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
84 JH/01 Rework error string handling in TLS interface so that the caller in
85 more cases is responsible for logging. This permits library-sourced
86 string to be attached to addresses during delivery, and collapses
87 pairs of long lines into single ones.
89 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
90 during configuration. Wildcards are allowed and expanded.
92 JH/02 Rework error string handling in DKIM to pass more info back to callers.
93 This permits better logging.
95 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
96 do not close it down and have the child transport start it up again on
97 the passed-on TCP connection. Instead, proxy the child (and any
98 subsequent ones) for TLS via a unix-domain socket channel. Logging is
99 affected: the continued delivery log lines do not have any DNSSEC, TLS
100 Certificate or OCSP information. TLS cipher information is still logged.
102 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
103 identical IP addresses on different listening ports. Will also affect
106 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
107 add noisy ifdef guards to special-case this sillyness.
108 Patch from Bernd Kuhls.
110 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
111 than 255 are no longer allowed.
113 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
114 Disable the session-cache too, which might reduce our load. Since we
115 currrectly use a new context for every connection, both as server and
116 client, there is no benefit for these.
117 GnuTLS appears to not support tickets server-side by default (we don't
118 call gnutls_session_ticket_enable_server()) but client side is enabled
119 by default on recent versions (3.1.3 +) unless the PFS priority string
122 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
123 <https://reproducible-builds.org/specs/source-date-epoch/>.
125 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
126 the check for any unsuccessful recipients did not notice the limit, and
127 erroneously found still-pending ones.
129 JH/08 Pipeline CHUNKING command and data together, on kernels that support
130 MSG_MORE. Only in-clear (not on TLS connections).
132 JH/09 Avoid using a temporary file during transport using dkim. Unless a
133 transport-filter is involved we can buffer the headers in memory for
134 creating the signature, and read the spool data file once for the
135 signature and again for transmission.
137 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
138 4.77 as the kernel support then wasn't solid, having issues in 64bit
139 mode. Now, it's been long enough. Add support for FreeBSD also.
141 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
142 case where the routing stage had gathered several addresses to send to
143 a host before calling the transport for the first, we previously failed
144 to close down TLS in the old transport process before passing the TCP
145 connection to the new process. The new one sent a STARTTLS command
146 which naturally failed, giving a failed delivery and bloating the retry
147 database. Investigation and fix prototype from Wolfgang Breyha.
149 JH/12 Fix check on SMTP command input synchronisation. Previously there were
150 false-negatives in the check that the sender had not preempted a response
151 or prompt from Exim (running as a server), due to that code's lack of
152 awareness of the SMTP input buffering.
154 PP/04 Add commandline_checks_require_admin option.
155 Exim drops privileges sanely, various checks such as -be aren't a
156 security problem, as long as you trust local users with access to their
157 own account. When invoked by services which pass untrusted data to
158 Exim, this might be an issue. Set this option in main configuration
159 AND make fixes to the calling application, such as using `--` to stop
162 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
163 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
164 responses to those, into a single TLS record each way (this usually means
165 a single packet). As a side issue, smtp_enforce_sync now works on TLS
168 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
169 affects you only if you're dancing at the edge of the param size limits.
170 If you are, and this message makes sense to you, then: raise the
171 configured limit or use OpenSSL 1.1. Nothing we can do for older
174 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
175 cmdline element to get the documented default one. Previously it was
178 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
179 are unsuable for non-message-specific reasons - eg. connection timeout,
180 banner-time rejection.
182 JH/16 Fix logging of delivery remote port, when specified by router, under
185 PP/06 Repair manualroute's ability to take options in any order, even if one
186 is the name of a transport.
189 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
191 JH/17 Change the list-building routines interface to use the expanding-string
192 triplet model, for better allocation and copying behaviour.
194 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
195 Previously it was constructed the first time a possibly-matching string
196 was met in the configuration file input during startup; now it is done
199 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
200 compatible one, to avoid the (poorly documented) possibility of a config
201 file in the working directory redirecting the DB files, possibly correpting
202 some existing file. CVE-2017-10140 assigned for BDB.
204 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
205 cache-hot. Previously, although the result was properly cached, the
206 initial verify call returned a defer.
208 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
209 the main verify for receipient in uncached-mode.
211 JH/22 Retire historical build files to an "unsupported" subdir. These are
212 defined as "ones for which we have no current evidence of testing".
214 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
215 if present. Previously it was ignored.
217 JH/24 Start using specified-initialisers in C structure init coding. This is
218 a C99 feature (it's 2017, so now considered safe).
220 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
221 if was a fixed-sized field and bitmask ops via macros; it is now more
224 PP/07 GitHub PR 56: Apply MariaDB build fix.
225 Patch provided by Jaroslav Škarvada.
227 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
228 during Coverity cleanups [4.87 JH/47]
229 Diagnosis and fix provided by Michael Fischer v. Mollard.
231 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
232 the right size to place the terminating semicolon on its own folded
233 line, the header hash was calculated to an incorrect value thanks to
234 the (relaxed) space the fold became.
236 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
237 and confused the parent.
239 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
240 which could crash as a result. This could lead to undeliverable messages.
242 JH/28 Logging: "next input sent too soon" now shows where input was truncated
245 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
246 matters on fast-turnover and PID-randomising systems, which were getting
247 out-of-order delivery.
249 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
250 a possibly-overlapping copy. The symptom was that "Remote host closed
251 connection in response to HELO" was logged instead of the actual 4xx
254 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
255 Previously only that bufferd was discarded, resulting in SYMTP command
258 JH/32 DKIM: when a message has multiple signatures matching an identity given
259 in dkim_verify_signers, run the dkim acl once for each. Previously only
260 one run was done. Bug 2189.
262 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
263 "panic the current process" to "deliberately defer". The panic log is
264 still written with the problem list name; the mail and reject logs now
265 get a temp-reject line for the message that was being handled, saying
266 something like "domains check lookup or other defer". The SMTP 451
267 message is still "Temporary local problem".
269 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
270 A crafted sequence of BDAT commands could result in in-use memory beeing
271 freed. CVE-2017-16943.
273 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
274 from SMTP input. Previously it was always done; now only done for DATA
275 and not BDAT commands. CVE-2017-16944.
277 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
278 to the message (such as an overlong header line). Previously this was
279 not done and we did not exit BDAT mode. Followon from the previous item
280 though a different problem.
286 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
287 than -2003 did; needs libidn2 in addition to libidn.
289 JH/02 The path option on a pipe transport is now expanded before use.
291 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
292 Patch provided by "Björn", documentation fix added too.
294 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
295 missing a wire-to-host endian conversion.
297 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
298 close after a BDAT command line could be taken as a following command,
299 giving a synch failure. Fix by only checking for synch immediately
300 before acknowledging the chunk.
302 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
303 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
304 macro. Patches provided by Josh Soref.
306 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
307 Previously we did not; the RFC seems ambiguous and VRFY is not listed
308 by IANA as a service extension. However, John Klensin suggests that we
311 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
312 the dkim code may be unix-mode line endings rather than smtp wire-format
313 CRLF, so prepend a CR to any bare LF.
315 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
316 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
318 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
319 the first were themselves being wrongly included in the feed into dkim
320 processing; with most chunk sizes in use this resulted in an incorrect
321 body hash calculated value.
323 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
324 DKIM signature block, for verification. Although advised against by
325 standards it is specifically not ruled illegal.
327 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
329 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
330 missing a body hash (the bh= tag).
332 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
333 It seems that HAProxy sends the Proxy Protocol information in clear and
334 only then does a TLS startup, so do the same.
336 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
337 TCP connections (such as for Spamd) unless the daemon successfully set
338 Fast Open mode on its listening sockets. This fixes breakage seen on
339 too-old kernels or those not configured for Fast Open, at the cost of
340 requiring both directions being enabled for TFO, and TFO never being used
341 by non-daemon-related Exim processes.
343 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
344 endings, at least on the first header line. Try to canonify any that get
345 past that check, despite the cost.
347 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
348 now limited to an arbitrary five deep, while parsing addresses with the
349 strip_excess_angle_brackets option enabled.
351 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
352 instead leave the unprompted TLS handshake in socket buffer for the
353 TLS library to consume.
355 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
357 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
359 JH/16 Drop variables when they go out of scope. Memory management drops a whole
360 region in one operation, for speed, and this leaves assigned pointers
361 dangling. Add checks run only under the testsuite which checks all
362 variables at a store-reset and panics on a dangling pointer; add code
363 explicitly nulling out all the variables discovered. Fixes one known
364 bug: a transport crash, where a dangling pointer for $sending_ip_address
365 originally assigned in a verify callout, is re-used.
367 PP/06 Drop '.' from @INC in various Perl scripts.
369 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
371 PP/08 Reduce a number of compilation warnings under clang; building with
372 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
373 should be warning-free.
375 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
377 HS/01 Fix portability problems introduced by PP/08 for platforms where
378 realloc(NULL) is not equivalent to malloc() [SunOS et al].
380 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
381 chunk. This allows us to accept broken chunked messages. We need a more
382 general solution here.
384 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
385 already-broken messages in the queue.
387 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
389 JH/19 Fix reference counting bug in routing-generated-address tracking.
395 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
396 supports it and a size is available (ie. the sending peer gave us one).
398 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
399 years of being deprecated). The replacements are the ACLs
400 acl_smtp_mime and acl_not_smtp_mime.
402 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
403 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
404 or even an in-clear connection were permitted. Now, if the host lookup
405 was dnssec and dane was requested then the host is only used if the
406 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
407 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
408 if one fails this test.
409 This means that a poorly-configured remote DNS will make it incommunicado;
410 but it protects against a DNS-interception attack on it.
412 JH/04 Bug 1810: make continued-use of an open smtp transport connection
413 non-noisy when a race steals the message being considered.
415 JH/05 If main configuration option tls_certificate is unset, generate a
416 self-signed certificate for inbound TLS connections.
418 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
419 in rewrites and routers.
421 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
422 and logged a warning sing 4.83; now they are a configuration file error.
424 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
425 (lacking @domain). Apply the same qualification processing as RCPT.
427 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
429 JH/10 Support ${sha256:} applied to a string (as well as the previous
432 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
433 a cutthrough deliver is pending, as we always want to make a connection.
434 This also avoids re-routing the message when later placing the cutthrough
435 connection after a verify cache hit.
436 Do not update it with the verify result either.
438 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
439 when routing results in more than one destination address.
441 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
442 signing (which inhibits the cutthrough capability). Previously only
443 the presence of an option was tested; now an expansion evaluating as
444 empty is permissible (obviously it should depend only on data available
445 when the cutthrough connection is made).
447 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
448 the relevant preceding SMTP command did not note the pipelining mode.
450 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
451 Previously they were not counted.
453 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
454 as one having no matching records. Previously we deferred the message
455 that needed the lookup.
457 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
458 distinguished as "(=".
460 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
461 for missing MX records. Previously it only worked for missing A records.
463 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
465 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
466 after the data-go-ahead and data-ack. Patch from Jason Betts.
468 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
469 even for a "none" policy. Patch from Tony Meyer.
471 JH/22 Fix continued use of a connection for further deliveries. If a port was
472 specified by a router, it must also match for the delivery to be
475 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
476 When one of the recipients of a message was unsuitable for the connection
477 (has no matching addresses), we lost track of needing to mark it
478 deferred. As a result mail would be lost.
480 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
482 JH/25 Decoding ACL controls is now done using a binary search; the source code
483 takes up less space and should be simpler to maintain. Merge the ACL
484 condition decode tables also, with similar effect.
486 JH/26 Fix problem with one_time used on a redirect router which returned the
487 parent address unchanged. A retry would see the parent address marked as
488 delivered, so not attempt the (identical) child. As a result mail would
491 JH/27 Fix a possible security hole, wherein a process operating with the Exim
492 UID can gain a root shell. Credit to http://www.halfdog.net/ for
493 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
496 JH/28 Enable {spool,log} filesystem space and inode checks as default.
497 Main config options check_{log,spool}_{inodes,space} are now
498 100 inodes, 10MB unless set otherwise in the configuration.
500 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
501 Previously it only applied to the main-section connection policy
504 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
506 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
507 by me. Added RFC7919 DH primes as an alternative.
509 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
510 are not in the system include path.
512 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
513 GnuTLS, when a session startup failed (eg because the client disconnected)
514 Exim did stdio operations after fclose. This was exposed by a recent
515 change which nulled out the file handle after the fclose.
517 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
518 signed directly by the cert-signing cert, rather than an intermediate
519 OCSP-signing cert. This is the model used by LetsEncrypt.
521 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
523 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
524 an incoming connection.
526 HS/02 Bug 1802: Do not half-close the connection after sending a request
529 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
530 fallback to "prime256v1".
532 JH/34 SECURITY: Use proper copy of DATA command in error message.
533 Could leak key material. Remotely exploitable. CVE-2016-9963.
539 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
540 and 3.4.4 - once the server is enabled to respond to an OCSP request
541 it does even when not requested, resulting in a stapling non-aware
542 client dropping the TLS connection.
544 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
545 support variable-length bit vectors. No functional change.
547 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
548 The I= interface field on outgoing lines is now after the H= remote
549 host field, same as incoming lines. There is a separate
550 outgoing_interface log selector which allows you to disable the
553 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
554 If not running log_selector +smtp_connection the mainlog would be held
555 open indefinitely after a "too many connections" event, including to a
556 deleted file after a log rotate. Leave the per net connection logging
557 leaving it open for efficiency as that will be quickly detected by the
558 check on the next write.
560 HS/01 Bug 1671: Fix post transport crash.
561 Processing the wait-<transport> messages could crash the delivery
562 process if the message IDs didn't exist for some reason. When
563 using 'split_spool_directory=yes' the construction of the spool
564 file name failed already, exposing the same netto behaviour.
566 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
567 mime_regex ACL conditions.
569 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
570 to DSN fail messages (bounces): remote IP, remote greeting, remote response
571 to HELO, local diagnostic string.
573 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
574 log line to debug. Even when configured with a tls authenticator many
575 client connections are expected to not authenticate in this way, so
576 an authenticate fail is not an error.
578 HS/02 Add the Exim version string to the process info. This way exiwhat
579 gives some more detail about the running daemon.
581 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
582 matter for fast-change records such as DNSBLs.
584 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
585 retry record, even if constant. There may be multiple transports with
586 different interface settings and the retry behaviour needs to be kept
589 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
591 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
593 JH/10 Bug 840: fix log_defer_output option of pipe transport
595 JH/11 Bug 830: use same host for all RCPTS of a message, even under
596 hosts_randomize. This matters a lot when combined with mua_wrapper.
598 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
599 ${quote_pgsql:<string>} operator.
601 JH/13 Bug 1708: avoid misaligned access in cached lookup.
603 JH/14 Change header file name for freeradius-client. Relevant if compiling
604 with Radius support; from the Gentoo tree and checked under Fedora.
606 JH/15 Bug 1712: Introduce $prdr_requested flag variable
608 JH/16 Bug 1714: Permit an empty string as expansion result for transport
609 option transport_filter, meaning no filtering.
611 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
613 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
614 defaults to "*" (all hosts). The variable is now available when not built
615 with TLS, default unset, mainly to enable keeping the testsuite sane.
616 If a server certificate is not supplied (via tls_certificate) an error is
617 logged, and clients will find TLS connections fail on startup. Presumably
618 they will retry in-clear.
619 Packagers of Exim are strongly encouraged to create a server certificate
620 at installation time.
622 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
623 with the $config_file variable.
625 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
626 in transport context, after the attempt, and per-recipient. The latter type
627 is per host attempted. The event data is the error message, and the errno
628 information encodes the lookup type (A vs. MX) used for the (first) host,
629 and the trailing two digits of the smtp 4xx response.
631 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
632 to write to mainlog (or rejectlog, paniclog) in the window between file
633 creation and permissions/ownership being changed. Particularly affects
634 installations where exicyclog is run as root, rather than exim user;
635 result is that the running daemon panics and dies.
637 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
639 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
640 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
641 "pri" and "weight". Note that the previous implicit priority given by the
642 list order is no longer honoured.
644 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
647 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
648 by defining SUPPORT_SOCKS.
650 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
651 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
652 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
653 variables are renamed to proxy_{local,external}_{address,port}.
655 JH/27 Move Internationalisation support from Experimental to mainline, enabled
656 for a build by defining SUPPORT_I18N
658 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
659 of the query string, and make ${quote_redis:} do that quoting.
661 JH/29 Move Events support from Experimental to mainline, enabled by default
662 and removable for a build by defining DISABLE_EVENT.
664 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
666 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
667 cached by the daemon.
669 JH/32 Move Redis support from Experimental to mainline, enabled for a build
670 by defining LOOKUP_REDIS. The libhiredis library is required.
672 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
673 keys are given for lookup.
675 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
676 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
677 only supported when built with TLS support. The PolarSSL SHA routines
678 are still used when the TLS library is too old for convenient support.
680 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
681 openssl_options), for security. OpenSSL forces this from version 1.1.0
682 server-side so match that on older versions.
684 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
685 allocation for $value could be released as the expansion processing
686 concluded, but leaving the global pointer active for it.
688 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
689 and to use the domains and local_parts ACL conditions.
691 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
692 incorrectly not doubled on cutthrough transmission, hence seen as a
693 body-termination at the receiving system - resulting in truncated mails.
694 Commonly the sender saw a TCP-level error, and retransmitted the message
695 via the normal store-and-forward channel. This could result in duplicates
696 received - but deduplicating mailstores were liable to retain only the
697 initial truncated version.
699 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
701 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
703 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
704 we're in there, support oversigning also; bug 1309.
706 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
708 HS/04 Add support for keep_environment and add_environment options.
710 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
711 either intentional arithmetic overflow during PRNG, or testing config-
714 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
715 delivery resulted in actual delivery. Cancel cutthrough before DATA
718 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
719 rejects a recipient: pass the reject to the originator.
721 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
722 Many were false-positives and ignorable, but it's worth fixing the
725 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
726 for the new environment-manipulation done at startup. Move the routines
727 from being local to tls.c to being global via the os.c file.
729 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
730 an extract embedded as result-arg for a map, the first arg for extract
731 is unavailable so we cannot tell if this is a numbered or keyed
732 extraction. Accept either.
738 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
741 JH/02 The smtp transport option "multi_domain" is now expanded.
743 JH/03 The smtp transport now requests PRDR by default, if the server offers
746 JH/04 Certificate name checking on server certificates, when exim is a client,
747 is now done by default. The transport option tls_verify_cert_hostnames
748 can be used to disable this per-host. The build option
749 EXPERIMENTAL_CERTNAMES is withdrawn.
751 JH/05 The value of the tls_verify_certificates smtp transport and main options
752 default to the word "system" to access the system default CA bundle.
753 For GnuTLS, only version 3.0.20 or later.
755 JH/06 Verification of the server certificate for a TLS connection is now tried
756 (but not required) by default. The verification status is now logged by
757 default, for both outbound TLS and client-certificate supplying inbound
760 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
763 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
764 Status Notification (bounce) messages are now MIME format per RFC 3464.
765 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
766 under the control of the dsn_advertise_hosts option, and routers may
767 have a dsn_lasthop option.
769 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
770 default, modifiable by a malware= option. The list separator for
771 the options can now be changed in the usual way. Bug 68.
773 JH/10 The smtp_receive_timeout main option is now expanded before use.
775 JH/11 The incoming_interface log option now also enables logging of the
776 local interface on delivery outgoing connections.
778 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
779 if the interface and destination host and port all match.
781 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
784 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
785 Patch from Andrew Lewis.
787 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
788 now supports optional time-restrictions, weighting, and priority
789 modifiers per server. Patch originally by <rommer@active.by>.
791 JH/16 The spamd_address main option now supports a mixed list of local
792 and remote servers. Remote servers can be IPv6 addresses, and
793 specify a port-range.
795 JH/17 Bug 68: The spamd_address main option now supports an optional
796 timeout value per server.
798 JH/18 Bug 1581: Router and transport options headers_add/remove can
799 now have the list separator specified.
801 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
804 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
807 JH/21 Support for the A6 type of dns record is withdrawn.
809 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
810 rather than the verbs used.
812 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
813 from 255 to 1024 chars.
815 JH/24 Verification callouts now attempt to use TLS by default.
817 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
818 are generic router options now. The defaults didn't change.
820 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
821 Original patch from Alexander Shikoff, worked over by JH.
823 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
824 files if ZCAT_COMMAND is not executable.
826 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
828 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
830 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
831 Normally benign, it bites when the pair was led to by a CNAME;
832 modern usage is to not canonicalize the domain to a CNAME target
833 (and we were inconsistent anyway for A-only vs AAAA+A).
835 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
837 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
838 when evaluating $sender_host_dnssec.
840 JH/31 Check the HELO verification lookup for DNSSEC, adding new
841 $sender_helo_dnssec variable.
843 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
845 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
847 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
849 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
850 documented as working, but never had. Support all but $spam_report.
852 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
853 added for tls authenticator.
855 HS/03 Add perl_taintmode main config option
861 TL/01 When running the test suite, the README says that variables such as
862 no_msglog_check are global and can be placed anywhere in a specific
863 test's script, however it was observed that placement needed to be near
864 the beginning for it to behave that way. Changed the runtest perl
865 script to read through the entire script once to detect and set these
866 variables, reset to the beginning of the script, and then run through
867 the script parsing/test process like normal.
869 TL/02 The BSD's have an arc4random API. One of the functions to induce
870 adding randomness was arc4random_stir(), but it has been removed in
871 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
872 function when detected.
874 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
875 cause callback expansion.
877 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
878 syntax errors in an expansion can be treated as a string instead of
879 logging or causing an error, due to the internal use of bool_lax
880 instead of bool when processing it.
882 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
883 server certificates when making smtp deliveries.
885 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
887 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
889 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
891 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
892 Merged patch from Sebastian Wiedenroth.
894 JH/05 Fix results-pipe from transport process. Several recipients, combined
895 with certificate use, exposed issues where response data items split
896 over buffer boundaries were not parsed properly. This eventually
897 resulted in duplicates being sent. This issue only became common enough
898 to notice due to the introduction of connection certificate information,
899 the item size being so much larger. Found and fixed by Wolfgang Breyha.
901 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
902 size buffer was used, resulting in syntax errors when an expansion
905 JH/07 Add support for directories of certificates when compiled with a GnuTLS
906 version 3.3.6 or later.
908 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
909 is EXPERIMENTAL_EVENT, the main-configuration and transport options
910 both become "event_action", the variables become $event_name, $event_data
911 and $event_defer_errno. There is a new variable $verify_mode, usable in
912 routers, transports and related events. The tls:cert event is now also
913 raised for inbound connections, if the main configuration event_action
916 TL/06 In test suite, disable OCSP for old versions of openssl which contained
917 early OCSP support, but no stapling (appears to be less than 1.0.0).
919 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
920 server certificate names available under the smtp transport option
921 "tls_verify_cert_hostname" now do not permit multi-component wildcard
924 JH/10 Time-related extraction expansions from certificates now use the main
925 option "timezone" setting for output formatting, and are consistent
926 between OpenSSL and GnuTLS compilations. Bug 1541.
928 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
929 encoded parameter in the incoming message. Bug 1558.
931 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
932 include certificate info, eximon was claiming there were spoolfile
935 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
937 JH/14 Log delivery-related information more consistently, using the sequence
938 "H=<name> [<ip>]" wherever possible.
940 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
941 are problematic for Debian distribution, omit them from the release
944 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
946 JH/16 Fix string representation of time values on 64bit time_t architectures.
949 JH/17 Fix a null-indirection in certextract expansions when a nondefault
950 output list separator was used.
955 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
956 checkers that were complaining about end of non-void function with no
959 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
960 This was a regression introduced in 4.83 by another bugfix.
962 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
964 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
965 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
971 TF/01 Correctly close the server side of TLS when forking for delivery.
973 When a message was received over SMTP with TLS, Exim failed to clear up
974 the incoming connection properly after forking off the child process to
975 deliver the message. In some situations the subsequent outgoing
976 delivery connection happened to have the same fd number as the incoming
977 connection previously had. Exim would try to use TLS and fail, logging
978 a "Bad file descriptor" error.
980 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
981 utilities have not been installed.
983 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
984 temporary space as the ACL may create new global variables.
986 TL/01 LDAP support uses per connection or global context settings, depending
987 upon the detected version of the libraries at build time.
989 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
990 to extract and use the src ip:port in logging and expansions as if it
991 were a direct connection from the outside internet. PPv2 support was
992 updated based on HAProxy spec change in May 2014.
994 JH/02 Add ${listextract {number}{list}{success}{fail}}.
996 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
997 Properly escape header and check for NULL return.
999 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1002 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1004 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1005 characters in header names, implemented as a verify condition.
1006 Contributed by Michael Fischer v. Mollard.
1008 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1009 results permerror and temperror. Previous values are deprecated but
1010 still accepted. In a future release, err_perm and err_temp will be
1011 completely removed, which will be a backward incompatibility if the
1012 ACL tests for either of these two old results. Patch contributed by
1013 user bes-internal on the mailing list.
1015 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1017 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1018 selectors, in both main and reject logs.
1020 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1023 JH/07 Add malware type "sock" for talking to simple daemon.
1025 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1027 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1028 routers/transports under cutthrough routing.
1030 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1031 numbers. Touch up "bool" conditional to keep the same definition.
1033 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1035 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1037 JH/12 Expand items in router/transport headers_add or headers_remove lists
1038 individually rather than the list as a whole. Bug 1452.
1040 Required for reasonable handling of multiple headers_ options when
1041 they may be empty; requires that headers_remove items with embedded
1042 colons must have them doubled (or the list-separator changed).
1044 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1045 view the policy declared in the DMARC record. Currently, $dmarc_status
1046 is a combined value of both the record presence and the result of the
1049 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1051 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1052 dnslookup router and the smtp transport (applying to the forward
1055 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1056 of ldap servers used for a specific lookup. Patch provided by Heiko
1059 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1060 New variable $lookup_dnssec_authenticated for observability.
1062 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1063 Patch submitted by Lars Timman.
1065 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1067 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1068 Requires trusted mode and valid format message id, aborts otherwise.
1069 Patch contributed by Heiko Schlichting.
1071 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1072 certextract with support for various fields. Bug 1358.
1074 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1075 is requested by default, modifiable by smtp transport option
1078 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1079 operate on certificate variables to give certificate fingerprints
1080 Also new ${sha256:cert_variable}.
1082 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1084 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1085 Christian Aistleitner.
1087 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1089 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1090 file. Patch from Wolfgang Breyha.
1092 JH/25 Expand the coverage of the delivery $host and $host_address to
1093 client authenticators run in verify callout. Bug 1476.
1095 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1096 align with daemon_smtp_ports. Bug 72.
1098 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1099 support and error reporting did not work properly.
1101 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1102 and is readable. Patch from Andrew Colin Kissa.
1104 TL/14 Enhance documentation of ${run expansion and how it parses the
1105 commandline after expansion, particularly in the case when an
1106 unquoted variable expansion results in an empty value.
1108 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1110 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1111 by renaming away from it. Addresses GH issue 15, reported by
1114 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1116 TL/15 SECURITY: prevent double expansion in math comparison functions
1117 (can expand unsanitized data). Not remotely exploitable.
1124 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1126 PP/02 Make -n do something, by making it not do something.
1127 When combined with -bP, the name of an option is not output.
1129 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1132 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1133 $sender_host_name and config options to manage this, and basic check
1136 PP/05 DSCP support for outbound connections and control modifier for inbound.
1138 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1139 (Only plugin which currently uses this is kerberos4, which nobody should
1140 be using, but we should make it available and other future plugins might
1141 conceivably use it, even though it would break NAT; stuff *should* be
1142 using channel bindings instead).
1144 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1145 name; added for Sendmail compatibility; requires admin caller.
1146 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1147 just ignore it); requires trusted caller.
1148 Also parse but ignore: -Ac -Am -X<logfile>
1151 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1153 TL/02 Add +smtp_confirmation as a default logging option.
1155 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1156 Patch by Magnus Holmgren from 2007-02-20.
1158 TL/04 Bugzilla 1281 - Spec typo.
1159 Bugzilla 1283 - Spec typo.
1160 Bugzilla 1290 - Spec grammar fixes.
1162 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1164 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1166 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1167 mailing list by by Dmitry Isaikin.
1169 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1171 JH/02 Support "G" suffix to numbers in ${if comparisons.
1173 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1175 NM/01 Bugzilla 1197 - Spec typo
1176 Bugzilla 1196 - Spec examples corrections
1178 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1180 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1181 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1184 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1185 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1187 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1188 "acl {{name}{arg}...}", and optional args on acl condition
1191 JH/05 Permit multiple router/transport headers_add/remove lines.
1193 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1195 JH/07 Avoid using a waiting database for a single-message-only transport.
1196 Performance patch from Paul Fisher. Bugzilla 1262.
1198 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1201 JH/09 Add $headers_added variable, with content from use of ACL modifier
1202 add_header (but not yet added to the message). Bugzilla 199.
1204 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1205 Pulled from Bugzilla 817 by Wolfgang Breyha.
1207 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1209 (nb: this is the same fix as in Exim 4.80.1)
1211 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1214 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1217 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1219 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1220 advertise SMTP AUTH mechanism to us, instead of a generic
1221 protocol violation error. Also, make Exim more robust to bad
1222 data from the Dovecot auth socket.
1224 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1226 When a queue runner is handling a message, Exim first routes the
1227 recipient addresses, during which it prunes them based on the retry
1228 hints database. After that it attempts to deliver the message to
1229 any remaining recipients. It then updates the hints database using
1232 So if a recipient address works intermittently, it can get repeatedly
1233 deferred at routing time. The retry hints record remains fresh so the
1234 address never reaches the final cutoff time.
1236 This is a fairly common occurrence when a user is bumping up against
1237 their storage quota. Exim had some logic in its local delivery code
1238 to deal with this. However it did not apply to per-recipient defers
1239 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1241 This change adds a proper retry rule check during routing so that the
1242 final cutoff time is checked against the message's age. We only do
1243 this check if there is an address retry record and there is not a
1244 domain retry record; this implies that previous attempts to handle
1245 the address had the retry_use_local_parts option turned on. We use
1246 this as an approximation for the destination being like a local
1247 delivery, as in LMTP.
1249 I suspect this new check makes the old local delivery cutoff check
1250 redundant, but I have not verified this so I left the code in place.
1252 TF/02 Correct gecos expansion when From: is a prefix of the username.
1254 Test 0254 submits a message to Exim with the header
1258 When I ran the test suite under the user fanf2, Exim expanded
1259 the header to contain my full name, whereas it should have added
1260 a Resent-Sender: header. It erroneously treats any prefix of the
1261 username as equal to the username.
1263 This change corrects that bug.
1265 GF/01 DCC debug and logging tidyup
1266 Error conditions log to paniclog rather than rejectlog.
1267 Debug lines prefixed by "DCC: " to remove any ambiguity.
1269 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1271 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1272 Bug spotted by Jeremy Harris; was flawed since initial commit.
1273 Would have resulted in OCSP responses post-SNI triggering an Exim
1274 NULL dereference and crash.
1276 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1278 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1279 Bug detection, analysis and fix by Samuel Thibault.
1280 Bugzilla 1331, Debian bug #698092.
1282 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1284 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1285 Server implementation by Todd Lyons, client by JH.
1286 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1287 config variable "prdr_enable" controls whether the server
1288 advertises the facility. If the client requests PRDR a new
1289 acl_data_smtp_prdr ACL is called once for each recipient, after
1290 the body content is received and before the acl_smtp_data ACL.
1291 The client is controlled by both of: a hosts_try_prdr option
1292 on the smtp transport, and the server advertisement.
1293 Default client logging of deliveries and rejections involving
1294 PRDR are flagged with the string "PRDR".
1296 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1297 fclose(). Diagnosis by Todd Lyons.
1299 PP/17 Update configure.default to handle IPv6 localhost better.
1300 Patch by Alain Williams (plus minor tweaks).
1303 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1304 This is now consistent with GnuTLS, and is now documented: the
1305 previous undocumented portable approach to treating the option as
1306 unset was to force an expansion failure. That still works, and
1307 an empty string is now equivalent.
1309 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1310 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1311 not performing validation itself.
1313 PP/20 Added force_command boolean option to pipe transport.
1314 Patch from Nick Koston, of cPanel Inc.
1316 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1319 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1321 PP/21 Fix eximon continuous updating with timestamped log-files.
1322 Broken in a format-string cleanup in 4.80, missed when I repaired the
1323 other false fix of the same issue.
1324 Report and fix from Heiko Schlichting.
1327 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1328 Report from Prashanth Katuri.
1330 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1331 It's SecureTransport, so affects any MacOS clients which use the
1332 system-integrated TLS libraries, including email clients.
1334 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1335 using a MIME ACL for non-SMTP local injection.
1336 Report and assistance in diagnosis by Warren Baker.
1338 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1340 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1342 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1343 last id that failed so it may be referenced in subsequent ACL's.
1345 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1348 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1349 ldap library initialization, allowing self-signed CA's to be
1350 used. Also properly sets require_cert option later in code by
1351 using NULL (global ldap config) instead of ldap handle (per
1352 session). Bug diagnosis and testing by alxgomz.
1354 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1355 the src/util/ subdirectory.
1357 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1358 renamed to Transport Post Delivery Action by Jeremy Harris, as
1361 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1362 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1363 redis_servers = needs to be configured which will be used by the redis
1364 lookup. Patch from Warren Baker, of The Packet Hub.
1366 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1368 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1369 hostname or reverse DNS when processing a host list. Used suggestions
1370 from multiple comments on this bug.
1372 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1374 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1375 Missed a few lines, added it to make the runtest require no keyboard
1378 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1379 contains upper case chars. Make router use caseful_local_part.
1381 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1382 support when GnuTLS has been built with p11-kit.
1388 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1390 This, or similar/improved, will also be change PP/11 of 4.82.
1396 PP/01 Handle short writes when writing local log-files.
1397 In practice, only affects FreeBSD (8 onwards).
1398 Bugzilla 1053, with thanks to Dmitry Isaikin.
1400 NM/01 Bugzilla 949 - Documentation tweak
1402 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1405 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1407 PP/02 Implemented gsasl authenticator.
1409 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1411 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1412 `pkg-config foo` for cflags/libs.
1414 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1415 with rest of GSASL and with heimdal_gssapi.
1417 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1418 `pkg-config foo` for cflags/libs for the TLS implementation.
1420 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1421 properties get this fed in as external SSF. A number of robustness
1422 and debugging improvements to the cyrus_sasl authenticator.
1424 PP/08 cyrus_sasl server now expands the server_realm option.
1426 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1427 Patch by Jeremy Harris.
1429 PP/10 Added dbmjz lookup type.
1431 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1433 PP/12 MAIL args handles TAB as well as SP, for better interop with
1434 non-compliant senders.
1435 Analysis and variant patch by Todd Lyons.
1437 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1438 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1439 Patch from Dirk Mueller <dmueller@suse.com>
1441 PP/13 tls_peerdn now print-escaped for spool files.
1442 Observed some $tls_peerdn in wild which contained \n, which resulted
1443 in spool file corruption.
1445 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1446 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1447 or write after TLS renegotiation, which otherwise led to messages
1450 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1451 as a tracking header (ie: a signed header comes before the signature).
1452 Patch from Wolfgang Breyha.
1454 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1455 comma-sep list; embedded commas doubled.
1457 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1459 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1461 Report and patch from Dmitry Banschikov.
1463 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1464 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1465 failures. We appear to now support TLS1.1+ with Exim.
1467 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1468 lets Exim select keys and certificates based upon TLS SNI from client.
1469 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1470 before an outbound SMTP session. New log_selector, +tls_sni.
1472 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1473 NULL dereference. Report and patch from Alun Jones.
1475 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1476 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1477 Not seeing resolver debug output on NetBSD, but suspect this is a
1478 resolver implementation change.
1480 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1481 Left warnings. Added "eximon gdb" invocation mode.
1483 PP/21 Defaulting "accept_8bitmime" to true, not false.
1485 PP/22 Added -bw for inetd wait mode support.
1487 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1488 locate the relevant includes and libraries. Made this the default.
1490 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1491 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1493 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1494 This may cause build issues on older platforms.
1496 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1497 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1498 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1499 Added SNI support via GnuTLS too.
1500 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1502 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1504 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1505 Applied second patch from Janne, implementing suggestion to default
1506 multiple-strings-in-record handling to match SPF spec.
1508 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1510 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1511 read-only, out of scope).
1512 Patch from Wolfgang Breyha, report from Stuart Northfield.
1514 PP/29 Fix three issues highlighted by clang analyser static analysis.
1515 Only crash-plausible issue would require the Cambridge-specific
1516 iplookup router and a misconfiguration.
1517 Report from Marcin Mirosław.
1519 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1521 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1522 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1523 As part of this, removing so much warning spew let me fix some minor
1524 real issues in debug logging.
1526 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1527 assignment on my part. Fixed.
1529 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1530 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1531 Janne Snabb (who went above and beyond: thank you).
1533 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1534 string otherwise requires a connection and a bunch more work and it's
1535 relatively easy to get wrong. Should also expose TLS library linkage
1538 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1539 64-bit ${eval} (JH/03).
1541 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1542 GNU libc to support some of the 64-bit stuff, should not lead to
1543 conflicts. Defined before os.h is pulled in, so if a given platform
1544 needs to override this, it can.
1546 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1547 protection layer was required, which is not implemented.
1548 Bugzilla 1254, patch from Wolfgang Breyha.
1550 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1551 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1552 tls_dhparam take prime identifiers. Also unbreak combination of
1553 OpenSSL+DH_params+TLSSNI.
1555 PP/39 Disable SSLv2 by default in OpenSSL support.
1561 PP/01 Solaris build fix for Oracle's LDAP libraries.
1562 Bugzilla 1109, patch from Stephen Usher.
1564 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1566 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1569 TF/02 Fix a couple more cases where we did not log the error message
1570 when unlink() failed. See also change 4.74-TF/03.
1572 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1573 lock up or crash if it happened to be inside a call to libc when it
1574 got a SIGUSR1 from exiwhat.
1576 The SIGUSR1 handler appends the current process status to the process
1577 log which is later printed by exiwhat. It used to use the general
1578 purpose logging code to do this, but several functions it calls are
1579 not safe for signals.
1581 The new output code in the SIGUSR1 handler is specific to the process
1582 log, and simple enough that it's easy to inspect for signal safety.
1583 Removing some special cases also simplifies the general logging code.
1584 Removing the spurious timestamps from the process log simplifies
1587 TF/04 Improved ratelimit ACL condition.
1589 The /noupdate option has been deprecated in favour of /readonly which
1590 has clearer semantics. The /leaky, /strict, and /readonly update modes
1591 are mutually exclusive. The update mode is no longer included in the
1592 database key; it just determines when the database is updated. (This
1593 means that when you upgrade Exim will forget old rate measurements.)
1595 Exim now checks that the per_* options are used with an update mode that
1596 makes sense for the current ACL. For example, when Exim is processing a
1597 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1598 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1599 must specify per_mail/readonly. If you omit the update mode it defaults to
1600 /leaky where that makes sense (as before) or /readonly where required.
1602 The /noupdate option is now undocumented but still supported for
1603 backwards compatibility. It is equivalent to /readonly except that in
1604 ACLs where /readonly is required you may specify /leaky/noupdate or
1605 /strict/noupdate which are treated the same as /readonly.
1607 A useful new feature is the /count= option. This is a generalization
1608 of the per_byte option, so that you can measure the throughput of other
1609 aggregate values. For example, the per_byte option is now equivalent
1610 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1612 The per_rcpt option has been generalized using the /count= mechanism
1613 (though it's more complicated than the per_byte equivalence). When it is
1614 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1615 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1616 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1617 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1618 Note that using per_rcpt with a non-readonly update mode in more than
1619 one ACL will cause the recipients to be double-counted. (The per_mail
1620 and per_byte options don't have this problem.)
1622 The handling of very low rates has changed slightly. If the computed rate
1623 is less than the event's count (usually one) then this event is the first
1624 after a long gap. In this case the rate is set to the same as this event's
1625 count, so that the first message of a spam run is counted properly.
1627 The major new feature is a mechanism for counting the rate of unique
1628 events. The new per_addr option counts the number of different
1629 recipients that someone has sent messages to in the last time period. It
1630 behaves like per_rcpt if all the recipient addresses are different, but
1631 duplicate recipient addresses do not increase the measured rate. Like
1632 the /count= option this is a general mechanism, so the per_addr option
1633 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1634 example, measure the rate that a client uses different sender addresses
1635 with the options per_mail/unique=$sender_address. There are further
1636 details in the main documentation.
1638 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1640 TF/06 Removed a few PCRE remnants.
1642 TF/07 Automatically extract Exim's version number from tags in the git
1643 repository when doing development or release builds.
1645 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1646 Bugzilla 879. Patch from Paul Fisher.
1648 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1649 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1652 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1654 PP/05 Variable $av_failed, true if the AV scanner deferred.
1655 Bugzilla 1078. Patch from John Horne.
1657 PP/06 Stop make process more reliably on build failure.
1658 Bugzilla 1087. Patch from Heiko Schlittermann.
1660 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1661 Bugzilla 1089. Patch from Heiko Schlittermann.
1663 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1664 Bugzilla 1131. Patch from Holger Weiß.
1666 PP/09 Handle IPv6 addresses with SPF.
1667 Bugzilla 860. Patch from Wolfgang Breyha.
1669 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1671 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1674 PP/11 match_* no longer expand right-hand-side by default.
1675 New compile-time build option, EXPAND_LISTMATCH_RHS.
1676 New expansion conditions, "inlist", "inlisti".
1678 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1680 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1682 PP/14 fix log_write() format string regression from TF/03.
1683 Bugzilla 1152. Patch from Dmitry Isaikin.
1689 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1691 PP/02 Harmonised TLS library version reporting; only show if debugging.
1692 Layout now matches that introduced for other libraries in 4.74 PP/03.
1694 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1696 PP/04 New "dns_use_edns0" global option.
1698 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1701 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1702 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1704 TK/01 Updated PolarSSL code to 0.14.2.
1705 Bugzilla 1097. Patch from Andreas Metzler.
1707 PP/07 Catch divide-by-zero in ${eval:...}.
1708 Fixes bugzilla 1102.
1710 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1713 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1714 format-string attack -- SECURITY: remote arbitrary code execution.
1716 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1717 time unintentionally subject to list matching rules, letting the header
1718 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1719 arbitrary string expansion). This allowed for information disclosure.
1721 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1722 INT_MIN/-1 -- value coerced to INT_MAX.
1728 NM/01 Workaround for PCRE version dependency in version reporting
1731 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1732 This fixes portability to compilers other than gcc, notably
1733 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1735 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1736 makefiles for portability to HP-UX and POSIX correctness.
1738 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1739 Also via indented variable definition in the Makefile.
1740 (Debugging by Oliver Heesakkers).
1742 PP/02 Restore caching of spamd results with expanded spamd_address.
1743 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1745 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1746 Improves build reliability. Fix from: Frank Elsner
1748 NM/02 Fix wide character breakage in the rfc2047 coding
1749 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1751 NM/03 Allow underscore in dnslist lookups
1752 Fixes bug 1026. Patch from Graeme Fowler
1754 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1755 Code patches from Adam Ciarcinski of NetBSD.
1757 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1760 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1761 is logged, to avoid truncation. Patch from John Horne.
1763 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1764 Patch from Jakob Hirsch.
1766 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1767 SQL string expansion failure details.
1768 Patch from Andrey Oktyabrski.
1770 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1771 Patch from Simon Arlott.
1773 PP/09 New lookups functionality failed to compile on old gcc which rejects
1774 extern declarations in function scope.
1775 Patch from Oliver Fleischmann
1777 PP/10 Use sig_atomic_t for flags set from signal handlers.
1778 Check getgroups() return and improve debugging.
1779 Fixed developed for diagnosis in bug 927 (which turned out to be
1782 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1783 Patch from Mark Zealey.
1785 PP/12 Bugzilla 1056: Improved spamd server selection.
1786 Patch from Mark Zealey.
1788 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1789 Based on patch from Heiko Schlittermann.
1791 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1792 Patch from Uwe Doering, sign-off by Michael Haardt.
1794 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1795 variable declaration deep within a block. Bug and patch from
1798 PP/15 lookups-Makefile IRIX compatibility coercion.
1800 PP/16 Make DISABLE_DKIM build knob functional.
1802 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1803 Patch by Simon Arlott
1805 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1806 variable argument macros. Our copy now differs from upstream.
1812 TF/01 Failure to get a lock on a hints database can have serious
1813 consequences so log it to the panic log.
1815 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1816 controlled using the smtp_confirmation log selector.
1818 TF/03 Include the error message when we fail to unlink a spool file.
1820 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1821 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1822 for maintaining out-of-tree patches for some time.
1824 PP/01 Bugzilla 139: Documentation and portability issues.
1825 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1826 Handle per-OS dynamic-module compilation flags.
1828 PP/02 Let /dev/null have normal permissions.
1829 The 4.73 fixes were a little too stringent and complained about the
1830 permissions on /dev/null. Exempt it from some checks.
1831 Reported by Andreas M. Kirchwitz.
1833 PP/03 Report version information for many libraries, including
1834 Exim version information for dynamically loaded libraries. Created
1835 version.h, now support a version extension string for distributors
1836 who patch heavily. Dynamic module ABI change.
1838 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1839 privilege escalation vulnerability whereby the Exim run-time user
1840 can cause root to append content of the attacker's choosing to
1843 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1846 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1847 If dropping privileges for untrusted macros, we disabled normal logging
1848 on the basis that it would fail; for the Exim run-time user, this is not
1849 the case, and it resulted in successful deliveries going unlogged.
1850 Fixed. Reported by Andreas Metzler.
1856 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1857 only prepend for the Resent-* case. Fixes regression introduced in
1858 Exim 4.70 by NM/22 for Bugzilla 607.
1860 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1861 increasing numbers of administrators be bitten by this.
1863 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1865 PP/03 Bugzilla 994: added openssl_options main configuration option.
1867 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1869 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1871 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1873 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1874 without a peer certificate, leading to a segfault because of an
1875 assumption that peers always have certificates. Be a little more
1876 paranoid. Problem reported by Martin Tscholak.
1878 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1879 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1880 NB: ClamAV planning to remove STREAM in "middle of 2010".
1881 CL also introduces -bmalware, various -d+acl logging additions and
1882 more caution in buffer sizes.
1884 PP/09 Implemented reverse_ip expansion operator.
1886 PP/10 Bugzilla 937: provide a "debug" ACL control.
1888 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1890 PP/12 Bugzilla 973: Implement --version.
1892 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1894 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1896 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1898 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1899 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1900 ignore trailing whitespace.
1902 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1904 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1907 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1908 Notification from Dr Andrew Aitchison.
1910 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1911 ExtendedDetectionInfo response format.
1912 Notification from John Horne.
1914 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1917 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1918 XSL and documented dependency on system catalogs, with examples of how
1921 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1924 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1925 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1926 configuration file which is writeable by the Exim user or group.
1928 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1929 of configuration files to cover files specified with the -C option if
1930 they are going to be used with root privileges, not just the default
1933 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1934 option (effectively making it always true).
1936 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1937 files to be used while preserving root privileges.
1939 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1940 that rogue child processes cannot use them.
1942 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1943 run-time user, instead of root.
1945 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1946 Exim run-time user without dropping privileges.
1948 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1949 result string, instead of calling string_vformat() twice with the same
1952 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1953 for other users. Others should always drop root privileges if they use
1954 -C on the command line, even for a whitelisted configure file.
1956 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1958 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1964 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1965 $data_path, and $header_path variables; fixed documentation bugs and
1968 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1969 exipick to access non-standard spools, including the "frozen" queue
1972 NM/01 Bugzilla 965: Support mysql stored procedures.
1973 Patch from Alain Williams
1975 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1977 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1978 Patch from Andreas Metzler
1980 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1981 Patch from Kirill Miazine
1983 NM/05 Bugzilla 671: Added umask to procmail example.
1985 JJ/03 installed exipick 20100323.0, fixing doc bug
1987 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1988 directory. Notification and patch from Dan Rosenberg.
1990 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1992 TK/02 Improve log output when DKIM signing operation fails.
1994 MH/01 Treat the transport option dkim_domain as a colon separated
1995 list, not as a single string, and sign the message with each element,
1996 omitting multiple occurences of the same signer.
1998 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1999 Bugzilla 985, 986. Patch by Simon Arlott
2001 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2002 Patch by Simon Arlott
2004 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2005 MBX locking. Notification from Dan Rosenberg.
2011 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2013 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2015 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2017 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2019 NM/04 Bugzilla 915: Flag broken perl installation during build.
2025 TK/01 Added patch by Johannes Berg that expands the main option
2026 "spamd_address" if it starts with a dollar sign.
2028 TK/02 Write list of recipients to X-Envelope-Sender header when building
2029 the mbox-format spool file for content scanning (suggested by Jakob
2032 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2033 (http://www.dcc-servers.net/) support via dccifd. Activated by
2034 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2036 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2037 by Mark Daniel Reidel <mr@df.eu>.
2039 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2040 When building exim an external PCRE library is now needed -
2041 PCRE is a system library on the majority of modern systems.
2042 See entry on PCRE_LIBS in EDITME file.
2044 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2045 conversation. Added nologin parameter to request.
2046 Patch contributed by Kirill Miazine.
2048 TF/01 Do not log submission mode rewrites if they do not change the address.
2050 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2052 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2053 log files in place. Contributed by Roberto Lima.
2055 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2057 TF/03 Bugzilla 615: When checking the local_parts router precondition
2058 after a local_part_suffix or local_part_prefix option, Exim now
2059 does not use the address's named list lookup cache, since this
2060 contains cached lookups for the whole local part.
2062 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2063 Robert Millan. Documentation is in experimental-spec.txt.
2065 TF/04 Bugzilla 668: Fix parallel build (make -j).
2067 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2069 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2070 Patch provided by Jan Srzednicki.
2072 TF/05 Leading white space used to be stripped from $spam_report which
2073 wrecked the formatting. Now it is preserved.
2075 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2076 that they are available at delivery time.
2078 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2080 TF/08 TLS error reporting now respects the incoming_interface and
2081 incoming_port log selectors.
2083 TF/09 Produce a more useful error message if an SMTP transport's hosts
2084 setting expands to an empty string.
2086 NM/06 Bugzilla 744: EXPN did not work under TLS.
2087 Patch provided by Phil Pennock.
2089 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2090 Patch provided by Richard Godbee.
2092 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2093 acl_smtp_notquit, added index entry.
2095 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2096 Patch provided by Eugene Bujak.
2098 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2099 accept(). Patch provided by Maxim Dounin.
2101 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2102 Patch provided by Phil Pennock.
2104 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2106 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2107 Patch provided by Brad "anomie" Jorsch.
2109 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2110 Patch provided by Dean Brooks.
2112 TK/05 Add native DKIM support (does not depend on external libraries).
2114 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2115 Patch provided by Graeme Fowler.
2117 NM/16 Bugzilla 851: Documentation example syntax fix.
2119 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2121 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2124 NM/19 Bugzilla 745: TLS version reporting.
2125 Patch provided by Phil Pennock.
2127 NM/20 Bugzilla 167: bool: condition support.
2128 Patch provided by Phil Pennock.
2130 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2131 clients. Patch provided by Phil Pennock.
2133 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2134 Patch provided by Brad "anomie" Jorsch.
2136 NM/23 Bugzilla 687: Fix misparses in eximstats.
2137 Patch provided by Heiko Schlittermann.
2139 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2140 Patch provided by Heiko Schlittermann.
2142 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2143 plus update to original patch.
2145 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2147 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2148 Patch provided by David Brownlee.
2150 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2152 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2154 NM/30 Bugzilla 888: TLS documentation bugfixes.
2156 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2158 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2159 Unlike the original bugzilla I have changed all shell scripts in src tree.
2161 NM/33 Bugzilla 898: Transport filter timeout fix.
2162 Patch by Todd Rinaldo.
2164 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2165 Patch by Serge Demonchaux.
2167 NM/35 Bugzilla 39: Base64 decode bug fixes.
2168 Patch by Jakob Hirsch.
2170 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2172 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2174 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2176 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2182 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2183 ALT-N's libdkim that I have put here:
2184 http://duncanthrax.net/exim-experimental/
2186 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2187 were called 'true' and it seems that C99 defines that as a reserved
2188 keyword to be used with 'bool' variable types. That means you could
2189 not include C99-style headers which use bools without triggering
2190 build errors in sieve.c.
2192 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2193 as mailq or other aliases. Changed the --help handling significantly
2194 to do whats expected. exim_usage() emits usage/help information.
2196 SC/01 Added the -bylocaldomain option to eximstats.
2198 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2200 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2202 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2208 PH/01 Another patch from the Sieve maintainer.
2210 PH/02 When an IPv6 address is converted to a string for single-key lookup
2211 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2212 dots are used instead of colons so that keys in lsearch files need not
2213 contain colons. This was done some time before quoting was made available
2214 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2215 (notated using the quote facility) so as to distinguish them from IPv4
2216 keys. This meant that lookups for IP addresses in host lists did not work
2217 for iplsearch lookups.
2219 This has been fixed by arranging for IPv6 addresses to be expressed with
2220 colons if the lookup type is iplsearch. This is not incompatible, because
2221 previously such lookups could never work.
2223 The situation is now rather anomalous, since one *can* have colons in
2224 ordinary lsearch keys. However, making the change in all cases is
2225 incompatible and would probably break a number of configurations.
2227 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2230 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2231 conversion specification without a maximum field width, thereby enabling
2232 a rogue spamd server to cause a buffer overflow. While nobody in their
2233 right mind would setup Exim to query an untrusted spamd server, an
2234 attacker that gains access to a server running spamd could potentially
2235 exploit this vulnerability to run arbitrary code as the Exim user.
2237 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2238 $primary_hostname instead of what libspf2 thinks the hosts name is.
2240 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2241 a directory entry by the name of the lookup key. Previously, if a
2242 symlink pointed to a non-existing file or a file in a directory that
2243 Exim lacked permissions to read, a lookup for a key matching that
2244 symlink would fail. Now it is enough that a matching directory entry
2245 exists, symlink or not. (Bugzilla 503.)
2247 PH/03 The body_linecount and body_zerocount variables are now exported in the
2250 PH/04 Added the $dnslist_matched variable.
2252 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2253 This means they are set thereafter only if the connection becomes
2256 PH/06 Added the client_condition to authenticators so that some can be skipped
2257 by clients under certain conditions.
2259 PH/07 The error message for a badly-placed control=no_multiline_responses left
2260 "_responses" off the end of the name.
2262 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2264 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2265 (without spaces) instead of just copying the configuration text.
2267 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2269 PH/11 Added $max_received_linelength.
2271 PH/12 Added +ignore_defer and +include_defer to host lists.
2273 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2274 way in which PCRE > 7.0 is built.
2276 PH/14 Implemented queue_only_load_latch.
2278 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2279 MAIL command. The effect was to mangle the value on 64-bit systems.
2281 PH/16 Another patch from the Sieve maintainer.
2283 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2285 PH/18 If a system quota error occurred while trying to create the file for
2286 a maildir delivery, the message "Mailbox is full" was not appended to the
2287 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2288 only to a quota excession during the actual writing of the file.
2290 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2291 characters?) which causes problems in log lines. The DN values are now
2292 passed through string_printing() before being added to log lines.
2294 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2295 and InterBase are left for another time.)
2297 PH/21 Added message_body_newlines option.
2299 PH/22 Guard against possible overflow in moan_check_errorcopy().
2301 PH/23 POSIX allows open() to be a macro; guard against that.
2303 PH/24 If the recipient of an error message contained an @ in the local part
2304 (suitably quoted, of course), incorrect values were put in $domain and
2305 $local_part during the evaluation of errors_copy.
2311 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2312 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2315 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2316 issue a MAIL command.
2318 PH/02 In an ACL statement such as
2320 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2322 if a client was not listed at all, or was listed with a value other than
2323 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2324 the condition was not true (as it should be), so access was not denied.
2325 The bug was that the ! inversion was incorrectly passed on to the second
2326 item. This has been fixed.
2328 PH/03 Added additional dnslists conditions == and =& which are different from
2329 = and & when the dns lookup returns more than one IP address.
2331 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2332 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2334 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2335 FSYNC, which compiles an option called disable_fsync that allows for
2336 bypassing fsync(). The documentation is heavily laced with warnings.
2338 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2340 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2341 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2342 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2343 including adding "make clean"; (3) Added -fPIC when compiling the test
2344 dynamically loaded module, to get rid of a warning.
2346 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2347 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2348 = 0s. The bug is otherwise harmless.
2350 PH/07 There was a bug in the dovecot authenticator such that the value of
2351 $auth1 could be overwritten, and so not correctly preserved, after a
2352 successful authentication. This usually meant that the value preserved by
2353 the server_setid option was incorrect.
2355 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2357 PH/09 Installed PCRE release 7.0.
2359 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2360 run for batched SMTP input. It is now run at the start of every message
2361 in the batch. While fixing this I discovered that the process information
2362 (output by running exiwhat) was not always getting set for -bs and -bS
2363 input. This is fixed, and it now also says "batched" for BSMTP.
2365 PH/11 Added control=no_pipelining.
2367 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2368 patch, slightly modified), and move the expansion of helo_data till after
2369 the connection is made in the smtp transport (so it can use these
2372 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2374 PH/14 Added log_selector = +pid.
2376 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2378 PH/16 Add ${if forany and ${if forall.
2380 PH/17 Added dsn_from option to vary the From: line in DSNs.
2382 PH/18 Flush SMTP output before performing a callout, unless control =
2383 no_callout_flush is set.
2385 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2386 was true (the default) a successful delivery failed to delete the retry
2387 item, thus causing premature timeout of the address. The bug is now
2390 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2392 PH/21 Long custom messages for fakedefer and fakereject are now split up
2393 into multiline responses in the same way that messages for "deny" and
2394 other ACL rejections are.
2396 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2397 with slight modification.
2399 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2400 draft, changing the syntax and factoring some duplicate code".
2402 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2403 for deliveries of the second and subsequent messages over the same SMTP
2406 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2407 ${reduce, with only minor "tidies".
2409 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2411 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2412 expansion side effects.
2414 PH/27 When a message times out after an over-quota error from an Exim-imposed
2415 quota, the bounce message says "mailbox is full". This message was not
2416 being given when it was a system quota that was exceeded. It now should
2419 MH/03 Made $recipients available in local_scan(). local_scan() already has
2420 better access to the recipient list through recipients_list[], but
2421 $recipients can be useful in postmaster-provided expansion strings.
2423 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2424 in the case of a MAIL command with additional options following the
2425 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2426 were accidentally chopped off.
2428 PH/29 SMTP synchronization checks are implemented when a command is read -
2429 there is a check that no more input is waiting when there shouldn't be
2430 any. However, for some commands, a delay in an ACL can mean that it is
2431 some time before the response is written. In this time, more input might
2432 arrive, invalidly. So now there are extra checks after an ACL has run for
2433 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2434 pipelining has not been advertised.
2436 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2438 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2439 returned in the SMTP response when smtp_return_error_details was false.
2440 This has been fixed.
2442 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2443 instead of the C I/O that was originally supplied, because problems were
2444 reported on Solaris.
2446 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2447 Exim which did not show up earlier: it was assuming that a call to
2448 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2449 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2450 was a macro that became an assignment, so it seemed to work. This has
2451 changed to a proper function call with a void return, hence the compile
2452 error. Exim's code has been fixed.
2454 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2457 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2459 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2461 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2462 because it turns out to be too big or there is a timeout) there is no
2463 "Completed" line in the log. When some messages of this type were
2464 selected by exigrep, they were listed as "not completed". Others were
2465 picked up by some special patterns. I have improved the selection
2466 criteria to be more general.
2468 PH/38 The host_find_failed option in the manualroute router can now be set
2469 to "ignore", to completely ignore a host whose IP address cannot be
2470 found. If all hosts are ignored, the behaviour is controlled by the new
2471 host_all_ignored option.
2473 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2474 homing or because of multiple MX records with /mx) generated more than
2475 one IP address, and the following item turned out to be the local host,
2476 all the secondary addresses of the first item were incorrectly removed
2477 from the list, along with the local host and any following hosts (which
2478 is what is supposed to happen).
2480 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2481 whoever called Exim into the -H file. In the case of the daemon it was
2482 behaving confusingly. When first started, it used values for whoever
2483 started the daemon, but after a SIGHUP it used the Exim user (because it
2484 calls itself on a restart). I have changed the code so that it now always
2487 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2488 message are rejected with the same error (e.g. no authentication or bad
2489 sender address), and a DATA command is nevertheless sent (as can happen
2490 with PIPELINING or a stupid MUA), the error message that was given to the
2491 RCPT commands is included in the rejection of the DATA command. This is
2492 intended to be helpful for MUAs that show only the final error to their
2495 PH/42 Another patch from the Sieve maintainer.
2497 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2498 Eximstats - Fixed some broken HTML links and added missing column headers
2500 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2501 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2503 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2505 PH/43 Yet another patch from the Sieve maintainer.
2507 PH/44 I found a way to check for a TCP/IP connection going away before sending
2508 the response to the final '.' that terminates a message, but only in the
2509 case where the client has not sent further data following the '.'
2510 (unfortunately, this is allowed). However, in many cases there won't be
2511 any further data because there won't be any more messages to send. A call
2512 to select() can be used: if it shows that the input is "ready", there is
2513 either input waiting, or the socket has been closed. An attempt to read
2514 the next input character can distinguish the two cases. Previously, Exim
2515 would have sent an OK response which the client would never have see.
2516 This could lead to message repetition. This fix should cure that, at
2517 least in a lot of common cases.
2519 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2520 advertised in response to EHLO.
2526 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2527 fixed by 4.65/MH/01 (is this a record?) are fixed:
2529 (i) An empty string was always treated as zero by the numeric comparison
2530 operators. This behaviour has been restored.
2532 (ii) It is documented that the numeric comparison operators always treat
2533 their arguments as decimal numbers. This was broken in that numbers
2534 starting with 0 were being interpreted as octal.
2536 While fixing these problems I realized that there was another issue that
2537 hadn't been noticed. Values of message_size_limit (both the global option
2538 and the transport option) were treated as octal if they started with 0.
2539 The documentation was vague. These values are now always treated as
2540 decimal, and I will make that clear in the documentation.
2546 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2547 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2550 MH/01 Don't check that the operands of numeric comparison operators are
2551 integers when their expansion is in "skipping" mode (fixes bug
2552 introduced by 4.64-PH/07).
2554 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2555 child addresses, Exim now panics and dies. Previously, because the count
2556 is held in a short int, deliveries were likely to be lost. As such a
2557 large number of recipients for a single message is ridiculous
2558 (performance will be very, very poor), I have chosen to impose a limit
2559 rather than extend the field.
2565 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2566 leftover -K file (the existence of which was triggered by #402).
2567 While we were at it, introduced process PID as part of the -K
2568 filename. This should rule out race conditions when creating
2571 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2572 processing considerably. Previous code took too long for large mails,
2573 triggering a timeout which in turn triggers #401.
2575 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2576 in the DK code in transports.c. sendfile() is not really portable,
2577 hence the _LINUX specificness.
2579 TF/01 In the add_headers option to the mail command in an Exim filter,
2580 there was a bug that Exim would claim a syntax error in any
2581 header after the first one which had an odd number of characters
2584 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2585 callout verification, Exim cached a "reject" for the entire domain. This
2586 is correct for most verifications, but it is not correct for a recipient
2587 verification with use_sender or use_postmaster set, because in that case
2588 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2589 case of MAIL FROM:<> rejection from other early rejections (e.g.
2590 rejection of HELO). When verifying a recipient using a non-null MAIL
2591 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2592 Whatever the result of the callout, the value of the domain cache is
2593 left unchanged (for any other kind of callout, getting as far as trying
2594 RCPT means that the domain itself is ok).
2596 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2599 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2600 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2601 session, the daemon ignores ECONNECT errors and logs others; it now
2602 ignores EPIPE as well.
2604 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2605 (quoted-printable decoding).
2607 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2608 later the small subsequent patch to fix an introduced bug.
2610 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2612 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2614 PH/08 An error is now given if message_size_limit is specified negative.
2616 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2617 to be given (somewhat) arbitrary names.
2619 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2622 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2623 miscellaneous code fixes
2625 PH/10 Added the log_reject_target ACL modifier to specify where to log
2628 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2629 hostname. This is wrong, because it relates to the incoming message (and
2630 probably the interface on which it is arriving) and not to the outgoing
2631 callout (which could be using a different interface). This has been
2632 changed to use the value of the helo_data option from the smtp transport
2633 instead - this is what is used when a message is actually being sent. If
2634 there is no remote transport (possible with a router that sets up host
2635 addresses), $smtp_active_hostname is used.
2637 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2638 tweaks were necessary in order to get it to work (see also 21 below):
2639 (a) The code assumed that strncpy() returns a negative number on buffer
2640 overflow, which isn't the case. Replaced with Exim's string_format()
2642 (b) There were several signed/unsigned issues. I just did the minimum
2643 hacking in of casts. There is scope for a larger refactoring.
2644 (c) The code used strcasecmp() which is not a standard C function.
2645 Replaced with Exim's strcmpic() function.
2646 (d) The code set only $1; it now sets $auth1 as well.
2647 (e) A simple test gave the error "authentication client didn't specify
2648 service in request". It would seem that Dovecot has changed its
2649 interface. Fortunately there's a specification; I followed it and
2650 changed what the client sends and it appears to be working now.
2652 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2655 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2656 address A is aliased to B and C, where B exists and C does not. Without
2657 -v the output is "A verified" because verification stops after a
2658 successful redirection if more than one address is generated. However,
2659 with -v the child addresses are also verified. Exim was outputting "A
2660 failed to verify" and then showing the successful verification for C,
2661 with its parentage. It now outputs "B failed to verify", showing B's
2662 parentage before showing the successful verification of C.
2664 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2665 look up a TXT record in a specific list after matching in a combined
2668 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2669 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2670 they consult the DNS. I had assumed they would set it the way they
2671 wanted; and indeed my experiments on Linux seem to show that in some
2672 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2673 To be on the safe side, however, I have now made the interface to
2674 host_find_byname() similar to host_find_bydns(), with an argument
2675 containing the DNS resolver options. The host_find_byname() function now
2676 sets these options at its start, just as host_find_bydns() does. The smtp
2677 transport options dns_qualify_single and dns_search_parents are passed to
2678 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2679 of host_find_byname() use the default settings of RES_DEFNAMES
2680 (qualify_single) but not RES_DNSRCH (search_parents).
2682 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2683 spool_read_header() do less string testing, by means of a preliminary
2684 switch on the second character of optional "-foo" lines. (This is
2685 overdue, caused by the large number of possibilities that now exist.
2686 Originally there were few.) While I was there, I also converted the
2687 str(n)cmp tests so they don't re-test the leading "-" and the first
2688 character, in the hope this might squeeze out yet more improvement.
2690 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2691 flag allowing group syntax was set by the header_syntax check but not
2692 turned off, possible causing trouble later; (2) The flag was not being
2693 set at all for the header_verify test, causing "group"-style headers to
2694 be rejected. I have now set it in this case, and also caused header_
2695 verify to ignore an empty address taken from a group. While doing this, I
2696 came across some other cases where the code for allowing group syntax
2697 while scanning a header line wasn't quite right (mostly, not resetting
2698 the flag correctly in the right place). These bugs could have caused
2699 trouble for malformed header lines. I hope it is now all correct.
2701 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2702 with the "reply" argument non-NULL. The code, however (which originally
2703 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2704 but it didn't always do it. This confused somebody who was copying the
2705 code for some other use. I have removed all the tests.
2707 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2708 feature that was used to support insecure browsers during the U.S. crypto
2709 embargo. It requires special client support, and Exim is probably the
2710 only MTA that supported it -- and would never use it because real RSA is
2711 always available. This code has been removed, because it had the bad
2712 effect of slowing Exim down by computing (never used) parameters for the
2713 RSA_EXPORT functionality.
2715 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2716 authenticator to fail if there's a tab character in the incoming data
2717 (there should never be unless someone is messing about, as it's supposed
2718 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2719 if the connection is using TLS or if the remote IP is the same as the
2720 local IP, and the "valid-client-cert option" if a client certificate has
2723 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2724 authenticators. This can be used for authorization after authentication
2725 succeeds. (In the case of plaintext, it servers for both authentication
2728 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2729 if any retry times were supplied.
2731 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2732 connection, where there is no client IP address to check. In this
2733 situation, the verify now always succeeds.
2735 PH/25 Applied John Jetmore's -Mset patch.
2737 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2739 PH/27 In a string expansion for a processed (not raw) header when multiple
2740 headers of the same name were present, leading whitespace was being
2741 removed from all of them, but trailing whitespace was being removed only
2742 from the last one. Now trailing whitespace is removed from each header
2743 before concatenation. Completely empty headers in a concatenation (as
2744 before) are ignored.
2746 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2747 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2749 PH/29 [Removed. This was a change that I later backed out, and forgot to
2750 correct the ChangeLog entry (that I had efficiently created) before
2751 committing the later change.]
2753 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2754 address errors (4xx response to RCPT) over the same connection as other
2755 messages routed to the same hosts. Such deliveries are always "forced",
2756 so retry times are not inspected. This resulted in far too many retries
2757 for the affected addresses. The effect occurred only when there were more
2758 hosts than the hosts_max_try setting in the smtp transport when it had
2759 the 4xx errors. Those hosts that it had tried were not added to the list
2760 of hosts for which the message was waiting, so if all were tried, there
2761 was no problem. Two fixes have been applied:
2763 (i) If there are any address or message errors in an SMTP delivery, none
2764 of the hosts (tried or untried) are now added to the list of hosts
2765 for which the message is waiting, so the message should not be a
2766 candidate for sending over the same connection that was used for a
2767 successful delivery of some other message. This seems entirely
2768 reasonable: after all the message is NOT "waiting for some host".
2769 This is so "obvious" that I'm not sure why it wasn't done
2770 previously. Hope I haven't missed anything, but it can't do any
2771 harm, as the worst effect is to miss an optimization.
2773 (ii) If, despite (i), such a delivery is accidentally attempted, the
2774 routing retry time is respected, so at least it doesn't keep
2775 hammering the server.
2777 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2778 in ${readsocket because some servers need this prod.
2780 PH/32 Added some extra debug output when updating a wait-xxx database.
2782 PH/33 The hint "could be header name not terminated by colon", which has been
2783 given for certain expansion errors for a long time, was not being given
2784 for the ${if def:h_colon_omitted{... case.
2786 PH/34 The spec says: "With one important exception, whenever a domain list is
2787 being scanned, $domain contains the subject domain." There was at least
2788 one case where this was not true.
2790 PH/35 The error "getsockname() failed: connection reset by peer" was being
2791 written to the panic log as well as the main log, but it isn't really
2792 panic-worthy as it just means the connection died rather early on. I have
2793 removed the panic log writing for the ECONNRESET error when getsockname()
2796 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2797 runs only) independently of the message's sender address. This meant
2798 that, if the 4xx error was in fact related to the sender, a different
2799 message to the same recipient with a different sender could confuse
2800 things. In particular, this can happen when sending to a greylisting
2801 server, but other circumstances could also provoke similar problems.
2802 I have changed the default so that the retry time for these errors is now
2803 based a combination of the sender and recipient addresses. This change
2804 can be overridden by setting address_retry_include_sender=false in the
2807 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2808 remote server are returned as part of bounce messages. This was not
2809 happening for LMTP over a pipe (the lmtp transport), but now it is the
2810 same for both kinds of LMTP.
2812 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2813 in header lines that were in fact CNAMEs. This is no longer the case.
2815 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2816 and queue runs started by the daemon processed all messages. This has
2817 been fixed so that -R and -S can now usefully be given with -q<time>.
2819 PH/40 Import PCRE release 6.7 (fixes some bugs).
2821 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2823 PH/42 Give an error if -q is specified more than once.
2825 PH/43 Renamed the variables $interface_address and $interface_port as
2826 $received_ip_address and $received_port, to make it clear that these
2827 values apply to message reception, and not to the outgoing interface when
2828 a message is delivered. (The old names remain recognized, of course.)
2830 PH/44 There was no timeout on the connect() call when using a Unix domain
2831 socket in the ${readsocket expansion. There now is.
2833 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2834 be meaningful with "accept".
2836 SC/01 Eximstats V1.43
2837 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2839 SC/02 Eximstats V1.44
2840 Use a glob alias rather than an array ref in the generated
2841 parser. This improves both readability and performance.
2843 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2844 Collect SpamAssassin and rejection statistics.
2845 Don't display local sender or destination tables unless
2846 there is data to show.
2847 Added average volumes into the top table text output.
2849 SC/04 Eximstats V1.46
2850 Collect data on the number of addresses (recipients)
2851 as well as the number of messages.
2853 SC/05 Eximstats V1.47
2854 Added 'Message too big' to the list of mail rejection
2855 reasons (thanks to Marco Gaiarin).
2857 SC/06 Eximstats V1.48
2858 Mainlog lines which have GMT offsets and are too short to
2859 have a flag are now skipped.
2861 SC/07 Eximstats V1.49 (Alain Williams)
2862 Added the -emptyok flag.
2864 SC/08 Eximstats V1.50
2865 Fixes for obtaining the IP address from reject messages.
2867 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2868 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2869 whitespace changes from 4.64-PH/27
2871 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2874 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2875 are found, allow negative numbers in numeric criteria)
2877 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2879 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2880 to match changes made in 4.64-PH/43
2882 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2884 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2885 "domain" is actually an IP address, and force a failure. This locks out
2886 those revolvers/nameservers that support "A-for-A" lookups, in
2887 contravention of the specifications.
2889 PH/48 When a host name was looked up from an IP address, and the subsequent
2890 forward lookup of the name timed out, the host name was left in
2891 $sender_host_name, contrary to the specification.
2893 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2894 restricted to single-key lookups, Exim was not diagnosing an error if
2895 * or *@ was used with a query-style lookup.
2897 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2899 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2900 long ago, but noone interested enough thought of it. Let's just say that
2901 the "1.1" means that there are some new functions that weren't there at
2902 some point in the past.
2904 PH/51 Error processing for expansion failure of helo_data from an smtp
2905 transport during callout processing was broken.
2907 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2908 tested/used via the -bh/-bhc/-bs options.
2910 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2911 bug, fixed in subsequent PCRE releases).
2913 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2914 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2916 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2922 SC/01 Use a glob alias rather than an array ref in eximstats generated
2923 parser. This improves both readability and performance.
2925 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2926 Don't display local sender or destination tables in eximstats unless
2927 there is data to show.
2928 Added average volumes into the eximstats top table text output.
2930 SC/03 Collect data on the number of addresses (recipients) as well
2931 as the number of messages in eximstats.
2933 TF/01 Correct an error in the documentation for the redirect router. Exim
2934 does (usually) call initgroups() when daemonizing.
2936 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2937 with consistent privilege compared to when running as a daemon.
2939 TF/03 Note in the spec that $authenticated_id is not set for local
2940 submissions from trusted users.
2942 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2943 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2945 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2946 by adding some example configuration directives to the default
2947 configuration file. A little bit of work is required to uncomment the
2948 directives and define how usernames and passwords are checked, but
2949 there is now a framework to start from.
2951 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2952 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2953 without this. I don't know how relevant this is to other LDAP libraries.
2955 PH/02 Add the verb name to the "unknown ACL verb" error.
2957 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2959 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2961 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2962 directory not expanded when it should be if an expanded home directory
2963 was set for the address (which is overridden by the transport).
2965 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2968 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2969 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2970 because it is too late at that time, and has no effect.
2972 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2973 security issue with \' (bugzilla #107). I could not use the
2974 PQescapeStringConn() function, because it needs a PGconn value as one of
2977 PH/08 When testing addresses using -bt, indicate those final addresses that
2978 are duplicates that would not cause an additional delivery. At least one
2979 person was confused, thinking that -bt output corresponded to deliveries.
2980 (Suppressing duplicates isn't a good idea as you lose the information
2981 about possibly different redirections that led to the duplicates.)
2983 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2984 systems where poll() doesn't work, in particular OS X.
2986 PH/10 Added more information to debugging output for retry time not reached.
2988 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2989 operations in malware.c.
2991 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2994 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2995 syslog_duplication set false, log lines that would normally be written
2996 both the the main log and to the reject log were not written to syslog at
2999 PH/14 In the default configuration, change the use of "message" in ACL warn
3000 statements to "add_header".
3002 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3003 not followed by a command (e.g. "seen endif").
3005 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3006 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3009 PH/17 Added extra conditions to the default value of delay_warning_condition
3013 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3014 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3015 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3018 The Auto-Submitted: and various List- headers are standardised, whereas I
3019 don't think Precedence: ever was.
3021 PH/18 Refactored debugging code in route_finduser() to show more information,
3022 in particular, the error code if getpwnam() issues one.
3024 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3025 This is apparently needed in addition to the PH/07 change above to avoid
3026 any possible encoding problems.
3028 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3029 but not after initializing Perl.
3031 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3032 output them only if debugging. By default they are written stderr,
3033 apparently, which is not desirable.
3035 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3038 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3041 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3043 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3044 authenticated or an ident call has been made. Suppress the default
3045 values for $authenticated_id and $authenticated_sender (but permit -oMai
3046 and -oMas) when testing with -bh.
3048 PH/24 Re-jigged the order of the tests in the default configuration so that the
3049 tests for valid domains and recipients precede the DNS black list and CSA
3050 tests, on the grounds that those ones are more expensive.
3052 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3053 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3054 This bug exists in every version of Exim that I still have, right back to
3057 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3058 However, an attempt to turn on case-sensitivity in a regex key by
3059 including (?-i) didn't work because the subject string was already
3060 lowercased, and the effects were non-intuitive. It turns out that a
3061 one-line patch can be used to allow (?-i) to work as expected.
3067 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3068 other effects) broke the use of negated acl sub-conditions.
3070 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3073 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3074 "Deny" causes Exim to reject the incoming connection with a 554 error.
3075 Unfortunately, if there is a major crisis, such as a disk failure,
3076 tcp-wrappers gives "deny", whereas what one would like would be some
3077 kind of temporary error. A kludge has been added to help with this.
3078 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3079 554 error is used if errno is still zero or contains ENOENT (which occurs
3080 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3083 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3085 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3086 errors. Otherwise a message that provokes a temporary error (when other
3087 messages do not) can cause a whole host to time out.
3089 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3090 addresses were routed directly to files or pipes from a redirect router.
3091 File deliveries just didn't batch; pipe deliveries might have suffered
3094 PH/06 A failure to get a lock for a hints database would erroneously always say
3095 "Failed to get write lock", even when it was really a read lock.
3097 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3098 of 0600. This has been changed to use the value of the lockfile_mode
3099 option (which defaults to 0600).
3101 PH/08 Applied small patch from the Sieve maintainer.
3103 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3104 folder from quota calculations, a direct delivery into this folder messed
3105 up the contents of the maildirsize file. This was because the regex was
3106 used only to exclude .Trash (or whatever) when the size of the mailbox
3107 was calculated. There was no check that a delivery was happening into an
3108 excluded directory. This bug has been fixed by ignoring all quota
3109 processing for deliveries into excluded directories.
3111 PH/10 Added the maildirfolder_create_regex option to appendfile.
3117 PH/01 The code for finding all the local interface addresses on a FreeBSD
3118 system running IPv6 was broken. This may well have applied to all BSD
3119 systems, as well as to others that have similar system calls. The broken
3120 code found IPv4 interfaces correctly, but gave incorrect values for the
3121 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3122 that it would not match correctly against @[] and not recognize the IPv6
3125 PH/02 The ipliteral router was not recognizing addresses of the form user@
3126 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3128 PH/03 Added disable_ipv6.
3130 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3131 decoded form, because it is most often used to construct To: headers
3132 lines in autoreplies, and the decoded form may well be syntactically
3133 invalid. However, $reply_address has leading white space removed, and all
3134 newlines turned into spaces so that the autoreply transport does not
3137 PH/05 If group was specified without a user on a router, and no group or user
3138 was specified on a transport, the group from the router was ignored.
3140 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3141 for visible compile-time settings that can be used to change these
3142 numbers, for those that want even more. Backwards compatibility with old
3143 spool files has been maintained. However, going back to a previous Exim
3144 release will lost any variables that are in spool files.
3146 PH/07 Two small changes when running in the test harness: increase delay when
3147 passing a TCP/IP connection to a new process, in case the original
3148 process has to generate a bounce, and remove special handling of
3149 127.0.0.2 (sic), which is no longer necessary.
3151 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3152 be the same on different OS.
3154 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3157 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3158 whether --show-vars was specified or not
3160 JJ/02 exipick: Added support for new ACL variable spool format introduced
3163 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3164 syntactically invalid From: or Reply-to: line, and a filter used this to
3165 generate an autoreply, and therefore failed to obtain an address for the
3166 autoreply, Exim could try to deliver to a non-existent relative file
3167 name, causing unrelated and misleading errors. What now happens is that
3168 it logs this as a hard delivery error, but does not attempt to create a
3171 PH/11 The exinext utility has a -C option for testing purposes, but although
3172 the given file was scanned by exinext itself; it wasn't being passed on
3173 when Exim was called.
3175 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3176 an end-of-file indication when reading a command response.
3178 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3179 compiled. In many other places in Exim, IPv6 addresses are always
3180 recognized, so I have changed this. It also means that IPv4 domain
3181 literals of the form [IPV4:n.n.n.n] are now always recognized.
3183 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3184 used if the router is not running as root, for example, when verifying at
3185 ACL time, or when using -bh. The debugging output from this situation was
3186 non-existent - all you got was a failure to exec. I have made two
3189 (a) Failures to set uid/gid, the current directory, or a process leader
3190 in a subprocess such as that created by queryprogram now generate
3191 suitable debugging output when -d is set.
3193 (b) The queryprogram router detects when it is not running as root,
3194 outputs suitable debugging information if -d is set, and then runs
3195 the subprocess without attempting to change uid/gid.
3197 PH/15 Minor change to Makefile for building test_host (undocumented testing
3200 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3201 additional section of a DNS packet that returns MX or SRV records.
3202 Instead, it always explicitly searches for A/AAAA records. This avoids
3203 major problems that occur when a DNS server includes only records of one
3204 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3205 fixed another bug: if SRV records were looked up and the corresponding
3206 address records were *not* found in the additional section, the port
3207 values from the SRV records were lost.
3209 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3210 using the correct key (the original address) when searching the retry
3211 rules in order to find which one to use for generating the retry hint.
3213 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3214 adding the default one. Similarly, if it contains a Reply-To: header, the
3215 errors_reply_to option, if set, is not used.
3217 PH/19 When calculating a retry time, Exim used to measure the "time since
3218 failure" by looking at the "first failed" field in the retry record. Now
3219 it does not use this if it is later than than the arrival time of the
3220 message. Instead it uses the arrival time. This makes for better
3221 behaviour in cases where some deliveries succeed, thus re-setting the
3222 "first failed" field. An example is a quota failure for a huge message
3223 when small messages continue to be delivered. Without this change, the
3224 "time since failure" will always be short, possible causing more frequent
3225 delivery attempts for the huge message than are intended.
3226 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3228 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3229 $1, $2, $3) because the numerical variables can be reset during some
3230 expansion items (e.g. "match"), thereby losing the authentication data.
3232 PH/21 Make -bV show the size of off_t variables so that the test suite can
3233 decide whether to run tests for quotas > 2G.
3235 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3236 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3237 filecount value is greater than 2G or if a quota value is greater than 2G
3238 on a system where the size of off_t is not greater than 4, a panic error
3241 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3242 never match. The debug and -bh output now contains an explicit error
3243 message indicating a malformed IPv4 address or mask.
3245 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3246 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3247 PH/23 above applies.
3249 PH/25 Do not write to syslog when running in the test harness. The only
3250 occasion when this arises is a failure to open the main or panic logs
3251 (for which there is an explicit test).
3253 PH/26 Added the /no_tell option to "control=freeze".
3255 PH/27 If a host name lookup failed very early in a connection, for example, if
3256 the IP address matched host_lookup and the reverse lookup yielded a name
3257 that did not have a forward lookup, an error message of the form "no IP
3258 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3259 could be logged. Now it outputs the IP address instead of "NULL".
3261 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3262 allows the sender and the authenticated sender to be set when
3263 submitting a message from within Exim. Since child_open_exim() is
3264 documented for local_scan(), the new function should be too.
3266 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3267 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3268 results in an empty string is now treated as unset.
3270 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3272 PH/31 Added sender_verify_fail logging option.
3274 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3275 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3276 tidied the source and removed it altogether.
3278 PH/33 When a queue run was abandoned because the load average was too high, a
3279 log line was always written; now it is written only if the queue_run log
3280 selector is set. In addition, the log line for abandonment now contains
3281 information about the queue run such as the pid. This is always present
3282 in "start" and "stop" lines but was omitted from the "abandon" line.
3284 PH/34 Omit spaces between a header name and the colon in the error message that
3285 is given when verify = headers_syntax fails (if there are lots of them,
3286 the message gets confusing).
3288 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3289 names, as there are now some PTR records that contain slashes. This check
3290 is only to protect against broken name servers that fall over on strange
3291 characters, so the fact that it applies to all lookups doesn't matter.
3293 PH/36 Now that the new test suite is complete, we can remove some of the
3294 special code in Exim that was needed for the old test suite. For example,
3295 sorting DNS records because real resolvers return them in an arbitrary
3296 order. The new test suite's fake resolver always returns records in the
3299 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3300 bounces) except when queue_only is set, to avoid logging races between
3301 the different processes.
3303 PH/38 Panic-die if .include specifies a non-absolute path.
3305 PH/39 A tweak to the "H" retry rule from its user.
3307 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3308 a label. They prevented compilation on older perls.
3310 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3311 a warning to be raised on newish perls.
3313 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3314 on queue. Changes to match documented behaviour of showing count of
3315 messages matching specified criteria.
3317 PH/40 Changed the default ident timeout from 30s to 5s.
3319 PH/41 Added support for the use of login_cap features, on those BSD systems
3320 that have them, for controlling the resources used by pipe deliveries.
3322 PH/42 The content-scanning code uses fopen() to create files in which to put
3323 message data. Previously it was not paying any attention to the mode of
3324 the files. Exim runs with umask(0) because the rest of the code creates
3325 files with open(), and sets the required mode explicitly. Thus, these
3326 files were ending up world-writeable. This was not a big issue, because,
3327 being within the spool directory, they were not world-accessible. I have
3328 created a function called modefopen, which takes an additional mode
3329 argument. It sets umask(777), creates the file, chmods it to the required
3330 mode, then resets the umask. All the relevant calls to fopen() in the
3331 content scanning code have been changed to use this function.
3333 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3334 to 24 hours. This avoids potential overflow problems when processing G
3335 and H retry rules. I suspect nobody ever tinkers with this value.
3337 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3339 PH/45 When the plaintext authenticator is running as a client, the server's
3340 challenges are checked to ensure they are valid base64 strings. By
3341 default, the authentication attempt is cancelled if an invalid string is
3342 received. Setting client_ignore_invalid_base64 true ignores these errors.
3343 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3344 they are received. Thus, the responses can be made to depend on the
3345 challenges. If an invalid string is ignored, an empty string is placed in
3348 PH/46 Messages that are created by the autoreply transport now contains a
3349 References: header, in accordance with RFCs 2822 and 3834.
3351 PH/47 Added authenticated_sender_force to the smtp transport.
3353 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3355 PH/49 Installed latest patch from the Sieve maintainer.
3357 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3358 was also set, the appendfile transport was unnecessarily scanning a
3359 directory of message files (e.g. for maildir delivery) to find the count
3360 of files (along with the size), even though it did not need this
3361 information. It now does the scan only if it needs to find either the
3362 size of the count of files.
3364 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3366 PH/52 Two bugs concerned with error handling when the smtp transport is
3369 (i) Exim was not creating retry information for temporary errors given
3370 for individual recipients after the DATA command when the smtp transport
3371 was used in LMTP mode. This meant that they could be retried too
3372 frequently, and not timed out correctly.
3374 (ii) Exim was setting the flag that allows error details to be returned
3375 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3376 recipients that were returned after the DATA command.
3378 PH/53 This is related to PH/52, but is more general: for any failing address,
3379 when detailed error information was permitted to be returned to the
3380 sender, but the error was temporary, then after the final timeout, only
3381 "retry timeout exceeded" was returned. Now it returns the full error as
3382 well as "retry timeout exceeded".
3384 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3385 do this, and (what is worse) MTAs that accept it.
3387 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3388 will now be deprecated.
3390 PH/56 New os.c-cygwin from the Cygwin maintainer.
3392 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3393 formats (previously only available in exim formats via -bpr, -bpru,
3394 and -bpra. Now also available in native and exiqgrep formats)
3396 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3397 with very large, slow to parse queues
3399 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3401 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3403 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3404 responses to authentication challenges, though it was showing the
3405 challenges; (ii) I've removed the CR characters from the debug output for
3408 PH/58 Allow for the insertion of a newline as well as a space when a string
3409 is turned into more than one encoded-word during RFC 2047 encoding. The
3410 Sieve code now uses this.
3412 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3413 data_4xx, lost_connection, tls_required.
3415 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3416 message was being sent as an SMTP response.
3418 PH/61 Add -l and -k options to exicyclog.
3420 PH/62 When verifying, if an address was redirected to one new address, so that
3421 verification continued, and the new address failed or deferred after
3422 having set something in $address_data, the value of $address_data was not
3423 passed back to the ACL. This was different to the case when no
3424 redirection occurred. The value is now passed back in both cases.
3426 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3427 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3428 use login_cap.h, so on its own it isn't the distinguishing feature. The
3429 new name refers directly to the setclassresources() function.
3431 PH/65 Added configuration files for NetBSD3.
3433 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3435 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3436 is preferred over IPv4.
3438 PH/68 The bounce_return_message and bounce_return_body options were not being
3439 honoured for bounces generated during the reception of non-SMTP messages.
3440 In particular, this applied to messages rejected by the ACL. This bug has
3441 been fixed. However, if bounce_return_message is true and bounce_return_
3442 body is false, the headers that are returned for a non-SMTP message
3443 include only those that have been read before the error was detected.
3444 (In the case of an ACL rejection, they have all been read.)
3446 PH/69 The HTML version of the specification is now built in a directory called
3447 spec_html instead of spec.html, because the latter looks like a path with
3448 a MIME-type, and this confuses some software.
3450 PH/70 Catch two compiler warnings in sieve.c.
3452 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3453 function verify_get_ident() calls ip_connect() to connect a socket, but
3454 if the "connect()" function timed out, ip_connect() used to close the
3455 socket. However, verify_get_ident() also closes the socket later, and in
3456 between Exim writes to the log, which may get opened at this point. When
3457 the socket was closed in ip_connect(), the log could get the same file
3458 descriptor number as the socket. This naturally causes chaos. The fix is
3459 not to close the socket in ip_connect(); the socket should be closed by
3460 the function that creates it. There was only one place in the code where
3461 this was missing, in the iplookup router, which I don't think anybody now
3462 uses, but I've fixed it anyway.
3464 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3465 well as to direct DNS lookups. Otherwise the handling of names in host
3466 lists is inconsistent and therefore confusing.
3472 PH/01 Two changes to the default runtime configuration:
3474 (1) Move the checks for relay_from_hosts and authenticated clients from
3475 after to before the (commented out) DNS black list checks.
3477 (2) Add control=submission to the relay_from_hosts and authenticated
3478 clients checks, on the grounds that messages accepted by these
3479 statements are most likely to be submissions.
3481 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3483 (1) Generate an error if the third argument for the ${prvs expansion is
3486 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3489 (3) Reset the variables that are obtained from the first argument of
3490 ${prvscheck and used in the second argument before leaving the code,
3491 because their memory is reclaimed, so using them afterwards may do
3494 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3495 one (it's much easier than Tom thought :-).
3497 (5) Because of (4), we can now allow for the use of $prvscheck_result
3498 inside the third argument.
3500 PH/03 For some reason, the default setting of PATH when running a command from
3501 a pipe transport was just "/usr/bin". I have changed it to
3504 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3505 anything to be listed in the output from -bV.
3507 PH/05 When a filter generated an autoreply, the entire To: header line was
3508 quoted in the delivery log line, like this:
3510 => >A.N.Other <ano@some.domain> <original@ddress> ...
3512 This has been changed so that it extracts the operative address. There
3513 may be more than one such address. If so, they are comma-separated, like
3516 => >ano@some.domain,ona@other.domain <original@ddress> ...
3518 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3519 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3520 not being looked up in the rDNS to get a host name, Exim was showing the
3521 IP address twice in Received: lines, even though the IP addresses were
3522 identical. For example:
3524 Received: from [1.2.3.4] (helo=[1.2.3.4])
3526 However, if the real host name was known, it was omitting the HELO data
3527 if it matched the actual IP address. This has been tidied up so that it
3528 doesn't show the same IP address twice.
3530 PH/07 When both +timestamp and +memory debugging was on, the value given by
3531 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3532 called by the debug printing, thereby overwriting the timestamp buffer.
3533 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3535 PH/08 When the original message was included in an autoreply transport, it
3536 always said "this is a copy of the message, including all the headers",
3537 even if body_only or headers_only was set. It now gives an appropriate
3540 PH/09 Applied a patch from the Sieve maintainer which:
3542 o fixes some comments
3543 o adds the (disabled) notify extension core
3544 o adds some debug output for the result of if/elsif tests
3545 o points to the current vacation draft in the documentation
3546 and documents the missing references header update
3550 o fixes a bug in processing the envelope test (when testing
3551 multiple envelope elements, the last element determined the
3554 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3555 Electronic Mail") by including:
3557 Auto-submitted: auto-generated
3559 in the messages that it generates (bounce messages and others, such as
3560 warnings). In the case of bounce messages for non-SMTP messages, there was
3561 also a typo: it was using "Auto_submitted" (underscore instead of
3562 hyphen). Since every message generated by Exim is necessarily in response
3563 to another message, thes have all been changed to:
3565 Auto-Submitted: auto-replied
3567 in accordance with these statements in the RFC:
3569 The auto-replied keyword:
3571 - SHOULD be used on messages sent in direct response to another
3572 message by an automatic process,
3574 - MUST NOT be used on manually-generated messages,
3576 - MAY be used on Delivery Status Notifications (DSNs) and Message
3577 Disposition Notifications (MDNs),
3579 - MUST NOT be used on messages generated by automatic or periodic
3580 processes, except for messages which are automatic responses to
3583 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3584 to the default Received: header definition.
3586 PH/12 Added log selector acl_warn_skipped (default on).
3588 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3589 variables because (a) they are in the wrong storage pool and (b) even if
3590 they were copied, it wouldn't work properly because of the caching.
3592 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3593 checking when decoding. Apparently there are clients that generate
3594 overlong encoded strings. Why am I not surprised?
3596 PH/15 If the first argument of "${if match_address" was not empty, but did not
3597 contain an "@" character, Exim crashed. Now it writes a panic log message
3598 and treats the condition as false.
3600 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3602 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3603 extension "envelope-auth". The code is finished and in agreement with
3604 other implementations, but there is no documentation so far and in fact,
3605 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3606 not changing the active code.
3608 Print executed "if" and "elsif" statements when debugging is used. This
3609 helps a great deal to understand what a filter does.
3611 Document more things not specified clearly in RFC3028. I had all this
3612 sorted out, when out of a sudden new issues came to my mind. Oops."
3614 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3617 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3618 canonical form (as documented). However, after a host name lookup from
3619 the IP address, check_host() was doing a simple string comparison with
3620 addresses acquired from the DNS when checking that the found name did
3621 have the original IP as one of its addresses. Since any found IPv6
3622 addresses are likely to be in abbreviated form, the comparison could
3623 fail. Luckily, there already exists a function for doing the comparison
3624 by converting both addresses to binary, so now that is used instead of
3625 the text comparison.
3627 PH/20 There was another similar case to PH/19, when a complete host name was
3628 given in a host list; looking up its IP address could give an abbreviated
3629 form, whereas the current host's name might or might not be abbreviated.
3630 The same fix has been applied.
3636 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3637 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3640 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3641 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3643 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3645 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3646 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3647 Also, for "Auto-Submitted", treat anything other than "no" as
3648 non-personal, in accordance with RFC 3834. (Previously it treated
3649 anything starting "auto-" as non-personal.)
3651 TF/01 The control=submission/name=... option had a problem with syntax
3652 errors if the name included a slash character. The /name= option
3653 now slurps the rest of the string, so it can include any characters
3654 but it must come last in the list of options (after /sender_retain
3657 PH/05 Some modifications to the interface to the fake nameserver for the new
3665 TK/01 Added the "success_on_redirect" address verification option. See
3666 NewStuff for rationale and an example.
3668 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3670 PH/02 Patch to exigrep to allow it to work on syslog lines.
3672 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3673 fread() to skip over the body file's header line, because in Cygwin the
3674 header line is locked and is inaccessible.
3676 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3677 co-exist for some time) to make it clear that it is the Exim ID that is
3678 referenced, not the Message-ID: header line.
3680 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3681 string_format() function, because snprintf() does not exist on all
3684 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3685 ${stat: expansion item.
3687 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3688 protocol synchronization error", to keep the pedants happy.
3690 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3691 well as for IRIX systems, when gcc is being used. See the host.c source
3694 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3696 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3699 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3700 transport and to the smtp transport in LMTP mode.
3702 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3704 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3705 run a filter in a subprocess. This could lead to confusion in subsequent
3706 lookups in the parent process. There should also be a search_tidyup() at
3707 the end of the subprocess.
3709 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3710 only if the host matched helo_try_verify_hosts, which caused the
3711 verification to occur when the EHLO/HELO command was issued. The ACL just
3712 tested the remembered result. Now, if a previous verification attempt has
3713 not happened, "verify = helo" does it there and then.
3715 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3717 TK/03 Fix log output including CR from clamd.
3719 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3720 exist provoked a memory error which could cause a segfault.
3722 PH/15 Installed PCRE 6.2
3724 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3726 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3727 of the problem. Specifically, suggested +O2 rather than +O1 for the
3730 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3732 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3733 record, the port was not correctly logged when the outgoing_port log
3734 selector was set (it logged the transort's default port).
3736 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3737 fallback_hosts, and "hosts" in the smtp transport.
3739 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3740 host errors such as "Connection refused".
3742 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3743 authentication with radiusclient 0.4.9:
3745 - Error returned from rc_read_config was caught wrongly
3746 - Username/password not passed on to radius server due to wrong length.
3748 The presumption is that some radiusclient API changes for 4.51/PH/17
3749 were not taken care of correctly. The code is still untested by me (my
3750 Linux distribution still has 0.3.2 of radiusclient), but it was
3751 contributed by a Radius user.
3753 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3754 expanding the "port" option of the smtp transport.
3756 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3757 while reading a MIME header. Thanks to Tom Hughes for a patch.
3759 PH/24 Include config.h inside local_scan.h so that configuration settings are
3762 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3763 that in an ACL for RCPT (for example), you can examine exactly what was
3766 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3767 commands, but it was not correctly comparing the address with the actual
3768 client host address. Thus, it would show the EHLO address in Received:
3769 header lines when this was not necessary.
3771 PH/27 Added the % operator to ${eval:}.
3773 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3774 it should be ignoring failures (because with -C, for example, it has lost
3775 privilege). It wasn't ignoring creation failures other than "already
3778 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3781 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3782 because an input line was too long, either on its own, or by virtue of
3783 too many continuations, the temporary file was not being removed, and the
3784 return code was incorrect.
3786 PH/31 Missing "BOOL" in function definition in filtertest.c.
3788 PH/32 Applied Sieve patches from the maintainer.
3790 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3792 PH/33 Added "verify = not_blind".
3794 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3795 Local/Makefile (with some defaults set). These are used in built scripts
3796 such as exicyclog, but they have never been used in the exim_install
3797 script (though there are many overriding facilities there). I have
3798 arranged that the exim_install script now takes note of these two
3801 PH/35 Installed configuration files for Dragonfly.
3803 PH/36 When a locally submitted message by a trusted user did not contain a
3804 From: header, and the sender address was obtained from -f or from an SMTP
3805 MAIL command, and the trusted user did not use -F to supply a sender
3806 name, $originator_name was incorrectly used when constructing a From:
3807 header. Furthermore, $originator_name was used for submission mode
3808 messages from external hosts without From: headers in a similar way,
3809 which is clearly wrong.
3811 PH/37 Added control=suppress_local_fixups.
3813 PH/38 When log_selector = +received_sender was set, and the addition of the
3814 sender made the log line's construction buffer exactly full, or one byte
3815 less than full, an overflow happened when the terminating "\n" was
3818 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3819 when the result of a list match is failure because a DNS lookup failed.
3821 PH/40 RM_COMMAND is now used in the building process.
3823 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3824 the "build-* directories that it finds.
3826 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3827 address in a domain literal was a prefix of an interface address.
3829 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3830 when verifying a sender address, unless rewrite_headers is false.
3832 PH/44 Wrote a long comment about why errors_to addresses are verified as
3833 recipients, not senders.
3835 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3836 the ratelimit ACL was added.
3838 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3840 PH/46 Added extra information about PostgreSQL errors to the error string.
3842 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3843 avoiding the need to install special zones in a real server. This is
3844 backwards compatible; if it can't find the fake resolver, it drops back.
3845 Thus, both old and new test suites can be run.
3847 TF/02 Added util/ratelimit.pl
3849 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3850 clock is set back in time.
3852 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3853 Candler <B.Candler@pobox.com>.
3855 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3856 OK for addresses that are the result of redirections.
3858 PH/48 A number of further additions for the benefit of the new test suite,
3859 including a fake gethostbyname() that interfaces to the fake DNS resolver
3862 TF/06 The fix for widen_domains has also been applied to qualify_single and
3863 search_parents which are the other dnslookup options that can cause
3866 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3869 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3871 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3872 service" if it encountered an MX record with an empty target hostname.
3873 The message is now "an MX or SRV record indicated no SMTP service".
3875 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3876 if the DNS of the sending site is misconfigured. This is quite a
3877 common situation. This change restores the behaviour of treating a
3878 helo verification defer as a failure.
3880 PH/51 If self=fail was set on a router, the bounce message did not include the
3881 actual error message.
3887 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3889 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3890 command itself ended in error, the underlying message about the transport
3891 filter timeout was being overwritten with the pipe command error. Now the
3892 underlying error message should be appended to the second error message.
3894 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3896 PH/02 Reduce the amount of output that "make" produces by default. Full output
3897 can still be requested.
3899 PH/03 The warning log line about a condition test deferring for a "warn" verb
3900 was being output only once per connection, rather than after each
3901 occurrence (because it was using the same function as for successful
3902 "warn" verbs). This seems wrong, so I have changed it.
3904 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3905 it should not have, which might have caused a crash in the right
3906 circumstances, but probably never did.
3908 PH/04 Installed a modified version of Tony Finch's patch to make submission
3909 mode fix the return path as well as the Sender: header line, and to
3910 add a /name= option so that you can make the user's friendly name appear
3913 TF/03 Added the control = fakedefer ACL modifier.
3915 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3916 Mark Lowes for thorough testing.
3918 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3920 TK/03 Merged latest SRS patch from Miles Wilton.
3922 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3923 with the definition in sysexits.h (which is #included earlier).
3924 Fortunately, Exim does not actually use EX_OK. The code used to try to
3925 preserve the sysexits.h value, by assuming that macro definitions were
3926 scanned for macro replacements. I have been disabused of this notion,
3927 so now the code just undefines EX_OK before #including unistd.h.
3929 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3930 in the smtp transport. When a block could not be written in a single
3931 write() function, the timeout was being re-applied to each part-write.
3932 This seems wrong - if the receiver was accepting one byte at a time it
3933 would take for ever. The timeout is now adjusted when this happens. It
3934 doesn't have to be particularly precise.
3936 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3937 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3939 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3940 without a domain if the check to <postmaster@domain> fails.
3942 SC/01 Eximstats: added -xls and the ability to specify output files
3943 (patch written by Frank Heydlauf).
3945 SC/02 Eximstats: use FileHandles for outputting results.
3947 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3949 SC/04 Eximstats: fixed display of large numbers with -nvr option
3951 SC/05 Eximstats: fixed merging of reports with empty tables.
3953 SC/06 Eximstats: added the -include_original_destination flag
3955 SC/07 Eximstats: removed tabs and trailing whitespace.
3957 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3959 TK/06 MBOX spool code: Add real "From " MBOX separator line
3960 so the .eml file is really in mbox format (even though
3961 most programs do not really care). Patch from Alex Miller.
3963 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3964 The latter is generated from $received_to and is only set if the
3965 message has one envelope recipient. SA can use these headers,
3966 obviously out-of-the-box. Patch from Alex Miller.
3968 PH/08 The ${def test on a variable was returning false if the variable's
3969 value was "0", contrary to what the specification has always said!
3970 The result should be true unless the variable is empty.
3972 PH/09 The syntax error of a character other than { following "${if
3973 def:variable_name" (after optional whitespace) was not being diagnosed.
3974 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3975 accidental colon was present, for example, could give incorrect results.
3977 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3978 result is used (not including appendfile, where other changes are about
3981 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3982 This involved changing a lot of size variables from int to off_t. It
3983 should work with maildirs and everything.
3985 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3986 spamd dying while we are connected to it.
3988 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3991 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3992 function for BDB 4.3.
3994 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3996 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3997 the sources for PCRE have been re-arranged and more files are now
4000 PH/15 The code I had for printing potentially long long variables in PH/11
4001 above was not the best (it lost precision). The length of off_t variables
4002 is now inspected at build time, and an appropriate printing format (%ld
4003 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4004 to be "long long int" or "long int". This is needed for the internal
4005 formatting function string_vformat().
4007 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4008 the configuration file to be ":syslog", then the script "guesses" where
4009 the logs files are, rather than using the compiled in default. In our
4010 case the guess is not the same as the compiled default, so the script
4011 suddenly stopped working when I started to use syslog. The patch checks
4012 to see if log_file_path is "". If so, it attempts to read it from exim
4013 with no configuration file to get the compiled in version, before it
4014 falls back to the previous guessing code."
4016 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4017 implementing BATV in an Exim configuration. See NewStuff for the gory
4020 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4021 Makefile that are specific to HP-UX.
4023 PH/18 If the "use_postmaster" option was set for a recipient callout together
4024 with the "random" option, the postmaster address was used as the MAIL
4025 FROM address for the random test, but not for the subsequent recipient
4026 test. It is now used for both.
4028 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4029 patch removes a few documentation additions to RFC 3028, because the
4030 latest draft now contains them. It adds the new en;ascii-case comparator
4031 and a new error check for 8bit text in MIME parts. Comparator and
4032 require names are now matched exactly. I enabled the subaddress
4033 extension, but it is not well tested yet (read: it works for me)."
4035 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4036 rework some of the code of TK/09 above to avoid the hardwired use of
4037 "%lld" and "long long". Replaced the call to snprintf() with a call to
4040 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4041 records point to non-existent hosts", "retry timeout exceeded", and
4042 "retry time not reached for any host after a long failure period".
4044 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4045 experimental DomainKeys support:
4047 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4048 (2) On an error such as an illegally used "control", the wrong name for
4049 the control was given.
4051 These problems did NOT occur unless DomainKeys support was compiled.
4053 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4055 PH/24 Added ${if match_ip condition.
4057 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4058 databases so that it will be absolutely obvious if a crash occurs in the
4059 DB library. This is a regular occurrence (often caused by mis-matched
4062 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4063 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4064 picked up on a user's system that detects such things. There doesn't seem
4065 to be a gcc warning option for this - only an attribute that has to be
4066 put on the function's prototype. It seems that in Fedora Core 4 they have
4067 set this on a number of new functions. No doubt there will be more in due
4070 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4071 specify a transport. However, if an address that was verified by such a
4072 router was the subject of a callout, Exim crashed because it tried to
4073 read the rcpt_include_affixes from the non-existent transport. Now it
4074 just assumes that the setting of that option is false. This bug was
4075 introduced by 4.51/PH/31.
4077 PH/28 Changed -d+all to exclude +memory, because that information is very
4078 rarely of interest, but it makes the output a lot bigger. People tend to
4079 do -d+all out of habit.
4081 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4082 code in os-type was giving problems when libc.so lives in lib64, like on
4085 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4086 aren't the modern standard, and it seems that some systems' include files
4087 don't always have them. Exim was already checking for some of the newer
4088 ones like T_AAAA, and defining it itself. I've added checks for all the
4089 record types that Exim uses.
4091 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4092 not automatically generating a new one, as it is supposed to. This
4093 prevented TLS from working. If the file did exist, but contained invalid
4094 data, a new version was generated, as expected. It was only the case of a
4095 non-existent file that was broken.
4097 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4098 with a change in libdomainkeys > 0.64.
4100 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4101 from DNS. If the selector record carries the flag, it now has
4102 precedence over the domain-wide flag.
4104 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4106 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4107 the use of a port name that isn't defined in /etc/services) occurred, the
4108 message was deferred as in a normal delivery, and thus remained on the
4109 spool, instead of being failed because of the mua_wrapper setting. This
4110 is now fixed, and I tidied up some of the mua_wrapper messages at the
4113 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4114 the messages in a hash of arrays rather than using individual hashes.
4115 This is a bit cleaner and results in dramatic memory savings, albeit
4116 at a slight CPU cost.
4118 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4119 as requested by Marc Sherman.
4121 SC/10 Eximstats: added histograms for user specified patterns as requested
4124 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4126 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4127 fopen() in the content-scanning modules that did not already have it.
4133 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4134 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4136 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4138 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4140 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4141 to test Sieve filters that use "vacation".
4143 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4144 that changes the way the GnuTLS parameters are stored in the cache file.
4145 The new format can be generated externally. For backward compatibility,
4146 if the data in the cache doesn't make sense, Exim assumes it has read an
4147 old-format file, and it generates new data and writes a new file. This
4148 means that you can't go back to an older release without removing the
4151 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4152 work if there are any delivery delays because "one_time" forces the
4153 parent to be marked "delivered", so its unseen clone is never tried
4154 again. For this reason, Exim now forbids the simultaneous setting of
4157 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4158 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4159 entry if you want to know the details. The fix, however, neglected to
4160 consider the case when local delivery batching is involved. The test for
4161 "previously delivered" was not happening when checking to see if an
4162 address could be batched with a previous (undelivered) one; under
4163 certain circumstances this could lead to multiple deliveries to the same
4166 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4167 in its include files, and this causes problems building Exim.
4169 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4170 header_syntax) but Exim was just ignoring anything given after a slash.
4171 In particular, this caused confusion with an attempt to use "verify =
4172 reverse_host_lookup/defer_ok". An error is now given when options are
4173 supplied for verify items that do not have them. (Maybe reverse_host_
4174 lookup should have a defer_ok option, but that's a different point.)
4176 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4177 defined by RFC 821) to 2048, because there were problems with some AUTH
4178 commands, and RFC 1869 says the size should be increased for extended
4179 SMTP commands that take arguments.
4181 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4184 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4185 "unknown" error; now it says that the functionality isn't in the binary.
4187 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4188 an address' error message when a string expansion fails (syntax or
4189 whatever). Otherwise the password may appear in the log. Following change
4190 PH/42 below, there is no longer a chance of it appearing in a bounce
4193 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4195 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4196 of its addresses was ever tried. (Bugzilla bug #2.)
4198 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4199 the result incorrectly in the debug output. (It correctly added a newline
4200 to what was transported.)
4202 TF/01 Added $received_time.
4204 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4205 commented out examples of how to interface to a virus scanner and to
4206 SpamAssassin. Also added commented examples of av_scanner and
4207 spamd_address settings.
4209 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4210 and controls are allowed in which ACLs. There were a couple of minor
4211 errors. Some of the entries in the conditions table (which is a table of
4212 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4213 negation of where the condition IS allowed.
4215 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4217 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4218 header file does not have a version number, so I've had to invent a new
4219 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4220 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4221 radiusclient), but it was contributed by a Radius user.
4223 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4224 files or directories, for OpenSSL.
4226 PH/19 When an Exim process that is running as root has to create an Exim log
4227 file, it does so in a subprocess that runs as exim:exim so as to get the
4228 ownership right at creation (otherwise, other Exim processes might see
4229 the file with the wrong ownership). There was no test for failure of this
4230 fork() call, which would lead to the process getting stuck as it waited
4231 for a non-existent subprocess. Forks do occasionally fail when resources
4232 run out. I reviewed all the other calls to fork(); they all seem to check
4235 PH/20 When checking for unexpected SMTP input at connect time (before writing
4236 the banner), Exim was not dealing correctly with a non-positive return
4237 from the read() function. If the client had disconnected by this time,
4238 the result was a log entry for a synchronization error with an empty
4239 string after "input=" when read() returned zero. If read() returned -1
4240 (an event I could not check), uninitialized data bytes were printed.
4241 There were reports of junk text (parts of files, etc) appearing after
4244 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4246 PH/22 Added support for macro redefinition, and (re)definition in between
4247 driver and ACL definitions.
4249 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4250 forgetting to use the resulting value; it was using the unexpanded value.
4252 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4253 hadn't been configured. The fix is from Juergen Kreileder, who
4254 understands it better than I do:
4256 "Here's what I see happening with three configured cyrus_sasl
4257 authenticators configured (plain, login, cram-md5):
4259 On startup auth_cyrus_sasl_init() gets called for each of these.
4260 This means three calls to sasl_listmech() without a specified mech_list.
4261 => SASL tests which mechs of all available mechs actually work
4262 => three warnings about OTP not working
4263 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4265 With the patch, sasl_listmech() also gets called three times. But now
4266 SASL's mech_list option is set to the server_mech specified in the the
4267 authenticator. Or in other words, the answer from sasl_listmech()
4268 gets limited to just the mech you're testing for (which is different
4270 => the return list contains just 'plain' or 'login', 'cram-md5' or
4271 nothing depending on the value of ob->server_mech.
4273 I've just tested the patch: Authentication still works fine,
4274 unavailable mechs specified in the exim configuration are still
4275 caught, and the auth.log warnings about OTP are gone."
4277 PH/25 When debugging is enabled, the contents of the command line are added
4278 to the debugging output, even when log_selector=+arguments is not
4281 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4282 answer is "GNU", and only if the return is "GNU/something" is the answer
4285 PH/27 $acl_verify_message is now set immediately after the failure of a
4286 verification in an ACL, and so is available in subsequent modifiers. In
4287 particular, the message can be preserved by coding like this:
4289 warn !verify = sender
4290 set acl_m0 = $acl_verify_message
4292 Previously, $acl_verify_message was set only while expanding "message"
4293 and "log_message" when a very denied access.
4295 PH/28 Modified OS/os.c-Linux with
4297 -#ifndef OS_LOAD_AVERAGE
4298 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4300 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4301 nomenclature these days.)
4303 PH/29 Installed patch from the Sieve maintainer that adds the options
4304 sieve_useraddress and sieve_subaddress to the redirect router.
4306 PH/30 In these circumstances:
4307 . Two addresses routed to the same list of hosts;
4308 . First host does not offer TLS;
4309 . First host accepts first address;
4310 . First host gives temporary error to second address;
4311 . Second host offers TLS and a TLS session is established;
4312 . Second host accepts second address.
4313 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4314 and peerdn, if requested) that were in fact used only for the second
4317 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4318 attention to any local part prefix or suffix that was matched by the
4319 router that accepted the address. It now behaves in the same way as it
4320 does for delivery: the affixes are removed from the local part unless
4321 rcpt_include_affixes is set on the transport.
4323 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4324 timeout during the DATA phase of an incoming message.
4326 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4327 applied a patch sanctioned by the Sieve maintainer.
4329 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4330 the uid or gid is negative. A case of a negative gid caused this to be
4331 noticed. The fix allows for either to be negative.
4333 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4334 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4337 PH/36 The change PH/12 above was broken. Fixed it.
4339 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4340 the grounds that routing the same address twice would always produce the
4341 same answer. This might have been true once, but it is certainly no
4342 longer true now. Routing a child address may depend on the previous
4343 routing that produced that child. Some complicated redirection strategies
4344 went wrong when messages had multiple recipients, and made Exim's
4345 behaviour dependent on the order in which the addresses were given.
4347 I have moved the duplicate checking until after the routing is complete.
4348 Exim scans the addresses that are assigned to local and remote
4349 transports, and removes any duplicates. This means that more work will be
4350 done, as duplicates will always all be routed, but duplicates are
4351 presumably rare, so I don't expect this is of any significance.
4353 For deliveries to pipes, files, and autoreplies, the duplicate checking
4354 still happens during the routing process, since they are not going to be
4357 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4358 It corrects a timeout issue with spamd. This is Ian's comment: "The
4359 background is that sometimes spamd either never reads data from a
4360 connection it has accepted, or it never writes response data. The exiscan
4361 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4362 blindly assumes that writes won't block so it may never time out."
4364 PH/39 Allow G after quota size as well as K and M.
4366 PH/40 The value set for $authenticated_id in an authenticator may not contain
4367 binary zeroes or newlines because the value is written to log lines and
4368 to spool files. There was no check on this. Now the value is run through
4369 the string_printing() function so that such characters are converted to
4370 printable escape sequences.
4372 PH/41 $message_linecount is a new variable that contains the total number of
4373 lines in the message. Compare $body_linecount, which is the count for the
4376 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4377 bounce and delay warning messages, except in certain special cases, which
4380 (a) An SMTP error message from a remote host;
4381 (b) A message specified in a :fail: redirection;
4382 (c) A message specified in a "fail" command in a system filter;
4383 (d) A message specified in a FAIL return from the queryprogram router;
4384 (e) A message specified by the cannot_route_message router option.
4386 In these cases only, Exim does include the error details in bounce and
4387 warning messages. There are also a few cases where bland messages such
4388 as "unrouteable address" or "local delivery error" are given.
4390 PH/43 $value is now also set for the "else" part of a ${run expansion.
4392 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4393 being worked on, but at least Exim now implements the latest version to
4396 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4397 process to complete was treated as a delivery failure, a timeout while
4398 writing the message to the pipe was logged, but erroneously treated as a
4399 successful delivery. Such timeouts include transport filter timeouts. For
4400 consistency with the overall process timeout, these timeouts are now
4401 treated as errors, giving rise to delivery failures by default. However,
4402 there is now a new Boolean option for the pipe transport called
4403 timeout_defer, which, if set TRUE, converts the failures into defers for
4404 both kinds of timeout. A transport filter timeout is now identified in
4407 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4408 systems where "make" and "gmake" are different, calling "gmake" at top
4409 level broke things. I've arranged for the value of $(MAKE) to be passed
4410 from the Makefile to this script so that it can call the same version of
4414 A note about Exim versions 4.44 and 4.50
4415 ----------------------------------------
4417 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4418 changes of various kinds. As a consequence, a big documentation update was
4419 needed. This delayed the release for rather longer than seemed good, especially
4420 in the light of a couple of (minor) security issues. Therefore, the changes
4421 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4422 release. So 4.44 and 4.50 are in effect two different branches that both start
4425 I have left the 4.50 change log unchanged; it contains all the changes since
4426 4.43. The change log for 4.44 is below; many of its items are identical to
4427 those for 4.50. This seems to be the most sensible way to preserve the
4428 historical information.
4434 1. Minor wording change to the doc/README.SIEVE file.
4436 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4437 computation of the current number of files was incorrect.
4439 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4440 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4443 4. Give more explanation in the error message when the command for a transport
4444 filter fails to execute.
4446 5. There are several places where Exim runs a non-Exim command in a
4447 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4448 was being done only for the command run by the queryprogram router. It is
4449 now done for all such subprocesses. The other cases are: ${run, transport
4450 filters, and the commands run by the lmtp and pipe transports.
4452 6. Added CONFIGURE_GROUP build-time option.
4454 7. Some older OS have a limit of 256 on the maximum number of file
4455 descriptors. Exim was using setrlimit() to set 1000 as a large value
4456 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4457 systems. I've change it so that if it can't get 1000, it tries for 256.
4459 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4460 was an oversight, and furthermore, ever since the addition of extra
4461 controls (e.g. 4.43/32), the checks on when to allow different forms of
4462 "control" were broken. There should now be diagnostics for all cases when a
4463 control that does not make sense is encountered.
4465 9. Added the /retain_sender option to "control=submission".
4467 10. $recipients is now available in the predata ACL (oversight).
4469 11. Tidy the search cache before the fork to do a delivery from a message
4470 received from the command line. Otherwise the child will trigger a lookup
4471 failure and thereby defer the delivery if it tries to use (for example) a
4472 cached ldap connection that the parent has called unbind on.
4474 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4475 of $address_data from the recipient verification was clobbered by the
4476 sender verification.
4478 13. The value of address_data from a sender verification is now available in
4479 $sender_address_data in subsequent conditions in the ACL statement.
4481 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4483 15. Added a new option "connect=<time>" to callout options, to set a different
4486 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4487 was its contents. (It was OK if the option was not defined at all.)
4489 17. A "Completed" log line is now written for messages that are removed from
4490 the spool by the -Mrm option.
4492 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4493 information about exactly what failed.
4495 19. Added -dd to debug only the daemon process.
4497 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4498 handles timeouts, both on the server side and network timeouts. Renamed the
4499 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4501 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4502 It is now set to "smtps".
4504 22. $host_address is now set to the target address during the checking of
4505 ignore_target_hosts.
4507 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4508 being passed; this would have caused $sender_host_name to have been used if
4509 matching the list had actually called for a host name (not very likely,
4510 since this list is usually IP addresses). A host name is now passed as
4513 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4514 code that specifies a non-restarting handler (typically sigaction() in
4515 modern systems) in an attempt to fix a rare and obscure crash bug.
4517 25. Narrowed the window for a race in the daemon that could cause it to ignore
4518 SIGCHLD signals. This is not a major problem, because they are used only to
4519 wake it up if nothing else does.
4521 26. A malformed maildirsize file could cause Exim to calculate negative values
4522 for the mailbox size or file count. Odd effects could occur as a result.
4523 The maildirsize information is now recalculated if the size or filecount
4526 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4527 support for a long time. Removed HAVE_SYS_VFS_H.
4529 28. Installed the latest version of exipick from John Jetmore.
4531 29. In an address list, if the pattern was not a regular expression, an empty
4532 subject address (from a bounce message) matched only if the pattern was an
4533 empty string. Non-empty patterns were not even tested. This was the wrong
4534 because it is perfectly reasonable to use an empty address as part of a
4535 database query. An empty address is now tested by patterns that are
4536 lookups. However, all the other forms of pattern expect the subject to
4537 contain a local part and a domain, and therefore, for them, an empty
4538 address still always fails if the pattern is not itself empty.
4540 30. Exim went into a mad DNS loop when attempting to do a callout where the
4541 host was specified on an smtp transport, and looking it up yielded more
4542 than one IP address.
4544 31. Re-factored the code for checking spool and log partition space into a
4545 function that finds that data and another that does the check. The former
4546 is then used to implement four new variables: $spool_space, $log_space,
4547 $spool_inodes, and $log_inodes.
4549 32. The RFC2047 encoding function was originally intended for short strings
4550 such as real names; it was not keeping to the 75-character limit for
4551 encoded words that the RFC imposes. It now respects the limit, and
4552 generates multiple encoded words if necessary. To be on the safe side, I
4553 have increased the buffer size for the ${rfc2047: expansion operator from
4556 33. It is now permitted to omit both strings after an "if" condition; if the
4557 condition is true, the result is "true". As before, when the second string
4558 is omitted, a false condition yields an empty string. This makes it less
4559 cumbersome to write custom ACL and router conditions.
4561 34. Failure to deliver a bounce message always caused it to be frozen, even if
4562 there was an errors_to setting on the router. The errors_to setting is now
4565 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4566 canonical form (fully expanded) before being placed in
4567 $sender_host_address.
4569 36. The table in the code that translates DNS record types into text (T_A to
4570 "A" for instance) was missing entries for NS and CNAME. It is just possible
4571 that this could have caused confusion if both these types were looked up
4572 for the same domain, because the text type is used as part of Exim's
4573 per-process caching. But the chance of anyone hitting this buglet seems
4576 37. The dnsdb lookup has been extended in a number of ways.
4578 (1) There is a new type, "zns", which walks up the domain tree until it
4579 finds some nameserver records. It should be used with care.
4581 (2) There is a new type, "mxh", which is like "mx" except that it returns
4582 just the host names, not the priorities.
4584 (3) It is now possible to give a list of domains (or IP addresses) to be
4585 looked up. The behaviour when one of the lookups defers can be
4586 controlled by a keyword.
4588 (4) It is now possible to specify the separator character for use when
4589 multiple records are returned.
4591 38. The dnslists ACL condition has been extended: it is now possible to supply
4592 a list of IP addresses and/or domains to be looked up in a particular DNS
4595 39. Added log_selector=+queue_time_overall.
4597 40. When running the queue in the test harness, wait just a tad after forking a
4598 delivery process, to get repeatability of debugging output.
4600 41. Include certificate and key file names in error message when GnuTLS fails
4601 to set them up, because the GnuTLS error message doesn't include the name
4602 of the failing file when there is a problem reading it.
4604 42. Allow both -bf and -bF in the same test run.
4606 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4608 44. The "Exiscan patch" is now merged into the mainline Exim source.
4610 45. Sometimes the final signoff response after QUIT could fail to get
4611 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4612 < 0 before doing a fflush(). This bug looks as though it goes back to the
4613 introduction of TLS in release 3.20, but "sometimes" must have been rare
4614 because the tests only now provoked it.
4616 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4617 (this can affect the format of dates).
4619 47. exim_tidydb, when checking for the continued existence of a message for
4620 which it has found a message-specific retry record, was not finding
4621 messages that were in split spool directories. Consequently, it was
4622 deleting retry records that should have stayed in existence.
4624 48. Steve fixed some bugs in eximstats.
4626 49. The SPA authentication driver was not abandoning authentication and moving
4627 on to the next authenticator when an expansion was forced to fail,
4628 contradicting the general specification for all authenticators. Instead it
4629 was generating a temporary error. It now behaves as specified.
4631 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4632 (the order specifies the preference for clients). The order is now AES256,
4633 AES128, 3DES, ARCFOUR128.
4635 51. Small patch to Sieve code - explicitly set From: when generating an
4638 52. Exim crashed if a remote delivery caused a very long error message to be
4639 recorded - for instance if somebody sent an entire SpamAssassin report back
4640 as a large number of 550 error lines. This bug was coincidentally fixed by
4641 increasing the size of one of Exim's internal buffers (big_buffer) that
4642 happened as part of the Exiscan merge. However, to be on the safe side, I
4643 have made the code more robust (and fixed the comments that describe what
4646 53. Now that there can be additional text after "Completed" in log lines (if
4647 the queue_time_overall log selector is set), a one-byte patch to exigrep
4648 was needed to allow it to recognize "Completed" as not the last thing in
4651 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4652 patch that reportedly fixes this has been added. I am not expert enough to
4653 create a test for it. This is what the patch creator wrote:
4655 "I found a little strange behaviour of ldap code when working with
4656 Windows 2003 AD Domain, where users was placed in more than one
4657 Organization Units. When I tried to give exim partial DN, the exit code
4658 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4659 But simultaneously result of request was absolutely normal ldap result,
4660 so I produce this patch..."
4662 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4663 REFERENCE, so I have modified the code to exclude the patch when that macro
4666 55. Some experimental protocols are using DNS PTR records for new purposes. The
4667 keys for these records are domain names, not reversed IP addresses. The
4668 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4669 leaves it alone. Component reversal etc. now happens only for IP addresses.
4672 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4674 57. Double the size of the debug message buffer (to 2048) so that more of very
4675 long debug lines gets shown.
4677 58. The exicyclog utility now does better if the number of log files to keep
4678 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4680 59. Two changes related to the smtp_active_hostname option:
4682 (1) $smtp_active_hostname is now available as a variable.
4683 (2) The default for smtp_banner uses $smtp_active_hostname instead
4684 of $primary_hostname.
4686 60. The host_aton() function is supposed to be passed a string that is known
4687 to be a valid IP address. However, in the case of IPv6 addresses, it was
4688 not checking this. This is a hostage to fortune. Exim now panics and dies
4689 if the condition is not met. A case was found where this could be provoked
4690 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4691 components; fortuitously, this particular loophole had already been fixed
4692 by change 4.50/55 above.
4694 If there are any other similar loopholes, the new check in host_aton()
4695 itself should stop them being exploited. The report I received stated that
4696 data on the command line could provoke the exploit when Exim was running as
4697 exim, but did not say which command line option was involved. All I could
4698 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4699 running as the user.
4702 61. There was a buffer overflow vulnerability in the SPA authentication code
4703 (which came originally from the Samba project). I have added a test to the
4704 spa_base64_to_bits() function which I hope fixes it.
4707 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4708 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4710 63. The daemon start-up calls getloadavg() while still root for those OS that
4711 need the first call to be done as root, but it missed one case: when
4712 deliver_queue_load_max is set with deliver_drop_privilege. This is
4713 necessary for the benefit of the queue runner, because there is no re-exec
4714 when deliver_drop_privilege is set.
4716 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4717 This has been fixed.
4719 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4720 was in use, was not putting the data itself into the right store pool;
4721 consequently, it could be overwritten for a subsequent message in the same
4722 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4725 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4727 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4728 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4729 the function were treating the return as a boolean value, which happened to
4730 work because 0=false and not-0=true, but is not correct code.
4732 68. The host_aton() function was not handling scoped IPv6 addresses (those
4733 with, for example, "%eth0" on the end) correctly.
4735 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4736 negated items (that is, ~something) in unsigned ints. Some compilers
4737 apparently mutter when there is no cast.
4739 70. If an address verification called from an ACL failed, and did not produce a
4740 user-specific message (i.e. there was only a "system" message), nothing was
4741 put in $acl_verify_message. In this situation, it now puts the system
4744 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4745 session; change 4.31/43 added the unwanted input to the log line - except
4746 that it did not do this in the start of session case. It now does.
4748 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4749 This is wrong and can cause the other end to generate a synchronization
4750 error if it is another Exim or anything else that does the synchronization
4751 check. A QUIT command is no longer sent after a timeout.
4753 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4754 during host lookups.
4756 74. The defer_ok option of callout verification was not working if it was used
4757 when verifying addresses in header lines, that is, for this case:
4759 verify = header_sender/callout=defer_ok
4761 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4762 those file descriptors could be used for SMTP connections. If anything
4763 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4764 could be sent to the SMTP client, causing chaos. The daemon now opens
4765 stdin, stdout, and stderr to /dev/null when it puts itself into the
4768 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4769 log by default. The user can override this with suitable Perl magic.
4771 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4772 the log message when discard triggers, was not working for the DATA ACL or
4773 for the non-SMTP ACL.
4775 78. Error message wording change in sieve.c.
4777 79. If smtp_accept_max_per_host was set, the number of connections could be
4778 restricted to fewer than expected, because the daemon was trying to set up
4779 a new connection before checking whether the processes handling previous
4780 connections had finished. The check for completed processes is now done
4781 earlier. On busy systems, this bug wouldn't be noticed because something
4782 else would have woken the daemon, and it would have reaped the completed
4785 80. If a message was submitted locally by a user whose login name contained one
4786 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4787 It caused a spool format error. I have fixed the spool reading code. A
4788 related problem was that the "from" clause in the Received: line became
4789 illegal because of the space(s). It is now covered by ${quote_local_part.
4791 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4794 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4796 83. Patch from Sieve maintainer to fix unterminated string problem in
4797 "vacation" handling.
4799 84. Some minor changes to the Linux configuration files to help with other
4800 OS variants using glibc.
4802 85. One more patch for Sieve to update vacation handling to latest spec.
4805 ----------------------------------------------------
4806 See the note above about the 4.44 and 4.50 releases.
4807 ----------------------------------------------------
4813 1. Change 4.43/35 introduced a bug that caused file counts to be
4814 incorrectly computed when quota_filecount was set in an appendfile
4817 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4818 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4821 3. Give more explanation in the error message when the command for a transport
4822 filter fails to execute.
4824 4. There are several places where Exim runs a non-Exim command in a
4825 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4826 was being done only for the command run by the queryprogram router. It is
4827 now done for all such subprocesses. The other cases are: ${run, transport
4828 filters, and the commands run by the lmtp and pipe transports.
4830 5. Some older OS have a limit of 256 on the maximum number of file
4831 descriptors. Exim was using setrlimit() to set 1000 as a large value
4832 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4833 systems. I've change it so that if it can't get 1000, it tries for 256.
4835 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4836 was an oversight, and furthermore, ever since the addition of extra
4837 controls (e.g. 4.43/32), the checks on when to allow different forms of
4838 "control" were broken. There should now be diagnostics for all cases when a
4839 control that does not make sense is encountered.
4841 7. $recipients is now available in the predata ACL (oversight).
4843 8. Tidy the search cache before the fork to do a delivery from a message
4844 received from the command line. Otherwise the child will trigger a lookup
4845 failure and thereby defer the delivery if it tries to use (for example) a
4846 cached ldap connection that the parent has called unbind on.
4848 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4849 of $address_data from the recipient verification was clobbered by the
4850 sender verification.
4852 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4853 was its contents. (It was OK if the option was not defined at all.)
4855 11. A "Completed" log line is now written for messages that are removed from
4856 the spool by the -Mrm option.
4858 12. $host_address is now set to the target address during the checking of
4859 ignore_target_hosts.
4861 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4862 being passed; this would have caused $sender_host_name to have been used if
4863 matching the list had actually called for a host name (not very likely,
4864 since this list is usually IP addresses). A host name is now passed as
4867 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4868 code that specifies a non-restarting handler (typically sigaction() in
4869 modern systems) in an attempt to fix a rare and obscure crash bug.
4871 15. Narrowed the window for a race in the daemon that could cause it to ignore
4872 SIGCHLD signals. This is not a major problem, because they are used only to
4873 wake it up if nothing else does.
4875 16. A malformed maildirsize file could cause Exim to calculate negative values
4876 for the mailbox size or file count. Odd effects could occur as a result.
4877 The maildirsize information is now recalculated if the size or filecount
4880 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4881 support for a long time. Removed HAVE_SYS_VFS_H.
4883 18. Updated exipick to current release from John Jetmore.
4885 19. Allow an empty sender to be matched against a lookup in an address list.
4886 Previously the only cases considered were a regular expression, or an
4889 20. Exim went into a mad DNS lookup loop when doing a callout where the
4890 host was specified on the transport, if the DNS lookup yielded more than
4893 21. The RFC2047 encoding function was originally intended for short strings
4894 such as real names; it was not keeping to the 75-character limit for
4895 encoded words that the RFC imposes. It now respects the limit, and
4896 generates multiple encoded words if necessary. To be on the safe side, I
4897 have increased the buffer size for the ${rfc2047: expansion operator from
4900 22. Failure to deliver a bounce message always caused it to be frozen, even if
4901 there was an errors_to setting on the router. The errors_to setting is now
4904 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4905 canonical form (fully expanded) before being placed in
4906 $sender_host_address.
4908 24. Updated eximstats to version 1.33
4910 25. Include certificate and key file names in error message when GnuTLS fails
4911 to set them up, because the GnuTLS error message doesn't include the name
4912 of the failing file when there is a problem reading it.
4914 26. Expand error message when OpenSSL has problems setting up cert/key files.
4917 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4918 (this can affect the format of dates).
4920 28. exim_tidydb, when checking for the continued existence of a message for
4921 which it has found a message-specific retry record, was not finding
4922 messages that were in split spool directories. Consequently, it was
4923 deleting retry records that should have stayed in existence.
4925 29. eximstats updated to version 1.35
4926 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4927 1.35 - bugfix such that pie charts by volume are generated correctly
4929 30. The SPA authentication driver was not abandoning authentication and moving
4930 on to the next authenticator when an expansion was forced to fail,
4931 contradicting the general specification for all authenticators. Instead it
4932 was generating a temporary error. It now behaves as specified.
4934 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4935 (the order specifies the preference for clients). The order is now AES256,
4936 AES128, 3DES, ARCFOUR128.
4938 31. Small patch to Sieve code - explicitly set From: when generating an
4941 32. Exim crashed if a remote delivery caused a very long error message to be
4942 recorded - for instance if somebody sent an entire SpamAssassin report back
4943 as a large number of 550 error lines. This bug was coincidentally fixed by
4944 increasing the size of one of Exim's internal buffers (big_buffer) that
4945 happened as part of the Exiscan merge. However, to be on the safe side, I
4946 have made the code more robust (and fixed the comments that describe what
4949 33. Some experimental protocols are using DNS PTR records for new purposes. The
4950 keys for these records are domain names, not reversed IP addresses. The
4951 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4952 leaves it alone. Component reversal etc. now happens only for IP addresses.
4955 34. The host_aton() function is supposed to be passed a string that is known
4956 to be a valid IP address. However, in the case of IPv6 addresses, it was
4957 not checking this. This is a hostage to fortune. Exim now panics and dies
4958 if the condition is not met. A case was found where this could be provoked
4959 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4960 components; fortuitously, this particular loophole had already been fixed
4961 by change 4.50/55 or 4.44/33 above.
4963 If there are any other similar loopholes, the new check in host_aton()
4964 itself should stop them being exploited. The report I received stated that
4965 data on the command line could provoke the exploit when Exim was running as
4966 exim, but did not say which command line option was involved. All I could
4967 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4968 running as the user.
4971 35. There was a buffer overflow vulnerability in the SPA authentication code
4972 (which came originally from the Samba project). I have added a test to the
4973 spa_base64_to_bits() function which I hope fixes it.
4976 36. The daemon start-up calls getloadavg() while still root for those OS that
4977 need the first call to be done as root, but it missed one case: when
4978 deliver_queue_load_max is set with deliver_drop_privilege. This is
4979 necessary for the benefit of the queue runner, because there is no re-exec
4980 when deliver_drop_privilege is set.
4982 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4983 was in use, was not putting the data itself into the right store pool;
4984 consequently, it could be overwritten for a subsequent message in the same
4985 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4988 38. Sometimes the final signoff response after QUIT could fail to get
4989 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4990 < 0 before doing a fflush(). This bug looks as though it goes back to the
4991 introduction of TLS in release 3.20, but "sometimes" must have been rare
4992 because the tests only now provoked it.
4998 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4999 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5000 Now it returns an integer. A number of calls were still expecting a T/F
5001 return. Fortuitously, in all cases, the tests worked in OK situations,
5002 which is the norm. However, things would have gone wrong on any write
5003 failures on the smtp file descriptor. This function is used when sending
5004 messages over SMTP and also when doing verify callouts.
5006 2. When Exim is called to do synchronous delivery of a locally submitted
5007 message (the -odf or -odi options), it no longer closes stderr before doing
5010 3. Implemented the mua_wrapper option.
5012 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5014 5. Implemented the functions header_remove(), header_testname(),
5015 header_add_at_position(), and receive_remove_recipient(), and exported them
5018 6. If an ACL "warn" statement specified the addition of headers, Exim already
5019 inserted X-ACL-Warn: at the start if there was no header name. However, it
5020 was not making this test for the second and subsequent header lines if
5021 there were newlines in the string. This meant that an invalid header could
5022 be inserted if Exim was badly configured.
5024 7. Allow an ACL "warn" statement to add header lines at the start or after all
5025 the Received: headers, as well as at the end.
5027 8. Added the rcpt_4xx retry error code.
5029 9. Added postmaster_mailfrom=xxx to callout verification option.
5031 10. Added mailfrom=xxxx to the callout verification option, for verify=
5034 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5035 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5037 12. Inserted some casts to stop certain compilers warning when using pointer
5038 differences as field lengths or precisions in printf-type calls (mostly
5039 affecting debugging statements).
5041 13. Added optional readline() support for -be (dynamically loaded).
5043 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5044 same clock tick as a message's arrival, so that its received time was the
5045 same as the "first fail" time on the retry record, and that message
5046 remained on the queue past the ultimate address timeout, every queue runner
5047 would try a delivery (because it was past the ultimate address timeout) but
5048 after another failure, the ultimate address timeout, which should have then
5049 bounced the address, did not kick in. This was a "< instead of <=" error;
5050 in most cases the first failure would have been in the next clock tick
5051 after the received time, and all would be well.
5053 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5054 being recognized when the domain list was tested by the match_domain
5055 condition in an expansion string.
5057 16. Added the ${str2b64: operator.
5059 17. Exim was always calling setrlimit() to set a large limit for the number of
5060 processes, without checking whether the existing limit was already
5061 adequate. (It did check for the limit on file descriptors.) Furthermore,
5062 errors from getrlimit() and setrlimit() were being ignored. Now they are
5063 logged to the main and panic logs, but Exim does carry on, to try to do its
5064 job under whatever limits there are.
5066 18. Imported PCRE 5.0.
5068 19. Trivial typo in log message " temporarily refused connection" (the leading
5071 20. If the log selector return_path_on_delivery was set and an address was
5072 redirected to /dev/null, the delivery process crashed because it assumed
5073 that a return path would always be set for a "successful" delivery. In this
5074 case, the whole delivery is bypassed as an optimization, and therefore no
5077 21. Internal re-arrangement: the function for sending a challenge and reading
5078 a response while authentication was assuming a zero-terminated challenge
5079 string. It's now changed to take a pointer and a length, to allow for
5080 binary data in such strings.
5082 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5084 23. Exim was not respecting finduser_retries when seeking the login of the
5085 uid under which it was called; it was always trying 10 times. (The default
5086 setting of finduser_retries is zero.) Also, it was sleeping after the final
5087 failure, which is pointless.
5089 24. Implemented tls_on_connect_ports.
5091 25. Implemented acl_smtp_predata.
5093 26. If the domain in control=submission is set empty, Exim assumes that the
5094 authenticated id is a complete email address when it generates From: or
5095 Sender: header lines.
5097 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5098 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5099 chown and chgrp in /bin and hostname in /usr/bin.
5101 28. Exim was keeping the "process log" file open after each use, just as it
5102 does for the main log. This opens the possibility of it remaining open for
5103 long periods when the USR1 signal hits a daemon. Occasional processlog
5104 errors were reported, that could have been caused by this. Anyway, it seems
5105 much more sensible not to leave this file open at all, so that is what now
5108 29. The long-running daemon process does not normally write to the log once it
5109 has entered its main loop, and it closes the log before doing so. This is
5110 so that log files can straightforwardly be renamed and moved. However,
5111 there are a couple of unusual error situations where the daemon does write
5112 log entries, and I had neglected to close the log afterwards.
5114 30. The text of an SMTP error response that was received during a remote
5115 delivery was being truncated at 512 bytes. This is too short for some of
5116 the long messages that one sometimes sees. I've increased the limit to
5119 31. It is now possible to make retry rules that apply only when a message has a
5120 specific sender, in particular, an empty sender.
5122 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5123 it possible to be selective about when SMTP synchronization is enforced.
5125 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5127 32. Implemented hosts_connection_nolog.
5129 33. Added an ACL for QUIT.
5131 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5134 35. Added mailbox_size and mailbox_filecount to appendfile.
5136 36. Added control = no_multiline_responses to ACLs.
5138 37. There was a bug in the logic of the code that waits for the clock to tick
5139 in the case where the clock went backwards by a substantial amount such
5140 that the microsecond fraction of "now" was more than the microsecond
5141 fraction of "then" (but the whole seconds number was less).
5143 38. Added support for the libradius Radius client library this is found on
5144 FreeBSD (previously only the radiusclient library was supported).
5150 1. When certain lookups returned multiple values in the form name=value, the
5151 quoting of the values was not always being done properly. Specifically:
5152 (a) If the value started with a double quote, but contained no whitespace,
5154 (b) If the value contained whitespace other than a space character (i.e.
5155 tabs or newlines or carriage returns) it was not quoted.
5156 This fix has been applied to the mysql and pgsql lookups by writing a
5157 separate quoting function and calling it from the lookup code. The fix
5158 should probably also be applied to nisplus, ibase and oracle lookups, but
5159 since I cannot test any of those, I have not disturbed their existing code.
5161 2. A hit in the callout cache for a specific address caused a log line with no
5162 reason for rejecting RCPT. Now it says "Previous (cached) callout
5163 verification failure".
5165 3. There was an off-by-one bug in the queryprogram router. An over-long
5166 return line was truncated at 256 instead of 255 characters, thereby
5167 overflowing its buffer with the terminating zero. As well as fixing this, I
5168 have increased the buffer size to 1024 (and made a note to document this).
5170 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5171 when Exim is waiting for an SMTP response from a remote server, Exim
5172 restarts its select() call on the socket, thereby resetting its timeout.
5173 This is not a problem when such interrupts are rare. Somebody set up a cron
5174 job to run exiwhat every 2 minutes, which is less than the normal select()
5175 timeout (5 or 10 minutes). This meant that the select() timeout never
5176 kicked in because it was always reset. I have fixed this by comparing the
5177 time when an interrupt arrives with the time at the start of the first call
5178 to select(). If more time than the timeout has elapsed, the interrupt is
5179 treated as a timeout.
5181 5. Some internal re-factoring in preparation for the addition of Sieve
5182 extensions (by MH). In particular, the "personal" test is moved to a
5183 separate function, and given an option for scanning Cc: and Bcc: (which is
5184 not set for Exim filters).
5186 6. When Exim created an email address using the login of the caller as the
5187 local part (e.g. when creating a From: or Sender: header line), it was not
5188 quoting the local part when it contained special characters such as @.
5190 7. Installed new OpenBSD configuration files.
5192 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5193 try to make them clearer.
5195 9. Callout options, other than the timeout value, were being ignored when
5196 verifying sender addresses in header lines. For example, when using
5198 verify = header_sender/callout=no_cache
5200 the cache was (incorrectly) being used.
5202 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5203 only the Cygwin environment.
5205 11. When return_path_on_delivery was set as a log selector, if different remote
5206 addresses in the same message used different return paths and parallel
5207 remote delivery occurred, the wrong values would sometimes be logged.
5208 (Whenever a remote delivery process finished, the return path value from
5209 the most recently started remote delivery process was logged.)
5211 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5212 lines when AUTH and/or TLS are in use. This is the "received protocol"
5213 field. Exim used to use "asmtp" for authenticated SMTP, without any
5214 indication (in the protocol name) for TLS use. Now it follows the RFC and
5215 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5216 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5217 names appear in log lines as well as in Received: header lines.
5219 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5220 extensions, and comparison tests, and to fix some bugs.
5222 14. Changes to the "personal" filter test:
5224 (1) The test was buggy in that it was just doing the equivalent of
5225 "contains" tests on header lines. For example, if a user's address was
5226 anne@some.where, the "personal" test would incorrectly be true for
5228 To: susanne@some.where
5230 This test is now done by extracting each address from the header in turn,
5231 and checking the entire address. Other tests that are part of "personal"
5232 are now done using regular expressions (for example, to check local parts
5233 of addresses in From: header lines).
5235 (2) The list of non-personal local parts in From: addresses has been
5236 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5237 taken from the Sieve specification recommendations.
5239 (3) If the message contains any header line starting with "List-" it is
5240 treated as non-personal.
5242 (4) The test for "circular" in the Subject: header line has been removed
5243 because it now seems ill-conceived.
5245 15. Minor typos in src/EDITME comments corrected.
5247 16. Installed latest exipick from John Jetmore.
5249 17. If headers_add on a router specified a text string that was too long for
5250 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5251 of string_sprintf() is now avoided.
5253 18. $message_body_size was not set (it was always zero) when running the DATA
5254 ACL and the local_scan() function.
5256 19. For the "mail" command in an Exim filter, no default was being set for
5257 the once_repeat time, causing a random time value to be used if "once" was
5258 specified. (If the value happened to be <= 0, no repeat happened.) The
5259 default is now 0s, meaning "never repeat". The "vacation" command was OK
5260 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5261 (I found it when inspecting the code).
5263 20. There is now an overall timeout for performing a callout verification. It
5264 defaults to 4 times the callout timeout, which applies to individual SMTP
5265 commands during the callout. The overall timeout applies when there is more
5266 than one host that can be tried. The timeout is checked before trying the
5267 next host. This prevents very long delays if there are a large number of
5268 hosts and all are timing out (e.g. when the network connections are timing
5269 out). The value of the overall timeout can be changed by specifying an
5270 additional sub-option for "callout", called "maxwait". For example:
5272 verify = sender/callout=5s,maxwait=20s
5274 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5275 to the end before writing, but this should make it even safer).
5277 22. Exim was forgetting that it had advertised PIPELINING for the second and
5278 subsequent messages on an SMTP connection. It was also not resetting its
5279 memory on STARTTLS and an internal HELO.
5281 23. When Exim logs an SMTP synchronization error within a session, it now
5282 records whether PIPELINING has been advertised or not.
5284 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5285 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5286 rather than long int.
5288 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5290 26. Added the never_mail option to autoreply.
5296 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5297 crash if the getsockname() call failed; this can happen if a connection is
5298 closed very soon after it is established. The problem was simply in the
5299 order in which certain operations were done, causing Exim to try to write
5300 to the SMTP stream before it had set up the file descriptor. The bug has
5301 been fixed by making things happen in the correct order.
5307 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5308 before the connection was closed, thus losing the rejection response.
5310 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5311 some early Solaris releases, but causes trouble in current releases where
5312 socklen_t is defined.
5314 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5317 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5320 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5321 that releasing the top of it at the end releases what was used for sub-
5322 expansions (unless the block got too big). However, discard this block if
5323 the first thing is a variable or header, so that we can use its block when
5324 it is dynamic (useful for very large $message_headers, for example).
5326 6. Lookups now cache *every* query, not just the most recent. A new, separate
5327 store pool is used for this. It can be recovered when all lookup caches are
5328 flushed. Lookups now release memory at the end of their result strings.
5329 This has involved some general refactoring of the lookup sources.
5331 7. Some code has been added to the store_xxx() functions to reduce the amount
5332 of flapping under certain conditions.
5334 8. log_incoming_interface used to affect only the <= reception log lines. Now
5335 it causes the local interface and port to be added to several more SMTP log
5336 lines, for example "SMTP connection from", and rejection lines.
5338 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5340 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5342 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5343 could be overwritten at the end of the current message (or the start of a
5344 new message if it was set in a HELO ACL). The value is now preserved for
5345 the duration of the SMTP connection.
5347 12. If a transport had a headers_rewrite setting, and a matching header line
5348 contained an unqualified address, that address was qualified, even if it
5349 did not match any rewriting rules. The underlying bug was that the values
5350 of the flags that permit the existence of unqualified sender and recipient
5351 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5352 non-local messages, and by -bnq for local messages) were not being
5353 preserved with the message after it was received.
5355 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5356 "next input=" as part of the text comprising the host identity instead of
5357 the correct text. The code was using the same buffer for two different
5358 strings. However, depending on which order the printing function evaluated
5359 its arguments, the bug did not always show up. Under Linux, for example, my
5360 test suite worked just fine.
5362 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5363 doesn't work with some older versions of Perl. It has been changed to "my",
5364 which in any case is probably the better facility to use.
5366 15. A really picky compiler found some instances of statements for creating
5367 error messages that either had too many or two few arguments for the format
5370 16. The size of the buffer for calls to the DNS resolver has been increased
5371 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5372 for addresses that have a lot of PTR records. This alleviates a problem; it
5373 does not fully solve it.
5375 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5376 buffer now truncates the list and logs the incident, which is the same
5377 action as happens when Exim is looking up a host name and its aliases.
5378 Previously in this situation something unpredictable would happen;
5379 sometimes it was "internal error: store_reset failed".
5381 18. If a server dropped the connection unexpectedly when an Exim client was
5382 using GnuTLS and trying to read a response, the client delivery process
5383 crashed while trying to generate an error log message.
5385 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5386 string, for example:
5388 warn message = H1: something\nH2: something
5390 the text was added as a single header line from Exim's point of view
5391 though it ended up OK in the delivered message. However, searching for the
5392 second and subsequent header lines using $h_h2: did not work. This has been
5393 fixed. Similarly, if a system filter added multiple headers in this way,
5394 the routers could not see them.
5396 20. Expanded the error message when iplsearch is called with an invalid key to
5397 suggest using net-iplsearch in a host list.
5399 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5400 ACLs are no longer actually imposed (and a message to that effect is
5403 22. If a "gecos" field in a passwd entry contained escaped characters, in
5404 particular, if it contained a \" sequence, Exim got it wrong when building
5405 a From: or a Sender: header from that name. A second bug also caused
5406 incorrect handling when an unquoted " was present following a character
5407 that needed quoting.
5409 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5410 was not being matched caselessly.
5412 24. Arranged for all hyphens in the exim.8 source to be escaped with
5415 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5416 properly. Recipient callout cache records were still being keyed to include
5417 the sender, even when use_sender was set false. This led to far more
5418 callouts that were necessary. The sender is no longer included in the key
5419 when use_sender is false.
5421 26. Added "control = submission" modifier to ACLs.
5423 27. Added the ${base62d: operator to decode base 62 numbers.
5425 28. dnsdb lookups can now access SRV records.
5427 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5428 the configuration file.
5430 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5431 (-v) mode. This makes the output for a verbose queue run more intelligible.
5433 31. Added a use_postmaster feature to recipient callouts.
5435 32. Added the $body_zerocount variable, containing the number of binary zero
5436 bytes in the message body.
5438 33. The time of last modification of the "new" subdirectory is now used as the
5439 "mailbox time last read" when there is a quota error for a maildir
5442 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5444 35. Added +ignore_unknown as a special item in host lists.
5446 36. Code for decoding IPv6 addresses in host lists is now included, even if
5447 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5448 address was recognized as an IP address, but was then not correctly decoded
5449 into binary, causing unexpected and incorrect effects when compared with
5456 1. Very minor rewording of debugging text in manualroute to say "list of
5457 hosts" instead of "hostlist".
5459 2. If verify=header_syntax was set, and a header line with an unqualified
5460 address (no domain) and a large number of spaces between the end of the
5461 name and the colon was received, the reception process suffered a buffer
5462 overflow, and (when I tested it) crashed. This was caused by some obsolete
5463 code that should have been removed. The fix is to remove it!
5465 3. When running in the test harness, delay a bit after writing a bounce
5466 message to get a bit more predictability in the log output.
5468 4. Added a call to search_tidyup() just before forking a reception process. In
5469 theory, someone could use a lookup in the expansion of smtp_accept_max_
5470 per_host which, without the tidyup, could leave open a database connection.
5472 5. Added the variables $recipient_data and $sender_data which get set from a
5473 lookup success in an ACL "recipients" or "senders" condition, or a router
5474 "senders" option, similar to $domain_data and $local_part_data.
5476 6. Moved the writing of debug_print from before to after the "senders" test
5479 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5480 problems for message scanning, either using a data ACL, or using
5481 local_scan() because the Received: header was not generated till after they
5482 were called (in order to set the time as the time of reception completion).
5483 I have revised the way this works. The header is now generated after the
5484 body is received, but before the ACL or local_scan() are called. After they
5485 are run, the timestamp in the header is updated.
5491 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5492 before starting a queue runner without re-exec. This happened only when
5493 deliver_drop_privilege was set or when the Exim user was set to root. The
5494 effect of the bug was that timeouts during subsequent deliveries caused
5495 crashes instead of being properly handled. The handler is now left at its
5496 default (and expected) setting.
5498 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5499 message, again when deliver_drop_privilege is set or Exim is run as root.
5500 The bug described in (1) was not present in this case, but the tidying up
5501 of the other signals was missing. I have made the two cases consistent.
5503 3. The ignore_target_hosts setting on a manualroute router was being ignored
5504 for hosts that were looked up using the /MX notation.
5506 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5509 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5510 operated on the sender address. After changing the $sender_address to <>
5511 for the sender address verify, Exim was re-instated it as the original
5512 (before rewriting) address, but remembering that it had rewritten it, so it
5513 wasn't rewriting it again. This bug also had the effect of breaking the
5514 sender address verification caching when the sender address was rewritten.
5516 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5517 This has been changed so that if the ip literal address matches
5518 ignore_target_hosts, the router declines.
5520 7. Added expansion conditions match_domain, match_address, and match_local_
5521 part (NOT match_host).
5523 8. The placeholder for the Received: header didn't have a length field set.
5525 9. Added code to Exim itself and to exim_lock to test for a specific race
5526 condition that could lead to file corruption when using MBX delivery. The
5527 issue is with the lockfile that is created in /tmp. If this file is removed
5528 after a process has opened it but before that process has acquired a lock,
5529 there is the potential for a second process to recreate the file and also
5530 acquire a lock. This could lead to two Exim processes writing to the file
5531 at the same time. The added code performs the same test as UW imapd; it
5532 checks after acquiring the lock that its file descriptor still refers to
5533 the same named file.
5535 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5536 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5539 11. Added the smtp_active_hostname option. If used, this will typically be made
5540 to depend on the incoming interface address. Because $interface_address is
5541 not set up until the daemon has forked a reception process, error responses
5542 that can happen earlier (such as "too many connections") no longer contain
5545 12. If an expansion in a condition on a "warn" statement fails because a lookup
5546 defers, the "warn" statement is abandoned, and the next ACL statement is
5547 processed. Previously this caused the whole ACL to be aborted.
5549 13. Added the iplsearch lookup type.
5551 14. Added ident_timeout as a log selector.
5553 15. Added tls_certificate_verified as a log selector.
5555 16. Added a global option tls_require_ciphers (compare the smtp transport
5556 option of the same name). This controls incoming TLS connections.
5558 17. I finally figured out how to make tls_require_ciphers do a similar thing
5559 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5560 before starting the TLS session.
5562 18. Tabs are now shown as \t in -bP output.
5564 19. If the log selector return_path_on_delivery was set, Exim crashed when
5565 bouncing a message because it had too many Received: header lines.
5567 20. If two routers both had headers_remove settings, and the first one included
5568 a superfluous trailing colon, the final name in the first list and the
5569 first name in the second list were incorrectly joined into one item (with a
5570 colon in the middle).
5576 1. Added -C and -D options to the exinext utility, mainly to make it easier
5577 to include in the automated testing, but these could be helpful when
5578 multiple configurations are in use.
5580 2. The exinext utility was not formatting the output nicely when there was
5581 an alternate port involved in the retry record key, nor when there was a
5582 message id as well (for retries that were specific to a specific message
5583 and a specific host). It was also confused by IPv6 addresses, because of
5584 the additional colons they contain. I have fixed the IPv4 problem, and
5585 patched it up to do a reasonable job for IPv6.
5587 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5588 delivery, the log line now contains "pipelined" if PIPELINING was used.
5590 4. An SMTP transport process used to panic and die if the bind() call to set
5591 an explicit outgoing interface failed. This has been changed; it is now
5592 treated in the same way as a connect() failure.
5594 5. A reference to $sender_host_name in the part of a conditional expansion
5595 that was being skipped was still causing a DNS lookup. This no longer
5598 6. The def: expansion condition was not recognizing references to header lines
5599 that used bh_ and bheader_.
5601 7. Added the _cache feature to named lists.
5603 8. The code for checking quota_filecount in the appendfile transport was
5604 allowing one more file than it should have been.
5606 9. For compatibility with Sendmail, the command line option
5614 and sets the incoming protocol and host name (for trusted callers). The
5615 host name and its colon can be omitted when only the protocol is to be set.
5616 Note the Exim already has two private options, -pd and -ps, that refer to
5617 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5618 "s", but I don't think that's a major issue.
5620 10. A number of refactoring changes to the code, none of which should affect
5623 (a) The number of logging options was getting close to filling up the
5624 32-bit word that was used as a bit map. I have split them into two classes:
5625 those that are passed in the argument to log_write(), and those that are
5626 only ever tested independently outside of that function. These are now in
5627 separate 32-bit words, so there is plenty of room for expansion again.
5628 There is no change in the user interface or the logging behaviour.
5630 (b) When building, for example, log lines, the code previously used a
5631 macro that called string_cat() twice, in order to add two strings. This is
5632 not really sufficiently general. Furthermore, there was one instance where
5633 it was actually wrong because one of the argument was used twice, and in
5634 one call a function was used. (As it happened, calling the function twice
5635 did not affect the overall behaviour.) The macro has been replaced by a
5636 function that can join an arbitrary number of extra strings onto a growing
5639 (c) The code for expansion conditions now uses a table and a binary chop
5640 instead of a serial search (which was left over from when there were very
5641 few conditions). Also, it now recognizes conditions like "pam" even when
5642 the relevant support is not compiled in: a suitably worded error message is
5643 given if an attempt is made to use such a condition.
5645 11. Added ${time_interval:xxxxx}.
5647 12. A bug was causing one of the ddress fields not to be passed back correctly
5648 from remote delivery subprocesses. The field in question was not being
5649 subsequently used, so this caused to problems in practice.
5651 13. Added new log selectors queue_time and deliver_time.
5653 14. Might have fixed a bug in maildirsizefile handling that threw up
5654 "unexpected character" debug warnings, and recalculated the data
5655 unnecessarily. In any case, I expanded the warning message to give more
5658 15. Added the message "Restricted characters in address" to the statements in
5659 the default ACL that block characters like @ and % in local parts.
5661 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5662 Three changes have been made:
5664 (a) There was a serious bug; a negative response to MAIL caused the whole
5665 recipient domain to be cached as invalid, thereby blocking all messages
5666 to all local parts at the same domain, from all senders. This bug has
5667 been fixed. The domain is no longer cached after a negative response to
5668 MAIL if the sender used is not empty.
5670 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5673 (c) A new callout option, "use_sender" has been added for people who want
5674 the modified behaviour.
5680 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5683 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5684 indeed breaks things for older releases.
5686 3. Added additional logging to the case where there is a problem reading data
5687 from a filter that is running in a subprocess using a pipe, in order to
5688 try to track down a specific problem.
5690 4. Testing facility fudge: when running in the test harness and attempting
5691 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5692 getting "No route to host". Convert this to a timeout.
5694 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5697 6. Some OS don't have socklen_t but use size_t instead. This affects the
5698 fifth argument of getsockopt() amongst other things. This is now
5699 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5700 can be set for individual OS. I have set it for SunOS5, OSF1, and
5701 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5702 some earlier ones do not.
5704 7. Change 4.30/15 was not doing the test caselessly.
5706 8. The standard form for an IPv6 address literal was being rejected by address
5707 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5708 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5709 this, as well as the form without the "IPv6" on the front (but only when
5710 address literals are enabled, of course).
5712 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5714 10. Exim crashed if a message with an empty sender address specified by -f
5715 encountered a router with an errors_to setting. This could be provoked only
5716 by a command such as
5720 where an empty string was supplied; "<>" did not hit this bug.
5722 11. Installed PCRE release 4.5.
5724 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5725 remained set. It is now erased.
5727 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5728 times from message ids (which are base 36 rather than the normal 62).
5730 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5731 were being counted as actual protocol errors, and logged if the log
5732 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5733 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5734 and DATA following a set of rejected RCPTs do not count as protocol errors.
5735 In other words, Exim assumes they were pipelined, though this may not
5736 actually be the case. Of course, in all cases the client gets an
5737 appropriate error code.
5739 15. If a lookup fails in an ACL condition, a message about the failure may
5740 be available; it is used if testing the ACL cannot continue, because most
5741 such messages specify what the cause of the deferral is. However, some
5742 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5743 that caused an old message to be retained and used if a later statement
5744 caused a defer, replacing the real cause of the deferral.
5746 16. If an IP address had so many PTR records that the DNS lookup buffer
5747 was not large enough to hold them, Exim could crash while trying to process
5748 the truncated data. It now detects and logs this case.
5750 17. Further to 4.21/58, another change has been made: if (and only if) the
5751 first line of a message (the first header line) ends with CRLF, a bare LF
5752 in a subsequent header line has a space inserted after it, so as not to
5753 terminate the header.
5755 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5756 unnecessarily, used atoi() instead of strtol(), and didn't check the
5757 termination when getting file sizes from file names by regex.
5759 19. Completely re-implemented the support for maildirsize files, in the light
5760 of a number of problems with the previous contributed implementation
5761 (4.30/29). In particular:
5763 . If the quota is zero, the maildirsize file is maintained, but no quota is
5766 . If the maildir directory does not exist, it is created before any attempt
5767 to write a maildirsize file.
5769 . The quota value in the file is just a cache; if the quota is changed in
5770 the transport, the new value overrides.
5772 . A regular expression is available for excluding directories from the
5775 20. The autoreply transport checks the characters in options that define the
5776 message's headers; it allows continued headers, but it was checking with
5777 isspace() after an embedded newline instead of explicitly looking for a
5780 21. If all the "regular" hosts to which an address was routed had passed their
5781 expiry times, and had not reached their retry times, the address was
5782 bounced, even if fallback hosts were defined. Now Exim should go on to try
5785 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5786 equivalent code in the SMTP transport. Some hosts send humungous responses
5787 to HELO/EHLO, more than 1024 it seems.
5789 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5790 gives compiler warnings in some environments. I've now done it "properly",
5793 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5794 (because of problems with the built-in one) was declared to return uschar *
5795 instead of char *, causing compiler failure.
5797 25. Fixed a file descriptor leak when processing alias/forward files.
5799 26. Fixed a minor format string issue in dbfn.c.
5801 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5803 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5804 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5807 29. When a sender address is verified, it is cached, to save repeating the test
5808 when there is more than one recipient in a message. However, when the
5809 verification involves a callout, it is possible for different callout
5810 options to be set for different recipients. It is too complicated to keep
5811 track of this in the cache, so now Exim always runs a verification when a
5812 callout is required, relying on the callout cache for the optimization.
5813 The overhead is duplication of the address routing, but this should not be
5816 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5817 to be verified with callout=postmaster, and the main callout worked but the
5818 postmaster check failed, the verification correctly failed. However, if a
5819 subsequent RCPT command asked for sender verification *without* the
5820 postmaster check, incorrect caching caused this verification also to fail,
5823 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5824 it was not caching the DNS options (qualify_single, search_parents) that
5825 were used when the lookup failed. A subsequent lookup with different
5826 options therefore always gave the same answer, though there were cases
5827 where it should not have. (Example: a "domains = !$mx_any" option on a
5828 dnslookup router: the "domains" option is always processed without any
5829 widening, but the router might have qualify_single set.) Now Exim uses the
5830 cached value only when the same options are set.
5832 32. Added John Jetmore's "exipick" utility to the distribution.
5834 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5835 than a timeout (e.g. a certificate is required, and is not provided), an
5836 Exim server now closes the connection immediately. Previously it waited for
5837 the client to close - but if the client is SSL, it seems that they each
5838 wait for each other, leading to a delay before one of them times out.
5840 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5841 maintained 0.8.x compatibility because I don't think many are using it, and
5842 it is clearly obsolete.
5844 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5847 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5848 client certificate was expired. A simple patch fixes this, though I don't
5849 understand the full logic of why the verify callback is called multiple
5852 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5853 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5854 which causes problems with some clients (such as the Certicom SSL Plus
5855 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5856 disables the coutermeasure allowing Eudora to connect."
5858 38. Exim was not checking that a write() to a log file succeeded. This could
5859 lead to Bad Things if a log got too big, in particular if it hit a file
5860 size limit. Exim now panics and dies if it cannot write to a log file, just
5861 as it does if it cannot open a log file.
5863 39. Modified OS/Makefile-Linux so that it now contains
5865 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5867 The two -D definitions ensure that Exim is compiled with large file
5868 support, which makes it possible to handle log files that are bigger than
5871 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5872 instance) a domain was checked against a named list that involved a lookup,
5873 causing $domain_data to be set, then another domain was checked against the
5874 same list, then the first domain was re-checked, the value of $domain_data
5875 after the final check could be wrong. In particular, if the second check
5876 failed, it could be set empty. This bug probably also applied to
5879 41. The strip_trailing_dot option was not being applied to the address given
5880 with the -f command-line option.
5882 42. The code for reading a message's header from the spool was incrementing
5883 $received_count, but never initializing it. This meant that the value was
5884 incorrect (doubled) while delivering a message in the same process in which
5885 it was received. In the most common configuration of Exim, this never
5886 happens - a fresh exec is done - but it can happen when
5887 deliver_drop_privilege is set.
5889 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5890 it now includes up to 150 characters of the unexpected data in the log
5893 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5894 and building data strings. The size of both of these buffers was 10 000
5895 bytes - far larger than anybody would *ever* want, thought I. Needless to
5896 say, somebody hit the limit. I have increased the maximum line length to
5897 20 000 and the maximum data length of concatenated lines to 100 000. I have
5898 also fixed two bugs, because there was no checking on these buffers. Tsk,
5899 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5900 buffer is too small.
5902 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5903 lsearch lookups. Now it does.
5905 46. When parsing a route_list item in a manualroute router, a fixed-length
5906 buffer was used for the list of hosts. I made this 1024 bytes long,
5907 thinking that nobody would ever have a list of hosts that long. Wrong.
5908 Somebody had a whole pile of complicated expansion conditions, and the
5909 string was silently truncated, leading to an expansion error. It turns out
5910 that it is easier to change to an unlimited length (owing to other changes
5911 that have happened since this code was originally written) than to build
5912 structure for giving a limitation error. The length of the item that
5913 expands into the list of hosts is now unlimited.
5915 47. The lsearch lookup could not handle data where the length of text line was
5916 more than 4095 characters. Such lines were truncated, leading to shortened
5917 data being returned. It should now handle lines of any length.
5919 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5920 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5923 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5924 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5925 no longer happens in comment lines. A list of replacements is now placed
5926 at the head of all of the source files, except those whose only change is
5927 to replace PERL_COMMAND in the very first #! line.
5929 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5930 messages on the queue, with a bottom-up merge sort, using code contributed
5931 by Michael Haardt. This should make operations like -bp somewhat faster on
5932 large queues. It won't affect queue runners, except when queue_run_in_order
5935 51. Installed eximstats 1.31 in the distribution.
5937 52. Added support for SRV lookups to the dnslookup router.
5939 53. If an ACL referred to $message_body or $message_body_end, the value was not
5940 reset for any messages that followed in the same SMTP session.
5942 54. The store-handling optimization for building very long strings was not
5943 differentiating between the different store pools. I don't think this
5944 actually made any difference in practice, but I've tidied it.
5946 55. While running the routers to verify a sender address, $sender_address
5947 was still set to the sender address. This is wrong, because when routing to
5948 send a bounce to the sender, it would be empty. Therefore, I have changed
5949 it so that, while verifying a sender address, $sender_address is set to <>.
5950 (There is no change to what happens when verifying a recipient address.)
5952 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5953 target A or AAAA records (if not already returned) without resetting the
5954 qualify_single or search_parents options of the DNS resolver. These are
5955 inappropriate in this case because the targets of MX and SRV records must
5956 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5957 target that, when qualified, matched something in the local domain. These
5958 two options are now turned off when doing these lookups.
5960 57. It seems that at least some releases of Reiserfs (which does not have the
5961 concept of a fixed number of inodes) returns zero and not -1 for the
5962 number of available inodes. This interacted badly with check_spool_inodes,
5963 which assumed that -1 was the "no such thing" setting. What I have done is
5964 to check that the total number of inodes is greater than zero before doing
5965 the test of how many are available.
5967 58. When a "warn" ACL statement has a log_message modifier, the message is
5968 remembered, and not repeated. This is to avoid a lot of repetition when a
5969 message has many recipients that cause the same warning to be written.
5970 However, Exim was preserving the list of already written lines for an
5971 entire SMTP session, which doesn't seem right. The memory is now reset if a
5972 new message is started.
5974 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5975 debugging output unless log_rewrite was also set.
5977 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5978 of (int)(handle) when we know that handle contains (void *)(-1).
5980 61. The Exim daemon panic-logs an error return when it closes the incoming
5981 connection. However "connection reset by peer" seems to be common, and
5982 isn't really an error worthy of noting specially, so that particular error
5985 62. When Exim is trying to find all the local interfaces, it used to panic and
5986 die if the ioctl to get the interface flags failed. However, it seems that
5987 on at least one OS (Solaris 9) it is possible to have an interface that is
5988 included in the list of interfaces, but for which you get a failure error
5989 for this call. This happens when the interface is not "plumbed" into a
5990 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5991 failure of the "get flags" call assumes that the interface is down.
5993 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5994 makes life easier for people who are doing arithmetic on fields extracted
5995 from dates, where you often get leading zeros that should not be
5996 interpreted as octal.
5998 64. Added qualify_domain to the redirect router, to override the global
6001 65. If a pathologically long header line contained very many addresses (the
6002 report of this problem mentioned 10 000) and each of them was rewritten,
6003 Exim could use up a very large amount of memory. (It kept on making new
6004 copies of the header line as it rewrote, and never released the old ones.)
6005 At the expense of a bit more processing, the header rewriting function has
6006 been changed so that it no longer eats memory in this way.
6008 66. The generation of the Received: header has been moved from the time that a
6009 message starts to be received, to the time that it finishes. The timestamp
6010 in the Received: header should now be very close to that of the <= log
6011 line. There are two side-effects of this change:
6013 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6014 logged header lines no longer include the local Received: line, because
6015 it has not yet been created. The same applies to a copy of the message
6016 that is returned to a non-SMTP sender when a message is rejected.
6018 (b) When a filter file is tested using -bf, no additional Received: header
6019 is added to the test message. After some thought, I decided that this
6022 This change does not affect the value of $received_for. It is still set
6023 after address rewriting, but before local_scan() is called.
6025 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6027 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6028 gave an unhelpful panic error message, and a defer error. I have managed to
6029 change this behaviour so that it now rejects any supplied certificate,
6030 which seems right, as the list of acceptable certificates is empty.
6032 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6033 gave an unhelpful defer error. I have not managed to make this reject any
6034 supplied certificates, but the error message it gives is "no certificate
6035 supplied", which is not helpful.
6037 70. exigrep's output now also includes lines that are not associated with any
6038 message, but which match the given pattern. Implemented by a patch from
6039 Martin Sluka, which also tidied up the Perl a bit.
6041 71. Recipient callout verification, like sender verification, was using <> in
6042 the MAIL FROM command. This isn't really the right thing, since the actual
6043 sender may affect whether the remote host accepts the recipient or not. I
6044 have changed it to use the actual sender in the callout; this means that
6045 the cache record is now keyed on a recipient/sender pair, not just the
6046 recipient address. There doesn't seem to be a real danger of callout loops,
6047 since a callout by the remote host to check the sender would use <>.
6048 [SEE ABOVE: changed after hitting problems.]
6050 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6051 temporary errors. However, in the case of such a code being given after
6052 the end of a data transmission (i.e. after ".") Exim was failing to write
6053 a retry record for the message. (Yes, there was some broken host that was
6054 actually sending 8xx at this point.)
6056 73. An unknown lookup type in a host list could cause Exim to panic-die when
6057 the list was checked. (An example that provoked this was putting <; in the
6058 middle of a list instead of at the start.) If this happened during a DATA
6059 ACL check, a -D file could be left lying around. This kind of configuration
6060 error no longer causes Exim to die; instead it causes a defer error. The
6061 incident is still logged to the main and panic logs.
6063 74. Buglet left over from Exim 3 conversion. The message "too many messages
6064 in one connection" was written to the rejectlog but not the mainlog, except
6065 when address rewriting (yes!) was being logged.
6067 75. Added write_rejectlog option.
6069 76. When a system filter was run not as root (that is, when system_filter_user
6070 was set), the values of the $n variables were not being returned to the
6071 main process; thus, they were not subsequently available in the $sn
6074 77. Added +return_path_on_delivery log selector.
6076 78. A connection timeout was being treated differently from recipients deferred
6077 when testing hosts_max_try with a message that was older than the host's
6078 retry timeout. (The host should not be counted, thus allowing all hosts to
6079 be tried at least once before bouncing.) This may have been the cause of an
6080 occasionally reported bug whereby a message would remain on the queue
6081 longer than the retry timeout, but would be bounced if a delivery was
6082 forced. I say "may" because I never totally pinned down the problem;
6083 setting up timeout/retry tests is difficult. See also the next item.
6085 79. The ultimate address timeout was not being applied to errors that involved
6086 a combination of host plus message (for example, a timeout on a MAIL
6087 command). When an address resolved to a number of possible hosts, and they
6088 were not all tried for each delivery (e.g. because of hosts_max_try), a
6089 message could remain on the queue longer than the retry timeout.
6091 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6094 81. Fixed an obscure SMTP outgoing bug which required at least the following
6095 conditions: (a) there was another message waiting for the same server;
6096 (b) the server returned 5xx to all RCPT commands in the first message so
6097 that the message was not completed; (c) the server dropped the connection
6098 or gave a negative response to the RSET that Exim sends to abort the
6099 transaction. The observed case was a dropped connection after DATA that had
6100 been sent in pipelining mode. That is, the server had advertised PIPELINING
6101 but was not implementing it correctly. The effect of the bug was incorrect
6102 behaviour, such as trying another host, and this could lead to a crash.
6108 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6109 and daemon.c were passed as pointers to ints; they should have been
6110 pointers to socklen_t variables (which are typically unsigned ints).
6112 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6115 3. Fixed a really odd bug that affected only the testing scheme; patching a
6116 certain fixed string in the binary changed the value of another string that
6117 happened to be identical to the end of the original first string.
6119 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6120 name", it returns that address as the IP address. On some operating
6121 systems (e.g. Solaris), it also passes back the IP address string as the
6122 "host name". However, on others (e.g. Linux), it passes back an empty
6123 string. Exim wasn't checking for this, and was changing the host name to an
6124 empty string, assuming it had been canonicalized.
6126 5. Although rare, it is permitted to have more than one PTR record for a given
6127 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6128 all the names associated with an address, because they do in Solaris.
6129 However, it seems that they do not in Linux for data that comes from the
6130 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6131 I found this out when I moved to a new Linux workstation and tried to run
6132 the Exim test suite.
6134 To get round this problem I have changed the code so that it now does its
6135 own call to the DNS to look up PTR records when searching for a host name.
6136 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6137 addresses that are only in /etc/hosts are still found.
6139 This behaviour is, however, controlled by an option called host_lookup_
6140 order, which defaults to "bydns:byaddr". If people want to use the other
6141 order, or indeed, just use one or the other means of lookup, they can
6142 specify it in this variable.
6144 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6145 some operating systems, this comes back from gethostbyaddr() as an empty
6146 string, and this is what Exim used to test for. However, it seems that in
6147 other systems, "." is yielded. Exim now tests for this case too.
6149 7. The values of check_spool_space and check_log_space are now held internally
6150 as a number of kilobytes instead of an absolute number of bytes. If a
6151 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6152 kilobyte. This means that much larger values can be stored.
6154 8. Exim monitor: an attempt to get the action menu when not actually pointing
6155 at a message produces an empty menu entitled "No message selected". This
6156 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6157 no entries in it ("Shell widget menu has zero width and/or height"). So I
6158 have added a single, blank menu entry in this case.
6160 9. Added ${quote_local_part.
6162 10. MIME decoding is now applied to the contents of Subject: header lines when
6165 11. Now that a reference to $sender_host_address automatically causes a reverse
6166 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6167 host lookup before query-style lookups in lists that might use this
6168 variable. This has therefore been abolished, and the "net-" prefix is no
6169 longer necessary for query-style lookups.
6171 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6172 have been a typo for LFLAGS, so it has been changed.
6174 13. The install script calls Exim with "-C /dev/null" in order to find the
6175 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6176 to be output. However, since Exim outputs its version number before the
6177 error, it didn't break the script. It just looked ugly. I fixed this by
6178 always allowing "-C /dev/null" if the caller is root.
6180 14. Ignore overlarge ACL variable number when reading spool file - insurance
6181 against a later release with more variables having written the file.
6183 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6184 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6185 this, as well as the form without the "IPv6" on the front.
6187 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6188 OS/Makefile-Darwin file.
6190 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6191 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6193 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6194 variable. However, some OS use ints for inodes. I've added cast to long int
6195 to get rid of the compiler warning.
6197 19. I had forgotten to lock out "/../" in configuration file names when
6198 ALT_CONFIG_PREFIX was set.
6200 20. Routers used for verification do not need to specify transports. However,
6201 if such a router generated a host list, and callout was configured, Exim
6202 crashed, because it could not find a port number from the (non-existent)
6203 transport. It now assumes port 25 in this circumstance.
6205 21. Added the -t option to exigrep.
6207 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6208 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6209 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6210 them is set, LOOKUP_LSEARCH is forced.
6212 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6214 24. Added sender and host information to the "rejected by local_scan()" log
6215 line; previously there was no indication of these.
6217 25. Added .include_if_exists.
6219 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6220 Linux. It turns out that not all file systems support this. Apparently some
6221 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6222 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6223 not supported on the file descriptor, is now ignored when Exim is trying to
6224 sync a directory. This applies only to Linux.
6226 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6228 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6229 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6232 29. Added support for maildirsize files from supplied patch (modified a bit).
6234 30. The use of :fail: followed by an empty string could lead Exim to respond to
6235 sender verification failures with (e.g.):
6237 550 Verification failed for <xxx>
6238 550 Sender verify failed
6240 where the first response line was missing the '-' that indicates it is not
6241 the final line of the response.
6243 31. The loop for finding the name of the user that called Exim had a hardwired
6244 limit of 10; it now uses the value of finduser_retries, which is used for
6245 all other user lookups.
6247 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6250 33. Exim was neglecting to zero errno before one call of strtol() when
6251 expanding a string and expecting an integer value. On some systems this
6252 resulted in spurious "integer overflow" errors. Also, it was casting the
6253 result into an int without checking.
6255 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6256 did not work. The code looks as if it has *never* worked, though it appears
6257 to have been documented since at least release 1.62. I have made it work.
6259 35. The "timeout_DNS" error in retry rules, also documented since at least
6260 1.62, also never worked. As it isn't clear exactly what this means, and
6261 clearly it isn't a major issue, I have abolished the feature by treating it
6262 as "timeout", and writing a warning to the main and panic logs.
6264 36. The display of retry rules for -brt wasn't always showing the error code
6267 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6268 timeout_connect_A, timeout_connect_MX.
6270 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6271 to the empty sender.
6273 39. The daemon was not analysing the content of -oX till after it had closed
6274 stderr and disconnected from the controlling terminal. This meant that any
6275 syntax errors were only noted on the panic log, and the return code from
6276 the command was 0. By re-arranging the code a little, I've made the
6277 decoding happen first, so such errors now appear on stderr, and the return
6278 code is 1. However, the actual setting up of the sockets still happens in
6279 the disconnected process, so errors there are still only recorded on the
6282 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6283 connections (as happens on some IP stacks) was logged at start up time as
6284 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6285 it from "IPv6 and IPv4", which means that two separate sockets are being
6288 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6289 says whether AF_INET or AF_INET6 was passed as an argument.
6291 42. Exiwhat output was messed up when time zones were included in log
6294 43. Exiwhat now gives more information about the daemon's listening ports,
6295 and whether -tls-on-connect was used.
6297 44. The "port" option of the smtp transport is now expanded.
6299 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6300 silently ignored. Now an error message is written to the main and panic
6303 46. There's a new ACL modifier called "logwrite" which writes to a log file
6304 as soon as it is encountered.
6306 47. Added $local_user_uid and $local_user_gid at routing time.
6308 48. Exim crashed when trying to verify a sender address that was being
6311 49. Exim was recognizing only a space character after ".include". It now also
6312 recognizes a tab character.
6314 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6315 extracting the relevant information from the specification. The man page no
6316 longer contains scrambled data for the -d option, and I've added a section
6317 at the front about calling Exim under different names.
6319 51. Added "extra_headers" argument to the "mail" command in filter files.
6321 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6324 53. Installed eximstats 1.29.
6326 54. Added transport_filter_timeout as a generic transport option.
6328 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6329 Cc: header lines. This was required by RFC 822, but it not required by RFC
6332 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6333 incoming messages that did not have them. Now it does so only if the
6334 message originates locally, that is, if there is no associated remote host
6335 address. When Resent- header lines are present, this applies to the Resent-
6336 lines rather than the non-Resent- lines.
6338 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6339 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6341 58. Messages for configuration errors now include the name of the main
6342 configuration file - useful now that there may be more than one file in a
6343 list (.included file names were always shown).
6345 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6346 for those rare installations that do not start the daemon as root or run it
6347 setuid root. I've cut out the call to initgroups() if the daemon is not
6350 60. The Exim user and group can now be bound into the binary as text strings
6351 that are looked up at the start of Exim's processing.
6353 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6355 62. Added $mailstore_basename variable.
6357 63. Installed patch to sieve.c from Michael Haardt.
6359 64. When Exim failed to open the panic log after failing to open the main log,
6360 the original message it was trying to log was written to stderr and debug
6361 output, but if they were not available (the usual case in production), it
6362 was lost. Now it is written to syslog before the two lines that record the
6363 failures to open the logs.
6365 65. Users' Exim filters run in subprocesses under the user's uid. It is
6366 possible for a "deliver" command or an alias in a "personal" command to
6367 provoke an address rewrite. If logging of address rewriting is configured,
6368 this fails because the process is not running as root or exim. There may be
6369 a better way of dealing with this, but for the moment (because 4.30 needs
6370 to be released), I have disabled address rewrite logging when running a
6371 filter in a non-root, non-exim process.
6377 1. The buildconfig auxiliary program wasn't quoting the value set for
6378 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6379 not defined. This bug was masked in 4.22 by the effect that was fixed in
6382 2. Some messages that were rejected after a message id was allocated were
6383 shown as "incomplete" by exigrep. It no longer does this for messages that
6384 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6386 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6387 have allow_domain_literals set, the ID did not get logged in the <= line.
6388 Domain literals are now always recognized in Message-ID: header lines.
6390 4. The first argument for a ${extract expansion item is the key name or field
6391 number. Leading and trailing spaces in this item were not being ignored,
6392 causing some misleading effects.
6394 5. When deliver_drop_privilege was set, single queue runner processes started
6395 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6396 same command in the process it spins off) were not dropping privilege.
6398 6. When the daemon running as "exim" started a queue runner, it always
6399 re-executed Exim in the spun-off process. This is a waste of effort when
6400 deliver_drop_privilege is set. The new process now just calls the
6401 queue-runner function directly.
6407 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6410 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6411 ignored. Though the use of -oP was forcing the writing of a pid file, it
6412 was always written to the default place.
6414 3. If the message "no IP address found for host xxxx" is generated during
6415 incoming verification, it is now followed by identification of the incoming
6416 connection (so you can more easily find what provoked it).
6418 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6420 5. Added some features to "harden" Exim a bit more against certain attacks:
6422 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6423 be put in Local/Makefile. This is like the never_users runtime option,
6424 but it cannot be overridden. The default setting is "root".
6426 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6427 prefix string with which any file named in a -C command line option
6430 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6431 is retained for -C and -D only if the caller of Exim is root. Without
6432 it, the exim user may also use -C and -D and retain privilege.
6434 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6435 command line option is disabled.
6437 6. Macro names set by the -D option must start with an upper case letter, just
6438 like macro names defined in the configuration file.
6440 7. Added "dereference=" facility to LDAP.
6442 8. Two instances of the typo "uknown" in the source files are fixed.
6444 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6445 the Configure-Makefile script screwed up while processing it.
6447 10. Incorporated PCRE 4.4.
6449 11. The SMTP synchronization check was not operating right at the start of an
6450 SMTP session. For example, it could not catch a HELO sent before the client
6451 waited for the greeting. There is now a check for outstanding input at the
6452 point when the greeting is written. Because of the duplex, asynchronous
6453 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6454 way, but not yet received, when the check is performed.
6456 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6457 on TCP/IP sockets, because this apparently causes some broken clients to
6460 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6461 unchanged) from the Cygwin maintainer.
6463 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6464 was supported instead of testing for mbx. Effectively a typo.
6466 15. The spa authenticator server code was not checking that the input it
6467 received was valid base64.
6469 16. The debug output line for the "set" modifier in ACLs was not showing the
6470 name of the variable that was being set.
6472 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6474 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6475 lookup on its own. Something else was needed to trigger the lookup. For
6476 example, a match in host_lookup or the need for a host name in a host list.
6477 Now, if $sender_host_name is referenced and the host name has not yet been
6478 looked up, a lookup is performed. If the lookup fails, the variable remains
6479 empty, and $host_lookup_failed is set to "1".
6481 19. Added "eqi" as a case-independent comparison operator.
6483 20. The saslauthd authentication condition could segfault if neither service
6484 nor realm was specified.
6486 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6487 error message that was logged was misleading, and incoming SMTP
6488 connections were dropped. The message is now more accurate, and temporary
6489 errors are given to SMTP connections.
6491 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6492 (except RSET) with a 421 error, until QUIT is received. However, it was
6493 failing to send a response to QUIT.
6495 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6496 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6497 ACL is now run after the helo_try_verify_hosts code.
6499 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6500 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6501 case-independent, so other case variants are also recognized). Apparently
6502 some systems use these upper case variants.
6504 25. If more than two messages were waiting for the same host, and a transport
6505 filter was specified for the transport, Exim sent two messages over the
6506 same TCP/IP connection, and then failed with "socket operation on non-
6507 socket" when it tried to send the third.
6509 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6511 27. The extern definition of crypt16() in expand.c was not being excluded when
6512 the OS had its own crypt16() function.
6514 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6515 as a preferred synonym for return_size_limit, both as an option and as an
6518 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6520 30. Changed the default configuration ACL to relax the local part checking rule
6521 for addresses that are not in any local domains. For these addresses,
6522 slashes and pipe symbols are allowed within local parts, but the sequence
6523 /../ is explicitly forbidden.
6525 31. SPA server authentication was not clearing the challenge buffer before
6528 32. log_message in a "warn" ACL statement was writing to the reject log as
6529 well as to the main log, which contradicts the documentation and doesn't
6530 seem right (because no rejection is happening). So I have stopped it.
6532 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6533 However, I am unable to do any testing of this.
6535 34. Fixed an infelicity in the appendfile transport. When checking directories
6536 for a mailbox, to see if any needed to be created, it was accidentally
6537 using path names with one or more superfluous leading slashes; tracing
6538 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6540 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6541 contents are added to the log line that is written for every discarded
6542 recipient. (Previously a log_message setting was ignored.)
6544 36. The ${quote: operator now quotes the string if it is empty.
6546 37. The install script runs exim in order to find its version number. If for
6547 some reason other than non-existence or emptiness, which it checks, it
6548 could not run './exim', it was installing it with an empty version number,
6549 i.e. as "exim-". This error state is now caught, and the installation is
6552 38. An argument was missing from the function that creates an error message
6553 when Exim fails to connect to the socket for saslauthd authentication.
6554 This could cause Exim to crash, or give a corrupted message.
6556 39. Added isip, isip4, and isip6 to ${if conditions.
6558 40. The ACL variables $acl_xx are now saved with the message, and can be
6559 accessed later in routers, transports, and filters.
6561 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6562 strings in the file are not string-expanded.
6564 42. If a MAIL command specified a SIZE value that was too large to fit into an
6565 int variable, the check against message_size_limit failed. Such values are
6566 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6567 day this will have to be increased, but I don't think I want to be around
6568 when emails are that large.
6575 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6576 iconv() is not standard in FreeBSD.
6578 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6579 IPv6 enabled. The observed symptom was a segmentation fault on return from
6580 the function os_common_find_running_interfaces() in src/os.c.
6582 3. In the check_special_case() function in daemon.c I had used "errno" as an
6583 argument name, which causes warnings on some systems. This was basically a
6584 typo, since it was named "eno" in the comments!
6586 4. The code that waits for the clock to tick (at a resolution of some fraction
6587 of a second) so as to ensure message-id uniqueness was always waiting for
6588 at least one whole tick, when it could have waited for less. [This is
6589 almost certainly not relevant at current processor speeds, where it is
6590 unlikely to ever wait at all. But we try to future-proof.]
6592 5. The function that sleeps for a time interval that includes fractions of a
6593 second contained a race. It did not block SIGALRM between setting the
6594 timer, and suspending (a couple of lines later). If the interval was short
6595 and the sigsuspend() was delayed until after it had expired, the suspension
6596 never ended. On busy systems this could lead to processes getting stuck for
6599 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6600 process, before it forks any delivery processes. The open lookup caching
6601 mechanism meant that the open file or database connection was passed into
6602 the delivery process. The problem was that delivery processes always tidy
6603 up cached lookup data. This could cause a problem for the next delivery
6604 process started by the queue runner, because the external queue runner
6605 process does not know about the closure. So the next delivery process
6606 still has data in the lookup cache. In the case of a file lookup, there was
6607 no problem because closing a file descriptor in a subprocess doesn't affect
6608 the parent. However, if the lookup was caching a connection to a database,
6609 the connection was closed, and the second delivery process was likely to
6610 see errors such as "PGSQL: query failed: server closed the connection
6611 unexpectedly". The problem has been fixed by closing all cached lookups
6612 in a queue runner before running a delivery process.
6614 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6615 seem to have the "const" qualifier which it has on other OS. I've
6618 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6619 *of the same type* (client or server) with the same public name that an
6620 error should be diagnosed.
6622 9. When Exim looked up a host name for an IP address, but failed to find the
6623 original IP address when looking up the host name (a safety check), it
6624 output the message "<ip address> does not match any IP for NULL", which was
6625 confusing, to say the least. The bug was that the host name should have
6626 appeared instead of "NULL".
6628 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6629 user that is built into the binary, and the -C or -D options is used, root
6630 privilege is dropped before the configuration file is read. In addition,
6631 logging is switched to stderr instead of the normal log files. If the
6632 configuration then re-defines the Exim user, the unprivileged environment
6633 is probably not what is expected, so Exim logs a panic warning message (but
6636 However, if deliver_drop_privilege is set, the unprivileged state may well
6637 be exactly what is intended, so the warning has been cut out in that case,
6638 and Exim is allowed to try to write to its normal log files.
6644 1. smtp_return_error_details was not giving details for temporary sender
6645 or receiver verification errors.
6647 2. Diagnose a configuration error if two authenticators have the same public
6650 3. Exim used not to create the message log file for a message until the first
6651 delivery attempt. This could be confusing when incoming messages were held
6652 for policy or load reasons. The message log file is now created at the time
6653 the message is received, and an initial "Received" line is written to it.
6655 4. The automatically generated man page for command line options had a minor
6656 bug that caused no ill effects; however, a more serious problem was that
6657 the procedure for building the man page automatically didn't always
6658 operate. Consequently, release 4.20 contains an out-of-date version. This
6659 shouldn't happen again.
6661 5. When building Exim with embedded Perl support, the script that builds the
6662 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6663 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6665 6. The freeze_tell option was not being used for messages that were frozen on
6666 arrival, either by an ACL or by local_scan().
6668 7. Added the smtp_incomplete_transaction log selector.
6670 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6671 was accepting AUTH without a new EHLO.
6673 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6674 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6677 10. Logging of TCP/IP connections (when configured) now happens in the main
6678 daemon process instead of the child process, so that the TCP/IP connection
6679 count is more accurate (but it can never be perfect).
6681 11. The use of "drop" in a nested ACL was not being handled correctly in the
6682 outer ACL. Now, if condition failure induced by the nested "drop" causes
6683 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6684 or "require"), the connection is dropped.
6686 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6687 that yield "discard" can now be used with an "accept" or a "discard" verb,
6688 but an error is generated for any others (because I can't see a useful way
6689 to define what should happen).
6691 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6692 are now processed in the same way as lines in the Exim configuration file.
6693 In particular, continuation lines are supported.
6695 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6697 15. Added -ti meaning -t -i.
6699 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6700 domains, and warn by logging if others are found.
6702 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6703 returned from GIFCONF when Exim is trying to find the list of interfaces on
6704 a host. The code in os.c has been modified to copy each ifreq to an aligned
6705 structure in all cases.
6707 Also, in some cases, the returned ifreq's were being copied to a 'struct
6708 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6709 means the last couple of bytes of an IPv6 address could be chopped if the
6710 ifreq contained only a normal sockaddr (14 bytes storage).
6712 18. Named domain lists were not supported in the hosts_treat_as_local option.
6713 An entry such as +xxxx was not recognized, and was treated as a literal
6716 19. Ensure that header lines added by a DATA ACL are included in the reject log
6717 if the ACL subsequently rejects the message.
6719 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6720 MD5 (which is deprecated).
6722 21. When testing a filter file using -bf, Exim was writing a message when it
6723 took the sender from a "From " line in the message, but it was not doing so
6724 when it took $return_path from a Return-Path: header line. It now does.
6726 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6727 with a valid header line field name (a series of printing characters
6728 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6730 23. Changed "disc" in the source to "disk" to conform to the documentation and
6731 the book and for uniformity.
6733 24. Ignore Sendmail's -Ooption=value command line item.
6735 25. When execve() failed while trying to run a command in a pipe transport,
6736 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6737 could be confused with a return value of 69 from the command itself. This
6738 has been changed to 127, the value the shell returns if it is asked to run
6739 a non-existent command. The wording for the related log line suggests a
6740 non-existent command as the problem.
6742 26. If received_header_text expands to an empty string, do not add a Received:
6743 header line to the message. (Well, it adds a token one on the spool, but
6744 marks it "old" so that it doesn't get used or transmitted.)
6746 27. Installed eximstats 1.28 (addition of -nt option).
6748 28. There was no check for failure on the call to getsockname() in the daemon
6749 code. This can fail if there is a shortage of resources on the system, with
6750 ENOMEM, for example. A temporary error is now given on failure.
6752 29. Contrary to the C standard, it seems that in some environments, the
6753 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6754 program. Exim now does this explicitly; it affects the formatting of
6755 timestamps using strftime().
6757 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6758 complaints. I've now initialized all the variables, to avoid this.
6760 32. Header lines added by a system filter were not being "seen" during
6761 transport-time rewrites.
6763 33. The info_callback() function passed to OpenSSL is set up with type void
6764 (*)(SSL *, int, int), as described somewhere. However, when calling the
6765 function (actually a macro) that sets it up, the type void(*)() is
6766 expected. I've put in a cast to prevent warnings from picky compilers.
6768 34. If a DNS black list lookup found a CNAME record, but there were no A
6769 records associated with the domain it pointed at, Exim crashed.
6771 35. If a DNS black list lookup returned more than one A record, Exim ignored
6772 all but the first. It now scans all returned addresses if a particular IP
6773 value is being sought. In this situation, the contents of the
6774 $dnslist_value variable are a list of all the addresses, separated by a
6777 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6778 to accept a host name and all its aliases if the forward lookup for any of
6779 them yielded the IP address of the incoming connection. Now it accepts only
6780 those names whose forward lookup yields the correct IP address. Any other
6781 names are discarded. This closes a loophole whereby a rogue DNS
6782 administrator could create reverse DNS records to break through a
6783 wildcarded host restriction in an ACL.
6785 37. If a user filter or a system filter that ran in a subprocess used any of
6786 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6787 the wrong values were passed to the pipe command ($thisaddress had the
6788 value of $0, $0 had the value of $1, etc). This bug was introduced by
6789 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6791 38. Improved the line breaking for long SMTP error messages from ACLs.
6792 Previously, if there was no break point between 40 and 75 characters, Exim
6793 left the rest of the message alone. Two changes have been made: (a) I've
6794 reduced the minimum length to 35 characters; (b) if it can't find a break
6795 point between 35 and 75 characters, it looks ahead and uses the first one
6796 that it finds. This may give the occasional overlong line, but at least the
6797 remaining text gets split now.
6799 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6800 file descriptors might be low, and that setting 1000 would always raise it.
6801 It turns out that in some environments, the limit is already over 1000 and
6802 that lowering it causes trouble. So now Exim takes care not to decrease it.
6804 40. When delivering a message, the value of $return_path is set to $sender_
6805 address at the start of routing (routers may change the value). By an
6806 oversight, this default was not being set up when an address was tested by
6807 -bt or -bv, which affected the outcome if any router or filter referred to
6810 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6811 log only when "message" or "log_message" are set. However, if one of the
6812 conditions was an address verification, or a call to a nested ACL, the
6813 messages generated by the underlying test were being passed through. This
6814 no longer happens. The underlying message is available in $acl_verify_
6815 message for both "message" and "log_message" expansions, so it can be
6816 passed through if needed.
6818 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6819 new expansion $bh_ to give the encoded byte string without charset
6820 translation. Translation happens only if iconv() is available; HAVE_ICONV
6821 indicates this at build time. HEADERS_CHARSET gives the charset to
6822 translate to; headers_charset can change it in the configuration, and
6823 "headers charset" can change it in an individual filter file.
6825 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6826 that creates RFC 2047 encoded "words" labels them as that charset instead
6827 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6828 expansion operator; (ii) when Exim creates a From: line for a local
6829 message; (iii) when a header line is rewritten to include a "phrase" part.
6831 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6832 buggy, causing it to skip the first lines of messages whose message ID
6833 ended in 'D'. This would not have bitten before Exim release 4.14, because
6834 message IDs were unlikely to end in 'D' before then. The effect was to have
6835 incorrect size information for certain domains.
6837 45. #include "config.h" was missing at the start of the crypt16.c module. This
6838 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6841 46. If there was a timeout during a "random" callout check, Exim treated it as
6842 a failure of the random address, and carried on sending RSET and the real
6843 address. If the delay was just some slowness somewhere, the response to the
6844 original RCPT would be taken as a response to RSET and so on, causing
6845 mayhem of various kinds.
6847 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6848 when I implemented it. It didn't allow for the fact that some option values
6849 may legitimately be negative (e.g. size_addition), and it didn't even do
6850 the right test for positive values.
6852 48. Domain names in DNS records are case-independent. Exim always looks them up
6853 in lower case. Some resolvers return domain names in exactly the case they
6854 appear in the zone file, that is, they may contain uppercase letters. Not
6855 all resolvers do this - some return always lower case. Exim was treating a
6856 change of case by a resolver as a change of domain, similar to a widening
6857 of a domain abbreviation. This triggered its re-routing code and so it was
6858 trying to route what was effectively the same domain again. This normally
6859 caused routing to fail (because the router wouldn't handle the domain
6860 twice). Now Exim checks for this case specially, and just changes the
6861 casing of the domain that it ultimately uses when it transmits the message
6864 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6867 50. If a filter generated a file delivery with a non-absolute name (possible if
6868 no home directory exists for the router), the forbid_file option was not
6871 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6872 the existing equality matching.
6874 52. Exim was using ints instead of ino_t variables in some places where it was
6875 dealing with inode numbers.
6877 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6878 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6879 called TMPDIR, and if it finds it is different, it changes its value.
6881 54. The smtp_printf() function is now made available to local_scan() so
6882 additional output lines can be written before returning. There is also an
6883 smtp_fflush() function to enable the detection of a dropped connection.
6884 The variables smtp_input and smtp_batched_input are exported to
6887 55. Changed the default runtime configuration: the message "Unknown user"
6888 has been removed from the ACL, and instead placed on the localuser router,
6889 using the cannot_route_message feature. This means that any verification
6890 failures that generate their own messages won't get overridden. Similarly,
6891 the "Unrouteable address" message that was in the ACL for unverifiable
6892 relay addresses has also been removed.
6894 56. Added hosts_avoid_esmtp to the smtp transport.
6896 57. The exicyclog script was not checking for the esoteric option
6897 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6898 will work only if exicyclog is run under the appropriate euid.
6900 58. Following a discussion on the list, the rules by which Exim recognises line
6901 endings on incoming messages have been changed. The -dropcr and drop_cr
6902 options are now no-ops, retained only for backwards compatibility. The
6903 following line terminators are recognized: LF CRLF CR. However, special
6904 processing applies to CR:
6906 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6907 nor a local message in the state where . is a terminator.
6909 (ii) If a bare CR is encountered in a header line, an extra space is added
6910 after the line terminator so as not to end the header. The reasoning
6911 behind this is that bare CRs in header lines are most likely either
6912 to be mistakes, or people trying to play silly games.
6914 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6915 was being incorrectly given as 18 bytes larger than it should have been.
6916 This is a VOB (very old bug).
6918 60. This may never have affected anything current, but just in case it has:
6919 When the local host is found other than at the start of a list of hosts,
6920 the local host, those with the same MX, and any that follow, are discarded.
6921 When the list in question was part of a longer list of hosts, the following
6922 hosts (not currently being processed) were also being discarded. This no
6923 longer happens. I'm not sure if this situation could ever has previously
6926 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6929 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6930 header. This is something that is recommended in a new Internet Draft, and
6931 is something that is documented as being done by Sendmail. There are two
6932 possible values. For messages generated by the autoreply transport, Exim
6935 Auto-Submitted: auto-replied
6937 whereas for all other generated messages (e.g. bounces) it adds
6939 Auto-Submitted: auto-generated
6941 63. The "personal" condition in filters now includes a test for the
6942 Auto-Submitted: header. If it contains the string "auto-" the message it
6943 not considered personal.
6945 64. Added rcpt_include_affixes as a generic transport option.
6947 65. Added queue_only_override (default true).
6949 66. Added the syslog_duplication option.
6951 67. If what should have been the first header line of a message consisted of
6952 a space followed by a colon, Exim was mis-interpreting it as a header line.
6953 It isn't of course - it is syntactically invalid and should therefore be
6954 treated as the start of the message body. The misbehaviour could have
6955 caused a number of strange effects, including loss of data in subsequent
6956 header lines, and spool format errors.
6958 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6959 client host had authenticated. This control can now be exercised by an ACL
6960 for more flexibility.
6962 69. By default, callouts do not happen when testing with -bh. There is now a
6963 variant, -bhc, which does actually run the callout code, including
6964 consulting and updating the callout cache.
6966 70. Added support for saslauthd authentication, courtesy of Alexander
6969 71. If statvfs() failed on the spool or log directories while checking their
6970 size for availability, Exim confusingly gave the error "space shortage".
6971 Furthermore, in debugging mode it crashed with a floating point exception.
6972 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6973 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6974 serious problem, Exim now writes to the main and panic logs when this
6975 happens, with details of the failure. It then refuses to accept the
6976 incoming message, giving the message "spool directory problem" or "log
6977 directory problem" with a 421 code for SMTP messages.
6979 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6980 0, 1, and 2 exist, because some OS complain for execs without them (see
6981 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6982 descriptors. However, the code omitted to check that the open succeeded,
6983 causing mysterious errors if for some reason the permissions on /dev/null
6984 got screwed. Now Exim writes a message to the main and panic logs, and
6985 bombs out if it can't open /dev/null.
6987 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6988 interact so that it is all more flexible. It is supposed to remain
6989 backwards compatible. Also added extra_local_interfaces.
6991 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6992 to bomb out with an assertion failure - to the client this appears as a
6993 connection drop. This problem occurs in the part of the code that was taken
6994 from the Samba project. Fortunately, the assertion is in a very simple
6995 function, so I have fixed this by reproducing the function inline in the
6996 one place where it is called, and arranging for authentication to fail
6997 instead of killing the process with assert().
6999 75. The SPA client code was not working when the server requested OEM rather
7000 than Unicode encoding.
7002 76. Added code to make require_files with a specific uid setting more usable in
7003 the case where statting the file as root fails - usually a non-root-mounted
7004 NFS file system. When this happens and the failure is EACCES, Exim now
7005 forks a subprocess and does the per-uid checking as the relevant uid.
7007 77. Added process_log_path.
7009 78. If log_file_path was not explicitly set, a setting of check_log_space or
7010 check_log_inodes was ignored.
7012 79. If a space check for the spool or log partitions fails, the incident is now
7013 logged. Of course, in the latter case the data may get lost...
7015 80. Added the %p formatting code to string_format() so that it can be used to
7016 print addresses in debug_print(). Adjusted all the address printing in the
7017 debugging in store.c to use %p rather than %d.
7019 81. There was a concern that a line of code in smtp_in.c could overflow a
7020 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7021 initially expressed, the concern was not well-founded, because trailing
7022 spaces are removed early. However, if the trailing spaces were followed by
7023 a NULL, they did not get removed, so the overflow was possible. Two fixes
7026 (a) I re-wrote the offending code in a cleaner fashion.
7027 (b) If an incoming SMTP command contains a NULL character, it is rejected
7030 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7031 runs initgroups(), so that if the Exim user is in any additional groups,
7032 they will be used during message reception.
7038 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob