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.
19 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
20 Previously only the last row was returned.
22 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
23 we assumed that tags in the header were well-formed, and parsed the
24 element content after inspecting only the first char of the tag.
25 Assumptions at that stage could crash the receive process on malformed
28 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
29 While running the DKIM ACL we operate on the Permanent memory pool so that
30 variables created with "set" persist to the DATA ACL. Also (at any time)
31 DNS lookups that fail create cache records using the Permanent pool. But
32 expansions release any allocations made on the current pool - so a dnsdb
33 lookup expansion done in the DKIM ACL releases the memory used for the
34 DNS negative-cache, and bad things result. Solution is to switch to the
35 Main pool for expansions.
36 While we're in that code, add checks on the DNS cache during store_reset,
37 active in the testsuite.
38 Problem spotted, and debugging aided, by Wolfgang Breyha.
40 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
41 When none of the hosts presented to a transport match an already-open
42 connection, close it and proceed with the list. Previously we would
43 queue the message. Spotted by Lena with Yahoo, probably involving
46 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
47 Previously a spurious "250 OK id=" response was appended to the proper
50 JH/08 The "support for" informational output now, which built with Content
51 Scanning support, has a line for the malware scanner interfaces compiled
52 in. Interface can be individually included or not at build time.
54 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
55 by the template makefile "src/EDITME". The "STREAM" support for an older
56 ClamAV interface method is removed.
58 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
59 rows affected is given instead).
61 JH/11 The runtime Berkeley DB library version is now additionally output by
62 "exim -d -bV". Previously only the compile-time version was shown.
64 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
65 SMTP connection. Previously, when one had more receipients than the
66 first, an abortive onward connection was made. Move to full support for
67 multiple onward connections in sequence, handling cutthrough connection
68 for all multi-message initiating connections.
70 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
71 routers. Previously, a multi-recipient message would fail to match the
72 onward-connection opened for the first recipient, and cause its closure.
74 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
75 a timeout on read on a GnuTLS initiating connection, resulting in the
76 initiating connection being dropped. This mattered most when the callout
77 was marked defer_ok. Fix to keep the two timeout-detection methods
80 JH/15 Relax results from ACL control request to enable cutthrough, in
81 unsupported situations, from error to silently (except under debug)
82 ignoring. This covers use with PRDR, frozen messages, queue-only and
85 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
86 metadata, resulting in a crash in free().
88 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
89 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
90 Broken also in d185889f4, with init system revamp.
92 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
93 Previously we abruptly closed the connection after reading a malware-
94 found indication; now we go on to read the "scan ok" response line,
97 JH/18 Convert macro handling to be tree-based, from the previous linear list.
98 With the number of builtin macros we now have this is worthwhile,
99 dropping the config-file read time (during which new macros are checked
100 and registered, and macros are expanded) from about 500 usec to about 180.
106 JH/01 Rework error string handling in TLS interface so that the caller in
107 more cases is responsible for logging. This permits library-sourced
108 string to be attached to addresses during delivery, and collapses
109 pairs of long lines into single ones.
111 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
112 during configuration. Wildcards are allowed and expanded.
114 JH/02 Rework error string handling in DKIM to pass more info back to callers.
115 This permits better logging.
117 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
118 do not close it down and have the child transport start it up again on
119 the passed-on TCP connection. Instead, proxy the child (and any
120 subsequent ones) for TLS via a unix-domain socket channel. Logging is
121 affected: the continued delivery log lines do not have any DNSSEC, TLS
122 Certificate or OCSP information. TLS cipher information is still logged.
124 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
125 identical IP addresses on different listening ports. Will also affect
128 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
129 add noisy ifdef guards to special-case this sillyness.
130 Patch from Bernd Kuhls.
132 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
133 than 255 are no longer allowed.
135 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
136 Disable the session-cache too, which might reduce our load. Since we
137 currrectly use a new context for every connection, both as server and
138 client, there is no benefit for these.
139 GnuTLS appears to not support tickets server-side by default (we don't
140 call gnutls_session_ticket_enable_server()) but client side is enabled
141 by default on recent versions (3.1.3 +) unless the PFS priority string
144 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
145 <https://reproducible-builds.org/specs/source-date-epoch/>.
147 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
148 the check for any unsuccessful recipients did not notice the limit, and
149 erroneously found still-pending ones.
151 JH/08 Pipeline CHUNKING command and data together, on kernels that support
152 MSG_MORE. Only in-clear (not on TLS connections).
154 JH/09 Avoid using a temporary file during transport using dkim. Unless a
155 transport-filter is involved we can buffer the headers in memory for
156 creating the signature, and read the spool data file once for the
157 signature and again for transmission.
159 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
160 4.77 as the kernel support then wasn't solid, having issues in 64bit
161 mode. Now, it's been long enough. Add support for FreeBSD also.
163 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
164 case where the routing stage had gathered several addresses to send to
165 a host before calling the transport for the first, we previously failed
166 to close down TLS in the old transport process before passing the TCP
167 connection to the new process. The new one sent a STARTTLS command
168 which naturally failed, giving a failed delivery and bloating the retry
169 database. Investigation and fix prototype from Wolfgang Breyha.
171 JH/12 Fix check on SMTP command input synchronisation. Previously there were
172 false-negatives in the check that the sender had not preempted a response
173 or prompt from Exim (running as a server), due to that code's lack of
174 awareness of the SMTP input buffering.
176 PP/04 Add commandline_checks_require_admin option.
177 Exim drops privileges sanely, various checks such as -be aren't a
178 security problem, as long as you trust local users with access to their
179 own account. When invoked by services which pass untrusted data to
180 Exim, this might be an issue. Set this option in main configuration
181 AND make fixes to the calling application, such as using `--` to stop
184 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
185 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
186 responses to those, into a single TLS record each way (this usually means
187 a single packet). As a side issue, smtp_enforce_sync now works on TLS
190 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
191 affects you only if you're dancing at the edge of the param size limits.
192 If you are, and this message makes sense to you, then: raise the
193 configured limit or use OpenSSL 1.1. Nothing we can do for older
196 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
197 cmdline element to get the documented default one. Previously it was
200 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
201 are unsuable for non-message-specific reasons - eg. connection timeout,
202 banner-time rejection.
204 JH/16 Fix logging of delivery remote port, when specified by router, under
207 PP/06 Repair manualroute's ability to take options in any order, even if one
208 is the name of a transport.
211 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
213 JH/17 Change the list-building routines interface to use the expanding-string
214 triplet model, for better allocation and copying behaviour.
216 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
217 Previously it was constructed the first time a possibly-matching string
218 was met in the configuration file input during startup; now it is done
221 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
222 compatible one, to avoid the (poorly documented) possibility of a config
223 file in the working directory redirecting the DB files, possibly correpting
224 some existing file. CVE-2017-10140 assigned for BDB.
226 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
227 cache-hot. Previously, although the result was properly cached, the
228 initial verify call returned a defer.
230 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
231 the main verify for receipient in uncached-mode.
233 JH/22 Retire historical build files to an "unsupported" subdir. These are
234 defined as "ones for which we have no current evidence of testing".
236 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
237 if present. Previously it was ignored.
239 JH/24 Start using specified-initialisers in C structure init coding. This is
240 a C99 feature (it's 2017, so now considered safe).
242 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
243 if was a fixed-sized field and bitmask ops via macros; it is now more
246 PP/07 GitHub PR 56: Apply MariaDB build fix.
247 Patch provided by Jaroslav Škarvada.
249 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
250 during Coverity cleanups [4.87 JH/47]
251 Diagnosis and fix provided by Michael Fischer v. Mollard.
253 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
254 the right size to place the terminating semicolon on its own folded
255 line, the header hash was calculated to an incorrect value thanks to
256 the (relaxed) space the fold became.
258 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
259 and confused the parent.
261 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
262 which could crash as a result. This could lead to undeliverable messages.
264 JH/28 Logging: "next input sent too soon" now shows where input was truncated
267 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
268 matters on fast-turnover and PID-randomising systems, which were getting
269 out-of-order delivery.
271 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
272 a possibly-overlapping copy. The symptom was that "Remote host closed
273 connection in response to HELO" was logged instead of the actual 4xx
276 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
277 Previously only that bufferd was discarded, resulting in SYMTP command
280 JH/32 DKIM: when a message has multiple signatures matching an identity given
281 in dkim_verify_signers, run the dkim acl once for each. Previously only
282 one run was done. Bug 2189.
284 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
285 "panic the current process" to "deliberately defer". The panic log is
286 still written with the problem list name; the mail and reject logs now
287 get a temp-reject line for the message that was being handled, saying
288 something like "domains check lookup or other defer". The SMTP 451
289 message is still "Temporary local problem".
291 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
292 A crafted sequence of BDAT commands could result in in-use memory beeing
293 freed. CVE-2017-16943.
295 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
296 from SMTP input. Previously it was always done; now only done for DATA
297 and not BDAT commands. CVE-2017-16944.
299 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
300 to the message (such as an overlong header line). Previously this was
301 not done and we did not exit BDAT mode. Followon from the previous item
302 though a different problem.
308 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
309 than -2003 did; needs libidn2 in addition to libidn.
311 JH/02 The path option on a pipe transport is now expanded before use.
313 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
314 Patch provided by "Björn", documentation fix added too.
316 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
317 missing a wire-to-host endian conversion.
319 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
320 close after a BDAT command line could be taken as a following command,
321 giving a synch failure. Fix by only checking for synch immediately
322 before acknowledging the chunk.
324 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
325 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
326 macro. Patches provided by Josh Soref.
328 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
329 Previously we did not; the RFC seems ambiguous and VRFY is not listed
330 by IANA as a service extension. However, John Klensin suggests that we
333 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
334 the dkim code may be unix-mode line endings rather than smtp wire-format
335 CRLF, so prepend a CR to any bare LF.
337 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
338 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
340 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
341 the first were themselves being wrongly included in the feed into dkim
342 processing; with most chunk sizes in use this resulted in an incorrect
343 body hash calculated value.
345 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
346 DKIM signature block, for verification. Although advised against by
347 standards it is specifically not ruled illegal.
349 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
351 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
352 missing a body hash (the bh= tag).
354 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
355 It seems that HAProxy sends the Proxy Protocol information in clear and
356 only then does a TLS startup, so do the same.
358 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
359 TCP connections (such as for Spamd) unless the daemon successfully set
360 Fast Open mode on its listening sockets. This fixes breakage seen on
361 too-old kernels or those not configured for Fast Open, at the cost of
362 requiring both directions being enabled for TFO, and TFO never being used
363 by non-daemon-related Exim processes.
365 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
366 endings, at least on the first header line. Try to canonify any that get
367 past that check, despite the cost.
369 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
370 now limited to an arbitrary five deep, while parsing addresses with the
371 strip_excess_angle_brackets option enabled.
373 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
374 instead leave the unprompted TLS handshake in socket buffer for the
375 TLS library to consume.
377 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
379 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
381 JH/16 Drop variables when they go out of scope. Memory management drops a whole
382 region in one operation, for speed, and this leaves assigned pointers
383 dangling. Add checks run only under the testsuite which checks all
384 variables at a store-reset and panics on a dangling pointer; add code
385 explicitly nulling out all the variables discovered. Fixes one known
386 bug: a transport crash, where a dangling pointer for $sending_ip_address
387 originally assigned in a verify callout, is re-used.
389 PP/06 Drop '.' from @INC in various Perl scripts.
391 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
393 PP/08 Reduce a number of compilation warnings under clang; building with
394 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
395 should be warning-free.
397 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
399 HS/01 Fix portability problems introduced by PP/08 for platforms where
400 realloc(NULL) is not equivalent to malloc() [SunOS et al].
402 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
403 chunk. This allows us to accept broken chunked messages. We need a more
404 general solution here.
406 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
407 already-broken messages in the queue.
409 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
411 JH/19 Fix reference counting bug in routing-generated-address tracking.
417 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
418 supports it and a size is available (ie. the sending peer gave us one).
420 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
421 years of being deprecated). The replacements are the ACLs
422 acl_smtp_mime and acl_not_smtp_mime.
424 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
425 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
426 or even an in-clear connection were permitted. Now, if the host lookup
427 was dnssec and dane was requested then the host is only used if the
428 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
429 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
430 if one fails this test.
431 This means that a poorly-configured remote DNS will make it incommunicado;
432 but it protects against a DNS-interception attack on it.
434 JH/04 Bug 1810: make continued-use of an open smtp transport connection
435 non-noisy when a race steals the message being considered.
437 JH/05 If main configuration option tls_certificate is unset, generate a
438 self-signed certificate for inbound TLS connections.
440 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
441 in rewrites and routers.
443 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
444 and logged a warning sing 4.83; now they are a configuration file error.
446 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
447 (lacking @domain). Apply the same qualification processing as RCPT.
449 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
451 JH/10 Support ${sha256:} applied to a string (as well as the previous
454 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
455 a cutthrough deliver is pending, as we always want to make a connection.
456 This also avoids re-routing the message when later placing the cutthrough
457 connection after a verify cache hit.
458 Do not update it with the verify result either.
460 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
461 when routing results in more than one destination address.
463 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
464 signing (which inhibits the cutthrough capability). Previously only
465 the presence of an option was tested; now an expansion evaluating as
466 empty is permissible (obviously it should depend only on data available
467 when the cutthrough connection is made).
469 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
470 the relevant preceding SMTP command did not note the pipelining mode.
472 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
473 Previously they were not counted.
475 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
476 as one having no matching records. Previously we deferred the message
477 that needed the lookup.
479 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
480 distinguished as "(=".
482 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
483 for missing MX records. Previously it only worked for missing A records.
485 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
487 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
488 after the data-go-ahead and data-ack. Patch from Jason Betts.
490 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
491 even for a "none" policy. Patch from Tony Meyer.
493 JH/22 Fix continued use of a connection for further deliveries. If a port was
494 specified by a router, it must also match for the delivery to be
497 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
498 When one of the recipients of a message was unsuitable for the connection
499 (has no matching addresses), we lost track of needing to mark it
500 deferred. As a result mail would be lost.
502 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
504 JH/25 Decoding ACL controls is now done using a binary search; the source code
505 takes up less space and should be simpler to maintain. Merge the ACL
506 condition decode tables also, with similar effect.
508 JH/26 Fix problem with one_time used on a redirect router which returned the
509 parent address unchanged. A retry would see the parent address marked as
510 delivered, so not attempt the (identical) child. As a result mail would
513 JH/27 Fix a possible security hole, wherein a process operating with the Exim
514 UID can gain a root shell. Credit to http://www.halfdog.net/ for
515 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
518 JH/28 Enable {spool,log} filesystem space and inode checks as default.
519 Main config options check_{log,spool}_{inodes,space} are now
520 100 inodes, 10MB unless set otherwise in the configuration.
522 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
523 Previously it only applied to the main-section connection policy
526 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
528 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
529 by me. Added RFC7919 DH primes as an alternative.
531 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
532 are not in the system include path.
534 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
535 GnuTLS, when a session startup failed (eg because the client disconnected)
536 Exim did stdio operations after fclose. This was exposed by a recent
537 change which nulled out the file handle after the fclose.
539 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
540 signed directly by the cert-signing cert, rather than an intermediate
541 OCSP-signing cert. This is the model used by LetsEncrypt.
543 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
545 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
546 an incoming connection.
548 HS/02 Bug 1802: Do not half-close the connection after sending a request
551 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
552 fallback to "prime256v1".
554 JH/34 SECURITY: Use proper copy of DATA command in error message.
555 Could leak key material. Remotely exploitable. CVE-2016-9963.
561 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
562 and 3.4.4 - once the server is enabled to respond to an OCSP request
563 it does even when not requested, resulting in a stapling non-aware
564 client dropping the TLS connection.
566 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
567 support variable-length bit vectors. No functional change.
569 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
570 The I= interface field on outgoing lines is now after the H= remote
571 host field, same as incoming lines. There is a separate
572 outgoing_interface log selector which allows you to disable the
575 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
576 If not running log_selector +smtp_connection the mainlog would be held
577 open indefinitely after a "too many connections" event, including to a
578 deleted file after a log rotate. Leave the per net connection logging
579 leaving it open for efficiency as that will be quickly detected by the
580 check on the next write.
582 HS/01 Bug 1671: Fix post transport crash.
583 Processing the wait-<transport> messages could crash the delivery
584 process if the message IDs didn't exist for some reason. When
585 using 'split_spool_directory=yes' the construction of the spool
586 file name failed already, exposing the same netto behaviour.
588 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
589 mime_regex ACL conditions.
591 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
592 to DSN fail messages (bounces): remote IP, remote greeting, remote response
593 to HELO, local diagnostic string.
595 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
596 log line to debug. Even when configured with a tls authenticator many
597 client connections are expected to not authenticate in this way, so
598 an authenticate fail is not an error.
600 HS/02 Add the Exim version string to the process info. This way exiwhat
601 gives some more detail about the running daemon.
603 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
604 matter for fast-change records such as DNSBLs.
606 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
607 retry record, even if constant. There may be multiple transports with
608 different interface settings and the retry behaviour needs to be kept
611 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
613 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
615 JH/10 Bug 840: fix log_defer_output option of pipe transport
617 JH/11 Bug 830: use same host for all RCPTS of a message, even under
618 hosts_randomize. This matters a lot when combined with mua_wrapper.
620 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
621 ${quote_pgsql:<string>} operator.
623 JH/13 Bug 1708: avoid misaligned access in cached lookup.
625 JH/14 Change header file name for freeradius-client. Relevant if compiling
626 with Radius support; from the Gentoo tree and checked under Fedora.
628 JH/15 Bug 1712: Introduce $prdr_requested flag variable
630 JH/16 Bug 1714: Permit an empty string as expansion result for transport
631 option transport_filter, meaning no filtering.
633 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
635 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
636 defaults to "*" (all hosts). The variable is now available when not built
637 with TLS, default unset, mainly to enable keeping the testsuite sane.
638 If a server certificate is not supplied (via tls_certificate) an error is
639 logged, and clients will find TLS connections fail on startup. Presumably
640 they will retry in-clear.
641 Packagers of Exim are strongly encouraged to create a server certificate
642 at installation time.
644 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
645 with the $config_file variable.
647 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
648 in transport context, after the attempt, and per-recipient. The latter type
649 is per host attempted. The event data is the error message, and the errno
650 information encodes the lookup type (A vs. MX) used for the (first) host,
651 and the trailing two digits of the smtp 4xx response.
653 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
654 to write to mainlog (or rejectlog, paniclog) in the window between file
655 creation and permissions/ownership being changed. Particularly affects
656 installations where exicyclog is run as root, rather than exim user;
657 result is that the running daemon panics and dies.
659 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
661 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
662 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
663 "pri" and "weight". Note that the previous implicit priority given by the
664 list order is no longer honoured.
666 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
669 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
670 by defining SUPPORT_SOCKS.
672 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
673 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
674 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
675 variables are renamed to proxy_{local,external}_{address,port}.
677 JH/27 Move Internationalisation support from Experimental to mainline, enabled
678 for a build by defining SUPPORT_I18N
680 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
681 of the query string, and make ${quote_redis:} do that quoting.
683 JH/29 Move Events support from Experimental to mainline, enabled by default
684 and removable for a build by defining DISABLE_EVENT.
686 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
688 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
689 cached by the daemon.
691 JH/32 Move Redis support from Experimental to mainline, enabled for a build
692 by defining LOOKUP_REDIS. The libhiredis library is required.
694 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
695 keys are given for lookup.
697 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
698 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
699 only supported when built with TLS support. The PolarSSL SHA routines
700 are still used when the TLS library is too old for convenient support.
702 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
703 openssl_options), for security. OpenSSL forces this from version 1.1.0
704 server-side so match that on older versions.
706 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
707 allocation for $value could be released as the expansion processing
708 concluded, but leaving the global pointer active for it.
710 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
711 and to use the domains and local_parts ACL conditions.
713 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
714 incorrectly not doubled on cutthrough transmission, hence seen as a
715 body-termination at the receiving system - resulting in truncated mails.
716 Commonly the sender saw a TCP-level error, and retransmitted the message
717 via the normal store-and-forward channel. This could result in duplicates
718 received - but deduplicating mailstores were liable to retain only the
719 initial truncated version.
721 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
723 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
725 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
726 we're in there, support oversigning also; bug 1309.
728 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
730 HS/04 Add support for keep_environment and add_environment options.
732 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
733 either intentional arithmetic overflow during PRNG, or testing config-
736 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
737 delivery resulted in actual delivery. Cancel cutthrough before DATA
740 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
741 rejects a recipient: pass the reject to the originator.
743 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
744 Many were false-positives and ignorable, but it's worth fixing the
747 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
748 for the new environment-manipulation done at startup. Move the routines
749 from being local to tls.c to being global via the os.c file.
751 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
752 an extract embedded as result-arg for a map, the first arg for extract
753 is unavailable so we cannot tell if this is a numbered or keyed
754 extraction. Accept either.
760 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
763 JH/02 The smtp transport option "multi_domain" is now expanded.
765 JH/03 The smtp transport now requests PRDR by default, if the server offers
768 JH/04 Certificate name checking on server certificates, when exim is a client,
769 is now done by default. The transport option tls_verify_cert_hostnames
770 can be used to disable this per-host. The build option
771 EXPERIMENTAL_CERTNAMES is withdrawn.
773 JH/05 The value of the tls_verify_certificates smtp transport and main options
774 default to the word "system" to access the system default CA bundle.
775 For GnuTLS, only version 3.0.20 or later.
777 JH/06 Verification of the server certificate for a TLS connection is now tried
778 (but not required) by default. The verification status is now logged by
779 default, for both outbound TLS and client-certificate supplying inbound
782 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
785 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
786 Status Notification (bounce) messages are now MIME format per RFC 3464.
787 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
788 under the control of the dsn_advertise_hosts option, and routers may
789 have a dsn_lasthop option.
791 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
792 default, modifiable by a malware= option. The list separator for
793 the options can now be changed in the usual way. Bug 68.
795 JH/10 The smtp_receive_timeout main option is now expanded before use.
797 JH/11 The incoming_interface log option now also enables logging of the
798 local interface on delivery outgoing connections.
800 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
801 if the interface and destination host and port all match.
803 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
806 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
807 Patch from Andrew Lewis.
809 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
810 now supports optional time-restrictions, weighting, and priority
811 modifiers per server. Patch originally by <rommer@active.by>.
813 JH/16 The spamd_address main option now supports a mixed list of local
814 and remote servers. Remote servers can be IPv6 addresses, and
815 specify a port-range.
817 JH/17 Bug 68: The spamd_address main option now supports an optional
818 timeout value per server.
820 JH/18 Bug 1581: Router and transport options headers_add/remove can
821 now have the list separator specified.
823 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
826 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
829 JH/21 Support for the A6 type of dns record is withdrawn.
831 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
832 rather than the verbs used.
834 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
835 from 255 to 1024 chars.
837 JH/24 Verification callouts now attempt to use TLS by default.
839 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
840 are generic router options now. The defaults didn't change.
842 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
843 Original patch from Alexander Shikoff, worked over by JH.
845 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
846 files if ZCAT_COMMAND is not executable.
848 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
850 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
852 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
853 Normally benign, it bites when the pair was led to by a CNAME;
854 modern usage is to not canonicalize the domain to a CNAME target
855 (and we were inconsistent anyway for A-only vs AAAA+A).
857 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
859 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
860 when evaluating $sender_host_dnssec.
862 JH/31 Check the HELO verification lookup for DNSSEC, adding new
863 $sender_helo_dnssec variable.
865 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
867 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
869 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
871 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
872 documented as working, but never had. Support all but $spam_report.
874 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
875 added for tls authenticator.
877 HS/03 Add perl_taintmode main config option
883 TL/01 When running the test suite, the README says that variables such as
884 no_msglog_check are global and can be placed anywhere in a specific
885 test's script, however it was observed that placement needed to be near
886 the beginning for it to behave that way. Changed the runtest perl
887 script to read through the entire script once to detect and set these
888 variables, reset to the beginning of the script, and then run through
889 the script parsing/test process like normal.
891 TL/02 The BSD's have an arc4random API. One of the functions to induce
892 adding randomness was arc4random_stir(), but it has been removed in
893 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
894 function when detected.
896 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
897 cause callback expansion.
899 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
900 syntax errors in an expansion can be treated as a string instead of
901 logging or causing an error, due to the internal use of bool_lax
902 instead of bool when processing it.
904 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
905 server certificates when making smtp deliveries.
907 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
909 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
911 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
913 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
914 Merged patch from Sebastian Wiedenroth.
916 JH/05 Fix results-pipe from transport process. Several recipients, combined
917 with certificate use, exposed issues where response data items split
918 over buffer boundaries were not parsed properly. This eventually
919 resulted in duplicates being sent. This issue only became common enough
920 to notice due to the introduction of connection certificate information,
921 the item size being so much larger. Found and fixed by Wolfgang Breyha.
923 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
924 size buffer was used, resulting in syntax errors when an expansion
927 JH/07 Add support for directories of certificates when compiled with a GnuTLS
928 version 3.3.6 or later.
930 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
931 is EXPERIMENTAL_EVENT, the main-configuration and transport options
932 both become "event_action", the variables become $event_name, $event_data
933 and $event_defer_errno. There is a new variable $verify_mode, usable in
934 routers, transports and related events. The tls:cert event is now also
935 raised for inbound connections, if the main configuration event_action
938 TL/06 In test suite, disable OCSP for old versions of openssl which contained
939 early OCSP support, but no stapling (appears to be less than 1.0.0).
941 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
942 server certificate names available under the smtp transport option
943 "tls_verify_cert_hostname" now do not permit multi-component wildcard
946 JH/10 Time-related extraction expansions from certificates now use the main
947 option "timezone" setting for output formatting, and are consistent
948 between OpenSSL and GnuTLS compilations. Bug 1541.
950 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
951 encoded parameter in the incoming message. Bug 1558.
953 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
954 include certificate info, eximon was claiming there were spoolfile
957 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
959 JH/14 Log delivery-related information more consistently, using the sequence
960 "H=<name> [<ip>]" wherever possible.
962 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
963 are problematic for Debian distribution, omit them from the release
966 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
968 JH/16 Fix string representation of time values on 64bit time_t architectures.
971 JH/17 Fix a null-indirection in certextract expansions when a nondefault
972 output list separator was used.
977 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
978 checkers that were complaining about end of non-void function with no
981 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
982 This was a regression introduced in 4.83 by another bugfix.
984 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
986 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
987 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
993 TF/01 Correctly close the server side of TLS when forking for delivery.
995 When a message was received over SMTP with TLS, Exim failed to clear up
996 the incoming connection properly after forking off the child process to
997 deliver the message. In some situations the subsequent outgoing
998 delivery connection happened to have the same fd number as the incoming
999 connection previously had. Exim would try to use TLS and fail, logging
1000 a "Bad file descriptor" error.
1002 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1003 utilities have not been installed.
1005 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1006 temporary space as the ACL may create new global variables.
1008 TL/01 LDAP support uses per connection or global context settings, depending
1009 upon the detected version of the libraries at build time.
1011 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1012 to extract and use the src ip:port in logging and expansions as if it
1013 were a direct connection from the outside internet. PPv2 support was
1014 updated based on HAProxy spec change in May 2014.
1016 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1018 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1019 Properly escape header and check for NULL return.
1021 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1024 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1026 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1027 characters in header names, implemented as a verify condition.
1028 Contributed by Michael Fischer v. Mollard.
1030 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1031 results permerror and temperror. Previous values are deprecated but
1032 still accepted. In a future release, err_perm and err_temp will be
1033 completely removed, which will be a backward incompatibility if the
1034 ACL tests for either of these two old results. Patch contributed by
1035 user bes-internal on the mailing list.
1037 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1039 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1040 selectors, in both main and reject logs.
1042 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1045 JH/07 Add malware type "sock" for talking to simple daemon.
1047 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1049 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1050 routers/transports under cutthrough routing.
1052 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1053 numbers. Touch up "bool" conditional to keep the same definition.
1055 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1057 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1059 JH/12 Expand items in router/transport headers_add or headers_remove lists
1060 individually rather than the list as a whole. Bug 1452.
1062 Required for reasonable handling of multiple headers_ options when
1063 they may be empty; requires that headers_remove items with embedded
1064 colons must have them doubled (or the list-separator changed).
1066 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1067 view the policy declared in the DMARC record. Currently, $dmarc_status
1068 is a combined value of both the record presence and the result of the
1071 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1073 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1074 dnslookup router and the smtp transport (applying to the forward
1077 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1078 of ldap servers used for a specific lookup. Patch provided by Heiko
1081 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1082 New variable $lookup_dnssec_authenticated for observability.
1084 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1085 Patch submitted by Lars Timman.
1087 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1089 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1090 Requires trusted mode and valid format message id, aborts otherwise.
1091 Patch contributed by Heiko Schlichting.
1093 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1094 certextract with support for various fields. Bug 1358.
1096 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1097 is requested by default, modifiable by smtp transport option
1100 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1101 operate on certificate variables to give certificate fingerprints
1102 Also new ${sha256:cert_variable}.
1104 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1106 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1107 Christian Aistleitner.
1109 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1111 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1112 file. Patch from Wolfgang Breyha.
1114 JH/25 Expand the coverage of the delivery $host and $host_address to
1115 client authenticators run in verify callout. Bug 1476.
1117 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1118 align with daemon_smtp_ports. Bug 72.
1120 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1121 support and error reporting did not work properly.
1123 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1124 and is readable. Patch from Andrew Colin Kissa.
1126 TL/14 Enhance documentation of ${run expansion and how it parses the
1127 commandline after expansion, particularly in the case when an
1128 unquoted variable expansion results in an empty value.
1130 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1132 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1133 by renaming away from it. Addresses GH issue 15, reported by
1136 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1138 TL/15 SECURITY: prevent double expansion in math comparison functions
1139 (can expand unsanitized data). Not remotely exploitable.
1146 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1148 PP/02 Make -n do something, by making it not do something.
1149 When combined with -bP, the name of an option is not output.
1151 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1154 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1155 $sender_host_name and config options to manage this, and basic check
1158 PP/05 DSCP support for outbound connections and control modifier for inbound.
1160 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1161 (Only plugin which currently uses this is kerberos4, which nobody should
1162 be using, but we should make it available and other future plugins might
1163 conceivably use it, even though it would break NAT; stuff *should* be
1164 using channel bindings instead).
1166 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1167 name; added for Sendmail compatibility; requires admin caller.
1168 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1169 just ignore it); requires trusted caller.
1170 Also parse but ignore: -Ac -Am -X<logfile>
1173 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1175 TL/02 Add +smtp_confirmation as a default logging option.
1177 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1178 Patch by Magnus Holmgren from 2007-02-20.
1180 TL/04 Bugzilla 1281 - Spec typo.
1181 Bugzilla 1283 - Spec typo.
1182 Bugzilla 1290 - Spec grammar fixes.
1184 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1186 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1188 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1189 mailing list by by Dmitry Isaikin.
1191 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1193 JH/02 Support "G" suffix to numbers in ${if comparisons.
1195 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1197 NM/01 Bugzilla 1197 - Spec typo
1198 Bugzilla 1196 - Spec examples corrections
1200 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1202 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1203 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1206 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1207 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1209 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1210 "acl {{name}{arg}...}", and optional args on acl condition
1213 JH/05 Permit multiple router/transport headers_add/remove lines.
1215 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1217 JH/07 Avoid using a waiting database for a single-message-only transport.
1218 Performance patch from Paul Fisher. Bugzilla 1262.
1220 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1223 JH/09 Add $headers_added variable, with content from use of ACL modifier
1224 add_header (but not yet added to the message). Bugzilla 199.
1226 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1227 Pulled from Bugzilla 817 by Wolfgang Breyha.
1229 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1231 (nb: this is the same fix as in Exim 4.80.1)
1233 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1236 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1239 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1241 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1242 advertise SMTP AUTH mechanism to us, instead of a generic
1243 protocol violation error. Also, make Exim more robust to bad
1244 data from the Dovecot auth socket.
1246 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1248 When a queue runner is handling a message, Exim first routes the
1249 recipient addresses, during which it prunes them based on the retry
1250 hints database. After that it attempts to deliver the message to
1251 any remaining recipients. It then updates the hints database using
1254 So if a recipient address works intermittently, it can get repeatedly
1255 deferred at routing time. The retry hints record remains fresh so the
1256 address never reaches the final cutoff time.
1258 This is a fairly common occurrence when a user is bumping up against
1259 their storage quota. Exim had some logic in its local delivery code
1260 to deal with this. However it did not apply to per-recipient defers
1261 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1263 This change adds a proper retry rule check during routing so that the
1264 final cutoff time is checked against the message's age. We only do
1265 this check if there is an address retry record and there is not a
1266 domain retry record; this implies that previous attempts to handle
1267 the address had the retry_use_local_parts option turned on. We use
1268 this as an approximation for the destination being like a local
1269 delivery, as in LMTP.
1271 I suspect this new check makes the old local delivery cutoff check
1272 redundant, but I have not verified this so I left the code in place.
1274 TF/02 Correct gecos expansion when From: is a prefix of the username.
1276 Test 0254 submits a message to Exim with the header
1280 When I ran the test suite under the user fanf2, Exim expanded
1281 the header to contain my full name, whereas it should have added
1282 a Resent-Sender: header. It erroneously treats any prefix of the
1283 username as equal to the username.
1285 This change corrects that bug.
1287 GF/01 DCC debug and logging tidyup
1288 Error conditions log to paniclog rather than rejectlog.
1289 Debug lines prefixed by "DCC: " to remove any ambiguity.
1291 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1293 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1294 Bug spotted by Jeremy Harris; was flawed since initial commit.
1295 Would have resulted in OCSP responses post-SNI triggering an Exim
1296 NULL dereference and crash.
1298 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1300 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1301 Bug detection, analysis and fix by Samuel Thibault.
1302 Bugzilla 1331, Debian bug #698092.
1304 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1306 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1307 Server implementation by Todd Lyons, client by JH.
1308 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1309 config variable "prdr_enable" controls whether the server
1310 advertises the facility. If the client requests PRDR a new
1311 acl_data_smtp_prdr ACL is called once for each recipient, after
1312 the body content is received and before the acl_smtp_data ACL.
1313 The client is controlled by both of: a hosts_try_prdr option
1314 on the smtp transport, and the server advertisement.
1315 Default client logging of deliveries and rejections involving
1316 PRDR are flagged with the string "PRDR".
1318 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1319 fclose(). Diagnosis by Todd Lyons.
1321 PP/17 Update configure.default to handle IPv6 localhost better.
1322 Patch by Alain Williams (plus minor tweaks).
1325 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1326 This is now consistent with GnuTLS, and is now documented: the
1327 previous undocumented portable approach to treating the option as
1328 unset was to force an expansion failure. That still works, and
1329 an empty string is now equivalent.
1331 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1332 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1333 not performing validation itself.
1335 PP/20 Added force_command boolean option to pipe transport.
1336 Patch from Nick Koston, of cPanel Inc.
1338 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1341 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1343 PP/21 Fix eximon continuous updating with timestamped log-files.
1344 Broken in a format-string cleanup in 4.80, missed when I repaired the
1345 other false fix of the same issue.
1346 Report and fix from Heiko Schlichting.
1349 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1350 Report from Prashanth Katuri.
1352 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1353 It's SecureTransport, so affects any MacOS clients which use the
1354 system-integrated TLS libraries, including email clients.
1356 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1357 using a MIME ACL for non-SMTP local injection.
1358 Report and assistance in diagnosis by Warren Baker.
1360 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1362 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1364 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1365 last id that failed so it may be referenced in subsequent ACL's.
1367 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1370 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1371 ldap library initialization, allowing self-signed CA's to be
1372 used. Also properly sets require_cert option later in code by
1373 using NULL (global ldap config) instead of ldap handle (per
1374 session). Bug diagnosis and testing by alxgomz.
1376 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1377 the src/util/ subdirectory.
1379 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1380 renamed to Transport Post Delivery Action by Jeremy Harris, as
1383 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1384 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1385 redis_servers = needs to be configured which will be used by the redis
1386 lookup. Patch from Warren Baker, of The Packet Hub.
1388 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1390 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1391 hostname or reverse DNS when processing a host list. Used suggestions
1392 from multiple comments on this bug.
1394 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1396 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1397 Missed a few lines, added it to make the runtest require no keyboard
1400 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1401 contains upper case chars. Make router use caseful_local_part.
1403 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1404 support when GnuTLS has been built with p11-kit.
1410 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1412 This, or similar/improved, will also be change PP/11 of 4.82.
1418 PP/01 Handle short writes when writing local log-files.
1419 In practice, only affects FreeBSD (8 onwards).
1420 Bugzilla 1053, with thanks to Dmitry Isaikin.
1422 NM/01 Bugzilla 949 - Documentation tweak
1424 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1427 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1429 PP/02 Implemented gsasl authenticator.
1431 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1433 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1434 `pkg-config foo` for cflags/libs.
1436 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1437 with rest of GSASL and with heimdal_gssapi.
1439 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1440 `pkg-config foo` for cflags/libs for the TLS implementation.
1442 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1443 properties get this fed in as external SSF. A number of robustness
1444 and debugging improvements to the cyrus_sasl authenticator.
1446 PP/08 cyrus_sasl server now expands the server_realm option.
1448 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1449 Patch by Jeremy Harris.
1451 PP/10 Added dbmjz lookup type.
1453 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1455 PP/12 MAIL args handles TAB as well as SP, for better interop with
1456 non-compliant senders.
1457 Analysis and variant patch by Todd Lyons.
1459 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1460 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1461 Patch from Dirk Mueller <dmueller@suse.com>
1463 PP/13 tls_peerdn now print-escaped for spool files.
1464 Observed some $tls_peerdn in wild which contained \n, which resulted
1465 in spool file corruption.
1467 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1468 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1469 or write after TLS renegotiation, which otherwise led to messages
1472 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1473 as a tracking header (ie: a signed header comes before the signature).
1474 Patch from Wolfgang Breyha.
1476 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1477 comma-sep list; embedded commas doubled.
1479 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1481 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1483 Report and patch from Dmitry Banschikov.
1485 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1486 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1487 failures. We appear to now support TLS1.1+ with Exim.
1489 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1490 lets Exim select keys and certificates based upon TLS SNI from client.
1491 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1492 before an outbound SMTP session. New log_selector, +tls_sni.
1494 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1495 NULL dereference. Report and patch from Alun Jones.
1497 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1498 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1499 Not seeing resolver debug output on NetBSD, but suspect this is a
1500 resolver implementation change.
1502 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1503 Left warnings. Added "eximon gdb" invocation mode.
1505 PP/21 Defaulting "accept_8bitmime" to true, not false.
1507 PP/22 Added -bw for inetd wait mode support.
1509 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1510 locate the relevant includes and libraries. Made this the default.
1512 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1513 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1515 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1516 This may cause build issues on older platforms.
1518 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1519 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1520 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1521 Added SNI support via GnuTLS too.
1522 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1524 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1526 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1527 Applied second patch from Janne, implementing suggestion to default
1528 multiple-strings-in-record handling to match SPF spec.
1530 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1532 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1533 read-only, out of scope).
1534 Patch from Wolfgang Breyha, report from Stuart Northfield.
1536 PP/29 Fix three issues highlighted by clang analyser static analysis.
1537 Only crash-plausible issue would require the Cambridge-specific
1538 iplookup router and a misconfiguration.
1539 Report from Marcin Mirosław.
1541 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1543 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1544 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1545 As part of this, removing so much warning spew let me fix some minor
1546 real issues in debug logging.
1548 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1549 assignment on my part. Fixed.
1551 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1552 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1553 Janne Snabb (who went above and beyond: thank you).
1555 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1556 string otherwise requires a connection and a bunch more work and it's
1557 relatively easy to get wrong. Should also expose TLS library linkage
1560 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1561 64-bit ${eval} (JH/03).
1563 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1564 GNU libc to support some of the 64-bit stuff, should not lead to
1565 conflicts. Defined before os.h is pulled in, so if a given platform
1566 needs to override this, it can.
1568 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1569 protection layer was required, which is not implemented.
1570 Bugzilla 1254, patch from Wolfgang Breyha.
1572 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1573 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1574 tls_dhparam take prime identifiers. Also unbreak combination of
1575 OpenSSL+DH_params+TLSSNI.
1577 PP/39 Disable SSLv2 by default in OpenSSL support.
1583 PP/01 Solaris build fix for Oracle's LDAP libraries.
1584 Bugzilla 1109, patch from Stephen Usher.
1586 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1588 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1591 TF/02 Fix a couple more cases where we did not log the error message
1592 when unlink() failed. See also change 4.74-TF/03.
1594 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1595 lock up or crash if it happened to be inside a call to libc when it
1596 got a SIGUSR1 from exiwhat.
1598 The SIGUSR1 handler appends the current process status to the process
1599 log which is later printed by exiwhat. It used to use the general
1600 purpose logging code to do this, but several functions it calls are
1601 not safe for signals.
1603 The new output code in the SIGUSR1 handler is specific to the process
1604 log, and simple enough that it's easy to inspect for signal safety.
1605 Removing some special cases also simplifies the general logging code.
1606 Removing the spurious timestamps from the process log simplifies
1609 TF/04 Improved ratelimit ACL condition.
1611 The /noupdate option has been deprecated in favour of /readonly which
1612 has clearer semantics. The /leaky, /strict, and /readonly update modes
1613 are mutually exclusive. The update mode is no longer included in the
1614 database key; it just determines when the database is updated. (This
1615 means that when you upgrade Exim will forget old rate measurements.)
1617 Exim now checks that the per_* options are used with an update mode that
1618 makes sense for the current ACL. For example, when Exim is processing a
1619 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1620 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1621 must specify per_mail/readonly. If you omit the update mode it defaults to
1622 /leaky where that makes sense (as before) or /readonly where required.
1624 The /noupdate option is now undocumented but still supported for
1625 backwards compatibility. It is equivalent to /readonly except that in
1626 ACLs where /readonly is required you may specify /leaky/noupdate or
1627 /strict/noupdate which are treated the same as /readonly.
1629 A useful new feature is the /count= option. This is a generalization
1630 of the per_byte option, so that you can measure the throughput of other
1631 aggregate values. For example, the per_byte option is now equivalent
1632 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1634 The per_rcpt option has been generalized using the /count= mechanism
1635 (though it's more complicated than the per_byte equivalence). When it is
1636 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1637 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1638 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1639 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1640 Note that using per_rcpt with a non-readonly update mode in more than
1641 one ACL will cause the recipients to be double-counted. (The per_mail
1642 and per_byte options don't have this problem.)
1644 The handling of very low rates has changed slightly. If the computed rate
1645 is less than the event's count (usually one) then this event is the first
1646 after a long gap. In this case the rate is set to the same as this event's
1647 count, so that the first message of a spam run is counted properly.
1649 The major new feature is a mechanism for counting the rate of unique
1650 events. The new per_addr option counts the number of different
1651 recipients that someone has sent messages to in the last time period. It
1652 behaves like per_rcpt if all the recipient addresses are different, but
1653 duplicate recipient addresses do not increase the measured rate. Like
1654 the /count= option this is a general mechanism, so the per_addr option
1655 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1656 example, measure the rate that a client uses different sender addresses
1657 with the options per_mail/unique=$sender_address. There are further
1658 details in the main documentation.
1660 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1662 TF/06 Removed a few PCRE remnants.
1664 TF/07 Automatically extract Exim's version number from tags in the git
1665 repository when doing development or release builds.
1667 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1668 Bugzilla 879. Patch from Paul Fisher.
1670 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1671 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1674 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1676 PP/05 Variable $av_failed, true if the AV scanner deferred.
1677 Bugzilla 1078. Patch from John Horne.
1679 PP/06 Stop make process more reliably on build failure.
1680 Bugzilla 1087. Patch from Heiko Schlittermann.
1682 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1683 Bugzilla 1089. Patch from Heiko Schlittermann.
1685 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1686 Bugzilla 1131. Patch from Holger Weiß.
1688 PP/09 Handle IPv6 addresses with SPF.
1689 Bugzilla 860. Patch from Wolfgang Breyha.
1691 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1693 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1696 PP/11 match_* no longer expand right-hand-side by default.
1697 New compile-time build option, EXPAND_LISTMATCH_RHS.
1698 New expansion conditions, "inlist", "inlisti".
1700 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1702 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1704 PP/14 fix log_write() format string regression from TF/03.
1705 Bugzilla 1152. Patch from Dmitry Isaikin.
1711 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1713 PP/02 Harmonised TLS library version reporting; only show if debugging.
1714 Layout now matches that introduced for other libraries in 4.74 PP/03.
1716 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1718 PP/04 New "dns_use_edns0" global option.
1720 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1723 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1724 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1726 TK/01 Updated PolarSSL code to 0.14.2.
1727 Bugzilla 1097. Patch from Andreas Metzler.
1729 PP/07 Catch divide-by-zero in ${eval:...}.
1730 Fixes bugzilla 1102.
1732 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1735 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1736 format-string attack -- SECURITY: remote arbitrary code execution.
1738 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1739 time unintentionally subject to list matching rules, letting the header
1740 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1741 arbitrary string expansion). This allowed for information disclosure.
1743 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1744 INT_MIN/-1 -- value coerced to INT_MAX.
1750 NM/01 Workaround for PCRE version dependency in version reporting
1753 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1754 This fixes portability to compilers other than gcc, notably
1755 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1757 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1758 makefiles for portability to HP-UX and POSIX correctness.
1760 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1761 Also via indented variable definition in the Makefile.
1762 (Debugging by Oliver Heesakkers).
1764 PP/02 Restore caching of spamd results with expanded spamd_address.
1765 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1767 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1768 Improves build reliability. Fix from: Frank Elsner
1770 NM/02 Fix wide character breakage in the rfc2047 coding
1771 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1773 NM/03 Allow underscore in dnslist lookups
1774 Fixes bug 1026. Patch from Graeme Fowler
1776 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1777 Code patches from Adam Ciarcinski of NetBSD.
1779 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1782 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1783 is logged, to avoid truncation. Patch from John Horne.
1785 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1786 Patch from Jakob Hirsch.
1788 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1789 SQL string expansion failure details.
1790 Patch from Andrey Oktyabrski.
1792 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1793 Patch from Simon Arlott.
1795 PP/09 New lookups functionality failed to compile on old gcc which rejects
1796 extern declarations in function scope.
1797 Patch from Oliver Fleischmann
1799 PP/10 Use sig_atomic_t for flags set from signal handlers.
1800 Check getgroups() return and improve debugging.
1801 Fixed developed for diagnosis in bug 927 (which turned out to be
1804 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1805 Patch from Mark Zealey.
1807 PP/12 Bugzilla 1056: Improved spamd server selection.
1808 Patch from Mark Zealey.
1810 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1811 Based on patch from Heiko Schlittermann.
1813 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1814 Patch from Uwe Doering, sign-off by Michael Haardt.
1816 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1817 variable declaration deep within a block. Bug and patch from
1820 PP/15 lookups-Makefile IRIX compatibility coercion.
1822 PP/16 Make DISABLE_DKIM build knob functional.
1824 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1825 Patch by Simon Arlott
1827 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1828 variable argument macros. Our copy now differs from upstream.
1834 TF/01 Failure to get a lock on a hints database can have serious
1835 consequences so log it to the panic log.
1837 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1838 controlled using the smtp_confirmation log selector.
1840 TF/03 Include the error message when we fail to unlink a spool file.
1842 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1843 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1844 for maintaining out-of-tree patches for some time.
1846 PP/01 Bugzilla 139: Documentation and portability issues.
1847 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1848 Handle per-OS dynamic-module compilation flags.
1850 PP/02 Let /dev/null have normal permissions.
1851 The 4.73 fixes were a little too stringent and complained about the
1852 permissions on /dev/null. Exempt it from some checks.
1853 Reported by Andreas M. Kirchwitz.
1855 PP/03 Report version information for many libraries, including
1856 Exim version information for dynamically loaded libraries. Created
1857 version.h, now support a version extension string for distributors
1858 who patch heavily. Dynamic module ABI change.
1860 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1861 privilege escalation vulnerability whereby the Exim run-time user
1862 can cause root to append content of the attacker's choosing to
1865 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1868 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1869 If dropping privileges for untrusted macros, we disabled normal logging
1870 on the basis that it would fail; for the Exim run-time user, this is not
1871 the case, and it resulted in successful deliveries going unlogged.
1872 Fixed. Reported by Andreas Metzler.
1878 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1879 only prepend for the Resent-* case. Fixes regression introduced in
1880 Exim 4.70 by NM/22 for Bugzilla 607.
1882 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1883 increasing numbers of administrators be bitten by this.
1885 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1887 PP/03 Bugzilla 994: added openssl_options main configuration option.
1889 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1891 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1893 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1895 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1896 without a peer certificate, leading to a segfault because of an
1897 assumption that peers always have certificates. Be a little more
1898 paranoid. Problem reported by Martin Tscholak.
1900 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1901 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1902 NB: ClamAV planning to remove STREAM in "middle of 2010".
1903 CL also introduces -bmalware, various -d+acl logging additions and
1904 more caution in buffer sizes.
1906 PP/09 Implemented reverse_ip expansion operator.
1908 PP/10 Bugzilla 937: provide a "debug" ACL control.
1910 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1912 PP/12 Bugzilla 973: Implement --version.
1914 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1916 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1918 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1920 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1921 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1922 ignore trailing whitespace.
1924 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1926 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1929 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1930 Notification from Dr Andrew Aitchison.
1932 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1933 ExtendedDetectionInfo response format.
1934 Notification from John Horne.
1936 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1939 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1940 XSL and documented dependency on system catalogs, with examples of how
1943 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1946 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1947 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1948 configuration file which is writeable by the Exim user or group.
1950 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1951 of configuration files to cover files specified with the -C option if
1952 they are going to be used with root privileges, not just the default
1955 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1956 option (effectively making it always true).
1958 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1959 files to be used while preserving root privileges.
1961 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1962 that rogue child processes cannot use them.
1964 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1965 run-time user, instead of root.
1967 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1968 Exim run-time user without dropping privileges.
1970 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1971 result string, instead of calling string_vformat() twice with the same
1974 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1975 for other users. Others should always drop root privileges if they use
1976 -C on the command line, even for a whitelisted configure file.
1978 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1980 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1986 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1987 $data_path, and $header_path variables; fixed documentation bugs and
1990 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1991 exipick to access non-standard spools, including the "frozen" queue
1994 NM/01 Bugzilla 965: Support mysql stored procedures.
1995 Patch from Alain Williams
1997 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1999 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2000 Patch from Andreas Metzler
2002 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2003 Patch from Kirill Miazine
2005 NM/05 Bugzilla 671: Added umask to procmail example.
2007 JJ/03 installed exipick 20100323.0, fixing doc bug
2009 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2010 directory. Notification and patch from Dan Rosenberg.
2012 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2014 TK/02 Improve log output when DKIM signing operation fails.
2016 MH/01 Treat the transport option dkim_domain as a colon separated
2017 list, not as a single string, and sign the message with each element,
2018 omitting multiple occurences of the same signer.
2020 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2021 Bugzilla 985, 986. Patch by Simon Arlott
2023 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2024 Patch by Simon Arlott
2026 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2027 MBX locking. Notification from Dan Rosenberg.
2033 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2035 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2037 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2039 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2041 NM/04 Bugzilla 915: Flag broken perl installation during build.
2047 TK/01 Added patch by Johannes Berg that expands the main option
2048 "spamd_address" if it starts with a dollar sign.
2050 TK/02 Write list of recipients to X-Envelope-Sender header when building
2051 the mbox-format spool file for content scanning (suggested by Jakob
2054 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2055 (http://www.dcc-servers.net/) support via dccifd. Activated by
2056 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2058 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2059 by Mark Daniel Reidel <mr@df.eu>.
2061 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2062 When building exim an external PCRE library is now needed -
2063 PCRE is a system library on the majority of modern systems.
2064 See entry on PCRE_LIBS in EDITME file.
2066 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2067 conversation. Added nologin parameter to request.
2068 Patch contributed by Kirill Miazine.
2070 TF/01 Do not log submission mode rewrites if they do not change the address.
2072 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2074 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2075 log files in place. Contributed by Roberto Lima.
2077 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2079 TF/03 Bugzilla 615: When checking the local_parts router precondition
2080 after a local_part_suffix or local_part_prefix option, Exim now
2081 does not use the address's named list lookup cache, since this
2082 contains cached lookups for the whole local part.
2084 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2085 Robert Millan. Documentation is in experimental-spec.txt.
2087 TF/04 Bugzilla 668: Fix parallel build (make -j).
2089 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2091 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2092 Patch provided by Jan Srzednicki.
2094 TF/05 Leading white space used to be stripped from $spam_report which
2095 wrecked the formatting. Now it is preserved.
2097 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2098 that they are available at delivery time.
2100 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2102 TF/08 TLS error reporting now respects the incoming_interface and
2103 incoming_port log selectors.
2105 TF/09 Produce a more useful error message if an SMTP transport's hosts
2106 setting expands to an empty string.
2108 NM/06 Bugzilla 744: EXPN did not work under TLS.
2109 Patch provided by Phil Pennock.
2111 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2112 Patch provided by Richard Godbee.
2114 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2115 acl_smtp_notquit, added index entry.
2117 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2118 Patch provided by Eugene Bujak.
2120 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2121 accept(). Patch provided by Maxim Dounin.
2123 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2124 Patch provided by Phil Pennock.
2126 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2128 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2129 Patch provided by Brad "anomie" Jorsch.
2131 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2132 Patch provided by Dean Brooks.
2134 TK/05 Add native DKIM support (does not depend on external libraries).
2136 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2137 Patch provided by Graeme Fowler.
2139 NM/16 Bugzilla 851: Documentation example syntax fix.
2141 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2143 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2146 NM/19 Bugzilla 745: TLS version reporting.
2147 Patch provided by Phil Pennock.
2149 NM/20 Bugzilla 167: bool: condition support.
2150 Patch provided by Phil Pennock.
2152 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2153 clients. Patch provided by Phil Pennock.
2155 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2156 Patch provided by Brad "anomie" Jorsch.
2158 NM/23 Bugzilla 687: Fix misparses in eximstats.
2159 Patch provided by Heiko Schlittermann.
2161 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2162 Patch provided by Heiko Schlittermann.
2164 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2165 plus update to original patch.
2167 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2169 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2170 Patch provided by David Brownlee.
2172 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2174 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2176 NM/30 Bugzilla 888: TLS documentation bugfixes.
2178 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2180 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2181 Unlike the original bugzilla I have changed all shell scripts in src tree.
2183 NM/33 Bugzilla 898: Transport filter timeout fix.
2184 Patch by Todd Rinaldo.
2186 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2187 Patch by Serge Demonchaux.
2189 NM/35 Bugzilla 39: Base64 decode bug fixes.
2190 Patch by Jakob Hirsch.
2192 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2194 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2196 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2198 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2204 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2205 ALT-N's libdkim that I have put here:
2206 http://duncanthrax.net/exim-experimental/
2208 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2209 were called 'true' and it seems that C99 defines that as a reserved
2210 keyword to be used with 'bool' variable types. That means you could
2211 not include C99-style headers which use bools without triggering
2212 build errors in sieve.c.
2214 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2215 as mailq or other aliases. Changed the --help handling significantly
2216 to do whats expected. exim_usage() emits usage/help information.
2218 SC/01 Added the -bylocaldomain option to eximstats.
2220 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2222 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2224 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2230 PH/01 Another patch from the Sieve maintainer.
2232 PH/02 When an IPv6 address is converted to a string for single-key lookup
2233 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2234 dots are used instead of colons so that keys in lsearch files need not
2235 contain colons. This was done some time before quoting was made available
2236 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2237 (notated using the quote facility) so as to distinguish them from IPv4
2238 keys. This meant that lookups for IP addresses in host lists did not work
2239 for iplsearch lookups.
2241 This has been fixed by arranging for IPv6 addresses to be expressed with
2242 colons if the lookup type is iplsearch. This is not incompatible, because
2243 previously such lookups could never work.
2245 The situation is now rather anomalous, since one *can* have colons in
2246 ordinary lsearch keys. However, making the change in all cases is
2247 incompatible and would probably break a number of configurations.
2249 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2252 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2253 conversion specification without a maximum field width, thereby enabling
2254 a rogue spamd server to cause a buffer overflow. While nobody in their
2255 right mind would setup Exim to query an untrusted spamd server, an
2256 attacker that gains access to a server running spamd could potentially
2257 exploit this vulnerability to run arbitrary code as the Exim user.
2259 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2260 $primary_hostname instead of what libspf2 thinks the hosts name is.
2262 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2263 a directory entry by the name of the lookup key. Previously, if a
2264 symlink pointed to a non-existing file or a file in a directory that
2265 Exim lacked permissions to read, a lookup for a key matching that
2266 symlink would fail. Now it is enough that a matching directory entry
2267 exists, symlink or not. (Bugzilla 503.)
2269 PH/03 The body_linecount and body_zerocount variables are now exported in the
2272 PH/04 Added the $dnslist_matched variable.
2274 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2275 This means they are set thereafter only if the connection becomes
2278 PH/06 Added the client_condition to authenticators so that some can be skipped
2279 by clients under certain conditions.
2281 PH/07 The error message for a badly-placed control=no_multiline_responses left
2282 "_responses" off the end of the name.
2284 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2286 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2287 (without spaces) instead of just copying the configuration text.
2289 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2291 PH/11 Added $max_received_linelength.
2293 PH/12 Added +ignore_defer and +include_defer to host lists.
2295 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2296 way in which PCRE > 7.0 is built.
2298 PH/14 Implemented queue_only_load_latch.
2300 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2301 MAIL command. The effect was to mangle the value on 64-bit systems.
2303 PH/16 Another patch from the Sieve maintainer.
2305 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2307 PH/18 If a system quota error occurred while trying to create the file for
2308 a maildir delivery, the message "Mailbox is full" was not appended to the
2309 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2310 only to a quota excession during the actual writing of the file.
2312 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2313 characters?) which causes problems in log lines. The DN values are now
2314 passed through string_printing() before being added to log lines.
2316 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2317 and InterBase are left for another time.)
2319 PH/21 Added message_body_newlines option.
2321 PH/22 Guard against possible overflow in moan_check_errorcopy().
2323 PH/23 POSIX allows open() to be a macro; guard against that.
2325 PH/24 If the recipient of an error message contained an @ in the local part
2326 (suitably quoted, of course), incorrect values were put in $domain and
2327 $local_part during the evaluation of errors_copy.
2333 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2334 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2337 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2338 issue a MAIL command.
2340 PH/02 In an ACL statement such as
2342 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2344 if a client was not listed at all, or was listed with a value other than
2345 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2346 the condition was not true (as it should be), so access was not denied.
2347 The bug was that the ! inversion was incorrectly passed on to the second
2348 item. This has been fixed.
2350 PH/03 Added additional dnslists conditions == and =& which are different from
2351 = and & when the dns lookup returns more than one IP address.
2353 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2354 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2356 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2357 FSYNC, which compiles an option called disable_fsync that allows for
2358 bypassing fsync(). The documentation is heavily laced with warnings.
2360 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2362 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2363 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2364 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2365 including adding "make clean"; (3) Added -fPIC when compiling the test
2366 dynamically loaded module, to get rid of a warning.
2368 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2369 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2370 = 0s. The bug is otherwise harmless.
2372 PH/07 There was a bug in the dovecot authenticator such that the value of
2373 $auth1 could be overwritten, and so not correctly preserved, after a
2374 successful authentication. This usually meant that the value preserved by
2375 the server_setid option was incorrect.
2377 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2379 PH/09 Installed PCRE release 7.0.
2381 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2382 run for batched SMTP input. It is now run at the start of every message
2383 in the batch. While fixing this I discovered that the process information
2384 (output by running exiwhat) was not always getting set for -bs and -bS
2385 input. This is fixed, and it now also says "batched" for BSMTP.
2387 PH/11 Added control=no_pipelining.
2389 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2390 patch, slightly modified), and move the expansion of helo_data till after
2391 the connection is made in the smtp transport (so it can use these
2394 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2396 PH/14 Added log_selector = +pid.
2398 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2400 PH/16 Add ${if forany and ${if forall.
2402 PH/17 Added dsn_from option to vary the From: line in DSNs.
2404 PH/18 Flush SMTP output before performing a callout, unless control =
2405 no_callout_flush is set.
2407 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2408 was true (the default) a successful delivery failed to delete the retry
2409 item, thus causing premature timeout of the address. The bug is now
2412 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2414 PH/21 Long custom messages for fakedefer and fakereject are now split up
2415 into multiline responses in the same way that messages for "deny" and
2416 other ACL rejections are.
2418 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2419 with slight modification.
2421 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2422 draft, changing the syntax and factoring some duplicate code".
2424 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2425 for deliveries of the second and subsequent messages over the same SMTP
2428 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2429 ${reduce, with only minor "tidies".
2431 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2433 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2434 expansion side effects.
2436 PH/27 When a message times out after an over-quota error from an Exim-imposed
2437 quota, the bounce message says "mailbox is full". This message was not
2438 being given when it was a system quota that was exceeded. It now should
2441 MH/03 Made $recipients available in local_scan(). local_scan() already has
2442 better access to the recipient list through recipients_list[], but
2443 $recipients can be useful in postmaster-provided expansion strings.
2445 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2446 in the case of a MAIL command with additional options following the
2447 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2448 were accidentally chopped off.
2450 PH/29 SMTP synchronization checks are implemented when a command is read -
2451 there is a check that no more input is waiting when there shouldn't be
2452 any. However, for some commands, a delay in an ACL can mean that it is
2453 some time before the response is written. In this time, more input might
2454 arrive, invalidly. So now there are extra checks after an ACL has run for
2455 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2456 pipelining has not been advertised.
2458 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2460 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2461 returned in the SMTP response when smtp_return_error_details was false.
2462 This has been fixed.
2464 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2465 instead of the C I/O that was originally supplied, because problems were
2466 reported on Solaris.
2468 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2469 Exim which did not show up earlier: it was assuming that a call to
2470 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2471 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2472 was a macro that became an assignment, so it seemed to work. This has
2473 changed to a proper function call with a void return, hence the compile
2474 error. Exim's code has been fixed.
2476 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2479 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2481 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2483 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2484 because it turns out to be too big or there is a timeout) there is no
2485 "Completed" line in the log. When some messages of this type were
2486 selected by exigrep, they were listed as "not completed". Others were
2487 picked up by some special patterns. I have improved the selection
2488 criteria to be more general.
2490 PH/38 The host_find_failed option in the manualroute router can now be set
2491 to "ignore", to completely ignore a host whose IP address cannot be
2492 found. If all hosts are ignored, the behaviour is controlled by the new
2493 host_all_ignored option.
2495 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2496 homing or because of multiple MX records with /mx) generated more than
2497 one IP address, and the following item turned out to be the local host,
2498 all the secondary addresses of the first item were incorrectly removed
2499 from the list, along with the local host and any following hosts (which
2500 is what is supposed to happen).
2502 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2503 whoever called Exim into the -H file. In the case of the daemon it was
2504 behaving confusingly. When first started, it used values for whoever
2505 started the daemon, but after a SIGHUP it used the Exim user (because it
2506 calls itself on a restart). I have changed the code so that it now always
2509 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2510 message are rejected with the same error (e.g. no authentication or bad
2511 sender address), and a DATA command is nevertheless sent (as can happen
2512 with PIPELINING or a stupid MUA), the error message that was given to the
2513 RCPT commands is included in the rejection of the DATA command. This is
2514 intended to be helpful for MUAs that show only the final error to their
2517 PH/42 Another patch from the Sieve maintainer.
2519 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2520 Eximstats - Fixed some broken HTML links and added missing column headers
2522 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2523 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2525 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2527 PH/43 Yet another patch from the Sieve maintainer.
2529 PH/44 I found a way to check for a TCP/IP connection going away before sending
2530 the response to the final '.' that terminates a message, but only in the
2531 case where the client has not sent further data following the '.'
2532 (unfortunately, this is allowed). However, in many cases there won't be
2533 any further data because there won't be any more messages to send. A call
2534 to select() can be used: if it shows that the input is "ready", there is
2535 either input waiting, or the socket has been closed. An attempt to read
2536 the next input character can distinguish the two cases. Previously, Exim
2537 would have sent an OK response which the client would never have see.
2538 This could lead to message repetition. This fix should cure that, at
2539 least in a lot of common cases.
2541 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2542 advertised in response to EHLO.
2548 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2549 fixed by 4.65/MH/01 (is this a record?) are fixed:
2551 (i) An empty string was always treated as zero by the numeric comparison
2552 operators. This behaviour has been restored.
2554 (ii) It is documented that the numeric comparison operators always treat
2555 their arguments as decimal numbers. This was broken in that numbers
2556 starting with 0 were being interpreted as octal.
2558 While fixing these problems I realized that there was another issue that
2559 hadn't been noticed. Values of message_size_limit (both the global option
2560 and the transport option) were treated as octal if they started with 0.
2561 The documentation was vague. These values are now always treated as
2562 decimal, and I will make that clear in the documentation.
2568 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2569 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2572 MH/01 Don't check that the operands of numeric comparison operators are
2573 integers when their expansion is in "skipping" mode (fixes bug
2574 introduced by 4.64-PH/07).
2576 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2577 child addresses, Exim now panics and dies. Previously, because the count
2578 is held in a short int, deliveries were likely to be lost. As such a
2579 large number of recipients for a single message is ridiculous
2580 (performance will be very, very poor), I have chosen to impose a limit
2581 rather than extend the field.
2587 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2588 leftover -K file (the existence of which was triggered by #402).
2589 While we were at it, introduced process PID as part of the -K
2590 filename. This should rule out race conditions when creating
2593 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2594 processing considerably. Previous code took too long for large mails,
2595 triggering a timeout which in turn triggers #401.
2597 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2598 in the DK code in transports.c. sendfile() is not really portable,
2599 hence the _LINUX specificness.
2601 TF/01 In the add_headers option to the mail command in an Exim filter,
2602 there was a bug that Exim would claim a syntax error in any
2603 header after the first one which had an odd number of characters
2606 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2607 callout verification, Exim cached a "reject" for the entire domain. This
2608 is correct for most verifications, but it is not correct for a recipient
2609 verification with use_sender or use_postmaster set, because in that case
2610 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2611 case of MAIL FROM:<> rejection from other early rejections (e.g.
2612 rejection of HELO). When verifying a recipient using a non-null MAIL
2613 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2614 Whatever the result of the callout, the value of the domain cache is
2615 left unchanged (for any other kind of callout, getting as far as trying
2616 RCPT means that the domain itself is ok).
2618 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2621 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2622 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2623 session, the daemon ignores ECONNECT errors and logs others; it now
2624 ignores EPIPE as well.
2626 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2627 (quoted-printable decoding).
2629 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2630 later the small subsequent patch to fix an introduced bug.
2632 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2634 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2636 PH/08 An error is now given if message_size_limit is specified negative.
2638 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2639 to be given (somewhat) arbitrary names.
2641 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2644 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2645 miscellaneous code fixes
2647 PH/10 Added the log_reject_target ACL modifier to specify where to log
2650 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2651 hostname. This is wrong, because it relates to the incoming message (and
2652 probably the interface on which it is arriving) and not to the outgoing
2653 callout (which could be using a different interface). This has been
2654 changed to use the value of the helo_data option from the smtp transport
2655 instead - this is what is used when a message is actually being sent. If
2656 there is no remote transport (possible with a router that sets up host
2657 addresses), $smtp_active_hostname is used.
2659 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2660 tweaks were necessary in order to get it to work (see also 21 below):
2661 (a) The code assumed that strncpy() returns a negative number on buffer
2662 overflow, which isn't the case. Replaced with Exim's string_format()
2664 (b) There were several signed/unsigned issues. I just did the minimum
2665 hacking in of casts. There is scope for a larger refactoring.
2666 (c) The code used strcasecmp() which is not a standard C function.
2667 Replaced with Exim's strcmpic() function.
2668 (d) The code set only $1; it now sets $auth1 as well.
2669 (e) A simple test gave the error "authentication client didn't specify
2670 service in request". It would seem that Dovecot has changed its
2671 interface. Fortunately there's a specification; I followed it and
2672 changed what the client sends and it appears to be working now.
2674 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2677 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2678 address A is aliased to B and C, where B exists and C does not. Without
2679 -v the output is "A verified" because verification stops after a
2680 successful redirection if more than one address is generated. However,
2681 with -v the child addresses are also verified. Exim was outputting "A
2682 failed to verify" and then showing the successful verification for C,
2683 with its parentage. It now outputs "B failed to verify", showing B's
2684 parentage before showing the successful verification of C.
2686 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2687 look up a TXT record in a specific list after matching in a combined
2690 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2691 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2692 they consult the DNS. I had assumed they would set it the way they
2693 wanted; and indeed my experiments on Linux seem to show that in some
2694 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2695 To be on the safe side, however, I have now made the interface to
2696 host_find_byname() similar to host_find_bydns(), with an argument
2697 containing the DNS resolver options. The host_find_byname() function now
2698 sets these options at its start, just as host_find_bydns() does. The smtp
2699 transport options dns_qualify_single and dns_search_parents are passed to
2700 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2701 of host_find_byname() use the default settings of RES_DEFNAMES
2702 (qualify_single) but not RES_DNSRCH (search_parents).
2704 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2705 spool_read_header() do less string testing, by means of a preliminary
2706 switch on the second character of optional "-foo" lines. (This is
2707 overdue, caused by the large number of possibilities that now exist.
2708 Originally there were few.) While I was there, I also converted the
2709 str(n)cmp tests so they don't re-test the leading "-" and the first
2710 character, in the hope this might squeeze out yet more improvement.
2712 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2713 flag allowing group syntax was set by the header_syntax check but not
2714 turned off, possible causing trouble later; (2) The flag was not being
2715 set at all for the header_verify test, causing "group"-style headers to
2716 be rejected. I have now set it in this case, and also caused header_
2717 verify to ignore an empty address taken from a group. While doing this, I
2718 came across some other cases where the code for allowing group syntax
2719 while scanning a header line wasn't quite right (mostly, not resetting
2720 the flag correctly in the right place). These bugs could have caused
2721 trouble for malformed header lines. I hope it is now all correct.
2723 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2724 with the "reply" argument non-NULL. The code, however (which originally
2725 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2726 but it didn't always do it. This confused somebody who was copying the
2727 code for some other use. I have removed all the tests.
2729 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2730 feature that was used to support insecure browsers during the U.S. crypto
2731 embargo. It requires special client support, and Exim is probably the
2732 only MTA that supported it -- and would never use it because real RSA is
2733 always available. This code has been removed, because it had the bad
2734 effect of slowing Exim down by computing (never used) parameters for the
2735 RSA_EXPORT functionality.
2737 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2738 authenticator to fail if there's a tab character in the incoming data
2739 (there should never be unless someone is messing about, as it's supposed
2740 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2741 if the connection is using TLS or if the remote IP is the same as the
2742 local IP, and the "valid-client-cert option" if a client certificate has
2745 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2746 authenticators. This can be used for authorization after authentication
2747 succeeds. (In the case of plaintext, it servers for both authentication
2750 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2751 if any retry times were supplied.
2753 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2754 connection, where there is no client IP address to check. In this
2755 situation, the verify now always succeeds.
2757 PH/25 Applied John Jetmore's -Mset patch.
2759 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2761 PH/27 In a string expansion for a processed (not raw) header when multiple
2762 headers of the same name were present, leading whitespace was being
2763 removed from all of them, but trailing whitespace was being removed only
2764 from the last one. Now trailing whitespace is removed from each header
2765 before concatenation. Completely empty headers in a concatenation (as
2766 before) are ignored.
2768 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2769 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2771 PH/29 [Removed. This was a change that I later backed out, and forgot to
2772 correct the ChangeLog entry (that I had efficiently created) before
2773 committing the later change.]
2775 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2776 address errors (4xx response to RCPT) over the same connection as other
2777 messages routed to the same hosts. Such deliveries are always "forced",
2778 so retry times are not inspected. This resulted in far too many retries
2779 for the affected addresses. The effect occurred only when there were more
2780 hosts than the hosts_max_try setting in the smtp transport when it had
2781 the 4xx errors. Those hosts that it had tried were not added to the list
2782 of hosts for which the message was waiting, so if all were tried, there
2783 was no problem. Two fixes have been applied:
2785 (i) If there are any address or message errors in an SMTP delivery, none
2786 of the hosts (tried or untried) are now added to the list of hosts
2787 for which the message is waiting, so the message should not be a
2788 candidate for sending over the same connection that was used for a
2789 successful delivery of some other message. This seems entirely
2790 reasonable: after all the message is NOT "waiting for some host".
2791 This is so "obvious" that I'm not sure why it wasn't done
2792 previously. Hope I haven't missed anything, but it can't do any
2793 harm, as the worst effect is to miss an optimization.
2795 (ii) If, despite (i), such a delivery is accidentally attempted, the
2796 routing retry time is respected, so at least it doesn't keep
2797 hammering the server.
2799 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2800 in ${readsocket because some servers need this prod.
2802 PH/32 Added some extra debug output when updating a wait-xxx database.
2804 PH/33 The hint "could be header name not terminated by colon", which has been
2805 given for certain expansion errors for a long time, was not being given
2806 for the ${if def:h_colon_omitted{... case.
2808 PH/34 The spec says: "With one important exception, whenever a domain list is
2809 being scanned, $domain contains the subject domain." There was at least
2810 one case where this was not true.
2812 PH/35 The error "getsockname() failed: connection reset by peer" was being
2813 written to the panic log as well as the main log, but it isn't really
2814 panic-worthy as it just means the connection died rather early on. I have
2815 removed the panic log writing for the ECONNRESET error when getsockname()
2818 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2819 runs only) independently of the message's sender address. This meant
2820 that, if the 4xx error was in fact related to the sender, a different
2821 message to the same recipient with a different sender could confuse
2822 things. In particular, this can happen when sending to a greylisting
2823 server, but other circumstances could also provoke similar problems.
2824 I have changed the default so that the retry time for these errors is now
2825 based a combination of the sender and recipient addresses. This change
2826 can be overridden by setting address_retry_include_sender=false in the
2829 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2830 remote server are returned as part of bounce messages. This was not
2831 happening for LMTP over a pipe (the lmtp transport), but now it is the
2832 same for both kinds of LMTP.
2834 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2835 in header lines that were in fact CNAMEs. This is no longer the case.
2837 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2838 and queue runs started by the daemon processed all messages. This has
2839 been fixed so that -R and -S can now usefully be given with -q<time>.
2841 PH/40 Import PCRE release 6.7 (fixes some bugs).
2843 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2845 PH/42 Give an error if -q is specified more than once.
2847 PH/43 Renamed the variables $interface_address and $interface_port as
2848 $received_ip_address and $received_port, to make it clear that these
2849 values apply to message reception, and not to the outgoing interface when
2850 a message is delivered. (The old names remain recognized, of course.)
2852 PH/44 There was no timeout on the connect() call when using a Unix domain
2853 socket in the ${readsocket expansion. There now is.
2855 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2856 be meaningful with "accept".
2858 SC/01 Eximstats V1.43
2859 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2861 SC/02 Eximstats V1.44
2862 Use a glob alias rather than an array ref in the generated
2863 parser. This improves both readability and performance.
2865 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2866 Collect SpamAssassin and rejection statistics.
2867 Don't display local sender or destination tables unless
2868 there is data to show.
2869 Added average volumes into the top table text output.
2871 SC/04 Eximstats V1.46
2872 Collect data on the number of addresses (recipients)
2873 as well as the number of messages.
2875 SC/05 Eximstats V1.47
2876 Added 'Message too big' to the list of mail rejection
2877 reasons (thanks to Marco Gaiarin).
2879 SC/06 Eximstats V1.48
2880 Mainlog lines which have GMT offsets and are too short to
2881 have a flag are now skipped.
2883 SC/07 Eximstats V1.49 (Alain Williams)
2884 Added the -emptyok flag.
2886 SC/08 Eximstats V1.50
2887 Fixes for obtaining the IP address from reject messages.
2889 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2890 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2891 whitespace changes from 4.64-PH/27
2893 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2896 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2897 are found, allow negative numbers in numeric criteria)
2899 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2901 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2902 to match changes made in 4.64-PH/43
2904 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2906 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2907 "domain" is actually an IP address, and force a failure. This locks out
2908 those revolvers/nameservers that support "A-for-A" lookups, in
2909 contravention of the specifications.
2911 PH/48 When a host name was looked up from an IP address, and the subsequent
2912 forward lookup of the name timed out, the host name was left in
2913 $sender_host_name, contrary to the specification.
2915 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2916 restricted to single-key lookups, Exim was not diagnosing an error if
2917 * or *@ was used with a query-style lookup.
2919 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2921 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2922 long ago, but noone interested enough thought of it. Let's just say that
2923 the "1.1" means that there are some new functions that weren't there at
2924 some point in the past.
2926 PH/51 Error processing for expansion failure of helo_data from an smtp
2927 transport during callout processing was broken.
2929 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2930 tested/used via the -bh/-bhc/-bs options.
2932 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2933 bug, fixed in subsequent PCRE releases).
2935 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2936 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2938 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2944 SC/01 Use a glob alias rather than an array ref in eximstats generated
2945 parser. This improves both readability and performance.
2947 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2948 Don't display local sender or destination tables in eximstats unless
2949 there is data to show.
2950 Added average volumes into the eximstats top table text output.
2952 SC/03 Collect data on the number of addresses (recipients) as well
2953 as the number of messages in eximstats.
2955 TF/01 Correct an error in the documentation for the redirect router. Exim
2956 does (usually) call initgroups() when daemonizing.
2958 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2959 with consistent privilege compared to when running as a daemon.
2961 TF/03 Note in the spec that $authenticated_id is not set for local
2962 submissions from trusted users.
2964 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2965 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2967 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2968 by adding some example configuration directives to the default
2969 configuration file. A little bit of work is required to uncomment the
2970 directives and define how usernames and passwords are checked, but
2971 there is now a framework to start from.
2973 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2974 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2975 without this. I don't know how relevant this is to other LDAP libraries.
2977 PH/02 Add the verb name to the "unknown ACL verb" error.
2979 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2981 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2983 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2984 directory not expanded when it should be if an expanded home directory
2985 was set for the address (which is overridden by the transport).
2987 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2990 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2991 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2992 because it is too late at that time, and has no effect.
2994 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2995 security issue with \' (bugzilla #107). I could not use the
2996 PQescapeStringConn() function, because it needs a PGconn value as one of
2999 PH/08 When testing addresses using -bt, indicate those final addresses that
3000 are duplicates that would not cause an additional delivery. At least one
3001 person was confused, thinking that -bt output corresponded to deliveries.
3002 (Suppressing duplicates isn't a good idea as you lose the information
3003 about possibly different redirections that led to the duplicates.)
3005 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3006 systems where poll() doesn't work, in particular OS X.
3008 PH/10 Added more information to debugging output for retry time not reached.
3010 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3011 operations in malware.c.
3013 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3016 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3017 syslog_duplication set false, log lines that would normally be written
3018 both the the main log and to the reject log were not written to syslog at
3021 PH/14 In the default configuration, change the use of "message" in ACL warn
3022 statements to "add_header".
3024 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3025 not followed by a command (e.g. "seen endif").
3027 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3028 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3031 PH/17 Added extra conditions to the default value of delay_warning_condition
3035 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3036 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3037 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3040 The Auto-Submitted: and various List- headers are standardised, whereas I
3041 don't think Precedence: ever was.
3043 PH/18 Refactored debugging code in route_finduser() to show more information,
3044 in particular, the error code if getpwnam() issues one.
3046 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3047 This is apparently needed in addition to the PH/07 change above to avoid
3048 any possible encoding problems.
3050 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3051 but not after initializing Perl.
3053 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3054 output them only if debugging. By default they are written stderr,
3055 apparently, which is not desirable.
3057 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3060 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3063 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3065 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3066 authenticated or an ident call has been made. Suppress the default
3067 values for $authenticated_id and $authenticated_sender (but permit -oMai
3068 and -oMas) when testing with -bh.
3070 PH/24 Re-jigged the order of the tests in the default configuration so that the
3071 tests for valid domains and recipients precede the DNS black list and CSA
3072 tests, on the grounds that those ones are more expensive.
3074 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3075 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3076 This bug exists in every version of Exim that I still have, right back to
3079 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3080 However, an attempt to turn on case-sensitivity in a regex key by
3081 including (?-i) didn't work because the subject string was already
3082 lowercased, and the effects were non-intuitive. It turns out that a
3083 one-line patch can be used to allow (?-i) to work as expected.
3089 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3090 other effects) broke the use of negated acl sub-conditions.
3092 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3095 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3096 "Deny" causes Exim to reject the incoming connection with a 554 error.
3097 Unfortunately, if there is a major crisis, such as a disk failure,
3098 tcp-wrappers gives "deny", whereas what one would like would be some
3099 kind of temporary error. A kludge has been added to help with this.
3100 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3101 554 error is used if errno is still zero or contains ENOENT (which occurs
3102 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3105 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3107 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3108 errors. Otherwise a message that provokes a temporary error (when other
3109 messages do not) can cause a whole host to time out.
3111 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3112 addresses were routed directly to files or pipes from a redirect router.
3113 File deliveries just didn't batch; pipe deliveries might have suffered
3116 PH/06 A failure to get a lock for a hints database would erroneously always say
3117 "Failed to get write lock", even when it was really a read lock.
3119 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3120 of 0600. This has been changed to use the value of the lockfile_mode
3121 option (which defaults to 0600).
3123 PH/08 Applied small patch from the Sieve maintainer.
3125 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3126 folder from quota calculations, a direct delivery into this folder messed
3127 up the contents of the maildirsize file. This was because the regex was
3128 used only to exclude .Trash (or whatever) when the size of the mailbox
3129 was calculated. There was no check that a delivery was happening into an
3130 excluded directory. This bug has been fixed by ignoring all quota
3131 processing for deliveries into excluded directories.
3133 PH/10 Added the maildirfolder_create_regex option to appendfile.
3139 PH/01 The code for finding all the local interface addresses on a FreeBSD
3140 system running IPv6 was broken. This may well have applied to all BSD
3141 systems, as well as to others that have similar system calls. The broken
3142 code found IPv4 interfaces correctly, but gave incorrect values for the
3143 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3144 that it would not match correctly against @[] and not recognize the IPv6
3147 PH/02 The ipliteral router was not recognizing addresses of the form user@
3148 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3150 PH/03 Added disable_ipv6.
3152 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3153 decoded form, because it is most often used to construct To: headers
3154 lines in autoreplies, and the decoded form may well be syntactically
3155 invalid. However, $reply_address has leading white space removed, and all
3156 newlines turned into spaces so that the autoreply transport does not
3159 PH/05 If group was specified without a user on a router, and no group or user
3160 was specified on a transport, the group from the router was ignored.
3162 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3163 for visible compile-time settings that can be used to change these
3164 numbers, for those that want even more. Backwards compatibility with old
3165 spool files has been maintained. However, going back to a previous Exim
3166 release will lost any variables that are in spool files.
3168 PH/07 Two small changes when running in the test harness: increase delay when
3169 passing a TCP/IP connection to a new process, in case the original
3170 process has to generate a bounce, and remove special handling of
3171 127.0.0.2 (sic), which is no longer necessary.
3173 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3174 be the same on different OS.
3176 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3179 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3180 whether --show-vars was specified or not
3182 JJ/02 exipick: Added support for new ACL variable spool format introduced
3185 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3186 syntactically invalid From: or Reply-to: line, and a filter used this to
3187 generate an autoreply, and therefore failed to obtain an address for the
3188 autoreply, Exim could try to deliver to a non-existent relative file
3189 name, causing unrelated and misleading errors. What now happens is that
3190 it logs this as a hard delivery error, but does not attempt to create a
3193 PH/11 The exinext utility has a -C option for testing purposes, but although
3194 the given file was scanned by exinext itself; it wasn't being passed on
3195 when Exim was called.
3197 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3198 an end-of-file indication when reading a command response.
3200 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3201 compiled. In many other places in Exim, IPv6 addresses are always
3202 recognized, so I have changed this. It also means that IPv4 domain
3203 literals of the form [IPV4:n.n.n.n] are now always recognized.
3205 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3206 used if the router is not running as root, for example, when verifying at
3207 ACL time, or when using -bh. The debugging output from this situation was
3208 non-existent - all you got was a failure to exec. I have made two
3211 (a) Failures to set uid/gid, the current directory, or a process leader
3212 in a subprocess such as that created by queryprogram now generate
3213 suitable debugging output when -d is set.
3215 (b) The queryprogram router detects when it is not running as root,
3216 outputs suitable debugging information if -d is set, and then runs
3217 the subprocess without attempting to change uid/gid.
3219 PH/15 Minor change to Makefile for building test_host (undocumented testing
3222 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3223 additional section of a DNS packet that returns MX or SRV records.
3224 Instead, it always explicitly searches for A/AAAA records. This avoids
3225 major problems that occur when a DNS server includes only records of one
3226 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3227 fixed another bug: if SRV records were looked up and the corresponding
3228 address records were *not* found in the additional section, the port
3229 values from the SRV records were lost.
3231 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3232 using the correct key (the original address) when searching the retry
3233 rules in order to find which one to use for generating the retry hint.
3235 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3236 adding the default one. Similarly, if it contains a Reply-To: header, the
3237 errors_reply_to option, if set, is not used.
3239 PH/19 When calculating a retry time, Exim used to measure the "time since
3240 failure" by looking at the "first failed" field in the retry record. Now
3241 it does not use this if it is later than than the arrival time of the
3242 message. Instead it uses the arrival time. This makes for better
3243 behaviour in cases where some deliveries succeed, thus re-setting the
3244 "first failed" field. An example is a quota failure for a huge message
3245 when small messages continue to be delivered. Without this change, the
3246 "time since failure" will always be short, possible causing more frequent
3247 delivery attempts for the huge message than are intended.
3248 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3250 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3251 $1, $2, $3) because the numerical variables can be reset during some
3252 expansion items (e.g. "match"), thereby losing the authentication data.
3254 PH/21 Make -bV show the size of off_t variables so that the test suite can
3255 decide whether to run tests for quotas > 2G.
3257 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3258 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3259 filecount value is greater than 2G or if a quota value is greater than 2G
3260 on a system where the size of off_t is not greater than 4, a panic error
3263 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3264 never match. The debug and -bh output now contains an explicit error
3265 message indicating a malformed IPv4 address or mask.
3267 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3268 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3269 PH/23 above applies.
3271 PH/25 Do not write to syslog when running in the test harness. The only
3272 occasion when this arises is a failure to open the main or panic logs
3273 (for which there is an explicit test).
3275 PH/26 Added the /no_tell option to "control=freeze".
3277 PH/27 If a host name lookup failed very early in a connection, for example, if
3278 the IP address matched host_lookup and the reverse lookup yielded a name
3279 that did not have a forward lookup, an error message of the form "no IP
3280 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3281 could be logged. Now it outputs the IP address instead of "NULL".
3283 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3284 allows the sender and the authenticated sender to be set when
3285 submitting a message from within Exim. Since child_open_exim() is
3286 documented for local_scan(), the new function should be too.
3288 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3289 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3290 results in an empty string is now treated as unset.
3292 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3294 PH/31 Added sender_verify_fail logging option.
3296 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3297 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3298 tidied the source and removed it altogether.
3300 PH/33 When a queue run was abandoned because the load average was too high, a
3301 log line was always written; now it is written only if the queue_run log
3302 selector is set. In addition, the log line for abandonment now contains
3303 information about the queue run such as the pid. This is always present
3304 in "start" and "stop" lines but was omitted from the "abandon" line.
3306 PH/34 Omit spaces between a header name and the colon in the error message that
3307 is given when verify = headers_syntax fails (if there are lots of them,
3308 the message gets confusing).
3310 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3311 names, as there are now some PTR records that contain slashes. This check
3312 is only to protect against broken name servers that fall over on strange
3313 characters, so the fact that it applies to all lookups doesn't matter.
3315 PH/36 Now that the new test suite is complete, we can remove some of the
3316 special code in Exim that was needed for the old test suite. For example,
3317 sorting DNS records because real resolvers return them in an arbitrary
3318 order. The new test suite's fake resolver always returns records in the
3321 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3322 bounces) except when queue_only is set, to avoid logging races between
3323 the different processes.
3325 PH/38 Panic-die if .include specifies a non-absolute path.
3327 PH/39 A tweak to the "H" retry rule from its user.
3329 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3330 a label. They prevented compilation on older perls.
3332 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3333 a warning to be raised on newish perls.
3335 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3336 on queue. Changes to match documented behaviour of showing count of
3337 messages matching specified criteria.
3339 PH/40 Changed the default ident timeout from 30s to 5s.
3341 PH/41 Added support for the use of login_cap features, on those BSD systems
3342 that have them, for controlling the resources used by pipe deliveries.
3344 PH/42 The content-scanning code uses fopen() to create files in which to put
3345 message data. Previously it was not paying any attention to the mode of
3346 the files. Exim runs with umask(0) because the rest of the code creates
3347 files with open(), and sets the required mode explicitly. Thus, these
3348 files were ending up world-writeable. This was not a big issue, because,
3349 being within the spool directory, they were not world-accessible. I have
3350 created a function called modefopen, which takes an additional mode
3351 argument. It sets umask(777), creates the file, chmods it to the required
3352 mode, then resets the umask. All the relevant calls to fopen() in the
3353 content scanning code have been changed to use this function.
3355 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3356 to 24 hours. This avoids potential overflow problems when processing G
3357 and H retry rules. I suspect nobody ever tinkers with this value.
3359 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3361 PH/45 When the plaintext authenticator is running as a client, the server's
3362 challenges are checked to ensure they are valid base64 strings. By
3363 default, the authentication attempt is cancelled if an invalid string is
3364 received. Setting client_ignore_invalid_base64 true ignores these errors.
3365 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3366 they are received. Thus, the responses can be made to depend on the
3367 challenges. If an invalid string is ignored, an empty string is placed in
3370 PH/46 Messages that are created by the autoreply transport now contains a
3371 References: header, in accordance with RFCs 2822 and 3834.
3373 PH/47 Added authenticated_sender_force to the smtp transport.
3375 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3377 PH/49 Installed latest patch from the Sieve maintainer.
3379 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3380 was also set, the appendfile transport was unnecessarily scanning a
3381 directory of message files (e.g. for maildir delivery) to find the count
3382 of files (along with the size), even though it did not need this
3383 information. It now does the scan only if it needs to find either the
3384 size of the count of files.
3386 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3388 PH/52 Two bugs concerned with error handling when the smtp transport is
3391 (i) Exim was not creating retry information for temporary errors given
3392 for individual recipients after the DATA command when the smtp transport
3393 was used in LMTP mode. This meant that they could be retried too
3394 frequently, and not timed out correctly.
3396 (ii) Exim was setting the flag that allows error details to be returned
3397 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3398 recipients that were returned after the DATA command.
3400 PH/53 This is related to PH/52, but is more general: for any failing address,
3401 when detailed error information was permitted to be returned to the
3402 sender, but the error was temporary, then after the final timeout, only
3403 "retry timeout exceeded" was returned. Now it returns the full error as
3404 well as "retry timeout exceeded".
3406 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3407 do this, and (what is worse) MTAs that accept it.
3409 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3410 will now be deprecated.
3412 PH/56 New os.c-cygwin from the Cygwin maintainer.
3414 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3415 formats (previously only available in exim formats via -bpr, -bpru,
3416 and -bpra. Now also available in native and exiqgrep formats)
3418 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3419 with very large, slow to parse queues
3421 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3423 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3425 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3426 responses to authentication challenges, though it was showing the
3427 challenges; (ii) I've removed the CR characters from the debug output for
3430 PH/58 Allow for the insertion of a newline as well as a space when a string
3431 is turned into more than one encoded-word during RFC 2047 encoding. The
3432 Sieve code now uses this.
3434 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3435 data_4xx, lost_connection, tls_required.
3437 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3438 message was being sent as an SMTP response.
3440 PH/61 Add -l and -k options to exicyclog.
3442 PH/62 When verifying, if an address was redirected to one new address, so that
3443 verification continued, and the new address failed or deferred after
3444 having set something in $address_data, the value of $address_data was not
3445 passed back to the ACL. This was different to the case when no
3446 redirection occurred. The value is now passed back in both cases.
3448 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3449 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3450 use login_cap.h, so on its own it isn't the distinguishing feature. The
3451 new name refers directly to the setclassresources() function.
3453 PH/65 Added configuration files for NetBSD3.
3455 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3457 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3458 is preferred over IPv4.
3460 PH/68 The bounce_return_message and bounce_return_body options were not being
3461 honoured for bounces generated during the reception of non-SMTP messages.
3462 In particular, this applied to messages rejected by the ACL. This bug has
3463 been fixed. However, if bounce_return_message is true and bounce_return_
3464 body is false, the headers that are returned for a non-SMTP message
3465 include only those that have been read before the error was detected.
3466 (In the case of an ACL rejection, they have all been read.)
3468 PH/69 The HTML version of the specification is now built in a directory called
3469 spec_html instead of spec.html, because the latter looks like a path with
3470 a MIME-type, and this confuses some software.
3472 PH/70 Catch two compiler warnings in sieve.c.
3474 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3475 function verify_get_ident() calls ip_connect() to connect a socket, but
3476 if the "connect()" function timed out, ip_connect() used to close the
3477 socket. However, verify_get_ident() also closes the socket later, and in
3478 between Exim writes to the log, which may get opened at this point. When
3479 the socket was closed in ip_connect(), the log could get the same file
3480 descriptor number as the socket. This naturally causes chaos. The fix is
3481 not to close the socket in ip_connect(); the socket should be closed by
3482 the function that creates it. There was only one place in the code where
3483 this was missing, in the iplookup router, which I don't think anybody now
3484 uses, but I've fixed it anyway.
3486 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3487 well as to direct DNS lookups. Otherwise the handling of names in host
3488 lists is inconsistent and therefore confusing.
3494 PH/01 Two changes to the default runtime configuration:
3496 (1) Move the checks for relay_from_hosts and authenticated clients from
3497 after to before the (commented out) DNS black list checks.
3499 (2) Add control=submission to the relay_from_hosts and authenticated
3500 clients checks, on the grounds that messages accepted by these
3501 statements are most likely to be submissions.
3503 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3505 (1) Generate an error if the third argument for the ${prvs expansion is
3508 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3511 (3) Reset the variables that are obtained from the first argument of
3512 ${prvscheck and used in the second argument before leaving the code,
3513 because their memory is reclaimed, so using them afterwards may do
3516 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3517 one (it's much easier than Tom thought :-).
3519 (5) Because of (4), we can now allow for the use of $prvscheck_result
3520 inside the third argument.
3522 PH/03 For some reason, the default setting of PATH when running a command from
3523 a pipe transport was just "/usr/bin". I have changed it to
3526 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3527 anything to be listed in the output from -bV.
3529 PH/05 When a filter generated an autoreply, the entire To: header line was
3530 quoted in the delivery log line, like this:
3532 => >A.N.Other <ano@some.domain> <original@ddress> ...
3534 This has been changed so that it extracts the operative address. There
3535 may be more than one such address. If so, they are comma-separated, like
3538 => >ano@some.domain,ona@other.domain <original@ddress> ...
3540 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3541 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3542 not being looked up in the rDNS to get a host name, Exim was showing the
3543 IP address twice in Received: lines, even though the IP addresses were
3544 identical. For example:
3546 Received: from [1.2.3.4] (helo=[1.2.3.4])
3548 However, if the real host name was known, it was omitting the HELO data
3549 if it matched the actual IP address. This has been tidied up so that it
3550 doesn't show the same IP address twice.
3552 PH/07 When both +timestamp and +memory debugging was on, the value given by
3553 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3554 called by the debug printing, thereby overwriting the timestamp buffer.
3555 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3557 PH/08 When the original message was included in an autoreply transport, it
3558 always said "this is a copy of the message, including all the headers",
3559 even if body_only or headers_only was set. It now gives an appropriate
3562 PH/09 Applied a patch from the Sieve maintainer which:
3564 o fixes some comments
3565 o adds the (disabled) notify extension core
3566 o adds some debug output for the result of if/elsif tests
3567 o points to the current vacation draft in the documentation
3568 and documents the missing references header update
3572 o fixes a bug in processing the envelope test (when testing
3573 multiple envelope elements, the last element determined the
3576 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3577 Electronic Mail") by including:
3579 Auto-submitted: auto-generated
3581 in the messages that it generates (bounce messages and others, such as
3582 warnings). In the case of bounce messages for non-SMTP messages, there was
3583 also a typo: it was using "Auto_submitted" (underscore instead of
3584 hyphen). Since every message generated by Exim is necessarily in response
3585 to another message, thes have all been changed to:
3587 Auto-Submitted: auto-replied
3589 in accordance with these statements in the RFC:
3591 The auto-replied keyword:
3593 - SHOULD be used on messages sent in direct response to another
3594 message by an automatic process,
3596 - MUST NOT be used on manually-generated messages,
3598 - MAY be used on Delivery Status Notifications (DSNs) and Message
3599 Disposition Notifications (MDNs),
3601 - MUST NOT be used on messages generated by automatic or periodic
3602 processes, except for messages which are automatic responses to
3605 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3606 to the default Received: header definition.
3608 PH/12 Added log selector acl_warn_skipped (default on).
3610 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3611 variables because (a) they are in the wrong storage pool and (b) even if
3612 they were copied, it wouldn't work properly because of the caching.
3614 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3615 checking when decoding. Apparently there are clients that generate
3616 overlong encoded strings. Why am I not surprised?
3618 PH/15 If the first argument of "${if match_address" was not empty, but did not
3619 contain an "@" character, Exim crashed. Now it writes a panic log message
3620 and treats the condition as false.
3622 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3624 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3625 extension "envelope-auth". The code is finished and in agreement with
3626 other implementations, but there is no documentation so far and in fact,
3627 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3628 not changing the active code.
3630 Print executed "if" and "elsif" statements when debugging is used. This
3631 helps a great deal to understand what a filter does.
3633 Document more things not specified clearly in RFC3028. I had all this
3634 sorted out, when out of a sudden new issues came to my mind. Oops."
3636 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3639 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3640 canonical form (as documented). However, after a host name lookup from
3641 the IP address, check_host() was doing a simple string comparison with
3642 addresses acquired from the DNS when checking that the found name did
3643 have the original IP as one of its addresses. Since any found IPv6
3644 addresses are likely to be in abbreviated form, the comparison could
3645 fail. Luckily, there already exists a function for doing the comparison
3646 by converting both addresses to binary, so now that is used instead of
3647 the text comparison.
3649 PH/20 There was another similar case to PH/19, when a complete host name was
3650 given in a host list; looking up its IP address could give an abbreviated
3651 form, whereas the current host's name might or might not be abbreviated.
3652 The same fix has been applied.
3658 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3659 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3662 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3663 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3665 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3667 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3668 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3669 Also, for "Auto-Submitted", treat anything other than "no" as
3670 non-personal, in accordance with RFC 3834. (Previously it treated
3671 anything starting "auto-" as non-personal.)
3673 TF/01 The control=submission/name=... option had a problem with syntax
3674 errors if the name included a slash character. The /name= option
3675 now slurps the rest of the string, so it can include any characters
3676 but it must come last in the list of options (after /sender_retain
3679 PH/05 Some modifications to the interface to the fake nameserver for the new
3687 TK/01 Added the "success_on_redirect" address verification option. See
3688 NewStuff for rationale and an example.
3690 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3692 PH/02 Patch to exigrep to allow it to work on syslog lines.
3694 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3695 fread() to skip over the body file's header line, because in Cygwin the
3696 header line is locked and is inaccessible.
3698 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3699 co-exist for some time) to make it clear that it is the Exim ID that is
3700 referenced, not the Message-ID: header line.
3702 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3703 string_format() function, because snprintf() does not exist on all
3706 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3707 ${stat: expansion item.
3709 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3710 protocol synchronization error", to keep the pedants happy.
3712 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3713 well as for IRIX systems, when gcc is being used. See the host.c source
3716 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3718 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3721 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3722 transport and to the smtp transport in LMTP mode.
3724 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3726 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3727 run a filter in a subprocess. This could lead to confusion in subsequent
3728 lookups in the parent process. There should also be a search_tidyup() at
3729 the end of the subprocess.
3731 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3732 only if the host matched helo_try_verify_hosts, which caused the
3733 verification to occur when the EHLO/HELO command was issued. The ACL just
3734 tested the remembered result. Now, if a previous verification attempt has
3735 not happened, "verify = helo" does it there and then.
3737 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3739 TK/03 Fix log output including CR from clamd.
3741 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3742 exist provoked a memory error which could cause a segfault.
3744 PH/15 Installed PCRE 6.2
3746 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3748 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3749 of the problem. Specifically, suggested +O2 rather than +O1 for the
3752 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3754 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3755 record, the port was not correctly logged when the outgoing_port log
3756 selector was set (it logged the transort's default port).
3758 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3759 fallback_hosts, and "hosts" in the smtp transport.
3761 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3762 host errors such as "Connection refused".
3764 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3765 authentication with radiusclient 0.4.9:
3767 - Error returned from rc_read_config was caught wrongly
3768 - Username/password not passed on to radius server due to wrong length.
3770 The presumption is that some radiusclient API changes for 4.51/PH/17
3771 were not taken care of correctly. The code is still untested by me (my
3772 Linux distribution still has 0.3.2 of radiusclient), but it was
3773 contributed by a Radius user.
3775 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3776 expanding the "port" option of the smtp transport.
3778 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3779 while reading a MIME header. Thanks to Tom Hughes for a patch.
3781 PH/24 Include config.h inside local_scan.h so that configuration settings are
3784 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3785 that in an ACL for RCPT (for example), you can examine exactly what was
3788 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3789 commands, but it was not correctly comparing the address with the actual
3790 client host address. Thus, it would show the EHLO address in Received:
3791 header lines when this was not necessary.
3793 PH/27 Added the % operator to ${eval:}.
3795 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3796 it should be ignoring failures (because with -C, for example, it has lost
3797 privilege). It wasn't ignoring creation failures other than "already
3800 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3803 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3804 because an input line was too long, either on its own, or by virtue of
3805 too many continuations, the temporary file was not being removed, and the
3806 return code was incorrect.
3808 PH/31 Missing "BOOL" in function definition in filtertest.c.
3810 PH/32 Applied Sieve patches from the maintainer.
3812 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3814 PH/33 Added "verify = not_blind".
3816 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3817 Local/Makefile (with some defaults set). These are used in built scripts
3818 such as exicyclog, but they have never been used in the exim_install
3819 script (though there are many overriding facilities there). I have
3820 arranged that the exim_install script now takes note of these two
3823 PH/35 Installed configuration files for Dragonfly.
3825 PH/36 When a locally submitted message by a trusted user did not contain a
3826 From: header, and the sender address was obtained from -f or from an SMTP
3827 MAIL command, and the trusted user did not use -F to supply a sender
3828 name, $originator_name was incorrectly used when constructing a From:
3829 header. Furthermore, $originator_name was used for submission mode
3830 messages from external hosts without From: headers in a similar way,
3831 which is clearly wrong.
3833 PH/37 Added control=suppress_local_fixups.
3835 PH/38 When log_selector = +received_sender was set, and the addition of the
3836 sender made the log line's construction buffer exactly full, or one byte
3837 less than full, an overflow happened when the terminating "\n" was
3840 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3841 when the result of a list match is failure because a DNS lookup failed.
3843 PH/40 RM_COMMAND is now used in the building process.
3845 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3846 the "build-* directories that it finds.
3848 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3849 address in a domain literal was a prefix of an interface address.
3851 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3852 when verifying a sender address, unless rewrite_headers is false.
3854 PH/44 Wrote a long comment about why errors_to addresses are verified as
3855 recipients, not senders.
3857 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3858 the ratelimit ACL was added.
3860 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3862 PH/46 Added extra information about PostgreSQL errors to the error string.
3864 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3865 avoiding the need to install special zones in a real server. This is
3866 backwards compatible; if it can't find the fake resolver, it drops back.
3867 Thus, both old and new test suites can be run.
3869 TF/02 Added util/ratelimit.pl
3871 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3872 clock is set back in time.
3874 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3875 Candler <B.Candler@pobox.com>.
3877 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3878 OK for addresses that are the result of redirections.
3880 PH/48 A number of further additions for the benefit of the new test suite,
3881 including a fake gethostbyname() that interfaces to the fake DNS resolver
3884 TF/06 The fix for widen_domains has also been applied to qualify_single and
3885 search_parents which are the other dnslookup options that can cause
3888 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3891 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3893 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3894 service" if it encountered an MX record with an empty target hostname.
3895 The message is now "an MX or SRV record indicated no SMTP service".
3897 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3898 if the DNS of the sending site is misconfigured. This is quite a
3899 common situation. This change restores the behaviour of treating a
3900 helo verification defer as a failure.
3902 PH/51 If self=fail was set on a router, the bounce message did not include the
3903 actual error message.
3909 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3911 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3912 command itself ended in error, the underlying message about the transport
3913 filter timeout was being overwritten with the pipe command error. Now the
3914 underlying error message should be appended to the second error message.
3916 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3918 PH/02 Reduce the amount of output that "make" produces by default. Full output
3919 can still be requested.
3921 PH/03 The warning log line about a condition test deferring for a "warn" verb
3922 was being output only once per connection, rather than after each
3923 occurrence (because it was using the same function as for successful
3924 "warn" verbs). This seems wrong, so I have changed it.
3926 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3927 it should not have, which might have caused a crash in the right
3928 circumstances, but probably never did.
3930 PH/04 Installed a modified version of Tony Finch's patch to make submission
3931 mode fix the return path as well as the Sender: header line, and to
3932 add a /name= option so that you can make the user's friendly name appear
3935 TF/03 Added the control = fakedefer ACL modifier.
3937 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3938 Mark Lowes for thorough testing.
3940 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3942 TK/03 Merged latest SRS patch from Miles Wilton.
3944 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3945 with the definition in sysexits.h (which is #included earlier).
3946 Fortunately, Exim does not actually use EX_OK. The code used to try to
3947 preserve the sysexits.h value, by assuming that macro definitions were
3948 scanned for macro replacements. I have been disabused of this notion,
3949 so now the code just undefines EX_OK before #including unistd.h.
3951 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3952 in the smtp transport. When a block could not be written in a single
3953 write() function, the timeout was being re-applied to each part-write.
3954 This seems wrong - if the receiver was accepting one byte at a time it
3955 would take for ever. The timeout is now adjusted when this happens. It
3956 doesn't have to be particularly precise.
3958 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3959 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3961 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3962 without a domain if the check to <postmaster@domain> fails.
3964 SC/01 Eximstats: added -xls and the ability to specify output files
3965 (patch written by Frank Heydlauf).
3967 SC/02 Eximstats: use FileHandles for outputting results.
3969 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3971 SC/04 Eximstats: fixed display of large numbers with -nvr option
3973 SC/05 Eximstats: fixed merging of reports with empty tables.
3975 SC/06 Eximstats: added the -include_original_destination flag
3977 SC/07 Eximstats: removed tabs and trailing whitespace.
3979 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3981 TK/06 MBOX spool code: Add real "From " MBOX separator line
3982 so the .eml file is really in mbox format (even though
3983 most programs do not really care). Patch from Alex Miller.
3985 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3986 The latter is generated from $received_to and is only set if the
3987 message has one envelope recipient. SA can use these headers,
3988 obviously out-of-the-box. Patch from Alex Miller.
3990 PH/08 The ${def test on a variable was returning false if the variable's
3991 value was "0", contrary to what the specification has always said!
3992 The result should be true unless the variable is empty.
3994 PH/09 The syntax error of a character other than { following "${if
3995 def:variable_name" (after optional whitespace) was not being diagnosed.
3996 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3997 accidental colon was present, for example, could give incorrect results.
3999 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4000 result is used (not including appendfile, where other changes are about
4003 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4004 This involved changing a lot of size variables from int to off_t. It
4005 should work with maildirs and everything.
4007 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4008 spamd dying while we are connected to it.
4010 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4013 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4014 function for BDB 4.3.
4016 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4018 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4019 the sources for PCRE have been re-arranged and more files are now
4022 PH/15 The code I had for printing potentially long long variables in PH/11
4023 above was not the best (it lost precision). The length of off_t variables
4024 is now inspected at build time, and an appropriate printing format (%ld
4025 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4026 to be "long long int" or "long int". This is needed for the internal
4027 formatting function string_vformat().
4029 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4030 the configuration file to be ":syslog", then the script "guesses" where
4031 the logs files are, rather than using the compiled in default. In our
4032 case the guess is not the same as the compiled default, so the script
4033 suddenly stopped working when I started to use syslog. The patch checks
4034 to see if log_file_path is "". If so, it attempts to read it from exim
4035 with no configuration file to get the compiled in version, before it
4036 falls back to the previous guessing code."
4038 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4039 implementing BATV in an Exim configuration. See NewStuff for the gory
4042 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4043 Makefile that are specific to HP-UX.
4045 PH/18 If the "use_postmaster" option was set for a recipient callout together
4046 with the "random" option, the postmaster address was used as the MAIL
4047 FROM address for the random test, but not for the subsequent recipient
4048 test. It is now used for both.
4050 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4051 patch removes a few documentation additions to RFC 3028, because the
4052 latest draft now contains them. It adds the new en;ascii-case comparator
4053 and a new error check for 8bit text in MIME parts. Comparator and
4054 require names are now matched exactly. I enabled the subaddress
4055 extension, but it is not well tested yet (read: it works for me)."
4057 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4058 rework some of the code of TK/09 above to avoid the hardwired use of
4059 "%lld" and "long long". Replaced the call to snprintf() with a call to
4062 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4063 records point to non-existent hosts", "retry timeout exceeded", and
4064 "retry time not reached for any host after a long failure period".
4066 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4067 experimental DomainKeys support:
4069 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4070 (2) On an error such as an illegally used "control", the wrong name for
4071 the control was given.
4073 These problems did NOT occur unless DomainKeys support was compiled.
4075 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4077 PH/24 Added ${if match_ip condition.
4079 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4080 databases so that it will be absolutely obvious if a crash occurs in the
4081 DB library. This is a regular occurrence (often caused by mis-matched
4084 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4085 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4086 picked up on a user's system that detects such things. There doesn't seem
4087 to be a gcc warning option for this - only an attribute that has to be
4088 put on the function's prototype. It seems that in Fedora Core 4 they have
4089 set this on a number of new functions. No doubt there will be more in due
4092 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4093 specify a transport. However, if an address that was verified by such a
4094 router was the subject of a callout, Exim crashed because it tried to
4095 read the rcpt_include_affixes from the non-existent transport. Now it
4096 just assumes that the setting of that option is false. This bug was
4097 introduced by 4.51/PH/31.
4099 PH/28 Changed -d+all to exclude +memory, because that information is very
4100 rarely of interest, but it makes the output a lot bigger. People tend to
4101 do -d+all out of habit.
4103 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4104 code in os-type was giving problems when libc.so lives in lib64, like on
4107 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4108 aren't the modern standard, and it seems that some systems' include files
4109 don't always have them. Exim was already checking for some of the newer
4110 ones like T_AAAA, and defining it itself. I've added checks for all the
4111 record types that Exim uses.
4113 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4114 not automatically generating a new one, as it is supposed to. This
4115 prevented TLS from working. If the file did exist, but contained invalid
4116 data, a new version was generated, as expected. It was only the case of a
4117 non-existent file that was broken.
4119 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4120 with a change in libdomainkeys > 0.64.
4122 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4123 from DNS. If the selector record carries the flag, it now has
4124 precedence over the domain-wide flag.
4126 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4128 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4129 the use of a port name that isn't defined in /etc/services) occurred, the
4130 message was deferred as in a normal delivery, and thus remained on the
4131 spool, instead of being failed because of the mua_wrapper setting. This
4132 is now fixed, and I tidied up some of the mua_wrapper messages at the
4135 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4136 the messages in a hash of arrays rather than using individual hashes.
4137 This is a bit cleaner and results in dramatic memory savings, albeit
4138 at a slight CPU cost.
4140 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4141 as requested by Marc Sherman.
4143 SC/10 Eximstats: added histograms for user specified patterns as requested
4146 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4148 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4149 fopen() in the content-scanning modules that did not already have it.
4155 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4156 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4158 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4160 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4162 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4163 to test Sieve filters that use "vacation".
4165 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4166 that changes the way the GnuTLS parameters are stored in the cache file.
4167 The new format can be generated externally. For backward compatibility,
4168 if the data in the cache doesn't make sense, Exim assumes it has read an
4169 old-format file, and it generates new data and writes a new file. This
4170 means that you can't go back to an older release without removing the
4173 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4174 work if there are any delivery delays because "one_time" forces the
4175 parent to be marked "delivered", so its unseen clone is never tried
4176 again. For this reason, Exim now forbids the simultaneous setting of
4179 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4180 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4181 entry if you want to know the details. The fix, however, neglected to
4182 consider the case when local delivery batching is involved. The test for
4183 "previously delivered" was not happening when checking to see if an
4184 address could be batched with a previous (undelivered) one; under
4185 certain circumstances this could lead to multiple deliveries to the same
4188 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4189 in its include files, and this causes problems building Exim.
4191 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4192 header_syntax) but Exim was just ignoring anything given after a slash.
4193 In particular, this caused confusion with an attempt to use "verify =
4194 reverse_host_lookup/defer_ok". An error is now given when options are
4195 supplied for verify items that do not have them. (Maybe reverse_host_
4196 lookup should have a defer_ok option, but that's a different point.)
4198 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4199 defined by RFC 821) to 2048, because there were problems with some AUTH
4200 commands, and RFC 1869 says the size should be increased for extended
4201 SMTP commands that take arguments.
4203 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4206 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4207 "unknown" error; now it says that the functionality isn't in the binary.
4209 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4210 an address' error message when a string expansion fails (syntax or
4211 whatever). Otherwise the password may appear in the log. Following change
4212 PH/42 below, there is no longer a chance of it appearing in a bounce
4215 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4217 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4218 of its addresses was ever tried. (Bugzilla bug #2.)
4220 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4221 the result incorrectly in the debug output. (It correctly added a newline
4222 to what was transported.)
4224 TF/01 Added $received_time.
4226 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4227 commented out examples of how to interface to a virus scanner and to
4228 SpamAssassin. Also added commented examples of av_scanner and
4229 spamd_address settings.
4231 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4232 and controls are allowed in which ACLs. There were a couple of minor
4233 errors. Some of the entries in the conditions table (which is a table of
4234 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4235 negation of where the condition IS allowed.
4237 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4239 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4240 header file does not have a version number, so I've had to invent a new
4241 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4242 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4243 radiusclient), but it was contributed by a Radius user.
4245 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4246 files or directories, for OpenSSL.
4248 PH/19 When an Exim process that is running as root has to create an Exim log
4249 file, it does so in a subprocess that runs as exim:exim so as to get the
4250 ownership right at creation (otherwise, other Exim processes might see
4251 the file with the wrong ownership). There was no test for failure of this
4252 fork() call, which would lead to the process getting stuck as it waited
4253 for a non-existent subprocess. Forks do occasionally fail when resources
4254 run out. I reviewed all the other calls to fork(); they all seem to check
4257 PH/20 When checking for unexpected SMTP input at connect time (before writing
4258 the banner), Exim was not dealing correctly with a non-positive return
4259 from the read() function. If the client had disconnected by this time,
4260 the result was a log entry for a synchronization error with an empty
4261 string after "input=" when read() returned zero. If read() returned -1
4262 (an event I could not check), uninitialized data bytes were printed.
4263 There were reports of junk text (parts of files, etc) appearing after
4266 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4268 PH/22 Added support for macro redefinition, and (re)definition in between
4269 driver and ACL definitions.
4271 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4272 forgetting to use the resulting value; it was using the unexpanded value.
4274 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4275 hadn't been configured. The fix is from Juergen Kreileder, who
4276 understands it better than I do:
4278 "Here's what I see happening with three configured cyrus_sasl
4279 authenticators configured (plain, login, cram-md5):
4281 On startup auth_cyrus_sasl_init() gets called for each of these.
4282 This means three calls to sasl_listmech() without a specified mech_list.
4283 => SASL tests which mechs of all available mechs actually work
4284 => three warnings about OTP not working
4285 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4287 With the patch, sasl_listmech() also gets called three times. But now
4288 SASL's mech_list option is set to the server_mech specified in the the
4289 authenticator. Or in other words, the answer from sasl_listmech()
4290 gets limited to just the mech you're testing for (which is different
4292 => the return list contains just 'plain' or 'login', 'cram-md5' or
4293 nothing depending on the value of ob->server_mech.
4295 I've just tested the patch: Authentication still works fine,
4296 unavailable mechs specified in the exim configuration are still
4297 caught, and the auth.log warnings about OTP are gone."
4299 PH/25 When debugging is enabled, the contents of the command line are added
4300 to the debugging output, even when log_selector=+arguments is not
4303 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4304 answer is "GNU", and only if the return is "GNU/something" is the answer
4307 PH/27 $acl_verify_message is now set immediately after the failure of a
4308 verification in an ACL, and so is available in subsequent modifiers. In
4309 particular, the message can be preserved by coding like this:
4311 warn !verify = sender
4312 set acl_m0 = $acl_verify_message
4314 Previously, $acl_verify_message was set only while expanding "message"
4315 and "log_message" when a very denied access.
4317 PH/28 Modified OS/os.c-Linux with
4319 -#ifndef OS_LOAD_AVERAGE
4320 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4322 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4323 nomenclature these days.)
4325 PH/29 Installed patch from the Sieve maintainer that adds the options
4326 sieve_useraddress and sieve_subaddress to the redirect router.
4328 PH/30 In these circumstances:
4329 . Two addresses routed to the same list of hosts;
4330 . First host does not offer TLS;
4331 . First host accepts first address;
4332 . First host gives temporary error to second address;
4333 . Second host offers TLS and a TLS session is established;
4334 . Second host accepts second address.
4335 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4336 and peerdn, if requested) that were in fact used only for the second
4339 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4340 attention to any local part prefix or suffix that was matched by the
4341 router that accepted the address. It now behaves in the same way as it
4342 does for delivery: the affixes are removed from the local part unless
4343 rcpt_include_affixes is set on the transport.
4345 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4346 timeout during the DATA phase of an incoming message.
4348 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4349 applied a patch sanctioned by the Sieve maintainer.
4351 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4352 the uid or gid is negative. A case of a negative gid caused this to be
4353 noticed. The fix allows for either to be negative.
4355 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4356 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4359 PH/36 The change PH/12 above was broken. Fixed it.
4361 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4362 the grounds that routing the same address twice would always produce the
4363 same answer. This might have been true once, but it is certainly no
4364 longer true now. Routing a child address may depend on the previous
4365 routing that produced that child. Some complicated redirection strategies
4366 went wrong when messages had multiple recipients, and made Exim's
4367 behaviour dependent on the order in which the addresses were given.
4369 I have moved the duplicate checking until after the routing is complete.
4370 Exim scans the addresses that are assigned to local and remote
4371 transports, and removes any duplicates. This means that more work will be
4372 done, as duplicates will always all be routed, but duplicates are
4373 presumably rare, so I don't expect this is of any significance.
4375 For deliveries to pipes, files, and autoreplies, the duplicate checking
4376 still happens during the routing process, since they are not going to be
4379 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4380 It corrects a timeout issue with spamd. This is Ian's comment: "The
4381 background is that sometimes spamd either never reads data from a
4382 connection it has accepted, or it never writes response data. The exiscan
4383 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4384 blindly assumes that writes won't block so it may never time out."
4386 PH/39 Allow G after quota size as well as K and M.
4388 PH/40 The value set for $authenticated_id in an authenticator may not contain
4389 binary zeroes or newlines because the value is written to log lines and
4390 to spool files. There was no check on this. Now the value is run through
4391 the string_printing() function so that such characters are converted to
4392 printable escape sequences.
4394 PH/41 $message_linecount is a new variable that contains the total number of
4395 lines in the message. Compare $body_linecount, which is the count for the
4398 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4399 bounce and delay warning messages, except in certain special cases, which
4402 (a) An SMTP error message from a remote host;
4403 (b) A message specified in a :fail: redirection;
4404 (c) A message specified in a "fail" command in a system filter;
4405 (d) A message specified in a FAIL return from the queryprogram router;
4406 (e) A message specified by the cannot_route_message router option.
4408 In these cases only, Exim does include the error details in bounce and
4409 warning messages. There are also a few cases where bland messages such
4410 as "unrouteable address" or "local delivery error" are given.
4412 PH/43 $value is now also set for the "else" part of a ${run expansion.
4414 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4415 being worked on, but at least Exim now implements the latest version to
4418 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4419 process to complete was treated as a delivery failure, a timeout while
4420 writing the message to the pipe was logged, but erroneously treated as a
4421 successful delivery. Such timeouts include transport filter timeouts. For
4422 consistency with the overall process timeout, these timeouts are now
4423 treated as errors, giving rise to delivery failures by default. However,
4424 there is now a new Boolean option for the pipe transport called
4425 timeout_defer, which, if set TRUE, converts the failures into defers for
4426 both kinds of timeout. A transport filter timeout is now identified in
4429 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4430 systems where "make" and "gmake" are different, calling "gmake" at top
4431 level broke things. I've arranged for the value of $(MAKE) to be passed
4432 from the Makefile to this script so that it can call the same version of
4436 A note about Exim versions 4.44 and 4.50
4437 ----------------------------------------
4439 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4440 changes of various kinds. As a consequence, a big documentation update was
4441 needed. This delayed the release for rather longer than seemed good, especially
4442 in the light of a couple of (minor) security issues. Therefore, the changes
4443 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4444 release. So 4.44 and 4.50 are in effect two different branches that both start
4447 I have left the 4.50 change log unchanged; it contains all the changes since
4448 4.43. The change log for 4.44 is below; many of its items are identical to
4449 those for 4.50. This seems to be the most sensible way to preserve the
4450 historical information.
4456 1. Minor wording change to the doc/README.SIEVE file.
4458 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4459 computation of the current number of files was incorrect.
4461 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4462 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4465 4. Give more explanation in the error message when the command for a transport
4466 filter fails to execute.
4468 5. There are several places where Exim runs a non-Exim command in a
4469 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4470 was being done only for the command run by the queryprogram router. It is
4471 now done for all such subprocesses. The other cases are: ${run, transport
4472 filters, and the commands run by the lmtp and pipe transports.
4474 6. Added CONFIGURE_GROUP build-time option.
4476 7. Some older OS have a limit of 256 on the maximum number of file
4477 descriptors. Exim was using setrlimit() to set 1000 as a large value
4478 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4479 systems. I've change it so that if it can't get 1000, it tries for 256.
4481 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4482 was an oversight, and furthermore, ever since the addition of extra
4483 controls (e.g. 4.43/32), the checks on when to allow different forms of
4484 "control" were broken. There should now be diagnostics for all cases when a
4485 control that does not make sense is encountered.
4487 9. Added the /retain_sender option to "control=submission".
4489 10. $recipients is now available in the predata ACL (oversight).
4491 11. Tidy the search cache before the fork to do a delivery from a message
4492 received from the command line. Otherwise the child will trigger a lookup
4493 failure and thereby defer the delivery if it tries to use (for example) a
4494 cached ldap connection that the parent has called unbind on.
4496 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4497 of $address_data from the recipient verification was clobbered by the
4498 sender verification.
4500 13. The value of address_data from a sender verification is now available in
4501 $sender_address_data in subsequent conditions in the ACL statement.
4503 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4505 15. Added a new option "connect=<time>" to callout options, to set a different
4508 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4509 was its contents. (It was OK if the option was not defined at all.)
4511 17. A "Completed" log line is now written for messages that are removed from
4512 the spool by the -Mrm option.
4514 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4515 information about exactly what failed.
4517 19. Added -dd to debug only the daemon process.
4519 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4520 handles timeouts, both on the server side and network timeouts. Renamed the
4521 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4523 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4524 It is now set to "smtps".
4526 22. $host_address is now set to the target address during the checking of
4527 ignore_target_hosts.
4529 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4530 being passed; this would have caused $sender_host_name to have been used if
4531 matching the list had actually called for a host name (not very likely,
4532 since this list is usually IP addresses). A host name is now passed as
4535 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4536 code that specifies a non-restarting handler (typically sigaction() in
4537 modern systems) in an attempt to fix a rare and obscure crash bug.
4539 25. Narrowed the window for a race in the daemon that could cause it to ignore
4540 SIGCHLD signals. This is not a major problem, because they are used only to
4541 wake it up if nothing else does.
4543 26. A malformed maildirsize file could cause Exim to calculate negative values
4544 for the mailbox size or file count. Odd effects could occur as a result.
4545 The maildirsize information is now recalculated if the size or filecount
4548 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4549 support for a long time. Removed HAVE_SYS_VFS_H.
4551 28. Installed the latest version of exipick from John Jetmore.
4553 29. In an address list, if the pattern was not a regular expression, an empty
4554 subject address (from a bounce message) matched only if the pattern was an
4555 empty string. Non-empty patterns were not even tested. This was the wrong
4556 because it is perfectly reasonable to use an empty address as part of a
4557 database query. An empty address is now tested by patterns that are
4558 lookups. However, all the other forms of pattern expect the subject to
4559 contain a local part and a domain, and therefore, for them, an empty
4560 address still always fails if the pattern is not itself empty.
4562 30. Exim went into a mad DNS loop when attempting to do a callout where the
4563 host was specified on an smtp transport, and looking it up yielded more
4564 than one IP address.
4566 31. Re-factored the code for checking spool and log partition space into a
4567 function that finds that data and another that does the check. The former
4568 is then used to implement four new variables: $spool_space, $log_space,
4569 $spool_inodes, and $log_inodes.
4571 32. The RFC2047 encoding function was originally intended for short strings
4572 such as real names; it was not keeping to the 75-character limit for
4573 encoded words that the RFC imposes. It now respects the limit, and
4574 generates multiple encoded words if necessary. To be on the safe side, I
4575 have increased the buffer size for the ${rfc2047: expansion operator from
4578 33. It is now permitted to omit both strings after an "if" condition; if the
4579 condition is true, the result is "true". As before, when the second string
4580 is omitted, a false condition yields an empty string. This makes it less
4581 cumbersome to write custom ACL and router conditions.
4583 34. Failure to deliver a bounce message always caused it to be frozen, even if
4584 there was an errors_to setting on the router. The errors_to setting is now
4587 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4588 canonical form (fully expanded) before being placed in
4589 $sender_host_address.
4591 36. The table in the code that translates DNS record types into text (T_A to
4592 "A" for instance) was missing entries for NS and CNAME. It is just possible
4593 that this could have caused confusion if both these types were looked up
4594 for the same domain, because the text type is used as part of Exim's
4595 per-process caching. But the chance of anyone hitting this buglet seems
4598 37. The dnsdb lookup has been extended in a number of ways.
4600 (1) There is a new type, "zns", which walks up the domain tree until it
4601 finds some nameserver records. It should be used with care.
4603 (2) There is a new type, "mxh", which is like "mx" except that it returns
4604 just the host names, not the priorities.
4606 (3) It is now possible to give a list of domains (or IP addresses) to be
4607 looked up. The behaviour when one of the lookups defers can be
4608 controlled by a keyword.
4610 (4) It is now possible to specify the separator character for use when
4611 multiple records are returned.
4613 38. The dnslists ACL condition has been extended: it is now possible to supply
4614 a list of IP addresses and/or domains to be looked up in a particular DNS
4617 39. Added log_selector=+queue_time_overall.
4619 40. When running the queue in the test harness, wait just a tad after forking a
4620 delivery process, to get repeatability of debugging output.
4622 41. Include certificate and key file names in error message when GnuTLS fails
4623 to set them up, because the GnuTLS error message doesn't include the name
4624 of the failing file when there is a problem reading it.
4626 42. Allow both -bf and -bF in the same test run.
4628 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4630 44. The "Exiscan patch" is now merged into the mainline Exim source.
4632 45. Sometimes the final signoff response after QUIT could fail to get
4633 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4634 < 0 before doing a fflush(). This bug looks as though it goes back to the
4635 introduction of TLS in release 3.20, but "sometimes" must have been rare
4636 because the tests only now provoked it.
4638 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4639 (this can affect the format of dates).
4641 47. exim_tidydb, when checking for the continued existence of a message for
4642 which it has found a message-specific retry record, was not finding
4643 messages that were in split spool directories. Consequently, it was
4644 deleting retry records that should have stayed in existence.
4646 48. Steve fixed some bugs in eximstats.
4648 49. The SPA authentication driver was not abandoning authentication and moving
4649 on to the next authenticator when an expansion was forced to fail,
4650 contradicting the general specification for all authenticators. Instead it
4651 was generating a temporary error. It now behaves as specified.
4653 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4654 (the order specifies the preference for clients). The order is now AES256,
4655 AES128, 3DES, ARCFOUR128.
4657 51. Small patch to Sieve code - explicitly set From: when generating an
4660 52. Exim crashed if a remote delivery caused a very long error message to be
4661 recorded - for instance if somebody sent an entire SpamAssassin report back
4662 as a large number of 550 error lines. This bug was coincidentally fixed by
4663 increasing the size of one of Exim's internal buffers (big_buffer) that
4664 happened as part of the Exiscan merge. However, to be on the safe side, I
4665 have made the code more robust (and fixed the comments that describe what
4668 53. Now that there can be additional text after "Completed" in log lines (if
4669 the queue_time_overall log selector is set), a one-byte patch to exigrep
4670 was needed to allow it to recognize "Completed" as not the last thing in
4673 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4674 patch that reportedly fixes this has been added. I am not expert enough to
4675 create a test for it. This is what the patch creator wrote:
4677 "I found a little strange behaviour of ldap code when working with
4678 Windows 2003 AD Domain, where users was placed in more than one
4679 Organization Units. When I tried to give exim partial DN, the exit code
4680 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4681 But simultaneously result of request was absolutely normal ldap result,
4682 so I produce this patch..."
4684 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4685 REFERENCE, so I have modified the code to exclude the patch when that macro
4688 55. Some experimental protocols are using DNS PTR records for new purposes. The
4689 keys for these records are domain names, not reversed IP addresses. The
4690 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4691 leaves it alone. Component reversal etc. now happens only for IP addresses.
4694 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4696 57. Double the size of the debug message buffer (to 2048) so that more of very
4697 long debug lines gets shown.
4699 58. The exicyclog utility now does better if the number of log files to keep
4700 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4702 59. Two changes related to the smtp_active_hostname option:
4704 (1) $smtp_active_hostname is now available as a variable.
4705 (2) The default for smtp_banner uses $smtp_active_hostname instead
4706 of $primary_hostname.
4708 60. The host_aton() function is supposed to be passed a string that is known
4709 to be a valid IP address. However, in the case of IPv6 addresses, it was
4710 not checking this. This is a hostage to fortune. Exim now panics and dies
4711 if the condition is not met. A case was found where this could be provoked
4712 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4713 components; fortuitously, this particular loophole had already been fixed
4714 by change 4.50/55 above.
4716 If there are any other similar loopholes, the new check in host_aton()
4717 itself should stop them being exploited. The report I received stated that
4718 data on the command line could provoke the exploit when Exim was running as
4719 exim, but did not say which command line option was involved. All I could
4720 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4721 running as the user.
4724 61. There was a buffer overflow vulnerability in the SPA authentication code
4725 (which came originally from the Samba project). I have added a test to the
4726 spa_base64_to_bits() function which I hope fixes it.
4729 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4730 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4732 63. The daemon start-up calls getloadavg() while still root for those OS that
4733 need the first call to be done as root, but it missed one case: when
4734 deliver_queue_load_max is set with deliver_drop_privilege. This is
4735 necessary for the benefit of the queue runner, because there is no re-exec
4736 when deliver_drop_privilege is set.
4738 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4739 This has been fixed.
4741 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4742 was in use, was not putting the data itself into the right store pool;
4743 consequently, it could be overwritten for a subsequent message in the same
4744 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4747 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4749 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4750 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4751 the function were treating the return as a boolean value, which happened to
4752 work because 0=false and not-0=true, but is not correct code.
4754 68. The host_aton() function was not handling scoped IPv6 addresses (those
4755 with, for example, "%eth0" on the end) correctly.
4757 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4758 negated items (that is, ~something) in unsigned ints. Some compilers
4759 apparently mutter when there is no cast.
4761 70. If an address verification called from an ACL failed, and did not produce a
4762 user-specific message (i.e. there was only a "system" message), nothing was
4763 put in $acl_verify_message. In this situation, it now puts the system
4766 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4767 session; change 4.31/43 added the unwanted input to the log line - except
4768 that it did not do this in the start of session case. It now does.
4770 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4771 This is wrong and can cause the other end to generate a synchronization
4772 error if it is another Exim or anything else that does the synchronization
4773 check. A QUIT command is no longer sent after a timeout.
4775 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4776 during host lookups.
4778 74. The defer_ok option of callout verification was not working if it was used
4779 when verifying addresses in header lines, that is, for this case:
4781 verify = header_sender/callout=defer_ok
4783 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4784 those file descriptors could be used for SMTP connections. If anything
4785 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4786 could be sent to the SMTP client, causing chaos. The daemon now opens
4787 stdin, stdout, and stderr to /dev/null when it puts itself into the
4790 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4791 log by default. The user can override this with suitable Perl magic.
4793 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4794 the log message when discard triggers, was not working for the DATA ACL or
4795 for the non-SMTP ACL.
4797 78. Error message wording change in sieve.c.
4799 79. If smtp_accept_max_per_host was set, the number of connections could be
4800 restricted to fewer than expected, because the daemon was trying to set up
4801 a new connection before checking whether the processes handling previous
4802 connections had finished. The check for completed processes is now done
4803 earlier. On busy systems, this bug wouldn't be noticed because something
4804 else would have woken the daemon, and it would have reaped the completed
4807 80. If a message was submitted locally by a user whose login name contained one
4808 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4809 It caused a spool format error. I have fixed the spool reading code. A
4810 related problem was that the "from" clause in the Received: line became
4811 illegal because of the space(s). It is now covered by ${quote_local_part.
4813 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4816 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4818 83. Patch from Sieve maintainer to fix unterminated string problem in
4819 "vacation" handling.
4821 84. Some minor changes to the Linux configuration files to help with other
4822 OS variants using glibc.
4824 85. One more patch for Sieve to update vacation handling to latest spec.
4827 ----------------------------------------------------
4828 See the note above about the 4.44 and 4.50 releases.
4829 ----------------------------------------------------
4835 1. Change 4.43/35 introduced a bug that caused file counts to be
4836 incorrectly computed when quota_filecount was set in an appendfile
4839 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4840 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4843 3. Give more explanation in the error message when the command for a transport
4844 filter fails to execute.
4846 4. There are several places where Exim runs a non-Exim command in a
4847 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4848 was being done only for the command run by the queryprogram router. It is
4849 now done for all such subprocesses. The other cases are: ${run, transport
4850 filters, and the commands run by the lmtp and pipe transports.
4852 5. Some older OS have a limit of 256 on the maximum number of file
4853 descriptors. Exim was using setrlimit() to set 1000 as a large value
4854 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4855 systems. I've change it so that if it can't get 1000, it tries for 256.
4857 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4858 was an oversight, and furthermore, ever since the addition of extra
4859 controls (e.g. 4.43/32), the checks on when to allow different forms of
4860 "control" were broken. There should now be diagnostics for all cases when a
4861 control that does not make sense is encountered.
4863 7. $recipients is now available in the predata ACL (oversight).
4865 8. Tidy the search cache before the fork to do a delivery from a message
4866 received from the command line. Otherwise the child will trigger a lookup
4867 failure and thereby defer the delivery if it tries to use (for example) a
4868 cached ldap connection that the parent has called unbind on.
4870 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4871 of $address_data from the recipient verification was clobbered by the
4872 sender verification.
4874 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4875 was its contents. (It was OK if the option was not defined at all.)
4877 11. A "Completed" log line is now written for messages that are removed from
4878 the spool by the -Mrm option.
4880 12. $host_address is now set to the target address during the checking of
4881 ignore_target_hosts.
4883 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4884 being passed; this would have caused $sender_host_name to have been used if
4885 matching the list had actually called for a host name (not very likely,
4886 since this list is usually IP addresses). A host name is now passed as
4889 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4890 code that specifies a non-restarting handler (typically sigaction() in
4891 modern systems) in an attempt to fix a rare and obscure crash bug.
4893 15. Narrowed the window for a race in the daemon that could cause it to ignore
4894 SIGCHLD signals. This is not a major problem, because they are used only to
4895 wake it up if nothing else does.
4897 16. A malformed maildirsize file could cause Exim to calculate negative values
4898 for the mailbox size or file count. Odd effects could occur as a result.
4899 The maildirsize information is now recalculated if the size or filecount
4902 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4903 support for a long time. Removed HAVE_SYS_VFS_H.
4905 18. Updated exipick to current release from John Jetmore.
4907 19. Allow an empty sender to be matched against a lookup in an address list.
4908 Previously the only cases considered were a regular expression, or an
4911 20. Exim went into a mad DNS lookup loop when doing a callout where the
4912 host was specified on the transport, if the DNS lookup yielded more than
4915 21. The RFC2047 encoding function was originally intended for short strings
4916 such as real names; it was not keeping to the 75-character limit for
4917 encoded words that the RFC imposes. It now respects the limit, and
4918 generates multiple encoded words if necessary. To be on the safe side, I
4919 have increased the buffer size for the ${rfc2047: expansion operator from
4922 22. Failure to deliver a bounce message always caused it to be frozen, even if
4923 there was an errors_to setting on the router. The errors_to setting is now
4926 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4927 canonical form (fully expanded) before being placed in
4928 $sender_host_address.
4930 24. Updated eximstats to version 1.33
4932 25. Include certificate and key file names in error message when GnuTLS fails
4933 to set them up, because the GnuTLS error message doesn't include the name
4934 of the failing file when there is a problem reading it.
4936 26. Expand error message when OpenSSL has problems setting up cert/key files.
4939 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4940 (this can affect the format of dates).
4942 28. exim_tidydb, when checking for the continued existence of a message for
4943 which it has found a message-specific retry record, was not finding
4944 messages that were in split spool directories. Consequently, it was
4945 deleting retry records that should have stayed in existence.
4947 29. eximstats updated to version 1.35
4948 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4949 1.35 - bugfix such that pie charts by volume are generated correctly
4951 30. The SPA authentication driver was not abandoning authentication and moving
4952 on to the next authenticator when an expansion was forced to fail,
4953 contradicting the general specification for all authenticators. Instead it
4954 was generating a temporary error. It now behaves as specified.
4956 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4957 (the order specifies the preference for clients). The order is now AES256,
4958 AES128, 3DES, ARCFOUR128.
4960 31. Small patch to Sieve code - explicitly set From: when generating an
4963 32. Exim crashed if a remote delivery caused a very long error message to be
4964 recorded - for instance if somebody sent an entire SpamAssassin report back
4965 as a large number of 550 error lines. This bug was coincidentally fixed by
4966 increasing the size of one of Exim's internal buffers (big_buffer) that
4967 happened as part of the Exiscan merge. However, to be on the safe side, I
4968 have made the code more robust (and fixed the comments that describe what
4971 33. Some experimental protocols are using DNS PTR records for new purposes. The
4972 keys for these records are domain names, not reversed IP addresses. The
4973 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4974 leaves it alone. Component reversal etc. now happens only for IP addresses.
4977 34. The host_aton() function is supposed to be passed a string that is known
4978 to be a valid IP address. However, in the case of IPv6 addresses, it was
4979 not checking this. This is a hostage to fortune. Exim now panics and dies
4980 if the condition is not met. A case was found where this could be provoked
4981 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4982 components; fortuitously, this particular loophole had already been fixed
4983 by change 4.50/55 or 4.44/33 above.
4985 If there are any other similar loopholes, the new check in host_aton()
4986 itself should stop them being exploited. The report I received stated that
4987 data on the command line could provoke the exploit when Exim was running as
4988 exim, but did not say which command line option was involved. All I could
4989 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4990 running as the user.
4993 35. There was a buffer overflow vulnerability in the SPA authentication code
4994 (which came originally from the Samba project). I have added a test to the
4995 spa_base64_to_bits() function which I hope fixes it.
4998 36. The daemon start-up calls getloadavg() while still root for those OS that
4999 need the first call to be done as root, but it missed one case: when
5000 deliver_queue_load_max is set with deliver_drop_privilege. This is
5001 necessary for the benefit of the queue runner, because there is no re-exec
5002 when deliver_drop_privilege is set.
5004 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5005 was in use, was not putting the data itself into the right store pool;
5006 consequently, it could be overwritten for a subsequent message in the same
5007 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5010 38. Sometimes the final signoff response after QUIT could fail to get
5011 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5012 < 0 before doing a fflush(). This bug looks as though it goes back to the
5013 introduction of TLS in release 3.20, but "sometimes" must have been rare
5014 because the tests only now provoked it.
5020 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5021 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5022 Now it returns an integer. A number of calls were still expecting a T/F
5023 return. Fortuitously, in all cases, the tests worked in OK situations,
5024 which is the norm. However, things would have gone wrong on any write
5025 failures on the smtp file descriptor. This function is used when sending
5026 messages over SMTP and also when doing verify callouts.
5028 2. When Exim is called to do synchronous delivery of a locally submitted
5029 message (the -odf or -odi options), it no longer closes stderr before doing
5032 3. Implemented the mua_wrapper option.
5034 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5036 5. Implemented the functions header_remove(), header_testname(),
5037 header_add_at_position(), and receive_remove_recipient(), and exported them
5040 6. If an ACL "warn" statement specified the addition of headers, Exim already
5041 inserted X-ACL-Warn: at the start if there was no header name. However, it
5042 was not making this test for the second and subsequent header lines if
5043 there were newlines in the string. This meant that an invalid header could
5044 be inserted if Exim was badly configured.
5046 7. Allow an ACL "warn" statement to add header lines at the start or after all
5047 the Received: headers, as well as at the end.
5049 8. Added the rcpt_4xx retry error code.
5051 9. Added postmaster_mailfrom=xxx to callout verification option.
5053 10. Added mailfrom=xxxx to the callout verification option, for verify=
5056 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5057 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5059 12. Inserted some casts to stop certain compilers warning when using pointer
5060 differences as field lengths or precisions in printf-type calls (mostly
5061 affecting debugging statements).
5063 13. Added optional readline() support for -be (dynamically loaded).
5065 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5066 same clock tick as a message's arrival, so that its received time was the
5067 same as the "first fail" time on the retry record, and that message
5068 remained on the queue past the ultimate address timeout, every queue runner
5069 would try a delivery (because it was past the ultimate address timeout) but
5070 after another failure, the ultimate address timeout, which should have then
5071 bounced the address, did not kick in. This was a "< instead of <=" error;
5072 in most cases the first failure would have been in the next clock tick
5073 after the received time, and all would be well.
5075 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5076 being recognized when the domain list was tested by the match_domain
5077 condition in an expansion string.
5079 16. Added the ${str2b64: operator.
5081 17. Exim was always calling setrlimit() to set a large limit for the number of
5082 processes, without checking whether the existing limit was already
5083 adequate. (It did check for the limit on file descriptors.) Furthermore,
5084 errors from getrlimit() and setrlimit() were being ignored. Now they are
5085 logged to the main and panic logs, but Exim does carry on, to try to do its
5086 job under whatever limits there are.
5088 18. Imported PCRE 5.0.
5090 19. Trivial typo in log message " temporarily refused connection" (the leading
5093 20. If the log selector return_path_on_delivery was set and an address was
5094 redirected to /dev/null, the delivery process crashed because it assumed
5095 that a return path would always be set for a "successful" delivery. In this
5096 case, the whole delivery is bypassed as an optimization, and therefore no
5099 21. Internal re-arrangement: the function for sending a challenge and reading
5100 a response while authentication was assuming a zero-terminated challenge
5101 string. It's now changed to take a pointer and a length, to allow for
5102 binary data in such strings.
5104 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5106 23. Exim was not respecting finduser_retries when seeking the login of the
5107 uid under which it was called; it was always trying 10 times. (The default
5108 setting of finduser_retries is zero.) Also, it was sleeping after the final
5109 failure, which is pointless.
5111 24. Implemented tls_on_connect_ports.
5113 25. Implemented acl_smtp_predata.
5115 26. If the domain in control=submission is set empty, Exim assumes that the
5116 authenticated id is a complete email address when it generates From: or
5117 Sender: header lines.
5119 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5120 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5121 chown and chgrp in /bin and hostname in /usr/bin.
5123 28. Exim was keeping the "process log" file open after each use, just as it
5124 does for the main log. This opens the possibility of it remaining open for
5125 long periods when the USR1 signal hits a daemon. Occasional processlog
5126 errors were reported, that could have been caused by this. Anyway, it seems
5127 much more sensible not to leave this file open at all, so that is what now
5130 29. The long-running daemon process does not normally write to the log once it
5131 has entered its main loop, and it closes the log before doing so. This is
5132 so that log files can straightforwardly be renamed and moved. However,
5133 there are a couple of unusual error situations where the daemon does write
5134 log entries, and I had neglected to close the log afterwards.
5136 30. The text of an SMTP error response that was received during a remote
5137 delivery was being truncated at 512 bytes. This is too short for some of
5138 the long messages that one sometimes sees. I've increased the limit to
5141 31. It is now possible to make retry rules that apply only when a message has a
5142 specific sender, in particular, an empty sender.
5144 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5145 it possible to be selective about when SMTP synchronization is enforced.
5147 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5149 32. Implemented hosts_connection_nolog.
5151 33. Added an ACL for QUIT.
5153 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5156 35. Added mailbox_size and mailbox_filecount to appendfile.
5158 36. Added control = no_multiline_responses to ACLs.
5160 37. There was a bug in the logic of the code that waits for the clock to tick
5161 in the case where the clock went backwards by a substantial amount such
5162 that the microsecond fraction of "now" was more than the microsecond
5163 fraction of "then" (but the whole seconds number was less).
5165 38. Added support for the libradius Radius client library this is found on
5166 FreeBSD (previously only the radiusclient library was supported).
5172 1. When certain lookups returned multiple values in the form name=value, the
5173 quoting of the values was not always being done properly. Specifically:
5174 (a) If the value started with a double quote, but contained no whitespace,
5176 (b) If the value contained whitespace other than a space character (i.e.
5177 tabs or newlines or carriage returns) it was not quoted.
5178 This fix has been applied to the mysql and pgsql lookups by writing a
5179 separate quoting function and calling it from the lookup code. The fix
5180 should probably also be applied to nisplus, ibase and oracle lookups, but
5181 since I cannot test any of those, I have not disturbed their existing code.
5183 2. A hit in the callout cache for a specific address caused a log line with no
5184 reason for rejecting RCPT. Now it says "Previous (cached) callout
5185 verification failure".
5187 3. There was an off-by-one bug in the queryprogram router. An over-long
5188 return line was truncated at 256 instead of 255 characters, thereby
5189 overflowing its buffer with the terminating zero. As well as fixing this, I
5190 have increased the buffer size to 1024 (and made a note to document this).
5192 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5193 when Exim is waiting for an SMTP response from a remote server, Exim
5194 restarts its select() call on the socket, thereby resetting its timeout.
5195 This is not a problem when such interrupts are rare. Somebody set up a cron
5196 job to run exiwhat every 2 minutes, which is less than the normal select()
5197 timeout (5 or 10 minutes). This meant that the select() timeout never
5198 kicked in because it was always reset. I have fixed this by comparing the
5199 time when an interrupt arrives with the time at the start of the first call
5200 to select(). If more time than the timeout has elapsed, the interrupt is
5201 treated as a timeout.
5203 5. Some internal re-factoring in preparation for the addition of Sieve
5204 extensions (by MH). In particular, the "personal" test is moved to a
5205 separate function, and given an option for scanning Cc: and Bcc: (which is
5206 not set for Exim filters).
5208 6. When Exim created an email address using the login of the caller as the
5209 local part (e.g. when creating a From: or Sender: header line), it was not
5210 quoting the local part when it contained special characters such as @.
5212 7. Installed new OpenBSD configuration files.
5214 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5215 try to make them clearer.
5217 9. Callout options, other than the timeout value, were being ignored when
5218 verifying sender addresses in header lines. For example, when using
5220 verify = header_sender/callout=no_cache
5222 the cache was (incorrectly) being used.
5224 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5225 only the Cygwin environment.
5227 11. When return_path_on_delivery was set as a log selector, if different remote
5228 addresses in the same message used different return paths and parallel
5229 remote delivery occurred, the wrong values would sometimes be logged.
5230 (Whenever a remote delivery process finished, the return path value from
5231 the most recently started remote delivery process was logged.)
5233 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5234 lines when AUTH and/or TLS are in use. This is the "received protocol"
5235 field. Exim used to use "asmtp" for authenticated SMTP, without any
5236 indication (in the protocol name) for TLS use. Now it follows the RFC and
5237 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5238 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5239 names appear in log lines as well as in Received: header lines.
5241 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5242 extensions, and comparison tests, and to fix some bugs.
5244 14. Changes to the "personal" filter test:
5246 (1) The test was buggy in that it was just doing the equivalent of
5247 "contains" tests on header lines. For example, if a user's address was
5248 anne@some.where, the "personal" test would incorrectly be true for
5250 To: susanne@some.where
5252 This test is now done by extracting each address from the header in turn,
5253 and checking the entire address. Other tests that are part of "personal"
5254 are now done using regular expressions (for example, to check local parts
5255 of addresses in From: header lines).
5257 (2) The list of non-personal local parts in From: addresses has been
5258 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5259 taken from the Sieve specification recommendations.
5261 (3) If the message contains any header line starting with "List-" it is
5262 treated as non-personal.
5264 (4) The test for "circular" in the Subject: header line has been removed
5265 because it now seems ill-conceived.
5267 15. Minor typos in src/EDITME comments corrected.
5269 16. Installed latest exipick from John Jetmore.
5271 17. If headers_add on a router specified a text string that was too long for
5272 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5273 of string_sprintf() is now avoided.
5275 18. $message_body_size was not set (it was always zero) when running the DATA
5276 ACL and the local_scan() function.
5278 19. For the "mail" command in an Exim filter, no default was being set for
5279 the once_repeat time, causing a random time value to be used if "once" was
5280 specified. (If the value happened to be <= 0, no repeat happened.) The
5281 default is now 0s, meaning "never repeat". The "vacation" command was OK
5282 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5283 (I found it when inspecting the code).
5285 20. There is now an overall timeout for performing a callout verification. It
5286 defaults to 4 times the callout timeout, which applies to individual SMTP
5287 commands during the callout. The overall timeout applies when there is more
5288 than one host that can be tried. The timeout is checked before trying the
5289 next host. This prevents very long delays if there are a large number of
5290 hosts and all are timing out (e.g. when the network connections are timing
5291 out). The value of the overall timeout can be changed by specifying an
5292 additional sub-option for "callout", called "maxwait". For example:
5294 verify = sender/callout=5s,maxwait=20s
5296 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5297 to the end before writing, but this should make it even safer).
5299 22. Exim was forgetting that it had advertised PIPELINING for the second and
5300 subsequent messages on an SMTP connection. It was also not resetting its
5301 memory on STARTTLS and an internal HELO.
5303 23. When Exim logs an SMTP synchronization error within a session, it now
5304 records whether PIPELINING has been advertised or not.
5306 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5307 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5308 rather than long int.
5310 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5312 26. Added the never_mail option to autoreply.
5318 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5319 crash if the getsockname() call failed; this can happen if a connection is
5320 closed very soon after it is established. The problem was simply in the
5321 order in which certain operations were done, causing Exim to try to write
5322 to the SMTP stream before it had set up the file descriptor. The bug has
5323 been fixed by making things happen in the correct order.
5329 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5330 before the connection was closed, thus losing the rejection response.
5332 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5333 some early Solaris releases, but causes trouble in current releases where
5334 socklen_t is defined.
5336 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5339 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5342 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5343 that releasing the top of it at the end releases what was used for sub-
5344 expansions (unless the block got too big). However, discard this block if
5345 the first thing is a variable or header, so that we can use its block when
5346 it is dynamic (useful for very large $message_headers, for example).
5348 6. Lookups now cache *every* query, not just the most recent. A new, separate
5349 store pool is used for this. It can be recovered when all lookup caches are
5350 flushed. Lookups now release memory at the end of their result strings.
5351 This has involved some general refactoring of the lookup sources.
5353 7. Some code has been added to the store_xxx() functions to reduce the amount
5354 of flapping under certain conditions.
5356 8. log_incoming_interface used to affect only the <= reception log lines. Now
5357 it causes the local interface and port to be added to several more SMTP log
5358 lines, for example "SMTP connection from", and rejection lines.
5360 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5362 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5364 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5365 could be overwritten at the end of the current message (or the start of a
5366 new message if it was set in a HELO ACL). The value is now preserved for
5367 the duration of the SMTP connection.
5369 12. If a transport had a headers_rewrite setting, and a matching header line
5370 contained an unqualified address, that address was qualified, even if it
5371 did not match any rewriting rules. The underlying bug was that the values
5372 of the flags that permit the existence of unqualified sender and recipient
5373 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5374 non-local messages, and by -bnq for local messages) were not being
5375 preserved with the message after it was received.
5377 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5378 "next input=" as part of the text comprising the host identity instead of
5379 the correct text. The code was using the same buffer for two different
5380 strings. However, depending on which order the printing function evaluated
5381 its arguments, the bug did not always show up. Under Linux, for example, my
5382 test suite worked just fine.
5384 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5385 doesn't work with some older versions of Perl. It has been changed to "my",
5386 which in any case is probably the better facility to use.
5388 15. A really picky compiler found some instances of statements for creating
5389 error messages that either had too many or two few arguments for the format
5392 16. The size of the buffer for calls to the DNS resolver has been increased
5393 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5394 for addresses that have a lot of PTR records. This alleviates a problem; it
5395 does not fully solve it.
5397 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5398 buffer now truncates the list and logs the incident, which is the same
5399 action as happens when Exim is looking up a host name and its aliases.
5400 Previously in this situation something unpredictable would happen;
5401 sometimes it was "internal error: store_reset failed".
5403 18. If a server dropped the connection unexpectedly when an Exim client was
5404 using GnuTLS and trying to read a response, the client delivery process
5405 crashed while trying to generate an error log message.
5407 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5408 string, for example:
5410 warn message = H1: something\nH2: something
5412 the text was added as a single header line from Exim's point of view
5413 though it ended up OK in the delivered message. However, searching for the
5414 second and subsequent header lines using $h_h2: did not work. This has been
5415 fixed. Similarly, if a system filter added multiple headers in this way,
5416 the routers could not see them.
5418 20. Expanded the error message when iplsearch is called with an invalid key to
5419 suggest using net-iplsearch in a host list.
5421 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5422 ACLs are no longer actually imposed (and a message to that effect is
5425 22. If a "gecos" field in a passwd entry contained escaped characters, in
5426 particular, if it contained a \" sequence, Exim got it wrong when building
5427 a From: or a Sender: header from that name. A second bug also caused
5428 incorrect handling when an unquoted " was present following a character
5429 that needed quoting.
5431 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5432 was not being matched caselessly.
5434 24. Arranged for all hyphens in the exim.8 source to be escaped with
5437 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5438 properly. Recipient callout cache records were still being keyed to include
5439 the sender, even when use_sender was set false. This led to far more
5440 callouts that were necessary. The sender is no longer included in the key
5441 when use_sender is false.
5443 26. Added "control = submission" modifier to ACLs.
5445 27. Added the ${base62d: operator to decode base 62 numbers.
5447 28. dnsdb lookups can now access SRV records.
5449 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5450 the configuration file.
5452 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5453 (-v) mode. This makes the output for a verbose queue run more intelligible.
5455 31. Added a use_postmaster feature to recipient callouts.
5457 32. Added the $body_zerocount variable, containing the number of binary zero
5458 bytes in the message body.
5460 33. The time of last modification of the "new" subdirectory is now used as the
5461 "mailbox time last read" when there is a quota error for a maildir
5464 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5466 35. Added +ignore_unknown as a special item in host lists.
5468 36. Code for decoding IPv6 addresses in host lists is now included, even if
5469 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5470 address was recognized as an IP address, but was then not correctly decoded
5471 into binary, causing unexpected and incorrect effects when compared with
5478 1. Very minor rewording of debugging text in manualroute to say "list of
5479 hosts" instead of "hostlist".
5481 2. If verify=header_syntax was set, and a header line with an unqualified
5482 address (no domain) and a large number of spaces between the end of the
5483 name and the colon was received, the reception process suffered a buffer
5484 overflow, and (when I tested it) crashed. This was caused by some obsolete
5485 code that should have been removed. The fix is to remove it!
5487 3. When running in the test harness, delay a bit after writing a bounce
5488 message to get a bit more predictability in the log output.
5490 4. Added a call to search_tidyup() just before forking a reception process. In
5491 theory, someone could use a lookup in the expansion of smtp_accept_max_
5492 per_host which, without the tidyup, could leave open a database connection.
5494 5. Added the variables $recipient_data and $sender_data which get set from a
5495 lookup success in an ACL "recipients" or "senders" condition, or a router
5496 "senders" option, similar to $domain_data and $local_part_data.
5498 6. Moved the writing of debug_print from before to after the "senders" test
5501 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5502 problems for message scanning, either using a data ACL, or using
5503 local_scan() because the Received: header was not generated till after they
5504 were called (in order to set the time as the time of reception completion).
5505 I have revised the way this works. The header is now generated after the
5506 body is received, but before the ACL or local_scan() are called. After they
5507 are run, the timestamp in the header is updated.
5513 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5514 before starting a queue runner without re-exec. This happened only when
5515 deliver_drop_privilege was set or when the Exim user was set to root. The
5516 effect of the bug was that timeouts during subsequent deliveries caused
5517 crashes instead of being properly handled. The handler is now left at its
5518 default (and expected) setting.
5520 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5521 message, again when deliver_drop_privilege is set or Exim is run as root.
5522 The bug described in (1) was not present in this case, but the tidying up
5523 of the other signals was missing. I have made the two cases consistent.
5525 3. The ignore_target_hosts setting on a manualroute router was being ignored
5526 for hosts that were looked up using the /MX notation.
5528 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5531 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5532 operated on the sender address. After changing the $sender_address to <>
5533 for the sender address verify, Exim was re-instated it as the original
5534 (before rewriting) address, but remembering that it had rewritten it, so it
5535 wasn't rewriting it again. This bug also had the effect of breaking the
5536 sender address verification caching when the sender address was rewritten.
5538 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5539 This has been changed so that if the ip literal address matches
5540 ignore_target_hosts, the router declines.
5542 7. Added expansion conditions match_domain, match_address, and match_local_
5543 part (NOT match_host).
5545 8. The placeholder for the Received: header didn't have a length field set.
5547 9. Added code to Exim itself and to exim_lock to test for a specific race
5548 condition that could lead to file corruption when using MBX delivery. The
5549 issue is with the lockfile that is created in /tmp. If this file is removed
5550 after a process has opened it but before that process has acquired a lock,
5551 there is the potential for a second process to recreate the file and also
5552 acquire a lock. This could lead to two Exim processes writing to the file
5553 at the same time. The added code performs the same test as UW imapd; it
5554 checks after acquiring the lock that its file descriptor still refers to
5555 the same named file.
5557 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5558 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5561 11. Added the smtp_active_hostname option. If used, this will typically be made
5562 to depend on the incoming interface address. Because $interface_address is
5563 not set up until the daemon has forked a reception process, error responses
5564 that can happen earlier (such as "too many connections") no longer contain
5567 12. If an expansion in a condition on a "warn" statement fails because a lookup
5568 defers, the "warn" statement is abandoned, and the next ACL statement is
5569 processed. Previously this caused the whole ACL to be aborted.
5571 13. Added the iplsearch lookup type.
5573 14. Added ident_timeout as a log selector.
5575 15. Added tls_certificate_verified as a log selector.
5577 16. Added a global option tls_require_ciphers (compare the smtp transport
5578 option of the same name). This controls incoming TLS connections.
5580 17. I finally figured out how to make tls_require_ciphers do a similar thing
5581 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5582 before starting the TLS session.
5584 18. Tabs are now shown as \t in -bP output.
5586 19. If the log selector return_path_on_delivery was set, Exim crashed when
5587 bouncing a message because it had too many Received: header lines.
5589 20. If two routers both had headers_remove settings, and the first one included
5590 a superfluous trailing colon, the final name in the first list and the
5591 first name in the second list were incorrectly joined into one item (with a
5592 colon in the middle).
5598 1. Added -C and -D options to the exinext utility, mainly to make it easier
5599 to include in the automated testing, but these could be helpful when
5600 multiple configurations are in use.
5602 2. The exinext utility was not formatting the output nicely when there was
5603 an alternate port involved in the retry record key, nor when there was a
5604 message id as well (for retries that were specific to a specific message
5605 and a specific host). It was also confused by IPv6 addresses, because of
5606 the additional colons they contain. I have fixed the IPv4 problem, and
5607 patched it up to do a reasonable job for IPv6.
5609 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5610 delivery, the log line now contains "pipelined" if PIPELINING was used.
5612 4. An SMTP transport process used to panic and die if the bind() call to set
5613 an explicit outgoing interface failed. This has been changed; it is now
5614 treated in the same way as a connect() failure.
5616 5. A reference to $sender_host_name in the part of a conditional expansion
5617 that was being skipped was still causing a DNS lookup. This no longer
5620 6. The def: expansion condition was not recognizing references to header lines
5621 that used bh_ and bheader_.
5623 7. Added the _cache feature to named lists.
5625 8. The code for checking quota_filecount in the appendfile transport was
5626 allowing one more file than it should have been.
5628 9. For compatibility with Sendmail, the command line option
5636 and sets the incoming protocol and host name (for trusted callers). The
5637 host name and its colon can be omitted when only the protocol is to be set.
5638 Note the Exim already has two private options, -pd and -ps, that refer to
5639 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5640 "s", but I don't think that's a major issue.
5642 10. A number of refactoring changes to the code, none of which should affect
5645 (a) The number of logging options was getting close to filling up the
5646 32-bit word that was used as a bit map. I have split them into two classes:
5647 those that are passed in the argument to log_write(), and those that are
5648 only ever tested independently outside of that function. These are now in
5649 separate 32-bit words, so there is plenty of room for expansion again.
5650 There is no change in the user interface or the logging behaviour.
5652 (b) When building, for example, log lines, the code previously used a
5653 macro that called string_cat() twice, in order to add two strings. This is
5654 not really sufficiently general. Furthermore, there was one instance where
5655 it was actually wrong because one of the argument was used twice, and in
5656 one call a function was used. (As it happened, calling the function twice
5657 did not affect the overall behaviour.) The macro has been replaced by a
5658 function that can join an arbitrary number of extra strings onto a growing
5661 (c) The code for expansion conditions now uses a table and a binary chop
5662 instead of a serial search (which was left over from when there were very
5663 few conditions). Also, it now recognizes conditions like "pam" even when
5664 the relevant support is not compiled in: a suitably worded error message is
5665 given if an attempt is made to use such a condition.
5667 11. Added ${time_interval:xxxxx}.
5669 12. A bug was causing one of the ddress fields not to be passed back correctly
5670 from remote delivery subprocesses. The field in question was not being
5671 subsequently used, so this caused to problems in practice.
5673 13. Added new log selectors queue_time and deliver_time.
5675 14. Might have fixed a bug in maildirsizefile handling that threw up
5676 "unexpected character" debug warnings, and recalculated the data
5677 unnecessarily. In any case, I expanded the warning message to give more
5680 15. Added the message "Restricted characters in address" to the statements in
5681 the default ACL that block characters like @ and % in local parts.
5683 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5684 Three changes have been made:
5686 (a) There was a serious bug; a negative response to MAIL caused the whole
5687 recipient domain to be cached as invalid, thereby blocking all messages
5688 to all local parts at the same domain, from all senders. This bug has
5689 been fixed. The domain is no longer cached after a negative response to
5690 MAIL if the sender used is not empty.
5692 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5695 (c) A new callout option, "use_sender" has been added for people who want
5696 the modified behaviour.
5702 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5705 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5706 indeed breaks things for older releases.
5708 3. Added additional logging to the case where there is a problem reading data
5709 from a filter that is running in a subprocess using a pipe, in order to
5710 try to track down a specific problem.
5712 4. Testing facility fudge: when running in the test harness and attempting
5713 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5714 getting "No route to host". Convert this to a timeout.
5716 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5719 6. Some OS don't have socklen_t but use size_t instead. This affects the
5720 fifth argument of getsockopt() amongst other things. This is now
5721 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5722 can be set for individual OS. I have set it for SunOS5, OSF1, and
5723 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5724 some earlier ones do not.
5726 7. Change 4.30/15 was not doing the test caselessly.
5728 8. The standard form for an IPv6 address literal was being rejected by address
5729 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5730 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5731 this, as well as the form without the "IPv6" on the front (but only when
5732 address literals are enabled, of course).
5734 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5736 10. Exim crashed if a message with an empty sender address specified by -f
5737 encountered a router with an errors_to setting. This could be provoked only
5738 by a command such as
5742 where an empty string was supplied; "<>" did not hit this bug.
5744 11. Installed PCRE release 4.5.
5746 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5747 remained set. It is now erased.
5749 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5750 times from message ids (which are base 36 rather than the normal 62).
5752 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5753 were being counted as actual protocol errors, and logged if the log
5754 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5755 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5756 and DATA following a set of rejected RCPTs do not count as protocol errors.
5757 In other words, Exim assumes they were pipelined, though this may not
5758 actually be the case. Of course, in all cases the client gets an
5759 appropriate error code.
5761 15. If a lookup fails in an ACL condition, a message about the failure may
5762 be available; it is used if testing the ACL cannot continue, because most
5763 such messages specify what the cause of the deferral is. However, some
5764 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5765 that caused an old message to be retained and used if a later statement
5766 caused a defer, replacing the real cause of the deferral.
5768 16. If an IP address had so many PTR records that the DNS lookup buffer
5769 was not large enough to hold them, Exim could crash while trying to process
5770 the truncated data. It now detects and logs this case.
5772 17. Further to 4.21/58, another change has been made: if (and only if) the
5773 first line of a message (the first header line) ends with CRLF, a bare LF
5774 in a subsequent header line has a space inserted after it, so as not to
5775 terminate the header.
5777 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5778 unnecessarily, used atoi() instead of strtol(), and didn't check the
5779 termination when getting file sizes from file names by regex.
5781 19. Completely re-implemented the support for maildirsize files, in the light
5782 of a number of problems with the previous contributed implementation
5783 (4.30/29). In particular:
5785 . If the quota is zero, the maildirsize file is maintained, but no quota is
5788 . If the maildir directory does not exist, it is created before any attempt
5789 to write a maildirsize file.
5791 . The quota value in the file is just a cache; if the quota is changed in
5792 the transport, the new value overrides.
5794 . A regular expression is available for excluding directories from the
5797 20. The autoreply transport checks the characters in options that define the
5798 message's headers; it allows continued headers, but it was checking with
5799 isspace() after an embedded newline instead of explicitly looking for a
5802 21. If all the "regular" hosts to which an address was routed had passed their
5803 expiry times, and had not reached their retry times, the address was
5804 bounced, even if fallback hosts were defined. Now Exim should go on to try
5807 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5808 equivalent code in the SMTP transport. Some hosts send humungous responses
5809 to HELO/EHLO, more than 1024 it seems.
5811 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5812 gives compiler warnings in some environments. I've now done it "properly",
5815 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5816 (because of problems with the built-in one) was declared to return uschar *
5817 instead of char *, causing compiler failure.
5819 25. Fixed a file descriptor leak when processing alias/forward files.
5821 26. Fixed a minor format string issue in dbfn.c.
5823 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5825 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5826 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5829 29. When a sender address is verified, it is cached, to save repeating the test
5830 when there is more than one recipient in a message. However, when the
5831 verification involves a callout, it is possible for different callout
5832 options to be set for different recipients. It is too complicated to keep
5833 track of this in the cache, so now Exim always runs a verification when a
5834 callout is required, relying on the callout cache for the optimization.
5835 The overhead is duplication of the address routing, but this should not be
5838 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5839 to be verified with callout=postmaster, and the main callout worked but the
5840 postmaster check failed, the verification correctly failed. However, if a
5841 subsequent RCPT command asked for sender verification *without* the
5842 postmaster check, incorrect caching caused this verification also to fail,
5845 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5846 it was not caching the DNS options (qualify_single, search_parents) that
5847 were used when the lookup failed. A subsequent lookup with different
5848 options therefore always gave the same answer, though there were cases
5849 where it should not have. (Example: a "domains = !$mx_any" option on a
5850 dnslookup router: the "domains" option is always processed without any
5851 widening, but the router might have qualify_single set.) Now Exim uses the
5852 cached value only when the same options are set.
5854 32. Added John Jetmore's "exipick" utility to the distribution.
5856 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5857 than a timeout (e.g. a certificate is required, and is not provided), an
5858 Exim server now closes the connection immediately. Previously it waited for
5859 the client to close - but if the client is SSL, it seems that they each
5860 wait for each other, leading to a delay before one of them times out.
5862 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5863 maintained 0.8.x compatibility because I don't think many are using it, and
5864 it is clearly obsolete.
5866 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5869 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5870 client certificate was expired. A simple patch fixes this, though I don't
5871 understand the full logic of why the verify callback is called multiple
5874 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5875 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5876 which causes problems with some clients (such as the Certicom SSL Plus
5877 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5878 disables the coutermeasure allowing Eudora to connect."
5880 38. Exim was not checking that a write() to a log file succeeded. This could
5881 lead to Bad Things if a log got too big, in particular if it hit a file
5882 size limit. Exim now panics and dies if it cannot write to a log file, just
5883 as it does if it cannot open a log file.
5885 39. Modified OS/Makefile-Linux so that it now contains
5887 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5889 The two -D definitions ensure that Exim is compiled with large file
5890 support, which makes it possible to handle log files that are bigger than
5893 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5894 instance) a domain was checked against a named list that involved a lookup,
5895 causing $domain_data to be set, then another domain was checked against the
5896 same list, then the first domain was re-checked, the value of $domain_data
5897 after the final check could be wrong. In particular, if the second check
5898 failed, it could be set empty. This bug probably also applied to
5901 41. The strip_trailing_dot option was not being applied to the address given
5902 with the -f command-line option.
5904 42. The code for reading a message's header from the spool was incrementing
5905 $received_count, but never initializing it. This meant that the value was
5906 incorrect (doubled) while delivering a message in the same process in which
5907 it was received. In the most common configuration of Exim, this never
5908 happens - a fresh exec is done - but it can happen when
5909 deliver_drop_privilege is set.
5911 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5912 it now includes up to 150 characters of the unexpected data in the log
5915 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5916 and building data strings. The size of both of these buffers was 10 000
5917 bytes - far larger than anybody would *ever* want, thought I. Needless to
5918 say, somebody hit the limit. I have increased the maximum line length to
5919 20 000 and the maximum data length of concatenated lines to 100 000. I have
5920 also fixed two bugs, because there was no checking on these buffers. Tsk,
5921 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5922 buffer is too small.
5924 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5925 lsearch lookups. Now it does.
5927 46. When parsing a route_list item in a manualroute router, a fixed-length
5928 buffer was used for the list of hosts. I made this 1024 bytes long,
5929 thinking that nobody would ever have a list of hosts that long. Wrong.
5930 Somebody had a whole pile of complicated expansion conditions, and the
5931 string was silently truncated, leading to an expansion error. It turns out
5932 that it is easier to change to an unlimited length (owing to other changes
5933 that have happened since this code was originally written) than to build
5934 structure for giving a limitation error. The length of the item that
5935 expands into the list of hosts is now unlimited.
5937 47. The lsearch lookup could not handle data where the length of text line was
5938 more than 4095 characters. Such lines were truncated, leading to shortened
5939 data being returned. It should now handle lines of any length.
5941 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5942 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5945 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5946 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5947 no longer happens in comment lines. A list of replacements is now placed
5948 at the head of all of the source files, except those whose only change is
5949 to replace PERL_COMMAND in the very first #! line.
5951 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5952 messages on the queue, with a bottom-up merge sort, using code contributed
5953 by Michael Haardt. This should make operations like -bp somewhat faster on
5954 large queues. It won't affect queue runners, except when queue_run_in_order
5957 51. Installed eximstats 1.31 in the distribution.
5959 52. Added support for SRV lookups to the dnslookup router.
5961 53. If an ACL referred to $message_body or $message_body_end, the value was not
5962 reset for any messages that followed in the same SMTP session.
5964 54. The store-handling optimization for building very long strings was not
5965 differentiating between the different store pools. I don't think this
5966 actually made any difference in practice, but I've tidied it.
5968 55. While running the routers to verify a sender address, $sender_address
5969 was still set to the sender address. This is wrong, because when routing to
5970 send a bounce to the sender, it would be empty. Therefore, I have changed
5971 it so that, while verifying a sender address, $sender_address is set to <>.
5972 (There is no change to what happens when verifying a recipient address.)
5974 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5975 target A or AAAA records (if not already returned) without resetting the
5976 qualify_single or search_parents options of the DNS resolver. These are
5977 inappropriate in this case because the targets of MX and SRV records must
5978 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5979 target that, when qualified, matched something in the local domain. These
5980 two options are now turned off when doing these lookups.
5982 57. It seems that at least some releases of Reiserfs (which does not have the
5983 concept of a fixed number of inodes) returns zero and not -1 for the
5984 number of available inodes. This interacted badly with check_spool_inodes,
5985 which assumed that -1 was the "no such thing" setting. What I have done is
5986 to check that the total number of inodes is greater than zero before doing
5987 the test of how many are available.
5989 58. When a "warn" ACL statement has a log_message modifier, the message is
5990 remembered, and not repeated. This is to avoid a lot of repetition when a
5991 message has many recipients that cause the same warning to be written.
5992 However, Exim was preserving the list of already written lines for an
5993 entire SMTP session, which doesn't seem right. The memory is now reset if a
5994 new message is started.
5996 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5997 debugging output unless log_rewrite was also set.
5999 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6000 of (int)(handle) when we know that handle contains (void *)(-1).
6002 61. The Exim daemon panic-logs an error return when it closes the incoming
6003 connection. However "connection reset by peer" seems to be common, and
6004 isn't really an error worthy of noting specially, so that particular error
6007 62. When Exim is trying to find all the local interfaces, it used to panic and
6008 die if the ioctl to get the interface flags failed. However, it seems that
6009 on at least one OS (Solaris 9) it is possible to have an interface that is
6010 included in the list of interfaces, but for which you get a failure error
6011 for this call. This happens when the interface is not "plumbed" into a
6012 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6013 failure of the "get flags" call assumes that the interface is down.
6015 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6016 makes life easier for people who are doing arithmetic on fields extracted
6017 from dates, where you often get leading zeros that should not be
6018 interpreted as octal.
6020 64. Added qualify_domain to the redirect router, to override the global
6023 65. If a pathologically long header line contained very many addresses (the
6024 report of this problem mentioned 10 000) and each of them was rewritten,
6025 Exim could use up a very large amount of memory. (It kept on making new
6026 copies of the header line as it rewrote, and never released the old ones.)
6027 At the expense of a bit more processing, the header rewriting function has
6028 been changed so that it no longer eats memory in this way.
6030 66. The generation of the Received: header has been moved from the time that a
6031 message starts to be received, to the time that it finishes. The timestamp
6032 in the Received: header should now be very close to that of the <= log
6033 line. There are two side-effects of this change:
6035 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6036 logged header lines no longer include the local Received: line, because
6037 it has not yet been created. The same applies to a copy of the message
6038 that is returned to a non-SMTP sender when a message is rejected.
6040 (b) When a filter file is tested using -bf, no additional Received: header
6041 is added to the test message. After some thought, I decided that this
6044 This change does not affect the value of $received_for. It is still set
6045 after address rewriting, but before local_scan() is called.
6047 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6049 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6050 gave an unhelpful panic error message, and a defer error. I have managed to
6051 change this behaviour so that it now rejects any supplied certificate,
6052 which seems right, as the list of acceptable certificates is empty.
6054 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6055 gave an unhelpful defer error. I have not managed to make this reject any
6056 supplied certificates, but the error message it gives is "no certificate
6057 supplied", which is not helpful.
6059 70. exigrep's output now also includes lines that are not associated with any
6060 message, but which match the given pattern. Implemented by a patch from
6061 Martin Sluka, which also tidied up the Perl a bit.
6063 71. Recipient callout verification, like sender verification, was using <> in
6064 the MAIL FROM command. This isn't really the right thing, since the actual
6065 sender may affect whether the remote host accepts the recipient or not. I
6066 have changed it to use the actual sender in the callout; this means that
6067 the cache record is now keyed on a recipient/sender pair, not just the
6068 recipient address. There doesn't seem to be a real danger of callout loops,
6069 since a callout by the remote host to check the sender would use <>.
6070 [SEE ABOVE: changed after hitting problems.]
6072 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6073 temporary errors. However, in the case of such a code being given after
6074 the end of a data transmission (i.e. after ".") Exim was failing to write
6075 a retry record for the message. (Yes, there was some broken host that was
6076 actually sending 8xx at this point.)
6078 73. An unknown lookup type in a host list could cause Exim to panic-die when
6079 the list was checked. (An example that provoked this was putting <; in the
6080 middle of a list instead of at the start.) If this happened during a DATA
6081 ACL check, a -D file could be left lying around. This kind of configuration
6082 error no longer causes Exim to die; instead it causes a defer error. The
6083 incident is still logged to the main and panic logs.
6085 74. Buglet left over from Exim 3 conversion. The message "too many messages
6086 in one connection" was written to the rejectlog but not the mainlog, except
6087 when address rewriting (yes!) was being logged.
6089 75. Added write_rejectlog option.
6091 76. When a system filter was run not as root (that is, when system_filter_user
6092 was set), the values of the $n variables were not being returned to the
6093 main process; thus, they were not subsequently available in the $sn
6096 77. Added +return_path_on_delivery log selector.
6098 78. A connection timeout was being treated differently from recipients deferred
6099 when testing hosts_max_try with a message that was older than the host's
6100 retry timeout. (The host should not be counted, thus allowing all hosts to
6101 be tried at least once before bouncing.) This may have been the cause of an
6102 occasionally reported bug whereby a message would remain on the queue
6103 longer than the retry timeout, but would be bounced if a delivery was
6104 forced. I say "may" because I never totally pinned down the problem;
6105 setting up timeout/retry tests is difficult. See also the next item.
6107 79. The ultimate address timeout was not being applied to errors that involved
6108 a combination of host plus message (for example, a timeout on a MAIL
6109 command). When an address resolved to a number of possible hosts, and they
6110 were not all tried for each delivery (e.g. because of hosts_max_try), a
6111 message could remain on the queue longer than the retry timeout.
6113 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6116 81. Fixed an obscure SMTP outgoing bug which required at least the following
6117 conditions: (a) there was another message waiting for the same server;
6118 (b) the server returned 5xx to all RCPT commands in the first message so
6119 that the message was not completed; (c) the server dropped the connection
6120 or gave a negative response to the RSET that Exim sends to abort the
6121 transaction. The observed case was a dropped connection after DATA that had
6122 been sent in pipelining mode. That is, the server had advertised PIPELINING
6123 but was not implementing it correctly. The effect of the bug was incorrect
6124 behaviour, such as trying another host, and this could lead to a crash.
6130 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6131 and daemon.c were passed as pointers to ints; they should have been
6132 pointers to socklen_t variables (which are typically unsigned ints).
6134 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6137 3. Fixed a really odd bug that affected only the testing scheme; patching a
6138 certain fixed string in the binary changed the value of another string that
6139 happened to be identical to the end of the original first string.
6141 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6142 name", it returns that address as the IP address. On some operating
6143 systems (e.g. Solaris), it also passes back the IP address string as the
6144 "host name". However, on others (e.g. Linux), it passes back an empty
6145 string. Exim wasn't checking for this, and was changing the host name to an
6146 empty string, assuming it had been canonicalized.
6148 5. Although rare, it is permitted to have more than one PTR record for a given
6149 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6150 all the names associated with an address, because they do in Solaris.
6151 However, it seems that they do not in Linux for data that comes from the
6152 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6153 I found this out when I moved to a new Linux workstation and tried to run
6154 the Exim test suite.
6156 To get round this problem I have changed the code so that it now does its
6157 own call to the DNS to look up PTR records when searching for a host name.
6158 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6159 addresses that are only in /etc/hosts are still found.
6161 This behaviour is, however, controlled by an option called host_lookup_
6162 order, which defaults to "bydns:byaddr". If people want to use the other
6163 order, or indeed, just use one or the other means of lookup, they can
6164 specify it in this variable.
6166 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6167 some operating systems, this comes back from gethostbyaddr() as an empty
6168 string, and this is what Exim used to test for. However, it seems that in
6169 other systems, "." is yielded. Exim now tests for this case too.
6171 7. The values of check_spool_space and check_log_space are now held internally
6172 as a number of kilobytes instead of an absolute number of bytes. If a
6173 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6174 kilobyte. This means that much larger values can be stored.
6176 8. Exim monitor: an attempt to get the action menu when not actually pointing
6177 at a message produces an empty menu entitled "No message selected". This
6178 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6179 no entries in it ("Shell widget menu has zero width and/or height"). So I
6180 have added a single, blank menu entry in this case.
6182 9. Added ${quote_local_part.
6184 10. MIME decoding is now applied to the contents of Subject: header lines when
6187 11. Now that a reference to $sender_host_address automatically causes a reverse
6188 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6189 host lookup before query-style lookups in lists that might use this
6190 variable. This has therefore been abolished, and the "net-" prefix is no
6191 longer necessary for query-style lookups.
6193 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6194 have been a typo for LFLAGS, so it has been changed.
6196 13. The install script calls Exim with "-C /dev/null" in order to find the
6197 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6198 to be output. However, since Exim outputs its version number before the
6199 error, it didn't break the script. It just looked ugly. I fixed this by
6200 always allowing "-C /dev/null" if the caller is root.
6202 14. Ignore overlarge ACL variable number when reading spool file - insurance
6203 against a later release with more variables having written the file.
6205 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6206 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6207 this, as well as the form without the "IPv6" on the front.
6209 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6210 OS/Makefile-Darwin file.
6212 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6213 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6215 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6216 variable. However, some OS use ints for inodes. I've added cast to long int
6217 to get rid of the compiler warning.
6219 19. I had forgotten to lock out "/../" in configuration file names when
6220 ALT_CONFIG_PREFIX was set.
6222 20. Routers used for verification do not need to specify transports. However,
6223 if such a router generated a host list, and callout was configured, Exim
6224 crashed, because it could not find a port number from the (non-existent)
6225 transport. It now assumes port 25 in this circumstance.
6227 21. Added the -t option to exigrep.
6229 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6230 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6231 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6232 them is set, LOOKUP_LSEARCH is forced.
6234 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6236 24. Added sender and host information to the "rejected by local_scan()" log
6237 line; previously there was no indication of these.
6239 25. Added .include_if_exists.
6241 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6242 Linux. It turns out that not all file systems support this. Apparently some
6243 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6244 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6245 not supported on the file descriptor, is now ignored when Exim is trying to
6246 sync a directory. This applies only to Linux.
6248 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6250 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6251 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6254 29. Added support for maildirsize files from supplied patch (modified a bit).
6256 30. The use of :fail: followed by an empty string could lead Exim to respond to
6257 sender verification failures with (e.g.):
6259 550 Verification failed for <xxx>
6260 550 Sender verify failed
6262 where the first response line was missing the '-' that indicates it is not
6263 the final line of the response.
6265 31. The loop for finding the name of the user that called Exim had a hardwired
6266 limit of 10; it now uses the value of finduser_retries, which is used for
6267 all other user lookups.
6269 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6272 33. Exim was neglecting to zero errno before one call of strtol() when
6273 expanding a string and expecting an integer value. On some systems this
6274 resulted in spurious "integer overflow" errors. Also, it was casting the
6275 result into an int without checking.
6277 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6278 did not work. The code looks as if it has *never* worked, though it appears
6279 to have been documented since at least release 1.62. I have made it work.
6281 35. The "timeout_DNS" error in retry rules, also documented since at least
6282 1.62, also never worked. As it isn't clear exactly what this means, and
6283 clearly it isn't a major issue, I have abolished the feature by treating it
6284 as "timeout", and writing a warning to the main and panic logs.
6286 36. The display of retry rules for -brt wasn't always showing the error code
6289 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6290 timeout_connect_A, timeout_connect_MX.
6292 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6293 to the empty sender.
6295 39. The daemon was not analysing the content of -oX till after it had closed
6296 stderr and disconnected from the controlling terminal. This meant that any
6297 syntax errors were only noted on the panic log, and the return code from
6298 the command was 0. By re-arranging the code a little, I've made the
6299 decoding happen first, so such errors now appear on stderr, and the return
6300 code is 1. However, the actual setting up of the sockets still happens in
6301 the disconnected process, so errors there are still only recorded on the
6304 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6305 connections (as happens on some IP stacks) was logged at start up time as
6306 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6307 it from "IPv6 and IPv4", which means that two separate sockets are being
6310 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6311 says whether AF_INET or AF_INET6 was passed as an argument.
6313 42. Exiwhat output was messed up when time zones were included in log
6316 43. Exiwhat now gives more information about the daemon's listening ports,
6317 and whether -tls-on-connect was used.
6319 44. The "port" option of the smtp transport is now expanded.
6321 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6322 silently ignored. Now an error message is written to the main and panic
6325 46. There's a new ACL modifier called "logwrite" which writes to a log file
6326 as soon as it is encountered.
6328 47. Added $local_user_uid and $local_user_gid at routing time.
6330 48. Exim crashed when trying to verify a sender address that was being
6333 49. Exim was recognizing only a space character after ".include". It now also
6334 recognizes a tab character.
6336 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6337 extracting the relevant information from the specification. The man page no
6338 longer contains scrambled data for the -d option, and I've added a section
6339 at the front about calling Exim under different names.
6341 51. Added "extra_headers" argument to the "mail" command in filter files.
6343 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6346 53. Installed eximstats 1.29.
6348 54. Added transport_filter_timeout as a generic transport option.
6350 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6351 Cc: header lines. This was required by RFC 822, but it not required by RFC
6354 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6355 incoming messages that did not have them. Now it does so only if the
6356 message originates locally, that is, if there is no associated remote host
6357 address. When Resent- header lines are present, this applies to the Resent-
6358 lines rather than the non-Resent- lines.
6360 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6361 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6363 58. Messages for configuration errors now include the name of the main
6364 configuration file - useful now that there may be more than one file in a
6365 list (.included file names were always shown).
6367 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6368 for those rare installations that do not start the daemon as root or run it
6369 setuid root. I've cut out the call to initgroups() if the daemon is not
6372 60. The Exim user and group can now be bound into the binary as text strings
6373 that are looked up at the start of Exim's processing.
6375 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6377 62. Added $mailstore_basename variable.
6379 63. Installed patch to sieve.c from Michael Haardt.
6381 64. When Exim failed to open the panic log after failing to open the main log,
6382 the original message it was trying to log was written to stderr and debug
6383 output, but if they were not available (the usual case in production), it
6384 was lost. Now it is written to syslog before the two lines that record the
6385 failures to open the logs.
6387 65. Users' Exim filters run in subprocesses under the user's uid. It is
6388 possible for a "deliver" command or an alias in a "personal" command to
6389 provoke an address rewrite. If logging of address rewriting is configured,
6390 this fails because the process is not running as root or exim. There may be
6391 a better way of dealing with this, but for the moment (because 4.30 needs
6392 to be released), I have disabled address rewrite logging when running a
6393 filter in a non-root, non-exim process.
6399 1. The buildconfig auxiliary program wasn't quoting the value set for
6400 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6401 not defined. This bug was masked in 4.22 by the effect that was fixed in
6404 2. Some messages that were rejected after a message id was allocated were
6405 shown as "incomplete" by exigrep. It no longer does this for messages that
6406 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6408 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6409 have allow_domain_literals set, the ID did not get logged in the <= line.
6410 Domain literals are now always recognized in Message-ID: header lines.
6412 4. The first argument for a ${extract expansion item is the key name or field
6413 number. Leading and trailing spaces in this item were not being ignored,
6414 causing some misleading effects.
6416 5. When deliver_drop_privilege was set, single queue runner processes started
6417 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6418 same command in the process it spins off) were not dropping privilege.
6420 6. When the daemon running as "exim" started a queue runner, it always
6421 re-executed Exim in the spun-off process. This is a waste of effort when
6422 deliver_drop_privilege is set. The new process now just calls the
6423 queue-runner function directly.
6429 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6432 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6433 ignored. Though the use of -oP was forcing the writing of a pid file, it
6434 was always written to the default place.
6436 3. If the message "no IP address found for host xxxx" is generated during
6437 incoming verification, it is now followed by identification of the incoming
6438 connection (so you can more easily find what provoked it).
6440 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6442 5. Added some features to "harden" Exim a bit more against certain attacks:
6444 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6445 be put in Local/Makefile. This is like the never_users runtime option,
6446 but it cannot be overridden. The default setting is "root".
6448 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6449 prefix string with which any file named in a -C command line option
6452 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6453 is retained for -C and -D only if the caller of Exim is root. Without
6454 it, the exim user may also use -C and -D and retain privilege.
6456 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6457 command line option is disabled.
6459 6. Macro names set by the -D option must start with an upper case letter, just
6460 like macro names defined in the configuration file.
6462 7. Added "dereference=" facility to LDAP.
6464 8. Two instances of the typo "uknown" in the source files are fixed.
6466 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6467 the Configure-Makefile script screwed up while processing it.
6469 10. Incorporated PCRE 4.4.
6471 11. The SMTP synchronization check was not operating right at the start of an
6472 SMTP session. For example, it could not catch a HELO sent before the client
6473 waited for the greeting. There is now a check for outstanding input at the
6474 point when the greeting is written. Because of the duplex, asynchronous
6475 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6476 way, but not yet received, when the check is performed.
6478 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6479 on TCP/IP sockets, because this apparently causes some broken clients to
6482 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6483 unchanged) from the Cygwin maintainer.
6485 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6486 was supported instead of testing for mbx. Effectively a typo.
6488 15. The spa authenticator server code was not checking that the input it
6489 received was valid base64.
6491 16. The debug output line for the "set" modifier in ACLs was not showing the
6492 name of the variable that was being set.
6494 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6496 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6497 lookup on its own. Something else was needed to trigger the lookup. For
6498 example, a match in host_lookup or the need for a host name in a host list.
6499 Now, if $sender_host_name is referenced and the host name has not yet been
6500 looked up, a lookup is performed. If the lookup fails, the variable remains
6501 empty, and $host_lookup_failed is set to "1".
6503 19. Added "eqi" as a case-independent comparison operator.
6505 20. The saslauthd authentication condition could segfault if neither service
6506 nor realm was specified.
6508 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6509 error message that was logged was misleading, and incoming SMTP
6510 connections were dropped. The message is now more accurate, and temporary
6511 errors are given to SMTP connections.
6513 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6514 (except RSET) with a 421 error, until QUIT is received. However, it was
6515 failing to send a response to QUIT.
6517 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6518 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6519 ACL is now run after the helo_try_verify_hosts code.
6521 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6522 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6523 case-independent, so other case variants are also recognized). Apparently
6524 some systems use these upper case variants.
6526 25. If more than two messages were waiting for the same host, and a transport
6527 filter was specified for the transport, Exim sent two messages over the
6528 same TCP/IP connection, and then failed with "socket operation on non-
6529 socket" when it tried to send the third.
6531 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6533 27. The extern definition of crypt16() in expand.c was not being excluded when
6534 the OS had its own crypt16() function.
6536 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6537 as a preferred synonym for return_size_limit, both as an option and as an
6540 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6542 30. Changed the default configuration ACL to relax the local part checking rule
6543 for addresses that are not in any local domains. For these addresses,
6544 slashes and pipe symbols are allowed within local parts, but the sequence
6545 /../ is explicitly forbidden.
6547 31. SPA server authentication was not clearing the challenge buffer before
6550 32. log_message in a "warn" ACL statement was writing to the reject log as
6551 well as to the main log, which contradicts the documentation and doesn't
6552 seem right (because no rejection is happening). So I have stopped it.
6554 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6555 However, I am unable to do any testing of this.
6557 34. Fixed an infelicity in the appendfile transport. When checking directories
6558 for a mailbox, to see if any needed to be created, it was accidentally
6559 using path names with one or more superfluous leading slashes; tracing
6560 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6562 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6563 contents are added to the log line that is written for every discarded
6564 recipient. (Previously a log_message setting was ignored.)
6566 36. The ${quote: operator now quotes the string if it is empty.
6568 37. The install script runs exim in order to find its version number. If for
6569 some reason other than non-existence or emptiness, which it checks, it
6570 could not run './exim', it was installing it with an empty version number,
6571 i.e. as "exim-". This error state is now caught, and the installation is
6574 38. An argument was missing from the function that creates an error message
6575 when Exim fails to connect to the socket for saslauthd authentication.
6576 This could cause Exim to crash, or give a corrupted message.
6578 39. Added isip, isip4, and isip6 to ${if conditions.
6580 40. The ACL variables $acl_xx are now saved with the message, and can be
6581 accessed later in routers, transports, and filters.
6583 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6584 strings in the file are not string-expanded.
6586 42. If a MAIL command specified a SIZE value that was too large to fit into an
6587 int variable, the check against message_size_limit failed. Such values are
6588 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6589 day this will have to be increased, but I don't think I want to be around
6590 when emails are that large.
6597 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6598 iconv() is not standard in FreeBSD.
6600 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6601 IPv6 enabled. The observed symptom was a segmentation fault on return from
6602 the function os_common_find_running_interfaces() in src/os.c.
6604 3. In the check_special_case() function in daemon.c I had used "errno" as an
6605 argument name, which causes warnings on some systems. This was basically a
6606 typo, since it was named "eno" in the comments!
6608 4. The code that waits for the clock to tick (at a resolution of some fraction
6609 of a second) so as to ensure message-id uniqueness was always waiting for
6610 at least one whole tick, when it could have waited for less. [This is
6611 almost certainly not relevant at current processor speeds, where it is
6612 unlikely to ever wait at all. But we try to future-proof.]
6614 5. The function that sleeps for a time interval that includes fractions of a
6615 second contained a race. It did not block SIGALRM between setting the
6616 timer, and suspending (a couple of lines later). If the interval was short
6617 and the sigsuspend() was delayed until after it had expired, the suspension
6618 never ended. On busy systems this could lead to processes getting stuck for
6621 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6622 process, before it forks any delivery processes. The open lookup caching
6623 mechanism meant that the open file or database connection was passed into
6624 the delivery process. The problem was that delivery processes always tidy
6625 up cached lookup data. This could cause a problem for the next delivery
6626 process started by the queue runner, because the external queue runner
6627 process does not know about the closure. So the next delivery process
6628 still has data in the lookup cache. In the case of a file lookup, there was
6629 no problem because closing a file descriptor in a subprocess doesn't affect
6630 the parent. However, if the lookup was caching a connection to a database,
6631 the connection was closed, and the second delivery process was likely to
6632 see errors such as "PGSQL: query failed: server closed the connection
6633 unexpectedly". The problem has been fixed by closing all cached lookups
6634 in a queue runner before running a delivery process.
6636 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6637 seem to have the "const" qualifier which it has on other OS. I've
6640 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6641 *of the same type* (client or server) with the same public name that an
6642 error should be diagnosed.
6644 9. When Exim looked up a host name for an IP address, but failed to find the
6645 original IP address when looking up the host name (a safety check), it
6646 output the message "<ip address> does not match any IP for NULL", which was
6647 confusing, to say the least. The bug was that the host name should have
6648 appeared instead of "NULL".
6650 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6651 user that is built into the binary, and the -C or -D options is used, root
6652 privilege is dropped before the configuration file is read. In addition,
6653 logging is switched to stderr instead of the normal log files. If the
6654 configuration then re-defines the Exim user, the unprivileged environment
6655 is probably not what is expected, so Exim logs a panic warning message (but
6658 However, if deliver_drop_privilege is set, the unprivileged state may well
6659 be exactly what is intended, so the warning has been cut out in that case,
6660 and Exim is allowed to try to write to its normal log files.
6666 1. smtp_return_error_details was not giving details for temporary sender
6667 or receiver verification errors.
6669 2. Diagnose a configuration error if two authenticators have the same public
6672 3. Exim used not to create the message log file for a message until the first
6673 delivery attempt. This could be confusing when incoming messages were held
6674 for policy or load reasons. The message log file is now created at the time
6675 the message is received, and an initial "Received" line is written to it.
6677 4. The automatically generated man page for command line options had a minor
6678 bug that caused no ill effects; however, a more serious problem was that
6679 the procedure for building the man page automatically didn't always
6680 operate. Consequently, release 4.20 contains an out-of-date version. This
6681 shouldn't happen again.
6683 5. When building Exim with embedded Perl support, the script that builds the
6684 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6685 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6687 6. The freeze_tell option was not being used for messages that were frozen on
6688 arrival, either by an ACL or by local_scan().
6690 7. Added the smtp_incomplete_transaction log selector.
6692 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6693 was accepting AUTH without a new EHLO.
6695 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6696 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6699 10. Logging of TCP/IP connections (when configured) now happens in the main
6700 daemon process instead of the child process, so that the TCP/IP connection
6701 count is more accurate (but it can never be perfect).
6703 11. The use of "drop" in a nested ACL was not being handled correctly in the
6704 outer ACL. Now, if condition failure induced by the nested "drop" causes
6705 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6706 or "require"), the connection is dropped.
6708 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6709 that yield "discard" can now be used with an "accept" or a "discard" verb,
6710 but an error is generated for any others (because I can't see a useful way
6711 to define what should happen).
6713 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6714 are now processed in the same way as lines in the Exim configuration file.
6715 In particular, continuation lines are supported.
6717 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6719 15. Added -ti meaning -t -i.
6721 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6722 domains, and warn by logging if others are found.
6724 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6725 returned from GIFCONF when Exim is trying to find the list of interfaces on
6726 a host. The code in os.c has been modified to copy each ifreq to an aligned
6727 structure in all cases.
6729 Also, in some cases, the returned ifreq's were being copied to a 'struct
6730 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6731 means the last couple of bytes of an IPv6 address could be chopped if the
6732 ifreq contained only a normal sockaddr (14 bytes storage).
6734 18. Named domain lists were not supported in the hosts_treat_as_local option.
6735 An entry such as +xxxx was not recognized, and was treated as a literal
6738 19. Ensure that header lines added by a DATA ACL are included in the reject log
6739 if the ACL subsequently rejects the message.
6741 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6742 MD5 (which is deprecated).
6744 21. When testing a filter file using -bf, Exim was writing a message when it
6745 took the sender from a "From " line in the message, but it was not doing so
6746 when it took $return_path from a Return-Path: header line. It now does.
6748 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6749 with a valid header line field name (a series of printing characters
6750 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6752 23. Changed "disc" in the source to "disk" to conform to the documentation and
6753 the book and for uniformity.
6755 24. Ignore Sendmail's -Ooption=value command line item.
6757 25. When execve() failed while trying to run a command in a pipe transport,
6758 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6759 could be confused with a return value of 69 from the command itself. This
6760 has been changed to 127, the value the shell returns if it is asked to run
6761 a non-existent command. The wording for the related log line suggests a
6762 non-existent command as the problem.
6764 26. If received_header_text expands to an empty string, do not add a Received:
6765 header line to the message. (Well, it adds a token one on the spool, but
6766 marks it "old" so that it doesn't get used or transmitted.)
6768 27. Installed eximstats 1.28 (addition of -nt option).
6770 28. There was no check for failure on the call to getsockname() in the daemon
6771 code. This can fail if there is a shortage of resources on the system, with
6772 ENOMEM, for example. A temporary error is now given on failure.
6774 29. Contrary to the C standard, it seems that in some environments, the
6775 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6776 program. Exim now does this explicitly; it affects the formatting of
6777 timestamps using strftime().
6779 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6780 complaints. I've now initialized all the variables, to avoid this.
6782 32. Header lines added by a system filter were not being "seen" during
6783 transport-time rewrites.
6785 33. The info_callback() function passed to OpenSSL is set up with type void
6786 (*)(SSL *, int, int), as described somewhere. However, when calling the
6787 function (actually a macro) that sets it up, the type void(*)() is
6788 expected. I've put in a cast to prevent warnings from picky compilers.
6790 34. If a DNS black list lookup found a CNAME record, but there were no A
6791 records associated with the domain it pointed at, Exim crashed.
6793 35. If a DNS black list lookup returned more than one A record, Exim ignored
6794 all but the first. It now scans all returned addresses if a particular IP
6795 value is being sought. In this situation, the contents of the
6796 $dnslist_value variable are a list of all the addresses, separated by a
6799 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6800 to accept a host name and all its aliases if the forward lookup for any of
6801 them yielded the IP address of the incoming connection. Now it accepts only
6802 those names whose forward lookup yields the correct IP address. Any other
6803 names are discarded. This closes a loophole whereby a rogue DNS
6804 administrator could create reverse DNS records to break through a
6805 wildcarded host restriction in an ACL.
6807 37. If a user filter or a system filter that ran in a subprocess used any of
6808 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6809 the wrong values were passed to the pipe command ($thisaddress had the
6810 value of $0, $0 had the value of $1, etc). This bug was introduced by
6811 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6813 38. Improved the line breaking for long SMTP error messages from ACLs.
6814 Previously, if there was no break point between 40 and 75 characters, Exim
6815 left the rest of the message alone. Two changes have been made: (a) I've
6816 reduced the minimum length to 35 characters; (b) if it can't find a break
6817 point between 35 and 75 characters, it looks ahead and uses the first one
6818 that it finds. This may give the occasional overlong line, but at least the
6819 remaining text gets split now.
6821 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6822 file descriptors might be low, and that setting 1000 would always raise it.
6823 It turns out that in some environments, the limit is already over 1000 and
6824 that lowering it causes trouble. So now Exim takes care not to decrease it.
6826 40. When delivering a message, the value of $return_path is set to $sender_
6827 address at the start of routing (routers may change the value). By an
6828 oversight, this default was not being set up when an address was tested by
6829 -bt or -bv, which affected the outcome if any router or filter referred to
6832 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6833 log only when "message" or "log_message" are set. However, if one of the
6834 conditions was an address verification, or a call to a nested ACL, the
6835 messages generated by the underlying test were being passed through. This
6836 no longer happens. The underlying message is available in $acl_verify_
6837 message for both "message" and "log_message" expansions, so it can be
6838 passed through if needed.
6840 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6841 new expansion $bh_ to give the encoded byte string without charset
6842 translation. Translation happens only if iconv() is available; HAVE_ICONV
6843 indicates this at build time. HEADERS_CHARSET gives the charset to
6844 translate to; headers_charset can change it in the configuration, and
6845 "headers charset" can change it in an individual filter file.
6847 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6848 that creates RFC 2047 encoded "words" labels them as that charset instead
6849 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6850 expansion operator; (ii) when Exim creates a From: line for a local
6851 message; (iii) when a header line is rewritten to include a "phrase" part.
6853 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6854 buggy, causing it to skip the first lines of messages whose message ID
6855 ended in 'D'. This would not have bitten before Exim release 4.14, because
6856 message IDs were unlikely to end in 'D' before then. The effect was to have
6857 incorrect size information for certain domains.
6859 45. #include "config.h" was missing at the start of the crypt16.c module. This
6860 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6863 46. If there was a timeout during a "random" callout check, Exim treated it as
6864 a failure of the random address, and carried on sending RSET and the real
6865 address. If the delay was just some slowness somewhere, the response to the
6866 original RCPT would be taken as a response to RSET and so on, causing
6867 mayhem of various kinds.
6869 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6870 when I implemented it. It didn't allow for the fact that some option values
6871 may legitimately be negative (e.g. size_addition), and it didn't even do
6872 the right test for positive values.
6874 48. Domain names in DNS records are case-independent. Exim always looks them up
6875 in lower case. Some resolvers return domain names in exactly the case they
6876 appear in the zone file, that is, they may contain uppercase letters. Not
6877 all resolvers do this - some return always lower case. Exim was treating a
6878 change of case by a resolver as a change of domain, similar to a widening
6879 of a domain abbreviation. This triggered its re-routing code and so it was
6880 trying to route what was effectively the same domain again. This normally
6881 caused routing to fail (because the router wouldn't handle the domain
6882 twice). Now Exim checks for this case specially, and just changes the
6883 casing of the domain that it ultimately uses when it transmits the message
6886 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6889 50. If a filter generated a file delivery with a non-absolute name (possible if
6890 no home directory exists for the router), the forbid_file option was not
6893 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6894 the existing equality matching.
6896 52. Exim was using ints instead of ino_t variables in some places where it was
6897 dealing with inode numbers.
6899 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6900 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6901 called TMPDIR, and if it finds it is different, it changes its value.
6903 54. The smtp_printf() function is now made available to local_scan() so
6904 additional output lines can be written before returning. There is also an
6905 smtp_fflush() function to enable the detection of a dropped connection.
6906 The variables smtp_input and smtp_batched_input are exported to
6909 55. Changed the default runtime configuration: the message "Unknown user"
6910 has been removed from the ACL, and instead placed on the localuser router,
6911 using the cannot_route_message feature. This means that any verification
6912 failures that generate their own messages won't get overridden. Similarly,
6913 the "Unrouteable address" message that was in the ACL for unverifiable
6914 relay addresses has also been removed.
6916 56. Added hosts_avoid_esmtp to the smtp transport.
6918 57. The exicyclog script was not checking for the esoteric option
6919 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6920 will work only if exicyclog is run under the appropriate euid.
6922 58. Following a discussion on the list, the rules by which Exim recognises line
6923 endings on incoming messages have been changed. The -dropcr and drop_cr
6924 options are now no-ops, retained only for backwards compatibility. The
6925 following line terminators are recognized: LF CRLF CR. However, special
6926 processing applies to CR:
6928 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6929 nor a local message in the state where . is a terminator.
6931 (ii) If a bare CR is encountered in a header line, an extra space is added
6932 after the line terminator so as not to end the header. The reasoning
6933 behind this is that bare CRs in header lines are most likely either
6934 to be mistakes, or people trying to play silly games.
6936 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6937 was being incorrectly given as 18 bytes larger than it should have been.
6938 This is a VOB (very old bug).
6940 60. This may never have affected anything current, but just in case it has:
6941 When the local host is found other than at the start of a list of hosts,
6942 the local host, those with the same MX, and any that follow, are discarded.
6943 When the list in question was part of a longer list of hosts, the following
6944 hosts (not currently being processed) were also being discarded. This no
6945 longer happens. I'm not sure if this situation could ever has previously
6948 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6951 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6952 header. This is something that is recommended in a new Internet Draft, and
6953 is something that is documented as being done by Sendmail. There are two
6954 possible values. For messages generated by the autoreply transport, Exim
6957 Auto-Submitted: auto-replied
6959 whereas for all other generated messages (e.g. bounces) it adds
6961 Auto-Submitted: auto-generated
6963 63. The "personal" condition in filters now includes a test for the
6964 Auto-Submitted: header. If it contains the string "auto-" the message it
6965 not considered personal.
6967 64. Added rcpt_include_affixes as a generic transport option.
6969 65. Added queue_only_override (default true).
6971 66. Added the syslog_duplication option.
6973 67. If what should have been the first header line of a message consisted of
6974 a space followed by a colon, Exim was mis-interpreting it as a header line.
6975 It isn't of course - it is syntactically invalid and should therefore be
6976 treated as the start of the message body. The misbehaviour could have
6977 caused a number of strange effects, including loss of data in subsequent
6978 header lines, and spool format errors.
6980 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6981 client host had authenticated. This control can now be exercised by an ACL
6982 for more flexibility.
6984 69. By default, callouts do not happen when testing with -bh. There is now a
6985 variant, -bhc, which does actually run the callout code, including
6986 consulting and updating the callout cache.
6988 70. Added support for saslauthd authentication, courtesy of Alexander
6991 71. If statvfs() failed on the spool or log directories while checking their
6992 size for availability, Exim confusingly gave the error "space shortage".
6993 Furthermore, in debugging mode it crashed with a floating point exception.
6994 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6995 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6996 serious problem, Exim now writes to the main and panic logs when this
6997 happens, with details of the failure. It then refuses to accept the
6998 incoming message, giving the message "spool directory problem" or "log
6999 directory problem" with a 421 code for SMTP messages.
7001 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7002 0, 1, and 2 exist, because some OS complain for execs without them (see
7003 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7004 descriptors. However, the code omitted to check that the open succeeded,
7005 causing mysterious errors if for some reason the permissions on /dev/null
7006 got screwed. Now Exim writes a message to the main and panic logs, and
7007 bombs out if it can't open /dev/null.
7009 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7010 interact so that it is all more flexible. It is supposed to remain
7011 backwards compatible. Also added extra_local_interfaces.
7013 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7014 to bomb out with an assertion failure - to the client this appears as a
7015 connection drop. This problem occurs in the part of the code that was taken
7016 from the Samba project. Fortunately, the assertion is in a very simple
7017 function, so I have fixed this by reproducing the function inline in the
7018 one place where it is called, and arranging for authentication to fail
7019 instead of killing the process with assert().
7021 75. The SPA client code was not working when the server requested OEM rather
7022 than Unicode encoding.
7024 76. Added code to make require_files with a specific uid setting more usable in
7025 the case where statting the file as root fails - usually a non-root-mounted
7026 NFS file system. When this happens and the failure is EACCES, Exim now
7027 forks a subprocess and does the per-uid checking as the relevant uid.
7029 77. Added process_log_path.
7031 78. If log_file_path was not explicitly set, a setting of check_log_space or
7032 check_log_inodes was ignored.
7034 79. If a space check for the spool or log partitions fails, the incident is now
7035 logged. Of course, in the latter case the data may get lost...
7037 80. Added the %p formatting code to string_format() so that it can be used to
7038 print addresses in debug_print(). Adjusted all the address printing in the
7039 debugging in store.c to use %p rather than %d.
7041 81. There was a concern that a line of code in smtp_in.c could overflow a
7042 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7043 initially expressed, the concern was not well-founded, because trailing
7044 spaces are removed early. However, if the trailing spaces were followed by
7045 a NULL, they did not get removed, so the overflow was possible. Two fixes
7048 (a) I re-wrote the offending code in a cleaner fashion.
7049 (b) If an incoming SMTP command contains a NULL character, it is rejected
7052 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7053 runs initgroups(), so that if the Exim user is in any additional groups,
7054 they will be used during message reception.
7060 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob