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 Replace the store_release() internal interface with store_newblock(),
12 which internalises the check required to safely use the old one, plus
13 the allocate and data copy operations duplicated in both (!) of the
16 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
17 modifier. This matches the restriction on the commandline.
22 JH/01 Rework error string handling in TLS interface so that the caller in
23 more cases is responsible for logging. This permits library-sourced
24 string to be attached to addresses during delivery, and collapses
25 pairs of long lines into single ones.
27 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
28 during configuration. Wildcards are allowed and expanded.
30 JH/02 Rework error string handling in DKIM to pass more info back to callers.
31 This permits better logging.
33 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
34 do not close it down and have the child transport start it up again on
35 the passed-on TCP connection. Instead, proxy the child (and any
36 subsequent ones) for TLS via a unix-domain socket channel. Logging is
37 affected: the continued delivery log lines do not have any DNSSEC, TLS
38 Certificate or OCSP information. TLS cipher information is still logged.
40 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
41 identical IP addresses on different listening ports. Will also affect
44 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
45 add noisy ifdef guards to special-case this sillyness.
46 Patch from Bernd Kuhls.
48 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
49 than 255 are no longer allowed.
51 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
52 Disable the session-cache too, which might reduce our load. Since we
53 currrectly use a new context for every connection, both as server and
54 client, there is no benefit for these.
55 GnuTLS appears to not support tickets server-side by default (we don't
56 call gnutls_session_ticket_enable_server()) but client side is enabled
57 by default on recent versions (3.1.3 +) unless the PFS priority string
60 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
61 <https://reproducible-builds.org/specs/source-date-epoch/>.
63 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
64 the check for any unsuccessful recipients did not notice the limit, and
65 erroneously found still-pending ones.
67 JH/08 Pipeline CHUNKING command and data together, on kernels that support
68 MSG_MORE. Only in-clear (not on TLS connections).
70 JH/09 Avoid using a temporary file during transport using dkim. Unless a
71 transport-filter is involved we can buffer the headers in memory for
72 creating the signature, and read the spool data file once for the
73 signature and again for transmission.
75 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
76 4.77 as the kernel support then wasn't solid, having issues in 64bit
77 mode. Now, it's been long enough. Add support for FreeBSD also.
79 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
80 case where the routing stage had gathered several addresses to send to
81 a host before calling the transport for the first, we previously failed
82 to close down TLS in the old transport process before passing the TCP
83 connection to the new process. The new one sent a STARTTLS command
84 which naturally failed, giving a failed delivery and bloating the retry
85 database. Investigation and fix prototype from Wolfgang Breyha.
87 JH/12 Fix check on SMTP command input synchronisation. Previously there were
88 false-negatives in the check that the sender had not preempted a response
89 or prompt from Exim (running as a server), due to that code's lack of
90 awareness of the SMTP input buffering.
92 PP/04 Add commandline_checks_require_admin option.
93 Exim drops privileges sanely, various checks such as -be aren't a
94 security problem, as long as you trust local users with access to their
95 own account. When invoked by services which pass untrusted data to
96 Exim, this might be an issue. Set this option in main configuration
97 AND make fixes to the calling application, such as using `--` to stop
100 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
101 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
102 responses to those, into a single TLS record each way (this usually means
103 a single packet). As a side issue, smtp_enforce_sync now works on TLS
106 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
107 affects you only if you're dancing at the edge of the param size limits.
108 If you are, and this message makes sense to you, then: raise the
109 configured limit or use OpenSSL 1.1. Nothing we can do for older
112 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
113 cmdline element to get the documented default one. Previously it was
116 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
117 are unsuable for non-message-specific reasons - eg. connection timeout,
118 banner-time rejection.
120 JH/16 Fix logging of delivery remote port, when specified by router, under
123 PP/06 Repair manualroute's ability to take options in any order, even if one
124 is the name of a transport.
127 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
129 JH/17 Change the list-building routines interface to use the expanding-string
130 triplet model, for better allocation and copying behaviour.
132 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
133 Previously it was constructed the first time a possibly-matching string
134 was met in the configuration file input during startup; now it is done
137 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
138 compatible one, to avoid the (poorly documented) possibility of a config
139 file in the working directory redirecting the DB files, possibly correpting
140 some existing file. CVE-2017-10140 assigned for BDB.
142 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
143 cache-hot. Previously, although the result was properly cached, the
144 initial verify call returned a defer.
146 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
147 the main verify for receipient in uncached-mode.
149 JH/22 Retire historical build files to an "unsupported" subdir. These are
150 defined as "ones for which we have no current evidence of testing".
152 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
153 if present. Previously it was ignored.
155 JH/24 Start using specified-initialisers in C structure init coding. This is
156 a C99 feature (it's 2017, so now considered safe).
158 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
159 if was a fixed-sized field and bitmask ops via macros; it is now more
162 PP/07 GitHub PR 56: Apply MariaDB build fix.
163 Patch provided by Jaroslav Škarvada.
165 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
166 during Coverity cleanups [4.87 JH/47]
167 Diagnosis and fix provided by Michael Fischer v. Mollard.
169 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
170 the right size to place the terminating semicolon on its own folded
171 line, the header hash was calculated to an incorrect value thanks to
172 the (relaxed) space the fold became.
174 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
175 and confused the parent.
177 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
178 which could crash as a result. This could lead to undeliverable messages.
180 JH/28 Logging: "next input sent too soon" now shows where input was truncated
183 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
184 matters on fast-turnover and PID-randomising systems, which were getting
185 out-of-order delivery.
187 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
188 a possibly-overlapping copy. The symptom was that "Remote host closed
189 connection in response to HELO" was logged instead of the actual 4xx
192 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
193 Previously only that bufferd was discarded, resulting in SYMTP command
196 JH/32 DKIM: when a message has multiple signatures matching an identity given
197 in dkim_verify_signers, run the dkim acl once for each. Previously only
198 one run was done. Bug 2189.
200 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
201 "panic the current process" to "deliberately defer". The panic log is
202 still written with the problem list name; the mail and reject logs now
203 get a temp-reject line for the message that was being handled, saying
204 something like "domains check lookup or other defer". The SMTP 451
205 message is still "Temporary local problem".
207 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
208 A crafted sequence of BDAT commands could result in in-use memory beeing
209 freed. CVE-2017-16943.
211 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
212 from SMTP input. Previously it was always done; now only done for DATA
213 and not BDAT commands. CVE-2017-16944.
215 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
216 to the message (such as an overlong header line). Previously this was
217 not done and we did not exit BDAT mode. Followon from the previous item
218 though a different problem.
224 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
225 than -2003 did; needs libidn2 in addition to libidn.
227 JH/02 The path option on a pipe transport is now expanded before use.
229 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
230 Patch provided by "Björn", documentation fix added too.
232 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
233 missing a wire-to-host endian conversion.
235 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
236 close after a BDAT command line could be taken as a following command,
237 giving a synch failure. Fix by only checking for synch immediately
238 before acknowledging the chunk.
240 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
241 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
242 macro. Patches provided by Josh Soref.
244 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
245 Previously we did not; the RFC seems ambiguous and VRFY is not listed
246 by IANA as a service extension. However, John Klensin suggests that we
249 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
250 the dkim code may be unix-mode line endings rather than smtp wire-format
251 CRLF, so prepend a CR to any bare LF.
253 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
254 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
256 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
257 the first were themselves being wrongly included in the feed into dkim
258 processing; with most chunk sizes in use this resulted in an incorrect
259 body hash calculated value.
261 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
262 DKIM signature block, for verification. Although advised against by
263 standards it is specifically not ruled illegal.
265 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
267 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
268 missing a body hash (the bh= tag).
270 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
271 It seems that HAProxy sends the Proxy Protocol information in clear and
272 only then does a TLS startup, so do the same.
274 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
275 TCP connections (such as for Spamd) unless the daemon successfully set
276 Fast Open mode on its listening sockets. This fixes breakage seen on
277 too-old kernels or those not configured for Fast Open, at the cost of
278 requiring both directions being enabled for TFO, and TFO never being used
279 by non-daemon-related Exim processes.
281 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
282 endings, at least on the first header line. Try to canonify any that get
283 past that check, despite the cost.
285 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
286 now limited to an arbitrary five deep, while parsing addresses with the
287 strip_excess_angle_brackets option enabled.
289 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
290 instead leave the unprompted TLS handshake in socket buffer for the
291 TLS library to consume.
293 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
295 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
297 JH/16 Drop variables when they go out of scope. Memory management drops a whole
298 region in one operation, for speed, and this leaves assigned pointers
299 dangling. Add checks run only under the testsuite which checks all
300 variables at a store-reset and panics on a dangling pointer; add code
301 explicitly nulling out all the variables discovered. Fixes one known
302 bug: a transport crash, where a dangling pointer for $sending_ip_address
303 originally assigned in a verify callout, is re-used.
305 PP/06 Drop '.' from @INC in various Perl scripts.
307 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
309 PP/08 Reduce a number of compilation warnings under clang; building with
310 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
311 should be warning-free.
313 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
315 HS/01 Fix portability problems introduced by PP/08 for platforms where
316 realloc(NULL) is not equivalent to malloc() [SunOS et al].
318 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
319 chunk. This allows us to accept broken chunked messages. We need a more
320 general solution here.
322 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
323 already-broken messages in the queue.
325 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
327 JH/19 Fix reference counting bug in routing-generated-address tracking.
333 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
334 supports it and a size is available (ie. the sending peer gave us one).
336 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
337 years of being deprecated). The replacements are the ACLs
338 acl_smtp_mime and acl_not_smtp_mime.
340 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
341 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
342 or even an in-clear connection were permitted. Now, if the host lookup
343 was dnssec and dane was requested then the host is only used if the
344 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
345 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
346 if one fails this test.
347 This means that a poorly-configured remote DNS will make it incommunicado;
348 but it protects against a DNS-interception attack on it.
350 JH/04 Bug 1810: make continued-use of an open smtp transport connection
351 non-noisy when a race steals the message being considered.
353 JH/05 If main configuration option tls_certificate is unset, generate a
354 self-signed certificate for inbound TLS connections.
356 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
357 in rewrites and routers.
359 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
360 and logged a warning sing 4.83; now they are a configuration file error.
362 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
363 (lacking @domain). Apply the same qualification processing as RCPT.
365 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
367 JH/10 Support ${sha256:} applied to a string (as well as the previous
370 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
371 a cutthrough deliver is pending, as we always want to make a connection.
372 This also avoids re-routing the message when later placing the cutthrough
373 connection after a verify cache hit.
374 Do not update it with the verify result either.
376 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
377 when routing results in more than one destination address.
379 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
380 signing (which inhibits the cutthrough capability). Previously only
381 the presence of an option was tested; now an expansion evaluating as
382 empty is permissible (obviously it should depend only on data available
383 when the cutthrough connection is made).
385 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
386 the relevant preceding SMTP command did not note the pipelining mode.
388 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
389 Previously they were not counted.
391 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
392 as one having no matching records. Previously we deferred the message
393 that needed the lookup.
395 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
396 distinguished as "(=".
398 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
399 for missing MX records. Previously it only worked for missing A records.
401 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
403 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
404 after the data-go-ahead and data-ack. Patch from Jason Betts.
406 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
407 even for a "none" policy. Patch from Tony Meyer.
409 JH/22 Fix continued use of a connection for further deliveries. If a port was
410 specified by a router, it must also match for the delivery to be
413 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
414 When one of the recipients of a message was unsuitable for the connection
415 (has no matching addresses), we lost track of needing to mark it
416 deferred. As a result mail would be lost.
418 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
420 JH/25 Decoding ACL controls is now done using a binary search; the source code
421 takes up less space and should be simpler to maintain. Merge the ACL
422 condition decode tables also, with similar effect.
424 JH/26 Fix problem with one_time used on a redirect router which returned the
425 parent address unchanged. A retry would see the parent address marked as
426 delivered, so not attempt the (identical) child. As a result mail would
429 JH/27 Fix a possible security hole, wherein a process operating with the Exim
430 UID can gain a root shell. Credit to http://www.halfdog.net/ for
431 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
434 JH/28 Enable {spool,log} filesystem space and inode checks as default.
435 Main config options check_{log,spool}_{inodes,space} are now
436 100 inodes, 10MB unless set otherwise in the configuration.
438 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
439 Previously it only applied to the main-section connection policy
442 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
444 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
445 by me. Added RFC7919 DH primes as an alternative.
447 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
448 are not in the system include path.
450 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
451 GnuTLS, when a session startup failed (eg because the client disconnected)
452 Exim did stdio operations after fclose. This was exposed by a recent
453 change which nulled out the file handle after the fclose.
455 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
456 signed directly by the cert-signing cert, rather than an intermediate
457 OCSP-signing cert. This is the model used by LetsEncrypt.
459 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
461 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
462 an incoming connection.
464 HS/02 Bug 1802: Do not half-close the connection after sending a request
467 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
468 fallback to "prime256v1".
470 JH/34 SECURITY: Use proper copy of DATA command in error message.
471 Could leak key material. Remotely exploitable. CVE-2016-9963.
477 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
478 and 3.4.4 - once the server is enabled to respond to an OCSP request
479 it does even when not requested, resulting in a stapling non-aware
480 client dropping the TLS connection.
482 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
483 support variable-length bit vectors. No functional change.
485 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
486 The I= interface field on outgoing lines is now after the H= remote
487 host field, same as incoming lines. There is a separate
488 outgoing_interface log selector which allows you to disable the
491 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
492 If not running log_selector +smtp_connection the mainlog would be held
493 open indefinitely after a "too many connections" event, including to a
494 deleted file after a log rotate. Leave the per net connection logging
495 leaving it open for efficiency as that will be quickly detected by the
496 check on the next write.
498 HS/01 Bug 1671: Fix post transport crash.
499 Processing the wait-<transport> messages could crash the delivery
500 process if the message IDs didn't exist for some reason. When
501 using 'split_spool_directory=yes' the construction of the spool
502 file name failed already, exposing the same netto behaviour.
504 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
505 mime_regex ACL conditions.
507 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
508 to DSN fail messages (bounces): remote IP, remote greeting, remote response
509 to HELO, local diagnostic string.
511 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
512 log line to debug. Even when configured with a tls authenticator many
513 client connections are expected to not authenticate in this way, so
514 an authenticate fail is not an error.
516 HS/02 Add the Exim version string to the process info. This way exiwhat
517 gives some more detail about the running daemon.
519 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
520 matter for fast-change records such as DNSBLs.
522 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
523 retry record, even if constant. There may be multiple transports with
524 different interface settings and the retry behaviour needs to be kept
527 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
529 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
531 JH/10 Bug 840: fix log_defer_output option of pipe transport
533 JH/11 Bug 830: use same host for all RCPTS of a message, even under
534 hosts_randomize. This matters a lot when combined with mua_wrapper.
536 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
537 ${quote_pgsql:<string>} operator.
539 JH/13 Bug 1708: avoid misaligned access in cached lookup.
541 JH/14 Change header file name for freeradius-client. Relevant if compiling
542 with Radius support; from the Gentoo tree and checked under Fedora.
544 JH/15 Bug 1712: Introduce $prdr_requested flag variable
546 JH/16 Bug 1714: Permit an empty string as expansion result for transport
547 option transport_filter, meaning no filtering.
549 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
551 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
552 defaults to "*" (all hosts). The variable is now available when not built
553 with TLS, default unset, mainly to enable keeping the testsuite sane.
554 If a server certificate is not supplied (via tls_certificate) an error is
555 logged, and clients will find TLS connections fail on startup. Presumably
556 they will retry in-clear.
557 Packagers of Exim are strongly encouraged to create a server certificate
558 at installation time.
560 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
561 with the $config_file variable.
563 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
564 in transport context, after the attempt, and per-recipient. The latter type
565 is per host attempted. The event data is the error message, and the errno
566 information encodes the lookup type (A vs. MX) used for the (first) host,
567 and the trailing two digits of the smtp 4xx response.
569 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
570 to write to mainlog (or rejectlog, paniclog) in the window between file
571 creation and permissions/ownership being changed. Particularly affects
572 installations where exicyclog is run as root, rather than exim user;
573 result is that the running daemon panics and dies.
575 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
577 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
578 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
579 "pri" and "weight". Note that the previous implicit priority given by the
580 list order is no longer honoured.
582 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
585 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
586 by defining SUPPORT_SOCKS.
588 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
589 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
590 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
591 variables are renamed to proxy_{local,external}_{address,port}.
593 JH/27 Move Internationalisation support from Experimental to mainline, enabled
594 for a build by defining SUPPORT_I18N
596 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
597 of the query string, and make ${quote_redis:} do that quoting.
599 JH/29 Move Events support from Experimental to mainline, enabled by default
600 and removable for a build by defining DISABLE_EVENT.
602 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
604 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
605 cached by the daemon.
607 JH/32 Move Redis support from Experimental to mainline, enabled for a build
608 by defining LOOKUP_REDIS. The libhiredis library is required.
610 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
611 keys are given for lookup.
613 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
614 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
615 only supported when built with TLS support. The PolarSSL SHA routines
616 are still used when the TLS library is too old for convenient support.
618 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
619 openssl_options), for security. OpenSSL forces this from version 1.1.0
620 server-side so match that on older versions.
622 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
623 allocation for $value could be released as the expansion processing
624 concluded, but leaving the global pointer active for it.
626 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
627 and to use the domains and local_parts ACL conditions.
629 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
630 incorrectly not doubled on cutthrough transmission, hence seen as a
631 body-termination at the receiving system - resulting in truncated mails.
632 Commonly the sender saw a TCP-level error, and retransmitted the message
633 via the normal store-and-forward channel. This could result in duplicates
634 received - but deduplicating mailstores were liable to retain only the
635 initial truncated version.
637 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
639 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
641 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
642 we're in there, support oversigning also; bug 1309.
644 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
646 HS/04 Add support for keep_environment and add_environment options.
648 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
649 either intentional arithmetic overflow during PRNG, or testing config-
652 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
653 delivery resulted in actual delivery. Cancel cutthrough before DATA
656 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
657 rejects a recipient: pass the reject to the originator.
659 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
660 Many were false-positives and ignorable, but it's worth fixing the
663 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
664 for the new environment-manipulation done at startup. Move the routines
665 from being local to tls.c to being global via the os.c file.
667 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
668 an extract embedded as result-arg for a map, the first arg for extract
669 is unavailable so we cannot tell if this is a numbered or keyed
670 extraction. Accept either.
676 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
679 JH/02 The smtp transport option "multi_domain" is now expanded.
681 JH/03 The smtp transport now requests PRDR by default, if the server offers
684 JH/04 Certificate name checking on server certificates, when exim is a client,
685 is now done by default. The transport option tls_verify_cert_hostnames
686 can be used to disable this per-host. The build option
687 EXPERIMENTAL_CERTNAMES is withdrawn.
689 JH/05 The value of the tls_verify_certificates smtp transport and main options
690 default to the word "system" to access the system default CA bundle.
691 For GnuTLS, only version 3.0.20 or later.
693 JH/06 Verification of the server certificate for a TLS connection is now tried
694 (but not required) by default. The verification status is now logged by
695 default, for both outbound TLS and client-certificate supplying inbound
698 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
701 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
702 Status Notification (bounce) messages are now MIME format per RFC 3464.
703 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
704 under the control of the dsn_advertise_hosts option, and routers may
705 have a dsn_lasthop option.
707 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
708 default, modifiable by a malware= option. The list separator for
709 the options can now be changed in the usual way. Bug 68.
711 JH/10 The smtp_receive_timeout main option is now expanded before use.
713 JH/11 The incoming_interface log option now also enables logging of the
714 local interface on delivery outgoing connections.
716 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
717 if the interface and destination host and port all match.
719 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
722 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
723 Patch from Andrew Lewis.
725 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
726 now supports optional time-restrictions, weighting, and priority
727 modifiers per server. Patch originally by <rommer@active.by>.
729 JH/16 The spamd_address main option now supports a mixed list of local
730 and remote servers. Remote servers can be IPv6 addresses, and
731 specify a port-range.
733 JH/17 Bug 68: The spamd_address main option now supports an optional
734 timeout value per server.
736 JH/18 Bug 1581: Router and transport options headers_add/remove can
737 now have the list separator specified.
739 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
742 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
745 JH/21 Support for the A6 type of dns record is withdrawn.
747 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
748 rather than the verbs used.
750 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
751 from 255 to 1024 chars.
753 JH/24 Verification callouts now attempt to use TLS by default.
755 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
756 are generic router options now. The defaults didn't change.
758 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
759 Original patch from Alexander Shikoff, worked over by JH.
761 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
762 files if ZCAT_COMMAND is not executable.
764 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
766 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
768 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
769 Normally benign, it bites when the pair was led to by a CNAME;
770 modern usage is to not canonicalize the domain to a CNAME target
771 (and we were inconsistent anyway for A-only vs AAAA+A).
773 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
775 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
776 when evaluating $sender_host_dnssec.
778 JH/31 Check the HELO verification lookup for DNSSEC, adding new
779 $sender_helo_dnssec variable.
781 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
783 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
785 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
787 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
788 documented as working, but never had. Support all but $spam_report.
790 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
791 added for tls authenticator.
793 HS/03 Add perl_taintmode main config option
799 TL/01 When running the test suite, the README says that variables such as
800 no_msglog_check are global and can be placed anywhere in a specific
801 test's script, however it was observed that placement needed to be near
802 the beginning for it to behave that way. Changed the runtest perl
803 script to read through the entire script once to detect and set these
804 variables, reset to the beginning of the script, and then run through
805 the script parsing/test process like normal.
807 TL/02 The BSD's have an arc4random API. One of the functions to induce
808 adding randomness was arc4random_stir(), but it has been removed in
809 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
810 function when detected.
812 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
813 cause callback expansion.
815 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
816 syntax errors in an expansion can be treated as a string instead of
817 logging or causing an error, due to the internal use of bool_lax
818 instead of bool when processing it.
820 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
821 server certificates when making smtp deliveries.
823 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
825 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
827 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
829 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
830 Merged patch from Sebastian Wiedenroth.
832 JH/05 Fix results-pipe from transport process. Several recipients, combined
833 with certificate use, exposed issues where response data items split
834 over buffer boundaries were not parsed properly. This eventually
835 resulted in duplicates being sent. This issue only became common enough
836 to notice due to the introduction of connection certificate information,
837 the item size being so much larger. Found and fixed by Wolfgang Breyha.
839 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
840 size buffer was used, resulting in syntax errors when an expansion
843 JH/07 Add support for directories of certificates when compiled with a GnuTLS
844 version 3.3.6 or later.
846 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
847 is EXPERIMENTAL_EVENT, the main-configuration and transport options
848 both become "event_action", the variables become $event_name, $event_data
849 and $event_defer_errno. There is a new variable $verify_mode, usable in
850 routers, transports and related events. The tls:cert event is now also
851 raised for inbound connections, if the main configuration event_action
854 TL/06 In test suite, disable OCSP for old versions of openssl which contained
855 early OCSP support, but no stapling (appears to be less than 1.0.0).
857 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
858 server certificate names available under the smtp transport option
859 "tls_verify_cert_hostname" now do not permit multi-component wildcard
862 JH/10 Time-related extraction expansions from certificates now use the main
863 option "timezone" setting for output formatting, and are consistent
864 between OpenSSL and GnuTLS compilations. Bug 1541.
866 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
867 encoded parameter in the incoming message. Bug 1558.
869 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
870 include certificate info, eximon was claiming there were spoolfile
873 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
875 JH/14 Log delivery-related information more consistently, using the sequence
876 "H=<name> [<ip>]" wherever possible.
878 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
879 are problematic for Debian distribution, omit them from the release
882 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
884 JH/16 Fix string representation of time values on 64bit time_t architectures.
887 JH/17 Fix a null-indirection in certextract expansions when a nondefault
888 output list separator was used.
893 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
894 checkers that were complaining about end of non-void function with no
897 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
898 This was a regression introduced in 4.83 by another bugfix.
900 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
902 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
903 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
909 TF/01 Correctly close the server side of TLS when forking for delivery.
911 When a message was received over SMTP with TLS, Exim failed to clear up
912 the incoming connection properly after forking off the child process to
913 deliver the message. In some situations the subsequent outgoing
914 delivery connection happened to have the same fd number as the incoming
915 connection previously had. Exim would try to use TLS and fail, logging
916 a "Bad file descriptor" error.
918 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
919 utilities have not been installed.
921 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
922 temporary space as the ACL may create new global variables.
924 TL/01 LDAP support uses per connection or global context settings, depending
925 upon the detected version of the libraries at build time.
927 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
928 to extract and use the src ip:port in logging and expansions as if it
929 were a direct connection from the outside internet. PPv2 support was
930 updated based on HAProxy spec change in May 2014.
932 JH/02 Add ${listextract {number}{list}{success}{fail}}.
934 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
935 Properly escape header and check for NULL return.
937 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
940 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
942 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
943 characters in header names, implemented as a verify condition.
944 Contributed by Michael Fischer v. Mollard.
946 TL/05 Rename SPF condition results err_perm and err_temp to standardized
947 results permerror and temperror. Previous values are deprecated but
948 still accepted. In a future release, err_perm and err_temp will be
949 completely removed, which will be a backward incompatibility if the
950 ACL tests for either of these two old results. Patch contributed by
951 user bes-internal on the mailing list.
953 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
955 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
956 selectors, in both main and reject logs.
958 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
961 JH/07 Add malware type "sock" for talking to simple daemon.
963 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
965 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
966 routers/transports under cutthrough routing.
968 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
969 numbers. Touch up "bool" conditional to keep the same definition.
971 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
973 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
975 JH/12 Expand items in router/transport headers_add or headers_remove lists
976 individually rather than the list as a whole. Bug 1452.
978 Required for reasonable handling of multiple headers_ options when
979 they may be empty; requires that headers_remove items with embedded
980 colons must have them doubled (or the list-separator changed).
982 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
983 view the policy declared in the DMARC record. Currently, $dmarc_status
984 is a combined value of both the record presence and the result of the
987 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
989 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
990 dnslookup router and the smtp transport (applying to the forward
993 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
994 of ldap servers used for a specific lookup. Patch provided by Heiko
997 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
998 New variable $lookup_dnssec_authenticated for observability.
1000 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1001 Patch submitted by Lars Timman.
1003 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1005 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1006 Requires trusted mode and valid format message id, aborts otherwise.
1007 Patch contributed by Heiko Schlichting.
1009 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1010 certextract with support for various fields. Bug 1358.
1012 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1013 is requested by default, modifiable by smtp transport option
1016 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1017 operate on certificate variables to give certificate fingerprints
1018 Also new ${sha256:cert_variable}.
1020 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1022 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1023 Christian Aistleitner.
1025 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1027 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1028 file. Patch from Wolfgang Breyha.
1030 JH/25 Expand the coverage of the delivery $host and $host_address to
1031 client authenticators run in verify callout. Bug 1476.
1033 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1034 align with daemon_smtp_ports. Bug 72.
1036 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1037 support and error reporting did not work properly.
1039 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1040 and is readable. Patch from Andrew Colin Kissa.
1042 TL/14 Enhance documentation of ${run expansion and how it parses the
1043 commandline after expansion, particularly in the case when an
1044 unquoted variable expansion results in an empty value.
1046 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1048 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1049 by renaming away from it. Addresses GH issue 15, reported by
1052 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1054 TL/15 SECURITY: prevent double expansion in math comparison functions
1055 (can expand unsanitized data). Not remotely exploitable.
1062 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1064 PP/02 Make -n do something, by making it not do something.
1065 When combined with -bP, the name of an option is not output.
1067 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1070 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1071 $sender_host_name and config options to manage this, and basic check
1074 PP/05 DSCP support for outbound connections and control modifier for inbound.
1076 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1077 (Only plugin which currently uses this is kerberos4, which nobody should
1078 be using, but we should make it available and other future plugins might
1079 conceivably use it, even though it would break NAT; stuff *should* be
1080 using channel bindings instead).
1082 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1083 name; added for Sendmail compatibility; requires admin caller.
1084 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1085 just ignore it); requires trusted caller.
1086 Also parse but ignore: -Ac -Am -X<logfile>
1089 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1091 TL/02 Add +smtp_confirmation as a default logging option.
1093 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1094 Patch by Magnus Holmgren from 2007-02-20.
1096 TL/04 Bugzilla 1281 - Spec typo.
1097 Bugzilla 1283 - Spec typo.
1098 Bugzilla 1290 - Spec grammar fixes.
1100 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1102 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1104 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1105 mailing list by by Dmitry Isaikin.
1107 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1109 JH/02 Support "G" suffix to numbers in ${if comparisons.
1111 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1113 NM/01 Bugzilla 1197 - Spec typo
1114 Bugzilla 1196 - Spec examples corrections
1116 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1118 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1119 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1122 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1123 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1125 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1126 "acl {{name}{arg}...}", and optional args on acl condition
1129 JH/05 Permit multiple router/transport headers_add/remove lines.
1131 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1133 JH/07 Avoid using a waiting database for a single-message-only transport.
1134 Performance patch from Paul Fisher. Bugzilla 1262.
1136 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1139 JH/09 Add $headers_added variable, with content from use of ACL modifier
1140 add_header (but not yet added to the message). Bugzilla 199.
1142 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1143 Pulled from Bugzilla 817 by Wolfgang Breyha.
1145 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1147 (nb: this is the same fix as in Exim 4.80.1)
1149 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1152 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1155 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1157 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1158 advertise SMTP AUTH mechanism to us, instead of a generic
1159 protocol violation error. Also, make Exim more robust to bad
1160 data from the Dovecot auth socket.
1162 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1164 When a queue runner is handling a message, Exim first routes the
1165 recipient addresses, during which it prunes them based on the retry
1166 hints database. After that it attempts to deliver the message to
1167 any remaining recipients. It then updates the hints database using
1170 So if a recipient address works intermittently, it can get repeatedly
1171 deferred at routing time. The retry hints record remains fresh so the
1172 address never reaches the final cutoff time.
1174 This is a fairly common occurrence when a user is bumping up against
1175 their storage quota. Exim had some logic in its local delivery code
1176 to deal with this. However it did not apply to per-recipient defers
1177 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1179 This change adds a proper retry rule check during routing so that the
1180 final cutoff time is checked against the message's age. We only do
1181 this check if there is an address retry record and there is not a
1182 domain retry record; this implies that previous attempts to handle
1183 the address had the retry_use_local_parts option turned on. We use
1184 this as an approximation for the destination being like a local
1185 delivery, as in LMTP.
1187 I suspect this new check makes the old local delivery cutoff check
1188 redundant, but I have not verified this so I left the code in place.
1190 TF/02 Correct gecos expansion when From: is a prefix of the username.
1192 Test 0254 submits a message to Exim with the header
1196 When I ran the test suite under the user fanf2, Exim expanded
1197 the header to contain my full name, whereas it should have added
1198 a Resent-Sender: header. It erroneously treats any prefix of the
1199 username as equal to the username.
1201 This change corrects that bug.
1203 GF/01 DCC debug and logging tidyup
1204 Error conditions log to paniclog rather than rejectlog.
1205 Debug lines prefixed by "DCC: " to remove any ambiguity.
1207 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1209 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1210 Bug spotted by Jeremy Harris; was flawed since initial commit.
1211 Would have resulted in OCSP responses post-SNI triggering an Exim
1212 NULL dereference and crash.
1214 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1216 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1217 Bug detection, analysis and fix by Samuel Thibault.
1218 Bugzilla 1331, Debian bug #698092.
1220 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1222 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1223 Server implementation by Todd Lyons, client by JH.
1224 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1225 config variable "prdr_enable" controls whether the server
1226 advertises the facility. If the client requests PRDR a new
1227 acl_data_smtp_prdr ACL is called once for each recipient, after
1228 the body content is received and before the acl_smtp_data ACL.
1229 The client is controlled by both of: a hosts_try_prdr option
1230 on the smtp transport, and the server advertisement.
1231 Default client logging of deliveries and rejections involving
1232 PRDR are flagged with the string "PRDR".
1234 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1235 fclose(). Diagnosis by Todd Lyons.
1237 PP/17 Update configure.default to handle IPv6 localhost better.
1238 Patch by Alain Williams (plus minor tweaks).
1241 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1242 This is now consistent with GnuTLS, and is now documented: the
1243 previous undocumented portable approach to treating the option as
1244 unset was to force an expansion failure. That still works, and
1245 an empty string is now equivalent.
1247 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1248 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1249 not performing validation itself.
1251 PP/20 Added force_command boolean option to pipe transport.
1252 Patch from Nick Koston, of cPanel Inc.
1254 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1257 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1259 PP/21 Fix eximon continuous updating with timestamped log-files.
1260 Broken in a format-string cleanup in 4.80, missed when I repaired the
1261 other false fix of the same issue.
1262 Report and fix from Heiko Schlichting.
1265 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1266 Report from Prashanth Katuri.
1268 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1269 It's SecureTransport, so affects any MacOS clients which use the
1270 system-integrated TLS libraries, including email clients.
1272 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1273 using a MIME ACL for non-SMTP local injection.
1274 Report and assistance in diagnosis by Warren Baker.
1276 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1278 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1280 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1281 last id that failed so it may be referenced in subsequent ACL's.
1283 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1286 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1287 ldap library initialization, allowing self-signed CA's to be
1288 used. Also properly sets require_cert option later in code by
1289 using NULL (global ldap config) instead of ldap handle (per
1290 session). Bug diagnosis and testing by alxgomz.
1292 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1293 the src/util/ subdirectory.
1295 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1296 renamed to Transport Post Delivery Action by Jeremy Harris, as
1299 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1300 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1301 redis_servers = needs to be configured which will be used by the redis
1302 lookup. Patch from Warren Baker, of The Packet Hub.
1304 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1306 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1307 hostname or reverse DNS when processing a host list. Used suggestions
1308 from multiple comments on this bug.
1310 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1312 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1313 Missed a few lines, added it to make the runtest require no keyboard
1316 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1317 contains upper case chars. Make router use caseful_local_part.
1319 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1320 support when GnuTLS has been built with p11-kit.
1326 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1328 This, or similar/improved, will also be change PP/11 of 4.82.
1334 PP/01 Handle short writes when writing local log-files.
1335 In practice, only affects FreeBSD (8 onwards).
1336 Bugzilla 1053, with thanks to Dmitry Isaikin.
1338 NM/01 Bugzilla 949 - Documentation tweak
1340 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1343 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1345 PP/02 Implemented gsasl authenticator.
1347 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1349 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1350 `pkg-config foo` for cflags/libs.
1352 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1353 with rest of GSASL and with heimdal_gssapi.
1355 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1356 `pkg-config foo` for cflags/libs for the TLS implementation.
1358 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1359 properties get this fed in as external SSF. A number of robustness
1360 and debugging improvements to the cyrus_sasl authenticator.
1362 PP/08 cyrus_sasl server now expands the server_realm option.
1364 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1365 Patch by Jeremy Harris.
1367 PP/10 Added dbmjz lookup type.
1369 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1371 PP/12 MAIL args handles TAB as well as SP, for better interop with
1372 non-compliant senders.
1373 Analysis and variant patch by Todd Lyons.
1375 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1376 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1377 Patch from Dirk Mueller <dmueller@suse.com>
1379 PP/13 tls_peerdn now print-escaped for spool files.
1380 Observed some $tls_peerdn in wild which contained \n, which resulted
1381 in spool file corruption.
1383 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1384 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1385 or write after TLS renegotiation, which otherwise led to messages
1388 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1389 as a tracking header (ie: a signed header comes before the signature).
1390 Patch from Wolfgang Breyha.
1392 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1393 comma-sep list; embedded commas doubled.
1395 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1397 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1399 Report and patch from Dmitry Banschikov.
1401 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1402 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1403 failures. We appear to now support TLS1.1+ with Exim.
1405 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1406 lets Exim select keys and certificates based upon TLS SNI from client.
1407 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1408 before an outbound SMTP session. New log_selector, +tls_sni.
1410 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1411 NULL dereference. Report and patch from Alun Jones.
1413 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1414 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1415 Not seeing resolver debug output on NetBSD, but suspect this is a
1416 resolver implementation change.
1418 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1419 Left warnings. Added "eximon gdb" invocation mode.
1421 PP/21 Defaulting "accept_8bitmime" to true, not false.
1423 PP/22 Added -bw for inetd wait mode support.
1425 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1426 locate the relevant includes and libraries. Made this the default.
1428 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1429 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1431 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1432 This may cause build issues on older platforms.
1434 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1435 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1436 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1437 Added SNI support via GnuTLS too.
1438 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1440 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1442 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1443 Applied second patch from Janne, implementing suggestion to default
1444 multiple-strings-in-record handling to match SPF spec.
1446 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1448 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1449 read-only, out of scope).
1450 Patch from Wolfgang Breyha, report from Stuart Northfield.
1452 PP/29 Fix three issues highlighted by clang analyser static analysis.
1453 Only crash-plausible issue would require the Cambridge-specific
1454 iplookup router and a misconfiguration.
1455 Report from Marcin Mirosław.
1457 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1459 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1460 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1461 As part of this, removing so much warning spew let me fix some minor
1462 real issues in debug logging.
1464 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1465 assignment on my part. Fixed.
1467 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1468 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1469 Janne Snabb (who went above and beyond: thank you).
1471 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1472 string otherwise requires a connection and a bunch more work and it's
1473 relatively easy to get wrong. Should also expose TLS library linkage
1476 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1477 64-bit ${eval} (JH/03).
1479 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1480 GNU libc to support some of the 64-bit stuff, should not lead to
1481 conflicts. Defined before os.h is pulled in, so if a given platform
1482 needs to override this, it can.
1484 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1485 protection layer was required, which is not implemented.
1486 Bugzilla 1254, patch from Wolfgang Breyha.
1488 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1489 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1490 tls_dhparam take prime identifiers. Also unbreak combination of
1491 OpenSSL+DH_params+TLSSNI.
1493 PP/39 Disable SSLv2 by default in OpenSSL support.
1499 PP/01 Solaris build fix for Oracle's LDAP libraries.
1500 Bugzilla 1109, patch from Stephen Usher.
1502 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1504 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1507 TF/02 Fix a couple more cases where we did not log the error message
1508 when unlink() failed. See also change 4.74-TF/03.
1510 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1511 lock up or crash if it happened to be inside a call to libc when it
1512 got a SIGUSR1 from exiwhat.
1514 The SIGUSR1 handler appends the current process status to the process
1515 log which is later printed by exiwhat. It used to use the general
1516 purpose logging code to do this, but several functions it calls are
1517 not safe for signals.
1519 The new output code in the SIGUSR1 handler is specific to the process
1520 log, and simple enough that it's easy to inspect for signal safety.
1521 Removing some special cases also simplifies the general logging code.
1522 Removing the spurious timestamps from the process log simplifies
1525 TF/04 Improved ratelimit ACL condition.
1527 The /noupdate option has been deprecated in favour of /readonly which
1528 has clearer semantics. The /leaky, /strict, and /readonly update modes
1529 are mutually exclusive. The update mode is no longer included in the
1530 database key; it just determines when the database is updated. (This
1531 means that when you upgrade Exim will forget old rate measurements.)
1533 Exim now checks that the per_* options are used with an update mode that
1534 makes sense for the current ACL. For example, when Exim is processing a
1535 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1536 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1537 must specify per_mail/readonly. If you omit the update mode it defaults to
1538 /leaky where that makes sense (as before) or /readonly where required.
1540 The /noupdate option is now undocumented but still supported for
1541 backwards compatibility. It is equivalent to /readonly except that in
1542 ACLs where /readonly is required you may specify /leaky/noupdate or
1543 /strict/noupdate which are treated the same as /readonly.
1545 A useful new feature is the /count= option. This is a generalization
1546 of the per_byte option, so that you can measure the throughput of other
1547 aggregate values. For example, the per_byte option is now equivalent
1548 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1550 The per_rcpt option has been generalized using the /count= mechanism
1551 (though it's more complicated than the per_byte equivalence). When it is
1552 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1553 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1554 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1555 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1556 Note that using per_rcpt with a non-readonly update mode in more than
1557 one ACL will cause the recipients to be double-counted. (The per_mail
1558 and per_byte options don't have this problem.)
1560 The handling of very low rates has changed slightly. If the computed rate
1561 is less than the event's count (usually one) then this event is the first
1562 after a long gap. In this case the rate is set to the same as this event's
1563 count, so that the first message of a spam run is counted properly.
1565 The major new feature is a mechanism for counting the rate of unique
1566 events. The new per_addr option counts the number of different
1567 recipients that someone has sent messages to in the last time period. It
1568 behaves like per_rcpt if all the recipient addresses are different, but
1569 duplicate recipient addresses do not increase the measured rate. Like
1570 the /count= option this is a general mechanism, so the per_addr option
1571 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1572 example, measure the rate that a client uses different sender addresses
1573 with the options per_mail/unique=$sender_address. There are further
1574 details in the main documentation.
1576 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1578 TF/06 Removed a few PCRE remnants.
1580 TF/07 Automatically extract Exim's version number from tags in the git
1581 repository when doing development or release builds.
1583 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1584 Bugzilla 879. Patch from Paul Fisher.
1586 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1587 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1590 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1592 PP/05 Variable $av_failed, true if the AV scanner deferred.
1593 Bugzilla 1078. Patch from John Horne.
1595 PP/06 Stop make process more reliably on build failure.
1596 Bugzilla 1087. Patch from Heiko Schlittermann.
1598 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1599 Bugzilla 1089. Patch from Heiko Schlittermann.
1601 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1602 Bugzilla 1131. Patch from Holger Weiß.
1604 PP/09 Handle IPv6 addresses with SPF.
1605 Bugzilla 860. Patch from Wolfgang Breyha.
1607 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1609 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1612 PP/11 match_* no longer expand right-hand-side by default.
1613 New compile-time build option, EXPAND_LISTMATCH_RHS.
1614 New expansion conditions, "inlist", "inlisti".
1616 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1618 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1620 PP/14 fix log_write() format string regression from TF/03.
1621 Bugzilla 1152. Patch from Dmitry Isaikin.
1627 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1629 PP/02 Harmonised TLS library version reporting; only show if debugging.
1630 Layout now matches that introduced for other libraries in 4.74 PP/03.
1632 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1634 PP/04 New "dns_use_edns0" global option.
1636 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1639 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1640 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1642 TK/01 Updated PolarSSL code to 0.14.2.
1643 Bugzilla 1097. Patch from Andreas Metzler.
1645 PP/07 Catch divide-by-zero in ${eval:...}.
1646 Fixes bugzilla 1102.
1648 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1651 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1652 format-string attack -- SECURITY: remote arbitrary code execution.
1654 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1655 time unintentionally subject to list matching rules, letting the header
1656 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1657 arbitrary string expansion). This allowed for information disclosure.
1659 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1660 INT_MIN/-1 -- value coerced to INT_MAX.
1666 NM/01 Workaround for PCRE version dependency in version reporting
1669 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1670 This fixes portability to compilers other than gcc, notably
1671 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1673 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1674 makefiles for portability to HP-UX and POSIX correctness.
1676 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1677 Also via indented variable definition in the Makefile.
1678 (Debugging by Oliver Heesakkers).
1680 PP/02 Restore caching of spamd results with expanded spamd_address.
1681 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1683 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1684 Improves build reliability. Fix from: Frank Elsner
1686 NM/02 Fix wide character breakage in the rfc2047 coding
1687 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1689 NM/03 Allow underscore in dnslist lookups
1690 Fixes bug 1026. Patch from Graeme Fowler
1692 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1693 Code patches from Adam Ciarcinski of NetBSD.
1695 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1698 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1699 is logged, to avoid truncation. Patch from John Horne.
1701 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1702 Patch from Jakob Hirsch.
1704 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1705 SQL string expansion failure details.
1706 Patch from Andrey Oktyabrski.
1708 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1709 Patch from Simon Arlott.
1711 PP/09 New lookups functionality failed to compile on old gcc which rejects
1712 extern declarations in function scope.
1713 Patch from Oliver Fleischmann
1715 PP/10 Use sig_atomic_t for flags set from signal handlers.
1716 Check getgroups() return and improve debugging.
1717 Fixed developed for diagnosis in bug 927 (which turned out to be
1720 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1721 Patch from Mark Zealey.
1723 PP/12 Bugzilla 1056: Improved spamd server selection.
1724 Patch from Mark Zealey.
1726 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1727 Based on patch from Heiko Schlittermann.
1729 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1730 Patch from Uwe Doering, sign-off by Michael Haardt.
1732 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1733 variable declaration deep within a block. Bug and patch from
1736 PP/15 lookups-Makefile IRIX compatibility coercion.
1738 PP/16 Make DISABLE_DKIM build knob functional.
1740 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1741 Patch by Simon Arlott
1743 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1744 variable argument macros. Our copy now differs from upstream.
1750 TF/01 Failure to get a lock on a hints database can have serious
1751 consequences so log it to the panic log.
1753 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1754 controlled using the smtp_confirmation log selector.
1756 TF/03 Include the error message when we fail to unlink a spool file.
1758 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1759 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1760 for maintaining out-of-tree patches for some time.
1762 PP/01 Bugzilla 139: Documentation and portability issues.
1763 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1764 Handle per-OS dynamic-module compilation flags.
1766 PP/02 Let /dev/null have normal permissions.
1767 The 4.73 fixes were a little too stringent and complained about the
1768 permissions on /dev/null. Exempt it from some checks.
1769 Reported by Andreas M. Kirchwitz.
1771 PP/03 Report version information for many libraries, including
1772 Exim version information for dynamically loaded libraries. Created
1773 version.h, now support a version extension string for distributors
1774 who patch heavily. Dynamic module ABI change.
1776 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1777 privilege escalation vulnerability whereby the Exim run-time user
1778 can cause root to append content of the attacker's choosing to
1781 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1784 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1785 If dropping privileges for untrusted macros, we disabled normal logging
1786 on the basis that it would fail; for the Exim run-time user, this is not
1787 the case, and it resulted in successful deliveries going unlogged.
1788 Fixed. Reported by Andreas Metzler.
1794 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1795 only prepend for the Resent-* case. Fixes regression introduced in
1796 Exim 4.70 by NM/22 for Bugzilla 607.
1798 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1799 increasing numbers of administrators be bitten by this.
1801 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1803 PP/03 Bugzilla 994: added openssl_options main configuration option.
1805 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1807 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1809 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1811 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1812 without a peer certificate, leading to a segfault because of an
1813 assumption that peers always have certificates. Be a little more
1814 paranoid. Problem reported by Martin Tscholak.
1816 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1817 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1818 NB: ClamAV planning to remove STREAM in "middle of 2010".
1819 CL also introduces -bmalware, various -d+acl logging additions and
1820 more caution in buffer sizes.
1822 PP/09 Implemented reverse_ip expansion operator.
1824 PP/10 Bugzilla 937: provide a "debug" ACL control.
1826 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1828 PP/12 Bugzilla 973: Implement --version.
1830 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1832 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1834 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1836 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1837 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1838 ignore trailing whitespace.
1840 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1842 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1845 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1846 Notification from Dr Andrew Aitchison.
1848 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1849 ExtendedDetectionInfo response format.
1850 Notification from John Horne.
1852 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1855 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1856 XSL and documented dependency on system catalogs, with examples of how
1859 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1862 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1863 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1864 configuration file which is writeable by the Exim user or group.
1866 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1867 of configuration files to cover files specified with the -C option if
1868 they are going to be used with root privileges, not just the default
1871 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1872 option (effectively making it always true).
1874 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1875 files to be used while preserving root privileges.
1877 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1878 that rogue child processes cannot use them.
1880 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1881 run-time user, instead of root.
1883 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1884 Exim run-time user without dropping privileges.
1886 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1887 result string, instead of calling string_vformat() twice with the same
1890 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1891 for other users. Others should always drop root privileges if they use
1892 -C on the command line, even for a whitelisted configure file.
1894 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1896 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1902 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1903 $data_path, and $header_path variables; fixed documentation bugs and
1906 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1907 exipick to access non-standard spools, including the "frozen" queue
1910 NM/01 Bugzilla 965: Support mysql stored procedures.
1911 Patch from Alain Williams
1913 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1915 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1916 Patch from Andreas Metzler
1918 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1919 Patch from Kirill Miazine
1921 NM/05 Bugzilla 671: Added umask to procmail example.
1923 JJ/03 installed exipick 20100323.0, fixing doc bug
1925 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1926 directory. Notification and patch from Dan Rosenberg.
1928 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1930 TK/02 Improve log output when DKIM signing operation fails.
1932 MH/01 Treat the transport option dkim_domain as a colon separated
1933 list, not as a single string, and sign the message with each element,
1934 omitting multiple occurences of the same signer.
1936 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1937 Bugzilla 985, 986. Patch by Simon Arlott
1939 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1940 Patch by Simon Arlott
1942 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1943 MBX locking. Notification from Dan Rosenberg.
1949 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1951 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1953 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1955 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1957 NM/04 Bugzilla 915: Flag broken perl installation during build.
1963 TK/01 Added patch by Johannes Berg that expands the main option
1964 "spamd_address" if it starts with a dollar sign.
1966 TK/02 Write list of recipients to X-Envelope-Sender header when building
1967 the mbox-format spool file for content scanning (suggested by Jakob
1970 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1971 (http://www.dcc-servers.net/) support via dccifd. Activated by
1972 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1974 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1975 by Mark Daniel Reidel <mr@df.eu>.
1977 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1978 When building exim an external PCRE library is now needed -
1979 PCRE is a system library on the majority of modern systems.
1980 See entry on PCRE_LIBS in EDITME file.
1982 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1983 conversation. Added nologin parameter to request.
1984 Patch contributed by Kirill Miazine.
1986 TF/01 Do not log submission mode rewrites if they do not change the address.
1988 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1990 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1991 log files in place. Contributed by Roberto Lima.
1993 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1995 TF/03 Bugzilla 615: When checking the local_parts router precondition
1996 after a local_part_suffix or local_part_prefix option, Exim now
1997 does not use the address's named list lookup cache, since this
1998 contains cached lookups for the whole local part.
2000 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2001 Robert Millan. Documentation is in experimental-spec.txt.
2003 TF/04 Bugzilla 668: Fix parallel build (make -j).
2005 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2007 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2008 Patch provided by Jan Srzednicki.
2010 TF/05 Leading white space used to be stripped from $spam_report which
2011 wrecked the formatting. Now it is preserved.
2013 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2014 that they are available at delivery time.
2016 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2018 TF/08 TLS error reporting now respects the incoming_interface and
2019 incoming_port log selectors.
2021 TF/09 Produce a more useful error message if an SMTP transport's hosts
2022 setting expands to an empty string.
2024 NM/06 Bugzilla 744: EXPN did not work under TLS.
2025 Patch provided by Phil Pennock.
2027 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2028 Patch provided by Richard Godbee.
2030 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2031 acl_smtp_notquit, added index entry.
2033 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2034 Patch provided by Eugene Bujak.
2036 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2037 accept(). Patch provided by Maxim Dounin.
2039 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2040 Patch provided by Phil Pennock.
2042 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2044 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2045 Patch provided by Brad "anomie" Jorsch.
2047 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2048 Patch provided by Dean Brooks.
2050 TK/05 Add native DKIM support (does not depend on external libraries).
2052 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2053 Patch provided by Graeme Fowler.
2055 NM/16 Bugzilla 851: Documentation example syntax fix.
2057 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2059 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2062 NM/19 Bugzilla 745: TLS version reporting.
2063 Patch provided by Phil Pennock.
2065 NM/20 Bugzilla 167: bool: condition support.
2066 Patch provided by Phil Pennock.
2068 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2069 clients. Patch provided by Phil Pennock.
2071 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2072 Patch provided by Brad "anomie" Jorsch.
2074 NM/23 Bugzilla 687: Fix misparses in eximstats.
2075 Patch provided by Heiko Schlittermann.
2077 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2078 Patch provided by Heiko Schlittermann.
2080 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2081 plus update to original patch.
2083 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2085 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2086 Patch provided by David Brownlee.
2088 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2090 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2092 NM/30 Bugzilla 888: TLS documentation bugfixes.
2094 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2096 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2097 Unlike the original bugzilla I have changed all shell scripts in src tree.
2099 NM/33 Bugzilla 898: Transport filter timeout fix.
2100 Patch by Todd Rinaldo.
2102 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2103 Patch by Serge Demonchaux.
2105 NM/35 Bugzilla 39: Base64 decode bug fixes.
2106 Patch by Jakob Hirsch.
2108 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2110 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2112 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2114 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2120 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2121 ALT-N's libdkim that I have put here:
2122 http://duncanthrax.net/exim-experimental/
2124 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2125 were called 'true' and it seems that C99 defines that as a reserved
2126 keyword to be used with 'bool' variable types. That means you could
2127 not include C99-style headers which use bools without triggering
2128 build errors in sieve.c.
2130 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2131 as mailq or other aliases. Changed the --help handling significantly
2132 to do whats expected. exim_usage() emits usage/help information.
2134 SC/01 Added the -bylocaldomain option to eximstats.
2136 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2138 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2140 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2146 PH/01 Another patch from the Sieve maintainer.
2148 PH/02 When an IPv6 address is converted to a string for single-key lookup
2149 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2150 dots are used instead of colons so that keys in lsearch files need not
2151 contain colons. This was done some time before quoting was made available
2152 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2153 (notated using the quote facility) so as to distinguish them from IPv4
2154 keys. This meant that lookups for IP addresses in host lists did not work
2155 for iplsearch lookups.
2157 This has been fixed by arranging for IPv6 addresses to be expressed with
2158 colons if the lookup type is iplsearch. This is not incompatible, because
2159 previously such lookups could never work.
2161 The situation is now rather anomalous, since one *can* have colons in
2162 ordinary lsearch keys. However, making the change in all cases is
2163 incompatible and would probably break a number of configurations.
2165 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2168 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2169 conversion specification without a maximum field width, thereby enabling
2170 a rogue spamd server to cause a buffer overflow. While nobody in their
2171 right mind would setup Exim to query an untrusted spamd server, an
2172 attacker that gains access to a server running spamd could potentially
2173 exploit this vulnerability to run arbitrary code as the Exim user.
2175 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2176 $primary_hostname instead of what libspf2 thinks the hosts name is.
2178 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2179 a directory entry by the name of the lookup key. Previously, if a
2180 symlink pointed to a non-existing file or a file in a directory that
2181 Exim lacked permissions to read, a lookup for a key matching that
2182 symlink would fail. Now it is enough that a matching directory entry
2183 exists, symlink or not. (Bugzilla 503.)
2185 PH/03 The body_linecount and body_zerocount variables are now exported in the
2188 PH/04 Added the $dnslist_matched variable.
2190 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2191 This means they are set thereafter only if the connection becomes
2194 PH/06 Added the client_condition to authenticators so that some can be skipped
2195 by clients under certain conditions.
2197 PH/07 The error message for a badly-placed control=no_multiline_responses left
2198 "_responses" off the end of the name.
2200 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2202 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2203 (without spaces) instead of just copying the configuration text.
2205 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2207 PH/11 Added $max_received_linelength.
2209 PH/12 Added +ignore_defer and +include_defer to host lists.
2211 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2212 way in which PCRE > 7.0 is built.
2214 PH/14 Implemented queue_only_load_latch.
2216 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2217 MAIL command. The effect was to mangle the value on 64-bit systems.
2219 PH/16 Another patch from the Sieve maintainer.
2221 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2223 PH/18 If a system quota error occurred while trying to create the file for
2224 a maildir delivery, the message "Mailbox is full" was not appended to the
2225 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2226 only to a quota excession during the actual writing of the file.
2228 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2229 characters?) which causes problems in log lines. The DN values are now
2230 passed through string_printing() before being added to log lines.
2232 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2233 and InterBase are left for another time.)
2235 PH/21 Added message_body_newlines option.
2237 PH/22 Guard against possible overflow in moan_check_errorcopy().
2239 PH/23 POSIX allows open() to be a macro; guard against that.
2241 PH/24 If the recipient of an error message contained an @ in the local part
2242 (suitably quoted, of course), incorrect values were put in $domain and
2243 $local_part during the evaluation of errors_copy.
2249 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2250 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2253 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2254 issue a MAIL command.
2256 PH/02 In an ACL statement such as
2258 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2260 if a client was not listed at all, or was listed with a value other than
2261 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2262 the condition was not true (as it should be), so access was not denied.
2263 The bug was that the ! inversion was incorrectly passed on to the second
2264 item. This has been fixed.
2266 PH/03 Added additional dnslists conditions == and =& which are different from
2267 = and & when the dns lookup returns more than one IP address.
2269 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2270 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2272 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2273 FSYNC, which compiles an option called disable_fsync that allows for
2274 bypassing fsync(). The documentation is heavily laced with warnings.
2276 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2278 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2279 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2280 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2281 including adding "make clean"; (3) Added -fPIC when compiling the test
2282 dynamically loaded module, to get rid of a warning.
2284 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2285 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2286 = 0s. The bug is otherwise harmless.
2288 PH/07 There was a bug in the dovecot authenticator such that the value of
2289 $auth1 could be overwritten, and so not correctly preserved, after a
2290 successful authentication. This usually meant that the value preserved by
2291 the server_setid option was incorrect.
2293 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2295 PH/09 Installed PCRE release 7.0.
2297 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2298 run for batched SMTP input. It is now run at the start of every message
2299 in the batch. While fixing this I discovered that the process information
2300 (output by running exiwhat) was not always getting set for -bs and -bS
2301 input. This is fixed, and it now also says "batched" for BSMTP.
2303 PH/11 Added control=no_pipelining.
2305 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2306 patch, slightly modified), and move the expansion of helo_data till after
2307 the connection is made in the smtp transport (so it can use these
2310 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2312 PH/14 Added log_selector = +pid.
2314 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2316 PH/16 Add ${if forany and ${if forall.
2318 PH/17 Added dsn_from option to vary the From: line in DSNs.
2320 PH/18 Flush SMTP output before performing a callout, unless control =
2321 no_callout_flush is set.
2323 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2324 was true (the default) a successful delivery failed to delete the retry
2325 item, thus causing premature timeout of the address. The bug is now
2328 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2330 PH/21 Long custom messages for fakedefer and fakereject are now split up
2331 into multiline responses in the same way that messages for "deny" and
2332 other ACL rejections are.
2334 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2335 with slight modification.
2337 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2338 draft, changing the syntax and factoring some duplicate code".
2340 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2341 for deliveries of the second and subsequent messages over the same SMTP
2344 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2345 ${reduce, with only minor "tidies".
2347 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2349 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2350 expansion side effects.
2352 PH/27 When a message times out after an over-quota error from an Exim-imposed
2353 quota, the bounce message says "mailbox is full". This message was not
2354 being given when it was a system quota that was exceeded. It now should
2357 MH/03 Made $recipients available in local_scan(). local_scan() already has
2358 better access to the recipient list through recipients_list[], but
2359 $recipients can be useful in postmaster-provided expansion strings.
2361 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2362 in the case of a MAIL command with additional options following the
2363 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2364 were accidentally chopped off.
2366 PH/29 SMTP synchronization checks are implemented when a command is read -
2367 there is a check that no more input is waiting when there shouldn't be
2368 any. However, for some commands, a delay in an ACL can mean that it is
2369 some time before the response is written. In this time, more input might
2370 arrive, invalidly. So now there are extra checks after an ACL has run for
2371 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2372 pipelining has not been advertised.
2374 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2376 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2377 returned in the SMTP response when smtp_return_error_details was false.
2378 This has been fixed.
2380 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2381 instead of the C I/O that was originally supplied, because problems were
2382 reported on Solaris.
2384 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2385 Exim which did not show up earlier: it was assuming that a call to
2386 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2387 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2388 was a macro that became an assignment, so it seemed to work. This has
2389 changed to a proper function call with a void return, hence the compile
2390 error. Exim's code has been fixed.
2392 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2395 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2397 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2399 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2400 because it turns out to be too big or there is a timeout) there is no
2401 "Completed" line in the log. When some messages of this type were
2402 selected by exigrep, they were listed as "not completed". Others were
2403 picked up by some special patterns. I have improved the selection
2404 criteria to be more general.
2406 PH/38 The host_find_failed option in the manualroute router can now be set
2407 to "ignore", to completely ignore a host whose IP address cannot be
2408 found. If all hosts are ignored, the behaviour is controlled by the new
2409 host_all_ignored option.
2411 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2412 homing or because of multiple MX records with /mx) generated more than
2413 one IP address, and the following item turned out to be the local host,
2414 all the secondary addresses of the first item were incorrectly removed
2415 from the list, along with the local host and any following hosts (which
2416 is what is supposed to happen).
2418 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2419 whoever called Exim into the -H file. In the case of the daemon it was
2420 behaving confusingly. When first started, it used values for whoever
2421 started the daemon, but after a SIGHUP it used the Exim user (because it
2422 calls itself on a restart). I have changed the code so that it now always
2425 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2426 message are rejected with the same error (e.g. no authentication or bad
2427 sender address), and a DATA command is nevertheless sent (as can happen
2428 with PIPELINING or a stupid MUA), the error message that was given to the
2429 RCPT commands is included in the rejection of the DATA command. This is
2430 intended to be helpful for MUAs that show only the final error to their
2433 PH/42 Another patch from the Sieve maintainer.
2435 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2436 Eximstats - Fixed some broken HTML links and added missing column headers
2438 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2439 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2441 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2443 PH/43 Yet another patch from the Sieve maintainer.
2445 PH/44 I found a way to check for a TCP/IP connection going away before sending
2446 the response to the final '.' that terminates a message, but only in the
2447 case where the client has not sent further data following the '.'
2448 (unfortunately, this is allowed). However, in many cases there won't be
2449 any further data because there won't be any more messages to send. A call
2450 to select() can be used: if it shows that the input is "ready", there is
2451 either input waiting, or the socket has been closed. An attempt to read
2452 the next input character can distinguish the two cases. Previously, Exim
2453 would have sent an OK response which the client would never have see.
2454 This could lead to message repetition. This fix should cure that, at
2455 least in a lot of common cases.
2457 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2458 advertised in response to EHLO.
2464 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2465 fixed by 4.65/MH/01 (is this a record?) are fixed:
2467 (i) An empty string was always treated as zero by the numeric comparison
2468 operators. This behaviour has been restored.
2470 (ii) It is documented that the numeric comparison operators always treat
2471 their arguments as decimal numbers. This was broken in that numbers
2472 starting with 0 were being interpreted as octal.
2474 While fixing these problems I realized that there was another issue that
2475 hadn't been noticed. Values of message_size_limit (both the global option
2476 and the transport option) were treated as octal if they started with 0.
2477 The documentation was vague. These values are now always treated as
2478 decimal, and I will make that clear in the documentation.
2484 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2485 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2488 MH/01 Don't check that the operands of numeric comparison operators are
2489 integers when their expansion is in "skipping" mode (fixes bug
2490 introduced by 4.64-PH/07).
2492 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2493 child addresses, Exim now panics and dies. Previously, because the count
2494 is held in a short int, deliveries were likely to be lost. As such a
2495 large number of recipients for a single message is ridiculous
2496 (performance will be very, very poor), I have chosen to impose a limit
2497 rather than extend the field.
2503 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2504 leftover -K file (the existence of which was triggered by #402).
2505 While we were at it, introduced process PID as part of the -K
2506 filename. This should rule out race conditions when creating
2509 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2510 processing considerably. Previous code took too long for large mails,
2511 triggering a timeout which in turn triggers #401.
2513 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2514 in the DK code in transports.c. sendfile() is not really portable,
2515 hence the _LINUX specificness.
2517 TF/01 In the add_headers option to the mail command in an Exim filter,
2518 there was a bug that Exim would claim a syntax error in any
2519 header after the first one which had an odd number of characters
2522 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2523 callout verification, Exim cached a "reject" for the entire domain. This
2524 is correct for most verifications, but it is not correct for a recipient
2525 verification with use_sender or use_postmaster set, because in that case
2526 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2527 case of MAIL FROM:<> rejection from other early rejections (e.g.
2528 rejection of HELO). When verifying a recipient using a non-null MAIL
2529 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2530 Whatever the result of the callout, the value of the domain cache is
2531 left unchanged (for any other kind of callout, getting as far as trying
2532 RCPT means that the domain itself is ok).
2534 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2537 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2538 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2539 session, the daemon ignores ECONNECT errors and logs others; it now
2540 ignores EPIPE as well.
2542 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2543 (quoted-printable decoding).
2545 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2546 later the small subsequent patch to fix an introduced bug.
2548 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2550 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2552 PH/08 An error is now given if message_size_limit is specified negative.
2554 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2555 to be given (somewhat) arbitrary names.
2557 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2560 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2561 miscellaneous code fixes
2563 PH/10 Added the log_reject_target ACL modifier to specify where to log
2566 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2567 hostname. This is wrong, because it relates to the incoming message (and
2568 probably the interface on which it is arriving) and not to the outgoing
2569 callout (which could be using a different interface). This has been
2570 changed to use the value of the helo_data option from the smtp transport
2571 instead - this is what is used when a message is actually being sent. If
2572 there is no remote transport (possible with a router that sets up host
2573 addresses), $smtp_active_hostname is used.
2575 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2576 tweaks were necessary in order to get it to work (see also 21 below):
2577 (a) The code assumed that strncpy() returns a negative number on buffer
2578 overflow, which isn't the case. Replaced with Exim's string_format()
2580 (b) There were several signed/unsigned issues. I just did the minimum
2581 hacking in of casts. There is scope for a larger refactoring.
2582 (c) The code used strcasecmp() which is not a standard C function.
2583 Replaced with Exim's strcmpic() function.
2584 (d) The code set only $1; it now sets $auth1 as well.
2585 (e) A simple test gave the error "authentication client didn't specify
2586 service in request". It would seem that Dovecot has changed its
2587 interface. Fortunately there's a specification; I followed it and
2588 changed what the client sends and it appears to be working now.
2590 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2593 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2594 address A is aliased to B and C, where B exists and C does not. Without
2595 -v the output is "A verified" because verification stops after a
2596 successful redirection if more than one address is generated. However,
2597 with -v the child addresses are also verified. Exim was outputting "A
2598 failed to verify" and then showing the successful verification for C,
2599 with its parentage. It now outputs "B failed to verify", showing B's
2600 parentage before showing the successful verification of C.
2602 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2603 look up a TXT record in a specific list after matching in a combined
2606 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2607 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2608 they consult the DNS. I had assumed they would set it the way they
2609 wanted; and indeed my experiments on Linux seem to show that in some
2610 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2611 To be on the safe side, however, I have now made the interface to
2612 host_find_byname() similar to host_find_bydns(), with an argument
2613 containing the DNS resolver options. The host_find_byname() function now
2614 sets these options at its start, just as host_find_bydns() does. The smtp
2615 transport options dns_qualify_single and dns_search_parents are passed to
2616 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2617 of host_find_byname() use the default settings of RES_DEFNAMES
2618 (qualify_single) but not RES_DNSRCH (search_parents).
2620 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2621 spool_read_header() do less string testing, by means of a preliminary
2622 switch on the second character of optional "-foo" lines. (This is
2623 overdue, caused by the large number of possibilities that now exist.
2624 Originally there were few.) While I was there, I also converted the
2625 str(n)cmp tests so they don't re-test the leading "-" and the first
2626 character, in the hope this might squeeze out yet more improvement.
2628 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2629 flag allowing group syntax was set by the header_syntax check but not
2630 turned off, possible causing trouble later; (2) The flag was not being
2631 set at all for the header_verify test, causing "group"-style headers to
2632 be rejected. I have now set it in this case, and also caused header_
2633 verify to ignore an empty address taken from a group. While doing this, I
2634 came across some other cases where the code for allowing group syntax
2635 while scanning a header line wasn't quite right (mostly, not resetting
2636 the flag correctly in the right place). These bugs could have caused
2637 trouble for malformed header lines. I hope it is now all correct.
2639 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2640 with the "reply" argument non-NULL. The code, however (which originally
2641 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2642 but it didn't always do it. This confused somebody who was copying the
2643 code for some other use. I have removed all the tests.
2645 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2646 feature that was used to support insecure browsers during the U.S. crypto
2647 embargo. It requires special client support, and Exim is probably the
2648 only MTA that supported it -- and would never use it because real RSA is
2649 always available. This code has been removed, because it had the bad
2650 effect of slowing Exim down by computing (never used) parameters for the
2651 RSA_EXPORT functionality.
2653 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2654 authenticator to fail if there's a tab character in the incoming data
2655 (there should never be unless someone is messing about, as it's supposed
2656 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2657 if the connection is using TLS or if the remote IP is the same as the
2658 local IP, and the "valid-client-cert option" if a client certificate has
2661 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2662 authenticators. This can be used for authorization after authentication
2663 succeeds. (In the case of plaintext, it servers for both authentication
2666 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2667 if any retry times were supplied.
2669 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2670 connection, where there is no client IP address to check. In this
2671 situation, the verify now always succeeds.
2673 PH/25 Applied John Jetmore's -Mset patch.
2675 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2677 PH/27 In a string expansion for a processed (not raw) header when multiple
2678 headers of the same name were present, leading whitespace was being
2679 removed from all of them, but trailing whitespace was being removed only
2680 from the last one. Now trailing whitespace is removed from each header
2681 before concatenation. Completely empty headers in a concatenation (as
2682 before) are ignored.
2684 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2685 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2687 PH/29 [Removed. This was a change that I later backed out, and forgot to
2688 correct the ChangeLog entry (that I had efficiently created) before
2689 committing the later change.]
2691 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2692 address errors (4xx response to RCPT) over the same connection as other
2693 messages routed to the same hosts. Such deliveries are always "forced",
2694 so retry times are not inspected. This resulted in far too many retries
2695 for the affected addresses. The effect occurred only when there were more
2696 hosts than the hosts_max_try setting in the smtp transport when it had
2697 the 4xx errors. Those hosts that it had tried were not added to the list
2698 of hosts for which the message was waiting, so if all were tried, there
2699 was no problem. Two fixes have been applied:
2701 (i) If there are any address or message errors in an SMTP delivery, none
2702 of the hosts (tried or untried) are now added to the list of hosts
2703 for which the message is waiting, so the message should not be a
2704 candidate for sending over the same connection that was used for a
2705 successful delivery of some other message. This seems entirely
2706 reasonable: after all the message is NOT "waiting for some host".
2707 This is so "obvious" that I'm not sure why it wasn't done
2708 previously. Hope I haven't missed anything, but it can't do any
2709 harm, as the worst effect is to miss an optimization.
2711 (ii) If, despite (i), such a delivery is accidentally attempted, the
2712 routing retry time is respected, so at least it doesn't keep
2713 hammering the server.
2715 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2716 in ${readsocket because some servers need this prod.
2718 PH/32 Added some extra debug output when updating a wait-xxx database.
2720 PH/33 The hint "could be header name not terminated by colon", which has been
2721 given for certain expansion errors for a long time, was not being given
2722 for the ${if def:h_colon_omitted{... case.
2724 PH/34 The spec says: "With one important exception, whenever a domain list is
2725 being scanned, $domain contains the subject domain." There was at least
2726 one case where this was not true.
2728 PH/35 The error "getsockname() failed: connection reset by peer" was being
2729 written to the panic log as well as the main log, but it isn't really
2730 panic-worthy as it just means the connection died rather early on. I have
2731 removed the panic log writing for the ECONNRESET error when getsockname()
2734 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2735 runs only) independently of the message's sender address. This meant
2736 that, if the 4xx error was in fact related to the sender, a different
2737 message to the same recipient with a different sender could confuse
2738 things. In particular, this can happen when sending to a greylisting
2739 server, but other circumstances could also provoke similar problems.
2740 I have changed the default so that the retry time for these errors is now
2741 based a combination of the sender and recipient addresses. This change
2742 can be overridden by setting address_retry_include_sender=false in the
2745 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2746 remote server are returned as part of bounce messages. This was not
2747 happening for LMTP over a pipe (the lmtp transport), but now it is the
2748 same for both kinds of LMTP.
2750 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2751 in header lines that were in fact CNAMEs. This is no longer the case.
2753 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2754 and queue runs started by the daemon processed all messages. This has
2755 been fixed so that -R and -S can now usefully be given with -q<time>.
2757 PH/40 Import PCRE release 6.7 (fixes some bugs).
2759 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2761 PH/42 Give an error if -q is specified more than once.
2763 PH/43 Renamed the variables $interface_address and $interface_port as
2764 $received_ip_address and $received_port, to make it clear that these
2765 values apply to message reception, and not to the outgoing interface when
2766 a message is delivered. (The old names remain recognized, of course.)
2768 PH/44 There was no timeout on the connect() call when using a Unix domain
2769 socket in the ${readsocket expansion. There now is.
2771 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2772 be meaningful with "accept".
2774 SC/01 Eximstats V1.43
2775 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2777 SC/02 Eximstats V1.44
2778 Use a glob alias rather than an array ref in the generated
2779 parser. This improves both readability and performance.
2781 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2782 Collect SpamAssassin and rejection statistics.
2783 Don't display local sender or destination tables unless
2784 there is data to show.
2785 Added average volumes into the top table text output.
2787 SC/04 Eximstats V1.46
2788 Collect data on the number of addresses (recipients)
2789 as well as the number of messages.
2791 SC/05 Eximstats V1.47
2792 Added 'Message too big' to the list of mail rejection
2793 reasons (thanks to Marco Gaiarin).
2795 SC/06 Eximstats V1.48
2796 Mainlog lines which have GMT offsets and are too short to
2797 have a flag are now skipped.
2799 SC/07 Eximstats V1.49 (Alain Williams)
2800 Added the -emptyok flag.
2802 SC/08 Eximstats V1.50
2803 Fixes for obtaining the IP address from reject messages.
2805 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2806 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2807 whitespace changes from 4.64-PH/27
2809 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2812 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2813 are found, allow negative numbers in numeric criteria)
2815 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2817 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2818 to match changes made in 4.64-PH/43
2820 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2822 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2823 "domain" is actually an IP address, and force a failure. This locks out
2824 those revolvers/nameservers that support "A-for-A" lookups, in
2825 contravention of the specifications.
2827 PH/48 When a host name was looked up from an IP address, and the subsequent
2828 forward lookup of the name timed out, the host name was left in
2829 $sender_host_name, contrary to the specification.
2831 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2832 restricted to single-key lookups, Exim was not diagnosing an error if
2833 * or *@ was used with a query-style lookup.
2835 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2837 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2838 long ago, but noone interested enough thought of it. Let's just say that
2839 the "1.1" means that there are some new functions that weren't there at
2840 some point in the past.
2842 PH/51 Error processing for expansion failure of helo_data from an smtp
2843 transport during callout processing was broken.
2845 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2846 tested/used via the -bh/-bhc/-bs options.
2848 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2849 bug, fixed in subsequent PCRE releases).
2851 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2852 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2854 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2860 SC/01 Use a glob alias rather than an array ref in eximstats generated
2861 parser. This improves both readability and performance.
2863 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2864 Don't display local sender or destination tables in eximstats unless
2865 there is data to show.
2866 Added average volumes into the eximstats top table text output.
2868 SC/03 Collect data on the number of addresses (recipients) as well
2869 as the number of messages in eximstats.
2871 TF/01 Correct an error in the documentation for the redirect router. Exim
2872 does (usually) call initgroups() when daemonizing.
2874 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2875 with consistent privilege compared to when running as a daemon.
2877 TF/03 Note in the spec that $authenticated_id is not set for local
2878 submissions from trusted users.
2880 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2881 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2883 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2884 by adding some example configuration directives to the default
2885 configuration file. A little bit of work is required to uncomment the
2886 directives and define how usernames and passwords are checked, but
2887 there is now a framework to start from.
2889 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2890 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2891 without this. I don't know how relevant this is to other LDAP libraries.
2893 PH/02 Add the verb name to the "unknown ACL verb" error.
2895 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2897 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2899 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2900 directory not expanded when it should be if an expanded home directory
2901 was set for the address (which is overridden by the transport).
2903 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2906 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2907 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2908 because it is too late at that time, and has no effect.
2910 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2911 security issue with \' (bugzilla #107). I could not use the
2912 PQescapeStringConn() function, because it needs a PGconn value as one of
2915 PH/08 When testing addresses using -bt, indicate those final addresses that
2916 are duplicates that would not cause an additional delivery. At least one
2917 person was confused, thinking that -bt output corresponded to deliveries.
2918 (Suppressing duplicates isn't a good idea as you lose the information
2919 about possibly different redirections that led to the duplicates.)
2921 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2922 systems where poll() doesn't work, in particular OS X.
2924 PH/10 Added more information to debugging output for retry time not reached.
2926 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2927 operations in malware.c.
2929 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2932 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2933 syslog_duplication set false, log lines that would normally be written
2934 both the the main log and to the reject log were not written to syslog at
2937 PH/14 In the default configuration, change the use of "message" in ACL warn
2938 statements to "add_header".
2940 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2941 not followed by a command (e.g. "seen endif").
2943 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2944 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2947 PH/17 Added extra conditions to the default value of delay_warning_condition
2951 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2952 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2953 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2956 The Auto-Submitted: and various List- headers are standardised, whereas I
2957 don't think Precedence: ever was.
2959 PH/18 Refactored debugging code in route_finduser() to show more information,
2960 in particular, the error code if getpwnam() issues one.
2962 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2963 This is apparently needed in addition to the PH/07 change above to avoid
2964 any possible encoding problems.
2966 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2967 but not after initializing Perl.
2969 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2970 output them only if debugging. By default they are written stderr,
2971 apparently, which is not desirable.
2973 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2976 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2979 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2981 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2982 authenticated or an ident call has been made. Suppress the default
2983 values for $authenticated_id and $authenticated_sender (but permit -oMai
2984 and -oMas) when testing with -bh.
2986 PH/24 Re-jigged the order of the tests in the default configuration so that the
2987 tests for valid domains and recipients precede the DNS black list and CSA
2988 tests, on the grounds that those ones are more expensive.
2990 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2991 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2992 This bug exists in every version of Exim that I still have, right back to
2995 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2996 However, an attempt to turn on case-sensitivity in a regex key by
2997 including (?-i) didn't work because the subject string was already
2998 lowercased, and the effects were non-intuitive. It turns out that a
2999 one-line patch can be used to allow (?-i) to work as expected.
3005 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3006 other effects) broke the use of negated acl sub-conditions.
3008 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3011 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3012 "Deny" causes Exim to reject the incoming connection with a 554 error.
3013 Unfortunately, if there is a major crisis, such as a disk failure,
3014 tcp-wrappers gives "deny", whereas what one would like would be some
3015 kind of temporary error. A kludge has been added to help with this.
3016 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3017 554 error is used if errno is still zero or contains ENOENT (which occurs
3018 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3021 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3023 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3024 errors. Otherwise a message that provokes a temporary error (when other
3025 messages do not) can cause a whole host to time out.
3027 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3028 addresses were routed directly to files or pipes from a redirect router.
3029 File deliveries just didn't batch; pipe deliveries might have suffered
3032 PH/06 A failure to get a lock for a hints database would erroneously always say
3033 "Failed to get write lock", even when it was really a read lock.
3035 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3036 of 0600. This has been changed to use the value of the lockfile_mode
3037 option (which defaults to 0600).
3039 PH/08 Applied small patch from the Sieve maintainer.
3041 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3042 folder from quota calculations, a direct delivery into this folder messed
3043 up the contents of the maildirsize file. This was because the regex was
3044 used only to exclude .Trash (or whatever) when the size of the mailbox
3045 was calculated. There was no check that a delivery was happening into an
3046 excluded directory. This bug has been fixed by ignoring all quota
3047 processing for deliveries into excluded directories.
3049 PH/10 Added the maildirfolder_create_regex option to appendfile.
3055 PH/01 The code for finding all the local interface addresses on a FreeBSD
3056 system running IPv6 was broken. This may well have applied to all BSD
3057 systems, as well as to others that have similar system calls. The broken
3058 code found IPv4 interfaces correctly, but gave incorrect values for the
3059 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3060 that it would not match correctly against @[] and not recognize the IPv6
3063 PH/02 The ipliteral router was not recognizing addresses of the form user@
3064 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3066 PH/03 Added disable_ipv6.
3068 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3069 decoded form, because it is most often used to construct To: headers
3070 lines in autoreplies, and the decoded form may well be syntactically
3071 invalid. However, $reply_address has leading white space removed, and all
3072 newlines turned into spaces so that the autoreply transport does not
3075 PH/05 If group was specified without a user on a router, and no group or user
3076 was specified on a transport, the group from the router was ignored.
3078 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3079 for visible compile-time settings that can be used to change these
3080 numbers, for those that want even more. Backwards compatibility with old
3081 spool files has been maintained. However, going back to a previous Exim
3082 release will lost any variables that are in spool files.
3084 PH/07 Two small changes when running in the test harness: increase delay when
3085 passing a TCP/IP connection to a new process, in case the original
3086 process has to generate a bounce, and remove special handling of
3087 127.0.0.2 (sic), which is no longer necessary.
3089 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3090 be the same on different OS.
3092 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3095 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3096 whether --show-vars was specified or not
3098 JJ/02 exipick: Added support for new ACL variable spool format introduced
3101 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3102 syntactically invalid From: or Reply-to: line, and a filter used this to
3103 generate an autoreply, and therefore failed to obtain an address for the
3104 autoreply, Exim could try to deliver to a non-existent relative file
3105 name, causing unrelated and misleading errors. What now happens is that
3106 it logs this as a hard delivery error, but does not attempt to create a
3109 PH/11 The exinext utility has a -C option for testing purposes, but although
3110 the given file was scanned by exinext itself; it wasn't being passed on
3111 when Exim was called.
3113 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3114 an end-of-file indication when reading a command response.
3116 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3117 compiled. In many other places in Exim, IPv6 addresses are always
3118 recognized, so I have changed this. It also means that IPv4 domain
3119 literals of the form [IPV4:n.n.n.n] are now always recognized.
3121 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3122 used if the router is not running as root, for example, when verifying at
3123 ACL time, or when using -bh. The debugging output from this situation was
3124 non-existent - all you got was a failure to exec. I have made two
3127 (a) Failures to set uid/gid, the current directory, or a process leader
3128 in a subprocess such as that created by queryprogram now generate
3129 suitable debugging output when -d is set.
3131 (b) The queryprogram router detects when it is not running as root,
3132 outputs suitable debugging information if -d is set, and then runs
3133 the subprocess without attempting to change uid/gid.
3135 PH/15 Minor change to Makefile for building test_host (undocumented testing
3138 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3139 additional section of a DNS packet that returns MX or SRV records.
3140 Instead, it always explicitly searches for A/AAAA records. This avoids
3141 major problems that occur when a DNS server includes only records of one
3142 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3143 fixed another bug: if SRV records were looked up and the corresponding
3144 address records were *not* found in the additional section, the port
3145 values from the SRV records were lost.
3147 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3148 using the correct key (the original address) when searching the retry
3149 rules in order to find which one to use for generating the retry hint.
3151 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3152 adding the default one. Similarly, if it contains a Reply-To: header, the
3153 errors_reply_to option, if set, is not used.
3155 PH/19 When calculating a retry time, Exim used to measure the "time since
3156 failure" by looking at the "first failed" field in the retry record. Now
3157 it does not use this if it is later than than the arrival time of the
3158 message. Instead it uses the arrival time. This makes for better
3159 behaviour in cases where some deliveries succeed, thus re-setting the
3160 "first failed" field. An example is a quota failure for a huge message
3161 when small messages continue to be delivered. Without this change, the
3162 "time since failure" will always be short, possible causing more frequent
3163 delivery attempts for the huge message than are intended.
3164 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3166 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3167 $1, $2, $3) because the numerical variables can be reset during some
3168 expansion items (e.g. "match"), thereby losing the authentication data.
3170 PH/21 Make -bV show the size of off_t variables so that the test suite can
3171 decide whether to run tests for quotas > 2G.
3173 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3174 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3175 filecount value is greater than 2G or if a quota value is greater than 2G
3176 on a system where the size of off_t is not greater than 4, a panic error
3179 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3180 never match. The debug and -bh output now contains an explicit error
3181 message indicating a malformed IPv4 address or mask.
3183 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3184 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3185 PH/23 above applies.
3187 PH/25 Do not write to syslog when running in the test harness. The only
3188 occasion when this arises is a failure to open the main or panic logs
3189 (for which there is an explicit test).
3191 PH/26 Added the /no_tell option to "control=freeze".
3193 PH/27 If a host name lookup failed very early in a connection, for example, if
3194 the IP address matched host_lookup and the reverse lookup yielded a name
3195 that did not have a forward lookup, an error message of the form "no IP
3196 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3197 could be logged. Now it outputs the IP address instead of "NULL".
3199 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3200 allows the sender and the authenticated sender to be set when
3201 submitting a message from within Exim. Since child_open_exim() is
3202 documented for local_scan(), the new function should be too.
3204 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3205 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3206 results in an empty string is now treated as unset.
3208 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3210 PH/31 Added sender_verify_fail logging option.
3212 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3213 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3214 tidied the source and removed it altogether.
3216 PH/33 When a queue run was abandoned because the load average was too high, a
3217 log line was always written; now it is written only if the queue_run log
3218 selector is set. In addition, the log line for abandonment now contains
3219 information about the queue run such as the pid. This is always present
3220 in "start" and "stop" lines but was omitted from the "abandon" line.
3222 PH/34 Omit spaces between a header name and the colon in the error message that
3223 is given when verify = headers_syntax fails (if there are lots of them,
3224 the message gets confusing).
3226 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3227 names, as there are now some PTR records that contain slashes. This check
3228 is only to protect against broken name servers that fall over on strange
3229 characters, so the fact that it applies to all lookups doesn't matter.
3231 PH/36 Now that the new test suite is complete, we can remove some of the
3232 special code in Exim that was needed for the old test suite. For example,
3233 sorting DNS records because real resolvers return them in an arbitrary
3234 order. The new test suite's fake resolver always returns records in the
3237 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3238 bounces) except when queue_only is set, to avoid logging races between
3239 the different processes.
3241 PH/38 Panic-die if .include specifies a non-absolute path.
3243 PH/39 A tweak to the "H" retry rule from its user.
3245 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3246 a label. They prevented compilation on older perls.
3248 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3249 a warning to be raised on newish perls.
3251 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3252 on queue. Changes to match documented behaviour of showing count of
3253 messages matching specified criteria.
3255 PH/40 Changed the default ident timeout from 30s to 5s.
3257 PH/41 Added support for the use of login_cap features, on those BSD systems
3258 that have them, for controlling the resources used by pipe deliveries.
3260 PH/42 The content-scanning code uses fopen() to create files in which to put
3261 message data. Previously it was not paying any attention to the mode of
3262 the files. Exim runs with umask(0) because the rest of the code creates
3263 files with open(), and sets the required mode explicitly. Thus, these
3264 files were ending up world-writeable. This was not a big issue, because,
3265 being within the spool directory, they were not world-accessible. I have
3266 created a function called modefopen, which takes an additional mode
3267 argument. It sets umask(777), creates the file, chmods it to the required
3268 mode, then resets the umask. All the relevant calls to fopen() in the
3269 content scanning code have been changed to use this function.
3271 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3272 to 24 hours. This avoids potential overflow problems when processing G
3273 and H retry rules. I suspect nobody ever tinkers with this value.
3275 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3277 PH/45 When the plaintext authenticator is running as a client, the server's
3278 challenges are checked to ensure they are valid base64 strings. By
3279 default, the authentication attempt is cancelled if an invalid string is
3280 received. Setting client_ignore_invalid_base64 true ignores these errors.
3281 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3282 they are received. Thus, the responses can be made to depend on the
3283 challenges. If an invalid string is ignored, an empty string is placed in
3286 PH/46 Messages that are created by the autoreply transport now contains a
3287 References: header, in accordance with RFCs 2822 and 3834.
3289 PH/47 Added authenticated_sender_force to the smtp transport.
3291 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3293 PH/49 Installed latest patch from the Sieve maintainer.
3295 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3296 was also set, the appendfile transport was unnecessarily scanning a
3297 directory of message files (e.g. for maildir delivery) to find the count
3298 of files (along with the size), even though it did not need this
3299 information. It now does the scan only if it needs to find either the
3300 size of the count of files.
3302 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3304 PH/52 Two bugs concerned with error handling when the smtp transport is
3307 (i) Exim was not creating retry information for temporary errors given
3308 for individual recipients after the DATA command when the smtp transport
3309 was used in LMTP mode. This meant that they could be retried too
3310 frequently, and not timed out correctly.
3312 (ii) Exim was setting the flag that allows error details to be returned
3313 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3314 recipients that were returned after the DATA command.
3316 PH/53 This is related to PH/52, but is more general: for any failing address,
3317 when detailed error information was permitted to be returned to the
3318 sender, but the error was temporary, then after the final timeout, only
3319 "retry timeout exceeded" was returned. Now it returns the full error as
3320 well as "retry timeout exceeded".
3322 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3323 do this, and (what is worse) MTAs that accept it.
3325 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3326 will now be deprecated.
3328 PH/56 New os.c-cygwin from the Cygwin maintainer.
3330 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3331 formats (previously only available in exim formats via -bpr, -bpru,
3332 and -bpra. Now also available in native and exiqgrep formats)
3334 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3335 with very large, slow to parse queues
3337 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3339 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3341 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3342 responses to authentication challenges, though it was showing the
3343 challenges; (ii) I've removed the CR characters from the debug output for
3346 PH/58 Allow for the insertion of a newline as well as a space when a string
3347 is turned into more than one encoded-word during RFC 2047 encoding. The
3348 Sieve code now uses this.
3350 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3351 data_4xx, lost_connection, tls_required.
3353 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3354 message was being sent as an SMTP response.
3356 PH/61 Add -l and -k options to exicyclog.
3358 PH/62 When verifying, if an address was redirected to one new address, so that
3359 verification continued, and the new address failed or deferred after
3360 having set something in $address_data, the value of $address_data was not
3361 passed back to the ACL. This was different to the case when no
3362 redirection occurred. The value is now passed back in both cases.
3364 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3365 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3366 use login_cap.h, so on its own it isn't the distinguishing feature. The
3367 new name refers directly to the setclassresources() function.
3369 PH/65 Added configuration files for NetBSD3.
3371 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3373 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3374 is preferred over IPv4.
3376 PH/68 The bounce_return_message and bounce_return_body options were not being
3377 honoured for bounces generated during the reception of non-SMTP messages.
3378 In particular, this applied to messages rejected by the ACL. This bug has
3379 been fixed. However, if bounce_return_message is true and bounce_return_
3380 body is false, the headers that are returned for a non-SMTP message
3381 include only those that have been read before the error was detected.
3382 (In the case of an ACL rejection, they have all been read.)
3384 PH/69 The HTML version of the specification is now built in a directory called
3385 spec_html instead of spec.html, because the latter looks like a path with
3386 a MIME-type, and this confuses some software.
3388 PH/70 Catch two compiler warnings in sieve.c.
3390 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3391 function verify_get_ident() calls ip_connect() to connect a socket, but
3392 if the "connect()" function timed out, ip_connect() used to close the
3393 socket. However, verify_get_ident() also closes the socket later, and in
3394 between Exim writes to the log, which may get opened at this point. When
3395 the socket was closed in ip_connect(), the log could get the same file
3396 descriptor number as the socket. This naturally causes chaos. The fix is
3397 not to close the socket in ip_connect(); the socket should be closed by
3398 the function that creates it. There was only one place in the code where
3399 this was missing, in the iplookup router, which I don't think anybody now
3400 uses, but I've fixed it anyway.
3402 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3403 well as to direct DNS lookups. Otherwise the handling of names in host
3404 lists is inconsistent and therefore confusing.
3410 PH/01 Two changes to the default runtime configuration:
3412 (1) Move the checks for relay_from_hosts and authenticated clients from
3413 after to before the (commented out) DNS black list checks.
3415 (2) Add control=submission to the relay_from_hosts and authenticated
3416 clients checks, on the grounds that messages accepted by these
3417 statements are most likely to be submissions.
3419 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3421 (1) Generate an error if the third argument for the ${prvs expansion is
3424 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3427 (3) Reset the variables that are obtained from the first argument of
3428 ${prvscheck and used in the second argument before leaving the code,
3429 because their memory is reclaimed, so using them afterwards may do
3432 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3433 one (it's much easier than Tom thought :-).
3435 (5) Because of (4), we can now allow for the use of $prvscheck_result
3436 inside the third argument.
3438 PH/03 For some reason, the default setting of PATH when running a command from
3439 a pipe transport was just "/usr/bin". I have changed it to
3442 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3443 anything to be listed in the output from -bV.
3445 PH/05 When a filter generated an autoreply, the entire To: header line was
3446 quoted in the delivery log line, like this:
3448 => >A.N.Other <ano@some.domain> <original@ddress> ...
3450 This has been changed so that it extracts the operative address. There
3451 may be more than one such address. If so, they are comma-separated, like
3454 => >ano@some.domain,ona@other.domain <original@ddress> ...
3456 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3457 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3458 not being looked up in the rDNS to get a host name, Exim was showing the
3459 IP address twice in Received: lines, even though the IP addresses were
3460 identical. For example:
3462 Received: from [1.2.3.4] (helo=[1.2.3.4])
3464 However, if the real host name was known, it was omitting the HELO data
3465 if it matched the actual IP address. This has been tidied up so that it
3466 doesn't show the same IP address twice.
3468 PH/07 When both +timestamp and +memory debugging was on, the value given by
3469 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3470 called by the debug printing, thereby overwriting the timestamp buffer.
3471 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3473 PH/08 When the original message was included in an autoreply transport, it
3474 always said "this is a copy of the message, including all the headers",
3475 even if body_only or headers_only was set. It now gives an appropriate
3478 PH/09 Applied a patch from the Sieve maintainer which:
3480 o fixes some comments
3481 o adds the (disabled) notify extension core
3482 o adds some debug output for the result of if/elsif tests
3483 o points to the current vacation draft in the documentation
3484 and documents the missing references header update
3488 o fixes a bug in processing the envelope test (when testing
3489 multiple envelope elements, the last element determined the
3492 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3493 Electronic Mail") by including:
3495 Auto-submitted: auto-generated
3497 in the messages that it generates (bounce messages and others, such as
3498 warnings). In the case of bounce messages for non-SMTP messages, there was
3499 also a typo: it was using "Auto_submitted" (underscore instead of
3500 hyphen). Since every message generated by Exim is necessarily in response
3501 to another message, thes have all been changed to:
3503 Auto-Submitted: auto-replied
3505 in accordance with these statements in the RFC:
3507 The auto-replied keyword:
3509 - SHOULD be used on messages sent in direct response to another
3510 message by an automatic process,
3512 - MUST NOT be used on manually-generated messages,
3514 - MAY be used on Delivery Status Notifications (DSNs) and Message
3515 Disposition Notifications (MDNs),
3517 - MUST NOT be used on messages generated by automatic or periodic
3518 processes, except for messages which are automatic responses to
3521 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3522 to the default Received: header definition.
3524 PH/12 Added log selector acl_warn_skipped (default on).
3526 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3527 variables because (a) they are in the wrong storage pool and (b) even if
3528 they were copied, it wouldn't work properly because of the caching.
3530 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3531 checking when decoding. Apparently there are clients that generate
3532 overlong encoded strings. Why am I not surprised?
3534 PH/15 If the first argument of "${if match_address" was not empty, but did not
3535 contain an "@" character, Exim crashed. Now it writes a panic log message
3536 and treats the condition as false.
3538 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3540 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3541 extension "envelope-auth". The code is finished and in agreement with
3542 other implementations, but there is no documentation so far and in fact,
3543 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3544 not changing the active code.
3546 Print executed "if" and "elsif" statements when debugging is used. This
3547 helps a great deal to understand what a filter does.
3549 Document more things not specified clearly in RFC3028. I had all this
3550 sorted out, when out of a sudden new issues came to my mind. Oops."
3552 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3555 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3556 canonical form (as documented). However, after a host name lookup from
3557 the IP address, check_host() was doing a simple string comparison with
3558 addresses acquired from the DNS when checking that the found name did
3559 have the original IP as one of its addresses. Since any found IPv6
3560 addresses are likely to be in abbreviated form, the comparison could
3561 fail. Luckily, there already exists a function for doing the comparison
3562 by converting both addresses to binary, so now that is used instead of
3563 the text comparison.
3565 PH/20 There was another similar case to PH/19, when a complete host name was
3566 given in a host list; looking up its IP address could give an abbreviated
3567 form, whereas the current host's name might or might not be abbreviated.
3568 The same fix has been applied.
3574 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3575 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3578 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3579 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3581 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3583 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3584 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3585 Also, for "Auto-Submitted", treat anything other than "no" as
3586 non-personal, in accordance with RFC 3834. (Previously it treated
3587 anything starting "auto-" as non-personal.)
3589 TF/01 The control=submission/name=... option had a problem with syntax
3590 errors if the name included a slash character. The /name= option
3591 now slurps the rest of the string, so it can include any characters
3592 but it must come last in the list of options (after /sender_retain
3595 PH/05 Some modifications to the interface to the fake nameserver for the new
3603 TK/01 Added the "success_on_redirect" address verification option. See
3604 NewStuff for rationale and an example.
3606 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3608 PH/02 Patch to exigrep to allow it to work on syslog lines.
3610 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3611 fread() to skip over the body file's header line, because in Cygwin the
3612 header line is locked and is inaccessible.
3614 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3615 co-exist for some time) to make it clear that it is the Exim ID that is
3616 referenced, not the Message-ID: header line.
3618 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3619 string_format() function, because snprintf() does not exist on all
3622 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3623 ${stat: expansion item.
3625 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3626 protocol synchronization error", to keep the pedants happy.
3628 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3629 well as for IRIX systems, when gcc is being used. See the host.c source
3632 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3634 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3637 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3638 transport and to the smtp transport in LMTP mode.
3640 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3642 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3643 run a filter in a subprocess. This could lead to confusion in subsequent
3644 lookups in the parent process. There should also be a search_tidyup() at
3645 the end of the subprocess.
3647 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3648 only if the host matched helo_try_verify_hosts, which caused the
3649 verification to occur when the EHLO/HELO command was issued. The ACL just
3650 tested the remembered result. Now, if a previous verification attempt has
3651 not happened, "verify = helo" does it there and then.
3653 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3655 TK/03 Fix log output including CR from clamd.
3657 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3658 exist provoked a memory error which could cause a segfault.
3660 PH/15 Installed PCRE 6.2
3662 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3664 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3665 of the problem. Specifically, suggested +O2 rather than +O1 for the
3668 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3670 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3671 record, the port was not correctly logged when the outgoing_port log
3672 selector was set (it logged the transort's default port).
3674 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3675 fallback_hosts, and "hosts" in the smtp transport.
3677 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3678 host errors such as "Connection refused".
3680 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3681 authentication with radiusclient 0.4.9:
3683 - Error returned from rc_read_config was caught wrongly
3684 - Username/password not passed on to radius server due to wrong length.
3686 The presumption is that some radiusclient API changes for 4.51/PH/17
3687 were not taken care of correctly. The code is still untested by me (my
3688 Linux distribution still has 0.3.2 of radiusclient), but it was
3689 contributed by a Radius user.
3691 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3692 expanding the "port" option of the smtp transport.
3694 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3695 while reading a MIME header. Thanks to Tom Hughes for a patch.
3697 PH/24 Include config.h inside local_scan.h so that configuration settings are
3700 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3701 that in an ACL for RCPT (for example), you can examine exactly what was
3704 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3705 commands, but it was not correctly comparing the address with the actual
3706 client host address. Thus, it would show the EHLO address in Received:
3707 header lines when this was not necessary.
3709 PH/27 Added the % operator to ${eval:}.
3711 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3712 it should be ignoring failures (because with -C, for example, it has lost
3713 privilege). It wasn't ignoring creation failures other than "already
3716 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3719 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3720 because an input line was too long, either on its own, or by virtue of
3721 too many continuations, the temporary file was not being removed, and the
3722 return code was incorrect.
3724 PH/31 Missing "BOOL" in function definition in filtertest.c.
3726 PH/32 Applied Sieve patches from the maintainer.
3728 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3730 PH/33 Added "verify = not_blind".
3732 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3733 Local/Makefile (with some defaults set). These are used in built scripts
3734 such as exicyclog, but they have never been used in the exim_install
3735 script (though there are many overriding facilities there). I have
3736 arranged that the exim_install script now takes note of these two
3739 PH/35 Installed configuration files for Dragonfly.
3741 PH/36 When a locally submitted message by a trusted user did not contain a
3742 From: header, and the sender address was obtained from -f or from an SMTP
3743 MAIL command, and the trusted user did not use -F to supply a sender
3744 name, $originator_name was incorrectly used when constructing a From:
3745 header. Furthermore, $originator_name was used for submission mode
3746 messages from external hosts without From: headers in a similar way,
3747 which is clearly wrong.
3749 PH/37 Added control=suppress_local_fixups.
3751 PH/38 When log_selector = +received_sender was set, and the addition of the
3752 sender made the log line's construction buffer exactly full, or one byte
3753 less than full, an overflow happened when the terminating "\n" was
3756 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3757 when the result of a list match is failure because a DNS lookup failed.
3759 PH/40 RM_COMMAND is now used in the building process.
3761 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3762 the "build-* directories that it finds.
3764 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3765 address in a domain literal was a prefix of an interface address.
3767 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3768 when verifying a sender address, unless rewrite_headers is false.
3770 PH/44 Wrote a long comment about why errors_to addresses are verified as
3771 recipients, not senders.
3773 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3774 the ratelimit ACL was added.
3776 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3778 PH/46 Added extra information about PostgreSQL errors to the error string.
3780 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3781 avoiding the need to install special zones in a real server. This is
3782 backwards compatible; if it can't find the fake resolver, it drops back.
3783 Thus, both old and new test suites can be run.
3785 TF/02 Added util/ratelimit.pl
3787 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3788 clock is set back in time.
3790 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3791 Candler <B.Candler@pobox.com>.
3793 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3794 OK for addresses that are the result of redirections.
3796 PH/48 A number of further additions for the benefit of the new test suite,
3797 including a fake gethostbyname() that interfaces to the fake DNS resolver
3800 TF/06 The fix for widen_domains has also been applied to qualify_single and
3801 search_parents which are the other dnslookup options that can cause
3804 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3807 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3809 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3810 service" if it encountered an MX record with an empty target hostname.
3811 The message is now "an MX or SRV record indicated no SMTP service".
3813 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3814 if the DNS of the sending site is misconfigured. This is quite a
3815 common situation. This change restores the behaviour of treating a
3816 helo verification defer as a failure.
3818 PH/51 If self=fail was set on a router, the bounce message did not include the
3819 actual error message.
3825 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3827 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3828 command itself ended in error, the underlying message about the transport
3829 filter timeout was being overwritten with the pipe command error. Now the
3830 underlying error message should be appended to the second error message.
3832 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3834 PH/02 Reduce the amount of output that "make" produces by default. Full output
3835 can still be requested.
3837 PH/03 The warning log line about a condition test deferring for a "warn" verb
3838 was being output only once per connection, rather than after each
3839 occurrence (because it was using the same function as for successful
3840 "warn" verbs). This seems wrong, so I have changed it.
3842 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3843 it should not have, which might have caused a crash in the right
3844 circumstances, but probably never did.
3846 PH/04 Installed a modified version of Tony Finch's patch to make submission
3847 mode fix the return path as well as the Sender: header line, and to
3848 add a /name= option so that you can make the user's friendly name appear
3851 TF/03 Added the control = fakedefer ACL modifier.
3853 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3854 Mark Lowes for thorough testing.
3856 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3858 TK/03 Merged latest SRS patch from Miles Wilton.
3860 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3861 with the definition in sysexits.h (which is #included earlier).
3862 Fortunately, Exim does not actually use EX_OK. The code used to try to
3863 preserve the sysexits.h value, by assuming that macro definitions were
3864 scanned for macro replacements. I have been disabused of this notion,
3865 so now the code just undefines EX_OK before #including unistd.h.
3867 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3868 in the smtp transport. When a block could not be written in a single
3869 write() function, the timeout was being re-applied to each part-write.
3870 This seems wrong - if the receiver was accepting one byte at a time it
3871 would take for ever. The timeout is now adjusted when this happens. It
3872 doesn't have to be particularly precise.
3874 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3875 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3877 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3878 without a domain if the check to <postmaster@domain> fails.
3880 SC/01 Eximstats: added -xls and the ability to specify output files
3881 (patch written by Frank Heydlauf).
3883 SC/02 Eximstats: use FileHandles for outputting results.
3885 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3887 SC/04 Eximstats: fixed display of large numbers with -nvr option
3889 SC/05 Eximstats: fixed merging of reports with empty tables.
3891 SC/06 Eximstats: added the -include_original_destination flag
3893 SC/07 Eximstats: removed tabs and trailing whitespace.
3895 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3897 TK/06 MBOX spool code: Add real "From " MBOX separator line
3898 so the .eml file is really in mbox format (even though
3899 most programs do not really care). Patch from Alex Miller.
3901 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3902 The latter is generated from $received_to and is only set if the
3903 message has one envelope recipient. SA can use these headers,
3904 obviously out-of-the-box. Patch from Alex Miller.
3906 PH/08 The ${def test on a variable was returning false if the variable's
3907 value was "0", contrary to what the specification has always said!
3908 The result should be true unless the variable is empty.
3910 PH/09 The syntax error of a character other than { following "${if
3911 def:variable_name" (after optional whitespace) was not being diagnosed.
3912 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3913 accidental colon was present, for example, could give incorrect results.
3915 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3916 result is used (not including appendfile, where other changes are about
3919 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3920 This involved changing a lot of size variables from int to off_t. It
3921 should work with maildirs and everything.
3923 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3924 spamd dying while we are connected to it.
3926 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3929 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3930 function for BDB 4.3.
3932 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3934 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3935 the sources for PCRE have been re-arranged and more files are now
3938 PH/15 The code I had for printing potentially long long variables in PH/11
3939 above was not the best (it lost precision). The length of off_t variables
3940 is now inspected at build time, and an appropriate printing format (%ld
3941 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3942 to be "long long int" or "long int". This is needed for the internal
3943 formatting function string_vformat().
3945 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3946 the configuration file to be ":syslog", then the script "guesses" where
3947 the logs files are, rather than using the compiled in default. In our
3948 case the guess is not the same as the compiled default, so the script
3949 suddenly stopped working when I started to use syslog. The patch checks
3950 to see if log_file_path is "". If so, it attempts to read it from exim
3951 with no configuration file to get the compiled in version, before it
3952 falls back to the previous guessing code."
3954 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3955 implementing BATV in an Exim configuration. See NewStuff for the gory
3958 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3959 Makefile that are specific to HP-UX.
3961 PH/18 If the "use_postmaster" option was set for a recipient callout together
3962 with the "random" option, the postmaster address was used as the MAIL
3963 FROM address for the random test, but not for the subsequent recipient
3964 test. It is now used for both.
3966 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3967 patch removes a few documentation additions to RFC 3028, because the
3968 latest draft now contains them. It adds the new en;ascii-case comparator
3969 and a new error check for 8bit text in MIME parts. Comparator and
3970 require names are now matched exactly. I enabled the subaddress
3971 extension, but it is not well tested yet (read: it works for me)."
3973 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3974 rework some of the code of TK/09 above to avoid the hardwired use of
3975 "%lld" and "long long". Replaced the call to snprintf() with a call to
3978 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3979 records point to non-existent hosts", "retry timeout exceeded", and
3980 "retry time not reached for any host after a long failure period".
3982 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3983 experimental DomainKeys support:
3985 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3986 (2) On an error such as an illegally used "control", the wrong name for
3987 the control was given.
3989 These problems did NOT occur unless DomainKeys support was compiled.
3991 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3993 PH/24 Added ${if match_ip condition.
3995 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3996 databases so that it will be absolutely obvious if a crash occurs in the
3997 DB library. This is a regular occurrence (often caused by mis-matched
4000 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4001 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4002 picked up on a user's system that detects such things. There doesn't seem
4003 to be a gcc warning option for this - only an attribute that has to be
4004 put on the function's prototype. It seems that in Fedora Core 4 they have
4005 set this on a number of new functions. No doubt there will be more in due
4008 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4009 specify a transport. However, if an address that was verified by such a
4010 router was the subject of a callout, Exim crashed because it tried to
4011 read the rcpt_include_affixes from the non-existent transport. Now it
4012 just assumes that the setting of that option is false. This bug was
4013 introduced by 4.51/PH/31.
4015 PH/28 Changed -d+all to exclude +memory, because that information is very
4016 rarely of interest, but it makes the output a lot bigger. People tend to
4017 do -d+all out of habit.
4019 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4020 code in os-type was giving problems when libc.so lives in lib64, like on
4023 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4024 aren't the modern standard, and it seems that some systems' include files
4025 don't always have them. Exim was already checking for some of the newer
4026 ones like T_AAAA, and defining it itself. I've added checks for all the
4027 record types that Exim uses.
4029 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4030 not automatically generating a new one, as it is supposed to. This
4031 prevented TLS from working. If the file did exist, but contained invalid
4032 data, a new version was generated, as expected. It was only the case of a
4033 non-existent file that was broken.
4035 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4036 with a change in libdomainkeys > 0.64.
4038 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4039 from DNS. If the selector record carries the flag, it now has
4040 precedence over the domain-wide flag.
4042 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4044 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4045 the use of a port name that isn't defined in /etc/services) occurred, the
4046 message was deferred as in a normal delivery, and thus remained on the
4047 spool, instead of being failed because of the mua_wrapper setting. This
4048 is now fixed, and I tidied up some of the mua_wrapper messages at the
4051 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4052 the messages in a hash of arrays rather than using individual hashes.
4053 This is a bit cleaner and results in dramatic memory savings, albeit
4054 at a slight CPU cost.
4056 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4057 as requested by Marc Sherman.
4059 SC/10 Eximstats: added histograms for user specified patterns as requested
4062 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4064 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4065 fopen() in the content-scanning modules that did not already have it.
4071 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4072 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4074 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4076 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4078 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4079 to test Sieve filters that use "vacation".
4081 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4082 that changes the way the GnuTLS parameters are stored in the cache file.
4083 The new format can be generated externally. For backward compatibility,
4084 if the data in the cache doesn't make sense, Exim assumes it has read an
4085 old-format file, and it generates new data and writes a new file. This
4086 means that you can't go back to an older release without removing the
4089 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4090 work if there are any delivery delays because "one_time" forces the
4091 parent to be marked "delivered", so its unseen clone is never tried
4092 again. For this reason, Exim now forbids the simultaneous setting of
4095 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4096 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4097 entry if you want to know the details. The fix, however, neglected to
4098 consider the case when local delivery batching is involved. The test for
4099 "previously delivered" was not happening when checking to see if an
4100 address could be batched with a previous (undelivered) one; under
4101 certain circumstances this could lead to multiple deliveries to the same
4104 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4105 in its include files, and this causes problems building Exim.
4107 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4108 header_syntax) but Exim was just ignoring anything given after a slash.
4109 In particular, this caused confusion with an attempt to use "verify =
4110 reverse_host_lookup/defer_ok". An error is now given when options are
4111 supplied for verify items that do not have them. (Maybe reverse_host_
4112 lookup should have a defer_ok option, but that's a different point.)
4114 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4115 defined by RFC 821) to 2048, because there were problems with some AUTH
4116 commands, and RFC 1869 says the size should be increased for extended
4117 SMTP commands that take arguments.
4119 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4122 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4123 "unknown" error; now it says that the functionality isn't in the binary.
4125 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4126 an address' error message when a string expansion fails (syntax or
4127 whatever). Otherwise the password may appear in the log. Following change
4128 PH/42 below, there is no longer a chance of it appearing in a bounce
4131 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4133 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4134 of its addresses was ever tried. (Bugzilla bug #2.)
4136 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4137 the result incorrectly in the debug output. (It correctly added a newline
4138 to what was transported.)
4140 TF/01 Added $received_time.
4142 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4143 commented out examples of how to interface to a virus scanner and to
4144 SpamAssassin. Also added commented examples of av_scanner and
4145 spamd_address settings.
4147 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4148 and controls are allowed in which ACLs. There were a couple of minor
4149 errors. Some of the entries in the conditions table (which is a table of
4150 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4151 negation of where the condition IS allowed.
4153 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4155 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4156 header file does not have a version number, so I've had to invent a new
4157 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4158 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4159 radiusclient), but it was contributed by a Radius user.
4161 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4162 files or directories, for OpenSSL.
4164 PH/19 When an Exim process that is running as root has to create an Exim log
4165 file, it does so in a subprocess that runs as exim:exim so as to get the
4166 ownership right at creation (otherwise, other Exim processes might see
4167 the file with the wrong ownership). There was no test for failure of this
4168 fork() call, which would lead to the process getting stuck as it waited
4169 for a non-existent subprocess. Forks do occasionally fail when resources
4170 run out. I reviewed all the other calls to fork(); they all seem to check
4173 PH/20 When checking for unexpected SMTP input at connect time (before writing
4174 the banner), Exim was not dealing correctly with a non-positive return
4175 from the read() function. If the client had disconnected by this time,
4176 the result was a log entry for a synchronization error with an empty
4177 string after "input=" when read() returned zero. If read() returned -1
4178 (an event I could not check), uninitialized data bytes were printed.
4179 There were reports of junk text (parts of files, etc) appearing after
4182 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4184 PH/22 Added support for macro redefinition, and (re)definition in between
4185 driver and ACL definitions.
4187 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4188 forgetting to use the resulting value; it was using the unexpanded value.
4190 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4191 hadn't been configured. The fix is from Juergen Kreileder, who
4192 understands it better than I do:
4194 "Here's what I see happening with three configured cyrus_sasl
4195 authenticators configured (plain, login, cram-md5):
4197 On startup auth_cyrus_sasl_init() gets called for each of these.
4198 This means three calls to sasl_listmech() without a specified mech_list.
4199 => SASL tests which mechs of all available mechs actually work
4200 => three warnings about OTP not working
4201 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4203 With the patch, sasl_listmech() also gets called three times. But now
4204 SASL's mech_list option is set to the server_mech specified in the the
4205 authenticator. Or in other words, the answer from sasl_listmech()
4206 gets limited to just the mech you're testing for (which is different
4208 => the return list contains just 'plain' or 'login', 'cram-md5' or
4209 nothing depending on the value of ob->server_mech.
4211 I've just tested the patch: Authentication still works fine,
4212 unavailable mechs specified in the exim configuration are still
4213 caught, and the auth.log warnings about OTP are gone."
4215 PH/25 When debugging is enabled, the contents of the command line are added
4216 to the debugging output, even when log_selector=+arguments is not
4219 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4220 answer is "GNU", and only if the return is "GNU/something" is the answer
4223 PH/27 $acl_verify_message is now set immediately after the failure of a
4224 verification in an ACL, and so is available in subsequent modifiers. In
4225 particular, the message can be preserved by coding like this:
4227 warn !verify = sender
4228 set acl_m0 = $acl_verify_message
4230 Previously, $acl_verify_message was set only while expanding "message"
4231 and "log_message" when a very denied access.
4233 PH/28 Modified OS/os.c-Linux with
4235 -#ifndef OS_LOAD_AVERAGE
4236 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4238 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4239 nomenclature these days.)
4241 PH/29 Installed patch from the Sieve maintainer that adds the options
4242 sieve_useraddress and sieve_subaddress to the redirect router.
4244 PH/30 In these circumstances:
4245 . Two addresses routed to the same list of hosts;
4246 . First host does not offer TLS;
4247 . First host accepts first address;
4248 . First host gives temporary error to second address;
4249 . Second host offers TLS and a TLS session is established;
4250 . Second host accepts second address.
4251 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4252 and peerdn, if requested) that were in fact used only for the second
4255 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4256 attention to any local part prefix or suffix that was matched by the
4257 router that accepted the address. It now behaves in the same way as it
4258 does for delivery: the affixes are removed from the local part unless
4259 rcpt_include_affixes is set on the transport.
4261 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4262 timeout during the DATA phase of an incoming message.
4264 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4265 applied a patch sanctioned by the Sieve maintainer.
4267 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4268 the uid or gid is negative. A case of a negative gid caused this to be
4269 noticed. The fix allows for either to be negative.
4271 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4272 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4275 PH/36 The change PH/12 above was broken. Fixed it.
4277 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4278 the grounds that routing the same address twice would always produce the
4279 same answer. This might have been true once, but it is certainly no
4280 longer true now. Routing a child address may depend on the previous
4281 routing that produced that child. Some complicated redirection strategies
4282 went wrong when messages had multiple recipients, and made Exim's
4283 behaviour dependent on the order in which the addresses were given.
4285 I have moved the duplicate checking until after the routing is complete.
4286 Exim scans the addresses that are assigned to local and remote
4287 transports, and removes any duplicates. This means that more work will be
4288 done, as duplicates will always all be routed, but duplicates are
4289 presumably rare, so I don't expect this is of any significance.
4291 For deliveries to pipes, files, and autoreplies, the duplicate checking
4292 still happens during the routing process, since they are not going to be
4295 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4296 It corrects a timeout issue with spamd. This is Ian's comment: "The
4297 background is that sometimes spamd either never reads data from a
4298 connection it has accepted, or it never writes response data. The exiscan
4299 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4300 blindly assumes that writes won't block so it may never time out."
4302 PH/39 Allow G after quota size as well as K and M.
4304 PH/40 The value set for $authenticated_id in an authenticator may not contain
4305 binary zeroes or newlines because the value is written to log lines and
4306 to spool files. There was no check on this. Now the value is run through
4307 the string_printing() function so that such characters are converted to
4308 printable escape sequences.
4310 PH/41 $message_linecount is a new variable that contains the total number of
4311 lines in the message. Compare $body_linecount, which is the count for the
4314 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4315 bounce and delay warning messages, except in certain special cases, which
4318 (a) An SMTP error message from a remote host;
4319 (b) A message specified in a :fail: redirection;
4320 (c) A message specified in a "fail" command in a system filter;
4321 (d) A message specified in a FAIL return from the queryprogram router;
4322 (e) A message specified by the cannot_route_message router option.
4324 In these cases only, Exim does include the error details in bounce and
4325 warning messages. There are also a few cases where bland messages such
4326 as "unrouteable address" or "local delivery error" are given.
4328 PH/43 $value is now also set for the "else" part of a ${run expansion.
4330 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4331 being worked on, but at least Exim now implements the latest version to
4334 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4335 process to complete was treated as a delivery failure, a timeout while
4336 writing the message to the pipe was logged, but erroneously treated as a
4337 successful delivery. Such timeouts include transport filter timeouts. For
4338 consistency with the overall process timeout, these timeouts are now
4339 treated as errors, giving rise to delivery failures by default. However,
4340 there is now a new Boolean option for the pipe transport called
4341 timeout_defer, which, if set TRUE, converts the failures into defers for
4342 both kinds of timeout. A transport filter timeout is now identified in
4345 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4346 systems where "make" and "gmake" are different, calling "gmake" at top
4347 level broke things. I've arranged for the value of $(MAKE) to be passed
4348 from the Makefile to this script so that it can call the same version of
4352 A note about Exim versions 4.44 and 4.50
4353 ----------------------------------------
4355 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4356 changes of various kinds. As a consequence, a big documentation update was
4357 needed. This delayed the release for rather longer than seemed good, especially
4358 in the light of a couple of (minor) security issues. Therefore, the changes
4359 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4360 release. So 4.44 and 4.50 are in effect two different branches that both start
4363 I have left the 4.50 change log unchanged; it contains all the changes since
4364 4.43. The change log for 4.44 is below; many of its items are identical to
4365 those for 4.50. This seems to be the most sensible way to preserve the
4366 historical information.
4372 1. Minor wording change to the doc/README.SIEVE file.
4374 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4375 computation of the current number of files was incorrect.
4377 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4378 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4381 4. Give more explanation in the error message when the command for a transport
4382 filter fails to execute.
4384 5. There are several places where Exim runs a non-Exim command in a
4385 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4386 was being done only for the command run by the queryprogram router. It is
4387 now done for all such subprocesses. The other cases are: ${run, transport
4388 filters, and the commands run by the lmtp and pipe transports.
4390 6. Added CONFIGURE_GROUP build-time option.
4392 7. Some older OS have a limit of 256 on the maximum number of file
4393 descriptors. Exim was using setrlimit() to set 1000 as a large value
4394 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4395 systems. I've change it so that if it can't get 1000, it tries for 256.
4397 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4398 was an oversight, and furthermore, ever since the addition of extra
4399 controls (e.g. 4.43/32), the checks on when to allow different forms of
4400 "control" were broken. There should now be diagnostics for all cases when a
4401 control that does not make sense is encountered.
4403 9. Added the /retain_sender option to "control=submission".
4405 10. $recipients is now available in the predata ACL (oversight).
4407 11. Tidy the search cache before the fork to do a delivery from a message
4408 received from the command line. Otherwise the child will trigger a lookup
4409 failure and thereby defer the delivery if it tries to use (for example) a
4410 cached ldap connection that the parent has called unbind on.
4412 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4413 of $address_data from the recipient verification was clobbered by the
4414 sender verification.
4416 13. The value of address_data from a sender verification is now available in
4417 $sender_address_data in subsequent conditions in the ACL statement.
4419 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4421 15. Added a new option "connect=<time>" to callout options, to set a different
4424 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4425 was its contents. (It was OK if the option was not defined at all.)
4427 17. A "Completed" log line is now written for messages that are removed from
4428 the spool by the -Mrm option.
4430 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4431 information about exactly what failed.
4433 19. Added -dd to debug only the daemon process.
4435 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4436 handles timeouts, both on the server side and network timeouts. Renamed the
4437 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4439 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4440 It is now set to "smtps".
4442 22. $host_address is now set to the target address during the checking of
4443 ignore_target_hosts.
4445 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4446 being passed; this would have caused $sender_host_name to have been used if
4447 matching the list had actually called for a host name (not very likely,
4448 since this list is usually IP addresses). A host name is now passed as
4451 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4452 code that specifies a non-restarting handler (typically sigaction() in
4453 modern systems) in an attempt to fix a rare and obscure crash bug.
4455 25. Narrowed the window for a race in the daemon that could cause it to ignore
4456 SIGCHLD signals. This is not a major problem, because they are used only to
4457 wake it up if nothing else does.
4459 26. A malformed maildirsize file could cause Exim to calculate negative values
4460 for the mailbox size or file count. Odd effects could occur as a result.
4461 The maildirsize information is now recalculated if the size or filecount
4464 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4465 support for a long time. Removed HAVE_SYS_VFS_H.
4467 28. Installed the latest version of exipick from John Jetmore.
4469 29. In an address list, if the pattern was not a regular expression, an empty
4470 subject address (from a bounce message) matched only if the pattern was an
4471 empty string. Non-empty patterns were not even tested. This was the wrong
4472 because it is perfectly reasonable to use an empty address as part of a
4473 database query. An empty address is now tested by patterns that are
4474 lookups. However, all the other forms of pattern expect the subject to
4475 contain a local part and a domain, and therefore, for them, an empty
4476 address still always fails if the pattern is not itself empty.
4478 30. Exim went into a mad DNS loop when attempting to do a callout where the
4479 host was specified on an smtp transport, and looking it up yielded more
4480 than one IP address.
4482 31. Re-factored the code for checking spool and log partition space into a
4483 function that finds that data and another that does the check. The former
4484 is then used to implement four new variables: $spool_space, $log_space,
4485 $spool_inodes, and $log_inodes.
4487 32. The RFC2047 encoding function was originally intended for short strings
4488 such as real names; it was not keeping to the 75-character limit for
4489 encoded words that the RFC imposes. It now respects the limit, and
4490 generates multiple encoded words if necessary. To be on the safe side, I
4491 have increased the buffer size for the ${rfc2047: expansion operator from
4494 33. It is now permitted to omit both strings after an "if" condition; if the
4495 condition is true, the result is "true". As before, when the second string
4496 is omitted, a false condition yields an empty string. This makes it less
4497 cumbersome to write custom ACL and router conditions.
4499 34. Failure to deliver a bounce message always caused it to be frozen, even if
4500 there was an errors_to setting on the router. The errors_to setting is now
4503 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4504 canonical form (fully expanded) before being placed in
4505 $sender_host_address.
4507 36. The table in the code that translates DNS record types into text (T_A to
4508 "A" for instance) was missing entries for NS and CNAME. It is just possible
4509 that this could have caused confusion if both these types were looked up
4510 for the same domain, because the text type is used as part of Exim's
4511 per-process caching. But the chance of anyone hitting this buglet seems
4514 37. The dnsdb lookup has been extended in a number of ways.
4516 (1) There is a new type, "zns", which walks up the domain tree until it
4517 finds some nameserver records. It should be used with care.
4519 (2) There is a new type, "mxh", which is like "mx" except that it returns
4520 just the host names, not the priorities.
4522 (3) It is now possible to give a list of domains (or IP addresses) to be
4523 looked up. The behaviour when one of the lookups defers can be
4524 controlled by a keyword.
4526 (4) It is now possible to specify the separator character for use when
4527 multiple records are returned.
4529 38. The dnslists ACL condition has been extended: it is now possible to supply
4530 a list of IP addresses and/or domains to be looked up in a particular DNS
4533 39. Added log_selector=+queue_time_overall.
4535 40. When running the queue in the test harness, wait just a tad after forking a
4536 delivery process, to get repeatability of debugging output.
4538 41. Include certificate and key file names in error message when GnuTLS fails
4539 to set them up, because the GnuTLS error message doesn't include the name
4540 of the failing file when there is a problem reading it.
4542 42. Allow both -bf and -bF in the same test run.
4544 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4546 44. The "Exiscan patch" is now merged into the mainline Exim source.
4548 45. Sometimes the final signoff response after QUIT could fail to get
4549 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4550 < 0 before doing a fflush(). This bug looks as though it goes back to the
4551 introduction of TLS in release 3.20, but "sometimes" must have been rare
4552 because the tests only now provoked it.
4554 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4555 (this can affect the format of dates).
4557 47. exim_tidydb, when checking for the continued existence of a message for
4558 which it has found a message-specific retry record, was not finding
4559 messages that were in split spool directories. Consequently, it was
4560 deleting retry records that should have stayed in existence.
4562 48. Steve fixed some bugs in eximstats.
4564 49. The SPA authentication driver was not abandoning authentication and moving
4565 on to the next authenticator when an expansion was forced to fail,
4566 contradicting the general specification for all authenticators. Instead it
4567 was generating a temporary error. It now behaves as specified.
4569 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4570 (the order specifies the preference for clients). The order is now AES256,
4571 AES128, 3DES, ARCFOUR128.
4573 51. Small patch to Sieve code - explicitly set From: when generating an
4576 52. Exim crashed if a remote delivery caused a very long error message to be
4577 recorded - for instance if somebody sent an entire SpamAssassin report back
4578 as a large number of 550 error lines. This bug was coincidentally fixed by
4579 increasing the size of one of Exim's internal buffers (big_buffer) that
4580 happened as part of the Exiscan merge. However, to be on the safe side, I
4581 have made the code more robust (and fixed the comments that describe what
4584 53. Now that there can be additional text after "Completed" in log lines (if
4585 the queue_time_overall log selector is set), a one-byte patch to exigrep
4586 was needed to allow it to recognize "Completed" as not the last thing in
4589 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4590 patch that reportedly fixes this has been added. I am not expert enough to
4591 create a test for it. This is what the patch creator wrote:
4593 "I found a little strange behaviour of ldap code when working with
4594 Windows 2003 AD Domain, where users was placed in more than one
4595 Organization Units. When I tried to give exim partial DN, the exit code
4596 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4597 But simultaneously result of request was absolutely normal ldap result,
4598 so I produce this patch..."
4600 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4601 REFERENCE, so I have modified the code to exclude the patch when that macro
4604 55. Some experimental protocols are using DNS PTR records for new purposes. The
4605 keys for these records are domain names, not reversed IP addresses. The
4606 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4607 leaves it alone. Component reversal etc. now happens only for IP addresses.
4610 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4612 57. Double the size of the debug message buffer (to 2048) so that more of very
4613 long debug lines gets shown.
4615 58. The exicyclog utility now does better if the number of log files to keep
4616 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4618 59. Two changes related to the smtp_active_hostname option:
4620 (1) $smtp_active_hostname is now available as a variable.
4621 (2) The default for smtp_banner uses $smtp_active_hostname instead
4622 of $primary_hostname.
4624 60. The host_aton() function is supposed to be passed a string that is known
4625 to be a valid IP address. However, in the case of IPv6 addresses, it was
4626 not checking this. This is a hostage to fortune. Exim now panics and dies
4627 if the condition is not met. A case was found where this could be provoked
4628 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4629 components; fortuitously, this particular loophole had already been fixed
4630 by change 4.50/55 above.
4632 If there are any other similar loopholes, the new check in host_aton()
4633 itself should stop them being exploited. The report I received stated that
4634 data on the command line could provoke the exploit when Exim was running as
4635 exim, but did not say which command line option was involved. All I could
4636 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4637 running as the user.
4640 61. There was a buffer overflow vulnerability in the SPA authentication code
4641 (which came originally from the Samba project). I have added a test to the
4642 spa_base64_to_bits() function which I hope fixes it.
4645 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4646 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4648 63. The daemon start-up calls getloadavg() while still root for those OS that
4649 need the first call to be done as root, but it missed one case: when
4650 deliver_queue_load_max is set with deliver_drop_privilege. This is
4651 necessary for the benefit of the queue runner, because there is no re-exec
4652 when deliver_drop_privilege is set.
4654 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4655 This has been fixed.
4657 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4658 was in use, was not putting the data itself into the right store pool;
4659 consequently, it could be overwritten for a subsequent message in the same
4660 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4663 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4665 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4666 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4667 the function were treating the return as a boolean value, which happened to
4668 work because 0=false and not-0=true, but is not correct code.
4670 68. The host_aton() function was not handling scoped IPv6 addresses (those
4671 with, for example, "%eth0" on the end) correctly.
4673 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4674 negated items (that is, ~something) in unsigned ints. Some compilers
4675 apparently mutter when there is no cast.
4677 70. If an address verification called from an ACL failed, and did not produce a
4678 user-specific message (i.e. there was only a "system" message), nothing was
4679 put in $acl_verify_message. In this situation, it now puts the system
4682 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4683 session; change 4.31/43 added the unwanted input to the log line - except
4684 that it did not do this in the start of session case. It now does.
4686 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4687 This is wrong and can cause the other end to generate a synchronization
4688 error if it is another Exim or anything else that does the synchronization
4689 check. A QUIT command is no longer sent after a timeout.
4691 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4692 during host lookups.
4694 74. The defer_ok option of callout verification was not working if it was used
4695 when verifying addresses in header lines, that is, for this case:
4697 verify = header_sender/callout=defer_ok
4699 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4700 those file descriptors could be used for SMTP connections. If anything
4701 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4702 could be sent to the SMTP client, causing chaos. The daemon now opens
4703 stdin, stdout, and stderr to /dev/null when it puts itself into the
4706 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4707 log by default. The user can override this with suitable Perl magic.
4709 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4710 the log message when discard triggers, was not working for the DATA ACL or
4711 for the non-SMTP ACL.
4713 78. Error message wording change in sieve.c.
4715 79. If smtp_accept_max_per_host was set, the number of connections could be
4716 restricted to fewer than expected, because the daemon was trying to set up
4717 a new connection before checking whether the processes handling previous
4718 connections had finished. The check for completed processes is now done
4719 earlier. On busy systems, this bug wouldn't be noticed because something
4720 else would have woken the daemon, and it would have reaped the completed
4723 80. If a message was submitted locally by a user whose login name contained one
4724 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4725 It caused a spool format error. I have fixed the spool reading code. A
4726 related problem was that the "from" clause in the Received: line became
4727 illegal because of the space(s). It is now covered by ${quote_local_part.
4729 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4732 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4734 83. Patch from Sieve maintainer to fix unterminated string problem in
4735 "vacation" handling.
4737 84. Some minor changes to the Linux configuration files to help with other
4738 OS variants using glibc.
4740 85. One more patch for Sieve to update vacation handling to latest spec.
4743 ----------------------------------------------------
4744 See the note above about the 4.44 and 4.50 releases.
4745 ----------------------------------------------------
4751 1. Change 4.43/35 introduced a bug that caused file counts to be
4752 incorrectly computed when quota_filecount was set in an appendfile
4755 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4756 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4759 3. Give more explanation in the error message when the command for a transport
4760 filter fails to execute.
4762 4. There are several places where Exim runs a non-Exim command in a
4763 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4764 was being done only for the command run by the queryprogram router. It is
4765 now done for all such subprocesses. The other cases are: ${run, transport
4766 filters, and the commands run by the lmtp and pipe transports.
4768 5. Some older OS have a limit of 256 on the maximum number of file
4769 descriptors. Exim was using setrlimit() to set 1000 as a large value
4770 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4771 systems. I've change it so that if it can't get 1000, it tries for 256.
4773 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4774 was an oversight, and furthermore, ever since the addition of extra
4775 controls (e.g. 4.43/32), the checks on when to allow different forms of
4776 "control" were broken. There should now be diagnostics for all cases when a
4777 control that does not make sense is encountered.
4779 7. $recipients is now available in the predata ACL (oversight).
4781 8. Tidy the search cache before the fork to do a delivery from a message
4782 received from the command line. Otherwise the child will trigger a lookup
4783 failure and thereby defer the delivery if it tries to use (for example) a
4784 cached ldap connection that the parent has called unbind on.
4786 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4787 of $address_data from the recipient verification was clobbered by the
4788 sender verification.
4790 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4791 was its contents. (It was OK if the option was not defined at all.)
4793 11. A "Completed" log line is now written for messages that are removed from
4794 the spool by the -Mrm option.
4796 12. $host_address is now set to the target address during the checking of
4797 ignore_target_hosts.
4799 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4800 being passed; this would have caused $sender_host_name to have been used if
4801 matching the list had actually called for a host name (not very likely,
4802 since this list is usually IP addresses). A host name is now passed as
4805 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4806 code that specifies a non-restarting handler (typically sigaction() in
4807 modern systems) in an attempt to fix a rare and obscure crash bug.
4809 15. Narrowed the window for a race in the daemon that could cause it to ignore
4810 SIGCHLD signals. This is not a major problem, because they are used only to
4811 wake it up if nothing else does.
4813 16. A malformed maildirsize file could cause Exim to calculate negative values
4814 for the mailbox size or file count. Odd effects could occur as a result.
4815 The maildirsize information is now recalculated if the size or filecount
4818 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4819 support for a long time. Removed HAVE_SYS_VFS_H.
4821 18. Updated exipick to current release from John Jetmore.
4823 19. Allow an empty sender to be matched against a lookup in an address list.
4824 Previously the only cases considered were a regular expression, or an
4827 20. Exim went into a mad DNS lookup loop when doing a callout where the
4828 host was specified on the transport, if the DNS lookup yielded more than
4831 21. The RFC2047 encoding function was originally intended for short strings
4832 such as real names; it was not keeping to the 75-character limit for
4833 encoded words that the RFC imposes. It now respects the limit, and
4834 generates multiple encoded words if necessary. To be on the safe side, I
4835 have increased the buffer size for the ${rfc2047: expansion operator from
4838 22. Failure to deliver a bounce message always caused it to be frozen, even if
4839 there was an errors_to setting on the router. The errors_to setting is now
4842 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4843 canonical form (fully expanded) before being placed in
4844 $sender_host_address.
4846 24. Updated eximstats to version 1.33
4848 25. Include certificate and key file names in error message when GnuTLS fails
4849 to set them up, because the GnuTLS error message doesn't include the name
4850 of the failing file when there is a problem reading it.
4852 26. Expand error message when OpenSSL has problems setting up cert/key files.
4855 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4856 (this can affect the format of dates).
4858 28. exim_tidydb, when checking for the continued existence of a message for
4859 which it has found a message-specific retry record, was not finding
4860 messages that were in split spool directories. Consequently, it was
4861 deleting retry records that should have stayed in existence.
4863 29. eximstats updated to version 1.35
4864 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4865 1.35 - bugfix such that pie charts by volume are generated correctly
4867 30. The SPA authentication driver was not abandoning authentication and moving
4868 on to the next authenticator when an expansion was forced to fail,
4869 contradicting the general specification for all authenticators. Instead it
4870 was generating a temporary error. It now behaves as specified.
4872 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4873 (the order specifies the preference for clients). The order is now AES256,
4874 AES128, 3DES, ARCFOUR128.
4876 31. Small patch to Sieve code - explicitly set From: when generating an
4879 32. Exim crashed if a remote delivery caused a very long error message to be
4880 recorded - for instance if somebody sent an entire SpamAssassin report back
4881 as a large number of 550 error lines. This bug was coincidentally fixed by
4882 increasing the size of one of Exim's internal buffers (big_buffer) that
4883 happened as part of the Exiscan merge. However, to be on the safe side, I
4884 have made the code more robust (and fixed the comments that describe what
4887 33. Some experimental protocols are using DNS PTR records for new purposes. The
4888 keys for these records are domain names, not reversed IP addresses. The
4889 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4890 leaves it alone. Component reversal etc. now happens only for IP addresses.
4893 34. The host_aton() function is supposed to be passed a string that is known
4894 to be a valid IP address. However, in the case of IPv6 addresses, it was
4895 not checking this. This is a hostage to fortune. Exim now panics and dies
4896 if the condition is not met. A case was found where this could be provoked
4897 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4898 components; fortuitously, this particular loophole had already been fixed
4899 by change 4.50/55 or 4.44/33 above.
4901 If there are any other similar loopholes, the new check in host_aton()
4902 itself should stop them being exploited. The report I received stated that
4903 data on the command line could provoke the exploit when Exim was running as
4904 exim, but did not say which command line option was involved. All I could
4905 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4906 running as the user.
4909 35. There was a buffer overflow vulnerability in the SPA authentication code
4910 (which came originally from the Samba project). I have added a test to the
4911 spa_base64_to_bits() function which I hope fixes it.
4914 36. The daemon start-up calls getloadavg() while still root for those OS that
4915 need the first call to be done as root, but it missed one case: when
4916 deliver_queue_load_max is set with deliver_drop_privilege. This is
4917 necessary for the benefit of the queue runner, because there is no re-exec
4918 when deliver_drop_privilege is set.
4920 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4921 was in use, was not putting the data itself into the right store pool;
4922 consequently, it could be overwritten for a subsequent message in the same
4923 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4926 38. Sometimes the final signoff response after QUIT could fail to get
4927 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4928 < 0 before doing a fflush(). This bug looks as though it goes back to the
4929 introduction of TLS in release 3.20, but "sometimes" must have been rare
4930 because the tests only now provoked it.
4936 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4937 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4938 Now it returns an integer. A number of calls were still expecting a T/F
4939 return. Fortuitously, in all cases, the tests worked in OK situations,
4940 which is the norm. However, things would have gone wrong on any write
4941 failures on the smtp file descriptor. This function is used when sending
4942 messages over SMTP and also when doing verify callouts.
4944 2. When Exim is called to do synchronous delivery of a locally submitted
4945 message (the -odf or -odi options), it no longer closes stderr before doing
4948 3. Implemented the mua_wrapper option.
4950 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4952 5. Implemented the functions header_remove(), header_testname(),
4953 header_add_at_position(), and receive_remove_recipient(), and exported them
4956 6. If an ACL "warn" statement specified the addition of headers, Exim already
4957 inserted X-ACL-Warn: at the start if there was no header name. However, it
4958 was not making this test for the second and subsequent header lines if
4959 there were newlines in the string. This meant that an invalid header could
4960 be inserted if Exim was badly configured.
4962 7. Allow an ACL "warn" statement to add header lines at the start or after all
4963 the Received: headers, as well as at the end.
4965 8. Added the rcpt_4xx retry error code.
4967 9. Added postmaster_mailfrom=xxx to callout verification option.
4969 10. Added mailfrom=xxxx to the callout verification option, for verify=
4972 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4973 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4975 12. Inserted some casts to stop certain compilers warning when using pointer
4976 differences as field lengths or precisions in printf-type calls (mostly
4977 affecting debugging statements).
4979 13. Added optional readline() support for -be (dynamically loaded).
4981 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4982 same clock tick as a message's arrival, so that its received time was the
4983 same as the "first fail" time on the retry record, and that message
4984 remained on the queue past the ultimate address timeout, every queue runner
4985 would try a delivery (because it was past the ultimate address timeout) but
4986 after another failure, the ultimate address timeout, which should have then
4987 bounced the address, did not kick in. This was a "< instead of <=" error;
4988 in most cases the first failure would have been in the next clock tick
4989 after the received time, and all would be well.
4991 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4992 being recognized when the domain list was tested by the match_domain
4993 condition in an expansion string.
4995 16. Added the ${str2b64: operator.
4997 17. Exim was always calling setrlimit() to set a large limit for the number of
4998 processes, without checking whether the existing limit was already
4999 adequate. (It did check for the limit on file descriptors.) Furthermore,
5000 errors from getrlimit() and setrlimit() were being ignored. Now they are
5001 logged to the main and panic logs, but Exim does carry on, to try to do its
5002 job under whatever limits there are.
5004 18. Imported PCRE 5.0.
5006 19. Trivial typo in log message " temporarily refused connection" (the leading
5009 20. If the log selector return_path_on_delivery was set and an address was
5010 redirected to /dev/null, the delivery process crashed because it assumed
5011 that a return path would always be set for a "successful" delivery. In this
5012 case, the whole delivery is bypassed as an optimization, and therefore no
5015 21. Internal re-arrangement: the function for sending a challenge and reading
5016 a response while authentication was assuming a zero-terminated challenge
5017 string. It's now changed to take a pointer and a length, to allow for
5018 binary data in such strings.
5020 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5022 23. Exim was not respecting finduser_retries when seeking the login of the
5023 uid under which it was called; it was always trying 10 times. (The default
5024 setting of finduser_retries is zero.) Also, it was sleeping after the final
5025 failure, which is pointless.
5027 24. Implemented tls_on_connect_ports.
5029 25. Implemented acl_smtp_predata.
5031 26. If the domain in control=submission is set empty, Exim assumes that the
5032 authenticated id is a complete email address when it generates From: or
5033 Sender: header lines.
5035 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5036 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5037 chown and chgrp in /bin and hostname in /usr/bin.
5039 28. Exim was keeping the "process log" file open after each use, just as it
5040 does for the main log. This opens the possibility of it remaining open for
5041 long periods when the USR1 signal hits a daemon. Occasional processlog
5042 errors were reported, that could have been caused by this. Anyway, it seems
5043 much more sensible not to leave this file open at all, so that is what now
5046 29. The long-running daemon process does not normally write to the log once it
5047 has entered its main loop, and it closes the log before doing so. This is
5048 so that log files can straightforwardly be renamed and moved. However,
5049 there are a couple of unusual error situations where the daemon does write
5050 log entries, and I had neglected to close the log afterwards.
5052 30. The text of an SMTP error response that was received during a remote
5053 delivery was being truncated at 512 bytes. This is too short for some of
5054 the long messages that one sometimes sees. I've increased the limit to
5057 31. It is now possible to make retry rules that apply only when a message has a
5058 specific sender, in particular, an empty sender.
5060 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5061 it possible to be selective about when SMTP synchronization is enforced.
5063 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5065 32. Implemented hosts_connection_nolog.
5067 33. Added an ACL for QUIT.
5069 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5072 35. Added mailbox_size and mailbox_filecount to appendfile.
5074 36. Added control = no_multiline_responses to ACLs.
5076 37. There was a bug in the logic of the code that waits for the clock to tick
5077 in the case where the clock went backwards by a substantial amount such
5078 that the microsecond fraction of "now" was more than the microsecond
5079 fraction of "then" (but the whole seconds number was less).
5081 38. Added support for the libradius Radius client library this is found on
5082 FreeBSD (previously only the radiusclient library was supported).
5088 1. When certain lookups returned multiple values in the form name=value, the
5089 quoting of the values was not always being done properly. Specifically:
5090 (a) If the value started with a double quote, but contained no whitespace,
5092 (b) If the value contained whitespace other than a space character (i.e.
5093 tabs or newlines or carriage returns) it was not quoted.
5094 This fix has been applied to the mysql and pgsql lookups by writing a
5095 separate quoting function and calling it from the lookup code. The fix
5096 should probably also be applied to nisplus, ibase and oracle lookups, but
5097 since I cannot test any of those, I have not disturbed their existing code.
5099 2. A hit in the callout cache for a specific address caused a log line with no
5100 reason for rejecting RCPT. Now it says "Previous (cached) callout
5101 verification failure".
5103 3. There was an off-by-one bug in the queryprogram router. An over-long
5104 return line was truncated at 256 instead of 255 characters, thereby
5105 overflowing its buffer with the terminating zero. As well as fixing this, I
5106 have increased the buffer size to 1024 (and made a note to document this).
5108 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5109 when Exim is waiting for an SMTP response from a remote server, Exim
5110 restarts its select() call on the socket, thereby resetting its timeout.
5111 This is not a problem when such interrupts are rare. Somebody set up a cron
5112 job to run exiwhat every 2 minutes, which is less than the normal select()
5113 timeout (5 or 10 minutes). This meant that the select() timeout never
5114 kicked in because it was always reset. I have fixed this by comparing the
5115 time when an interrupt arrives with the time at the start of the first call
5116 to select(). If more time than the timeout has elapsed, the interrupt is
5117 treated as a timeout.
5119 5. Some internal re-factoring in preparation for the addition of Sieve
5120 extensions (by MH). In particular, the "personal" test is moved to a
5121 separate function, and given an option for scanning Cc: and Bcc: (which is
5122 not set for Exim filters).
5124 6. When Exim created an email address using the login of the caller as the
5125 local part (e.g. when creating a From: or Sender: header line), it was not
5126 quoting the local part when it contained special characters such as @.
5128 7. Installed new OpenBSD configuration files.
5130 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5131 try to make them clearer.
5133 9. Callout options, other than the timeout value, were being ignored when
5134 verifying sender addresses in header lines. For example, when using
5136 verify = header_sender/callout=no_cache
5138 the cache was (incorrectly) being used.
5140 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5141 only the Cygwin environment.
5143 11. When return_path_on_delivery was set as a log selector, if different remote
5144 addresses in the same message used different return paths and parallel
5145 remote delivery occurred, the wrong values would sometimes be logged.
5146 (Whenever a remote delivery process finished, the return path value from
5147 the most recently started remote delivery process was logged.)
5149 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5150 lines when AUTH and/or TLS are in use. This is the "received protocol"
5151 field. Exim used to use "asmtp" for authenticated SMTP, without any
5152 indication (in the protocol name) for TLS use. Now it follows the RFC and
5153 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5154 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5155 names appear in log lines as well as in Received: header lines.
5157 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5158 extensions, and comparison tests, and to fix some bugs.
5160 14. Changes to the "personal" filter test:
5162 (1) The test was buggy in that it was just doing the equivalent of
5163 "contains" tests on header lines. For example, if a user's address was
5164 anne@some.where, the "personal" test would incorrectly be true for
5166 To: susanne@some.where
5168 This test is now done by extracting each address from the header in turn,
5169 and checking the entire address. Other tests that are part of "personal"
5170 are now done using regular expressions (for example, to check local parts
5171 of addresses in From: header lines).
5173 (2) The list of non-personal local parts in From: addresses has been
5174 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5175 taken from the Sieve specification recommendations.
5177 (3) If the message contains any header line starting with "List-" it is
5178 treated as non-personal.
5180 (4) The test for "circular" in the Subject: header line has been removed
5181 because it now seems ill-conceived.
5183 15. Minor typos in src/EDITME comments corrected.
5185 16. Installed latest exipick from John Jetmore.
5187 17. If headers_add on a router specified a text string that was too long for
5188 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5189 of string_sprintf() is now avoided.
5191 18. $message_body_size was not set (it was always zero) when running the DATA
5192 ACL and the local_scan() function.
5194 19. For the "mail" command in an Exim filter, no default was being set for
5195 the once_repeat time, causing a random time value to be used if "once" was
5196 specified. (If the value happened to be <= 0, no repeat happened.) The
5197 default is now 0s, meaning "never repeat". The "vacation" command was OK
5198 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5199 (I found it when inspecting the code).
5201 20. There is now an overall timeout for performing a callout verification. It
5202 defaults to 4 times the callout timeout, which applies to individual SMTP
5203 commands during the callout. The overall timeout applies when there is more
5204 than one host that can be tried. The timeout is checked before trying the
5205 next host. This prevents very long delays if there are a large number of
5206 hosts and all are timing out (e.g. when the network connections are timing
5207 out). The value of the overall timeout can be changed by specifying an
5208 additional sub-option for "callout", called "maxwait". For example:
5210 verify = sender/callout=5s,maxwait=20s
5212 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5213 to the end before writing, but this should make it even safer).
5215 22. Exim was forgetting that it had advertised PIPELINING for the second and
5216 subsequent messages on an SMTP connection. It was also not resetting its
5217 memory on STARTTLS and an internal HELO.
5219 23. When Exim logs an SMTP synchronization error within a session, it now
5220 records whether PIPELINING has been advertised or not.
5222 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5223 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5224 rather than long int.
5226 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5228 26. Added the never_mail option to autoreply.
5234 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5235 crash if the getsockname() call failed; this can happen if a connection is
5236 closed very soon after it is established. The problem was simply in the
5237 order in which certain operations were done, causing Exim to try to write
5238 to the SMTP stream before it had set up the file descriptor. The bug has
5239 been fixed by making things happen in the correct order.
5245 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5246 before the connection was closed, thus losing the rejection response.
5248 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5249 some early Solaris releases, but causes trouble in current releases where
5250 socklen_t is defined.
5252 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5255 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5258 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5259 that releasing the top of it at the end releases what was used for sub-
5260 expansions (unless the block got too big). However, discard this block if
5261 the first thing is a variable or header, so that we can use its block when
5262 it is dynamic (useful for very large $message_headers, for example).
5264 6. Lookups now cache *every* query, not just the most recent. A new, separate
5265 store pool is used for this. It can be recovered when all lookup caches are
5266 flushed. Lookups now release memory at the end of their result strings.
5267 This has involved some general refactoring of the lookup sources.
5269 7. Some code has been added to the store_xxx() functions to reduce the amount
5270 of flapping under certain conditions.
5272 8. log_incoming_interface used to affect only the <= reception log lines. Now
5273 it causes the local interface and port to be added to several more SMTP log
5274 lines, for example "SMTP connection from", and rejection lines.
5276 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5278 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5280 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5281 could be overwritten at the end of the current message (or the start of a
5282 new message if it was set in a HELO ACL). The value is now preserved for
5283 the duration of the SMTP connection.
5285 12. If a transport had a headers_rewrite setting, and a matching header line
5286 contained an unqualified address, that address was qualified, even if it
5287 did not match any rewriting rules. The underlying bug was that the values
5288 of the flags that permit the existence of unqualified sender and recipient
5289 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5290 non-local messages, and by -bnq for local messages) were not being
5291 preserved with the message after it was received.
5293 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5294 "next input=" as part of the text comprising the host identity instead of
5295 the correct text. The code was using the same buffer for two different
5296 strings. However, depending on which order the printing function evaluated
5297 its arguments, the bug did not always show up. Under Linux, for example, my
5298 test suite worked just fine.
5300 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5301 doesn't work with some older versions of Perl. It has been changed to "my",
5302 which in any case is probably the better facility to use.
5304 15. A really picky compiler found some instances of statements for creating
5305 error messages that either had too many or two few arguments for the format
5308 16. The size of the buffer for calls to the DNS resolver has been increased
5309 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5310 for addresses that have a lot of PTR records. This alleviates a problem; it
5311 does not fully solve it.
5313 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5314 buffer now truncates the list and logs the incident, which is the same
5315 action as happens when Exim is looking up a host name and its aliases.
5316 Previously in this situation something unpredictable would happen;
5317 sometimes it was "internal error: store_reset failed".
5319 18. If a server dropped the connection unexpectedly when an Exim client was
5320 using GnuTLS and trying to read a response, the client delivery process
5321 crashed while trying to generate an error log message.
5323 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5324 string, for example:
5326 warn message = H1: something\nH2: something
5328 the text was added as a single header line from Exim's point of view
5329 though it ended up OK in the delivered message. However, searching for the
5330 second and subsequent header lines using $h_h2: did not work. This has been
5331 fixed. Similarly, if a system filter added multiple headers in this way,
5332 the routers could not see them.
5334 20. Expanded the error message when iplsearch is called with an invalid key to
5335 suggest using net-iplsearch in a host list.
5337 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5338 ACLs are no longer actually imposed (and a message to that effect is
5341 22. If a "gecos" field in a passwd entry contained escaped characters, in
5342 particular, if it contained a \" sequence, Exim got it wrong when building
5343 a From: or a Sender: header from that name. A second bug also caused
5344 incorrect handling when an unquoted " was present following a character
5345 that needed quoting.
5347 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5348 was not being matched caselessly.
5350 24. Arranged for all hyphens in the exim.8 source to be escaped with
5353 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5354 properly. Recipient callout cache records were still being keyed to include
5355 the sender, even when use_sender was set false. This led to far more
5356 callouts that were necessary. The sender is no longer included in the key
5357 when use_sender is false.
5359 26. Added "control = submission" modifier to ACLs.
5361 27. Added the ${base62d: operator to decode base 62 numbers.
5363 28. dnsdb lookups can now access SRV records.
5365 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5366 the configuration file.
5368 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5369 (-v) mode. This makes the output for a verbose queue run more intelligible.
5371 31. Added a use_postmaster feature to recipient callouts.
5373 32. Added the $body_zerocount variable, containing the number of binary zero
5374 bytes in the message body.
5376 33. The time of last modification of the "new" subdirectory is now used as the
5377 "mailbox time last read" when there is a quota error for a maildir
5380 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5382 35. Added +ignore_unknown as a special item in host lists.
5384 36. Code for decoding IPv6 addresses in host lists is now included, even if
5385 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5386 address was recognized as an IP address, but was then not correctly decoded
5387 into binary, causing unexpected and incorrect effects when compared with
5394 1. Very minor rewording of debugging text in manualroute to say "list of
5395 hosts" instead of "hostlist".
5397 2. If verify=header_syntax was set, and a header line with an unqualified
5398 address (no domain) and a large number of spaces between the end of the
5399 name and the colon was received, the reception process suffered a buffer
5400 overflow, and (when I tested it) crashed. This was caused by some obsolete
5401 code that should have been removed. The fix is to remove it!
5403 3. When running in the test harness, delay a bit after writing a bounce
5404 message to get a bit more predictability in the log output.
5406 4. Added a call to search_tidyup() just before forking a reception process. In
5407 theory, someone could use a lookup in the expansion of smtp_accept_max_
5408 per_host which, without the tidyup, could leave open a database connection.
5410 5. Added the variables $recipient_data and $sender_data which get set from a
5411 lookup success in an ACL "recipients" or "senders" condition, or a router
5412 "senders" option, similar to $domain_data and $local_part_data.
5414 6. Moved the writing of debug_print from before to after the "senders" test
5417 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5418 problems for message scanning, either using a data ACL, or using
5419 local_scan() because the Received: header was not generated till after they
5420 were called (in order to set the time as the time of reception completion).
5421 I have revised the way this works. The header is now generated after the
5422 body is received, but before the ACL or local_scan() are called. After they
5423 are run, the timestamp in the header is updated.
5429 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5430 before starting a queue runner without re-exec. This happened only when
5431 deliver_drop_privilege was set or when the Exim user was set to root. The
5432 effect of the bug was that timeouts during subsequent deliveries caused
5433 crashes instead of being properly handled. The handler is now left at its
5434 default (and expected) setting.
5436 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5437 message, again when deliver_drop_privilege is set or Exim is run as root.
5438 The bug described in (1) was not present in this case, but the tidying up
5439 of the other signals was missing. I have made the two cases consistent.
5441 3. The ignore_target_hosts setting on a manualroute router was being ignored
5442 for hosts that were looked up using the /MX notation.
5444 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5447 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5448 operated on the sender address. After changing the $sender_address to <>
5449 for the sender address verify, Exim was re-instated it as the original
5450 (before rewriting) address, but remembering that it had rewritten it, so it
5451 wasn't rewriting it again. This bug also had the effect of breaking the
5452 sender address verification caching when the sender address was rewritten.
5454 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5455 This has been changed so that if the ip literal address matches
5456 ignore_target_hosts, the router declines.
5458 7. Added expansion conditions match_domain, match_address, and match_local_
5459 part (NOT match_host).
5461 8. The placeholder for the Received: header didn't have a length field set.
5463 9. Added code to Exim itself and to exim_lock to test for a specific race
5464 condition that could lead to file corruption when using MBX delivery. The
5465 issue is with the lockfile that is created in /tmp. If this file is removed
5466 after a process has opened it but before that process has acquired a lock,
5467 there is the potential for a second process to recreate the file and also
5468 acquire a lock. This could lead to two Exim processes writing to the file
5469 at the same time. The added code performs the same test as UW imapd; it
5470 checks after acquiring the lock that its file descriptor still refers to
5471 the same named file.
5473 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5474 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5477 11. Added the smtp_active_hostname option. If used, this will typically be made
5478 to depend on the incoming interface address. Because $interface_address is
5479 not set up until the daemon has forked a reception process, error responses
5480 that can happen earlier (such as "too many connections") no longer contain
5483 12. If an expansion in a condition on a "warn" statement fails because a lookup
5484 defers, the "warn" statement is abandoned, and the next ACL statement is
5485 processed. Previously this caused the whole ACL to be aborted.
5487 13. Added the iplsearch lookup type.
5489 14. Added ident_timeout as a log selector.
5491 15. Added tls_certificate_verified as a log selector.
5493 16. Added a global option tls_require_ciphers (compare the smtp transport
5494 option of the same name). This controls incoming TLS connections.
5496 17. I finally figured out how to make tls_require_ciphers do a similar thing
5497 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5498 before starting the TLS session.
5500 18. Tabs are now shown as \t in -bP output.
5502 19. If the log selector return_path_on_delivery was set, Exim crashed when
5503 bouncing a message because it had too many Received: header lines.
5505 20. If two routers both had headers_remove settings, and the first one included
5506 a superfluous trailing colon, the final name in the first list and the
5507 first name in the second list were incorrectly joined into one item (with a
5508 colon in the middle).
5514 1. Added -C and -D options to the exinext utility, mainly to make it easier
5515 to include in the automated testing, but these could be helpful when
5516 multiple configurations are in use.
5518 2. The exinext utility was not formatting the output nicely when there was
5519 an alternate port involved in the retry record key, nor when there was a
5520 message id as well (for retries that were specific to a specific message
5521 and a specific host). It was also confused by IPv6 addresses, because of
5522 the additional colons they contain. I have fixed the IPv4 problem, and
5523 patched it up to do a reasonable job for IPv6.
5525 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5526 delivery, the log line now contains "pipelined" if PIPELINING was used.
5528 4. An SMTP transport process used to panic and die if the bind() call to set
5529 an explicit outgoing interface failed. This has been changed; it is now
5530 treated in the same way as a connect() failure.
5532 5. A reference to $sender_host_name in the part of a conditional expansion
5533 that was being skipped was still causing a DNS lookup. This no longer
5536 6. The def: expansion condition was not recognizing references to header lines
5537 that used bh_ and bheader_.
5539 7. Added the _cache feature to named lists.
5541 8. The code for checking quota_filecount in the appendfile transport was
5542 allowing one more file than it should have been.
5544 9. For compatibility with Sendmail, the command line option
5552 and sets the incoming protocol and host name (for trusted callers). The
5553 host name and its colon can be omitted when only the protocol is to be set.
5554 Note the Exim already has two private options, -pd and -ps, that refer to
5555 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5556 "s", but I don't think that's a major issue.
5558 10. A number of refactoring changes to the code, none of which should affect
5561 (a) The number of logging options was getting close to filling up the
5562 32-bit word that was used as a bit map. I have split them into two classes:
5563 those that are passed in the argument to log_write(), and those that are
5564 only ever tested independently outside of that function. These are now in
5565 separate 32-bit words, so there is plenty of room for expansion again.
5566 There is no change in the user interface or the logging behaviour.
5568 (b) When building, for example, log lines, the code previously used a
5569 macro that called string_cat() twice, in order to add two strings. This is
5570 not really sufficiently general. Furthermore, there was one instance where
5571 it was actually wrong because one of the argument was used twice, and in
5572 one call a function was used. (As it happened, calling the function twice
5573 did not affect the overall behaviour.) The macro has been replaced by a
5574 function that can join an arbitrary number of extra strings onto a growing
5577 (c) The code for expansion conditions now uses a table and a binary chop
5578 instead of a serial search (which was left over from when there were very
5579 few conditions). Also, it now recognizes conditions like "pam" even when
5580 the relevant support is not compiled in: a suitably worded error message is
5581 given if an attempt is made to use such a condition.
5583 11. Added ${time_interval:xxxxx}.
5585 12. A bug was causing one of the ddress fields not to be passed back correctly
5586 from remote delivery subprocesses. The field in question was not being
5587 subsequently used, so this caused to problems in practice.
5589 13. Added new log selectors queue_time and deliver_time.
5591 14. Might have fixed a bug in maildirsizefile handling that threw up
5592 "unexpected character" debug warnings, and recalculated the data
5593 unnecessarily. In any case, I expanded the warning message to give more
5596 15. Added the message "Restricted characters in address" to the statements in
5597 the default ACL that block characters like @ and % in local parts.
5599 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5600 Three changes have been made:
5602 (a) There was a serious bug; a negative response to MAIL caused the whole
5603 recipient domain to be cached as invalid, thereby blocking all messages
5604 to all local parts at the same domain, from all senders. This bug has
5605 been fixed. The domain is no longer cached after a negative response to
5606 MAIL if the sender used is not empty.
5608 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5611 (c) A new callout option, "use_sender" has been added for people who want
5612 the modified behaviour.
5618 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5621 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5622 indeed breaks things for older releases.
5624 3. Added additional logging to the case where there is a problem reading data
5625 from a filter that is running in a subprocess using a pipe, in order to
5626 try to track down a specific problem.
5628 4. Testing facility fudge: when running in the test harness and attempting
5629 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5630 getting "No route to host". Convert this to a timeout.
5632 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5635 6. Some OS don't have socklen_t but use size_t instead. This affects the
5636 fifth argument of getsockopt() amongst other things. This is now
5637 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5638 can be set for individual OS. I have set it for SunOS5, OSF1, and
5639 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5640 some earlier ones do not.
5642 7. Change 4.30/15 was not doing the test caselessly.
5644 8. The standard form for an IPv6 address literal was being rejected by address
5645 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5646 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5647 this, as well as the form without the "IPv6" on the front (but only when
5648 address literals are enabled, of course).
5650 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5652 10. Exim crashed if a message with an empty sender address specified by -f
5653 encountered a router with an errors_to setting. This could be provoked only
5654 by a command such as
5658 where an empty string was supplied; "<>" did not hit this bug.
5660 11. Installed PCRE release 4.5.
5662 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5663 remained set. It is now erased.
5665 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5666 times from message ids (which are base 36 rather than the normal 62).
5668 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5669 were being counted as actual protocol errors, and logged if the log
5670 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5671 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5672 and DATA following a set of rejected RCPTs do not count as protocol errors.
5673 In other words, Exim assumes they were pipelined, though this may not
5674 actually be the case. Of course, in all cases the client gets an
5675 appropriate error code.
5677 15. If a lookup fails in an ACL condition, a message about the failure may
5678 be available; it is used if testing the ACL cannot continue, because most
5679 such messages specify what the cause of the deferral is. However, some
5680 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5681 that caused an old message to be retained and used if a later statement
5682 caused a defer, replacing the real cause of the deferral.
5684 16. If an IP address had so many PTR records that the DNS lookup buffer
5685 was not large enough to hold them, Exim could crash while trying to process
5686 the truncated data. It now detects and logs this case.
5688 17. Further to 4.21/58, another change has been made: if (and only if) the
5689 first line of a message (the first header line) ends with CRLF, a bare LF
5690 in a subsequent header line has a space inserted after it, so as not to
5691 terminate the header.
5693 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5694 unnecessarily, used atoi() instead of strtol(), and didn't check the
5695 termination when getting file sizes from file names by regex.
5697 19. Completely re-implemented the support for maildirsize files, in the light
5698 of a number of problems with the previous contributed implementation
5699 (4.30/29). In particular:
5701 . If the quota is zero, the maildirsize file is maintained, but no quota is
5704 . If the maildir directory does not exist, it is created before any attempt
5705 to write a maildirsize file.
5707 . The quota value in the file is just a cache; if the quota is changed in
5708 the transport, the new value overrides.
5710 . A regular expression is available for excluding directories from the
5713 20. The autoreply transport checks the characters in options that define the
5714 message's headers; it allows continued headers, but it was checking with
5715 isspace() after an embedded newline instead of explicitly looking for a
5718 21. If all the "regular" hosts to which an address was routed had passed their
5719 expiry times, and had not reached their retry times, the address was
5720 bounced, even if fallback hosts were defined. Now Exim should go on to try
5723 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5724 equivalent code in the SMTP transport. Some hosts send humungous responses
5725 to HELO/EHLO, more than 1024 it seems.
5727 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5728 gives compiler warnings in some environments. I've now done it "properly",
5731 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5732 (because of problems with the built-in one) was declared to return uschar *
5733 instead of char *, causing compiler failure.
5735 25. Fixed a file descriptor leak when processing alias/forward files.
5737 26. Fixed a minor format string issue in dbfn.c.
5739 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5741 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5742 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5745 29. When a sender address is verified, it is cached, to save repeating the test
5746 when there is more than one recipient in a message. However, when the
5747 verification involves a callout, it is possible for different callout
5748 options to be set for different recipients. It is too complicated to keep
5749 track of this in the cache, so now Exim always runs a verification when a
5750 callout is required, relying on the callout cache for the optimization.
5751 The overhead is duplication of the address routing, but this should not be
5754 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5755 to be verified with callout=postmaster, and the main callout worked but the
5756 postmaster check failed, the verification correctly failed. However, if a
5757 subsequent RCPT command asked for sender verification *without* the
5758 postmaster check, incorrect caching caused this verification also to fail,
5761 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5762 it was not caching the DNS options (qualify_single, search_parents) that
5763 were used when the lookup failed. A subsequent lookup with different
5764 options therefore always gave the same answer, though there were cases
5765 where it should not have. (Example: a "domains = !$mx_any" option on a
5766 dnslookup router: the "domains" option is always processed without any
5767 widening, but the router might have qualify_single set.) Now Exim uses the
5768 cached value only when the same options are set.
5770 32. Added John Jetmore's "exipick" utility to the distribution.
5772 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5773 than a timeout (e.g. a certificate is required, and is not provided), an
5774 Exim server now closes the connection immediately. Previously it waited for
5775 the client to close - but if the client is SSL, it seems that they each
5776 wait for each other, leading to a delay before one of them times out.
5778 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5779 maintained 0.8.x compatibility because I don't think many are using it, and
5780 it is clearly obsolete.
5782 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5785 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5786 client certificate was expired. A simple patch fixes this, though I don't
5787 understand the full logic of why the verify callback is called multiple
5790 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5791 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5792 which causes problems with some clients (such as the Certicom SSL Plus
5793 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5794 disables the coutermeasure allowing Eudora to connect."
5796 38. Exim was not checking that a write() to a log file succeeded. This could
5797 lead to Bad Things if a log got too big, in particular if it hit a file
5798 size limit. Exim now panics and dies if it cannot write to a log file, just
5799 as it does if it cannot open a log file.
5801 39. Modified OS/Makefile-Linux so that it now contains
5803 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5805 The two -D definitions ensure that Exim is compiled with large file
5806 support, which makes it possible to handle log files that are bigger than
5809 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5810 instance) a domain was checked against a named list that involved a lookup,
5811 causing $domain_data to be set, then another domain was checked against the
5812 same list, then the first domain was re-checked, the value of $domain_data
5813 after the final check could be wrong. In particular, if the second check
5814 failed, it could be set empty. This bug probably also applied to
5817 41. The strip_trailing_dot option was not being applied to the address given
5818 with the -f command-line option.
5820 42. The code for reading a message's header from the spool was incrementing
5821 $received_count, but never initializing it. This meant that the value was
5822 incorrect (doubled) while delivering a message in the same process in which
5823 it was received. In the most common configuration of Exim, this never
5824 happens - a fresh exec is done - but it can happen when
5825 deliver_drop_privilege is set.
5827 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5828 it now includes up to 150 characters of the unexpected data in the log
5831 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5832 and building data strings. The size of both of these buffers was 10 000
5833 bytes - far larger than anybody would *ever* want, thought I. Needless to
5834 say, somebody hit the limit. I have increased the maximum line length to
5835 20 000 and the maximum data length of concatenated lines to 100 000. I have
5836 also fixed two bugs, because there was no checking on these buffers. Tsk,
5837 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5838 buffer is too small.
5840 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5841 lsearch lookups. Now it does.
5843 46. When parsing a route_list item in a manualroute router, a fixed-length
5844 buffer was used for the list of hosts. I made this 1024 bytes long,
5845 thinking that nobody would ever have a list of hosts that long. Wrong.
5846 Somebody had a whole pile of complicated expansion conditions, and the
5847 string was silently truncated, leading to an expansion error. It turns out
5848 that it is easier to change to an unlimited length (owing to other changes
5849 that have happened since this code was originally written) than to build
5850 structure for giving a limitation error. The length of the item that
5851 expands into the list of hosts is now unlimited.
5853 47. The lsearch lookup could not handle data where the length of text line was
5854 more than 4095 characters. Such lines were truncated, leading to shortened
5855 data being returned. It should now handle lines of any length.
5857 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5858 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5861 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5862 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5863 no longer happens in comment lines. A list of replacements is now placed
5864 at the head of all of the source files, except those whose only change is
5865 to replace PERL_COMMAND in the very first #! line.
5867 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5868 messages on the queue, with a bottom-up merge sort, using code contributed
5869 by Michael Haardt. This should make operations like -bp somewhat faster on
5870 large queues. It won't affect queue runners, except when queue_run_in_order
5873 51. Installed eximstats 1.31 in the distribution.
5875 52. Added support for SRV lookups to the dnslookup router.
5877 53. If an ACL referred to $message_body or $message_body_end, the value was not
5878 reset for any messages that followed in the same SMTP session.
5880 54. The store-handling optimization for building very long strings was not
5881 differentiating between the different store pools. I don't think this
5882 actually made any difference in practice, but I've tidied it.
5884 55. While running the routers to verify a sender address, $sender_address
5885 was still set to the sender address. This is wrong, because when routing to
5886 send a bounce to the sender, it would be empty. Therefore, I have changed
5887 it so that, while verifying a sender address, $sender_address is set to <>.
5888 (There is no change to what happens when verifying a recipient address.)
5890 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5891 target A or AAAA records (if not already returned) without resetting the
5892 qualify_single or search_parents options of the DNS resolver. These are
5893 inappropriate in this case because the targets of MX and SRV records must
5894 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5895 target that, when qualified, matched something in the local domain. These
5896 two options are now turned off when doing these lookups.
5898 57. It seems that at least some releases of Reiserfs (which does not have the
5899 concept of a fixed number of inodes) returns zero and not -1 for the
5900 number of available inodes. This interacted badly with check_spool_inodes,
5901 which assumed that -1 was the "no such thing" setting. What I have done is
5902 to check that the total number of inodes is greater than zero before doing
5903 the test of how many are available.
5905 58. When a "warn" ACL statement has a log_message modifier, the message is
5906 remembered, and not repeated. This is to avoid a lot of repetition when a
5907 message has many recipients that cause the same warning to be written.
5908 However, Exim was preserving the list of already written lines for an
5909 entire SMTP session, which doesn't seem right. The memory is now reset if a
5910 new message is started.
5912 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5913 debugging output unless log_rewrite was also set.
5915 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5916 of (int)(handle) when we know that handle contains (void *)(-1).
5918 61. The Exim daemon panic-logs an error return when it closes the incoming
5919 connection. However "connection reset by peer" seems to be common, and
5920 isn't really an error worthy of noting specially, so that particular error
5923 62. When Exim is trying to find all the local interfaces, it used to panic and
5924 die if the ioctl to get the interface flags failed. However, it seems that
5925 on at least one OS (Solaris 9) it is possible to have an interface that is
5926 included in the list of interfaces, but for which you get a failure error
5927 for this call. This happens when the interface is not "plumbed" into a
5928 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5929 failure of the "get flags" call assumes that the interface is down.
5931 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5932 makes life easier for people who are doing arithmetic on fields extracted
5933 from dates, where you often get leading zeros that should not be
5934 interpreted as octal.
5936 64. Added qualify_domain to the redirect router, to override the global
5939 65. If a pathologically long header line contained very many addresses (the
5940 report of this problem mentioned 10 000) and each of them was rewritten,
5941 Exim could use up a very large amount of memory. (It kept on making new
5942 copies of the header line as it rewrote, and never released the old ones.)
5943 At the expense of a bit more processing, the header rewriting function has
5944 been changed so that it no longer eats memory in this way.
5946 66. The generation of the Received: header has been moved from the time that a
5947 message starts to be received, to the time that it finishes. The timestamp
5948 in the Received: header should now be very close to that of the <= log
5949 line. There are two side-effects of this change:
5951 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5952 logged header lines no longer include the local Received: line, because
5953 it has not yet been created. The same applies to a copy of the message
5954 that is returned to a non-SMTP sender when a message is rejected.
5956 (b) When a filter file is tested using -bf, no additional Received: header
5957 is added to the test message. After some thought, I decided that this
5960 This change does not affect the value of $received_for. It is still set
5961 after address rewriting, but before local_scan() is called.
5963 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5965 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5966 gave an unhelpful panic error message, and a defer error. I have managed to
5967 change this behaviour so that it now rejects any supplied certificate,
5968 which seems right, as the list of acceptable certificates is empty.
5970 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5971 gave an unhelpful defer error. I have not managed to make this reject any
5972 supplied certificates, but the error message it gives is "no certificate
5973 supplied", which is not helpful.
5975 70. exigrep's output now also includes lines that are not associated with any
5976 message, but which match the given pattern. Implemented by a patch from
5977 Martin Sluka, which also tidied up the Perl a bit.
5979 71. Recipient callout verification, like sender verification, was using <> in
5980 the MAIL FROM command. This isn't really the right thing, since the actual
5981 sender may affect whether the remote host accepts the recipient or not. I
5982 have changed it to use the actual sender in the callout; this means that
5983 the cache record is now keyed on a recipient/sender pair, not just the
5984 recipient address. There doesn't seem to be a real danger of callout loops,
5985 since a callout by the remote host to check the sender would use <>.
5986 [SEE ABOVE: changed after hitting problems.]
5988 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5989 temporary errors. However, in the case of such a code being given after
5990 the end of a data transmission (i.e. after ".") Exim was failing to write
5991 a retry record for the message. (Yes, there was some broken host that was
5992 actually sending 8xx at this point.)
5994 73. An unknown lookup type in a host list could cause Exim to panic-die when
5995 the list was checked. (An example that provoked this was putting <; in the
5996 middle of a list instead of at the start.) If this happened during a DATA
5997 ACL check, a -D file could be left lying around. This kind of configuration
5998 error no longer causes Exim to die; instead it causes a defer error. The
5999 incident is still logged to the main and panic logs.
6001 74. Buglet left over from Exim 3 conversion. The message "too many messages
6002 in one connection" was written to the rejectlog but not the mainlog, except
6003 when address rewriting (yes!) was being logged.
6005 75. Added write_rejectlog option.
6007 76. When a system filter was run not as root (that is, when system_filter_user
6008 was set), the values of the $n variables were not being returned to the
6009 main process; thus, they were not subsequently available in the $sn
6012 77. Added +return_path_on_delivery log selector.
6014 78. A connection timeout was being treated differently from recipients deferred
6015 when testing hosts_max_try with a message that was older than the host's
6016 retry timeout. (The host should not be counted, thus allowing all hosts to
6017 be tried at least once before bouncing.) This may have been the cause of an
6018 occasionally reported bug whereby a message would remain on the queue
6019 longer than the retry timeout, but would be bounced if a delivery was
6020 forced. I say "may" because I never totally pinned down the problem;
6021 setting up timeout/retry tests is difficult. See also the next item.
6023 79. The ultimate address timeout was not being applied to errors that involved
6024 a combination of host plus message (for example, a timeout on a MAIL
6025 command). When an address resolved to a number of possible hosts, and they
6026 were not all tried for each delivery (e.g. because of hosts_max_try), a
6027 message could remain on the queue longer than the retry timeout.
6029 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6032 81. Fixed an obscure SMTP outgoing bug which required at least the following
6033 conditions: (a) there was another message waiting for the same server;
6034 (b) the server returned 5xx to all RCPT commands in the first message so
6035 that the message was not completed; (c) the server dropped the connection
6036 or gave a negative response to the RSET that Exim sends to abort the
6037 transaction. The observed case was a dropped connection after DATA that had
6038 been sent in pipelining mode. That is, the server had advertised PIPELINING
6039 but was not implementing it correctly. The effect of the bug was incorrect
6040 behaviour, such as trying another host, and this could lead to a crash.
6046 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6047 and daemon.c were passed as pointers to ints; they should have been
6048 pointers to socklen_t variables (which are typically unsigned ints).
6050 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6053 3. Fixed a really odd bug that affected only the testing scheme; patching a
6054 certain fixed string in the binary changed the value of another string that
6055 happened to be identical to the end of the original first string.
6057 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6058 name", it returns that address as the IP address. On some operating
6059 systems (e.g. Solaris), it also passes back the IP address string as the
6060 "host name". However, on others (e.g. Linux), it passes back an empty
6061 string. Exim wasn't checking for this, and was changing the host name to an
6062 empty string, assuming it had been canonicalized.
6064 5. Although rare, it is permitted to have more than one PTR record for a given
6065 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6066 all the names associated with an address, because they do in Solaris.
6067 However, it seems that they do not in Linux for data that comes from the
6068 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6069 I found this out when I moved to a new Linux workstation and tried to run
6070 the Exim test suite.
6072 To get round this problem I have changed the code so that it now does its
6073 own call to the DNS to look up PTR records when searching for a host name.
6074 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6075 addresses that are only in /etc/hosts are still found.
6077 This behaviour is, however, controlled by an option called host_lookup_
6078 order, which defaults to "bydns:byaddr". If people want to use the other
6079 order, or indeed, just use one or the other means of lookup, they can
6080 specify it in this variable.
6082 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6083 some operating systems, this comes back from gethostbyaddr() as an empty
6084 string, and this is what Exim used to test for. However, it seems that in
6085 other systems, "." is yielded. Exim now tests for this case too.
6087 7. The values of check_spool_space and check_log_space are now held internally
6088 as a number of kilobytes instead of an absolute number of bytes. If a
6089 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6090 kilobyte. This means that much larger values can be stored.
6092 8. Exim monitor: an attempt to get the action menu when not actually pointing
6093 at a message produces an empty menu entitled "No message selected". This
6094 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6095 no entries in it ("Shell widget menu has zero width and/or height"). So I
6096 have added a single, blank menu entry in this case.
6098 9. Added ${quote_local_part.
6100 10. MIME decoding is now applied to the contents of Subject: header lines when
6103 11. Now that a reference to $sender_host_address automatically causes a reverse
6104 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6105 host lookup before query-style lookups in lists that might use this
6106 variable. This has therefore been abolished, and the "net-" prefix is no
6107 longer necessary for query-style lookups.
6109 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6110 have been a typo for LFLAGS, so it has been changed.
6112 13. The install script calls Exim with "-C /dev/null" in order to find the
6113 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6114 to be output. However, since Exim outputs its version number before the
6115 error, it didn't break the script. It just looked ugly. I fixed this by
6116 always allowing "-C /dev/null" if the caller is root.
6118 14. Ignore overlarge ACL variable number when reading spool file - insurance
6119 against a later release with more variables having written the file.
6121 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6122 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6123 this, as well as the form without the "IPv6" on the front.
6125 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6126 OS/Makefile-Darwin file.
6128 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6129 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6131 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6132 variable. However, some OS use ints for inodes. I've added cast to long int
6133 to get rid of the compiler warning.
6135 19. I had forgotten to lock out "/../" in configuration file names when
6136 ALT_CONFIG_PREFIX was set.
6138 20. Routers used for verification do not need to specify transports. However,
6139 if such a router generated a host list, and callout was configured, Exim
6140 crashed, because it could not find a port number from the (non-existent)
6141 transport. It now assumes port 25 in this circumstance.
6143 21. Added the -t option to exigrep.
6145 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6146 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6147 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6148 them is set, LOOKUP_LSEARCH is forced.
6150 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6152 24. Added sender and host information to the "rejected by local_scan()" log
6153 line; previously there was no indication of these.
6155 25. Added .include_if_exists.
6157 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6158 Linux. It turns out that not all file systems support this. Apparently some
6159 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6160 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6161 not supported on the file descriptor, is now ignored when Exim is trying to
6162 sync a directory. This applies only to Linux.
6164 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6166 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6167 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6170 29. Added support for maildirsize files from supplied patch (modified a bit).
6172 30. The use of :fail: followed by an empty string could lead Exim to respond to
6173 sender verification failures with (e.g.):
6175 550 Verification failed for <xxx>
6176 550 Sender verify failed
6178 where the first response line was missing the '-' that indicates it is not
6179 the final line of the response.
6181 31. The loop for finding the name of the user that called Exim had a hardwired
6182 limit of 10; it now uses the value of finduser_retries, which is used for
6183 all other user lookups.
6185 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6188 33. Exim was neglecting to zero errno before one call of strtol() when
6189 expanding a string and expecting an integer value. On some systems this
6190 resulted in spurious "integer overflow" errors. Also, it was casting the
6191 result into an int without checking.
6193 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6194 did not work. The code looks as if it has *never* worked, though it appears
6195 to have been documented since at least release 1.62. I have made it work.
6197 35. The "timeout_DNS" error in retry rules, also documented since at least
6198 1.62, also never worked. As it isn't clear exactly what this means, and
6199 clearly it isn't a major issue, I have abolished the feature by treating it
6200 as "timeout", and writing a warning to the main and panic logs.
6202 36. The display of retry rules for -brt wasn't always showing the error code
6205 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6206 timeout_connect_A, timeout_connect_MX.
6208 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6209 to the empty sender.
6211 39. The daemon was not analysing the content of -oX till after it had closed
6212 stderr and disconnected from the controlling terminal. This meant that any
6213 syntax errors were only noted on the panic log, and the return code from
6214 the command was 0. By re-arranging the code a little, I've made the
6215 decoding happen first, so such errors now appear on stderr, and the return
6216 code is 1. However, the actual setting up of the sockets still happens in
6217 the disconnected process, so errors there are still only recorded on the
6220 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6221 connections (as happens on some IP stacks) was logged at start up time as
6222 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6223 it from "IPv6 and IPv4", which means that two separate sockets are being
6226 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6227 says whether AF_INET or AF_INET6 was passed as an argument.
6229 42. Exiwhat output was messed up when time zones were included in log
6232 43. Exiwhat now gives more information about the daemon's listening ports,
6233 and whether -tls-on-connect was used.
6235 44. The "port" option of the smtp transport is now expanded.
6237 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6238 silently ignored. Now an error message is written to the main and panic
6241 46. There's a new ACL modifier called "logwrite" which writes to a log file
6242 as soon as it is encountered.
6244 47. Added $local_user_uid and $local_user_gid at routing time.
6246 48. Exim crashed when trying to verify a sender address that was being
6249 49. Exim was recognizing only a space character after ".include". It now also
6250 recognizes a tab character.
6252 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6253 extracting the relevant information from the specification. The man page no
6254 longer contains scrambled data for the -d option, and I've added a section
6255 at the front about calling Exim under different names.
6257 51. Added "extra_headers" argument to the "mail" command in filter files.
6259 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6262 53. Installed eximstats 1.29.
6264 54. Added transport_filter_timeout as a generic transport option.
6266 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6267 Cc: header lines. This was required by RFC 822, but it not required by RFC
6270 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6271 incoming messages that did not have them. Now it does so only if the
6272 message originates locally, that is, if there is no associated remote host
6273 address. When Resent- header lines are present, this applies to the Resent-
6274 lines rather than the non-Resent- lines.
6276 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6277 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6279 58. Messages for configuration errors now include the name of the main
6280 configuration file - useful now that there may be more than one file in a
6281 list (.included file names were always shown).
6283 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6284 for those rare installations that do not start the daemon as root or run it
6285 setuid root. I've cut out the call to initgroups() if the daemon is not
6288 60. The Exim user and group can now be bound into the binary as text strings
6289 that are looked up at the start of Exim's processing.
6291 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6293 62. Added $mailstore_basename variable.
6295 63. Installed patch to sieve.c from Michael Haardt.
6297 64. When Exim failed to open the panic log after failing to open the main log,
6298 the original message it was trying to log was written to stderr and debug
6299 output, but if they were not available (the usual case in production), it
6300 was lost. Now it is written to syslog before the two lines that record the
6301 failures to open the logs.
6303 65. Users' Exim filters run in subprocesses under the user's uid. It is
6304 possible for a "deliver" command or an alias in a "personal" command to
6305 provoke an address rewrite. If logging of address rewriting is configured,
6306 this fails because the process is not running as root or exim. There may be
6307 a better way of dealing with this, but for the moment (because 4.30 needs
6308 to be released), I have disabled address rewrite logging when running a
6309 filter in a non-root, non-exim process.
6315 1. The buildconfig auxiliary program wasn't quoting the value set for
6316 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6317 not defined. This bug was masked in 4.22 by the effect that was fixed in
6320 2. Some messages that were rejected after a message id was allocated were
6321 shown as "incomplete" by exigrep. It no longer does this for messages that
6322 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6324 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6325 have allow_domain_literals set, the ID did not get logged in the <= line.
6326 Domain literals are now always recognized in Message-ID: header lines.
6328 4. The first argument for a ${extract expansion item is the key name or field
6329 number. Leading and trailing spaces in this item were not being ignored,
6330 causing some misleading effects.
6332 5. When deliver_drop_privilege was set, single queue runner processes started
6333 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6334 same command in the process it spins off) were not dropping privilege.
6336 6. When the daemon running as "exim" started a queue runner, it always
6337 re-executed Exim in the spun-off process. This is a waste of effort when
6338 deliver_drop_privilege is set. The new process now just calls the
6339 queue-runner function directly.
6345 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6348 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6349 ignored. Though the use of -oP was forcing the writing of a pid file, it
6350 was always written to the default place.
6352 3. If the message "no IP address found for host xxxx" is generated during
6353 incoming verification, it is now followed by identification of the incoming
6354 connection (so you can more easily find what provoked it).
6356 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6358 5. Added some features to "harden" Exim a bit more against certain attacks:
6360 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6361 be put in Local/Makefile. This is like the never_users runtime option,
6362 but it cannot be overridden. The default setting is "root".
6364 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6365 prefix string with which any file named in a -C command line option
6368 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6369 is retained for -C and -D only if the caller of Exim is root. Without
6370 it, the exim user may also use -C and -D and retain privilege.
6372 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6373 command line option is disabled.
6375 6. Macro names set by the -D option must start with an upper case letter, just
6376 like macro names defined in the configuration file.
6378 7. Added "dereference=" facility to LDAP.
6380 8. Two instances of the typo "uknown" in the source files are fixed.
6382 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6383 the Configure-Makefile script screwed up while processing it.
6385 10. Incorporated PCRE 4.4.
6387 11. The SMTP synchronization check was not operating right at the start of an
6388 SMTP session. For example, it could not catch a HELO sent before the client
6389 waited for the greeting. There is now a check for outstanding input at the
6390 point when the greeting is written. Because of the duplex, asynchronous
6391 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6392 way, but not yet received, when the check is performed.
6394 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6395 on TCP/IP sockets, because this apparently causes some broken clients to
6398 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6399 unchanged) from the Cygwin maintainer.
6401 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6402 was supported instead of testing for mbx. Effectively a typo.
6404 15. The spa authenticator server code was not checking that the input it
6405 received was valid base64.
6407 16. The debug output line for the "set" modifier in ACLs was not showing the
6408 name of the variable that was being set.
6410 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6412 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6413 lookup on its own. Something else was needed to trigger the lookup. For
6414 example, a match in host_lookup or the need for a host name in a host list.
6415 Now, if $sender_host_name is referenced and the host name has not yet been
6416 looked up, a lookup is performed. If the lookup fails, the variable remains
6417 empty, and $host_lookup_failed is set to "1".
6419 19. Added "eqi" as a case-independent comparison operator.
6421 20. The saslauthd authentication condition could segfault if neither service
6422 nor realm was specified.
6424 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6425 error message that was logged was misleading, and incoming SMTP
6426 connections were dropped. The message is now more accurate, and temporary
6427 errors are given to SMTP connections.
6429 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6430 (except RSET) with a 421 error, until QUIT is received. However, it was
6431 failing to send a response to QUIT.
6433 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6434 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6435 ACL is now run after the helo_try_verify_hosts code.
6437 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6438 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6439 case-independent, so other case variants are also recognized). Apparently
6440 some systems use these upper case variants.
6442 25. If more than two messages were waiting for the same host, and a transport
6443 filter was specified for the transport, Exim sent two messages over the
6444 same TCP/IP connection, and then failed with "socket operation on non-
6445 socket" when it tried to send the third.
6447 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6449 27. The extern definition of crypt16() in expand.c was not being excluded when
6450 the OS had its own crypt16() function.
6452 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6453 as a preferred synonym for return_size_limit, both as an option and as an
6456 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6458 30. Changed the default configuration ACL to relax the local part checking rule
6459 for addresses that are not in any local domains. For these addresses,
6460 slashes and pipe symbols are allowed within local parts, but the sequence
6461 /../ is explicitly forbidden.
6463 31. SPA server authentication was not clearing the challenge buffer before
6466 32. log_message in a "warn" ACL statement was writing to the reject log as
6467 well as to the main log, which contradicts the documentation and doesn't
6468 seem right (because no rejection is happening). So I have stopped it.
6470 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6471 However, I am unable to do any testing of this.
6473 34. Fixed an infelicity in the appendfile transport. When checking directories
6474 for a mailbox, to see if any needed to be created, it was accidentally
6475 using path names with one or more superfluous leading slashes; tracing
6476 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6478 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6479 contents are added to the log line that is written for every discarded
6480 recipient. (Previously a log_message setting was ignored.)
6482 36. The ${quote: operator now quotes the string if it is empty.
6484 37. The install script runs exim in order to find its version number. If for
6485 some reason other than non-existence or emptiness, which it checks, it
6486 could not run './exim', it was installing it with an empty version number,
6487 i.e. as "exim-". This error state is now caught, and the installation is
6490 38. An argument was missing from the function that creates an error message
6491 when Exim fails to connect to the socket for saslauthd authentication.
6492 This could cause Exim to crash, or give a corrupted message.
6494 39. Added isip, isip4, and isip6 to ${if conditions.
6496 40. The ACL variables $acl_xx are now saved with the message, and can be
6497 accessed later in routers, transports, and filters.
6499 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6500 strings in the file are not string-expanded.
6502 42. If a MAIL command specified a SIZE value that was too large to fit into an
6503 int variable, the check against message_size_limit failed. Such values are
6504 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6505 day this will have to be increased, but I don't think I want to be around
6506 when emails are that large.
6513 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6514 iconv() is not standard in FreeBSD.
6516 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6517 IPv6 enabled. The observed symptom was a segmentation fault on return from
6518 the function os_common_find_running_interfaces() in src/os.c.
6520 3. In the check_special_case() function in daemon.c I had used "errno" as an
6521 argument name, which causes warnings on some systems. This was basically a
6522 typo, since it was named "eno" in the comments!
6524 4. The code that waits for the clock to tick (at a resolution of some fraction
6525 of a second) so as to ensure message-id uniqueness was always waiting for
6526 at least one whole tick, when it could have waited for less. [This is
6527 almost certainly not relevant at current processor speeds, where it is
6528 unlikely to ever wait at all. But we try to future-proof.]
6530 5. The function that sleeps for a time interval that includes fractions of a
6531 second contained a race. It did not block SIGALRM between setting the
6532 timer, and suspending (a couple of lines later). If the interval was short
6533 and the sigsuspend() was delayed until after it had expired, the suspension
6534 never ended. On busy systems this could lead to processes getting stuck for
6537 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6538 process, before it forks any delivery processes. The open lookup caching
6539 mechanism meant that the open file or database connection was passed into
6540 the delivery process. The problem was that delivery processes always tidy
6541 up cached lookup data. This could cause a problem for the next delivery
6542 process started by the queue runner, because the external queue runner
6543 process does not know about the closure. So the next delivery process
6544 still has data in the lookup cache. In the case of a file lookup, there was
6545 no problem because closing a file descriptor in a subprocess doesn't affect
6546 the parent. However, if the lookup was caching a connection to a database,
6547 the connection was closed, and the second delivery process was likely to
6548 see errors such as "PGSQL: query failed: server closed the connection
6549 unexpectedly". The problem has been fixed by closing all cached lookups
6550 in a queue runner before running a delivery process.
6552 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6553 seem to have the "const" qualifier which it has on other OS. I've
6556 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6557 *of the same type* (client or server) with the same public name that an
6558 error should be diagnosed.
6560 9. When Exim looked up a host name for an IP address, but failed to find the
6561 original IP address when looking up the host name (a safety check), it
6562 output the message "<ip address> does not match any IP for NULL", which was
6563 confusing, to say the least. The bug was that the host name should have
6564 appeared instead of "NULL".
6566 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6567 user that is built into the binary, and the -C or -D options is used, root
6568 privilege is dropped before the configuration file is read. In addition,
6569 logging is switched to stderr instead of the normal log files. If the
6570 configuration then re-defines the Exim user, the unprivileged environment
6571 is probably not what is expected, so Exim logs a panic warning message (but
6574 However, if deliver_drop_privilege is set, the unprivileged state may well
6575 be exactly what is intended, so the warning has been cut out in that case,
6576 and Exim is allowed to try to write to its normal log files.
6582 1. smtp_return_error_details was not giving details for temporary sender
6583 or receiver verification errors.
6585 2. Diagnose a configuration error if two authenticators have the same public
6588 3. Exim used not to create the message log file for a message until the first
6589 delivery attempt. This could be confusing when incoming messages were held
6590 for policy or load reasons. The message log file is now created at the time
6591 the message is received, and an initial "Received" line is written to it.
6593 4. The automatically generated man page for command line options had a minor
6594 bug that caused no ill effects; however, a more serious problem was that
6595 the procedure for building the man page automatically didn't always
6596 operate. Consequently, release 4.20 contains an out-of-date version. This
6597 shouldn't happen again.
6599 5. When building Exim with embedded Perl support, the script that builds the
6600 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6601 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6603 6. The freeze_tell option was not being used for messages that were frozen on
6604 arrival, either by an ACL or by local_scan().
6606 7. Added the smtp_incomplete_transaction log selector.
6608 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6609 was accepting AUTH without a new EHLO.
6611 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6612 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6615 10. Logging of TCP/IP connections (when configured) now happens in the main
6616 daemon process instead of the child process, so that the TCP/IP connection
6617 count is more accurate (but it can never be perfect).
6619 11. The use of "drop" in a nested ACL was not being handled correctly in the
6620 outer ACL. Now, if condition failure induced by the nested "drop" causes
6621 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6622 or "require"), the connection is dropped.
6624 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6625 that yield "discard" can now be used with an "accept" or a "discard" verb,
6626 but an error is generated for any others (because I can't see a useful way
6627 to define what should happen).
6629 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6630 are now processed in the same way as lines in the Exim configuration file.
6631 In particular, continuation lines are supported.
6633 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6635 15. Added -ti meaning -t -i.
6637 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6638 domains, and warn by logging if others are found.
6640 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6641 returned from GIFCONF when Exim is trying to find the list of interfaces on
6642 a host. The code in os.c has been modified to copy each ifreq to an aligned
6643 structure in all cases.
6645 Also, in some cases, the returned ifreq's were being copied to a 'struct
6646 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6647 means the last couple of bytes of an IPv6 address could be chopped if the
6648 ifreq contained only a normal sockaddr (14 bytes storage).
6650 18. Named domain lists were not supported in the hosts_treat_as_local option.
6651 An entry such as +xxxx was not recognized, and was treated as a literal
6654 19. Ensure that header lines added by a DATA ACL are included in the reject log
6655 if the ACL subsequently rejects the message.
6657 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6658 MD5 (which is deprecated).
6660 21. When testing a filter file using -bf, Exim was writing a message when it
6661 took the sender from a "From " line in the message, but it was not doing so
6662 when it took $return_path from a Return-Path: header line. It now does.
6664 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6665 with a valid header line field name (a series of printing characters
6666 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6668 23. Changed "disc" in the source to "disk" to conform to the documentation and
6669 the book and for uniformity.
6671 24. Ignore Sendmail's -Ooption=value command line item.
6673 25. When execve() failed while trying to run a command in a pipe transport,
6674 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6675 could be confused with a return value of 69 from the command itself. This
6676 has been changed to 127, the value the shell returns if it is asked to run
6677 a non-existent command. The wording for the related log line suggests a
6678 non-existent command as the problem.
6680 26. If received_header_text expands to an empty string, do not add a Received:
6681 header line to the message. (Well, it adds a token one on the spool, but
6682 marks it "old" so that it doesn't get used or transmitted.)
6684 27. Installed eximstats 1.28 (addition of -nt option).
6686 28. There was no check for failure on the call to getsockname() in the daemon
6687 code. This can fail if there is a shortage of resources on the system, with
6688 ENOMEM, for example. A temporary error is now given on failure.
6690 29. Contrary to the C standard, it seems that in some environments, the
6691 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6692 program. Exim now does this explicitly; it affects the formatting of
6693 timestamps using strftime().
6695 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6696 complaints. I've now initialized all the variables, to avoid this.
6698 32. Header lines added by a system filter were not being "seen" during
6699 transport-time rewrites.
6701 33. The info_callback() function passed to OpenSSL is set up with type void
6702 (*)(SSL *, int, int), as described somewhere. However, when calling the
6703 function (actually a macro) that sets it up, the type void(*)() is
6704 expected. I've put in a cast to prevent warnings from picky compilers.
6706 34. If a DNS black list lookup found a CNAME record, but there were no A
6707 records associated with the domain it pointed at, Exim crashed.
6709 35. If a DNS black list lookup returned more than one A record, Exim ignored
6710 all but the first. It now scans all returned addresses if a particular IP
6711 value is being sought. In this situation, the contents of the
6712 $dnslist_value variable are a list of all the addresses, separated by a
6715 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6716 to accept a host name and all its aliases if the forward lookup for any of
6717 them yielded the IP address of the incoming connection. Now it accepts only
6718 those names whose forward lookup yields the correct IP address. Any other
6719 names are discarded. This closes a loophole whereby a rogue DNS
6720 administrator could create reverse DNS records to break through a
6721 wildcarded host restriction in an ACL.
6723 37. If a user filter or a system filter that ran in a subprocess used any of
6724 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6725 the wrong values were passed to the pipe command ($thisaddress had the
6726 value of $0, $0 had the value of $1, etc). This bug was introduced by
6727 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6729 38. Improved the line breaking for long SMTP error messages from ACLs.
6730 Previously, if there was no break point between 40 and 75 characters, Exim
6731 left the rest of the message alone. Two changes have been made: (a) I've
6732 reduced the minimum length to 35 characters; (b) if it can't find a break
6733 point between 35 and 75 characters, it looks ahead and uses the first one
6734 that it finds. This may give the occasional overlong line, but at least the
6735 remaining text gets split now.
6737 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6738 file descriptors might be low, and that setting 1000 would always raise it.
6739 It turns out that in some environments, the limit is already over 1000 and
6740 that lowering it causes trouble. So now Exim takes care not to decrease it.
6742 40. When delivering a message, the value of $return_path is set to $sender_
6743 address at the start of routing (routers may change the value). By an
6744 oversight, this default was not being set up when an address was tested by
6745 -bt or -bv, which affected the outcome if any router or filter referred to
6748 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6749 log only when "message" or "log_message" are set. However, if one of the
6750 conditions was an address verification, or a call to a nested ACL, the
6751 messages generated by the underlying test were being passed through. This
6752 no longer happens. The underlying message is available in $acl_verify_
6753 message for both "message" and "log_message" expansions, so it can be
6754 passed through if needed.
6756 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6757 new expansion $bh_ to give the encoded byte string without charset
6758 translation. Translation happens only if iconv() is available; HAVE_ICONV
6759 indicates this at build time. HEADERS_CHARSET gives the charset to
6760 translate to; headers_charset can change it in the configuration, and
6761 "headers charset" can change it in an individual filter file.
6763 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6764 that creates RFC 2047 encoded "words" labels them as that charset instead
6765 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6766 expansion operator; (ii) when Exim creates a From: line for a local
6767 message; (iii) when a header line is rewritten to include a "phrase" part.
6769 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6770 buggy, causing it to skip the first lines of messages whose message ID
6771 ended in 'D'. This would not have bitten before Exim release 4.14, because
6772 message IDs were unlikely to end in 'D' before then. The effect was to have
6773 incorrect size information for certain domains.
6775 45. #include "config.h" was missing at the start of the crypt16.c module. This
6776 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6779 46. If there was a timeout during a "random" callout check, Exim treated it as
6780 a failure of the random address, and carried on sending RSET and the real
6781 address. If the delay was just some slowness somewhere, the response to the
6782 original RCPT would be taken as a response to RSET and so on, causing
6783 mayhem of various kinds.
6785 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6786 when I implemented it. It didn't allow for the fact that some option values
6787 may legitimately be negative (e.g. size_addition), and it didn't even do
6788 the right test for positive values.
6790 48. Domain names in DNS records are case-independent. Exim always looks them up
6791 in lower case. Some resolvers return domain names in exactly the case they
6792 appear in the zone file, that is, they may contain uppercase letters. Not
6793 all resolvers do this - some return always lower case. Exim was treating a
6794 change of case by a resolver as a change of domain, similar to a widening
6795 of a domain abbreviation. This triggered its re-routing code and so it was
6796 trying to route what was effectively the same domain again. This normally
6797 caused routing to fail (because the router wouldn't handle the domain
6798 twice). Now Exim checks for this case specially, and just changes the
6799 casing of the domain that it ultimately uses when it transmits the message
6802 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6805 50. If a filter generated a file delivery with a non-absolute name (possible if
6806 no home directory exists for the router), the forbid_file option was not
6809 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6810 the existing equality matching.
6812 52. Exim was using ints instead of ino_t variables in some places where it was
6813 dealing with inode numbers.
6815 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6816 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6817 called TMPDIR, and if it finds it is different, it changes its value.
6819 54. The smtp_printf() function is now made available to local_scan() so
6820 additional output lines can be written before returning. There is also an
6821 smtp_fflush() function to enable the detection of a dropped connection.
6822 The variables smtp_input and smtp_batched_input are exported to
6825 55. Changed the default runtime configuration: the message "Unknown user"
6826 has been removed from the ACL, and instead placed on the localuser router,
6827 using the cannot_route_message feature. This means that any verification
6828 failures that generate their own messages won't get overridden. Similarly,
6829 the "Unrouteable address" message that was in the ACL for unverifiable
6830 relay addresses has also been removed.
6832 56. Added hosts_avoid_esmtp to the smtp transport.
6834 57. The exicyclog script was not checking for the esoteric option
6835 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6836 will work only if exicyclog is run under the appropriate euid.
6838 58. Following a discussion on the list, the rules by which Exim recognises line
6839 endings on incoming messages have been changed. The -dropcr and drop_cr
6840 options are now no-ops, retained only for backwards compatibility. The
6841 following line terminators are recognized: LF CRLF CR. However, special
6842 processing applies to CR:
6844 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6845 nor a local message in the state where . is a terminator.
6847 (ii) If a bare CR is encountered in a header line, an extra space is added
6848 after the line terminator so as not to end the header. The reasoning
6849 behind this is that bare CRs in header lines are most likely either
6850 to be mistakes, or people trying to play silly games.
6852 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6853 was being incorrectly given as 18 bytes larger than it should have been.
6854 This is a VOB (very old bug).
6856 60. This may never have affected anything current, but just in case it has:
6857 When the local host is found other than at the start of a list of hosts,
6858 the local host, those with the same MX, and any that follow, are discarded.
6859 When the list in question was part of a longer list of hosts, the following
6860 hosts (not currently being processed) were also being discarded. This no
6861 longer happens. I'm not sure if this situation could ever has previously
6864 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6867 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6868 header. This is something that is recommended in a new Internet Draft, and
6869 is something that is documented as being done by Sendmail. There are two
6870 possible values. For messages generated by the autoreply transport, Exim
6873 Auto-Submitted: auto-replied
6875 whereas for all other generated messages (e.g. bounces) it adds
6877 Auto-Submitted: auto-generated
6879 63. The "personal" condition in filters now includes a test for the
6880 Auto-Submitted: header. If it contains the string "auto-" the message it
6881 not considered personal.
6883 64. Added rcpt_include_affixes as a generic transport option.
6885 65. Added queue_only_override (default true).
6887 66. Added the syslog_duplication option.
6889 67. If what should have been the first header line of a message consisted of
6890 a space followed by a colon, Exim was mis-interpreting it as a header line.
6891 It isn't of course - it is syntactically invalid and should therefore be
6892 treated as the start of the message body. The misbehaviour could have
6893 caused a number of strange effects, including loss of data in subsequent
6894 header lines, and spool format errors.
6896 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6897 client host had authenticated. This control can now be exercised by an ACL
6898 for more flexibility.
6900 69. By default, callouts do not happen when testing with -bh. There is now a
6901 variant, -bhc, which does actually run the callout code, including
6902 consulting and updating the callout cache.
6904 70. Added support for saslauthd authentication, courtesy of Alexander
6907 71. If statvfs() failed on the spool or log directories while checking their
6908 size for availability, Exim confusingly gave the error "space shortage".
6909 Furthermore, in debugging mode it crashed with a floating point exception.
6910 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6911 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6912 serious problem, Exim now writes to the main and panic logs when this
6913 happens, with details of the failure. It then refuses to accept the
6914 incoming message, giving the message "spool directory problem" or "log
6915 directory problem" with a 421 code for SMTP messages.
6917 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6918 0, 1, and 2 exist, because some OS complain for execs without them (see
6919 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6920 descriptors. However, the code omitted to check that the open succeeded,
6921 causing mysterious errors if for some reason the permissions on /dev/null
6922 got screwed. Now Exim writes a message to the main and panic logs, and
6923 bombs out if it can't open /dev/null.
6925 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6926 interact so that it is all more flexible. It is supposed to remain
6927 backwards compatible. Also added extra_local_interfaces.
6929 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6930 to bomb out with an assertion failure - to the client this appears as a
6931 connection drop. This problem occurs in the part of the code that was taken
6932 from the Samba project. Fortunately, the assertion is in a very simple
6933 function, so I have fixed this by reproducing the function inline in the
6934 one place where it is called, and arranging for authentication to fail
6935 instead of killing the process with assert().
6937 75. The SPA client code was not working when the server requested OEM rather
6938 than Unicode encoding.
6940 76. Added code to make require_files with a specific uid setting more usable in
6941 the case where statting the file as root fails - usually a non-root-mounted
6942 NFS file system. When this happens and the failure is EACCES, Exim now
6943 forks a subprocess and does the per-uid checking as the relevant uid.
6945 77. Added process_log_path.
6947 78. If log_file_path was not explicitly set, a setting of check_log_space or
6948 check_log_inodes was ignored.
6950 79. If a space check for the spool or log partitions fails, the incident is now
6951 logged. Of course, in the latter case the data may get lost...
6953 80. Added the %p formatting code to string_format() so that it can be used to
6954 print addresses in debug_print(). Adjusted all the address printing in the
6955 debugging in store.c to use %p rather than %d.
6957 81. There was a concern that a line of code in smtp_in.c could overflow a
6958 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6959 initially expressed, the concern was not well-founded, because trailing
6960 spaces are removed early. However, if the trailing spaces were followed by
6961 a NULL, they did not get removed, so the overflow was possible. Two fixes
6964 (a) I re-wrote the offending code in a cleaner fashion.
6965 (b) If an incoming SMTP command contains a NULL character, it is rejected
6968 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6969 runs initgroups(), so that if the Exim user is in any additional groups,
6970 they will be used during message reception.
6976 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob