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.
11 JH/01 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
40 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
41 Disable the session-cache too, which might reduce our load. Since we
42 currrectly use a new context for every connection, both as server and
43 client, there is no benefit for these.
44 GnuTLS appears to not support tickets server-side by default (we don't
45 call gnutls_session_ticket_enable_server()) but client side is enabled
46 by default on recent versions (3.1.3 +) unless the PFS priority string
49 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
50 <https://reproducible-builds.org/specs/source-date-epoch/>.
52 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
53 the check for any unsuccessful recipients did not notice the limit, and
54 erroneously found still-pending ones.
56 JH/08 Pipeline CHUNKING command and data together, on kernels that support
57 MSG_MORE. Only in-clear (not on TLS connections).
59 JH/09 Avoid using a temporary file during transport using dkim. Unless a
60 transport-filter is involved we can buffer the headers in memory for
61 creating the signature, and read the spool data file once for the
62 signature and again for transmission.
64 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
65 4.77 as the kernel support then wasn't solid, having issues in 64bit
66 mode. Now, it's been long enough. Add support for FreeBSD also.
68 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
69 case where the routing stage had gathered several addresses to send to
70 a host before calling the transport for the first, we previously failed
71 to close down TLS in the old transport process before passing the TCP
72 connection to the new process. The new one sent a STARTTLS command
73 which naturally failed, giving a failed delivery and bloating the retry
74 database. Investigation and fix prototype from Wolfgang Breyha.
76 JH/12 Fix check on SMTP command input synchronisation. Previously there were
77 false-negatives in the check that the sender had not preempted a response
78 or prompt from Exim (running as a server), due to that code's lack of
79 awareness of the SMTP input buffering.
81 PP/04 Add commandline_checks_require_admin option.
82 Exim drops privileges sanely, various checks such as -be aren't a
83 security problem, as long as you trust local users with access to their
84 own account. When invoked by services which pass untrusted data to
85 Exim, this might be an issue. Set this option in main configuration
86 AND make fixes to the calling application, such as using `--` to stop
89 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
90 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
91 responses to those, into a single TLS record each way (this usually means
92 a single packet). As a side issue, smtp_enforce_sync now works on TLS
95 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
96 affects you only if you're dancing at the edge of the param size limits.
97 If you are, and this message makes sense to you, then: raise the
98 configured limit or use OpenSSL 1.1. Nothing we can do for older
101 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
102 cmdline element to get the documented default one. Previously it was
105 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
106 are unsuable for non-message-specific reasons - eg. connection timeout,
107 banner-time rejection.
109 JH/16 Fix logging of delivery remote port, when specified by router, under
112 PP/06 Repair manualroute's ability to take options in any order, even if one
113 is the name of a transport.
116 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
118 JH/17 Change the list-building routines interface to use the expanding-string
119 triplet model, for better allocation and copying behaviour.
121 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
122 Previously it was constructed the first time a possibly-matching string
123 was met in the configuration file input during startup; now it is done
126 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
127 compatible one, to avoid the (poorly documented) possibility of a config
128 file in the working directory redirecting the DB files, possibly correpting
135 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
136 than -2003 did; needs libidn2 in addition to libidn.
138 JH/02 The path option on a pipe transport is now expanded before use.
140 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
141 Patch provided by "Björn", documentation fix added too.
143 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
144 missing a wire-to-host endian conversion.
146 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
147 close after a BDAT command line could be taken as a following command,
148 giving a synch failure. Fix by only checking for synch immediately
149 before acknowledging the chunk.
151 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
152 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
153 macro. Patches provided by Josh Soref.
155 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
156 Previously we did not; the RFC seems ambiguous and VRFY is not listed
157 by IANA as a service extension. However, John Klensin suggests that we
160 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
161 the dkim code may be unix-mode line endings rather than smtp wire-format
162 CRLF, so prepend a CR to any bare LF.
164 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
165 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
167 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
168 the first were themselves being wrongly included in the feed into dkim
169 processing; with most chunk sizes in use this resulted in an incorrect
170 body hash calculated value.
172 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
173 DKIM signature block, for verification. Although advised against by
174 standards it is specifically not ruled illegal.
176 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
178 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
179 missing a body hash (the bh= tag).
181 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
182 It seems that HAProxy sends the Proxy Protocol information in clear and
183 only then does a TLS startup, so do the same.
185 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
186 TCP connections (such as for Spamd) unless the daemon successfully set
187 Fast Open mode on its listening sockets. This fixes breakage seen on
188 too-old kernels or those not configured for Fast Open, at the cost of
189 requiring both directions being enabled for TFO, and TFO never being used
190 by non-daemon-related Exim processes.
192 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
193 endings, at least on the first header line. Try to canonify any that get
194 past that check, despite the cost.
196 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
197 now limited to an arbitrary five deep, while parsing addresses with the
198 strip_excess_angle_brackets option enabled.
200 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
201 instead leave the unprompted TLS handshake in socket buffer for the
202 TLS library to consume.
204 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
206 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
208 JH/16 Drop variables when they go out of scope. Memory management drops a whole
209 region in one operation, for speed, and this leaves assigned pointers
210 dangling. Add checks run only under the testsuite which checks all
211 variables at a store-reset and panics on a dangling pointer; add code
212 explicitly nulling out all the variables discovered. Fixes one known
213 bug: a transport crash, where a dangling pointer for $sending_ip_address
214 originally assigned in a verify callout, is re-used.
216 PP/06 Drop '.' from @INC in various Perl scripts.
218 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
220 PP/08 Reduce a number of compilation warnings under clang; building with
221 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
222 should be warning-free.
224 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
226 HS/01 Fix portability problems introduced by PP/08 for platforms where
227 realloc(NULL) is not equivalent to malloc() [SunOS et al].
229 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
230 chunk. This allows us to accept broken chunked messages. We need a more
231 general solution here.
233 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
234 already-broken messages in the queue.
236 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
238 JH/19 Fix reference counting bug in routing-generated-address tracking.
244 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
245 supports it and a size is available (ie. the sending peer gave us one).
247 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
248 years of being deprecated). The replacements are the ACLs
249 acl_smtp_mime and acl_not_smtp_mime.
251 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
252 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
253 or even an in-clear connection were permitted. Now, if the host lookup
254 was dnssec and dane was requested then the host is only used if the
255 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
256 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
257 if one fails this test.
258 This means that a poorly-configured remote DNS will make it incommunicado;
259 but it protects against a DNS-interception attack on it.
261 JH/04 Bug 1810: make continued-use of an open smtp transport connection
262 non-noisy when a race steals the message being considered.
264 JH/05 If main configuration option tls_certificate is unset, generate a
265 self-signed certificate for inbound TLS connections.
267 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
268 in rewrites and routers.
270 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
271 and logged a warning sing 4.83; now they are a configuration file error.
273 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
274 (lacking @domain). Apply the same qualification processing as RCPT.
276 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
278 JH/10 Support ${sha256:} applied to a string (as well as the previous
281 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
282 a cutthrough deliver is pending, as we always want to make a connection.
283 This also avoids re-routing the message when later placing the cutthrough
284 connection after a verify cache hit.
285 Do not update it with the verify result either.
287 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
288 when routing results in more than one destination address.
290 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
291 signing (which inhibits the cutthrough capability). Previously only
292 the presence of an option was tested; now an expansion evaluating as
293 empty is permissible (obviously it should depend only on data available
294 when the cutthrough connection is made).
296 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
297 the relevant preceding SMTP command did not note the pipelining mode.
299 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
300 Previously they were not counted.
302 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
303 as one having no matching records. Previously we deferred the message
304 that needed the lookup.
306 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
307 distinguished as "(=".
309 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
310 for missing MX records. Previously it only worked for missing A records.
312 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
314 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
315 after the data-go-ahead and data-ack. Patch from Jason Betts.
317 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
318 even for a "none" policy. Patch from Tony Meyer.
320 JH/22 Fix continued use of a connection for further deliveries. If a port was
321 specified by a router, it must also match for the delivery to be
324 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
325 When one of the recipients of a message was unsuitable for the connection
326 (has no matching addresses), we lost track of needing to mark it
327 deferred. As a result mail would be lost.
329 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
331 JH/25 Decoding ACL controls is now done using a binary search; the source code
332 takes up less space and should be simpler to maintain. Merge the ACL
333 condition decode tables also, with similar effect.
335 JH/26 Fix problem with one_time used on a redirect router which returned the
336 parent address unchanged. A retry would see the parent address marked as
337 delivered, so not attempt the (identical) child. As a result mail would
340 JH/27 Fix a possible security hole, wherein a process operating with the Exim
341 UID can gain a root shell. Credit to http://www.halfdog.net/ for
342 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
345 JH/28 Enable {spool,log} filesystem space and inode checks as default.
346 Main config options check_{log,spool}_{inodes,space} are now
347 100 inodes, 10MB unless set otherwise in the configuration.
349 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
350 Previously it only applied to the main-section connection policy
353 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
355 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
356 by me. Added RFC7919 DH primes as an alternative.
358 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
359 are not in the system include path.
361 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
362 GnuTLS, when a session startup failed (eg because the client disconnected)
363 Exim did stdio operations after fclose. This was exposed by a recent
364 change which nulled out the file handle after the fclose.
366 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
367 signed directly by the cert-signing cert, rather than an intermediate
368 OCSP-signing cert. This is the model used by LetsEncrypt.
370 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
372 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
373 an incoming connection.
375 HS/02 Bug 1802: Do not half-close the connection after sending a request
378 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
379 fallback to "prime256v1".
381 JH/34 SECURITY: Use proper copy of DATA command in error message.
382 Could leak key material. Remotely exploitable. CVE-2016-9963.
388 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
389 and 3.4.4 - once the server is enabled to respond to an OCSP request
390 it does even when not requested, resulting in a stapling non-aware
391 client dropping the TLS connection.
393 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
394 support variable-length bit vectors. No functional change.
396 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
397 The I= interface field on outgoing lines is now after the H= remote
398 host field, same as incoming lines. There is a separate
399 outgoing_interface log selector which allows you to disable the
402 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
403 If not running log_selector +smtp_connection the mainlog would be held
404 open indefinitely after a "too many connections" event, including to a
405 deleted file after a log rotate. Leave the per net connection logging
406 leaving it open for efficiency as that will be quickly detected by the
407 check on the next write.
409 HS/01 Bug 1671: Fix post transport crash.
410 Processing the wait-<transport> messages could crash the delivery
411 process if the message IDs didn't exist for some reason. When
412 using 'split_spool_directory=yes' the construction of the spool
413 file name failed already, exposing the same netto behaviour.
415 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
416 mime_regex ACL conditions.
418 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
419 to DSN fail messages (bounces): remote IP, remote greeting, remote response
420 to HELO, local diagnostic string.
422 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
423 log line to debug. Even when configured with a tls authenticator many
424 client connections are expected to not authenticate in this way, so
425 an authenticate fail is not an error.
427 HS/02 Add the Exim version string to the process info. This way exiwhat
428 gives some more detail about the running daemon.
430 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
431 matter for fast-change records such as DNSBLs.
433 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
434 retry record, even if constant. There may be multiple transports with
435 different interface settings and the retry behaviour needs to be kept
438 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
440 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
442 JH/10 Bug 840: fix log_defer_output option of pipe transport
444 JH/11 Bug 830: use same host for all RCPTS of a message, even under
445 hosts_randomize. This matters a lot when combined with mua_wrapper.
447 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
448 ${quote_pgsql:<string>} operator.
450 JH/13 Bug 1708: avoid misaligned access in cached lookup.
452 JH/14 Change header file name for freeradius-client. Relevant if compiling
453 with Radius support; from the Gentoo tree and checked under Fedora.
455 JH/15 Bug 1712: Introduce $prdr_requested flag variable
457 JH/16 Bug 1714: Permit an empty string as expansion result for transport
458 option transport_filter, meaning no filtering.
460 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
462 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
463 defaults to "*" (all hosts). The variable is now available when not built
464 with TLS, default unset, mainly to enable keeping the testsuite sane.
465 If a server certificate is not supplied (via tls_certificate) an error is
466 logged, and clients will find TLS connections fail on startup. Presumably
467 they will retry in-clear.
468 Packagers of Exim are strongly encouraged to create a server certificate
469 at installation time.
471 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
472 with the $config_file variable.
474 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
475 in transport context, after the attempt, and per-recipient. The latter type
476 is per host attempted. The event data is the error message, and the errno
477 information encodes the lookup type (A vs. MX) used for the (first) host,
478 and the trailing two digits of the smtp 4xx response.
480 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
481 to write to mainlog (or rejectlog, paniclog) in the window between file
482 creation and permissions/ownership being changed. Particularly affects
483 installations where exicyclog is run as root, rather than exim user;
484 result is that the running daemon panics and dies.
486 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
488 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
489 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
490 "pri" and "weight". Note that the previous implicit priority given by the
491 list order is no longer honoured.
493 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
496 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
497 by defining SUPPORT_SOCKS.
499 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
500 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
501 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
502 variables are renamed to proxy_{local,external}_{address,port}.
504 JH/27 Move Internationalisation support from Experimental to mainline, enabled
505 for a build by defining SUPPORT_I18N
507 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
508 of the query string, and make ${quote_redis:} do that quoting.
510 JH/29 Move Events support from Experimental to mainline, enabled by default
511 and removable for a build by defining DISABLE_EVENT.
513 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
515 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
516 cached by the daemon.
518 JH/32 Move Redis support from Experimental to mainline, enabled for a build
519 by defining LOOKUP_REDIS. The libhiredis library is required.
521 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
522 keys are given for lookup.
524 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
525 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
526 only supported when built with TLS support. The PolarSSL SHA routines
527 are still used when the TLS library is too old for convenient support.
529 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
530 openssl_options), for security. OpenSSL forces this from version 1.1.0
531 server-side so match that on older versions.
533 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
534 allocation for $value could be released as the expansion processing
535 concluded, but leaving the global pointer active for it.
537 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
538 and to use the domains and local_parts ACL conditions.
540 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
541 incorrectly not doubled on cutthrough transmission, hence seen as a
542 body-termination at the receiving system - resulting in truncated mails.
543 Commonly the sender saw a TCP-level error, and retransmitted the message
544 via the normal store-and-forward channel. This could result in duplicates
545 received - but deduplicating mailstores were liable to retain only the
546 initial truncated version.
548 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
550 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
552 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
553 we're in there, support oversigning also; bug 1309.
555 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
557 HS/04 Add support for keep_environment and add_environment options.
559 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
560 either intentional arithmetic overflow during PRNG, or testing config-
563 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
564 delivery resulted in actual delivery. Cancel cutthrough before DATA
567 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
568 rejects a recipient: pass the reject to the originator.
570 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
571 Many were false-positives and ignorable, but it's worth fixing the
574 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
575 for the new environment-manipulation done at startup. Move the routines
576 from being local to tls.c to being global via the os.c file.
578 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
579 an extract embedded as result-arg for a map, the first arg for extract
580 is unavailable so we cannot tell if this is a numbered or keyed
581 extraction. Accept either.
587 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
590 JH/02 The smtp transport option "multi_domain" is now expanded.
592 JH/03 The smtp transport now requests PRDR by default, if the server offers
595 JH/04 Certificate name checking on server certificates, when exim is a client,
596 is now done by default. The transport option tls_verify_cert_hostnames
597 can be used to disable this per-host. The build option
598 EXPERIMENTAL_CERTNAMES is withdrawn.
600 JH/05 The value of the tls_verify_certificates smtp transport and main options
601 default to the word "system" to access the system default CA bundle.
602 For GnuTLS, only version 3.0.20 or later.
604 JH/06 Verification of the server certificate for a TLS connection is now tried
605 (but not required) by default. The verification status is now logged by
606 default, for both outbound TLS and client-certificate supplying inbound
609 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
612 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
613 Status Notification (bounce) messages are now MIME format per RFC 3464.
614 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
615 under the control of the dsn_advertise_hosts option, and routers may
616 have a dsn_lasthop option.
618 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
619 default, modifiable by a malware= option. The list separator for
620 the options can now be changed in the usual way. Bug 68.
622 JH/10 The smtp_receive_timeout main option is now expanded before use.
624 JH/11 The incoming_interface log option now also enables logging of the
625 local interface on delivery outgoing connections.
627 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
628 if the interface and destination host and port all match.
630 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
633 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
634 Patch from Andrew Lewis.
636 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
637 now supports optional time-restrictions, weighting, and priority
638 modifiers per server. Patch originally by <rommer@active.by>.
640 JH/16 The spamd_address main option now supports a mixed list of local
641 and remote servers. Remote servers can be IPv6 addresses, and
642 specify a port-range.
644 JH/17 Bug 68: The spamd_address main option now supports an optional
645 timeout value per server.
647 JH/18 Bug 1581: Router and transport options headers_add/remove can
648 now have the list separator specified.
650 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
653 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
656 JH/21 Support for the A6 type of dns record is withdrawn.
658 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
659 rather than the verbs used.
661 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
662 from 255 to 1024 chars.
664 JH/24 Verification callouts now attempt to use TLS by default.
666 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
667 are generic router options now. The defaults didn't change.
669 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
670 Original patch from Alexander Shikoff, worked over by JH.
672 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
673 files if ZCAT_COMMAND is not executable.
675 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
677 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
679 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
680 Normally benign, it bites when the pair was led to by a CNAME;
681 modern usage is to not canonicalize the domain to a CNAME target
682 (and we were inconsistent anyway for A-only vs AAAA+A).
684 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
686 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
687 when evaluating $sender_host_dnssec.
689 JH/31 Check the HELO verification lookup for DNSSEC, adding new
690 $sender_helo_dnssec variable.
692 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
694 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
696 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
698 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
699 documented as working, but never had. Support all but $spam_report.
701 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
702 added for tls authenticator.
704 HS/03 Add perl_taintmode main config option
710 TL/01 When running the test suite, the README says that variables such as
711 no_msglog_check are global and can be placed anywhere in a specific
712 test's script, however it was observed that placement needed to be near
713 the beginning for it to behave that way. Changed the runtest perl
714 script to read through the entire script once to detect and set these
715 variables, reset to the beginning of the script, and then run through
716 the script parsing/test process like normal.
718 TL/02 The BSD's have an arc4random API. One of the functions to induce
719 adding randomness was arc4random_stir(), but it has been removed in
720 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
721 function when detected.
723 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
724 cause callback expansion.
726 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
727 syntax errors in an expansion can be treated as a string instead of
728 logging or causing an error, due to the internal use of bool_lax
729 instead of bool when processing it.
731 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
732 server certificates when making smtp deliveries.
734 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
736 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
738 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
740 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
741 Merged patch from Sebastian Wiedenroth.
743 JH/05 Fix results-pipe from transport process. Several recipients, combined
744 with certificate use, exposed issues where response data items split
745 over buffer boundaries were not parsed properly. This eventually
746 resulted in duplicates being sent. This issue only became common enough
747 to notice due to the introduction of connection certificate information,
748 the item size being so much larger. Found and fixed by Wolfgang Breyha.
750 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
751 size buffer was used, resulting in syntax errors when an expansion
754 JH/07 Add support for directories of certificates when compiled with a GnuTLS
755 version 3.3.6 or later.
757 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
758 is EXPERIMENTAL_EVENT, the main-configuration and transport options
759 both become "event_action", the variables become $event_name, $event_data
760 and $event_defer_errno. There is a new variable $verify_mode, usable in
761 routers, transports and related events. The tls:cert event is now also
762 raised for inbound connections, if the main configuration event_action
765 TL/06 In test suite, disable OCSP for old versions of openssl which contained
766 early OCSP support, but no stapling (appears to be less than 1.0.0).
768 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
769 server certificate names available under the smtp transport option
770 "tls_verify_cert_hostname" now do not permit multi-component wildcard
773 JH/10 Time-related extraction expansions from certificates now use the main
774 option "timezone" setting for output formatting, and are consistent
775 between OpenSSL and GnuTLS compilations. Bug 1541.
777 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
778 encoded parameter in the incoming message. Bug 1558.
780 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
781 include certificate info, eximon was claiming there were spoolfile
784 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
786 JH/14 Log delivery-related information more consistently, using the sequence
787 "H=<name> [<ip>]" wherever possible.
789 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
790 are problematic for Debian distribution, omit them from the release
793 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
795 JH/16 Fix string representation of time values on 64bit time_t architectures.
798 JH/17 Fix a null-indirection in certextract expansions when a nondefault
799 output list separator was used.
804 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
805 checkers that were complaining about end of non-void function with no
808 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
809 This was a regression introduced in 4.83 by another bugfix.
811 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
813 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
814 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
820 TF/01 Correctly close the server side of TLS when forking for delivery.
822 When a message was received over SMTP with TLS, Exim failed to clear up
823 the incoming connection properly after forking off the child process to
824 deliver the message. In some situations the subsequent outgoing
825 delivery connection happened to have the same fd number as the incoming
826 connection previously had. Exim would try to use TLS and fail, logging
827 a "Bad file descriptor" error.
829 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
830 utilities have not been installed.
832 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
833 temporary space as the ACL may create new global variables.
835 TL/01 LDAP support uses per connection or global context settings, depending
836 upon the detected version of the libraries at build time.
838 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
839 to extract and use the src ip:port in logging and expansions as if it
840 were a direct connection from the outside internet. PPv2 support was
841 updated based on HAProxy spec change in May 2014.
843 JH/02 Add ${listextract {number}{list}{success}{fail}}.
845 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
846 Properly escape header and check for NULL return.
848 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
851 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
853 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
854 characters in header names, implemented as a verify condition.
855 Contributed by Michael Fischer v. Mollard.
857 TL/05 Rename SPF condition results err_perm and err_temp to standardized
858 results permerror and temperror. Previous values are deprecated but
859 still accepted. In a future release, err_perm and err_temp will be
860 completely removed, which will be a backward incompatibility if the
861 ACL tests for either of these two old results. Patch contributed by
862 user bes-internal on the mailing list.
864 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
866 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
867 selectors, in both main and reject logs.
869 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
872 JH/07 Add malware type "sock" for talking to simple daemon.
874 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
876 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
877 routers/transports under cutthrough routing.
879 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
880 numbers. Touch up "bool" conditional to keep the same definition.
882 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
884 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
886 JH/12 Expand items in router/transport headers_add or headers_remove lists
887 individually rather than the list as a whole. Bug 1452.
889 Required for reasonable handling of multiple headers_ options when
890 they may be empty; requires that headers_remove items with embedded
891 colons must have them doubled (or the list-separator changed).
893 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
894 view the policy declared in the DMARC record. Currently, $dmarc_status
895 is a combined value of both the record presence and the result of the
898 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
900 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
901 dnslookup router and the smtp transport (applying to the forward
904 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
905 of ldap servers used for a specific lookup. Patch provided by Heiko
908 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
909 New variable $lookup_dnssec_authenticated for observability.
911 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
912 Patch submitted by Lars Timman.
914 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
916 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
917 Requires trusted mode and valid format message id, aborts otherwise.
918 Patch contributed by Heiko Schlichting.
920 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
921 certextract with support for various fields. Bug 1358.
923 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
924 is requested by default, modifiable by smtp transport option
927 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
928 operate on certificate variables to give certificate fingerprints
929 Also new ${sha256:cert_variable}.
931 JH/23 The PRDR feature is moved from being Experimental into the mainline.
933 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
934 Christian Aistleitner.
936 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
938 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
939 file. Patch from Wolfgang Breyha.
941 JH/25 Expand the coverage of the delivery $host and $host_address to
942 client authenticators run in verify callout. Bug 1476.
944 JH/26 Port service names are now accepted for tls_on_connect_ports, to
945 align with daemon_smtp_ports. Bug 72.
947 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
948 support and error reporting did not work properly.
950 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
951 and is readable. Patch from Andrew Colin Kissa.
953 TL/14 Enhance documentation of ${run expansion and how it parses the
954 commandline after expansion, particularly in the case when an
955 unquoted variable expansion results in an empty value.
957 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
959 PP/02 Fix internal collision of T_APL on systems which support RFC3123
960 by renaming away from it. Addresses GH issue 15, reported by
963 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
965 TL/15 SECURITY: prevent double expansion in math comparison functions
966 (can expand unsanitized data). Not remotely exploitable.
973 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
975 PP/02 Make -n do something, by making it not do something.
976 When combined with -bP, the name of an option is not output.
978 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
981 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
982 $sender_host_name and config options to manage this, and basic check
985 PP/05 DSCP support for outbound connections and control modifier for inbound.
987 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
988 (Only plugin which currently uses this is kerberos4, which nobody should
989 be using, but we should make it available and other future plugins might
990 conceivably use it, even though it would break NAT; stuff *should* be
991 using channel bindings instead).
993 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
994 name; added for Sendmail compatibility; requires admin caller.
995 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
996 just ignore it); requires trusted caller.
997 Also parse but ignore: -Ac -Am -X<logfile>
1000 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1002 TL/02 Add +smtp_confirmation as a default logging option.
1004 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1005 Patch by Magnus Holmgren from 2007-02-20.
1007 TL/04 Bugzilla 1281 - Spec typo.
1008 Bugzilla 1283 - Spec typo.
1009 Bugzilla 1290 - Spec grammar fixes.
1011 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1013 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1015 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1016 mailing list by by Dmitry Isaikin.
1018 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1020 JH/02 Support "G" suffix to numbers in ${if comparisons.
1022 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1024 NM/01 Bugzilla 1197 - Spec typo
1025 Bugzilla 1196 - Spec examples corrections
1027 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1029 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1030 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1033 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1034 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1036 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1037 "acl {{name}{arg}...}", and optional args on acl condition
1040 JH/05 Permit multiple router/transport headers_add/remove lines.
1042 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1044 JH/07 Avoid using a waiting database for a single-message-only transport.
1045 Performance patch from Paul Fisher. Bugzilla 1262.
1047 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1050 JH/09 Add $headers_added variable, with content from use of ACL modifier
1051 add_header (but not yet added to the message). Bugzilla 199.
1053 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1054 Pulled from Bugzilla 817 by Wolfgang Breyha.
1056 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1058 (nb: this is the same fix as in Exim 4.80.1)
1060 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1063 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1066 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1068 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1069 advertise SMTP AUTH mechanism to us, instead of a generic
1070 protocol violation error. Also, make Exim more robust to bad
1071 data from the Dovecot auth socket.
1073 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1075 When a queue runner is handling a message, Exim first routes the
1076 recipient addresses, during which it prunes them based on the retry
1077 hints database. After that it attempts to deliver the message to
1078 any remaining recipients. It then updates the hints database using
1081 So if a recipient address works intermittently, it can get repeatedly
1082 deferred at routing time. The retry hints record remains fresh so the
1083 address never reaches the final cutoff time.
1085 This is a fairly common occurrence when a user is bumping up against
1086 their storage quota. Exim had some logic in its local delivery code
1087 to deal with this. However it did not apply to per-recipient defers
1088 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1090 This change adds a proper retry rule check during routing so that the
1091 final cutoff time is checked against the message's age. We only do
1092 this check if there is an address retry record and there is not a
1093 domain retry record; this implies that previous attempts to handle
1094 the address had the retry_use_local_parts option turned on. We use
1095 this as an approximation for the destination being like a local
1096 delivery, as in LMTP.
1098 I suspect this new check makes the old local delivery cutoff check
1099 redundant, but I have not verified this so I left the code in place.
1101 TF/02 Correct gecos expansion when From: is a prefix of the username.
1103 Test 0254 submits a message to Exim with the header
1107 When I ran the test suite under the user fanf2, Exim expanded
1108 the header to contain my full name, whereas it should have added
1109 a Resent-Sender: header. It erroneously treats any prefix of the
1110 username as equal to the username.
1112 This change corrects that bug.
1114 GF/01 DCC debug and logging tidyup
1115 Error conditions log to paniclog rather than rejectlog.
1116 Debug lines prefixed by "DCC: " to remove any ambiguity.
1118 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1120 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1121 Bug spotted by Jeremy Harris; was flawed since initial commit.
1122 Would have resulted in OCSP responses post-SNI triggering an Exim
1123 NULL dereference and crash.
1125 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1127 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1128 Bug detection, analysis and fix by Samuel Thibault.
1129 Bugzilla 1331, Debian bug #698092.
1131 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1133 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1134 Server implementation by Todd Lyons, client by JH.
1135 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1136 config variable "prdr_enable" controls whether the server
1137 advertises the facility. If the client requests PRDR a new
1138 acl_data_smtp_prdr ACL is called once for each recipient, after
1139 the body content is received and before the acl_smtp_data ACL.
1140 The client is controlled by both of: a hosts_try_prdr option
1141 on the smtp transport, and the server advertisement.
1142 Default client logging of deliveries and rejections involving
1143 PRDR are flagged with the string "PRDR".
1145 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1146 fclose(). Diagnosis by Todd Lyons.
1148 PP/17 Update configure.default to handle IPv6 localhost better.
1149 Patch by Alain Williams (plus minor tweaks).
1152 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1153 This is now consistent with GnuTLS, and is now documented: the
1154 previous undocumented portable approach to treating the option as
1155 unset was to force an expansion failure. That still works, and
1156 an empty string is now equivalent.
1158 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1159 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1160 not performing validation itself.
1162 PP/20 Added force_command boolean option to pipe transport.
1163 Patch from Nick Koston, of cPanel Inc.
1165 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1168 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1170 PP/21 Fix eximon continuous updating with timestamped log-files.
1171 Broken in a format-string cleanup in 4.80, missed when I repaired the
1172 other false fix of the same issue.
1173 Report and fix from Heiko Schlichting.
1176 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1177 Report from Prashanth Katuri.
1179 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1180 It's SecureTransport, so affects any MacOS clients which use the
1181 system-integrated TLS libraries, including email clients.
1183 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1184 using a MIME ACL for non-SMTP local injection.
1185 Report and assistance in diagnosis by Warren Baker.
1187 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1189 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1191 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1192 last id that failed so it may be referenced in subsequent ACL's.
1194 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1197 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1198 ldap library initialization, allowing self-signed CA's to be
1199 used. Also properly sets require_cert option later in code by
1200 using NULL (global ldap config) instead of ldap handle (per
1201 session). Bug diagnosis and testing by alxgomz.
1203 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1204 the src/util/ subdirectory.
1206 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1207 renamed to Transport Post Delivery Action by Jeremy Harris, as
1210 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1211 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1212 redis_servers = needs to be configured which will be used by the redis
1213 lookup. Patch from Warren Baker, of The Packet Hub.
1215 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1217 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1218 hostname or reverse DNS when processing a host list. Used suggestions
1219 from multiple comments on this bug.
1221 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1223 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1224 Missed a few lines, added it to make the runtest require no keyboard
1227 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1228 contains upper case chars. Make router use caseful_local_part.
1230 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1231 support when GnuTLS has been built with p11-kit.
1237 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1239 This, or similar/improved, will also be change PP/11 of 4.82.
1245 PP/01 Handle short writes when writing local log-files.
1246 In practice, only affects FreeBSD (8 onwards).
1247 Bugzilla 1053, with thanks to Dmitry Isaikin.
1249 NM/01 Bugzilla 949 - Documentation tweak
1251 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1254 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1256 PP/02 Implemented gsasl authenticator.
1258 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1260 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1261 `pkg-config foo` for cflags/libs.
1263 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1264 with rest of GSASL and with heimdal_gssapi.
1266 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1267 `pkg-config foo` for cflags/libs for the TLS implementation.
1269 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1270 properties get this fed in as external SSF. A number of robustness
1271 and debugging improvements to the cyrus_sasl authenticator.
1273 PP/08 cyrus_sasl server now expands the server_realm option.
1275 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1276 Patch by Jeremy Harris.
1278 PP/10 Added dbmjz lookup type.
1280 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1282 PP/12 MAIL args handles TAB as well as SP, for better interop with
1283 non-compliant senders.
1284 Analysis and variant patch by Todd Lyons.
1286 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1287 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1288 Patch from Dirk Mueller <dmueller@suse.com>
1290 PP/13 tls_peerdn now print-escaped for spool files.
1291 Observed some $tls_peerdn in wild which contained \n, which resulted
1292 in spool file corruption.
1294 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1295 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1296 or write after TLS renegotiation, which otherwise led to messages
1299 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1300 as a tracking header (ie: a signed header comes before the signature).
1301 Patch from Wolfgang Breyha.
1303 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1304 comma-sep list; embedded commas doubled.
1306 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1308 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1310 Report and patch from Dmitry Banschikov.
1312 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1313 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1314 failures. We appear to now support TLS1.1+ with Exim.
1316 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1317 lets Exim select keys and certificates based upon TLS SNI from client.
1318 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1319 before an outbound SMTP session. New log_selector, +tls_sni.
1321 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1322 NULL dereference. Report and patch from Alun Jones.
1324 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1325 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1326 Not seeing resolver debug output on NetBSD, but suspect this is a
1327 resolver implementation change.
1329 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1330 Left warnings. Added "eximon gdb" invocation mode.
1332 PP/21 Defaulting "accept_8bitmime" to true, not false.
1334 PP/22 Added -bw for inetd wait mode support.
1336 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1337 locate the relevant includes and libraries. Made this the default.
1339 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1340 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1342 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1343 This may cause build issues on older platforms.
1345 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1346 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1347 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1348 Added SNI support via GnuTLS too.
1349 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1351 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1353 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1354 Applied second patch from Janne, implementing suggestion to default
1355 multiple-strings-in-record handling to match SPF spec.
1357 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1359 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1360 read-only, out of scope).
1361 Patch from Wolfgang Breyha, report from Stuart Northfield.
1363 PP/29 Fix three issues highlighted by clang analyser static analysis.
1364 Only crash-plausible issue would require the Cambridge-specific
1365 iplookup router and a misconfiguration.
1366 Report from Marcin Mirosław.
1368 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1370 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1371 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1372 As part of this, removing so much warning spew let me fix some minor
1373 real issues in debug logging.
1375 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1376 assignment on my part. Fixed.
1378 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1379 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1380 Janne Snabb (who went above and beyond: thank you).
1382 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1383 string otherwise requires a connection and a bunch more work and it's
1384 relatively easy to get wrong. Should also expose TLS library linkage
1387 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1388 64-bit ${eval} (JH/03).
1390 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1391 GNU libc to support some of the 64-bit stuff, should not lead to
1392 conflicts. Defined before os.h is pulled in, so if a given platform
1393 needs to override this, it can.
1395 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1396 protection layer was required, which is not implemented.
1397 Bugzilla 1254, patch from Wolfgang Breyha.
1399 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1400 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1401 tls_dhparam take prime identifiers. Also unbreak combination of
1402 OpenSSL+DH_params+TLSSNI.
1404 PP/39 Disable SSLv2 by default in OpenSSL support.
1410 PP/01 Solaris build fix for Oracle's LDAP libraries.
1411 Bugzilla 1109, patch from Stephen Usher.
1413 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1415 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1418 TF/02 Fix a couple more cases where we did not log the error message
1419 when unlink() failed. See also change 4.74-TF/03.
1421 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1422 lock up or crash if it happened to be inside a call to libc when it
1423 got a SIGUSR1 from exiwhat.
1425 The SIGUSR1 handler appends the current process status to the process
1426 log which is later printed by exiwhat. It used to use the general
1427 purpose logging code to do this, but several functions it calls are
1428 not safe for signals.
1430 The new output code in the SIGUSR1 handler is specific to the process
1431 log, and simple enough that it's easy to inspect for signal safety.
1432 Removing some special cases also simplifies the general logging code.
1433 Removing the spurious timestamps from the process log simplifies
1436 TF/04 Improved ratelimit ACL condition.
1438 The /noupdate option has been deprecated in favour of /readonly which
1439 has clearer semantics. The /leaky, /strict, and /readonly update modes
1440 are mutually exclusive. The update mode is no longer included in the
1441 database key; it just determines when the database is updated. (This
1442 means that when you upgrade Exim will forget old rate measurements.)
1444 Exim now checks that the per_* options are used with an update mode that
1445 makes sense for the current ACL. For example, when Exim is processing a
1446 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1447 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1448 must specify per_mail/readonly. If you omit the update mode it defaults to
1449 /leaky where that makes sense (as before) or /readonly where required.
1451 The /noupdate option is now undocumented but still supported for
1452 backwards compatibility. It is equivalent to /readonly except that in
1453 ACLs where /readonly is required you may specify /leaky/noupdate or
1454 /strict/noupdate which are treated the same as /readonly.
1456 A useful new feature is the /count= option. This is a generalization
1457 of the per_byte option, so that you can measure the throughput of other
1458 aggregate values. For example, the per_byte option is now equivalent
1459 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1461 The per_rcpt option has been generalized using the /count= mechanism
1462 (though it's more complicated than the per_byte equivalence). When it is
1463 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1464 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1465 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1466 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1467 Note that using per_rcpt with a non-readonly update mode in more than
1468 one ACL will cause the recipients to be double-counted. (The per_mail
1469 and per_byte options don't have this problem.)
1471 The handling of very low rates has changed slightly. If the computed rate
1472 is less than the event's count (usually one) then this event is the first
1473 after a long gap. In this case the rate is set to the same as this event's
1474 count, so that the first message of a spam run is counted properly.
1476 The major new feature is a mechanism for counting the rate of unique
1477 events. The new per_addr option counts the number of different
1478 recipients that someone has sent messages to in the last time period. It
1479 behaves like per_rcpt if all the recipient addresses are different, but
1480 duplicate recipient addresses do not increase the measured rate. Like
1481 the /count= option this is a general mechanism, so the per_addr option
1482 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1483 example, measure the rate that a client uses different sender addresses
1484 with the options per_mail/unique=$sender_address. There are further
1485 details in the main documentation.
1487 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1489 TF/06 Removed a few PCRE remnants.
1491 TF/07 Automatically extract Exim's version number from tags in the git
1492 repository when doing development or release builds.
1494 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1495 Bugzilla 879. Patch from Paul Fisher.
1497 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1498 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1501 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1503 PP/05 Variable $av_failed, true if the AV scanner deferred.
1504 Bugzilla 1078. Patch from John Horne.
1506 PP/06 Stop make process more reliably on build failure.
1507 Bugzilla 1087. Patch from Heiko Schlittermann.
1509 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1510 Bugzilla 1089. Patch from Heiko Schlittermann.
1512 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1513 Bugzilla 1131. Patch from Holger Weiß.
1515 PP/09 Handle IPv6 addresses with SPF.
1516 Bugzilla 860. Patch from Wolfgang Breyha.
1518 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1520 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1523 PP/11 match_* no longer expand right-hand-side by default.
1524 New compile-time build option, EXPAND_LISTMATCH_RHS.
1525 New expansion conditions, "inlist", "inlisti".
1527 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1529 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1531 PP/14 fix log_write() format string regression from TF/03.
1532 Bugzilla 1152. Patch from Dmitry Isaikin.
1538 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1540 PP/02 Harmonised TLS library version reporting; only show if debugging.
1541 Layout now matches that introduced for other libraries in 4.74 PP/03.
1543 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1545 PP/04 New "dns_use_edns0" global option.
1547 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1550 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1551 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1553 TK/01 Updated PolarSSL code to 0.14.2.
1554 Bugzilla 1097. Patch from Andreas Metzler.
1556 PP/07 Catch divide-by-zero in ${eval:...}.
1557 Fixes bugzilla 1102.
1559 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1562 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1563 format-string attack -- SECURITY: remote arbitrary code execution.
1565 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1566 time unintentionally subject to list matching rules, letting the header
1567 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1568 arbitrary string expansion). This allowed for information disclosure.
1570 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1571 INT_MIN/-1 -- value coerced to INT_MAX.
1577 NM/01 Workaround for PCRE version dependency in version reporting
1580 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1581 This fixes portability to compilers other than gcc, notably
1582 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1584 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1585 makefiles for portability to HP-UX and POSIX correctness.
1587 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1588 Also via indented variable definition in the Makefile.
1589 (Debugging by Oliver Heesakkers).
1591 PP/02 Restore caching of spamd results with expanded spamd_address.
1592 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1594 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1595 Improves build reliability. Fix from: Frank Elsner
1597 NM/02 Fix wide character breakage in the rfc2047 coding
1598 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1600 NM/03 Allow underscore in dnslist lookups
1601 Fixes bug 1026. Patch from Graeme Fowler
1603 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1604 Code patches from Adam Ciarcinski of NetBSD.
1606 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1609 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1610 is logged, to avoid truncation. Patch from John Horne.
1612 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1613 Patch from Jakob Hirsch.
1615 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1616 SQL string expansion failure details.
1617 Patch from Andrey Oktyabrski.
1619 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1620 Patch from Simon Arlott.
1622 PP/09 New lookups functionality failed to compile on old gcc which rejects
1623 extern declarations in function scope.
1624 Patch from Oliver Fleischmann
1626 PP/10 Use sig_atomic_t for flags set from signal handlers.
1627 Check getgroups() return and improve debugging.
1628 Fixed developed for diagnosis in bug 927 (which turned out to be
1631 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1632 Patch from Mark Zealey.
1634 PP/12 Bugzilla 1056: Improved spamd server selection.
1635 Patch from Mark Zealey.
1637 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1638 Based on patch from Heiko Schlittermann.
1640 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1641 Patch from Uwe Doering, sign-off by Michael Haardt.
1643 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1644 variable declaration deep within a block. Bug and patch from
1647 PP/15 lookups-Makefile IRIX compatibility coercion.
1649 PP/16 Make DISABLE_DKIM build knob functional.
1651 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1652 Patch by Simon Arlott
1654 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1655 variable argument macros. Our copy now differs from upstream.
1661 TF/01 Failure to get a lock on a hints database can have serious
1662 consequences so log it to the panic log.
1664 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1665 controlled using the smtp_confirmation log selector.
1667 TF/03 Include the error message when we fail to unlink a spool file.
1669 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1670 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1671 for maintaining out-of-tree patches for some time.
1673 PP/01 Bugzilla 139: Documentation and portability issues.
1674 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1675 Handle per-OS dynamic-module compilation flags.
1677 PP/02 Let /dev/null have normal permissions.
1678 The 4.73 fixes were a little too stringent and complained about the
1679 permissions on /dev/null. Exempt it from some checks.
1680 Reported by Andreas M. Kirchwitz.
1682 PP/03 Report version information for many libraries, including
1683 Exim version information for dynamically loaded libraries. Created
1684 version.h, now support a version extension string for distributors
1685 who patch heavily. Dynamic module ABI change.
1687 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1688 privilege escalation vulnerability whereby the Exim run-time user
1689 can cause root to append content of the attacker's choosing to
1692 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1695 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1696 If dropping privileges for untrusted macros, we disabled normal logging
1697 on the basis that it would fail; for the Exim run-time user, this is not
1698 the case, and it resulted in successful deliveries going unlogged.
1699 Fixed. Reported by Andreas Metzler.
1705 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1706 only prepend for the Resent-* case. Fixes regression introduced in
1707 Exim 4.70 by NM/22 for Bugzilla 607.
1709 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1710 increasing numbers of administrators be bitten by this.
1712 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1714 PP/03 Bugzilla 994: added openssl_options main configuration option.
1716 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1718 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1720 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1722 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1723 without a peer certificate, leading to a segfault because of an
1724 assumption that peers always have certificates. Be a little more
1725 paranoid. Problem reported by Martin Tscholak.
1727 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1728 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1729 NB: ClamAV planning to remove STREAM in "middle of 2010".
1730 CL also introduces -bmalware, various -d+acl logging additions and
1731 more caution in buffer sizes.
1733 PP/09 Implemented reverse_ip expansion operator.
1735 PP/10 Bugzilla 937: provide a "debug" ACL control.
1737 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1739 PP/12 Bugzilla 973: Implement --version.
1741 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1743 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1745 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1747 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1748 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1749 ignore trailing whitespace.
1751 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1753 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1756 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1757 Notification from Dr Andrew Aitchison.
1759 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1760 ExtendedDetectionInfo response format.
1761 Notification from John Horne.
1763 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1766 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1767 XSL and documented dependency on system catalogs, with examples of how
1770 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1773 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1774 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1775 configuration file which is writeable by the Exim user or group.
1777 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1778 of configuration files to cover files specified with the -C option if
1779 they are going to be used with root privileges, not just the default
1782 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1783 option (effectively making it always true).
1785 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1786 files to be used while preserving root privileges.
1788 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1789 that rogue child processes cannot use them.
1791 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1792 run-time user, instead of root.
1794 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1795 Exim run-time user without dropping privileges.
1797 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1798 result string, instead of calling string_vformat() twice with the same
1801 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1802 for other users. Others should always drop root privileges if they use
1803 -C on the command line, even for a whitelisted configure file.
1805 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1807 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1813 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1814 $data_path, and $header_path variables; fixed documentation bugs and
1817 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1818 exipick to access non-standard spools, including the "frozen" queue
1821 NM/01 Bugzilla 965: Support mysql stored procedures.
1822 Patch from Alain Williams
1824 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1826 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1827 Patch from Andreas Metzler
1829 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1830 Patch from Kirill Miazine
1832 NM/05 Bugzilla 671: Added umask to procmail example.
1834 JJ/03 installed exipick 20100323.0, fixing doc bug
1836 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1837 directory. Notification and patch from Dan Rosenberg.
1839 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1841 TK/02 Improve log output when DKIM signing operation fails.
1843 MH/01 Treat the transport option dkim_domain as a colon separated
1844 list, not as a single string, and sign the message with each element,
1845 omitting multiple occurences of the same signer.
1847 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1848 Bugzilla 985, 986. Patch by Simon Arlott
1850 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1851 Patch by Simon Arlott
1853 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1854 MBX locking. Notification from Dan Rosenberg.
1860 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1862 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1864 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1866 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1868 NM/04 Bugzilla 915: Flag broken perl installation during build.
1874 TK/01 Added patch by Johannes Berg that expands the main option
1875 "spamd_address" if it starts with a dollar sign.
1877 TK/02 Write list of recipients to X-Envelope-Sender header when building
1878 the mbox-format spool file for content scanning (suggested by Jakob
1881 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1882 (http://www.dcc-servers.net/) support via dccifd. Activated by
1883 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1885 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1886 by Mark Daniel Reidel <mr@df.eu>.
1888 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1889 When building exim an external PCRE library is now needed -
1890 PCRE is a system library on the majority of modern systems.
1891 See entry on PCRE_LIBS in EDITME file.
1893 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1894 conversation. Added nologin parameter to request.
1895 Patch contributed by Kirill Miazine.
1897 TF/01 Do not log submission mode rewrites if they do not change the address.
1899 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1901 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1902 log files in place. Contributed by Roberto Lima.
1904 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1906 TF/03 Bugzilla 615: When checking the local_parts router precondition
1907 after a local_part_suffix or local_part_prefix option, Exim now
1908 does not use the address's named list lookup cache, since this
1909 contains cached lookups for the whole local part.
1911 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1912 Robert Millan. Documentation is in experimental-spec.txt.
1914 TF/04 Bugzilla 668: Fix parallel build (make -j).
1916 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1918 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1919 Patch provided by Jan Srzednicki.
1921 TF/05 Leading white space used to be stripped from $spam_report which
1922 wrecked the formatting. Now it is preserved.
1924 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1925 that they are available at delivery time.
1927 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1929 TF/08 TLS error reporting now respects the incoming_interface and
1930 incoming_port log selectors.
1932 TF/09 Produce a more useful error message if an SMTP transport's hosts
1933 setting expands to an empty string.
1935 NM/06 Bugzilla 744: EXPN did not work under TLS.
1936 Patch provided by Phil Pennock.
1938 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1939 Patch provided by Richard Godbee.
1941 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1942 acl_smtp_notquit, added index entry.
1944 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1945 Patch provided by Eugene Bujak.
1947 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1948 accept(). Patch provided by Maxim Dounin.
1950 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1951 Patch provided by Phil Pennock.
1953 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1955 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1956 Patch provided by Brad "anomie" Jorsch.
1958 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1959 Patch provided by Dean Brooks.
1961 TK/05 Add native DKIM support (does not depend on external libraries).
1963 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1964 Patch provided by Graeme Fowler.
1966 NM/16 Bugzilla 851: Documentation example syntax fix.
1968 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1970 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1973 NM/19 Bugzilla 745: TLS version reporting.
1974 Patch provided by Phil Pennock.
1976 NM/20 Bugzilla 167: bool: condition support.
1977 Patch provided by Phil Pennock.
1979 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1980 clients. Patch provided by Phil Pennock.
1982 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1983 Patch provided by Brad "anomie" Jorsch.
1985 NM/23 Bugzilla 687: Fix misparses in eximstats.
1986 Patch provided by Heiko Schlittermann.
1988 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1989 Patch provided by Heiko Schlittermann.
1991 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1992 plus update to original patch.
1994 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1996 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1997 Patch provided by David Brownlee.
1999 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2001 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2003 NM/30 Bugzilla 888: TLS documentation bugfixes.
2005 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2007 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2008 Unlike the original bugzilla I have changed all shell scripts in src tree.
2010 NM/33 Bugzilla 898: Transport filter timeout fix.
2011 Patch by Todd Rinaldo.
2013 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2014 Patch by Serge Demonchaux.
2016 NM/35 Bugzilla 39: Base64 decode bug fixes.
2017 Patch by Jakob Hirsch.
2019 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2021 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2023 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2025 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2031 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2032 ALT-N's libdkim that I have put here:
2033 http://duncanthrax.net/exim-experimental/
2035 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2036 were called 'true' and it seems that C99 defines that as a reserved
2037 keyword to be used with 'bool' variable types. That means you could
2038 not include C99-style headers which use bools without triggering
2039 build errors in sieve.c.
2041 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2042 as mailq or other aliases. Changed the --help handling significantly
2043 to do whats expected. exim_usage() emits usage/help information.
2045 SC/01 Added the -bylocaldomain option to eximstats.
2047 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2049 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2051 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2057 PH/01 Another patch from the Sieve maintainer.
2059 PH/02 When an IPv6 address is converted to a string for single-key lookup
2060 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2061 dots are used instead of colons so that keys in lsearch files need not
2062 contain colons. This was done some time before quoting was made available
2063 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2064 (notated using the quote facility) so as to distinguish them from IPv4
2065 keys. This meant that lookups for IP addresses in host lists did not work
2066 for iplsearch lookups.
2068 This has been fixed by arranging for IPv6 addresses to be expressed with
2069 colons if the lookup type is iplsearch. This is not incompatible, because
2070 previously such lookups could never work.
2072 The situation is now rather anomalous, since one *can* have colons in
2073 ordinary lsearch keys. However, making the change in all cases is
2074 incompatible and would probably break a number of configurations.
2076 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2079 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2080 conversion specification without a maximum field width, thereby enabling
2081 a rogue spamd server to cause a buffer overflow. While nobody in their
2082 right mind would setup Exim to query an untrusted spamd server, an
2083 attacker that gains access to a server running spamd could potentially
2084 exploit this vulnerability to run arbitrary code as the Exim user.
2086 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2087 $primary_hostname instead of what libspf2 thinks the hosts name is.
2089 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2090 a directory entry by the name of the lookup key. Previously, if a
2091 symlink pointed to a non-existing file or a file in a directory that
2092 Exim lacked permissions to read, a lookup for a key matching that
2093 symlink would fail. Now it is enough that a matching directory entry
2094 exists, symlink or not. (Bugzilla 503.)
2096 PH/03 The body_linecount and body_zerocount variables are now exported in the
2099 PH/04 Added the $dnslist_matched variable.
2101 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2102 This means they are set thereafter only if the connection becomes
2105 PH/06 Added the client_condition to authenticators so that some can be skipped
2106 by clients under certain conditions.
2108 PH/07 The error message for a badly-placed control=no_multiline_responses left
2109 "_responses" off the end of the name.
2111 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2113 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2114 (without spaces) instead of just copying the configuration text.
2116 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2118 PH/11 Added $max_received_linelength.
2120 PH/12 Added +ignore_defer and +include_defer to host lists.
2122 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2123 way in which PCRE > 7.0 is built.
2125 PH/14 Implemented queue_only_load_latch.
2127 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2128 MAIL command. The effect was to mangle the value on 64-bit systems.
2130 PH/16 Another patch from the Sieve maintainer.
2132 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2134 PH/18 If a system quota error occurred while trying to create the file for
2135 a maildir delivery, the message "Mailbox is full" was not appended to the
2136 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2137 only to a quota excession during the actual writing of the file.
2139 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2140 characters?) which causes problems in log lines. The DN values are now
2141 passed through string_printing() before being added to log lines.
2143 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2144 and InterBase are left for another time.)
2146 PH/21 Added message_body_newlines option.
2148 PH/22 Guard against possible overflow in moan_check_errorcopy().
2150 PH/23 POSIX allows open() to be a macro; guard against that.
2152 PH/24 If the recipient of an error message contained an @ in the local part
2153 (suitably quoted, of course), incorrect values were put in $domain and
2154 $local_part during the evaluation of errors_copy.
2160 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2161 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2164 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2165 issue a MAIL command.
2167 PH/02 In an ACL statement such as
2169 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2171 if a client was not listed at all, or was listed with a value other than
2172 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2173 the condition was not true (as it should be), so access was not denied.
2174 The bug was that the ! inversion was incorrectly passed on to the second
2175 item. This has been fixed.
2177 PH/03 Added additional dnslists conditions == and =& which are different from
2178 = and & when the dns lookup returns more than one IP address.
2180 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2181 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2183 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2184 FSYNC, which compiles an option called disable_fsync that allows for
2185 bypassing fsync(). The documentation is heavily laced with warnings.
2187 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2189 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2190 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2191 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2192 including adding "make clean"; (3) Added -fPIC when compiling the test
2193 dynamically loaded module, to get rid of a warning.
2195 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2196 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2197 = 0s. The bug is otherwise harmless.
2199 PH/07 There was a bug in the dovecot authenticator such that the value of
2200 $auth1 could be overwritten, and so not correctly preserved, after a
2201 successful authentication. This usually meant that the value preserved by
2202 the server_setid option was incorrect.
2204 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2206 PH/09 Installed PCRE release 7.0.
2208 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2209 run for batched SMTP input. It is now run at the start of every message
2210 in the batch. While fixing this I discovered that the process information
2211 (output by running exiwhat) was not always getting set for -bs and -bS
2212 input. This is fixed, and it now also says "batched" for BSMTP.
2214 PH/11 Added control=no_pipelining.
2216 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2217 patch, slightly modified), and move the expansion of helo_data till after
2218 the connection is made in the smtp transport (so it can use these
2221 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2223 PH/14 Added log_selector = +pid.
2225 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2227 PH/16 Add ${if forany and ${if forall.
2229 PH/17 Added dsn_from option to vary the From: line in DSNs.
2231 PH/18 Flush SMTP output before performing a callout, unless control =
2232 no_callout_flush is set.
2234 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2235 was true (the default) a successful delivery failed to delete the retry
2236 item, thus causing premature timeout of the address. The bug is now
2239 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2241 PH/21 Long custom messages for fakedefer and fakereject are now split up
2242 into multiline responses in the same way that messages for "deny" and
2243 other ACL rejections are.
2245 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2246 with slight modification.
2248 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2249 draft, changing the syntax and factoring some duplicate code".
2251 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2252 for deliveries of the second and subsequent messages over the same SMTP
2255 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2256 ${reduce, with only minor "tidies".
2258 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2260 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2261 expansion side effects.
2263 PH/27 When a message times out after an over-quota error from an Exim-imposed
2264 quota, the bounce message says "mailbox is full". This message was not
2265 being given when it was a system quota that was exceeded. It now should
2268 MH/03 Made $recipients available in local_scan(). local_scan() already has
2269 better access to the recipient list through recipients_list[], but
2270 $recipients can be useful in postmaster-provided expansion strings.
2272 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2273 in the case of a MAIL command with additional options following the
2274 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2275 were accidentally chopped off.
2277 PH/29 SMTP synchronization checks are implemented when a command is read -
2278 there is a check that no more input is waiting when there shouldn't be
2279 any. However, for some commands, a delay in an ACL can mean that it is
2280 some time before the response is written. In this time, more input might
2281 arrive, invalidly. So now there are extra checks after an ACL has run for
2282 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2283 pipelining has not been advertised.
2285 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2287 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2288 returned in the SMTP response when smtp_return_error_details was false.
2289 This has been fixed.
2291 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2292 instead of the C I/O that was originally supplied, because problems were
2293 reported on Solaris.
2295 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2296 Exim which did not show up earlier: it was assuming that a call to
2297 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2298 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2299 was a macro that became an assignment, so it seemed to work. This has
2300 changed to a proper function call with a void return, hence the compile
2301 error. Exim's code has been fixed.
2303 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2306 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2308 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2310 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2311 because it turns out to be too big or there is a timeout) there is no
2312 "Completed" line in the log. When some messages of this type were
2313 selected by exigrep, they were listed as "not completed". Others were
2314 picked up by some special patterns. I have improved the selection
2315 criteria to be more general.
2317 PH/38 The host_find_failed option in the manualroute router can now be set
2318 to "ignore", to completely ignore a host whose IP address cannot be
2319 found. If all hosts are ignored, the behaviour is controlled by the new
2320 host_all_ignored option.
2322 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2323 homing or because of multiple MX records with /mx) generated more than
2324 one IP address, and the following item turned out to be the local host,
2325 all the secondary addresses of the first item were incorrectly removed
2326 from the list, along with the local host and any following hosts (which
2327 is what is supposed to happen).
2329 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2330 whoever called Exim into the -H file. In the case of the daemon it was
2331 behaving confusingly. When first started, it used values for whoever
2332 started the daemon, but after a SIGHUP it used the Exim user (because it
2333 calls itself on a restart). I have changed the code so that it now always
2336 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2337 message are rejected with the same error (e.g. no authentication or bad
2338 sender address), and a DATA command is nevertheless sent (as can happen
2339 with PIPELINING or a stupid MUA), the error message that was given to the
2340 RCPT commands is included in the rejection of the DATA command. This is
2341 intended to be helpful for MUAs that show only the final error to their
2344 PH/42 Another patch from the Sieve maintainer.
2346 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2347 Eximstats - Fixed some broken HTML links and added missing column headers
2349 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2350 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2352 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2354 PH/43 Yet another patch from the Sieve maintainer.
2356 PH/44 I found a way to check for a TCP/IP connection going away before sending
2357 the response to the final '.' that terminates a message, but only in the
2358 case where the client has not sent further data following the '.'
2359 (unfortunately, this is allowed). However, in many cases there won't be
2360 any further data because there won't be any more messages to send. A call
2361 to select() can be used: if it shows that the input is "ready", there is
2362 either input waiting, or the socket has been closed. An attempt to read
2363 the next input character can distinguish the two cases. Previously, Exim
2364 would have sent an OK response which the client would never have see.
2365 This could lead to message repetition. This fix should cure that, at
2366 least in a lot of common cases.
2368 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2369 advertised in response to EHLO.
2375 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2376 fixed by 4.65/MH/01 (is this a record?) are fixed:
2378 (i) An empty string was always treated as zero by the numeric comparison
2379 operators. This behaviour has been restored.
2381 (ii) It is documented that the numeric comparison operators always treat
2382 their arguments as decimal numbers. This was broken in that numbers
2383 starting with 0 were being interpreted as octal.
2385 While fixing these problems I realized that there was another issue that
2386 hadn't been noticed. Values of message_size_limit (both the global option
2387 and the transport option) were treated as octal if they started with 0.
2388 The documentation was vague. These values are now always treated as
2389 decimal, and I will make that clear in the documentation.
2395 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2396 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2399 MH/01 Don't check that the operands of numeric comparison operators are
2400 integers when their expansion is in "skipping" mode (fixes bug
2401 introduced by 4.64-PH/07).
2403 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2404 child addresses, Exim now panics and dies. Previously, because the count
2405 is held in a short int, deliveries were likely to be lost. As such a
2406 large number of recipients for a single message is ridiculous
2407 (performance will be very, very poor), I have chosen to impose a limit
2408 rather than extend the field.
2414 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2415 leftover -K file (the existence of which was triggered by #402).
2416 While we were at it, introduced process PID as part of the -K
2417 filename. This should rule out race conditions when creating
2420 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2421 processing considerably. Previous code took too long for large mails,
2422 triggering a timeout which in turn triggers #401.
2424 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2425 in the DK code in transports.c. sendfile() is not really portable,
2426 hence the _LINUX specificness.
2428 TF/01 In the add_headers option to the mail command in an Exim filter,
2429 there was a bug that Exim would claim a syntax error in any
2430 header after the first one which had an odd number of characters
2433 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2434 callout verification, Exim cached a "reject" for the entire domain. This
2435 is correct for most verifications, but it is not correct for a recipient
2436 verification with use_sender or use_postmaster set, because in that case
2437 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2438 case of MAIL FROM:<> rejection from other early rejections (e.g.
2439 rejection of HELO). When verifying a recipient using a non-null MAIL
2440 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2441 Whatever the result of the callout, the value of the domain cache is
2442 left unchanged (for any other kind of callout, getting as far as trying
2443 RCPT means that the domain itself is ok).
2445 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2448 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2449 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2450 session, the daemon ignores ECONNECT errors and logs others; it now
2451 ignores EPIPE as well.
2453 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2454 (quoted-printable decoding).
2456 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2457 later the small subsequent patch to fix an introduced bug.
2459 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2461 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2463 PH/08 An error is now given if message_size_limit is specified negative.
2465 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2466 to be given (somewhat) arbitrary names.
2468 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2471 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2472 miscellaneous code fixes
2474 PH/10 Added the log_reject_target ACL modifier to specify where to log
2477 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2478 hostname. This is wrong, because it relates to the incoming message (and
2479 probably the interface on which it is arriving) and not to the outgoing
2480 callout (which could be using a different interface). This has been
2481 changed to use the value of the helo_data option from the smtp transport
2482 instead - this is what is used when a message is actually being sent. If
2483 there is no remote transport (possible with a router that sets up host
2484 addresses), $smtp_active_hostname is used.
2486 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2487 tweaks were necessary in order to get it to work (see also 21 below):
2488 (a) The code assumed that strncpy() returns a negative number on buffer
2489 overflow, which isn't the case. Replaced with Exim's string_format()
2491 (b) There were several signed/unsigned issues. I just did the minimum
2492 hacking in of casts. There is scope for a larger refactoring.
2493 (c) The code used strcasecmp() which is not a standard C function.
2494 Replaced with Exim's strcmpic() function.
2495 (d) The code set only $1; it now sets $auth1 as well.
2496 (e) A simple test gave the error "authentication client didn't specify
2497 service in request". It would seem that Dovecot has changed its
2498 interface. Fortunately there's a specification; I followed it and
2499 changed what the client sends and it appears to be working now.
2501 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2504 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2505 address A is aliased to B and C, where B exists and C does not. Without
2506 -v the output is "A verified" because verification stops after a
2507 successful redirection if more than one address is generated. However,
2508 with -v the child addresses are also verified. Exim was outputting "A
2509 failed to verify" and then showing the successful verification for C,
2510 with its parentage. It now outputs "B failed to verify", showing B's
2511 parentage before showing the successful verification of C.
2513 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2514 look up a TXT record in a specific list after matching in a combined
2517 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2518 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2519 they consult the DNS. I had assumed they would set it the way they
2520 wanted; and indeed my experiments on Linux seem to show that in some
2521 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2522 To be on the safe side, however, I have now made the interface to
2523 host_find_byname() similar to host_find_bydns(), with an argument
2524 containing the DNS resolver options. The host_find_byname() function now
2525 sets these options at its start, just as host_find_bydns() does. The smtp
2526 transport options dns_qualify_single and dns_search_parents are passed to
2527 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2528 of host_find_byname() use the default settings of RES_DEFNAMES
2529 (qualify_single) but not RES_DNSRCH (search_parents).
2531 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2532 spool_read_header() do less string testing, by means of a preliminary
2533 switch on the second character of optional "-foo" lines. (This is
2534 overdue, caused by the large number of possibilities that now exist.
2535 Originally there were few.) While I was there, I also converted the
2536 str(n)cmp tests so they don't re-test the leading "-" and the first
2537 character, in the hope this might squeeze out yet more improvement.
2539 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2540 flag allowing group syntax was set by the header_syntax check but not
2541 turned off, possible causing trouble later; (2) The flag was not being
2542 set at all for the header_verify test, causing "group"-style headers to
2543 be rejected. I have now set it in this case, and also caused header_
2544 verify to ignore an empty address taken from a group. While doing this, I
2545 came across some other cases where the code for allowing group syntax
2546 while scanning a header line wasn't quite right (mostly, not resetting
2547 the flag correctly in the right place). These bugs could have caused
2548 trouble for malformed header lines. I hope it is now all correct.
2550 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2551 with the "reply" argument non-NULL. The code, however (which originally
2552 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2553 but it didn't always do it. This confused somebody who was copying the
2554 code for some other use. I have removed all the tests.
2556 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2557 feature that was used to support insecure browsers during the U.S. crypto
2558 embargo. It requires special client support, and Exim is probably the
2559 only MTA that supported it -- and would never use it because real RSA is
2560 always available. This code has been removed, because it had the bad
2561 effect of slowing Exim down by computing (never used) parameters for the
2562 RSA_EXPORT functionality.
2564 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2565 authenticator to fail if there's a tab character in the incoming data
2566 (there should never be unless someone is messing about, as it's supposed
2567 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2568 if the connection is using TLS or if the remote IP is the same as the
2569 local IP, and the "valid-client-cert option" if a client certificate has
2572 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2573 authenticators. This can be used for authorization after authentication
2574 succeeds. (In the case of plaintext, it servers for both authentication
2577 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2578 if any retry times were supplied.
2580 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2581 connection, where there is no client IP address to check. In this
2582 situation, the verify now always succeeds.
2584 PH/25 Applied John Jetmore's -Mset patch.
2586 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2588 PH/27 In a string expansion for a processed (not raw) header when multiple
2589 headers of the same name were present, leading whitespace was being
2590 removed from all of them, but trailing whitespace was being removed only
2591 from the last one. Now trailing whitespace is removed from each header
2592 before concatenation. Completely empty headers in a concatenation (as
2593 before) are ignored.
2595 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2596 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2598 PH/29 [Removed. This was a change that I later backed out, and forgot to
2599 correct the ChangeLog entry (that I had efficiently created) before
2600 committing the later change.]
2602 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2603 address errors (4xx response to RCPT) over the same connection as other
2604 messages routed to the same hosts. Such deliveries are always "forced",
2605 so retry times are not inspected. This resulted in far too many retries
2606 for the affected addresses. The effect occurred only when there were more
2607 hosts than the hosts_max_try setting in the smtp transport when it had
2608 the 4xx errors. Those hosts that it had tried were not added to the list
2609 of hosts for which the message was waiting, so if all were tried, there
2610 was no problem. Two fixes have been applied:
2612 (i) If there are any address or message errors in an SMTP delivery, none
2613 of the hosts (tried or untried) are now added to the list of hosts
2614 for which the message is waiting, so the message should not be a
2615 candidate for sending over the same connection that was used for a
2616 successful delivery of some other message. This seems entirely
2617 reasonable: after all the message is NOT "waiting for some host".
2618 This is so "obvious" that I'm not sure why it wasn't done
2619 previously. Hope I haven't missed anything, but it can't do any
2620 harm, as the worst effect is to miss an optimization.
2622 (ii) If, despite (i), such a delivery is accidentally attempted, the
2623 routing retry time is respected, so at least it doesn't keep
2624 hammering the server.
2626 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2627 in ${readsocket because some servers need this prod.
2629 PH/32 Added some extra debug output when updating a wait-xxx database.
2631 PH/33 The hint "could be header name not terminated by colon", which has been
2632 given for certain expansion errors for a long time, was not being given
2633 for the ${if def:h_colon_omitted{... case.
2635 PH/34 The spec says: "With one important exception, whenever a domain list is
2636 being scanned, $domain contains the subject domain." There was at least
2637 one case where this was not true.
2639 PH/35 The error "getsockname() failed: connection reset by peer" was being
2640 written to the panic log as well as the main log, but it isn't really
2641 panic-worthy as it just means the connection died rather early on. I have
2642 removed the panic log writing for the ECONNRESET error when getsockname()
2645 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2646 runs only) independently of the message's sender address. This meant
2647 that, if the 4xx error was in fact related to the sender, a different
2648 message to the same recipient with a different sender could confuse
2649 things. In particular, this can happen when sending to a greylisting
2650 server, but other circumstances could also provoke similar problems.
2651 I have changed the default so that the retry time for these errors is now
2652 based a combination of the sender and recipient addresses. This change
2653 can be overridden by setting address_retry_include_sender=false in the
2656 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2657 remote server are returned as part of bounce messages. This was not
2658 happening for LMTP over a pipe (the lmtp transport), but now it is the
2659 same for both kinds of LMTP.
2661 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2662 in header lines that were in fact CNAMEs. This is no longer the case.
2664 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2665 and queue runs started by the daemon processed all messages. This has
2666 been fixed so that -R and -S can now usefully be given with -q<time>.
2668 PH/40 Import PCRE release 6.7 (fixes some bugs).
2670 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2672 PH/42 Give an error if -q is specified more than once.
2674 PH/43 Renamed the variables $interface_address and $interface_port as
2675 $received_ip_address and $received_port, to make it clear that these
2676 values apply to message reception, and not to the outgoing interface when
2677 a message is delivered. (The old names remain recognized, of course.)
2679 PH/44 There was no timeout on the connect() call when using a Unix domain
2680 socket in the ${readsocket expansion. There now is.
2682 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2683 be meaningful with "accept".
2685 SC/01 Eximstats V1.43
2686 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2688 SC/02 Eximstats V1.44
2689 Use a glob alias rather than an array ref in the generated
2690 parser. This improves both readability and performance.
2692 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2693 Collect SpamAssassin and rejection statistics.
2694 Don't display local sender or destination tables unless
2695 there is data to show.
2696 Added average volumes into the top table text output.
2698 SC/04 Eximstats V1.46
2699 Collect data on the number of addresses (recipients)
2700 as well as the number of messages.
2702 SC/05 Eximstats V1.47
2703 Added 'Message too big' to the list of mail rejection
2704 reasons (thanks to Marco Gaiarin).
2706 SC/06 Eximstats V1.48
2707 Mainlog lines which have GMT offsets and are too short to
2708 have a flag are now skipped.
2710 SC/07 Eximstats V1.49 (Alain Williams)
2711 Added the -emptyok flag.
2713 SC/08 Eximstats V1.50
2714 Fixes for obtaining the IP address from reject messages.
2716 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2717 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2718 whitespace changes from 4.64-PH/27
2720 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2723 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2724 are found, allow negative numbers in numeric criteria)
2726 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2728 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2729 to match changes made in 4.64-PH/43
2731 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2733 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2734 "domain" is actually an IP address, and force a failure. This locks out
2735 those revolvers/nameservers that support "A-for-A" lookups, in
2736 contravention of the specifications.
2738 PH/48 When a host name was looked up from an IP address, and the subsequent
2739 forward lookup of the name timed out, the host name was left in
2740 $sender_host_name, contrary to the specification.
2742 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2743 restricted to single-key lookups, Exim was not diagnosing an error if
2744 * or *@ was used with a query-style lookup.
2746 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2748 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2749 long ago, but noone interested enough thought of it. Let's just say that
2750 the "1.1" means that there are some new functions that weren't there at
2751 some point in the past.
2753 PH/51 Error processing for expansion failure of helo_data from an smtp
2754 transport during callout processing was broken.
2756 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2757 tested/used via the -bh/-bhc/-bs options.
2759 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2760 bug, fixed in subsequent PCRE releases).
2762 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2763 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2765 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2771 SC/01 Use a glob alias rather than an array ref in eximstats generated
2772 parser. This improves both readability and performance.
2774 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2775 Don't display local sender or destination tables in eximstats unless
2776 there is data to show.
2777 Added average volumes into the eximstats top table text output.
2779 SC/03 Collect data on the number of addresses (recipients) as well
2780 as the number of messages in eximstats.
2782 TF/01 Correct an error in the documentation for the redirect router. Exim
2783 does (usually) call initgroups() when daemonizing.
2785 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2786 with consistent privilege compared to when running as a daemon.
2788 TF/03 Note in the spec that $authenticated_id is not set for local
2789 submissions from trusted users.
2791 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2792 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2794 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2795 by adding some example configuration directives to the default
2796 configuration file. A little bit of work is required to uncomment the
2797 directives and define how usernames and passwords are checked, but
2798 there is now a framework to start from.
2800 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2801 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2802 without this. I don't know how relevant this is to other LDAP libraries.
2804 PH/02 Add the verb name to the "unknown ACL verb" error.
2806 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2808 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2810 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2811 directory not expanded when it should be if an expanded home directory
2812 was set for the address (which is overridden by the transport).
2814 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2817 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2818 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2819 because it is too late at that time, and has no effect.
2821 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2822 security issue with \' (bugzilla #107). I could not use the
2823 PQescapeStringConn() function, because it needs a PGconn value as one of
2826 PH/08 When testing addresses using -bt, indicate those final addresses that
2827 are duplicates that would not cause an additional delivery. At least one
2828 person was confused, thinking that -bt output corresponded to deliveries.
2829 (Suppressing duplicates isn't a good idea as you lose the information
2830 about possibly different redirections that led to the duplicates.)
2832 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2833 systems where poll() doesn't work, in particular OS X.
2835 PH/10 Added more information to debugging output for retry time not reached.
2837 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2838 operations in malware.c.
2840 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2843 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2844 syslog_duplication set false, log lines that would normally be written
2845 both the the main log and to the reject log were not written to syslog at
2848 PH/14 In the default configuration, change the use of "message" in ACL warn
2849 statements to "add_header".
2851 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2852 not followed by a command (e.g. "seen endif").
2854 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2855 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2858 PH/17 Added extra conditions to the default value of delay_warning_condition
2862 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2863 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2864 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2867 The Auto-Submitted: and various List- headers are standardised, whereas I
2868 don't think Precedence: ever was.
2870 PH/18 Refactored debugging code in route_finduser() to show more information,
2871 in particular, the error code if getpwnam() issues one.
2873 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2874 This is apparently needed in addition to the PH/07 change above to avoid
2875 any possible encoding problems.
2877 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2878 but not after initializing Perl.
2880 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2881 output them only if debugging. By default they are written stderr,
2882 apparently, which is not desirable.
2884 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2887 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2890 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2892 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2893 authenticated or an ident call has been made. Suppress the default
2894 values for $authenticated_id and $authenticated_sender (but permit -oMai
2895 and -oMas) when testing with -bh.
2897 PH/24 Re-jigged the order of the tests in the default configuration so that the
2898 tests for valid domains and recipients precede the DNS black list and CSA
2899 tests, on the grounds that those ones are more expensive.
2901 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2902 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2903 This bug exists in every version of Exim that I still have, right back to
2906 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2907 However, an attempt to turn on case-sensitivity in a regex key by
2908 including (?-i) didn't work because the subject string was already
2909 lowercased, and the effects were non-intuitive. It turns out that a
2910 one-line patch can be used to allow (?-i) to work as expected.
2916 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2917 other effects) broke the use of negated acl sub-conditions.
2919 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2922 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2923 "Deny" causes Exim to reject the incoming connection with a 554 error.
2924 Unfortunately, if there is a major crisis, such as a disk failure,
2925 tcp-wrappers gives "deny", whereas what one would like would be some
2926 kind of temporary error. A kludge has been added to help with this.
2927 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2928 554 error is used if errno is still zero or contains ENOENT (which occurs
2929 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2932 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2934 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2935 errors. Otherwise a message that provokes a temporary error (when other
2936 messages do not) can cause a whole host to time out.
2938 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2939 addresses were routed directly to files or pipes from a redirect router.
2940 File deliveries just didn't batch; pipe deliveries might have suffered
2943 PH/06 A failure to get a lock for a hints database would erroneously always say
2944 "Failed to get write lock", even when it was really a read lock.
2946 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2947 of 0600. This has been changed to use the value of the lockfile_mode
2948 option (which defaults to 0600).
2950 PH/08 Applied small patch from the Sieve maintainer.
2952 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2953 folder from quota calculations, a direct delivery into this folder messed
2954 up the contents of the maildirsize file. This was because the regex was
2955 used only to exclude .Trash (or whatever) when the size of the mailbox
2956 was calculated. There was no check that a delivery was happening into an
2957 excluded directory. This bug has been fixed by ignoring all quota
2958 processing for deliveries into excluded directories.
2960 PH/10 Added the maildirfolder_create_regex option to appendfile.
2966 PH/01 The code for finding all the local interface addresses on a FreeBSD
2967 system running IPv6 was broken. This may well have applied to all BSD
2968 systems, as well as to others that have similar system calls. The broken
2969 code found IPv4 interfaces correctly, but gave incorrect values for the
2970 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2971 that it would not match correctly against @[] and not recognize the IPv6
2974 PH/02 The ipliteral router was not recognizing addresses of the form user@
2975 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2977 PH/03 Added disable_ipv6.
2979 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2980 decoded form, because it is most often used to construct To: headers
2981 lines in autoreplies, and the decoded form may well be syntactically
2982 invalid. However, $reply_address has leading white space removed, and all
2983 newlines turned into spaces so that the autoreply transport does not
2986 PH/05 If group was specified without a user on a router, and no group or user
2987 was specified on a transport, the group from the router was ignored.
2989 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2990 for visible compile-time settings that can be used to change these
2991 numbers, for those that want even more. Backwards compatibility with old
2992 spool files has been maintained. However, going back to a previous Exim
2993 release will lost any variables that are in spool files.
2995 PH/07 Two small changes when running in the test harness: increase delay when
2996 passing a TCP/IP connection to a new process, in case the original
2997 process has to generate a bounce, and remove special handling of
2998 127.0.0.2 (sic), which is no longer necessary.
3000 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3001 be the same on different OS.
3003 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3006 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3007 whether --show-vars was specified or not
3009 JJ/02 exipick: Added support for new ACL variable spool format introduced
3012 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3013 syntactically invalid From: or Reply-to: line, and a filter used this to
3014 generate an autoreply, and therefore failed to obtain an address for the
3015 autoreply, Exim could try to deliver to a non-existent relative file
3016 name, causing unrelated and misleading errors. What now happens is that
3017 it logs this as a hard delivery error, but does not attempt to create a
3020 PH/11 The exinext utility has a -C option for testing purposes, but although
3021 the given file was scanned by exinext itself; it wasn't being passed on
3022 when Exim was called.
3024 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3025 an end-of-file indication when reading a command response.
3027 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3028 compiled. In many other places in Exim, IPv6 addresses are always
3029 recognized, so I have changed this. It also means that IPv4 domain
3030 literals of the form [IPV4:n.n.n.n] are now always recognized.
3032 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3033 used if the router is not running as root, for example, when verifying at
3034 ACL time, or when using -bh. The debugging output from this situation was
3035 non-existent - all you got was a failure to exec. I have made two
3038 (a) Failures to set uid/gid, the current directory, or a process leader
3039 in a subprocess such as that created by queryprogram now generate
3040 suitable debugging output when -d is set.
3042 (b) The queryprogram router detects when it is not running as root,
3043 outputs suitable debugging information if -d is set, and then runs
3044 the subprocess without attempting to change uid/gid.
3046 PH/15 Minor change to Makefile for building test_host (undocumented testing
3049 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3050 additional section of a DNS packet that returns MX or SRV records.
3051 Instead, it always explicitly searches for A/AAAA records. This avoids
3052 major problems that occur when a DNS server includes only records of one
3053 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3054 fixed another bug: if SRV records were looked up and the corresponding
3055 address records were *not* found in the additional section, the port
3056 values from the SRV records were lost.
3058 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3059 using the correct key (the original address) when searching the retry
3060 rules in order to find which one to use for generating the retry hint.
3062 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3063 adding the default one. Similarly, if it contains a Reply-To: header, the
3064 errors_reply_to option, if set, is not used.
3066 PH/19 When calculating a retry time, Exim used to measure the "time since
3067 failure" by looking at the "first failed" field in the retry record. Now
3068 it does not use this if it is later than than the arrival time of the
3069 message. Instead it uses the arrival time. This makes for better
3070 behaviour in cases where some deliveries succeed, thus re-setting the
3071 "first failed" field. An example is a quota failure for a huge message
3072 when small messages continue to be delivered. Without this change, the
3073 "time since failure" will always be short, possible causing more frequent
3074 delivery attempts for the huge message than are intended.
3075 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3077 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3078 $1, $2, $3) because the numerical variables can be reset during some
3079 expansion items (e.g. "match"), thereby losing the authentication data.
3081 PH/21 Make -bV show the size of off_t variables so that the test suite can
3082 decide whether to run tests for quotas > 2G.
3084 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3085 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3086 filecount value is greater than 2G or if a quota value is greater than 2G
3087 on a system where the size of off_t is not greater than 4, a panic error
3090 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3091 never match. The debug and -bh output now contains an explicit error
3092 message indicating a malformed IPv4 address or mask.
3094 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3095 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3096 PH/23 above applies.
3098 PH/25 Do not write to syslog when running in the test harness. The only
3099 occasion when this arises is a failure to open the main or panic logs
3100 (for which there is an explicit test).
3102 PH/26 Added the /no_tell option to "control=freeze".
3104 PH/27 If a host name lookup failed very early in a connection, for example, if
3105 the IP address matched host_lookup and the reverse lookup yielded a name
3106 that did not have a forward lookup, an error message of the form "no IP
3107 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3108 could be logged. Now it outputs the IP address instead of "NULL".
3110 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3111 allows the sender and the authenticated sender to be set when
3112 submitting a message from within Exim. Since child_open_exim() is
3113 documented for local_scan(), the new function should be too.
3115 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3116 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3117 results in an empty string is now treated as unset.
3119 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3121 PH/31 Added sender_verify_fail logging option.
3123 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3124 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3125 tidied the source and removed it altogether.
3127 PH/33 When a queue run was abandoned because the load average was too high, a
3128 log line was always written; now it is written only if the queue_run log
3129 selector is set. In addition, the log line for abandonment now contains
3130 information about the queue run such as the pid. This is always present
3131 in "start" and "stop" lines but was omitted from the "abandon" line.
3133 PH/34 Omit spaces between a header name and the colon in the error message that
3134 is given when verify = headers_syntax fails (if there are lots of them,
3135 the message gets confusing).
3137 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3138 names, as there are now some PTR records that contain slashes. This check
3139 is only to protect against broken name servers that fall over on strange
3140 characters, so the fact that it applies to all lookups doesn't matter.
3142 PH/36 Now that the new test suite is complete, we can remove some of the
3143 special code in Exim that was needed for the old test suite. For example,
3144 sorting DNS records because real resolvers return them in an arbitrary
3145 order. The new test suite's fake resolver always returns records in the
3148 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3149 bounces) except when queue_only is set, to avoid logging races between
3150 the different processes.
3152 PH/38 Panic-die if .include specifies a non-absolute path.
3154 PH/39 A tweak to the "H" retry rule from its user.
3156 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3157 a label. They prevented compilation on older perls.
3159 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3160 a warning to be raised on newish perls.
3162 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3163 on queue. Changes to match documented behaviour of showing count of
3164 messages matching specified criteria.
3166 PH/40 Changed the default ident timeout from 30s to 5s.
3168 PH/41 Added support for the use of login_cap features, on those BSD systems
3169 that have them, for controlling the resources used by pipe deliveries.
3171 PH/42 The content-scanning code uses fopen() to create files in which to put
3172 message data. Previously it was not paying any attention to the mode of
3173 the files. Exim runs with umask(0) because the rest of the code creates
3174 files with open(), and sets the required mode explicitly. Thus, these
3175 files were ending up world-writeable. This was not a big issue, because,
3176 being within the spool directory, they were not world-accessible. I have
3177 created a function called modefopen, which takes an additional mode
3178 argument. It sets umask(777), creates the file, chmods it to the required
3179 mode, then resets the umask. All the relevant calls to fopen() in the
3180 content scanning code have been changed to use this function.
3182 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3183 to 24 hours. This avoids potential overflow problems when processing G
3184 and H retry rules. I suspect nobody ever tinkers with this value.
3186 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3188 PH/45 When the plaintext authenticator is running as a client, the server's
3189 challenges are checked to ensure they are valid base64 strings. By
3190 default, the authentication attempt is cancelled if an invalid string is
3191 received. Setting client_ignore_invalid_base64 true ignores these errors.
3192 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3193 they are received. Thus, the responses can be made to depend on the
3194 challenges. If an invalid string is ignored, an empty string is placed in
3197 PH/46 Messages that are created by the autoreply transport now contains a
3198 References: header, in accordance with RFCs 2822 and 3834.
3200 PH/47 Added authenticated_sender_force to the smtp transport.
3202 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3204 PH/49 Installed latest patch from the Sieve maintainer.
3206 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3207 was also set, the appendfile transport was unnecessarily scanning a
3208 directory of message files (e.g. for maildir delivery) to find the count
3209 of files (along with the size), even though it did not need this
3210 information. It now does the scan only if it needs to find either the
3211 size of the count of files.
3213 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3215 PH/52 Two bugs concerned with error handling when the smtp transport is
3218 (i) Exim was not creating retry information for temporary errors given
3219 for individual recipients after the DATA command when the smtp transport
3220 was used in LMTP mode. This meant that they could be retried too
3221 frequently, and not timed out correctly.
3223 (ii) Exim was setting the flag that allows error details to be returned
3224 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3225 recipients that were returned after the DATA command.
3227 PH/53 This is related to PH/52, but is more general: for any failing address,
3228 when detailed error information was permitted to be returned to the
3229 sender, but the error was temporary, then after the final timeout, only
3230 "retry timeout exceeded" was returned. Now it returns the full error as
3231 well as "retry timeout exceeded".
3233 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3234 do this, and (what is worse) MTAs that accept it.
3236 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3237 will now be deprecated.
3239 PH/56 New os.c-cygwin from the Cygwin maintainer.
3241 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3242 formats (previously only available in exim formats via -bpr, -bpru,
3243 and -bpra. Now also available in native and exiqgrep formats)
3245 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3246 with very large, slow to parse queues
3248 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3250 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3252 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3253 responses to authentication challenges, though it was showing the
3254 challenges; (ii) I've removed the CR characters from the debug output for
3257 PH/58 Allow for the insertion of a newline as well as a space when a string
3258 is turned into more than one encoded-word during RFC 2047 encoding. The
3259 Sieve code now uses this.
3261 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3262 data_4xx, lost_connection, tls_required.
3264 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3265 message was being sent as an SMTP response.
3267 PH/61 Add -l and -k options to exicyclog.
3269 PH/62 When verifying, if an address was redirected to one new address, so that
3270 verification continued, and the new address failed or deferred after
3271 having set something in $address_data, the value of $address_data was not
3272 passed back to the ACL. This was different to the case when no
3273 redirection occurred. The value is now passed back in both cases.
3275 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3276 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3277 use login_cap.h, so on its own it isn't the distinguishing feature. The
3278 new name refers directly to the setclassresources() function.
3280 PH/65 Added configuration files for NetBSD3.
3282 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3284 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3285 is preferred over IPv4.
3287 PH/68 The bounce_return_message and bounce_return_body options were not being
3288 honoured for bounces generated during the reception of non-SMTP messages.
3289 In particular, this applied to messages rejected by the ACL. This bug has
3290 been fixed. However, if bounce_return_message is true and bounce_return_
3291 body is false, the headers that are returned for a non-SMTP message
3292 include only those that have been read before the error was detected.
3293 (In the case of an ACL rejection, they have all been read.)
3295 PH/69 The HTML version of the specification is now built in a directory called
3296 spec_html instead of spec.html, because the latter looks like a path with
3297 a MIME-type, and this confuses some software.
3299 PH/70 Catch two compiler warnings in sieve.c.
3301 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3302 function verify_get_ident() calls ip_connect() to connect a socket, but
3303 if the "connect()" function timed out, ip_connect() used to close the
3304 socket. However, verify_get_ident() also closes the socket later, and in
3305 between Exim writes to the log, which may get opened at this point. When
3306 the socket was closed in ip_connect(), the log could get the same file
3307 descriptor number as the socket. This naturally causes chaos. The fix is
3308 not to close the socket in ip_connect(); the socket should be closed by
3309 the function that creates it. There was only one place in the code where
3310 this was missing, in the iplookup router, which I don't think anybody now
3311 uses, but I've fixed it anyway.
3313 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3314 well as to direct DNS lookups. Otherwise the handling of names in host
3315 lists is inconsistent and therefore confusing.
3321 PH/01 Two changes to the default runtime configuration:
3323 (1) Move the checks for relay_from_hosts and authenticated clients from
3324 after to before the (commented out) DNS black list checks.
3326 (2) Add control=submission to the relay_from_hosts and authenticated
3327 clients checks, on the grounds that messages accepted by these
3328 statements are most likely to be submissions.
3330 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3332 (1) Generate an error if the third argument for the ${prvs expansion is
3335 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3338 (3) Reset the variables that are obtained from the first argument of
3339 ${prvscheck and used in the second argument before leaving the code,
3340 because their memory is reclaimed, so using them afterwards may do
3343 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3344 one (it's much easier than Tom thought :-).
3346 (5) Because of (4), we can now allow for the use of $prvscheck_result
3347 inside the third argument.
3349 PH/03 For some reason, the default setting of PATH when running a command from
3350 a pipe transport was just "/usr/bin". I have changed it to
3353 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3354 anything to be listed in the output from -bV.
3356 PH/05 When a filter generated an autoreply, the entire To: header line was
3357 quoted in the delivery log line, like this:
3359 => >A.N.Other <ano@some.domain> <original@ddress> ...
3361 This has been changed so that it extracts the operative address. There
3362 may be more than one such address. If so, they are comma-separated, like
3365 => >ano@some.domain,ona@other.domain <original@ddress> ...
3367 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3368 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3369 not being looked up in the rDNS to get a host name, Exim was showing the
3370 IP address twice in Received: lines, even though the IP addresses were
3371 identical. For example:
3373 Received: from [1.2.3.4] (helo=[1.2.3.4])
3375 However, if the real host name was known, it was omitting the HELO data
3376 if it matched the actual IP address. This has been tidied up so that it
3377 doesn't show the same IP address twice.
3379 PH/07 When both +timestamp and +memory debugging was on, the value given by
3380 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3381 called by the debug printing, thereby overwriting the timestamp buffer.
3382 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3384 PH/08 When the original message was included in an autoreply transport, it
3385 always said "this is a copy of the message, including all the headers",
3386 even if body_only or headers_only was set. It now gives an appropriate
3389 PH/09 Applied a patch from the Sieve maintainer which:
3391 o fixes some comments
3392 o adds the (disabled) notify extension core
3393 o adds some debug output for the result of if/elsif tests
3394 o points to the current vacation draft in the documentation
3395 and documents the missing references header update
3399 o fixes a bug in processing the envelope test (when testing
3400 multiple envelope elements, the last element determined the
3403 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3404 Electronic Mail") by including:
3406 Auto-submitted: auto-generated
3408 in the messages that it generates (bounce messages and others, such as
3409 warnings). In the case of bounce messages for non-SMTP messages, there was
3410 also a typo: it was using "Auto_submitted" (underscore instead of
3411 hyphen). Since every message generated by Exim is necessarily in response
3412 to another message, thes have all been changed to:
3414 Auto-Submitted: auto-replied
3416 in accordance with these statements in the RFC:
3418 The auto-replied keyword:
3420 - SHOULD be used on messages sent in direct response to another
3421 message by an automatic process,
3423 - MUST NOT be used on manually-generated messages,
3425 - MAY be used on Delivery Status Notifications (DSNs) and Message
3426 Disposition Notifications (MDNs),
3428 - MUST NOT be used on messages generated by automatic or periodic
3429 processes, except for messages which are automatic responses to
3432 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3433 to the default Received: header definition.
3435 PH/12 Added log selector acl_warn_skipped (default on).
3437 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3438 variables because (a) they are in the wrong storage pool and (b) even if
3439 they were copied, it wouldn't work properly because of the caching.
3441 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3442 checking when decoding. Apparently there are clients that generate
3443 overlong encoded strings. Why am I not surprised?
3445 PH/15 If the first argument of "${if match_address" was not empty, but did not
3446 contain an "@" character, Exim crashed. Now it writes a panic log message
3447 and treats the condition as false.
3449 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3451 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3452 extension "envelope-auth". The code is finished and in agreement with
3453 other implementations, but there is no documentation so far and in fact,
3454 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3455 not changing the active code.
3457 Print executed "if" and "elsif" statements when debugging is used. This
3458 helps a great deal to understand what a filter does.
3460 Document more things not specified clearly in RFC3028. I had all this
3461 sorted out, when out of a sudden new issues came to my mind. Oops."
3463 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3466 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3467 canonical form (as documented). However, after a host name lookup from
3468 the IP address, check_host() was doing a simple string comparison with
3469 addresses acquired from the DNS when checking that the found name did
3470 have the original IP as one of its addresses. Since any found IPv6
3471 addresses are likely to be in abbreviated form, the comparison could
3472 fail. Luckily, there already exists a function for doing the comparison
3473 by converting both addresses to binary, so now that is used instead of
3474 the text comparison.
3476 PH/20 There was another similar case to PH/19, when a complete host name was
3477 given in a host list; looking up its IP address could give an abbreviated
3478 form, whereas the current host's name might or might not be abbreviated.
3479 The same fix has been applied.
3485 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3486 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3489 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3490 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3492 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3494 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3495 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3496 Also, for "Auto-Submitted", treat anything other than "no" as
3497 non-personal, in accordance with RFC 3834. (Previously it treated
3498 anything starting "auto-" as non-personal.)
3500 TF/01 The control=submission/name=... option had a problem with syntax
3501 errors if the name included a slash character. The /name= option
3502 now slurps the rest of the string, so it can include any characters
3503 but it must come last in the list of options (after /sender_retain
3506 PH/05 Some modifications to the interface to the fake nameserver for the new
3514 TK/01 Added the "success_on_redirect" address verification option. See
3515 NewStuff for rationale and an example.
3517 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3519 PH/02 Patch to exigrep to allow it to work on syslog lines.
3521 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3522 fread() to skip over the body file's header line, because in Cygwin the
3523 header line is locked and is inaccessible.
3525 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3526 co-exist for some time) to make it clear that it is the Exim ID that is
3527 referenced, not the Message-ID: header line.
3529 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3530 string_format() function, because snprintf() does not exist on all
3533 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3534 ${stat: expansion item.
3536 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3537 protocol synchronization error", to keep the pedants happy.
3539 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3540 well as for IRIX systems, when gcc is being used. See the host.c source
3543 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3545 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3548 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3549 transport and to the smtp transport in LMTP mode.
3551 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3553 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3554 run a filter in a subprocess. This could lead to confusion in subsequent
3555 lookups in the parent process. There should also be a search_tidyup() at
3556 the end of the subprocess.
3558 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3559 only if the host matched helo_try_verify_hosts, which caused the
3560 verification to occur when the EHLO/HELO command was issued. The ACL just
3561 tested the remembered result. Now, if a previous verification attempt has
3562 not happened, "verify = helo" does it there and then.
3564 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3566 TK/03 Fix log output including CR from clamd.
3568 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3569 exist provoked a memory error which could cause a segfault.
3571 PH/15 Installed PCRE 6.2
3573 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3575 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3576 of the problem. Specifically, suggested +O2 rather than +O1 for the
3579 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3581 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3582 record, the port was not correctly logged when the outgoing_port log
3583 selector was set (it logged the transort's default port).
3585 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3586 fallback_hosts, and "hosts" in the smtp transport.
3588 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3589 host errors such as "Connection refused".
3591 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3592 authentication with radiusclient 0.4.9:
3594 - Error returned from rc_read_config was caught wrongly
3595 - Username/password not passed on to radius server due to wrong length.
3597 The presumption is that some radiusclient API changes for 4.51/PH/17
3598 were not taken care of correctly. The code is still untested by me (my
3599 Linux distribution still has 0.3.2 of radiusclient), but it was
3600 contributed by a Radius user.
3602 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3603 expanding the "port" option of the smtp transport.
3605 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3606 while reading a MIME header. Thanks to Tom Hughes for a patch.
3608 PH/24 Include config.h inside local_scan.h so that configuration settings are
3611 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3612 that in an ACL for RCPT (for example), you can examine exactly what was
3615 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3616 commands, but it was not correctly comparing the address with the actual
3617 client host address. Thus, it would show the EHLO address in Received:
3618 header lines when this was not necessary.
3620 PH/27 Added the % operator to ${eval:}.
3622 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3623 it should be ignoring failures (because with -C, for example, it has lost
3624 privilege). It wasn't ignoring creation failures other than "already
3627 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3630 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3631 because an input line was too long, either on its own, or by virtue of
3632 too many continuations, the temporary file was not being removed, and the
3633 return code was incorrect.
3635 PH/31 Missing "BOOL" in function definition in filtertest.c.
3637 PH/32 Applied Sieve patches from the maintainer.
3639 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3641 PH/33 Added "verify = not_blind".
3643 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3644 Local/Makefile (with some defaults set). These are used in built scripts
3645 such as exicyclog, but they have never been used in the exim_install
3646 script (though there are many overriding facilities there). I have
3647 arranged that the exim_install script now takes note of these two
3650 PH/35 Installed configuration files for Dragonfly.
3652 PH/36 When a locally submitted message by a trusted user did not contain a
3653 From: header, and the sender address was obtained from -f or from an SMTP
3654 MAIL command, and the trusted user did not use -F to supply a sender
3655 name, $originator_name was incorrectly used when constructing a From:
3656 header. Furthermore, $originator_name was used for submission mode
3657 messages from external hosts without From: headers in a similar way,
3658 which is clearly wrong.
3660 PH/37 Added control=suppress_local_fixups.
3662 PH/38 When log_selector = +received_sender was set, and the addition of the
3663 sender made the log line's construction buffer exactly full, or one byte
3664 less than full, an overflow happened when the terminating "\n" was
3667 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3668 when the result of a list match is failure because a DNS lookup failed.
3670 PH/40 RM_COMMAND is now used in the building process.
3672 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3673 the "build-* directories that it finds.
3675 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3676 address in a domain literal was a prefix of an interface address.
3678 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3679 when verifying a sender address, unless rewrite_headers is false.
3681 PH/44 Wrote a long comment about why errors_to addresses are verified as
3682 recipients, not senders.
3684 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3685 the ratelimit ACL was added.
3687 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3689 PH/46 Added extra information about PostgreSQL errors to the error string.
3691 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3692 avoiding the need to install special zones in a real server. This is
3693 backwards compatible; if it can't find the fake resolver, it drops back.
3694 Thus, both old and new test suites can be run.
3696 TF/02 Added util/ratelimit.pl
3698 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3699 clock is set back in time.
3701 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3702 Candler <B.Candler@pobox.com>.
3704 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3705 OK for addresses that are the result of redirections.
3707 PH/48 A number of further additions for the benefit of the new test suite,
3708 including a fake gethostbyname() that interfaces to the fake DNS resolver
3711 TF/06 The fix for widen_domains has also been applied to qualify_single and
3712 search_parents which are the other dnslookup options that can cause
3715 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3718 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3720 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3721 service" if it encountered an MX record with an empty target hostname.
3722 The message is now "an MX or SRV record indicated no SMTP service".
3724 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3725 if the DNS of the sending site is misconfigured. This is quite a
3726 common situation. This change restores the behaviour of treating a
3727 helo verification defer as a failure.
3729 PH/51 If self=fail was set on a router, the bounce message did not include the
3730 actual error message.
3736 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3738 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3739 command itself ended in error, the underlying message about the transport
3740 filter timeout was being overwritten with the pipe command error. Now the
3741 underlying error message should be appended to the second error message.
3743 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3745 PH/02 Reduce the amount of output that "make" produces by default. Full output
3746 can still be requested.
3748 PH/03 The warning log line about a condition test deferring for a "warn" verb
3749 was being output only once per connection, rather than after each
3750 occurrence (because it was using the same function as for successful
3751 "warn" verbs). This seems wrong, so I have changed it.
3753 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3754 it should not have, which might have caused a crash in the right
3755 circumstances, but probably never did.
3757 PH/04 Installed a modified version of Tony Finch's patch to make submission
3758 mode fix the return path as well as the Sender: header line, and to
3759 add a /name= option so that you can make the user's friendly name appear
3762 TF/03 Added the control = fakedefer ACL modifier.
3764 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3765 Mark Lowes for thorough testing.
3767 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3769 TK/03 Merged latest SRS patch from Miles Wilton.
3771 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3772 with the definition in sysexits.h (which is #included earlier).
3773 Fortunately, Exim does not actually use EX_OK. The code used to try to
3774 preserve the sysexits.h value, by assuming that macro definitions were
3775 scanned for macro replacements. I have been disabused of this notion,
3776 so now the code just undefines EX_OK before #including unistd.h.
3778 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3779 in the smtp transport. When a block could not be written in a single
3780 write() function, the timeout was being re-applied to each part-write.
3781 This seems wrong - if the receiver was accepting one byte at a time it
3782 would take for ever. The timeout is now adjusted when this happens. It
3783 doesn't have to be particularly precise.
3785 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3786 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3788 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3789 without a domain if the check to <postmaster@domain> fails.
3791 SC/01 Eximstats: added -xls and the ability to specify output files
3792 (patch written by Frank Heydlauf).
3794 SC/02 Eximstats: use FileHandles for outputting results.
3796 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3798 SC/04 Eximstats: fixed display of large numbers with -nvr option
3800 SC/05 Eximstats: fixed merging of reports with empty tables.
3802 SC/06 Eximstats: added the -include_original_destination flag
3804 SC/07 Eximstats: removed tabs and trailing whitespace.
3806 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3808 TK/06 MBOX spool code: Add real "From " MBOX separator line
3809 so the .eml file is really in mbox format (even though
3810 most programs do not really care). Patch from Alex Miller.
3812 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3813 The latter is generated from $received_to and is only set if the
3814 message has one envelope recipient. SA can use these headers,
3815 obviously out-of-the-box. Patch from Alex Miller.
3817 PH/08 The ${def test on a variable was returning false if the variable's
3818 value was "0", contrary to what the specification has always said!
3819 The result should be true unless the variable is empty.
3821 PH/09 The syntax error of a character other than { following "${if
3822 def:variable_name" (after optional whitespace) was not being diagnosed.
3823 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3824 accidental colon was present, for example, could give incorrect results.
3826 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3827 result is used (not including appendfile, where other changes are about
3830 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3831 This involved changing a lot of size variables from int to off_t. It
3832 should work with maildirs and everything.
3834 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3835 spamd dying while we are connected to it.
3837 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3840 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3841 function for BDB 4.3.
3843 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3845 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3846 the sources for PCRE have been re-arranged and more files are now
3849 PH/15 The code I had for printing potentially long long variables in PH/11
3850 above was not the best (it lost precision). The length of off_t variables
3851 is now inspected at build time, and an appropriate printing format (%ld
3852 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3853 to be "long long int" or "long int". This is needed for the internal
3854 formatting function string_vformat().
3856 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3857 the configuration file to be ":syslog", then the script "guesses" where
3858 the logs files are, rather than using the compiled in default. In our
3859 case the guess is not the same as the compiled default, so the script
3860 suddenly stopped working when I started to use syslog. The patch checks
3861 to see if log_file_path is "". If so, it attempts to read it from exim
3862 with no configuration file to get the compiled in version, before it
3863 falls back to the previous guessing code."
3865 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3866 implementing BATV in an Exim configuration. See NewStuff for the gory
3869 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3870 Makefile that are specific to HP-UX.
3872 PH/18 If the "use_postmaster" option was set for a recipient callout together
3873 with the "random" option, the postmaster address was used as the MAIL
3874 FROM address for the random test, but not for the subsequent recipient
3875 test. It is now used for both.
3877 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3878 patch removes a few documentation additions to RFC 3028, because the
3879 latest draft now contains them. It adds the new en;ascii-case comparator
3880 and a new error check for 8bit text in MIME parts. Comparator and
3881 require names are now matched exactly. I enabled the subaddress
3882 extension, but it is not well tested yet (read: it works for me)."
3884 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3885 rework some of the code of TK/09 above to avoid the hardwired use of
3886 "%lld" and "long long". Replaced the call to snprintf() with a call to
3889 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3890 records point to non-existent hosts", "retry timeout exceeded", and
3891 "retry time not reached for any host after a long failure period".
3893 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3894 experimental DomainKeys support:
3896 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3897 (2) On an error such as an illegally used "control", the wrong name for
3898 the control was given.
3900 These problems did NOT occur unless DomainKeys support was compiled.
3902 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3904 PH/24 Added ${if match_ip condition.
3906 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3907 databases so that it will be absolutely obvious if a crash occurs in the
3908 DB library. This is a regular occurrence (often caused by mis-matched
3911 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3912 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3913 picked up on a user's system that detects such things. There doesn't seem
3914 to be a gcc warning option for this - only an attribute that has to be
3915 put on the function's prototype. It seems that in Fedora Core 4 they have
3916 set this on a number of new functions. No doubt there will be more in due
3919 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3920 specify a transport. However, if an address that was verified by such a
3921 router was the subject of a callout, Exim crashed because it tried to
3922 read the rcpt_include_affixes from the non-existent transport. Now it
3923 just assumes that the setting of that option is false. This bug was
3924 introduced by 4.51/PH/31.
3926 PH/28 Changed -d+all to exclude +memory, because that information is very
3927 rarely of interest, but it makes the output a lot bigger. People tend to
3928 do -d+all out of habit.
3930 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3931 code in os-type was giving problems when libc.so lives in lib64, like on
3934 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3935 aren't the modern standard, and it seems that some systems' include files
3936 don't always have them. Exim was already checking for some of the newer
3937 ones like T_AAAA, and defining it itself. I've added checks for all the
3938 record types that Exim uses.
3940 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3941 not automatically generating a new one, as it is supposed to. This
3942 prevented TLS from working. If the file did exist, but contained invalid
3943 data, a new version was generated, as expected. It was only the case of a
3944 non-existent file that was broken.
3946 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3947 with a change in libdomainkeys > 0.64.
3949 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3950 from DNS. If the selector record carries the flag, it now has
3951 precedence over the domain-wide flag.
3953 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3955 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3956 the use of a port name that isn't defined in /etc/services) occurred, the
3957 message was deferred as in a normal delivery, and thus remained on the
3958 spool, instead of being failed because of the mua_wrapper setting. This
3959 is now fixed, and I tidied up some of the mua_wrapper messages at the
3962 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3963 the messages in a hash of arrays rather than using individual hashes.
3964 This is a bit cleaner and results in dramatic memory savings, albeit
3965 at a slight CPU cost.
3967 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3968 as requested by Marc Sherman.
3970 SC/10 Eximstats: added histograms for user specified patterns as requested
3973 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3975 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3976 fopen() in the content-scanning modules that did not already have it.
3982 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3983 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3985 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3987 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3989 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3990 to test Sieve filters that use "vacation".
3992 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3993 that changes the way the GnuTLS parameters are stored in the cache file.
3994 The new format can be generated externally. For backward compatibility,
3995 if the data in the cache doesn't make sense, Exim assumes it has read an
3996 old-format file, and it generates new data and writes a new file. This
3997 means that you can't go back to an older release without removing the
4000 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4001 work if there are any delivery delays because "one_time" forces the
4002 parent to be marked "delivered", so its unseen clone is never tried
4003 again. For this reason, Exim now forbids the simultaneous setting of
4006 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4007 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4008 entry if you want to know the details. The fix, however, neglected to
4009 consider the case when local delivery batching is involved. The test for
4010 "previously delivered" was not happening when checking to see if an
4011 address could be batched with a previous (undelivered) one; under
4012 certain circumstances this could lead to multiple deliveries to the same
4015 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4016 in its include files, and this causes problems building Exim.
4018 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4019 header_syntax) but Exim was just ignoring anything given after a slash.
4020 In particular, this caused confusion with an attempt to use "verify =
4021 reverse_host_lookup/defer_ok". An error is now given when options are
4022 supplied for verify items that do not have them. (Maybe reverse_host_
4023 lookup should have a defer_ok option, but that's a different point.)
4025 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4026 defined by RFC 821) to 2048, because there were problems with some AUTH
4027 commands, and RFC 1869 says the size should be increased for extended
4028 SMTP commands that take arguments.
4030 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4033 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4034 "unknown" error; now it says that the functionality isn't in the binary.
4036 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4037 an address' error message when a string expansion fails (syntax or
4038 whatever). Otherwise the password may appear in the log. Following change
4039 PH/42 below, there is no longer a chance of it appearing in a bounce
4042 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4044 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4045 of its addresses was ever tried. (Bugzilla bug #2.)
4047 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4048 the result incorrectly in the debug output. (It correctly added a newline
4049 to what was transported.)
4051 TF/01 Added $received_time.
4053 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4054 commented out examples of how to interface to a virus scanner and to
4055 SpamAssassin. Also added commented examples of av_scanner and
4056 spamd_address settings.
4058 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4059 and controls are allowed in which ACLs. There were a couple of minor
4060 errors. Some of the entries in the conditions table (which is a table of
4061 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4062 negation of where the condition IS allowed.
4064 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4066 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4067 header file does not have a version number, so I've had to invent a new
4068 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4069 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4070 radiusclient), but it was contributed by a Radius user.
4072 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4073 files or directories, for OpenSSL.
4075 PH/19 When an Exim process that is running as root has to create an Exim log
4076 file, it does so in a subprocess that runs as exim:exim so as to get the
4077 ownership right at creation (otherwise, other Exim processes might see
4078 the file with the wrong ownership). There was no test for failure of this
4079 fork() call, which would lead to the process getting stuck as it waited
4080 for a non-existent subprocess. Forks do occasionally fail when resources
4081 run out. I reviewed all the other calls to fork(); they all seem to check
4084 PH/20 When checking for unexpected SMTP input at connect time (before writing
4085 the banner), Exim was not dealing correctly with a non-positive return
4086 from the read() function. If the client had disconnected by this time,
4087 the result was a log entry for a synchronization error with an empty
4088 string after "input=" when read() returned zero. If read() returned -1
4089 (an event I could not check), uninitialized data bytes were printed.
4090 There were reports of junk text (parts of files, etc) appearing after
4093 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4095 PH/22 Added support for macro redefinition, and (re)definition in between
4096 driver and ACL definitions.
4098 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4099 forgetting to use the resulting value; it was using the unexpanded value.
4101 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4102 hadn't been configured. The fix is from Juergen Kreileder, who
4103 understands it better than I do:
4105 "Here's what I see happening with three configured cyrus_sasl
4106 authenticators configured (plain, login, cram-md5):
4108 On startup auth_cyrus_sasl_init() gets called for each of these.
4109 This means three calls to sasl_listmech() without a specified mech_list.
4110 => SASL tests which mechs of all available mechs actually work
4111 => three warnings about OTP not working
4112 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4114 With the patch, sasl_listmech() also gets called three times. But now
4115 SASL's mech_list option is set to the server_mech specified in the the
4116 authenticator. Or in other words, the answer from sasl_listmech()
4117 gets limited to just the mech you're testing for (which is different
4119 => the return list contains just 'plain' or 'login', 'cram-md5' or
4120 nothing depending on the value of ob->server_mech.
4122 I've just tested the patch: Authentication still works fine,
4123 unavailable mechs specified in the exim configuration are still
4124 caught, and the auth.log warnings about OTP are gone."
4126 PH/25 When debugging is enabled, the contents of the command line are added
4127 to the debugging output, even when log_selector=+arguments is not
4130 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4131 answer is "GNU", and only if the return is "GNU/something" is the answer
4134 PH/27 $acl_verify_message is now set immediately after the failure of a
4135 verification in an ACL, and so is available in subsequent modifiers. In
4136 particular, the message can be preserved by coding like this:
4138 warn !verify = sender
4139 set acl_m0 = $acl_verify_message
4141 Previously, $acl_verify_message was set only while expanding "message"
4142 and "log_message" when a very denied access.
4144 PH/28 Modified OS/os.c-Linux with
4146 -#ifndef OS_LOAD_AVERAGE
4147 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4149 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4150 nomenclature these days.)
4152 PH/29 Installed patch from the Sieve maintainer that adds the options
4153 sieve_useraddress and sieve_subaddress to the redirect router.
4155 PH/30 In these circumstances:
4156 . Two addresses routed to the same list of hosts;
4157 . First host does not offer TLS;
4158 . First host accepts first address;
4159 . First host gives temporary error to second address;
4160 . Second host offers TLS and a TLS session is established;
4161 . Second host accepts second address.
4162 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4163 and peerdn, if requested) that were in fact used only for the second
4166 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4167 attention to any local part prefix or suffix that was matched by the
4168 router that accepted the address. It now behaves in the same way as it
4169 does for delivery: the affixes are removed from the local part unless
4170 rcpt_include_affixes is set on the transport.
4172 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4173 timeout during the DATA phase of an incoming message.
4175 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4176 applied a patch sanctioned by the Sieve maintainer.
4178 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4179 the uid or gid is negative. A case of a negative gid caused this to be
4180 noticed. The fix allows for either to be negative.
4182 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4183 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4186 PH/36 The change PH/12 above was broken. Fixed it.
4188 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4189 the grounds that routing the same address twice would always produce the
4190 same answer. This might have been true once, but it is certainly no
4191 longer true now. Routing a child address may depend on the previous
4192 routing that produced that child. Some complicated redirection strategies
4193 went wrong when messages had multiple recipients, and made Exim's
4194 behaviour dependent on the order in which the addresses were given.
4196 I have moved the duplicate checking until after the routing is complete.
4197 Exim scans the addresses that are assigned to local and remote
4198 transports, and removes any duplicates. This means that more work will be
4199 done, as duplicates will always all be routed, but duplicates are
4200 presumably rare, so I don't expect this is of any significance.
4202 For deliveries to pipes, files, and autoreplies, the duplicate checking
4203 still happens during the routing process, since they are not going to be
4206 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4207 It corrects a timeout issue with spamd. This is Ian's comment: "The
4208 background is that sometimes spamd either never reads data from a
4209 connection it has accepted, or it never writes response data. The exiscan
4210 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4211 blindly assumes that writes won't block so it may never time out."
4213 PH/39 Allow G after quota size as well as K and M.
4215 PH/40 The value set for $authenticated_id in an authenticator may not contain
4216 binary zeroes or newlines because the value is written to log lines and
4217 to spool files. There was no check on this. Now the value is run through
4218 the string_printing() function so that such characters are converted to
4219 printable escape sequences.
4221 PH/41 $message_linecount is a new variable that contains the total number of
4222 lines in the message. Compare $body_linecount, which is the count for the
4225 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4226 bounce and delay warning messages, except in certain special cases, which
4229 (a) An SMTP error message from a remote host;
4230 (b) A message specified in a :fail: redirection;
4231 (c) A message specified in a "fail" command in a system filter;
4232 (d) A message specified in a FAIL return from the queryprogram router;
4233 (e) A message specified by the cannot_route_message router option.
4235 In these cases only, Exim does include the error details in bounce and
4236 warning messages. There are also a few cases where bland messages such
4237 as "unrouteable address" or "local delivery error" are given.
4239 PH/43 $value is now also set for the "else" part of a ${run expansion.
4241 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4242 being worked on, but at least Exim now implements the latest version to
4245 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4246 process to complete was treated as a delivery failure, a timeout while
4247 writing the message to the pipe was logged, but erroneously treated as a
4248 successful delivery. Such timeouts include transport filter timeouts. For
4249 consistency with the overall process timeout, these timeouts are now
4250 treated as errors, giving rise to delivery failures by default. However,
4251 there is now a new Boolean option for the pipe transport called
4252 timeout_defer, which, if set TRUE, converts the failures into defers for
4253 both kinds of timeout. A transport filter timeout is now identified in
4256 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4257 systems where "make" and "gmake" are different, calling "gmake" at top
4258 level broke things. I've arranged for the value of $(MAKE) to be passed
4259 from the Makefile to this script so that it can call the same version of
4263 A note about Exim versions 4.44 and 4.50
4264 ----------------------------------------
4266 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4267 changes of various kinds. As a consequence, a big documentation update was
4268 needed. This delayed the release for rather longer than seemed good, especially
4269 in the light of a couple of (minor) security issues. Therefore, the changes
4270 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4271 release. So 4.44 and 4.50 are in effect two different branches that both start
4274 I have left the 4.50 change log unchanged; it contains all the changes since
4275 4.43. The change log for 4.44 is below; many of its items are identical to
4276 those for 4.50. This seems to be the most sensible way to preserve the
4277 historical information.
4283 1. Minor wording change to the doc/README.SIEVE file.
4285 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4286 computation of the current number of files was incorrect.
4288 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4289 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4292 4. Give more explanation in the error message when the command for a transport
4293 filter fails to execute.
4295 5. There are several places where Exim runs a non-Exim command in a
4296 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4297 was being done only for the command run by the queryprogram router. It is
4298 now done for all such subprocesses. The other cases are: ${run, transport
4299 filters, and the commands run by the lmtp and pipe transports.
4301 6. Added CONFIGURE_GROUP build-time option.
4303 7. Some older OS have a limit of 256 on the maximum number of file
4304 descriptors. Exim was using setrlimit() to set 1000 as a large value
4305 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4306 systems. I've change it so that if it can't get 1000, it tries for 256.
4308 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4309 was an oversight, and furthermore, ever since the addition of extra
4310 controls (e.g. 4.43/32), the checks on when to allow different forms of
4311 "control" were broken. There should now be diagnostics for all cases when a
4312 control that does not make sense is encountered.
4314 9. Added the /retain_sender option to "control=submission".
4316 10. $recipients is now available in the predata ACL (oversight).
4318 11. Tidy the search cache before the fork to do a delivery from a message
4319 received from the command line. Otherwise the child will trigger a lookup
4320 failure and thereby defer the delivery if it tries to use (for example) a
4321 cached ldap connection that the parent has called unbind on.
4323 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4324 of $address_data from the recipient verification was clobbered by the
4325 sender verification.
4327 13. The value of address_data from a sender verification is now available in
4328 $sender_address_data in subsequent conditions in the ACL statement.
4330 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4332 15. Added a new option "connect=<time>" to callout options, to set a different
4335 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4336 was its contents. (It was OK if the option was not defined at all.)
4338 17. A "Completed" log line is now written for messages that are removed from
4339 the spool by the -Mrm option.
4341 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4342 information about exactly what failed.
4344 19. Added -dd to debug only the daemon process.
4346 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4347 handles timeouts, both on the server side and network timeouts. Renamed the
4348 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4350 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4351 It is now set to "smtps".
4353 22. $host_address is now set to the target address during the checking of
4354 ignore_target_hosts.
4356 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4357 being passed; this would have caused $sender_host_name to have been used if
4358 matching the list had actually called for a host name (not very likely,
4359 since this list is usually IP addresses). A host name is now passed as
4362 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4363 code that specifies a non-restarting handler (typically sigaction() in
4364 modern systems) in an attempt to fix a rare and obscure crash bug.
4366 25. Narrowed the window for a race in the daemon that could cause it to ignore
4367 SIGCHLD signals. This is not a major problem, because they are used only to
4368 wake it up if nothing else does.
4370 26. A malformed maildirsize file could cause Exim to calculate negative values
4371 for the mailbox size or file count. Odd effects could occur as a result.
4372 The maildirsize information is now recalculated if the size or filecount
4375 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4376 support for a long time. Removed HAVE_SYS_VFS_H.
4378 28. Installed the latest version of exipick from John Jetmore.
4380 29. In an address list, if the pattern was not a regular expression, an empty
4381 subject address (from a bounce message) matched only if the pattern was an
4382 empty string. Non-empty patterns were not even tested. This was the wrong
4383 because it is perfectly reasonable to use an empty address as part of a
4384 database query. An empty address is now tested by patterns that are
4385 lookups. However, all the other forms of pattern expect the subject to
4386 contain a local part and a domain, and therefore, for them, an empty
4387 address still always fails if the pattern is not itself empty.
4389 30. Exim went into a mad DNS loop when attempting to do a callout where the
4390 host was specified on an smtp transport, and looking it up yielded more
4391 than one IP address.
4393 31. Re-factored the code for checking spool and log partition space into a
4394 function that finds that data and another that does the check. The former
4395 is then used to implement four new variables: $spool_space, $log_space,
4396 $spool_inodes, and $log_inodes.
4398 32. The RFC2047 encoding function was originally intended for short strings
4399 such as real names; it was not keeping to the 75-character limit for
4400 encoded words that the RFC imposes. It now respects the limit, and
4401 generates multiple encoded words if necessary. To be on the safe side, I
4402 have increased the buffer size for the ${rfc2047: expansion operator from
4405 33. It is now permitted to omit both strings after an "if" condition; if the
4406 condition is true, the result is "true". As before, when the second string
4407 is omitted, a false condition yields an empty string. This makes it less
4408 cumbersome to write custom ACL and router conditions.
4410 34. Failure to deliver a bounce message always caused it to be frozen, even if
4411 there was an errors_to setting on the router. The errors_to setting is now
4414 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4415 canonical form (fully expanded) before being placed in
4416 $sender_host_address.
4418 36. The table in the code that translates DNS record types into text (T_A to
4419 "A" for instance) was missing entries for NS and CNAME. It is just possible
4420 that this could have caused confusion if both these types were looked up
4421 for the same domain, because the text type is used as part of Exim's
4422 per-process caching. But the chance of anyone hitting this buglet seems
4425 37. The dnsdb lookup has been extended in a number of ways.
4427 (1) There is a new type, "zns", which walks up the domain tree until it
4428 finds some nameserver records. It should be used with care.
4430 (2) There is a new type, "mxh", which is like "mx" except that it returns
4431 just the host names, not the priorities.
4433 (3) It is now possible to give a list of domains (or IP addresses) to be
4434 looked up. The behaviour when one of the lookups defers can be
4435 controlled by a keyword.
4437 (4) It is now possible to specify the separator character for use when
4438 multiple records are returned.
4440 38. The dnslists ACL condition has been extended: it is now possible to supply
4441 a list of IP addresses and/or domains to be looked up in a particular DNS
4444 39. Added log_selector=+queue_time_overall.
4446 40. When running the queue in the test harness, wait just a tad after forking a
4447 delivery process, to get repeatability of debugging output.
4449 41. Include certificate and key file names in error message when GnuTLS fails
4450 to set them up, because the GnuTLS error message doesn't include the name
4451 of the failing file when there is a problem reading it.
4453 42. Allow both -bf and -bF in the same test run.
4455 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4457 44. The "Exiscan patch" is now merged into the mainline Exim source.
4459 45. Sometimes the final signoff response after QUIT could fail to get
4460 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4461 < 0 before doing a fflush(). This bug looks as though it goes back to the
4462 introduction of TLS in release 3.20, but "sometimes" must have been rare
4463 because the tests only now provoked it.
4465 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4466 (this can affect the format of dates).
4468 47. exim_tidydb, when checking for the continued existence of a message for
4469 which it has found a message-specific retry record, was not finding
4470 messages that were in split spool directories. Consequently, it was
4471 deleting retry records that should have stayed in existence.
4473 48. Steve fixed some bugs in eximstats.
4475 49. The SPA authentication driver was not abandoning authentication and moving
4476 on to the next authenticator when an expansion was forced to fail,
4477 contradicting the general specification for all authenticators. Instead it
4478 was generating a temporary error. It now behaves as specified.
4480 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4481 (the order specifies the preference for clients). The order is now AES256,
4482 AES128, 3DES, ARCFOUR128.
4484 51. Small patch to Sieve code - explicitly set From: when generating an
4487 52. Exim crashed if a remote delivery caused a very long error message to be
4488 recorded - for instance if somebody sent an entire SpamAssassin report back
4489 as a large number of 550 error lines. This bug was coincidentally fixed by
4490 increasing the size of one of Exim's internal buffers (big_buffer) that
4491 happened as part of the Exiscan merge. However, to be on the safe side, I
4492 have made the code more robust (and fixed the comments that describe what
4495 53. Now that there can be additional text after "Completed" in log lines (if
4496 the queue_time_overall log selector is set), a one-byte patch to exigrep
4497 was needed to allow it to recognize "Completed" as not the last thing in
4500 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4501 patch that reportedly fixes this has been added. I am not expert enough to
4502 create a test for it. This is what the patch creator wrote:
4504 "I found a little strange behaviour of ldap code when working with
4505 Windows 2003 AD Domain, where users was placed in more than one
4506 Organization Units. When I tried to give exim partial DN, the exit code
4507 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4508 But simultaneously result of request was absolutely normal ldap result,
4509 so I produce this patch..."
4511 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4512 REFERENCE, so I have modified the code to exclude the patch when that macro
4515 55. Some experimental protocols are using DNS PTR records for new purposes. The
4516 keys for these records are domain names, not reversed IP addresses. The
4517 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4518 leaves it alone. Component reversal etc. now happens only for IP addresses.
4521 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4523 57. Double the size of the debug message buffer (to 2048) so that more of very
4524 long debug lines gets shown.
4526 58. The exicyclog utility now does better if the number of log files to keep
4527 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4529 59. Two changes related to the smtp_active_hostname option:
4531 (1) $smtp_active_hostname is now available as a variable.
4532 (2) The default for smtp_banner uses $smtp_active_hostname instead
4533 of $primary_hostname.
4535 60. The host_aton() function is supposed to be passed a string that is known
4536 to be a valid IP address. However, in the case of IPv6 addresses, it was
4537 not checking this. This is a hostage to fortune. Exim now panics and dies
4538 if the condition is not met. A case was found where this could be provoked
4539 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4540 components; fortuitously, this particular loophole had already been fixed
4541 by change 4.50/55 above.
4543 If there are any other similar loopholes, the new check in host_aton()
4544 itself should stop them being exploited. The report I received stated that
4545 data on the command line could provoke the exploit when Exim was running as
4546 exim, but did not say which command line option was involved. All I could
4547 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4548 running as the user.
4551 61. There was a buffer overflow vulnerability in the SPA authentication code
4552 (which came originally from the Samba project). I have added a test to the
4553 spa_base64_to_bits() function which I hope fixes it.
4556 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4557 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4559 63. The daemon start-up calls getloadavg() while still root for those OS that
4560 need the first call to be done as root, but it missed one case: when
4561 deliver_queue_load_max is set with deliver_drop_privilege. This is
4562 necessary for the benefit of the queue runner, because there is no re-exec
4563 when deliver_drop_privilege is set.
4565 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4566 This has been fixed.
4568 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4569 was in use, was not putting the data itself into the right store pool;
4570 consequently, it could be overwritten for a subsequent message in the same
4571 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4574 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4576 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4577 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4578 the function were treating the return as a boolean value, which happened to
4579 work because 0=false and not-0=true, but is not correct code.
4581 68. The host_aton() function was not handling scoped IPv6 addresses (those
4582 with, for example, "%eth0" on the end) correctly.
4584 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4585 negated items (that is, ~something) in unsigned ints. Some compilers
4586 apparently mutter when there is no cast.
4588 70. If an address verification called from an ACL failed, and did not produce a
4589 user-specific message (i.e. there was only a "system" message), nothing was
4590 put in $acl_verify_message. In this situation, it now puts the system
4593 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4594 session; change 4.31/43 added the unwanted input to the log line - except
4595 that it did not do this in the start of session case. It now does.
4597 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4598 This is wrong and can cause the other end to generate a synchronization
4599 error if it is another Exim or anything else that does the synchronization
4600 check. A QUIT command is no longer sent after a timeout.
4602 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4603 during host lookups.
4605 74. The defer_ok option of callout verification was not working if it was used
4606 when verifying addresses in header lines, that is, for this case:
4608 verify = header_sender/callout=defer_ok
4610 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4611 those file descriptors could be used for SMTP connections. If anything
4612 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4613 could be sent to the SMTP client, causing chaos. The daemon now opens
4614 stdin, stdout, and stderr to /dev/null when it puts itself into the
4617 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4618 log by default. The user can override this with suitable Perl magic.
4620 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4621 the log message when discard triggers, was not working for the DATA ACL or
4622 for the non-SMTP ACL.
4624 78. Error message wording change in sieve.c.
4626 79. If smtp_accept_max_per_host was set, the number of connections could be
4627 restricted to fewer than expected, because the daemon was trying to set up
4628 a new connection before checking whether the processes handling previous
4629 connections had finished. The check for completed processes is now done
4630 earlier. On busy systems, this bug wouldn't be noticed because something
4631 else would have woken the daemon, and it would have reaped the completed
4634 80. If a message was submitted locally by a user whose login name contained one
4635 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4636 It caused a spool format error. I have fixed the spool reading code. A
4637 related problem was that the "from" clause in the Received: line became
4638 illegal because of the space(s). It is now covered by ${quote_local_part.
4640 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4643 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4645 83. Patch from Sieve maintainer to fix unterminated string problem in
4646 "vacation" handling.
4648 84. Some minor changes to the Linux configuration files to help with other
4649 OS variants using glibc.
4651 85. One more patch for Sieve to update vacation handling to latest spec.
4654 ----------------------------------------------------
4655 See the note above about the 4.44 and 4.50 releases.
4656 ----------------------------------------------------
4662 1. Change 4.43/35 introduced a bug that caused file counts to be
4663 incorrectly computed when quota_filecount was set in an appendfile
4666 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4667 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4670 3. Give more explanation in the error message when the command for a transport
4671 filter fails to execute.
4673 4. There are several places where Exim runs a non-Exim command in a
4674 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4675 was being done only for the command run by the queryprogram router. It is
4676 now done for all such subprocesses. The other cases are: ${run, transport
4677 filters, and the commands run by the lmtp and pipe transports.
4679 5. Some older OS have a limit of 256 on the maximum number of file
4680 descriptors. Exim was using setrlimit() to set 1000 as a large value
4681 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4682 systems. I've change it so that if it can't get 1000, it tries for 256.
4684 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4685 was an oversight, and furthermore, ever since the addition of extra
4686 controls (e.g. 4.43/32), the checks on when to allow different forms of
4687 "control" were broken. There should now be diagnostics for all cases when a
4688 control that does not make sense is encountered.
4690 7. $recipients is now available in the predata ACL (oversight).
4692 8. Tidy the search cache before the fork to do a delivery from a message
4693 received from the command line. Otherwise the child will trigger a lookup
4694 failure and thereby defer the delivery if it tries to use (for example) a
4695 cached ldap connection that the parent has called unbind on.
4697 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4698 of $address_data from the recipient verification was clobbered by the
4699 sender verification.
4701 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4702 was its contents. (It was OK if the option was not defined at all.)
4704 11. A "Completed" log line is now written for messages that are removed from
4705 the spool by the -Mrm option.
4707 12. $host_address is now set to the target address during the checking of
4708 ignore_target_hosts.
4710 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4711 being passed; this would have caused $sender_host_name to have been used if
4712 matching the list had actually called for a host name (not very likely,
4713 since this list is usually IP addresses). A host name is now passed as
4716 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4717 code that specifies a non-restarting handler (typically sigaction() in
4718 modern systems) in an attempt to fix a rare and obscure crash bug.
4720 15. Narrowed the window for a race in the daemon that could cause it to ignore
4721 SIGCHLD signals. This is not a major problem, because they are used only to
4722 wake it up if nothing else does.
4724 16. A malformed maildirsize file could cause Exim to calculate negative values
4725 for the mailbox size or file count. Odd effects could occur as a result.
4726 The maildirsize information is now recalculated if the size or filecount
4729 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4730 support for a long time. Removed HAVE_SYS_VFS_H.
4732 18. Updated exipick to current release from John Jetmore.
4734 19. Allow an empty sender to be matched against a lookup in an address list.
4735 Previously the only cases considered were a regular expression, or an
4738 20. Exim went into a mad DNS lookup loop when doing a callout where the
4739 host was specified on the transport, if the DNS lookup yielded more than
4742 21. The RFC2047 encoding function was originally intended for short strings
4743 such as real names; it was not keeping to the 75-character limit for
4744 encoded words that the RFC imposes. It now respects the limit, and
4745 generates multiple encoded words if necessary. To be on the safe side, I
4746 have increased the buffer size for the ${rfc2047: expansion operator from
4749 22. Failure to deliver a bounce message always caused it to be frozen, even if
4750 there was an errors_to setting on the router. The errors_to setting is now
4753 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4754 canonical form (fully expanded) before being placed in
4755 $sender_host_address.
4757 24. Updated eximstats to version 1.33
4759 25. Include certificate and key file names in error message when GnuTLS fails
4760 to set them up, because the GnuTLS error message doesn't include the name
4761 of the failing file when there is a problem reading it.
4763 26. Expand error message when OpenSSL has problems setting up cert/key files.
4766 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4767 (this can affect the format of dates).
4769 28. exim_tidydb, when checking for the continued existence of a message for
4770 which it has found a message-specific retry record, was not finding
4771 messages that were in split spool directories. Consequently, it was
4772 deleting retry records that should have stayed in existence.
4774 29. eximstats updated to version 1.35
4775 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4776 1.35 - bugfix such that pie charts by volume are generated correctly
4778 30. The SPA authentication driver was not abandoning authentication and moving
4779 on to the next authenticator when an expansion was forced to fail,
4780 contradicting the general specification for all authenticators. Instead it
4781 was generating a temporary error. It now behaves as specified.
4783 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4784 (the order specifies the preference for clients). The order is now AES256,
4785 AES128, 3DES, ARCFOUR128.
4787 31. Small patch to Sieve code - explicitly set From: when generating an
4790 32. Exim crashed if a remote delivery caused a very long error message to be
4791 recorded - for instance if somebody sent an entire SpamAssassin report back
4792 as a large number of 550 error lines. This bug was coincidentally fixed by
4793 increasing the size of one of Exim's internal buffers (big_buffer) that
4794 happened as part of the Exiscan merge. However, to be on the safe side, I
4795 have made the code more robust (and fixed the comments that describe what
4798 33. Some experimental protocols are using DNS PTR records for new purposes. The
4799 keys for these records are domain names, not reversed IP addresses. The
4800 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4801 leaves it alone. Component reversal etc. now happens only for IP addresses.
4804 34. The host_aton() function is supposed to be passed a string that is known
4805 to be a valid IP address. However, in the case of IPv6 addresses, it was
4806 not checking this. This is a hostage to fortune. Exim now panics and dies
4807 if the condition is not met. A case was found where this could be provoked
4808 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4809 components; fortuitously, this particular loophole had already been fixed
4810 by change 4.50/55 or 4.44/33 above.
4812 If there are any other similar loopholes, the new check in host_aton()
4813 itself should stop them being exploited. The report I received stated that
4814 data on the command line could provoke the exploit when Exim was running as
4815 exim, but did not say which command line option was involved. All I could
4816 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4817 running as the user.
4820 35. There was a buffer overflow vulnerability in the SPA authentication code
4821 (which came originally from the Samba project). I have added a test to the
4822 spa_base64_to_bits() function which I hope fixes it.
4825 36. The daemon start-up calls getloadavg() while still root for those OS that
4826 need the first call to be done as root, but it missed one case: when
4827 deliver_queue_load_max is set with deliver_drop_privilege. This is
4828 necessary for the benefit of the queue runner, because there is no re-exec
4829 when deliver_drop_privilege is set.
4831 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4832 was in use, was not putting the data itself into the right store pool;
4833 consequently, it could be overwritten for a subsequent message in the same
4834 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4837 38. Sometimes the final signoff response after QUIT could fail to get
4838 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4839 < 0 before doing a fflush(). This bug looks as though it goes back to the
4840 introduction of TLS in release 3.20, but "sometimes" must have been rare
4841 because the tests only now provoked it.
4847 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4848 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4849 Now it returns an integer. A number of calls were still expecting a T/F
4850 return. Fortuitously, in all cases, the tests worked in OK situations,
4851 which is the norm. However, things would have gone wrong on any write
4852 failures on the smtp file descriptor. This function is used when sending
4853 messages over SMTP and also when doing verify callouts.
4855 2. When Exim is called to do synchronous delivery of a locally submitted
4856 message (the -odf or -odi options), it no longer closes stderr before doing
4859 3. Implemented the mua_wrapper option.
4861 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4863 5. Implemented the functions header_remove(), header_testname(),
4864 header_add_at_position(), and receive_remove_recipient(), and exported them
4867 6. If an ACL "warn" statement specified the addition of headers, Exim already
4868 inserted X-ACL-Warn: at the start if there was no header name. However, it
4869 was not making this test for the second and subsequent header lines if
4870 there were newlines in the string. This meant that an invalid header could
4871 be inserted if Exim was badly configured.
4873 7. Allow an ACL "warn" statement to add header lines at the start or after all
4874 the Received: headers, as well as at the end.
4876 8. Added the rcpt_4xx retry error code.
4878 9. Added postmaster_mailfrom=xxx to callout verification option.
4880 10. Added mailfrom=xxxx to the callout verification option, for verify=
4883 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4884 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4886 12. Inserted some casts to stop certain compilers warning when using pointer
4887 differences as field lengths or precisions in printf-type calls (mostly
4888 affecting debugging statements).
4890 13. Added optional readline() support for -be (dynamically loaded).
4892 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4893 same clock tick as a message's arrival, so that its received time was the
4894 same as the "first fail" time on the retry record, and that message
4895 remained on the queue past the ultimate address timeout, every queue runner
4896 would try a delivery (because it was past the ultimate address timeout) but
4897 after another failure, the ultimate address timeout, which should have then
4898 bounced the address, did not kick in. This was a "< instead of <=" error;
4899 in most cases the first failure would have been in the next clock tick
4900 after the received time, and all would be well.
4902 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4903 being recognized when the domain list was tested by the match_domain
4904 condition in an expansion string.
4906 16. Added the ${str2b64: operator.
4908 17. Exim was always calling setrlimit() to set a large limit for the number of
4909 processes, without checking whether the existing limit was already
4910 adequate. (It did check for the limit on file descriptors.) Furthermore,
4911 errors from getrlimit() and setrlimit() were being ignored. Now they are
4912 logged to the main and panic logs, but Exim does carry on, to try to do its
4913 job under whatever limits there are.
4915 18. Imported PCRE 5.0.
4917 19. Trivial typo in log message " temporarily refused connection" (the leading
4920 20. If the log selector return_path_on_delivery was set and an address was
4921 redirected to /dev/null, the delivery process crashed because it assumed
4922 that a return path would always be set for a "successful" delivery. In this
4923 case, the whole delivery is bypassed as an optimization, and therefore no
4926 21. Internal re-arrangement: the function for sending a challenge and reading
4927 a response while authentication was assuming a zero-terminated challenge
4928 string. It's now changed to take a pointer and a length, to allow for
4929 binary data in such strings.
4931 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4933 23. Exim was not respecting finduser_retries when seeking the login of the
4934 uid under which it was called; it was always trying 10 times. (The default
4935 setting of finduser_retries is zero.) Also, it was sleeping after the final
4936 failure, which is pointless.
4938 24. Implemented tls_on_connect_ports.
4940 25. Implemented acl_smtp_predata.
4942 26. If the domain in control=submission is set empty, Exim assumes that the
4943 authenticated id is a complete email address when it generates From: or
4944 Sender: header lines.
4946 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4947 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4948 chown and chgrp in /bin and hostname in /usr/bin.
4950 28. Exim was keeping the "process log" file open after each use, just as it
4951 does for the main log. This opens the possibility of it remaining open for
4952 long periods when the USR1 signal hits a daemon. Occasional processlog
4953 errors were reported, that could have been caused by this. Anyway, it seems
4954 much more sensible not to leave this file open at all, so that is what now
4957 29. The long-running daemon process does not normally write to the log once it
4958 has entered its main loop, and it closes the log before doing so. This is
4959 so that log files can straightforwardly be renamed and moved. However,
4960 there are a couple of unusual error situations where the daemon does write
4961 log entries, and I had neglected to close the log afterwards.
4963 30. The text of an SMTP error response that was received during a remote
4964 delivery was being truncated at 512 bytes. This is too short for some of
4965 the long messages that one sometimes sees. I've increased the limit to
4968 31. It is now possible to make retry rules that apply only when a message has a
4969 specific sender, in particular, an empty sender.
4971 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4972 it possible to be selective about when SMTP synchronization is enforced.
4974 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4976 32. Implemented hosts_connection_nolog.
4978 33. Added an ACL for QUIT.
4980 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4983 35. Added mailbox_size and mailbox_filecount to appendfile.
4985 36. Added control = no_multiline_responses to ACLs.
4987 37. There was a bug in the logic of the code that waits for the clock to tick
4988 in the case where the clock went backwards by a substantial amount such
4989 that the microsecond fraction of "now" was more than the microsecond
4990 fraction of "then" (but the whole seconds number was less).
4992 38. Added support for the libradius Radius client library this is found on
4993 FreeBSD (previously only the radiusclient library was supported).
4999 1. When certain lookups returned multiple values in the form name=value, the
5000 quoting of the values was not always being done properly. Specifically:
5001 (a) If the value started with a double quote, but contained no whitespace,
5003 (b) If the value contained whitespace other than a space character (i.e.
5004 tabs or newlines or carriage returns) it was not quoted.
5005 This fix has been applied to the mysql and pgsql lookups by writing a
5006 separate quoting function and calling it from the lookup code. The fix
5007 should probably also be applied to nisplus, ibase and oracle lookups, but
5008 since I cannot test any of those, I have not disturbed their existing code.
5010 2. A hit in the callout cache for a specific address caused a log line with no
5011 reason for rejecting RCPT. Now it says "Previous (cached) callout
5012 verification failure".
5014 3. There was an off-by-one bug in the queryprogram router. An over-long
5015 return line was truncated at 256 instead of 255 characters, thereby
5016 overflowing its buffer with the terminating zero. As well as fixing this, I
5017 have increased the buffer size to 1024 (and made a note to document this).
5019 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5020 when Exim is waiting for an SMTP response from a remote server, Exim
5021 restarts its select() call on the socket, thereby resetting its timeout.
5022 This is not a problem when such interrupts are rare. Somebody set up a cron
5023 job to run exiwhat every 2 minutes, which is less than the normal select()
5024 timeout (5 or 10 minutes). This meant that the select() timeout never
5025 kicked in because it was always reset. I have fixed this by comparing the
5026 time when an interrupt arrives with the time at the start of the first call
5027 to select(). If more time than the timeout has elapsed, the interrupt is
5028 treated as a timeout.
5030 5. Some internal re-factoring in preparation for the addition of Sieve
5031 extensions (by MH). In particular, the "personal" test is moved to a
5032 separate function, and given an option for scanning Cc: and Bcc: (which is
5033 not set for Exim filters).
5035 6. When Exim created an email address using the login of the caller as the
5036 local part (e.g. when creating a From: or Sender: header line), it was not
5037 quoting the local part when it contained special characters such as @.
5039 7. Installed new OpenBSD configuration files.
5041 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5042 try to make them clearer.
5044 9. Callout options, other than the timeout value, were being ignored when
5045 verifying sender addresses in header lines. For example, when using
5047 verify = header_sender/callout=no_cache
5049 the cache was (incorrectly) being used.
5051 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5052 only the Cygwin environment.
5054 11. When return_path_on_delivery was set as a log selector, if different remote
5055 addresses in the same message used different return paths and parallel
5056 remote delivery occurred, the wrong values would sometimes be logged.
5057 (Whenever a remote delivery process finished, the return path value from
5058 the most recently started remote delivery process was logged.)
5060 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5061 lines when AUTH and/or TLS are in use. This is the "received protocol"
5062 field. Exim used to use "asmtp" for authenticated SMTP, without any
5063 indication (in the protocol name) for TLS use. Now it follows the RFC and
5064 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5065 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5066 names appear in log lines as well as in Received: header lines.
5068 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5069 extensions, and comparison tests, and to fix some bugs.
5071 14. Changes to the "personal" filter test:
5073 (1) The test was buggy in that it was just doing the equivalent of
5074 "contains" tests on header lines. For example, if a user's address was
5075 anne@some.where, the "personal" test would incorrectly be true for
5077 To: susanne@some.where
5079 This test is now done by extracting each address from the header in turn,
5080 and checking the entire address. Other tests that are part of "personal"
5081 are now done using regular expressions (for example, to check local parts
5082 of addresses in From: header lines).
5084 (2) The list of non-personal local parts in From: addresses has been
5085 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5086 taken from the Sieve specification recommendations.
5088 (3) If the message contains any header line starting with "List-" it is
5089 treated as non-personal.
5091 (4) The test for "circular" in the Subject: header line has been removed
5092 because it now seems ill-conceived.
5094 15. Minor typos in src/EDITME comments corrected.
5096 16. Installed latest exipick from John Jetmore.
5098 17. If headers_add on a router specified a text string that was too long for
5099 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5100 of string_sprintf() is now avoided.
5102 18. $message_body_size was not set (it was always zero) when running the DATA
5103 ACL and the local_scan() function.
5105 19. For the "mail" command in an Exim filter, no default was being set for
5106 the once_repeat time, causing a random time value to be used if "once" was
5107 specified. (If the value happened to be <= 0, no repeat happened.) The
5108 default is now 0s, meaning "never repeat". The "vacation" command was OK
5109 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5110 (I found it when inspecting the code).
5112 20. There is now an overall timeout for performing a callout verification. It
5113 defaults to 4 times the callout timeout, which applies to individual SMTP
5114 commands during the callout. The overall timeout applies when there is more
5115 than one host that can be tried. The timeout is checked before trying the
5116 next host. This prevents very long delays if there are a large number of
5117 hosts and all are timing out (e.g. when the network connections are timing
5118 out). The value of the overall timeout can be changed by specifying an
5119 additional sub-option for "callout", called "maxwait". For example:
5121 verify = sender/callout=5s,maxwait=20s
5123 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5124 to the end before writing, but this should make it even safer).
5126 22. Exim was forgetting that it had advertised PIPELINING for the second and
5127 subsequent messages on an SMTP connection. It was also not resetting its
5128 memory on STARTTLS and an internal HELO.
5130 23. When Exim logs an SMTP synchronization error within a session, it now
5131 records whether PIPELINING has been advertised or not.
5133 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5134 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5135 rather than long int.
5137 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5139 26. Added the never_mail option to autoreply.
5145 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5146 crash if the getsockname() call failed; this can happen if a connection is
5147 closed very soon after it is established. The problem was simply in the
5148 order in which certain operations were done, causing Exim to try to write
5149 to the SMTP stream before it had set up the file descriptor. The bug has
5150 been fixed by making things happen in the correct order.
5156 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5157 before the connection was closed, thus losing the rejection response.
5159 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5160 some early Solaris releases, but causes trouble in current releases where
5161 socklen_t is defined.
5163 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5166 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5169 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5170 that releasing the top of it at the end releases what was used for sub-
5171 expansions (unless the block got too big). However, discard this block if
5172 the first thing is a variable or header, so that we can use its block when
5173 it is dynamic (useful for very large $message_headers, for example).
5175 6. Lookups now cache *every* query, not just the most recent. A new, separate
5176 store pool is used for this. It can be recovered when all lookup caches are
5177 flushed. Lookups now release memory at the end of their result strings.
5178 This has involved some general refactoring of the lookup sources.
5180 7. Some code has been added to the store_xxx() functions to reduce the amount
5181 of flapping under certain conditions.
5183 8. log_incoming_interface used to affect only the <= reception log lines. Now
5184 it causes the local interface and port to be added to several more SMTP log
5185 lines, for example "SMTP connection from", and rejection lines.
5187 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5189 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5191 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5192 could be overwritten at the end of the current message (or the start of a
5193 new message if it was set in a HELO ACL). The value is now preserved for
5194 the duration of the SMTP connection.
5196 12. If a transport had a headers_rewrite setting, and a matching header line
5197 contained an unqualified address, that address was qualified, even if it
5198 did not match any rewriting rules. The underlying bug was that the values
5199 of the flags that permit the existence of unqualified sender and recipient
5200 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5201 non-local messages, and by -bnq for local messages) were not being
5202 preserved with the message after it was received.
5204 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5205 "next input=" as part of the text comprising the host identity instead of
5206 the correct text. The code was using the same buffer for two different
5207 strings. However, depending on which order the printing function evaluated
5208 its arguments, the bug did not always show up. Under Linux, for example, my
5209 test suite worked just fine.
5211 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5212 doesn't work with some older versions of Perl. It has been changed to "my",
5213 which in any case is probably the better facility to use.
5215 15. A really picky compiler found some instances of statements for creating
5216 error messages that either had too many or two few arguments for the format
5219 16. The size of the buffer for calls to the DNS resolver has been increased
5220 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5221 for addresses that have a lot of PTR records. This alleviates a problem; it
5222 does not fully solve it.
5224 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5225 buffer now truncates the list and logs the incident, which is the same
5226 action as happens when Exim is looking up a host name and its aliases.
5227 Previously in this situation something unpredictable would happen;
5228 sometimes it was "internal error: store_reset failed".
5230 18. If a server dropped the connection unexpectedly when an Exim client was
5231 using GnuTLS and trying to read a response, the client delivery process
5232 crashed while trying to generate an error log message.
5234 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5235 string, for example:
5237 warn message = H1: something\nH2: something
5239 the text was added as a single header line from Exim's point of view
5240 though it ended up OK in the delivered message. However, searching for the
5241 second and subsequent header lines using $h_h2: did not work. This has been
5242 fixed. Similarly, if a system filter added multiple headers in this way,
5243 the routers could not see them.
5245 20. Expanded the error message when iplsearch is called with an invalid key to
5246 suggest using net-iplsearch in a host list.
5248 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5249 ACLs are no longer actually imposed (and a message to that effect is
5252 22. If a "gecos" field in a passwd entry contained escaped characters, in
5253 particular, if it contained a \" sequence, Exim got it wrong when building
5254 a From: or a Sender: header from that name. A second bug also caused
5255 incorrect handling when an unquoted " was present following a character
5256 that needed quoting.
5258 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5259 was not being matched caselessly.
5261 24. Arranged for all hyphens in the exim.8 source to be escaped with
5264 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5265 properly. Recipient callout cache records were still being keyed to include
5266 the sender, even when use_sender was set false. This led to far more
5267 callouts that were necessary. The sender is no longer included in the key
5268 when use_sender is false.
5270 26. Added "control = submission" modifier to ACLs.
5272 27. Added the ${base62d: operator to decode base 62 numbers.
5274 28. dnsdb lookups can now access SRV records.
5276 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5277 the configuration file.
5279 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5280 (-v) mode. This makes the output for a verbose queue run more intelligible.
5282 31. Added a use_postmaster feature to recipient callouts.
5284 32. Added the $body_zerocount variable, containing the number of binary zero
5285 bytes in the message body.
5287 33. The time of last modification of the "new" subdirectory is now used as the
5288 "mailbox time last read" when there is a quota error for a maildir
5291 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5293 35. Added +ignore_unknown as a special item in host lists.
5295 36. Code for decoding IPv6 addresses in host lists is now included, even if
5296 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5297 address was recognized as an IP address, but was then not correctly decoded
5298 into binary, causing unexpected and incorrect effects when compared with
5305 1. Very minor rewording of debugging text in manualroute to say "list of
5306 hosts" instead of "hostlist".
5308 2. If verify=header_syntax was set, and a header line with an unqualified
5309 address (no domain) and a large number of spaces between the end of the
5310 name and the colon was received, the reception process suffered a buffer
5311 overflow, and (when I tested it) crashed. This was caused by some obsolete
5312 code that should have been removed. The fix is to remove it!
5314 3. When running in the test harness, delay a bit after writing a bounce
5315 message to get a bit more predictability in the log output.
5317 4. Added a call to search_tidyup() just before forking a reception process. In
5318 theory, someone could use a lookup in the expansion of smtp_accept_max_
5319 per_host which, without the tidyup, could leave open a database connection.
5321 5. Added the variables $recipient_data and $sender_data which get set from a
5322 lookup success in an ACL "recipients" or "senders" condition, or a router
5323 "senders" option, similar to $domain_data and $local_part_data.
5325 6. Moved the writing of debug_print from before to after the "senders" test
5328 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5329 problems for message scanning, either using a data ACL, or using
5330 local_scan() because the Received: header was not generated till after they
5331 were called (in order to set the time as the time of reception completion).
5332 I have revised the way this works. The header is now generated after the
5333 body is received, but before the ACL or local_scan() are called. After they
5334 are run, the timestamp in the header is updated.
5340 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5341 before starting a queue runner without re-exec. This happened only when
5342 deliver_drop_privilege was set or when the Exim user was set to root. The
5343 effect of the bug was that timeouts during subsequent deliveries caused
5344 crashes instead of being properly handled. The handler is now left at its
5345 default (and expected) setting.
5347 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5348 message, again when deliver_drop_privilege is set or Exim is run as root.
5349 The bug described in (1) was not present in this case, but the tidying up
5350 of the other signals was missing. I have made the two cases consistent.
5352 3. The ignore_target_hosts setting on a manualroute router was being ignored
5353 for hosts that were looked up using the /MX notation.
5355 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5358 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5359 operated on the sender address. After changing the $sender_address to <>
5360 for the sender address verify, Exim was re-instated it as the original
5361 (before rewriting) address, but remembering that it had rewritten it, so it
5362 wasn't rewriting it again. This bug also had the effect of breaking the
5363 sender address verification caching when the sender address was rewritten.
5365 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5366 This has been changed so that if the ip literal address matches
5367 ignore_target_hosts, the router declines.
5369 7. Added expansion conditions match_domain, match_address, and match_local_
5370 part (NOT match_host).
5372 8. The placeholder for the Received: header didn't have a length field set.
5374 9. Added code to Exim itself and to exim_lock to test for a specific race
5375 condition that could lead to file corruption when using MBX delivery. The
5376 issue is with the lockfile that is created in /tmp. If this file is removed
5377 after a process has opened it but before that process has acquired a lock,
5378 there is the potential for a second process to recreate the file and also
5379 acquire a lock. This could lead to two Exim processes writing to the file
5380 at the same time. The added code performs the same test as UW imapd; it
5381 checks after acquiring the lock that its file descriptor still refers to
5382 the same named file.
5384 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5385 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5388 11. Added the smtp_active_hostname option. If used, this will typically be made
5389 to depend on the incoming interface address. Because $interface_address is
5390 not set up until the daemon has forked a reception process, error responses
5391 that can happen earlier (such as "too many connections") no longer contain
5394 12. If an expansion in a condition on a "warn" statement fails because a lookup
5395 defers, the "warn" statement is abandoned, and the next ACL statement is
5396 processed. Previously this caused the whole ACL to be aborted.
5398 13. Added the iplsearch lookup type.
5400 14. Added ident_timeout as a log selector.
5402 15. Added tls_certificate_verified as a log selector.
5404 16. Added a global option tls_require_ciphers (compare the smtp transport
5405 option of the same name). This controls incoming TLS connections.
5407 17. I finally figured out how to make tls_require_ciphers do a similar thing
5408 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5409 before starting the TLS session.
5411 18. Tabs are now shown as \t in -bP output.
5413 19. If the log selector return_path_on_delivery was set, Exim crashed when
5414 bouncing a message because it had too many Received: header lines.
5416 20. If two routers both had headers_remove settings, and the first one included
5417 a superfluous trailing colon, the final name in the first list and the
5418 first name in the second list were incorrectly joined into one item (with a
5419 colon in the middle).
5425 1. Added -C and -D options to the exinext utility, mainly to make it easier
5426 to include in the automated testing, but these could be helpful when
5427 multiple configurations are in use.
5429 2. The exinext utility was not formatting the output nicely when there was
5430 an alternate port involved in the retry record key, nor when there was a
5431 message id as well (for retries that were specific to a specific message
5432 and a specific host). It was also confused by IPv6 addresses, because of
5433 the additional colons they contain. I have fixed the IPv4 problem, and
5434 patched it up to do a reasonable job for IPv6.
5436 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5437 delivery, the log line now contains "pipelined" if PIPELINING was used.
5439 4. An SMTP transport process used to panic and die if the bind() call to set
5440 an explicit outgoing interface failed. This has been changed; it is now
5441 treated in the same way as a connect() failure.
5443 5. A reference to $sender_host_name in the part of a conditional expansion
5444 that was being skipped was still causing a DNS lookup. This no longer
5447 6. The def: expansion condition was not recognizing references to header lines
5448 that used bh_ and bheader_.
5450 7. Added the _cache feature to named lists.
5452 8. The code for checking quota_filecount in the appendfile transport was
5453 allowing one more file than it should have been.
5455 9. For compatibility with Sendmail, the command line option
5463 and sets the incoming protocol and host name (for trusted callers). The
5464 host name and its colon can be omitted when only the protocol is to be set.
5465 Note the Exim already has two private options, -pd and -ps, that refer to
5466 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5467 "s", but I don't think that's a major issue.
5469 10. A number of refactoring changes to the code, none of which should affect
5472 (a) The number of logging options was getting close to filling up the
5473 32-bit word that was used as a bit map. I have split them into two classes:
5474 those that are passed in the argument to log_write(), and those that are
5475 only ever tested independently outside of that function. These are now in
5476 separate 32-bit words, so there is plenty of room for expansion again.
5477 There is no change in the user interface or the logging behaviour.
5479 (b) When building, for example, log lines, the code previously used a
5480 macro that called string_cat() twice, in order to add two strings. This is
5481 not really sufficiently general. Furthermore, there was one instance where
5482 it was actually wrong because one of the argument was used twice, and in
5483 one call a function was used. (As it happened, calling the function twice
5484 did not affect the overall behaviour.) The macro has been replaced by a
5485 function that can join an arbitrary number of extra strings onto a growing
5488 (c) The code for expansion conditions now uses a table and a binary chop
5489 instead of a serial search (which was left over from when there were very
5490 few conditions). Also, it now recognizes conditions like "pam" even when
5491 the relevant support is not compiled in: a suitably worded error message is
5492 given if an attempt is made to use such a condition.
5494 11. Added ${time_interval:xxxxx}.
5496 12. A bug was causing one of the ddress fields not to be passed back correctly
5497 from remote delivery subprocesses. The field in question was not being
5498 subsequently used, so this caused to problems in practice.
5500 13. Added new log selectors queue_time and deliver_time.
5502 14. Might have fixed a bug in maildirsizefile handling that threw up
5503 "unexpected character" debug warnings, and recalculated the data
5504 unnecessarily. In any case, I expanded the warning message to give more
5507 15. Added the message "Restricted characters in address" to the statements in
5508 the default ACL that block characters like @ and % in local parts.
5510 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5511 Three changes have been made:
5513 (a) There was a serious bug; a negative response to MAIL caused the whole
5514 recipient domain to be cached as invalid, thereby blocking all messages
5515 to all local parts at the same domain, from all senders. This bug has
5516 been fixed. The domain is no longer cached after a negative response to
5517 MAIL if the sender used is not empty.
5519 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5522 (c) A new callout option, "use_sender" has been added for people who want
5523 the modified behaviour.
5529 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5532 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5533 indeed breaks things for older releases.
5535 3. Added additional logging to the case where there is a problem reading data
5536 from a filter that is running in a subprocess using a pipe, in order to
5537 try to track down a specific problem.
5539 4. Testing facility fudge: when running in the test harness and attempting
5540 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5541 getting "No route to host". Convert this to a timeout.
5543 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5546 6. Some OS don't have socklen_t but use size_t instead. This affects the
5547 fifth argument of getsockopt() amongst other things. This is now
5548 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5549 can be set for individual OS. I have set it for SunOS5, OSF1, and
5550 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5551 some earlier ones do not.
5553 7. Change 4.30/15 was not doing the test caselessly.
5555 8. The standard form for an IPv6 address literal was being rejected by address
5556 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5557 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5558 this, as well as the form without the "IPv6" on the front (but only when
5559 address literals are enabled, of course).
5561 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5563 10. Exim crashed if a message with an empty sender address specified by -f
5564 encountered a router with an errors_to setting. This could be provoked only
5565 by a command such as
5569 where an empty string was supplied; "<>" did not hit this bug.
5571 11. Installed PCRE release 4.5.
5573 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5574 remained set. It is now erased.
5576 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5577 times from message ids (which are base 36 rather than the normal 62).
5579 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5580 were being counted as actual protocol errors, and logged if the log
5581 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5582 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5583 and DATA following a set of rejected RCPTs do not count as protocol errors.
5584 In other words, Exim assumes they were pipelined, though this may not
5585 actually be the case. Of course, in all cases the client gets an
5586 appropriate error code.
5588 15. If a lookup fails in an ACL condition, a message about the failure may
5589 be available; it is used if testing the ACL cannot continue, because most
5590 such messages specify what the cause of the deferral is. However, some
5591 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5592 that caused an old message to be retained and used if a later statement
5593 caused a defer, replacing the real cause of the deferral.
5595 16. If an IP address had so many PTR records that the DNS lookup buffer
5596 was not large enough to hold them, Exim could crash while trying to process
5597 the truncated data. It now detects and logs this case.
5599 17. Further to 4.21/58, another change has been made: if (and only if) the
5600 first line of a message (the first header line) ends with CRLF, a bare LF
5601 in a subsequent header line has a space inserted after it, so as not to
5602 terminate the header.
5604 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5605 unnecessarily, used atoi() instead of strtol(), and didn't check the
5606 termination when getting file sizes from file names by regex.
5608 19. Completely re-implemented the support for maildirsize files, in the light
5609 of a number of problems with the previous contributed implementation
5610 (4.30/29). In particular:
5612 . If the quota is zero, the maildirsize file is maintained, but no quota is
5615 . If the maildir directory does not exist, it is created before any attempt
5616 to write a maildirsize file.
5618 . The quota value in the file is just a cache; if the quota is changed in
5619 the transport, the new value overrides.
5621 . A regular expression is available for excluding directories from the
5624 20. The autoreply transport checks the characters in options that define the
5625 message's headers; it allows continued headers, but it was checking with
5626 isspace() after an embedded newline instead of explicitly looking for a
5629 21. If all the "regular" hosts to which an address was routed had passed their
5630 expiry times, and had not reached their retry times, the address was
5631 bounced, even if fallback hosts were defined. Now Exim should go on to try
5634 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5635 equivalent code in the SMTP transport. Some hosts send humungous responses
5636 to HELO/EHLO, more than 1024 it seems.
5638 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5639 gives compiler warnings in some environments. I've now done it "properly",
5642 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5643 (because of problems with the built-in one) was declared to return uschar *
5644 instead of char *, causing compiler failure.
5646 25. Fixed a file descriptor leak when processing alias/forward files.
5648 26. Fixed a minor format string issue in dbfn.c.
5650 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5652 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5653 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5656 29. When a sender address is verified, it is cached, to save repeating the test
5657 when there is more than one recipient in a message. However, when the
5658 verification involves a callout, it is possible for different callout
5659 options to be set for different recipients. It is too complicated to keep
5660 track of this in the cache, so now Exim always runs a verification when a
5661 callout is required, relying on the callout cache for the optimization.
5662 The overhead is duplication of the address routing, but this should not be
5665 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5666 to be verified with callout=postmaster, and the main callout worked but the
5667 postmaster check failed, the verification correctly failed. However, if a
5668 subsequent RCPT command asked for sender verification *without* the
5669 postmaster check, incorrect caching caused this verification also to fail,
5672 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5673 it was not caching the DNS options (qualify_single, search_parents) that
5674 were used when the lookup failed. A subsequent lookup with different
5675 options therefore always gave the same answer, though there were cases
5676 where it should not have. (Example: a "domains = !$mx_any" option on a
5677 dnslookup router: the "domains" option is always processed without any
5678 widening, but the router might have qualify_single set.) Now Exim uses the
5679 cached value only when the same options are set.
5681 32. Added John Jetmore's "exipick" utility to the distribution.
5683 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5684 than a timeout (e.g. a certificate is required, and is not provided), an
5685 Exim server now closes the connection immediately. Previously it waited for
5686 the client to close - but if the client is SSL, it seems that they each
5687 wait for each other, leading to a delay before one of them times out.
5689 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5690 maintained 0.8.x compatibility because I don't think many are using it, and
5691 it is clearly obsolete.
5693 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5696 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5697 client certificate was expired. A simple patch fixes this, though I don't
5698 understand the full logic of why the verify callback is called multiple
5701 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5702 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5703 which causes problems with some clients (such as the Certicom SSL Plus
5704 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5705 disables the coutermeasure allowing Eudora to connect."
5707 38. Exim was not checking that a write() to a log file succeeded. This could
5708 lead to Bad Things if a log got too big, in particular if it hit a file
5709 size limit. Exim now panics and dies if it cannot write to a log file, just
5710 as it does if it cannot open a log file.
5712 39. Modified OS/Makefile-Linux so that it now contains
5714 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5716 The two -D definitions ensure that Exim is compiled with large file
5717 support, which makes it possible to handle log files that are bigger than
5720 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5721 instance) a domain was checked against a named list that involved a lookup,
5722 causing $domain_data to be set, then another domain was checked against the
5723 same list, then the first domain was re-checked, the value of $domain_data
5724 after the final check could be wrong. In particular, if the second check
5725 failed, it could be set empty. This bug probably also applied to
5728 41. The strip_trailing_dot option was not being applied to the address given
5729 with the -f command-line option.
5731 42. The code for reading a message's header from the spool was incrementing
5732 $received_count, but never initializing it. This meant that the value was
5733 incorrect (doubled) while delivering a message in the same process in which
5734 it was received. In the most common configuration of Exim, this never
5735 happens - a fresh exec is done - but it can happen when
5736 deliver_drop_privilege is set.
5738 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5739 it now includes up to 150 characters of the unexpected data in the log
5742 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5743 and building data strings. The size of both of these buffers was 10 000
5744 bytes - far larger than anybody would *ever* want, thought I. Needless to
5745 say, somebody hit the limit. I have increased the maximum line length to
5746 20 000 and the maximum data length of concatenated lines to 100 000. I have
5747 also fixed two bugs, because there was no checking on these buffers. Tsk,
5748 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5749 buffer is too small.
5751 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5752 lsearch lookups. Now it does.
5754 46. When parsing a route_list item in a manualroute router, a fixed-length
5755 buffer was used for the list of hosts. I made this 1024 bytes long,
5756 thinking that nobody would ever have a list of hosts that long. Wrong.
5757 Somebody had a whole pile of complicated expansion conditions, and the
5758 string was silently truncated, leading to an expansion error. It turns out
5759 that it is easier to change to an unlimited length (owing to other changes
5760 that have happened since this code was originally written) than to build
5761 structure for giving a limitation error. The length of the item that
5762 expands into the list of hosts is now unlimited.
5764 47. The lsearch lookup could not handle data where the length of text line was
5765 more than 4095 characters. Such lines were truncated, leading to shortened
5766 data being returned. It should now handle lines of any length.
5768 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5769 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5772 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5773 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5774 no longer happens in comment lines. A list of replacements is now placed
5775 at the head of all of the source files, except those whose only change is
5776 to replace PERL_COMMAND in the very first #! line.
5778 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5779 messages on the queue, with a bottom-up merge sort, using code contributed
5780 by Michael Haardt. This should make operations like -bp somewhat faster on
5781 large queues. It won't affect queue runners, except when queue_run_in_order
5784 51. Installed eximstats 1.31 in the distribution.
5786 52. Added support for SRV lookups to the dnslookup router.
5788 53. If an ACL referred to $message_body or $message_body_end, the value was not
5789 reset for any messages that followed in the same SMTP session.
5791 54. The store-handling optimization for building very long strings was not
5792 differentiating between the different store pools. I don't think this
5793 actually made any difference in practice, but I've tidied it.
5795 55. While running the routers to verify a sender address, $sender_address
5796 was still set to the sender address. This is wrong, because when routing to
5797 send a bounce to the sender, it would be empty. Therefore, I have changed
5798 it so that, while verifying a sender address, $sender_address is set to <>.
5799 (There is no change to what happens when verifying a recipient address.)
5801 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5802 target A or AAAA records (if not already returned) without resetting the
5803 qualify_single or search_parents options of the DNS resolver. These are
5804 inappropriate in this case because the targets of MX and SRV records must
5805 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5806 target that, when qualified, matched something in the local domain. These
5807 two options are now turned off when doing these lookups.
5809 57. It seems that at least some releases of Reiserfs (which does not have the
5810 concept of a fixed number of inodes) returns zero and not -1 for the
5811 number of available inodes. This interacted badly with check_spool_inodes,
5812 which assumed that -1 was the "no such thing" setting. What I have done is
5813 to check that the total number of inodes is greater than zero before doing
5814 the test of how many are available.
5816 58. When a "warn" ACL statement has a log_message modifier, the message is
5817 remembered, and not repeated. This is to avoid a lot of repetition when a
5818 message has many recipients that cause the same warning to be written.
5819 However, Exim was preserving the list of already written lines for an
5820 entire SMTP session, which doesn't seem right. The memory is now reset if a
5821 new message is started.
5823 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5824 debugging output unless log_rewrite was also set.
5826 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5827 of (int)(handle) when we know that handle contains (void *)(-1).
5829 61. The Exim daemon panic-logs an error return when it closes the incoming
5830 connection. However "connection reset by peer" seems to be common, and
5831 isn't really an error worthy of noting specially, so that particular error
5834 62. When Exim is trying to find all the local interfaces, it used to panic and
5835 die if the ioctl to get the interface flags failed. However, it seems that
5836 on at least one OS (Solaris 9) it is possible to have an interface that is
5837 included in the list of interfaces, but for which you get a failure error
5838 for this call. This happens when the interface is not "plumbed" into a
5839 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5840 failure of the "get flags" call assumes that the interface is down.
5842 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5843 makes life easier for people who are doing arithmetic on fields extracted
5844 from dates, where you often get leading zeros that should not be
5845 interpreted as octal.
5847 64. Added qualify_domain to the redirect router, to override the global
5850 65. If a pathologically long header line contained very many addresses (the
5851 report of this problem mentioned 10 000) and each of them was rewritten,
5852 Exim could use up a very large amount of memory. (It kept on making new
5853 copies of the header line as it rewrote, and never released the old ones.)
5854 At the expense of a bit more processing, the header rewriting function has
5855 been changed so that it no longer eats memory in this way.
5857 66. The generation of the Received: header has been moved from the time that a
5858 message starts to be received, to the time that it finishes. The timestamp
5859 in the Received: header should now be very close to that of the <= log
5860 line. There are two side-effects of this change:
5862 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5863 logged header lines no longer include the local Received: line, because
5864 it has not yet been created. The same applies to a copy of the message
5865 that is returned to a non-SMTP sender when a message is rejected.
5867 (b) When a filter file is tested using -bf, no additional Received: header
5868 is added to the test message. After some thought, I decided that this
5871 This change does not affect the value of $received_for. It is still set
5872 after address rewriting, but before local_scan() is called.
5874 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5876 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5877 gave an unhelpful panic error message, and a defer error. I have managed to
5878 change this behaviour so that it now rejects any supplied certificate,
5879 which seems right, as the list of acceptable certificates is empty.
5881 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5882 gave an unhelpful defer error. I have not managed to make this reject any
5883 supplied certificates, but the error message it gives is "no certificate
5884 supplied", which is not helpful.
5886 70. exigrep's output now also includes lines that are not associated with any
5887 message, but which match the given pattern. Implemented by a patch from
5888 Martin Sluka, which also tidied up the Perl a bit.
5890 71. Recipient callout verification, like sender verification, was using <> in
5891 the MAIL FROM command. This isn't really the right thing, since the actual
5892 sender may affect whether the remote host accepts the recipient or not. I
5893 have changed it to use the actual sender in the callout; this means that
5894 the cache record is now keyed on a recipient/sender pair, not just the
5895 recipient address. There doesn't seem to be a real danger of callout loops,
5896 since a callout by the remote host to check the sender would use <>.
5897 [SEE ABOVE: changed after hitting problems.]
5899 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5900 temporary errors. However, in the case of such a code being given after
5901 the end of a data transmission (i.e. after ".") Exim was failing to write
5902 a retry record for the message. (Yes, there was some broken host that was
5903 actually sending 8xx at this point.)
5905 73. An unknown lookup type in a host list could cause Exim to panic-die when
5906 the list was checked. (An example that provoked this was putting <; in the
5907 middle of a list instead of at the start.) If this happened during a DATA
5908 ACL check, a -D file could be left lying around. This kind of configuration
5909 error no longer causes Exim to die; instead it causes a defer error. The
5910 incident is still logged to the main and panic logs.
5912 74. Buglet left over from Exim 3 conversion. The message "too many messages
5913 in one connection" was written to the rejectlog but not the mainlog, except
5914 when address rewriting (yes!) was being logged.
5916 75. Added write_rejectlog option.
5918 76. When a system filter was run not as root (that is, when system_filter_user
5919 was set), the values of the $n variables were not being returned to the
5920 main process; thus, they were not subsequently available in the $sn
5923 77. Added +return_path_on_delivery log selector.
5925 78. A connection timeout was being treated differently from recipients deferred
5926 when testing hosts_max_try with a message that was older than the host's
5927 retry timeout. (The host should not be counted, thus allowing all hosts to
5928 be tried at least once before bouncing.) This may have been the cause of an
5929 occasionally reported bug whereby a message would remain on the queue
5930 longer than the retry timeout, but would be bounced if a delivery was
5931 forced. I say "may" because I never totally pinned down the problem;
5932 setting up timeout/retry tests is difficult. See also the next item.
5934 79. The ultimate address timeout was not being applied to errors that involved
5935 a combination of host plus message (for example, a timeout on a MAIL
5936 command). When an address resolved to a number of possible hosts, and they
5937 were not all tried for each delivery (e.g. because of hosts_max_try), a
5938 message could remain on the queue longer than the retry timeout.
5940 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5943 81. Fixed an obscure SMTP outgoing bug which required at least the following
5944 conditions: (a) there was another message waiting for the same server;
5945 (b) the server returned 5xx to all RCPT commands in the first message so
5946 that the message was not completed; (c) the server dropped the connection
5947 or gave a negative response to the RSET that Exim sends to abort the
5948 transaction. The observed case was a dropped connection after DATA that had
5949 been sent in pipelining mode. That is, the server had advertised PIPELINING
5950 but was not implementing it correctly. The effect of the bug was incorrect
5951 behaviour, such as trying another host, and this could lead to a crash.
5957 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5958 and daemon.c were passed as pointers to ints; they should have been
5959 pointers to socklen_t variables (which are typically unsigned ints).
5961 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5964 3. Fixed a really odd bug that affected only the testing scheme; patching a
5965 certain fixed string in the binary changed the value of another string that
5966 happened to be identical to the end of the original first string.
5968 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5969 name", it returns that address as the IP address. On some operating
5970 systems (e.g. Solaris), it also passes back the IP address string as the
5971 "host name". However, on others (e.g. Linux), it passes back an empty
5972 string. Exim wasn't checking for this, and was changing the host name to an
5973 empty string, assuming it had been canonicalized.
5975 5. Although rare, it is permitted to have more than one PTR record for a given
5976 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5977 all the names associated with an address, because they do in Solaris.
5978 However, it seems that they do not in Linux for data that comes from the
5979 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5980 I found this out when I moved to a new Linux workstation and tried to run
5981 the Exim test suite.
5983 To get round this problem I have changed the code so that it now does its
5984 own call to the DNS to look up PTR records when searching for a host name.
5985 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5986 addresses that are only in /etc/hosts are still found.
5988 This behaviour is, however, controlled by an option called host_lookup_
5989 order, which defaults to "bydns:byaddr". If people want to use the other
5990 order, or indeed, just use one or the other means of lookup, they can
5991 specify it in this variable.
5993 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5994 some operating systems, this comes back from gethostbyaddr() as an empty
5995 string, and this is what Exim used to test for. However, it seems that in
5996 other systems, "." is yielded. Exim now tests for this case too.
5998 7. The values of check_spool_space and check_log_space are now held internally
5999 as a number of kilobytes instead of an absolute number of bytes. If a
6000 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6001 kilobyte. This means that much larger values can be stored.
6003 8. Exim monitor: an attempt to get the action menu when not actually pointing
6004 at a message produces an empty menu entitled "No message selected". This
6005 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6006 no entries in it ("Shell widget menu has zero width and/or height"). So I
6007 have added a single, blank menu entry in this case.
6009 9. Added ${quote_local_part.
6011 10. MIME decoding is now applied to the contents of Subject: header lines when
6014 11. Now that a reference to $sender_host_address automatically causes a reverse
6015 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6016 host lookup before query-style lookups in lists that might use this
6017 variable. This has therefore been abolished, and the "net-" prefix is no
6018 longer necessary for query-style lookups.
6020 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6021 have been a typo for LFLAGS, so it has been changed.
6023 13. The install script calls Exim with "-C /dev/null" in order to find the
6024 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6025 to be output. However, since Exim outputs its version number before the
6026 error, it didn't break the script. It just looked ugly. I fixed this by
6027 always allowing "-C /dev/null" if the caller is root.
6029 14. Ignore overlarge ACL variable number when reading spool file - insurance
6030 against a later release with more variables having written the file.
6032 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6033 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6034 this, as well as the form without the "IPv6" on the front.
6036 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6037 OS/Makefile-Darwin file.
6039 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6040 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6042 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6043 variable. However, some OS use ints for inodes. I've added cast to long int
6044 to get rid of the compiler warning.
6046 19. I had forgotten to lock out "/../" in configuration file names when
6047 ALT_CONFIG_PREFIX was set.
6049 20. Routers used for verification do not need to specify transports. However,
6050 if such a router generated a host list, and callout was configured, Exim
6051 crashed, because it could not find a port number from the (non-existent)
6052 transport. It now assumes port 25 in this circumstance.
6054 21. Added the -t option to exigrep.
6056 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6057 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6058 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6059 them is set, LOOKUP_LSEARCH is forced.
6061 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6063 24. Added sender and host information to the "rejected by local_scan()" log
6064 line; previously there was no indication of these.
6066 25. Added .include_if_exists.
6068 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6069 Linux. It turns out that not all file systems support this. Apparently some
6070 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6071 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6072 not supported on the file descriptor, is now ignored when Exim is trying to
6073 sync a directory. This applies only to Linux.
6075 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6077 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6078 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6081 29. Added support for maildirsize files from supplied patch (modified a bit).
6083 30. The use of :fail: followed by an empty string could lead Exim to respond to
6084 sender verification failures with (e.g.):
6086 550 Verification failed for <xxx>
6087 550 Sender verify failed
6089 where the first response line was missing the '-' that indicates it is not
6090 the final line of the response.
6092 31. The loop for finding the name of the user that called Exim had a hardwired
6093 limit of 10; it now uses the value of finduser_retries, which is used for
6094 all other user lookups.
6096 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6099 33. Exim was neglecting to zero errno before one call of strtol() when
6100 expanding a string and expecting an integer value. On some systems this
6101 resulted in spurious "integer overflow" errors. Also, it was casting the
6102 result into an int without checking.
6104 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6105 did not work. The code looks as if it has *never* worked, though it appears
6106 to have been documented since at least release 1.62. I have made it work.
6108 35. The "timeout_DNS" error in retry rules, also documented since at least
6109 1.62, also never worked. As it isn't clear exactly what this means, and
6110 clearly it isn't a major issue, I have abolished the feature by treating it
6111 as "timeout", and writing a warning to the main and panic logs.
6113 36. The display of retry rules for -brt wasn't always showing the error code
6116 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6117 timeout_connect_A, timeout_connect_MX.
6119 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6120 to the empty sender.
6122 39. The daemon was not analysing the content of -oX till after it had closed
6123 stderr and disconnected from the controlling terminal. This meant that any
6124 syntax errors were only noted on the panic log, and the return code from
6125 the command was 0. By re-arranging the code a little, I've made the
6126 decoding happen first, so such errors now appear on stderr, and the return
6127 code is 1. However, the actual setting up of the sockets still happens in
6128 the disconnected process, so errors there are still only recorded on the
6131 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6132 connections (as happens on some IP stacks) was logged at start up time as
6133 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6134 it from "IPv6 and IPv4", which means that two separate sockets are being
6137 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6138 says whether AF_INET or AF_INET6 was passed as an argument.
6140 42. Exiwhat output was messed up when time zones were included in log
6143 43. Exiwhat now gives more information about the daemon's listening ports,
6144 and whether -tls-on-connect was used.
6146 44. The "port" option of the smtp transport is now expanded.
6148 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6149 silently ignored. Now an error message is written to the main and panic
6152 46. There's a new ACL modifier called "logwrite" which writes to a log file
6153 as soon as it is encountered.
6155 47. Added $local_user_uid and $local_user_gid at routing time.
6157 48. Exim crashed when trying to verify a sender address that was being
6160 49. Exim was recognizing only a space character after ".include". It now also
6161 recognizes a tab character.
6163 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6164 extracting the relevant information from the specification. The man page no
6165 longer contains scrambled data for the -d option, and I've added a section
6166 at the front about calling Exim under different names.
6168 51. Added "extra_headers" argument to the "mail" command in filter files.
6170 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6173 53. Installed eximstats 1.29.
6175 54. Added transport_filter_timeout as a generic transport option.
6177 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6178 Cc: header lines. This was required by RFC 822, but it not required by RFC
6181 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6182 incoming messages that did not have them. Now it does so only if the
6183 message originates locally, that is, if there is no associated remote host
6184 address. When Resent- header lines are present, this applies to the Resent-
6185 lines rather than the non-Resent- lines.
6187 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6188 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6190 58. Messages for configuration errors now include the name of the main
6191 configuration file - useful now that there may be more than one file in a
6192 list (.included file names were always shown).
6194 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6195 for those rare installations that do not start the daemon as root or run it
6196 setuid root. I've cut out the call to initgroups() if the daemon is not
6199 60. The Exim user and group can now be bound into the binary as text strings
6200 that are looked up at the start of Exim's processing.
6202 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6204 62. Added $mailstore_basename variable.
6206 63. Installed patch to sieve.c from Michael Haardt.
6208 64. When Exim failed to open the panic log after failing to open the main log,
6209 the original message it was trying to log was written to stderr and debug
6210 output, but if they were not available (the usual case in production), it
6211 was lost. Now it is written to syslog before the two lines that record the
6212 failures to open the logs.
6214 65. Users' Exim filters run in subprocesses under the user's uid. It is
6215 possible for a "deliver" command or an alias in a "personal" command to
6216 provoke an address rewrite. If logging of address rewriting is configured,
6217 this fails because the process is not running as root or exim. There may be
6218 a better way of dealing with this, but for the moment (because 4.30 needs
6219 to be released), I have disabled address rewrite logging when running a
6220 filter in a non-root, non-exim process.
6226 1. The buildconfig auxiliary program wasn't quoting the value set for
6227 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6228 not defined. This bug was masked in 4.22 by the effect that was fixed in
6231 2. Some messages that were rejected after a message id was allocated were
6232 shown as "incomplete" by exigrep. It no longer does this for messages that
6233 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6235 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6236 have allow_domain_literals set, the ID did not get logged in the <= line.
6237 Domain literals are now always recognized in Message-ID: header lines.
6239 4. The first argument for a ${extract expansion item is the key name or field
6240 number. Leading and trailing spaces in this item were not being ignored,
6241 causing some misleading effects.
6243 5. When deliver_drop_privilege was set, single queue runner processes started
6244 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6245 same command in the process it spins off) were not dropping privilege.
6247 6. When the daemon running as "exim" started a queue runner, it always
6248 re-executed Exim in the spun-off process. This is a waste of effort when
6249 deliver_drop_privilege is set. The new process now just calls the
6250 queue-runner function directly.
6256 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6259 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6260 ignored. Though the use of -oP was forcing the writing of a pid file, it
6261 was always written to the default place.
6263 3. If the message "no IP address found for host xxxx" is generated during
6264 incoming verification, it is now followed by identification of the incoming
6265 connection (so you can more easily find what provoked it).
6267 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6269 5. Added some features to "harden" Exim a bit more against certain attacks:
6271 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6272 be put in Local/Makefile. This is like the never_users runtime option,
6273 but it cannot be overridden. The default setting is "root".
6275 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6276 prefix string with which any file named in a -C command line option
6279 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6280 is retained for -C and -D only if the caller of Exim is root. Without
6281 it, the exim user may also use -C and -D and retain privilege.
6283 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6284 command line option is disabled.
6286 6. Macro names set by the -D option must start with an upper case letter, just
6287 like macro names defined in the configuration file.
6289 7. Added "dereference=" facility to LDAP.
6291 8. Two instances of the typo "uknown" in the source files are fixed.
6293 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6294 the Configure-Makefile script screwed up while processing it.
6296 10. Incorporated PCRE 4.4.
6298 11. The SMTP synchronization check was not operating right at the start of an
6299 SMTP session. For example, it could not catch a HELO sent before the client
6300 waited for the greeting. There is now a check for outstanding input at the
6301 point when the greeting is written. Because of the duplex, asynchronous
6302 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6303 way, but not yet received, when the check is performed.
6305 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6306 on TCP/IP sockets, because this apparently causes some broken clients to
6309 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6310 unchanged) from the Cygwin maintainer.
6312 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6313 was supported instead of testing for mbx. Effectively a typo.
6315 15. The spa authenticator server code was not checking that the input it
6316 received was valid base64.
6318 16. The debug output line for the "set" modifier in ACLs was not showing the
6319 name of the variable that was being set.
6321 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6323 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6324 lookup on its own. Something else was needed to trigger the lookup. For
6325 example, a match in host_lookup or the need for a host name in a host list.
6326 Now, if $sender_host_name is referenced and the host name has not yet been
6327 looked up, a lookup is performed. If the lookup fails, the variable remains
6328 empty, and $host_lookup_failed is set to "1".
6330 19. Added "eqi" as a case-independent comparison operator.
6332 20. The saslauthd authentication condition could segfault if neither service
6333 nor realm was specified.
6335 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6336 error message that was logged was misleading, and incoming SMTP
6337 connections were dropped. The message is now more accurate, and temporary
6338 errors are given to SMTP connections.
6340 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6341 (except RSET) with a 421 error, until QUIT is received. However, it was
6342 failing to send a response to QUIT.
6344 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6345 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6346 ACL is now run after the helo_try_verify_hosts code.
6348 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6349 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6350 case-independent, so other case variants are also recognized). Apparently
6351 some systems use these upper case variants.
6353 25. If more than two messages were waiting for the same host, and a transport
6354 filter was specified for the transport, Exim sent two messages over the
6355 same TCP/IP connection, and then failed with "socket operation on non-
6356 socket" when it tried to send the third.
6358 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6360 27. The extern definition of crypt16() in expand.c was not being excluded when
6361 the OS had its own crypt16() function.
6363 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6364 as a preferred synonym for return_size_limit, both as an option and as an
6367 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6369 30. Changed the default configuration ACL to relax the local part checking rule
6370 for addresses that are not in any local domains. For these addresses,
6371 slashes and pipe symbols are allowed within local parts, but the sequence
6372 /../ is explicitly forbidden.
6374 31. SPA server authentication was not clearing the challenge buffer before
6377 32. log_message in a "warn" ACL statement was writing to the reject log as
6378 well as to the main log, which contradicts the documentation and doesn't
6379 seem right (because no rejection is happening). So I have stopped it.
6381 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6382 However, I am unable to do any testing of this.
6384 34. Fixed an infelicity in the appendfile transport. When checking directories
6385 for a mailbox, to see if any needed to be created, it was accidentally
6386 using path names with one or more superfluous leading slashes; tracing
6387 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6389 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6390 contents are added to the log line that is written for every discarded
6391 recipient. (Previously a log_message setting was ignored.)
6393 36. The ${quote: operator now quotes the string if it is empty.
6395 37. The install script runs exim in order to find its version number. If for
6396 some reason other than non-existence or emptiness, which it checks, it
6397 could not run './exim', it was installing it with an empty version number,
6398 i.e. as "exim-". This error state is now caught, and the installation is
6401 38. An argument was missing from the function that creates an error message
6402 when Exim fails to connect to the socket for saslauthd authentication.
6403 This could cause Exim to crash, or give a corrupted message.
6405 39. Added isip, isip4, and isip6 to ${if conditions.
6407 40. The ACL variables $acl_xx are now saved with the message, and can be
6408 accessed later in routers, transports, and filters.
6410 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6411 strings in the file are not string-expanded.
6413 42. If a MAIL command specified a SIZE value that was too large to fit into an
6414 int variable, the check against message_size_limit failed. Such values are
6415 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6416 day this will have to be increased, but I don't think I want to be around
6417 when emails are that large.
6424 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6425 iconv() is not standard in FreeBSD.
6427 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6428 IPv6 enabled. The observed symptom was a segmentation fault on return from
6429 the function os_common_find_running_interfaces() in src/os.c.
6431 3. In the check_special_case() function in daemon.c I had used "errno" as an
6432 argument name, which causes warnings on some systems. This was basically a
6433 typo, since it was named "eno" in the comments!
6435 4. The code that waits for the clock to tick (at a resolution of some fraction
6436 of a second) so as to ensure message-id uniqueness was always waiting for
6437 at least one whole tick, when it could have waited for less. [This is
6438 almost certainly not relevant at current processor speeds, where it is
6439 unlikely to ever wait at all. But we try to future-proof.]
6441 5. The function that sleeps for a time interval that includes fractions of a
6442 second contained a race. It did not block SIGALRM between setting the
6443 timer, and suspending (a couple of lines later). If the interval was short
6444 and the sigsuspend() was delayed until after it had expired, the suspension
6445 never ended. On busy systems this could lead to processes getting stuck for
6448 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6449 process, before it forks any delivery processes. The open lookup caching
6450 mechanism meant that the open file or database connection was passed into
6451 the delivery process. The problem was that delivery processes always tidy
6452 up cached lookup data. This could cause a problem for the next delivery
6453 process started by the queue runner, because the external queue runner
6454 process does not know about the closure. So the next delivery process
6455 still has data in the lookup cache. In the case of a file lookup, there was
6456 no problem because closing a file descriptor in a subprocess doesn't affect
6457 the parent. However, if the lookup was caching a connection to a database,
6458 the connection was closed, and the second delivery process was likely to
6459 see errors such as "PGSQL: query failed: server closed the connection
6460 unexpectedly". The problem has been fixed by closing all cached lookups
6461 in a queue runner before running a delivery process.
6463 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6464 seem to have the "const" qualifier which it has on other OS. I've
6467 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6468 *of the same type* (client or server) with the same public name that an
6469 error should be diagnosed.
6471 9. When Exim looked up a host name for an IP address, but failed to find the
6472 original IP address when looking up the host name (a safety check), it
6473 output the message "<ip address> does not match any IP for NULL", which was
6474 confusing, to say the least. The bug was that the host name should have
6475 appeared instead of "NULL".
6477 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6478 user that is built into the binary, and the -C or -D options is used, root
6479 privilege is dropped before the configuration file is read. In addition,
6480 logging is switched to stderr instead of the normal log files. If the
6481 configuration then re-defines the Exim user, the unprivileged environment
6482 is probably not what is expected, so Exim logs a panic warning message (but
6485 However, if deliver_drop_privilege is set, the unprivileged state may well
6486 be exactly what is intended, so the warning has been cut out in that case,
6487 and Exim is allowed to try to write to its normal log files.
6493 1. smtp_return_error_details was not giving details for temporary sender
6494 or receiver verification errors.
6496 2. Diagnose a configuration error if two authenticators have the same public
6499 3. Exim used not to create the message log file for a message until the first
6500 delivery attempt. This could be confusing when incoming messages were held
6501 for policy or load reasons. The message log file is now created at the time
6502 the message is received, and an initial "Received" line is written to it.
6504 4. The automatically generated man page for command line options had a minor
6505 bug that caused no ill effects; however, a more serious problem was that
6506 the procedure for building the man page automatically didn't always
6507 operate. Consequently, release 4.20 contains an out-of-date version. This
6508 shouldn't happen again.
6510 5. When building Exim with embedded Perl support, the script that builds the
6511 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6512 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6514 6. The freeze_tell option was not being used for messages that were frozen on
6515 arrival, either by an ACL or by local_scan().
6517 7. Added the smtp_incomplete_transaction log selector.
6519 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6520 was accepting AUTH without a new EHLO.
6522 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6523 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6526 10. Logging of TCP/IP connections (when configured) now happens in the main
6527 daemon process instead of the child process, so that the TCP/IP connection
6528 count is more accurate (but it can never be perfect).
6530 11. The use of "drop" in a nested ACL was not being handled correctly in the
6531 outer ACL. Now, if condition failure induced by the nested "drop" causes
6532 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6533 or "require"), the connection is dropped.
6535 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6536 that yield "discard" can now be used with an "accept" or a "discard" verb,
6537 but an error is generated for any others (because I can't see a useful way
6538 to define what should happen).
6540 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6541 are now processed in the same way as lines in the Exim configuration file.
6542 In particular, continuation lines are supported.
6544 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6546 15. Added -ti meaning -t -i.
6548 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6549 domains, and warn by logging if others are found.
6551 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6552 returned from GIFCONF when Exim is trying to find the list of interfaces on
6553 a host. The code in os.c has been modified to copy each ifreq to an aligned
6554 structure in all cases.
6556 Also, in some cases, the returned ifreq's were being copied to a 'struct
6557 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6558 means the last couple of bytes of an IPv6 address could be chopped if the
6559 ifreq contained only a normal sockaddr (14 bytes storage).
6561 18. Named domain lists were not supported in the hosts_treat_as_local option.
6562 An entry such as +xxxx was not recognized, and was treated as a literal
6565 19. Ensure that header lines added by a DATA ACL are included in the reject log
6566 if the ACL subsequently rejects the message.
6568 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6569 MD5 (which is deprecated).
6571 21. When testing a filter file using -bf, Exim was writing a message when it
6572 took the sender from a "From " line in the message, but it was not doing so
6573 when it took $return_path from a Return-Path: header line. It now does.
6575 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6576 with a valid header line field name (a series of printing characters
6577 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6579 23. Changed "disc" in the source to "disk" to conform to the documentation and
6580 the book and for uniformity.
6582 24. Ignore Sendmail's -Ooption=value command line item.
6584 25. When execve() failed while trying to run a command in a pipe transport,
6585 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6586 could be confused with a return value of 69 from the command itself. This
6587 has been changed to 127, the value the shell returns if it is asked to run
6588 a non-existent command. The wording for the related log line suggests a
6589 non-existent command as the problem.
6591 26. If received_header_text expands to an empty string, do not add a Received:
6592 header line to the message. (Well, it adds a token one on the spool, but
6593 marks it "old" so that it doesn't get used or transmitted.)
6595 27. Installed eximstats 1.28 (addition of -nt option).
6597 28. There was no check for failure on the call to getsockname() in the daemon
6598 code. This can fail if there is a shortage of resources on the system, with
6599 ENOMEM, for example. A temporary error is now given on failure.
6601 29. Contrary to the C standard, it seems that in some environments, the
6602 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6603 program. Exim now does this explicitly; it affects the formatting of
6604 timestamps using strftime().
6606 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6607 complaints. I've now initialized all the variables, to avoid this.
6609 32. Header lines added by a system filter were not being "seen" during
6610 transport-time rewrites.
6612 33. The info_callback() function passed to OpenSSL is set up with type void
6613 (*)(SSL *, int, int), as described somewhere. However, when calling the
6614 function (actually a macro) that sets it up, the type void(*)() is
6615 expected. I've put in a cast to prevent warnings from picky compilers.
6617 34. If a DNS black list lookup found a CNAME record, but there were no A
6618 records associated with the domain it pointed at, Exim crashed.
6620 35. If a DNS black list lookup returned more than one A record, Exim ignored
6621 all but the first. It now scans all returned addresses if a particular IP
6622 value is being sought. In this situation, the contents of the
6623 $dnslist_value variable are a list of all the addresses, separated by a
6626 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6627 to accept a host name and all its aliases if the forward lookup for any of
6628 them yielded the IP address of the incoming connection. Now it accepts only
6629 those names whose forward lookup yields the correct IP address. Any other
6630 names are discarded. This closes a loophole whereby a rogue DNS
6631 administrator could create reverse DNS records to break through a
6632 wildcarded host restriction in an ACL.
6634 37. If a user filter or a system filter that ran in a subprocess used any of
6635 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6636 the wrong values were passed to the pipe command ($thisaddress had the
6637 value of $0, $0 had the value of $1, etc). This bug was introduced by
6638 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6640 38. Improved the line breaking for long SMTP error messages from ACLs.
6641 Previously, if there was no break point between 40 and 75 characters, Exim
6642 left the rest of the message alone. Two changes have been made: (a) I've
6643 reduced the minimum length to 35 characters; (b) if it can't find a break
6644 point between 35 and 75 characters, it looks ahead and uses the first one
6645 that it finds. This may give the occasional overlong line, but at least the
6646 remaining text gets split now.
6648 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6649 file descriptors might be low, and that setting 1000 would always raise it.
6650 It turns out that in some environments, the limit is already over 1000 and
6651 that lowering it causes trouble. So now Exim takes care not to decrease it.
6653 40. When delivering a message, the value of $return_path is set to $sender_
6654 address at the start of routing (routers may change the value). By an
6655 oversight, this default was not being set up when an address was tested by
6656 -bt or -bv, which affected the outcome if any router or filter referred to
6659 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6660 log only when "message" or "log_message" are set. However, if one of the
6661 conditions was an address verification, or a call to a nested ACL, the
6662 messages generated by the underlying test were being passed through. This
6663 no longer happens. The underlying message is available in $acl_verify_
6664 message for both "message" and "log_message" expansions, so it can be
6665 passed through if needed.
6667 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6668 new expansion $bh_ to give the encoded byte string without charset
6669 translation. Translation happens only if iconv() is available; HAVE_ICONV
6670 indicates this at build time. HEADERS_CHARSET gives the charset to
6671 translate to; headers_charset can change it in the configuration, and
6672 "headers charset" can change it in an individual filter file.
6674 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6675 that creates RFC 2047 encoded "words" labels them as that charset instead
6676 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6677 expansion operator; (ii) when Exim creates a From: line for a local
6678 message; (iii) when a header line is rewritten to include a "phrase" part.
6680 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6681 buggy, causing it to skip the first lines of messages whose message ID
6682 ended in 'D'. This would not have bitten before Exim release 4.14, because
6683 message IDs were unlikely to end in 'D' before then. The effect was to have
6684 incorrect size information for certain domains.
6686 45. #include "config.h" was missing at the start of the crypt16.c module. This
6687 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6690 46. If there was a timeout during a "random" callout check, Exim treated it as
6691 a failure of the random address, and carried on sending RSET and the real
6692 address. If the delay was just some slowness somewhere, the response to the
6693 original RCPT would be taken as a response to RSET and so on, causing
6694 mayhem of various kinds.
6696 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6697 when I implemented it. It didn't allow for the fact that some option values
6698 may legitimately be negative (e.g. size_addition), and it didn't even do
6699 the right test for positive values.
6701 48. Domain names in DNS records are case-independent. Exim always looks them up
6702 in lower case. Some resolvers return domain names in exactly the case they
6703 appear in the zone file, that is, they may contain uppercase letters. Not
6704 all resolvers do this - some return always lower case. Exim was treating a
6705 change of case by a resolver as a change of domain, similar to a widening
6706 of a domain abbreviation. This triggered its re-routing code and so it was
6707 trying to route what was effectively the same domain again. This normally
6708 caused routing to fail (because the router wouldn't handle the domain
6709 twice). Now Exim checks for this case specially, and just changes the
6710 casing of the domain that it ultimately uses when it transmits the message
6713 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6716 50. If a filter generated a file delivery with a non-absolute name (possible if
6717 no home directory exists for the router), the forbid_file option was not
6720 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6721 the existing equality matching.
6723 52. Exim was using ints instead of ino_t variables in some places where it was
6724 dealing with inode numbers.
6726 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6727 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6728 called TMPDIR, and if it finds it is different, it changes its value.
6730 54. The smtp_printf() function is now made available to local_scan() so
6731 additional output lines can be written before returning. There is also an
6732 smtp_fflush() function to enable the detection of a dropped connection.
6733 The variables smtp_input and smtp_batched_input are exported to
6736 55. Changed the default runtime configuration: the message "Unknown user"
6737 has been removed from the ACL, and instead placed on the localuser router,
6738 using the cannot_route_message feature. This means that any verification
6739 failures that generate their own messages won't get overridden. Similarly,
6740 the "Unrouteable address" message that was in the ACL for unverifiable
6741 relay addresses has also been removed.
6743 56. Added hosts_avoid_esmtp to the smtp transport.
6745 57. The exicyclog script was not checking for the esoteric option
6746 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6747 will work only if exicyclog is run under the appropriate euid.
6749 58. Following a discussion on the list, the rules by which Exim recognises line
6750 endings on incoming messages have been changed. The -dropcr and drop_cr
6751 options are now no-ops, retained only for backwards compatibility. The
6752 following line terminators are recognized: LF CRLF CR. However, special
6753 processing applies to CR:
6755 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6756 nor a local message in the state where . is a terminator.
6758 (ii) If a bare CR is encountered in a header line, an extra space is added
6759 after the line terminator so as not to end the header. The reasoning
6760 behind this is that bare CRs in header lines are most likely either
6761 to be mistakes, or people trying to play silly games.
6763 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6764 was being incorrectly given as 18 bytes larger than it should have been.
6765 This is a VOB (very old bug).
6767 60. This may never have affected anything current, but just in case it has:
6768 When the local host is found other than at the start of a list of hosts,
6769 the local host, those with the same MX, and any that follow, are discarded.
6770 When the list in question was part of a longer list of hosts, the following
6771 hosts (not currently being processed) were also being discarded. This no
6772 longer happens. I'm not sure if this situation could ever has previously
6775 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6778 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6779 header. This is something that is recommended in a new Internet Draft, and
6780 is something that is documented as being done by Sendmail. There are two
6781 possible values. For messages generated by the autoreply transport, Exim
6784 Auto-Submitted: auto-replied
6786 whereas for all other generated messages (e.g. bounces) it adds
6788 Auto-Submitted: auto-generated
6790 63. The "personal" condition in filters now includes a test for the
6791 Auto-Submitted: header. If it contains the string "auto-" the message it
6792 not considered personal.
6794 64. Added rcpt_include_affixes as a generic transport option.
6796 65. Added queue_only_override (default true).
6798 66. Added the syslog_duplication option.
6800 67. If what should have been the first header line of a message consisted of
6801 a space followed by a colon, Exim was mis-interpreting it as a header line.
6802 It isn't of course - it is syntactically invalid and should therefore be
6803 treated as the start of the message body. The misbehaviour could have
6804 caused a number of strange effects, including loss of data in subsequent
6805 header lines, and spool format errors.
6807 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6808 client host had authenticated. This control can now be exercised by an ACL
6809 for more flexibility.
6811 69. By default, callouts do not happen when testing with -bh. There is now a
6812 variant, -bhc, which does actually run the callout code, including
6813 consulting and updating the callout cache.
6815 70. Added support for saslauthd authentication, courtesy of Alexander
6818 71. If statvfs() failed on the spool or log directories while checking their
6819 size for availability, Exim confusingly gave the error "space shortage".
6820 Furthermore, in debugging mode it crashed with a floating point exception.
6821 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6822 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6823 serious problem, Exim now writes to the main and panic logs when this
6824 happens, with details of the failure. It then refuses to accept the
6825 incoming message, giving the message "spool directory problem" or "log
6826 directory problem" with a 421 code for SMTP messages.
6828 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6829 0, 1, and 2 exist, because some OS complain for execs without them (see
6830 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6831 descriptors. However, the code omitted to check that the open succeeded,
6832 causing mysterious errors if for some reason the permissions on /dev/null
6833 got screwed. Now Exim writes a message to the main and panic logs, and
6834 bombs out if it can't open /dev/null.
6836 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6837 interact so that it is all more flexible. It is supposed to remain
6838 backwards compatible. Also added extra_local_interfaces.
6840 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6841 to bomb out with an assertion failure - to the client this appears as a
6842 connection drop. This problem occurs in the part of the code that was taken
6843 from the Samba project. Fortunately, the assertion is in a very simple
6844 function, so I have fixed this by reproducing the function inline in the
6845 one place where it is called, and arranging for authentication to fail
6846 instead of killing the process with assert().
6848 75. The SPA client code was not working when the server requested OEM rather
6849 than Unicode encoding.
6851 76. Added code to make require_files with a specific uid setting more usable in
6852 the case where statting the file as root fails - usually a non-root-mounted
6853 NFS file system. When this happens and the failure is EACCES, Exim now
6854 forks a subprocess and does the per-uid checking as the relevant uid.
6856 77. Added process_log_path.
6858 78. If log_file_path was not explicitly set, a setting of check_log_space or
6859 check_log_inodes was ignored.
6861 79. If a space check for the spool or log partitions fails, the incident is now
6862 logged. Of course, in the latter case the data may get lost...
6864 80. Added the %p formatting code to string_format() so that it can be used to
6865 print addresses in debug_print(). Adjusted all the address printing in the
6866 debugging in store.c to use %p rather than %d.
6868 81. There was a concern that a line of code in smtp_in.c could overflow a
6869 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6870 initially expressed, the concern was not well-founded, because trailing
6871 spaces are removed early. However, if the trailing spaces were followed by
6872 a NULL, they did not get removed, so the overflow was possible. Two fixes
6875 (a) I re-wrote the offending code in a cleaner fashion.
6876 (b) If an incoming SMTP command contains a NULL character, it is rejected
6879 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6880 runs initgroups(), so that if the Exim user is in any additional groups,
6881 they will be used during message reception.
6887 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob