop_privilege is set or Exim is run as root.
5452 The bug described in (1) was not present in this case, but the tidying up
5453 of the other signals was missing. I have made the two cases consistent.
5455 3. The ignore_target_hosts setting on a manualroute router was being ignored
5456 for hosts that were looked up using the /MX notation.
5458 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5461 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5462 operated on the sender address. After changing the $sender_address to <>
5463 for the sender address verify, Exim was re-instated it as the original
5464 (before rewriting) address, but remembering that it had rewritten it, so it
5465 wasn't rewriting it again. This bug also had the effect of breaking the
5466 sender address verification caching when the sender address was rewritten.
5468 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5469 This has been changed so that if the ip literal address matches
5470 ignore_target_hosts, the router declines.
5472 7. Added expansion conditions match_domain, match_address, and match_local_
5473 part (NOT match_host).
5475 8. The placeholder for the Received: header didn't have a length field set.
5477 9. Added code to Exim itself and to exim_lock to test for a specific race
5478 condition that could lead to file corruption when using MBX delivery. The
5479 issue is with the lockfile that is created in /tmp. If this file is removed
5480 after a process has opened it but before that process has acquired a lock,
5481 there is the potential for a second process to recreate the file and also
5482 acquire a lock. This could lead to two Exim processes writing to the file
5483 at the same time. The added code performs the same test as UW imapd; it
5484 checks after acquiring the lock that its file descriptor still refers to
5485 the same named file.
5487 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5488 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5491 11. Added the smtp_active_hostname option. If used, this will typically be made
5492 to depend on the incoming interface address. Because $interface_address is
5493 not set up until the daemon has forked a reception process, error responses
5494 that can happen earlier (such as "too many connections") no longer contain
5497 12. If an expansion in a condition on a "warn" statement fails because a lookup
5498 defers, the "warn" statement is abandoned, and the next ACL statement is
5499 processed. Previously this caused the whole ACL to be aborted.
5501 13. Added the iplsearch lookup type.
5503 14. Added ident_timeout as a log selector.
5505 15. Added tls_certificate_verified as a log selector.
5507 16. Added a global option tls_require_ciphers (compare the smtp transport
5508 option of the same name). This controls incoming TLS connections.
5510 17. I finally figured out how to make tls_require_ciphers do a similar thing
5511 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5512 before starting the TLS session.
5514 18. Tabs are now shown as \t in -bP output.
5516 19. If the log selector return_path_on_delivery was set, Exim crashed when
5517 bouncing a message because it had too many Received: header lines.
5519 20. If two routers both had headers_remove settings, and the first one included
5520 a superfluous trailing colon, the final name in the first list and the
5521 first name in the second list were incorrectly joined into one item (with a
5522 colon in the middle).
5528 1. Added -C and -D options to the exinext utility, mainly to make it easier
5529 to include in the automated testing, but these could be helpful when
5530 multiple configurations are in use.
5532 2. The exinext utility was not formatting the output nicely when there was
5533 an alternate port involved in the retry record key, nor when there was a
5534 message id as well (for retries that were specific to a specific message
5535 and a specific host). It was also confused by IPv6 addresses, because of
5536 the additional colons they contain. I have fixed the IPv4 problem, and
5537 patched it up to do a reasonable job for IPv6.
5539 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5540 delivery, the log line now contains "pipelined" if PIPELINING was used.
5542 4. An SMTP transport process used to panic and die if the bind() call to set
5543 an explicit outgoing interface failed. This has been changed; it is now
5544 treated in the same way as a connect() failure.
5546 5. A reference to $sender_host_name in the part of a conditional expansion
5547 that was being skipped was still causing a DNS lookup. This no longer
5550 6. The def: expansion condition was not recognizing references to header lines
5551 that used bh_ and bheader_.
5553 7. Added the _cache feature to named lists.
5555 8. The code for checking quota_filecount in the appendfile transport was
5556 allowing one more file than it should have been.
5558 9. For compatibility with Sendmail, the command line option
5566 and sets the incoming protocol and host name (for trusted callers). The
5567 host name and its colon can be omitted when only the protocol is to be set.
5568 Note the Exim already has two private options, -pd and -ps, that refer to
5569 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5570 "s", but I don't think that's a major issue.
5572 10. A number of refactoring changes to the code, none of which should affect
5575 (a) The number of logging options was getting close to filling up the
5576 32-bit word that was used as a bit map. I have split them into two classes:
5577 those that are passed in the argument to log_write(), and those that are
5578 only ever tested independently outside of that function. These are now in
5579 separate 32-bit words, so there is plenty of room for expansion again.
5580 There is no change in the user interface or the logging behaviour.
5582 (b) When building, for example, log lines, the code previously used a
5583 macro that called string_cat() twice, in order to add two strings. This is
5584 not really sufficiently general. Furthermore, there was one instance where
5585 it was actually wrong because one of the argument was used twice, and in
5586 one call a function was used. (As it happened, calling the function twice
5587 did not affect the overall behaviour.) The macro has been replaced by a
5588 function that can join an arbitrary number of extra strings onto a growing
5591 (c) The code for expansion conditions now uses a table and a binary chop
5592 instead of a serial search (which was left over from when there were very
5593 few conditions). Also, it now recognizes conditions like "pam" even when
5594 the relevant support is not compiled in: a suitably worded error message is
5595 given if an attempt is made to use such a condition.
5597 11. Added ${time_interval:xxxxx}.
5599 12. A bug was causing one of the ddress fields not to be passed back correctly
5600 from remote delivery subprocesses. The field in question was not being
5601 subsequently used, so this caused to problems in practice.
5603 13. Added new log selectors queue_time and deliver_time.
5605 14. Might have fixed a bug in maildirsizefile handling that threw up
5606 "unexpected character" debug warnings, and recalculated the data
5607 unnecessarily. In any case, I expanded the warning message to give more
5610 15. Added the message "Restricted characters in address" to the statements in
5611 the default ACL that block characters like @ and % in local parts.
5613 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5614 Three changes have been made:
5616 (a) There was a serious bug; a negative response to MAIL caused the whole
5617 recipient domain to be cached as invalid, thereby blocking all messages
5618 to all local parts at the same domain, from all senders. This bug has
5619 been fixed. The domain is no longer cached after a negative response to
5620 MAIL if the sender used is not empty.
5622 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5625 (c) A new callout option, "use_sender" has been added for people who want
5626 the modified behaviour.
5632 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5635 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5636 indeed breaks things for older releases.
5638 3. Added additional logging to the case where there is a problem reading data
5639 from a filter that is running in a subprocess using a pipe, in order to
5640 try to track down a specific problem.
5642 4. Testing facility fudge: when running in the test harness and attempting
5643 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5644 getting "No route to host". Convert this to a timeout.
5646 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5649 6. Some OS don't have socklen_t but use size_t instead. This affects the
5650 fifth argument of getsockopt() amongst other things. This is now
5651 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5652 can be set for individual OS. I have set it for SunOS5, OSF1, and
5653 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5654 some earlier ones do not.
5656 7. Change 4.30/15 was not doing the test caselessly.
5658 8. The standard form for an IPv6 address literal was being rejected by address
5659 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5660 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5661 this, as well as the form without the "IPv6" on the front (but only when
5662 address literals are enabled, of course).
5664 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5666 10. Exim crashed if a message with an empty sender address specified by -f
5667 encountered a router with an errors_to setting. This could be provoked only
5668 by a command such as
5672 where an empty string was supplied; "<>" did not hit this bug.
5674 11. Installed PCRE release 4.5.
5676 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5677 remained set. It is now erased.
5679 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5680 times from message ids (which are base 36 rather than the normal 62).
5682 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5683 were being counted as actual protocol errors, and logged if the log
5684 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5685 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5686 and DATA following a set of rejected RCPTs do not count as protocol errors.
5687 In other words, Exim assumes they were pipelined, though this may not
5688 actually be the case. Of course, in all cases the client gets an
5689 appropriate error code.
5691 15. If a lookup fails in an ACL condition, a message about the failure may
5692 be available; it is used if testing the ACL cannot continue, because most
5693 such messages specify what the cause of the deferral is. However, some
5694 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5695 that caused an old message to be retained and used if a later statement
5696 caused a defer, replacing the real cause of the deferral.
5698 16. If an IP address had so many PTR records that the DNS lookup buffer
5699 was not large enough to hold them, Exim could crash while trying to process
5700 the truncated data. It now detects and logs this case.
5702 17. Further to 4.21/58, another change has been made: if (and only if) the
5703 first line of a message (the first header line) ends with CRLF, a bare LF
5704 in a subsequent header line has a space inserted after it, so as not to
5705 terminate the header.
5707 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5708 unnecessarily, used atoi() instead of strtol(), and didn't check the
5709 termination when getting file sizes from file names by regex.
5711 19. Completely re-implemented the support for maildirsize files, in the light
5712 of a number of problems with the previous contributed implementation
5713 (4.30/29). In particular:
5715 . If the quota is zero, the maildirsize file is maintained, but no quota is
5718 . If the maildir directory does not exist, it is created before any attempt
5719 to write a maildirsize file.
5721 . The quota value in the file is just a cache; if the quota is changed in
5722 the transport, the new value overrides.
5724 . A regular expression is available for excluding directories from the
5727 20. The autoreply transport checks the characters in options that define the
5728 message's headers; it allows continued headers, but it was checking with
5729 isspace() after an embedded newline instead of explicitly looking for a
5732 21. If all the "regular" hosts to which an address was routed had passed their
5733 expiry times, and had not reached their retry times, the address was
5734 bounced, even if fallback hosts were defined. Now Exim should go on to try
5737 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5738 equivalent code in the SMTP transport. Some hosts send humungous responses
5739 to HELO/EHLO, more than 1024 it seems.
5741 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5742 gives compiler warnings in some environments. I've now done it "properly",
5745 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5746 (because of problems with the built-in one) was declared to return uschar *
5747 instead of char *, causing compiler failure.
5749 25. Fixed a file descriptor leak when processing alias/forward files.
5751 26. Fixed a minor format string issue in dbfn.c.
5753 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5755 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5756 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5759 29. When a sender address is verified, it is cached, to save repeating the test
5760 when there is more than one recipient in a message. However, when the
5761 verification involves a callout, it is possible for different callout
5762 options to be set for different recipients. It is too complicated to keep
5763 track of this in the cache, so now Exim always runs a verification when a
5764 callout is required, relying on the callout cache for the optimization.
5765 The overhead is duplication of the address routing, but this should not be
5768 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5769 to be verified with callout=postmaster, and the main callout worked but the
5770 postmaster check failed, the verification correctly failed. However, if a
5771 subsequent RCPT command asked for sender verification *without* the
5772 postmaster check, incorrect caching caused this verification also to fail,
5775 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5776 it was not caching the DNS options (qualify_single, search_parents) that
5777 were used when the lookup failed. A subsequent lookup with different
5778 options therefore always gave the same answer, though there were cases
5779 where it should not have. (Example: a "domains = !$mx_any" option on a
5780 dnslookup router: the "domains" option is always processed without any
5781 widening, but the router might have qualify_single set.) Now Exim uses the
5782 cached value only when the same options are set.
5784 32. Added John Jetmore's "exipick" utility to the distribution.
5786 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5787 than a timeout (e.g. a certificate is required, and is not provided), an
5788 Exim server now closes the connection immediately. Previously it waited for
5789 the client to close - but if the client is SSL, it seems that they each
5790 wait for each other, leading to a delay before one of them times out.
5792 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5793 maintained 0.8.x compatibility because I don't think many are using it, and
5794 it is clearly obsolete.
5796 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5799 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5800 client certificate was expired. A simple patch fixes this, though I don't
5801 understand the full logic of why the verify callback is called multiple
5804 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5805 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5806 which causes problems with some clients (such as the Certicom SSL Plus
5807 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5808 disables the coutermeasure allowing Eudora to connect."
5810 38. Exim was not checking that a write() to a log file succeeded. This could
5811 lead to Bad Things if a log got too big, in particular if it hit a file
5812 size limit. Exim now panics and dies if it cannot write to a log file, just
5813 as it does if it cannot open a log file.
5815 39. Modified OS/Makefile-Linux so that it now contains
5817 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5819 The two -D definitions ensure that Exim is compiled with large file
5820 support, which makes it possible to handle log files that are bigger than
5823 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5824 instance) a domain was checked against a named list that involved a lookup,
5825 causing $domain_data to be set, then another domain was checked against the
5826 same list, then the first domain was re-checked, the value of $domain_data
5827 after the final check could be wrong. In particular, if the second check
5828 failed, it could be set empty. This bug probably also applied to
5831 41. The strip_trailing_dot option was not being applied to the address given
5832 with the -f command-line option.
5834 42. The code for reading a message's header from the spool was incrementing
5835 $received_count, but never initializing it. This meant that the value was
5836 incorrect (doubled) while delivering a message in the same process in which
5837 it was received. In the most common configuration of Exim, this never
5838 happens - a fresh exec is done - but it can happen when
5839 deliver_drop_privilege is set.
5841 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5842 it now includes up to 150 characters of the unexpected data in the log
5845 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5846 and building data strings. The size of both of these buffers was 10 000
5847 bytes - far larger than anybody would *ever* want, thought I. Needless to
5848 say, somebody hit the limit. I have increased the maximum line length to
5849 20 000 and the maximum data length of concatenated lines to 100 000. I have
5850 also fixed two bugs, because there was no checking on these buffers. Tsk,
5851 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5852 buffer is too small.
5854 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5855 lsearch lookups. Now it does.
5857 46. When parsing a route_list item in a manualroute router, a fixed-length
5858 buffer was used for the list of hosts. I made this 1024 bytes long,
5859 thinking that nobody would ever have a list of hosts that long. Wrong.
5860 Somebody had a whole pile of complicated expansion conditions, and the
5861 string was silently truncated, leading to an expansion error. It turns out
5862 that it is easier to change to an unlimited length (owing to other changes
5863 that have happened since this code was originally written) than to build
5864 structure for giving a limitation error. The length of the item that
5865 expands into the list of hosts is now unlimited.
5867 47. The lsearch lookup could not handle data where the length of text line was
5868 more than 4095 characters. Such lines were truncated, leading to shortened
5869 data being returned. It should now handle lines of any length.
5871 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5872 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5875 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5876 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5877 no longer happens in comment lines. A list of replacements is now placed
5878 at the head of all of the source files, except those whose only change is
5879 to replace PERL_COMMAND in the very first #! line.
5881 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5882 messages on the queue, with a bottom-up merge sort, using code contributed
5883 by Michael Haardt. This should make operations like -bp somewhat faster on
5884 large queues. It won't affect queue runners, except when queue_run_in_order
5887 51. Installed eximstats 1.31 in the distribution.
5889 52. Added support for SRV lookups to the dnslookup router.
5891 53. If an ACL referred to $message_body or $message_body_end, the value was not
5892 reset for any messages that followed in the same SMTP session.
5894 54. The store-handling optimization for building very long strings was not
5895 differentiating between the different store pools. I don't think this
5896 actually made any difference in practice, but I've tidied it.
5898 55. While running the routers to verify a sender address, $sender_address
5899 was still set to the sender address. This is wrong, because when routing to
5900 send a bounce to the sender, it would be empty. Therefore, I have changed
5901 it so that, while verifying a sender address, $sender_address is set to <>.
5902 (There is no change to what happens when verifying a recipient address.)
5904 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5905 target A or AAAA records (if not already returned) without resetting the
5906 qualify_single or search_parents options of the DNS resolver. These are
5907 inappropriate in this case because the targets of MX and SRV records must
5908 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5909 target that, when qualified, matched something in the local domain. These
5910 two options are now turned off when doing these lookups.
5912 57. It seems that at least some releases of Reiserfs (which does not have the
5913 concept of a fixed number of inodes) returns zero and not -1 for the
5914 number of available inodes. This interacted badly with check_spool_inodes,
5915 which assumed that -1 was the "no such thing" setting. What I have done is
5916 to check that the total number of inodes is greater than zero before doing
5917 the test of how many are available.
5919 58. When a "warn" ACL statement has a log_message modifier, the message is
5920 remembered, and not repeated. This is to avoid a lot of repetition when a
5921 message has many recipients that cause the same warning to be written.
5922 However, Exim was preserving the list of already written lines for an
5923 entire SMTP session, which doesn't seem right. The memory is now reset if a
5924 new message is started.
5926 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5927 debugging output unless log_rewrite was also set.
5929 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5930 of (int)(handle) when we know that handle contains (void *)(-1).
5932 61. The Exim daemon panic-logs an error return when it closes the incoming
5933 connection. However "connection reset by peer" seems to be common, and
5934 isn't really an error worthy of noting specially, so that particular error
5937 62. When Exim is trying to find all the local interfaces, it used to panic and
5938 die if the ioctl to get the interface flags failed. However, it seems that
5939 on at least one OS (Solaris 9) it is possible to have an interface that is
5940 included in the list of interfaces, but for which you get a failure error
5941 for this call. This happens when the interface is not "plumbed" into a
5942 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5943 failure of the "get flags" call assumes that the interface is down.
5945 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5946 makes life easier for people who are doing arithmetic on fields extracted
5947 from dates, where you often get leading zeros that should not be
5948 interpreted as octal.
5950 64. Added qualify_domain to the redirect router, to override the global
5953 65. If a pathologically long header line contained very many addresses (the
5954 report of this problem mentioned 10 000) and each of them was rewritten,
5955 Exim could use up a very large amount of memory. (It kept on making new
5956 copies of the header line as it rewrote, and never released the old ones.)
5957 At the expense of a bit more processing, the header rewriting function has
5958 been changed so that it no longer eats memory in this way.
5960 66. The generation of the Received: header has been moved from the time that a
5961 message starts to be received, to the time that it finishes. The timestamp
5962 in the Received: header should now be very close to that of the <= log
5963 line. There are two side-effects of this change:
5965 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5966 logged header lines no longer include the local Received: line, because
5967 it has not yet been created. The same applies to a copy of the message
5968 that is returned to a non-SMTP sender when a message is rejected.
5970 (b) When a filter file is tested using -bf, no additional Received: header
5971 is added to the test message. After some thought, I decided that this
5974 This change does not affect the value of $received_for. It is still set
5975 after address rewriting, but before local_scan() is called.
5977 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5979 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5980 gave an unhelpful panic error message, and a defer error. I have managed to
5981 change this behaviour so that it now rejects any supplied certificate,
5982 which seems right, as the list of acceptable certificates is empty.
5984 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5985 gave an unhelpful defer error. I have not managed to make this reject any
5986 supplied certificates, but the error message it gives is "no certificate
5987 supplied", which is not helpful.
5989 70. exigrep's output now also includes lines that are not associated with any
5990 message, but which match the given pattern. Implemented by a patch from
5991 Martin Sluka, which also tidied up the Perl a bit.
5993 71. Recipient callout verification, like sender verification, was using <> in
5994 the MAIL FROM command. This isn't really the right thing, since the actual
5995 sender may affect whether the remote host accepts the recipient or not. I
5996 have changed it to use the actual sender in the callout; this means that
5997 the cache record is now keyed on a recipient/sender pair, not just the
5998 recipient address. There doesn't seem to be a real danger of callout loops,
5999 since a callout by the remote host to check the sender would use <>.
6000 [SEE ABOVE: changed after hitting problems.]
6002 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6003 temporary errors. However, in the case of such a code being given after
6004 the end of a data transmission (i.e. after ".") Exim was failing to write
6005 a retry record for the message. (Yes, there was some broken host that was
6006 actually sending 8xx at this point.)
6008 73. An unknown lookup type in a host list could cause Exim to panic-die when
6009 the list was checked. (An example that provoked this was putting <; in the
6010 middle of a list instead of at the start.) If this happened during a DATA
6011 ACL check, a -D file could be left lying around. This kind of configuration
6012 error no longer causes Exim to die; instead it causes a defer error. The
6013 incident is still logged to the main and panic logs.
6015 74. Buglet left over from Exim 3 conversion. The message "too many messages
6016 in one connection" was written to the rejectlog but not the mainlog, except
6017 when address rewriting (yes!) was being logged.
6019 75. Added write_rejectlog option.
6021 76. When a system filter was run not as root (that is, when system_filter_user
6022 was set), the values of the $n variables were not being returned to the
6023 main process; thus, they were not subsequently available in the $sn
6026 77. Added +return_path_on_delivery log selector.
6028 78. A connection timeout was being treated differently from recipients deferred
6029 when testing hosts_max_try with a message that was older than the host's
6030 retry timeout. (The host should not be counted, thus allowing all hosts to
6031 be tried at least once before bouncing.) This may have been the cause of an
6032 occasionally reported bug whereby a message would remain on the queue
6033 longer than the retry timeout, but would be bounced if a delivery was
6034 forced. I say "may" because I never totally pinned down the problem;
6035 setting up timeout/retry tests is difficult. See also the next item.
6037 79. The ultimate address timeout was not being applied to errors that involved
6038 a combination of host plus message (for example, a timeout on a MAIL
6039 command). When an address resolved to a number of possible hosts, and they
6040 were not all tried for each delivery (e.g. because of hosts_max_try), a
6041 message could remain on the queue longer than the retry timeout.
6043 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6046 81. Fixed an obscure SMTP outgoing bug which required at least the following
6047 conditions: (a) there was another message waiting for the same server;
6048 (b) the server returned 5xx to all RCPT commands in the first message so
6049 that the message was not completed; (c) the server dropped the connection
6050 or gave a negative response to the RSET that Exim sends to abort the
6051 transaction. The observed case was a dropped connection after DATA that had
6052 been sent in pipelining mode. That is, the server had advertised PIPELINING
6053 but was not implementing it correctly. The effect of the bug was incorrect
6054 behaviour, such as trying another host, and this could lead to a crash.
6060 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6061 and daemon.c were passed as pointers to ints; they should have been
6062 pointers to socklen_t variables (which are typically unsigned ints).
6064 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6067 3. Fixed a really odd bug that affected only the testing scheme; patching a
6068 certain fixed string in the binary changed the value of another string that
6069 happened to be identical to the end of the original first string.
6071 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6072 name", it returns that address as the IP address. On some operating
6073 systems (e.g. Solaris), it also passes back the IP address string as the
6074 "host name". However, on others (e.g. Linux), it passes back an empty
6075 string. Exim wasn't checking for this, and was changing the host name to an
6076 empty string, assuming it had been canonicalized.
6078 5. Although rare, it is permitted to have more than one PTR record for a given
6079 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6080 all the names associated with an address, because they do in Solaris.
6081 However, it seems that they do not in Linux for data that comes from the
6082 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6083 I found this out when I moved to a new Linux workstation and tried to run
6084 the Exim test suite.
6086 To get round this problem I have changed the code so that it now does its
6087 own call to the DNS to look up PTR records when searching for a host name.
6088 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6089 addresses that are only in /etc/hosts are still found.
6091 This behaviour is, however, controlled by an option called host_lookup_
6092 order, which defaults to "bydns:byaddr". If people want to use the other
6093 order, or indeed, just use one or the other means of lookup, they can
6094 specify it in this variable.
6096 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6097 some operating systems, this comes back from gethostbyaddr() as an empty
6098 string, and this is what Exim used to test for. However, it seems that in
6099 other systems, "." is yielded. Exim now tests for this case too.
6101 7. The values of check_spool_space and check_log_space are now held internally
6102 as a number of kilobytes instead of an absolute number of bytes. If a
6103 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6104 kilobyte. This means that much larger values can be stored.
6106 8. Exim monitor: an attempt to get the action menu when not actually pointing
6107 at a message produces an empty menu entitled "No message selected". This
6108 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6109 no entries in it ("Shell widget menu has zero width and/or height"). So I
6110 have added a single, blank menu entry in this case.
6112 9. Added ${quote_local_part.
6114 10. MIME decoding is now applied to the contents of Subject: header lines when
6117 11. Now that a reference to $sender_host_address automatically causes a reverse
6118 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6119 host lookup before query-style lookups in lists that might use this
6120 variable. This has therefore been abolished, and the "net-" prefix is no
6121 longer necessary for query-style lookups.
6123 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6124 have been a typo for LFLAGS, so it has been changed.
6126 13. The install script calls Exim with "-C /dev/null" in order to find the
6127 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6128 to be output. However, since Exim outputs its version number before the
6129 error, it didn't break the script. It just looked ugly. I fixed this by
6130 always allowing "-C /dev/null" if the caller is root.
6132 14. Ignore overlarge ACL variable number when reading spool file - insurance
6133 against a later release with more variables having written the file.
6135 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6136 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6137 this, as well as the form without the "IPv6" on the front.
6139 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6140 OS/Makefile-Darwin file.
6142 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6143 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6145 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6146 variable. However, some OS use ints for inodes. I've added cast to long int
6147 to get rid of the compiler warning.
6149 19. I had forgotten to lock out "/../" in configuration file names when
6150 ALT_CONFIG_PREFIX was set.
6152 20. Routers used for verification do not need to specify transports. However,
6153 if such a router generated a host list, and callout was configured, Exim
6154 crashed, because it could not find a port number from the (non-existent)
6155 transport. It now assumes port 25 in this circumstance.
6157 21. Added the -t option to exigrep.
6159 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6160 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6161 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6162 them is set, LOOKUP_LSEARCH is forced.
6164 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6166 24. Added sender and host information to the "rejected by local_scan()" log
6167 line; previously there was no indication of these.
6169 25. Added .include_if_exists.
6171 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6172 Linux. It turns out that not all file systems support this. Apparently some
6173 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6174 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6175 not supported on the file descriptor, is now ignored when Exim is trying to
6176 sync a directory. This applies only to Linux.
6178 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6180 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6181 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6184 29. Added support for maildirsize files from supplied patch (modified a bit).
6186 30. The use of :fail: followed by an empty string could lead Exim to respond to
6187 sender verification failures with (e.g.):
6189 550 Verification failed for <xxx>
6190 550 Sender verify failed
6192 where the first response line was missing the '-' that indicates it is not
6193 the final line of the response.
6195 31. The loop for finding the name of the user that called Exim had a hardwired
6196 limit of 10; it now uses the value of finduser_retries, which is used for
6197 all other user lookups.
6199 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6202 33. Exim was neglecting to zero errno before one call of strtol() when
6203 expanding a string and expecting an integer value. On some systems this
6204 resulted in spurious "integer overflow" errors. Also, it was casting the
6205 result into an int without checking.
6207 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6208 did not work. The code looks as if it has *never* worked, though it appears
6209 to have been documented since at least release 1.62. I have made it work.
6211 35. The "timeout_DNS" error in retry rules, also documented since at least
6212 1.62, also never worked. As it isn't clear exactly what this means, and
6213 clearly it isn't a major issue, I have abolished the feature by treating it
6214 as "timeout", and writing a warning to the main and panic logs.
6216 36. The display of retry rules for -brt wasn't always showing the error code
6219 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6220 timeout_connect_A, timeout_connect_MX.
6222 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6223 to the empty sender.
6225 39. The daemon was not analysing the content of -oX till after it had closed
6226 stderr and disconnected from the controlling terminal. This meant that any
6227 syntax errors were only noted on the panic log, and the return code from
6228 the command was 0. By re-arranging the code a little, I've made the
6229 decoding happen first, so such errors now appear on stderr, and the return
6230 code is 1. However, the actual setting up of the sockets still happens in
6231 the disconnected process, so errors there are still only recorded on the
6234 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6235 connections (as happens on some IP stacks) was logged at start up time as
6236 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6237 it from "IPv6 and IPv4", which means that two separate sockets are being
6240 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6241 says whether AF_INET or AF_INET6 was passed as an argument.
6243 42. Exiwhat output was messed up when time zones were included in log
6246 43. Exiwhat now gives more information about the daemon's listening ports,
6247 and whether -tls-on-connect was used.
6249 44. The "port" option of the smtp transport is now expanded.
6251 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6252 silently ignored. Now an error message is written to the main and panic
6255 46. There's a new ACL modifier called "logwrite" which writes to a log file
6256 as soon as it is encountered.
6258 47. Added $local_user_uid and $local_user_gid at routing time.
6260 48. Exim crashed when trying to verify a sender address that was being
6263 49. Exim was recognizing only a space character after ".include". It now also
6264 recognizes a tab character.
6266 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6267 extracting the relevant information from the specification. The man page no
6268 longer contains scrambled data for the -d option, and I've added a section
6269 at the front about calling Exim under different names.
6271 51. Added "extra_headers" argument to the "mail" command in filter files.
6273 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6276 53. Installed eximstats 1.29.
6278 54. Added transport_filter_timeout as a generic transport option.
6280 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6281 Cc: header lines. This was required by RFC 822, but it not required by RFC
6284 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6285 incoming messages that did not have them. Now it does so only if the
6286 message originates locally, that is, if there is no associated remote host
6287 address. When Resent- header lines are present, this applies to the Resent-
6288 lines rather than the non-Resent- lines.
6290 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6291 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6293 58. Messages for configuration errors now include the name of the main
6294 configuration file - useful now that there may be more than one file in a
6295 list (.included file names were always shown).
6297 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6298 for those rare installations that do not start the daemon as root or run it
6299 setuid root. I've cut out the call to initgroups() if the daemon is not
6302 60. The Exim user and group can now be bound into the binary as text strings
6303 that are looked up at the start of Exim's processing.
6305 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6307 62. Added $mailstore_basename variable.
6309 63. Installed patch to sieve.c from Michael Haardt.
6311 64. When Exim failed to open the panic log after failing to open the main log,
6312 the original message it was trying to log was written to stderr and debug
6313 output, but if they were not available (the usual case in production), it
6314 was lost. Now it is written to syslog before the two lines that record the
6315 failures to open the logs.
6317 65. Users' Exim filters run in subprocesses under the user's uid. It is
6318 possible for a "deliver" command or an alias in a "personal" command to
6319 provoke an address rewrite. If logging of address rewriting is configured,
6320 this fails because the process is not running as root or exim. There may be
6321 a better way of dealing with this, but for the moment (because 4.30 needs
6322 to be released), I have disabled address rewrite logging when running a
6323 filter in a non-root, non-exim process.
6329 1. The buildconfig auxiliary program wasn't quoting the value set for
6330 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6331 not defined. This bug was masked in 4.22 by the effect that was fixed in
6334 2. Some messages that were rejected after a message id was allocated were
6335 shown as "incomplete" by exigrep. It no longer does this for messages that
6336 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6338 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6339 have allow_domain_literals set, the ID did not get logged in the <= line.
6340 Domain literals are now always recognized in Message-ID: header lines.
6342 4. The first argument for a ${extract expansion item is the key name or field
6343 number. Leading and trailing spaces in this item were not being ignored,
6344 causing some misleading effects.
6346 5. When deliver_drop_privilege was set, single queue runner processes started
6347 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6348 same command in the process it spins off) were not dropping privilege.
6350 6. When the daemon running as "exim" started a queue runner, it always
6351 re-executed Exim in the spun-off process. This is a waste of effort when
6352 deliver_drop_privilege is set. The new process now just calls the
6353 queue-runner function directly.
6359 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6362 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6363 ignored. Though the use of -oP was forcing the writing of a pid file, it
6364 was always written to the default place.
6366 3. If the message "no IP address found for host xxxx" is generated during
6367 incoming verification, it is now followed by identification of the incoming
6368 connection (so you can more easily find what provoked it).
6370 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6372 5. Added some features to "harden" Exim a bit more against certain attacks:
6374 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6375 be put in Local/Makefile. This is like the never_users runtime option,
6376 but it cannot be overridden. The default setting is "root".
6378 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6379 prefix string with which any file named in a -C command line option
6382 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6383 is retained for -C and -D only if the caller of Exim is root. Without
6384 it, the exim user may also use -C and -D and retain privilege.
6386 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6387 command line option is disabled.
6389 6. Macro names set by the -D option must start with an upper case letter, just
6390 like macro names defined in the configuration file.
6392 7. Added "dereference=" facility to LDAP.
6394 8. Two instances of the typo "uknown" in the source files are fixed.
6396 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6397 the Configure-Makefile script screwed up while processing it.
6399 10. Incorporated PCRE 4.4.
6401 11. The SMTP synchronization check was not operating right at the start of an
6402 SMTP session. For example, it could not catch a HELO sent before the client
6403 waited for the greeting. There is now a check for outstanding input at the
6404 point when the greeting is written. Because of the duplex, asynchronous
6405 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6406 way, but not yet received, when the check is performed.
6408 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6409 on TCP/IP sockets, because this apparently causes some broken clients to
6412 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6413 unchanged) from the Cygwin maintainer.
6415 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6416 was supported instead of testing for mbx. Effectively a typo.
6418 15. The spa authenticator server code was not checking that the input it
6419 received was valid base64.
6421 16. The debug output line for the "set" modifier in ACLs was not showing the
6422 name of the variable that was being set.
6424 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6426 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6427 lookup on its own. Something else was needed to trigger the lookup. For
6428 example, a match in host_lookup or the need for a host name in a host list.
6429 Now, if $sender_host_name is referenced and the host name has not yet been
6430 looked up, a lookup is performed. If the lookup fails, the variable remains
6431 empty, and $host_lookup_failed is set to "1".
6433 19. Added "eqi" as a case-independent comparison operator.
6435 20. The saslauthd authentication condition could segfault if neither service
6436 nor realm was specified.
6438 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6439 error message that was logged was misleading, and incoming SMTP
6440 connections were dropped. The message is now more accurate, and temporary
6441 errors are given to SMTP connections.
6443 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6444 (except RSET) with a 421 error, until QUIT is received. However, it was
6445 failing to send a response to QUIT.
6447 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6448 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6449 ACL is now run after the helo_try_verify_hosts code.
6451 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6452 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6453 case-independent, so other case variants are also recognized). Apparently
6454 some systems use these upper case variants.
6456 25. If more than two messages were waiting for the same host, and a transport
6457 filter was specified for the transport, Exim sent two messages over the
6458 same TCP/IP connection, and then failed with "socket operation on non-
6459 socket" when it tried to send the third.
6461 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6463 27. The extern definition of crypt16() in expand.c was not being excluded when
6464 the OS had its own crypt16() function.
6466 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6467 as a preferred synonym for return_size_limit, both as an option and as an
6470 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6472 30. Changed the default configuration ACL to relax the local part checking rule
6473 for addresses that are not in any local domains. For these addresses,
6474 slashes and pipe symbols are allowed within local parts, but the sequence
6475 /../ is explicitly forbidden.
6477 31. SPA server authentication was not clearing the challenge buffer before
6480 32. log_message in a "warn" ACL statement was writing to the reject log as
6481 well as to the main log, which contradicts the documentation and doesn't
6482 seem right (because no rejection is happening). So I have stopped it.
6484 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6485 However, I am unable to do any testing of this.
6487 34. Fixed an infelicity in the appendfile transport. When checking directories
6488 for a mailbox, to see if any needed to be created, it was accidentally
6489 using path names with one or more superfluous leading slashes; tracing
6490 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6492 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6493 contents are added to the log line that is written for every discarded
6494 recipient. (Previously a log_message setting was ignored.)
6496 36. The ${quote: operator now quotes the string if it is empty.
6498 37. The install script runs exim in order to find its version number. If for
6499 some reason other than non-existence or emptiness, which it checks, it
6500 could not run './exim', it was installing it with an empty version number,
6501 i.e. as "exim-". This error state is now caught, and the installation is
6504 38. An argument was missing from the function that creates an error message
6505 when Exim fails to connect to the socket for saslauthd authentication.
6506 This could cause Exim to crash, or give a corrupted message.
6508 39. Added isip, isip4, and isip6 to ${if conditions.
6510 40. The ACL variables $acl_xx are now saved with the message, and can be
6511 accessed later in routers, transports, and filters.
6513 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6514 strings in the file are not string-expanded.
6516 42. If a MAIL command specified a SIZE value that was too large to fit into an
6517 int variable, the check against message_size_limit failed. Such values are
6518 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6519 day this will have to be increased, but I don't think I want to be around
6520 when emails are that large.
6527 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6528 iconv() is not standard in FreeBSD.
6530 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6531 IPv6 enabled. The observed symptom was a segmentation fault on return from
6532 the function os_common_find_running_interfaces() in src/os.c.
6534 3. In the check_special_case() function in daemon.c I had used "errno" as an
6535 argument name, which causes warnings on some systems. This was basically a
6536 typo, since it was named "eno" in the comments!
6538 4. The code that waits for the clock to tick (at a resolution of some fraction
6539 of a second) so as to ensure message-id uniqueness was always waiting for
6540 at least one whole tick, when it could have waited for less. [This is
6541 almost certainly not relevant at current processor speeds, where it is
6542 unlikely to ever wait at all. But we try to future-proof.]
6544 5. The function that sleeps for a time interval that includes fractions of a
6545 second contained a race. It did not block SIGALRM between setting the
6546 timer, and suspending (a couple of lines later). If the interval was short
6547 and the sigsuspend() was delayed until after it had expired, the suspension
6548 never ended. On busy systems this could lead to processes getting stuck for
6551 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6552 process, before it forks any delivery processes. The open lookup caching
6553 mechanism meant that the open file or database connection was passed into
6554 the delivery process. The problem was that delivery processes always tidy
6555 up cached lookup data. This could cause a problem for the next delivery
6556 process started by the queue runner, because the external queue runner
6557 process does not know about the closure. So the next delivery process
6558 still has data in the lookup cache. In the case of a file lookup, there was
6559 no problem because closing a file descriptor in a subprocess doesn't affect
6560 the parent. However, if the lookup was caching a connection to a database,
6561 the connection was closed, and the second delivery process was likely to
6562 see errors such as "PGSQL: query failed: server closed the connection
6563 unexpectedly". The problem has been fixed by closing all cached lookups
6564 in a queue runner before running a delivery process.
6566 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6567 seem to have the "const" qualifier which it has on other OS. I've
6570 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6571 *of the same type* (client or server) with the same public name that an
6572 error should be diagnosed.
6574 9. When Exim looked up a host name for an IP address, but failed to find the
6575 original IP address when looking up the host name (a safety check), it
6576 output the message "<ip address> does not match any IP for NULL", which was
6577 confusing, to say the least. The bug was that the host name should have
6578 appeared instead of "NULL".
6580 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6581 user that is built into the binary, and the -C or -D options is used, root
6582 privilege is dropped before the configuration file is read. In addition,
6583 logging is switched to stderr instead of the normal log files. If the
6584 configuration then re-defines the Exim user, the unprivileged environment
6585 is probably not what is expected, so Exim logs a panic warning message (but
6588 However, if deliver_drop_privilege is set, the unprivileged state may well
6589 be exactly what is intended, so the warning has been cut out in that case,
6590 and Exim is allowed to try to write to its normal log files.
6596 1. smtp_return_error_details was not giving details for temporary sender
6597 or receiver verification errors.
6599 2. Diagnose a configuration error if two authenticators have the same public
6602 3. Exim used not to create the message log file for a message until the first
6603 delivery attempt. This could be confusing when incoming messages were held
6604 for policy or load reasons. The message log file is now created at the time
6605 the message is received, and an initial "Received" line is written to it.
6607 4. The automatically generated man page for command line options had a minor
6608 bug that caused no ill effects; however, a more serious problem was that
6609 the procedure for building the man page automatically didn't always
6610 operate. Consequently, release 4.20 contains an out-of-date version. This
6611 shouldn't happen again.
6613 5. When building Exim with embedded Perl support, the script that builds the
6614 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6615 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6617 6. The freeze_tell option was not being used for messages that were frozen on
6618 arrival, either by an ACL or by local_scan().
6620 7. Added the smtp_incomplete_transaction log selector.
6622 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6623 was accepting AUTH without a new EHLO.
6625 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6626 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6629 10. Logging of TCP/IP connections (when configured) now happens in the main
6630 daemon process instead of the child process, so that the TCP/IP connection
6631 count is more accurate (but it can never be perfect).
6633 11. The use of "drop" in a nested ACL was not being handled correctly in the
6634 outer ACL. Now, if condition failure induced by the nested "drop" causes
6635 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6636 or "require"), the connection is dropped.
6638 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6639 that yield "discard" can now be used with an "accept" or a "discard" verb,
6640 but an error is generated for any others (because I can't see a useful way
6641 to define what should happen).
6643 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6644 are now processed in the same way as lines in the Exim configuration file.
6645 In particular, continuation lines are supported.
6647 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6649 15. Added -ti meaning -t -i.
6651 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6652 domains, and warn by logging if others are found.
6654 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6655 returned from GIFCONF when Exim is trying to find the list of interfaces on
6656 a host. The code in os.c has been modified to copy each ifreq to an aligned
6657 structure in all cases.
6659 Also, in some cases, the returned ifreq's were being copied to a 'struct
6660 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6661 means the last couple of bytes of an IPv6 address could be chopped if the
6662 ifreq contained only a normal sockaddr (14 bytes storage).
6664 18. Named domain lists were not supported in the hosts_treat_as_local option.
6665 An entry such as +xxxx was not recognized, and was treated as a literal
6668 19. Ensure that header lines added by a DATA ACL are included in the reject log
6669 if the ACL subsequently rejects the message.
6671 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6672 MD5 (which is deprecated).
6674 21. When testing a filter file using -bf, Exim was writing a message when it
6675 took the sender from a "From " line in the message, but it was not doing so
6676 when it took $return_path from a Return-Path: header line. It now does.
6678 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6679 with a valid header line field name (a series of printing characters
6680 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6682 23. Changed "disc" in the source to "disk" to conform to the documentation and
6683 the book and for uniformity.
6685 24. Ignore Sendmail's -Ooption=value command line item.
6687 25. When execve() failed while trying to run a command in a pipe transport,
6688 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6689 could be confused with a return value of 69 from the command itself. This
6690 has been changed to 127, the value the shell returns if it is asked to run
6691 a non-existent command. The wording for the related log line suggests a
6692 non-existent command as the problem.
6694 26. If received_header_text expands to an empty string, do not add a Received:
6695 header line to the message. (Well, it adds a token one on the spool, but
6696 marks it "old" so that it doesn't get used or transmitted.)
6698 27. Installed eximstats 1.28 (addition of -nt option).
6700 28. There was no check for failure on the call to getsockname() in the daemon
6701 code. This can fail if there is a shortage of resources on the system, with
6702 ENOMEM, for example. A temporary error is now given on failure.
6704 29. Contrary to the C standard, it seems that in some environments, the
6705 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6706 program. Exim now does this explicitly; it affects the formatting of
6707 timestamps using strftime().
6709 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6710 complaints. I've now initialized all the variables, to avoid this.
6712 32. Header lines added by a system filter were not being "seen" during
6713 transport-time rewrites.
6715 33. The info_callback() function passed to OpenSSL is set up with type void
6716 (*)(SSL *, int, int), as described somewhere. However, when calling the
6717 function (actually a macro) that sets it up, the type void(*)() is
6718 expected. I've put in a cast to prevent warnings from picky compilers.
6720 34. If a DNS black list lookup found a CNAME record, but there were no A
6721 records associated with the domain it pointed at, Exim crashed.
6723 35. If a DNS black list lookup returned more than one A record, Exim ignored
6724 all but the first. It now scans all returned addresses if a particular IP
6725 value is being sought. In this situation, the contents of the
6726 $dnslist_value variable are a list of all the addresses, separated by a
6729 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6730 to accept a host name and all its aliases if the forward lookup for any of
6731 them yielded the IP address of the incoming connection. Now it accepts only
6732 those names whose forward lookup yields the correct IP address. Any other
6733 names are discarded. This closes a loophole whereby a rogue DNS
6734 administrator could create reverse DNS records to break through a
6735 wildcarded host restriction in an ACL.
6737 37. If a user filter or a system filter that ran in a subprocess used any of
6738 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6739 the wrong values were passed to the pipe command ($thisaddress had the
6740 value of $0, $0 had the value of $1, etc). This bug was introduced by
6741 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6743 38. Improved the line breaking for long SMTP error messages from ACLs.
6744 Previously, if there was no break point between 40 and 75 characters, Exim
6745 left the rest of the message alone. Two changes have been made: (a) I've
6746 reduced the minimum length to 35 characters; (b) if it can't find a break
6747 point between 35 and 75 characters, it looks ahead and uses the first one
6748 that it finds. This may give the occasional overlong line, but at least the
6749 remaining text gets split now.
6751 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6752 file descriptors might be low, and that setting 1000 would always raise it.
6753 It turns out that in some environments, the limit is already over 1000 and
6754 that lowering it causes trouble. So now Exim takes care not to decrease it.
6756 40. When delivering a message, the value of $return_path is set to $sender_
6757 address at the start of routing (routers may change the value). By an
6758 oversight, this default was not being set up when an address was tested by
6759 -bt or -bv, which affected the outcome if any router or filter referred to
6762 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6763 log only when "message" or "log_message" are set. However, if one of the
6764 conditions was an address verification, or a call to a nested ACL, the
6765 messages generated by the underlying test were being passed through. This
6766 no longer happens. The underlying message is available in $acl_verify_
6767 message for both "message" and "log_message" expansions, so it can be
6768 passed through if needed.
6770 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6771 new expansion $bh_ to give the encoded byte string without charset
6772 translation. Translation happens only if iconv() is available; HAVE_ICONV
6773 indicates this at build time. HEADERS_CHARSET gives the charset to
6774 translate to; headers_charset can change it in the configuration, and
6775 "headers charset" can change it in an individual filter file.
6777 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6778 that creates RFC 2047 encoded "words" labels them as that charset instead
6779 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6780 expansion operator; (ii) when Exim creates a From: line for a local
6781 message; (iii) when a header line is rewritten to include a "phrase" part.
6783 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6784 buggy, causing it to skip the first lines of messages whose message ID
6785 ended in 'D'. This would not have bitten before Exim release 4.14, because
6786 message IDs were unlikely to end in 'D' before then. The effect was to have
6787 incorrect size information for certain domains.
6789 45. #include "config.h" was missing at the start of the crypt16.c module. This
6790 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6793 46. If there was a timeout during a "random" callout check, Exim treated it as
6794 a failure of the random address, and carried on sending RSET and the real
6795 address. If the delay was just some slowness somewhere, the response to the
6796 original RCPT would be taken as a response to RSET and so on, causing
6797 mayhem of various kinds.
6799 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6800 when I implemented it. It didn't allow for the fact that some option values
6801 may legitimately be negative (e.g. size_addition), and it didn't even do
6802 the right test for positive values.
6804 48. Domain names in DNS records are case-independent. Exim always looks them up
6805 in lower case. Some resolvers return domain names in exactly the case they
6806 appear in the zone file, that is, they may contain uppercase letters. Not
6807 all resolvers do this - some return always lower case. Exim was treating a
6808 change of case by a resolver as a change of domain, similar to a widening
6809 of a domain abbreviation. This triggered its re-routing code and so it was
6810 trying to route what was effectively the same domain again. This normally
6811 caused routing to fail (because the router wouldn't handle the domain
6812 twice). Now Exim checks for this case specially, and just changes the
6813 casing of the domain that it ultimately uses when it transmits the message
6816 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6819 50. If a filter generated a file delivery with a non-absolute name (possible if
6820 no home directory exists for the router), the forbid_file option was not
6823 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6824 the existing equality matching.
6826 52. Exim was using ints instead of ino_t variables in some places where it was
6827 dealing with inode numbers.
6829 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6830 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6831 called TMPDIR, and if it finds it is different, it changes its value.
6833 54. The smtp_printf() function is now made available to local_scan() so
6834 additional output lines can be written before returning. There is also an
6835 smtp_fflush() function to enable the detection of a dropped connection.
6836 The variables smtp_input and smtp_batched_input are exported to
6839 55. Changed the default runtime configuration: the message "Unknown user"
6840 has been removed from the ACL, and instead placed on the localuser router,
6841 using the cannot_route_message feature. This means that any verification
6842 failures that generate their own messages won't get overridden. Similarly,
6843 the "Unrouteable address" message that was in the ACL for unverifiable
6844 relay addresses has also been removed.
6846 56. Added hosts_avoid_esmtp to the smtp transport.
6848 57. The exicyclog script was not checking for the esoteric option
6849 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6850 will work only if exicyclog is run under the appropriate euid.
6852 58. Following a discussion on the list, the rules by which Exim recognises line
6853 endings on incoming messages have been changed. The -dropcr and drop_cr
6854 options are now no-ops, retained only for backwards compatibility. The
6855 following line terminators are recognized: LF CRLF CR. However, special
6856 processing applies to CR:
6858 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6859 nor a local message in the state where . is a terminator.
6861 (ii) If a bare CR is encountered in a header line, an extra space is added
6862 after the line terminator so as not to end the header. The reasoning
6863 behind this is that bare CRs in header lines are most likely either
6864 to be mistakes, or people trying to play silly games.
6866 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6867 was being incorrectly given as 18 bytes larger than it should have been.
6868 This is a VOB (very old bug).
6870 60. This may never have affected anything current, but just in case it has:
6871 When the local host is found other than at the start of a list of hosts,
6872 the local host, those with the same MX, and any that follow, are discarded.
6873 When the list in question was part of a longer list of hosts, the following
6874 hosts (not currently being processed) were also being discarded. This no
6875 longer happens. I'm not sure if this situation could ever has previously
6878 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6881 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6882 header. This is something that is recommended in a new Internet Draft, and
6883 is something that is documented as being done by Sendmail. There are two
6884 possible values. For messages generated by the autoreply transport, Exim
6887 Auto-Submitted: auto-replied
6889 whereas for all other generated messages (e.g. bounces) it adds
6891 Auto-Submitted: auto-generated
6893 63. The "personal" condition in filters now includes a test for the
6894 Auto-Submitted: header. If it contains the string "auto-" the message it
6895 not considered personal.
6897 64. Added rcpt_include_affixes as a generic transport option.
6899 65. Added queue_only_override (default true).
6901 66. Added the syslog_duplication option.
6903 67. If what should have been the first header line of a message consisted of
6904 a space followed by a colon, Exim was mis-interpreting it as a header line.
6905 It isn't of course - it is syntactically invalid and should therefore be
6906 treated as the start of the message body. The misbehaviour could have
6907 caused a number of strange effects, including loss of data in subsequent
6908 header lines, and spool format errors.
6910 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6911 client host had authenticated. This control can now be exercised by an ACL
6912 for more flexibility.
6914 69. By default, callouts do not happen when testing with -bh. There is now a
6915 variant, -bhc, which does actually run the callout code, including
6916 consulting and updating the callout cache.
6918 70. Added support for saslauthd authentication, courtesy of Alexander
6921 71. If statvfs() failed on the spool or log directories while checking their
6922 size for availability, Exim confusingly gave the error "space shortage".
6923 Furthermore, in debugging mode it crashed with a floating point exception.
6924 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6925 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6926 serious problem, Exim now writes to the main and panic logs when this
6927 happens, with details of the failure. It then refuses to accept the
6928 incoming message, giving the message "spool directory problem" or "log
6929 directory problem" with a 421 code for SMTP messages.
6931 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6932 0, 1, and 2 exist, because some OS complain for execs without them (see
6933 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6934 descriptors. However, the code omitted to check that the open succeeded,
6935 causing mysterious errors if for some reason the permissions on /dev/null
6936 got screwed. Now Exim writes a message to the main and panic logs, and
6937 bombs out if it can't open /dev/null.
6939 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6940 interact so that it is all more flexible. It is supposed to remain
6941 backwards compatible. Also added extra_local_interfaces.
6943 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6944 to bomb out with an assertion failure - to the client this appears as a
6945 connection drop. This problem occurs in the part of the code that was taken
6946 from the Samba project. Fortunately, the assertion is in a very simple
6947 function, so I have fixed this by reproducing the function inline in the
6948 one place where it is called, and arranging for authentication to fail
6949 instead of killing the process with assert().
6951 75. The SPA client code was not working when the server requested OEM rather
6952 than Unicode encoding.
6954 76. Added code to make require_files with a specific uid setting more usable in
6955 the case where statting the file as root fails - usually a non-root-mounted
6956 NFS file system. When this happens and the failure is EACCES, Exim now
6957 forks a subprocess and does the per-uid checking as the relevant uid.
6959 77. Added process_log_path.
6961 78. If log_file_path was not explicitly set, a setting of check_log_space or
6962 check_log_inodes was ignored.
6964 79. If a space check for the spool or log partitions fails, the incident is now
6965 logged. Of course, in the latter case the data may get lost...
6967 80. Added the %p formatting code to string_format() so that it can be used to
6968 print addresses in debug_print(). Adjusted all the address printing in the
6969 debugging in store.c to use %p rather than %d.
6971 81. There was a concern that a line of code in smtp_in.c could overflow a
6972 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6973 initially expressed, the concern was not well-founded, because trailing
6974 spaces are removed early. However, if the trailing spaces were followed by
6975 a NULL, they did not get removed, so the overflow was possible. Two fixes
6978 (a) I re-wrote the offending code in a cleaner fashion.
6979 (b) If an incoming SMTP command contains a NULL character, it is rejected
6982 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6983 runs initgroups(), so that if the Exim user is in any additional groups,
6984 they will be used during message reception.
6990 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob