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.
62 JH/01 Rework error string handling in TLS interface so that the caller in
63 more cases is responsible for logging. This permits library-sourced
64 string to be attached to addresses during delivery, and collapses
65 pairs of long lines into single ones.
67 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
68 during configuration. Wildcards are allowed and expanded.
70 JH/02 Rework error string handling in DKIM to pass more info back to callers.
71 This permits better logging.
73 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
74 do not close it down and have the child transport start it up again on
75 the passed-on TCP connection. Instead, proxy the child (and any
76 subsequent ones) for TLS via a unix-domain socket channel. Logging is
77 affected: the continued delivery log lines do not have any DNSSEC, TLS
78 Certificate or OCSP information. TLS cipher information is still logged.
80 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
81 identical IP addresses on different listening ports. Will also affect
84 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
85 add noisy ifdef guards to special-case this sillyness.
86 Patch from Bernd Kuhls.
88 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
89 than 255 are no longer allowed.
91 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
92 Disable the session-cache too, which might reduce our load. Since we
93 currrectly use a new context for every connection, both as server and
94 client, there is no benefit for these.
95 GnuTLS appears to not support tickets server-side by default (we don't
96 call gnutls_session_ticket_enable_server()) but client side is enabled
97 by default on recent versions (3.1.3 +) unless the PFS priority string
100 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
101 <https://reproducible-builds.org/specs/source-date-epoch/>.
103 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
104 the check for any unsuccessful recipients did not notice the limit, and
105 erroneously found still-pending ones.
107 JH/08 Pipeline CHUNKING command and data together, on kernels that support
108 MSG_MORE. Only in-clear (not on TLS connections).
110 JH/09 Avoid using a temporary file during transport using dkim. Unless a
111 transport-filter is involved we can buffer the headers in memory for
112 creating the signature, and read the spool data file once for the
113 signature and again for transmission.
115 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
116 4.77 as the kernel support then wasn't solid, having issues in 64bit
117 mode. Now, it's been long enough. Add support for FreeBSD also.
119 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
120 case where the routing stage had gathered several addresses to send to
121 a host before calling the transport for the first, we previously failed
122 to close down TLS in the old transport process before passing the TCP
123 connection to the new process. The new one sent a STARTTLS command
124 which naturally failed, giving a failed delivery and bloating the retry
125 database. Investigation and fix prototype from Wolfgang Breyha.
127 JH/12 Fix check on SMTP command input synchronisation. Previously there were
128 false-negatives in the check that the sender had not preempted a response
129 or prompt from Exim (running as a server), due to that code's lack of
130 awareness of the SMTP input buffering.
132 PP/04 Add commandline_checks_require_admin option.
133 Exim drops privileges sanely, various checks such as -be aren't a
134 security problem, as long as you trust local users with access to their
135 own account. When invoked by services which pass untrusted data to
136 Exim, this might be an issue. Set this option in main configuration
137 AND make fixes to the calling application, such as using `--` to stop
140 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
141 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
142 responses to those, into a single TLS record each way (this usually means
143 a single packet). As a side issue, smtp_enforce_sync now works on TLS
146 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
147 affects you only if you're dancing at the edge of the param size limits.
148 If you are, and this message makes sense to you, then: raise the
149 configured limit or use OpenSSL 1.1. Nothing we can do for older
152 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
153 cmdline element to get the documented default one. Previously it was
156 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
157 are unsuable for non-message-specific reasons - eg. connection timeout,
158 banner-time rejection.
160 JH/16 Fix logging of delivery remote port, when specified by router, under
163 PP/06 Repair manualroute's ability to take options in any order, even if one
164 is the name of a transport.
167 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
169 JH/17 Change the list-building routines interface to use the expanding-string
170 triplet model, for better allocation and copying behaviour.
172 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
173 Previously it was constructed the first time a possibly-matching string
174 was met in the configuration file input during startup; now it is done
177 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
178 compatible one, to avoid the (poorly documented) possibility of a config
179 file in the working directory redirecting the DB files, possibly correpting
180 some existing file. CVE-2017-10140 assigned for BDB.
182 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
183 cache-hot. Previously, although the result was properly cached, the
184 initial verify call returned a defer.
186 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
187 the main verify for receipient in uncached-mode.
189 JH/22 Retire historical build files to an "unsupported" subdir. These are
190 defined as "ones for which we have no current evidence of testing".
192 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
193 if present. Previously it was ignored.
195 JH/24 Start using specified-initialisers in C structure init coding. This is
196 a C99 feature (it's 2017, so now considered safe).
198 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
199 if was a fixed-sized field and bitmask ops via macros; it is now more
202 PP/07 GitHub PR 56: Apply MariaDB build fix.
203 Patch provided by Jaroslav Škarvada.
205 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
206 during Coverity cleanups [4.87 JH/47]
207 Diagnosis and fix provided by Michael Fischer v. Mollard.
209 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
210 the right size to place the terminating semicolon on its own folded
211 line, the header hash was calculated to an incorrect value thanks to
212 the (relaxed) space the fold became.
214 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
215 and confused the parent.
217 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
218 which could crash as a result. This could lead to undeliverable messages.
220 JH/28 Logging: "next input sent too soon" now shows where input was truncated
223 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
224 matters on fast-turnover and PID-randomising systems, which were getting
225 out-of-order delivery.
227 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
228 a possibly-overlapping copy. The symptom was that "Remote host closed
229 connection in response to HELO" was logged instead of the actual 4xx
232 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
233 Previously only that bufferd was discarded, resulting in SYMTP command
236 JH/32 DKIM: when a message has multiple signatures matching an identity given
237 in dkim_verify_signers, run the dkim acl once for each. Previously only
238 one run was done. Bug 2189.
240 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
241 "panic the current process" to "deliberately defer". The panic log is
242 still written with the problem list name; the mail and reject logs now
243 get a temp-reject line for the message that was being handled, saying
244 something like "domains check lookup or other defer". The SMTP 451
245 message is still "Temporary local problem".
247 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
248 A crafted sequence of BDAT commands could result in in-use memory beeing
249 freed. CVE-2017-16943.
251 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
252 from SMTP input. Previously it was always done; now only done for DATA
253 and not BDAT commands. CVE-2017-16944.
255 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
256 to the message (such as an overlong header line). Previously this was
257 not done and we did not exit BDAT mode. Followon from the previous item
258 though a different problem.
264 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
265 than -2003 did; needs libidn2 in addition to libidn.
267 JH/02 The path option on a pipe transport is now expanded before use.
269 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
270 Patch provided by "Björn", documentation fix added too.
272 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
273 missing a wire-to-host endian conversion.
275 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
276 close after a BDAT command line could be taken as a following command,
277 giving a synch failure. Fix by only checking for synch immediately
278 before acknowledging the chunk.
280 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
281 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
282 macro. Patches provided by Josh Soref.
284 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
285 Previously we did not; the RFC seems ambiguous and VRFY is not listed
286 by IANA as a service extension. However, John Klensin suggests that we
289 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
290 the dkim code may be unix-mode line endings rather than smtp wire-format
291 CRLF, so prepend a CR to any bare LF.
293 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
294 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
296 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
297 the first were themselves being wrongly included in the feed into dkim
298 processing; with most chunk sizes in use this resulted in an incorrect
299 body hash calculated value.
301 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
302 DKIM signature block, for verification. Although advised against by
303 standards it is specifically not ruled illegal.
305 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
307 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
308 missing a body hash (the bh= tag).
310 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
311 It seems that HAProxy sends the Proxy Protocol information in clear and
312 only then does a TLS startup, so do the same.
314 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
315 TCP connections (such as for Spamd) unless the daemon successfully set
316 Fast Open mode on its listening sockets. This fixes breakage seen on
317 too-old kernels or those not configured for Fast Open, at the cost of
318 requiring both directions being enabled for TFO, and TFO never being used
319 by non-daemon-related Exim processes.
321 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
322 endings, at least on the first header line. Try to canonify any that get
323 past that check, despite the cost.
325 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
326 now limited to an arbitrary five deep, while parsing addresses with the
327 strip_excess_angle_brackets option enabled.
329 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
330 instead leave the unprompted TLS handshake in socket buffer for the
331 TLS library to consume.
333 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
335 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
337 JH/16 Drop variables when they go out of scope. Memory management drops a whole
338 region in one operation, for speed, and this leaves assigned pointers
339 dangling. Add checks run only under the testsuite which checks all
340 variables at a store-reset and panics on a dangling pointer; add code
341 explicitly nulling out all the variables discovered. Fixes one known
342 bug: a transport crash, where a dangling pointer for $sending_ip_address
343 originally assigned in a verify callout, is re-used.
345 PP/06 Drop '.' from @INC in various Perl scripts.
347 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
349 PP/08 Reduce a number of compilation warnings under clang; building with
350 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
351 should be warning-free.
353 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
355 HS/01 Fix portability problems introduced by PP/08 for platforms where
356 realloc(NULL) is not equivalent to malloc() [SunOS et al].
358 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
359 chunk. This allows us to accept broken chunked messages. We need a more
360 general solution here.
362 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
363 already-broken messages in the queue.
365 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
367 JH/19 Fix reference counting bug in routing-generated-address tracking.
373 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
374 supports it and a size is available (ie. the sending peer gave us one).
376 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
377 years of being deprecated). The replacements are the ACLs
378 acl_smtp_mime and acl_not_smtp_mime.
380 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
381 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
382 or even an in-clear connection were permitted. Now, if the host lookup
383 was dnssec and dane was requested then the host is only used if the
384 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
385 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
386 if one fails this test.
387 This means that a poorly-configured remote DNS will make it incommunicado;
388 but it protects against a DNS-interception attack on it.
390 JH/04 Bug 1810: make continued-use of an open smtp transport connection
391 non-noisy when a race steals the message being considered.
393 JH/05 If main configuration option tls_certificate is unset, generate a
394 self-signed certificate for inbound TLS connections.
396 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
397 in rewrites and routers.
399 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
400 and logged a warning sing 4.83; now they are a configuration file error.
402 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
403 (lacking @domain). Apply the same qualification processing as RCPT.
405 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
407 JH/10 Support ${sha256:} applied to a string (as well as the previous
410 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
411 a cutthrough deliver is pending, as we always want to make a connection.
412 This also avoids re-routing the message when later placing the cutthrough
413 connection after a verify cache hit.
414 Do not update it with the verify result either.
416 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
417 when routing results in more than one destination address.
419 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
420 signing (which inhibits the cutthrough capability). Previously only
421 the presence of an option was tested; now an expansion evaluating as
422 empty is permissible (obviously it should depend only on data available
423 when the cutthrough connection is made).
425 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
426 the relevant preceding SMTP command did not note the pipelining mode.
428 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
429 Previously they were not counted.
431 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
432 as one having no matching records. Previously we deferred the message
433 that needed the lookup.
435 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
436 distinguished as "(=".
438 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
439 for missing MX records. Previously it only worked for missing A records.
441 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
443 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
444 after the data-go-ahead and data-ack. Patch from Jason Betts.
446 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
447 even for a "none" policy. Patch from Tony Meyer.
449 JH/22 Fix continued use of a connection for further deliveries. If a port was
450 specified by a router, it must also match for the delivery to be
453 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
454 When one of the recipients of a message was unsuitable for the connection
455 (has no matching addresses), we lost track of needing to mark it
456 deferred. As a result mail would be lost.
458 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
460 JH/25 Decoding ACL controls is now done using a binary search; the source code
461 takes up less space and should be simpler to maintain. Merge the ACL
462 condition decode tables also, with similar effect.
464 JH/26 Fix problem with one_time used on a redirect router which returned the
465 parent address unchanged. A retry would see the parent address marked as
466 delivered, so not attempt the (identical) child. As a result mail would
469 JH/27 Fix a possible security hole, wherein a process operating with the Exim
470 UID can gain a root shell. Credit to http://www.halfdog.net/ for
471 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
474 JH/28 Enable {spool,log} filesystem space and inode checks as default.
475 Main config options check_{log,spool}_{inodes,space} are now
476 100 inodes, 10MB unless set otherwise in the configuration.
478 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
479 Previously it only applied to the main-section connection policy
482 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
484 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
485 by me. Added RFC7919 DH primes as an alternative.
487 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
488 are not in the system include path.
490 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
491 GnuTLS, when a session startup failed (eg because the client disconnected)
492 Exim did stdio operations after fclose. This was exposed by a recent
493 change which nulled out the file handle after the fclose.
495 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
496 signed directly by the cert-signing cert, rather than an intermediate
497 OCSP-signing cert. This is the model used by LetsEncrypt.
499 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
501 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
502 an incoming connection.
504 HS/02 Bug 1802: Do not half-close the connection after sending a request
507 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
508 fallback to "prime256v1".
510 JH/34 SECURITY: Use proper copy of DATA command in error message.
511 Could leak key material. Remotely exploitable. CVE-2016-9963.
517 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
518 and 3.4.4 - once the server is enabled to respond to an OCSP request
519 it does even when not requested, resulting in a stapling non-aware
520 client dropping the TLS connection.
522 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
523 support variable-length bit vectors. No functional change.
525 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
526 The I= interface field on outgoing lines is now after the H= remote
527 host field, same as incoming lines. There is a separate
528 outgoing_interface log selector which allows you to disable the
531 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
532 If not running log_selector +smtp_connection the mainlog would be held
533 open indefinitely after a "too many connections" event, including to a
534 deleted file after a log rotate. Leave the per net connection logging
535 leaving it open for efficiency as that will be quickly detected by the
536 check on the next write.
538 HS/01 Bug 1671: Fix post transport crash.
539 Processing the wait-<transport> messages could crash the delivery
540 process if the message IDs didn't exist for some reason. When
541 using 'split_spool_directory=yes' the construction of the spool
542 file name failed already, exposing the same netto behaviour.
544 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
545 mime_regex ACL conditions.
547 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
548 to DSN fail messages (bounces): remote IP, remote greeting, remote response
549 to HELO, local diagnostic string.
551 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
552 log line to debug. Even when configured with a tls authenticator many
553 client connections are expected to not authenticate in this way, so
554 an authenticate fail is not an error.
556 HS/02 Add the Exim version string to the process info. This way exiwhat
557 gives some more detail about the running daemon.
559 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
560 matter for fast-change records such as DNSBLs.
562 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
563 retry record, even if constant. There may be multiple transports with
564 different interface settings and the retry behaviour needs to be kept
567 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
569 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
571 JH/10 Bug 840: fix log_defer_output option of pipe transport
573 JH/11 Bug 830: use same host for all RCPTS of a message, even under
574 hosts_randomize. This matters a lot when combined with mua_wrapper.
576 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
577 ${quote_pgsql:<string>} operator.
579 JH/13 Bug 1708: avoid misaligned access in cached lookup.
581 JH/14 Change header file name for freeradius-client. Relevant if compiling
582 with Radius support; from the Gentoo tree and checked under Fedora.
584 JH/15 Bug 1712: Introduce $prdr_requested flag variable
586 JH/16 Bug 1714: Permit an empty string as expansion result for transport
587 option transport_filter, meaning no filtering.
589 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
591 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
592 defaults to "*" (all hosts). The variable is now available when not built
593 with TLS, default unset, mainly to enable keeping the testsuite sane.
594 If a server certificate is not supplied (via tls_certificate) an error is
595 logged, and clients will find TLS connections fail on startup. Presumably
596 they will retry in-clear.
597 Packagers of Exim are strongly encouraged to create a server certificate
598 at installation time.
600 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
601 with the $config_file variable.
603 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
604 in transport context, after the attempt, and per-recipient. The latter type
605 is per host attempted. The event data is the error message, and the errno
606 information encodes the lookup type (A vs. MX) used for the (first) host,
607 and the trailing two digits of the smtp 4xx response.
609 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
610 to write to mainlog (or rejectlog, paniclog) in the window between file
611 creation and permissions/ownership being changed. Particularly affects
612 installations where exicyclog is run as root, rather than exim user;
613 result is that the running daemon panics and dies.
615 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
617 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
618 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
619 "pri" and "weight". Note that the previous implicit priority given by the
620 list order is no longer honoured.
622 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
625 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
626 by defining SUPPORT_SOCKS.
628 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
629 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
630 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
631 variables are renamed to proxy_{local,external}_{address,port}.
633 JH/27 Move Internationalisation support from Experimental to mainline, enabled
634 for a build by defining SUPPORT_I18N
636 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
637 of the query string, and make ${quote_redis:} do that quoting.
639 JH/29 Move Events support from Experimental to mainline, enabled by default
640 and removable for a build by defining DISABLE_EVENT.
642 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
644 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
645 cached by the daemon.
647 JH/32 Move Redis support from Experimental to mainline, enabled for a build
648 by defining LOOKUP_REDIS. The libhiredis library is required.
650 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
651 keys are given for lookup.
653 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
654 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
655 only supported when built with TLS support. The PolarSSL SHA routines
656 are still used when the TLS library is too old for convenient support.
658 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
659 openssl_options), for security. OpenSSL forces this from version 1.1.0
660 server-side so match that on older versions.
662 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
663 allocation for $value could be released as the expansion processing
664 concluded, but leaving the global pointer active for it.
666 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
667 and to use the domains and local_parts ACL conditions.
669 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
670 incorrectly not doubled on cutthrough transmission, hence seen as a
671 body-termination at the receiving system - resulting in truncated mails.
672 Commonly the sender saw a TCP-level error, and retransmitted the message
673 via the normal store-and-forward channel. This could result in duplicates
674 received - but deduplicating mailstores were liable to retain only the
675 initial truncated version.
677 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
679 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
681 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
682 we're in there, support oversigning also; bug 1309.
684 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
686 HS/04 Add support for keep_environment and add_environment options.
688 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
689 either intentional arithmetic overflow during PRNG, or testing config-
692 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
693 delivery resulted in actual delivery. Cancel cutthrough before DATA
696 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
697 rejects a recipient: pass the reject to the originator.
699 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
700 Many were false-positives and ignorable, but it's worth fixing the
703 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
704 for the new environment-manipulation done at startup. Move the routines
705 from being local to tls.c to being global via the os.c file.
707 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
708 an extract embedded as result-arg for a map, the first arg for extract
709 is unavailable so we cannot tell if this is a numbered or keyed
710 extraction. Accept either.
716 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
719 JH/02 The smtp transport option "multi_domain" is now expanded.
721 JH/03 The smtp transport now requests PRDR by default, if the server offers
724 JH/04 Certificate name checking on server certificates, when exim is a client,
725 is now done by default. The transport option tls_verify_cert_hostnames
726 can be used to disable this per-host. The build option
727 EXPERIMENTAL_CERTNAMES is withdrawn.
729 JH/05 The value of the tls_verify_certificates smtp transport and main options
730 default to the word "system" to access the system default CA bundle.
731 For GnuTLS, only version 3.0.20 or later.
733 JH/06 Verification of the server certificate for a TLS connection is now tried
734 (but not required) by default. The verification status is now logged by
735 default, for both outbound TLS and client-certificate supplying inbound
738 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
741 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
742 Status Notification (bounce) messages are now MIME format per RFC 3464.
743 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
744 under the control of the dsn_advertise_hosts option, and routers may
745 have a dsn_lasthop option.
747 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
748 default, modifiable by a malware= option. The list separator for
749 the options can now be changed in the usual way. Bug 68.
751 JH/10 The smtp_receive_timeout main option is now expanded before use.
753 JH/11 The incoming_interface log option now also enables logging of the
754 local interface on delivery outgoing connections.
756 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
757 if the interface and destination host and port all match.
759 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
762 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
763 Patch from Andrew Lewis.
765 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
766 now supports optional time-restrictions, weighting, and priority
767 modifiers per server. Patch originally by <rommer@active.by>.
769 JH/16 The spamd_address main option now supports a mixed list of local
770 and remote servers. Remote servers can be IPv6 addresses, and
771 specify a port-range.
773 JH/17 Bug 68: The spamd_address main option now supports an optional
774 timeout value per server.
776 JH/18 Bug 1581: Router and transport options headers_add/remove can
777 now have the list separator specified.
779 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
782 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
785 JH/21 Support for the A6 type of dns record is withdrawn.
787 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
788 rather than the verbs used.
790 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
791 from 255 to 1024 chars.
793 JH/24 Verification callouts now attempt to use TLS by default.
795 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
796 are generic router options now. The defaults didn't change.
798 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
799 Original patch from Alexander Shikoff, worked over by JH.
801 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
802 files if ZCAT_COMMAND is not executable.
804 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
806 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
808 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
809 Normally benign, it bites when the pair was led to by a CNAME;
810 modern usage is to not canonicalize the domain to a CNAME target
811 (and we were inconsistent anyway for A-only vs AAAA+A).
813 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
815 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
816 when evaluating $sender_host_dnssec.
818 JH/31 Check the HELO verification lookup for DNSSEC, adding new
819 $sender_helo_dnssec variable.
821 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
823 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
825 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
827 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
828 documented as working, but never had. Support all but $spam_report.
830 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
831 added for tls authenticator.
833 HS/03 Add perl_taintmode main config option
839 TL/01 When running the test suite, the README says that variables such as
840 no_msglog_check are global and can be placed anywhere in a specific
841 test's script, however it was observed that placement needed to be near
842 the beginning for it to behave that way. Changed the runtest perl
843 script to read through the entire script once to detect and set these
844 variables, reset to the beginning of the script, and then run through
845 the script parsing/test process like normal.
847 TL/02 The BSD's have an arc4random API. One of the functions to induce
848 adding randomness was arc4random_stir(), but it has been removed in
849 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
850 function when detected.
852 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
853 cause callback expansion.
855 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
856 syntax errors in an expansion can be treated as a string instead of
857 logging or causing an error, due to the internal use of bool_lax
858 instead of bool when processing it.
860 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
861 server certificates when making smtp deliveries.
863 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
865 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
867 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
869 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
870 Merged patch from Sebastian Wiedenroth.
872 JH/05 Fix results-pipe from transport process. Several recipients, combined
873 with certificate use, exposed issues where response data items split
874 over buffer boundaries were not parsed properly. This eventually
875 resulted in duplicates being sent. This issue only became common enough
876 to notice due to the introduction of connection certificate information,
877 the item size being so much larger. Found and fixed by Wolfgang Breyha.
879 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
880 size buffer was used, resulting in syntax errors when an expansion
883 JH/07 Add support for directories of certificates when compiled with a GnuTLS
884 version 3.3.6 or later.
886 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
887 is EXPERIMENTAL_EVENT, the main-configuration and transport options
888 both become "event_action", the variables become $event_name, $event_data
889 and $event_defer_errno. There is a new variable $verify_mode, usable in
890 routers, transports and related events. The tls:cert event is now also
891 raised for inbound connections, if the main configuration event_action
894 TL/06 In test suite, disable OCSP for old versions of openssl which contained
895 early OCSP support, but no stapling (appears to be less than 1.0.0).
897 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
898 server certificate names available under the smtp transport option
899 "tls_verify_cert_hostname" now do not permit multi-component wildcard
902 JH/10 Time-related extraction expansions from certificates now use the main
903 option "timezone" setting for output formatting, and are consistent
904 between OpenSSL and GnuTLS compilations. Bug 1541.
906 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
907 encoded parameter in the incoming message. Bug 1558.
909 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
910 include certificate info, eximon was claiming there were spoolfile
913 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
915 JH/14 Log delivery-related information more consistently, using the sequence
916 "H=<name> [<ip>]" wherever possible.
918 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
919 are problematic for Debian distribution, omit them from the release
922 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
924 JH/16 Fix string representation of time values on 64bit time_t architectures.
927 JH/17 Fix a null-indirection in certextract expansions when a nondefault
928 output list separator was used.
933 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
934 checkers that were complaining about end of non-void function with no
937 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
938 This was a regression introduced in 4.83 by another bugfix.
940 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
942 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
943 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
949 TF/01 Correctly close the server side of TLS when forking for delivery.
951 When a message was received over SMTP with TLS, Exim failed to clear up
952 the incoming connection properly after forking off the child process to
953 deliver the message. In some situations the subsequent outgoing
954 delivery connection happened to have the same fd number as the incoming
955 connection previously had. Exim would try to use TLS and fail, logging
956 a "Bad file descriptor" error.
958 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
959 utilities have not been installed.
961 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
962 temporary space as the ACL may create new global variables.
964 TL/01 LDAP support uses per connection or global context settings, depending
965 upon the detected version of the libraries at build time.
967 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
968 to extract and use the src ip:port in logging and expansions as if it
969 were a direct connection from the outside internet. PPv2 support was
970 updated based on HAProxy spec change in May 2014.
972 JH/02 Add ${listextract {number}{list}{success}{fail}}.
974 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
975 Properly escape header and check for NULL return.
977 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
980 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
982 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
983 characters in header names, implemented as a verify condition.
984 Contributed by Michael Fischer v. Mollard.
986 TL/05 Rename SPF condition results err_perm and err_temp to standardized
987 results permerror and temperror. Previous values are deprecated but
988 still accepted. In a future release, err_perm and err_temp will be
989 completely removed, which will be a backward incompatibility if the
990 ACL tests for either of these two old results. Patch contributed by
991 user bes-internal on the mailing list.
993 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
995 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
996 selectors, in both main and reject logs.
998 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1001 JH/07 Add malware type "sock" for talking to simple daemon.
1003 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1005 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1006 routers/transports under cutthrough routing.
1008 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1009 numbers. Touch up "bool" conditional to keep the same definition.
1011 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1013 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1015 JH/12 Expand items in router/transport headers_add or headers_remove lists
1016 individually rather than the list as a whole. Bug 1452.
1018 Required for reasonable handling of multiple headers_ options when
1019 they may be empty; requires that headers_remove items with embedded
1020 colons must have them doubled (or the list-separator changed).
1022 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1023 view the policy declared in the DMARC record. Currently, $dmarc_status
1024 is a combined value of both the record presence and the result of the
1027 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1029 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1030 dnslookup router and the smtp transport (applying to the forward
1033 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1034 of ldap servers used for a specific lookup. Patch provided by Heiko
1037 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1038 New variable $lookup_dnssec_authenticated for observability.
1040 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1041 Patch submitted by Lars Timman.
1043 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1045 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1046 Requires trusted mode and valid format message id, aborts otherwise.
1047 Patch contributed by Heiko Schlichting.
1049 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1050 certextract with support for various fields. Bug 1358.
1052 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1053 is requested by default, modifiable by smtp transport option
1056 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1057 operate on certificate variables to give certificate fingerprints
1058 Also new ${sha256:cert_variable}.
1060 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1062 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1063 Christian Aistleitner.
1065 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1067 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1068 file. Patch from Wolfgang Breyha.
1070 JH/25 Expand the coverage of the delivery $host and $host_address to
1071 client authenticators run in verify callout. Bug 1476.
1073 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1074 align with daemon_smtp_ports. Bug 72.
1076 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1077 support and error reporting did not work properly.
1079 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1080 and is readable. Patch from Andrew Colin Kissa.
1082 TL/14 Enhance documentation of ${run expansion and how it parses the
1083 commandline after expansion, particularly in the case when an
1084 unquoted variable expansion results in an empty value.
1086 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1088 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1089 by renaming away from it. Addresses GH issue 15, reported by
1092 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1094 TL/15 SECURITY: prevent double expansion in math comparison functions
1095 (can expand unsanitized data). Not remotely exploitable.
1102 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1104 PP/02 Make -n do something, by making it not do something.
1105 When combined with -bP, the name of an option is not output.
1107 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1110 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1111 $sender_host_name and config options to manage this, and basic check
1114 PP/05 DSCP support for outbound connections and control modifier for inbound.
1116 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1117 (Only plugin which currently uses this is kerberos4, which nobody should
1118 be using, but we should make it available and other future plugins might
1119 conceivably use it, even though it would break NAT; stuff *should* be
1120 using channel bindings instead).
1122 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1123 name; added for Sendmail compatibility; requires admin caller.
1124 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1125 just ignore it); requires trusted caller.
1126 Also parse but ignore: -Ac -Am -X<logfile>
1129 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1131 TL/02 Add +smtp_confirmation as a default logging option.
1133 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1134 Patch by Magnus Holmgren from 2007-02-20.
1136 TL/04 Bugzilla 1281 - Spec typo.
1137 Bugzilla 1283 - Spec typo.
1138 Bugzilla 1290 - Spec grammar fixes.
1140 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1142 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1144 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1145 mailing list by by Dmitry Isaikin.
1147 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1149 JH/02 Support "G" suffix to numbers in ${if comparisons.
1151 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1153 NM/01 Bugzilla 1197 - Spec typo
1154 Bugzilla 1196 - Spec examples corrections
1156 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1158 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1159 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1162 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1163 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1165 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1166 "acl {{name}{arg}...}", and optional args on acl condition
1169 JH/05 Permit multiple router/transport headers_add/remove lines.
1171 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1173 JH/07 Avoid using a waiting database for a single-message-only transport.
1174 Performance patch from Paul Fisher. Bugzilla 1262.
1176 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1179 JH/09 Add $headers_added variable, with content from use of ACL modifier
1180 add_header (but not yet added to the message). Bugzilla 199.
1182 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1183 Pulled from Bugzilla 817 by Wolfgang Breyha.
1185 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1187 (nb: this is the same fix as in Exim 4.80.1)
1189 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1192 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1195 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1197 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1198 advertise SMTP AUTH mechanism to us, instead of a generic
1199 protocol violation error. Also, make Exim more robust to bad
1200 data from the Dovecot auth socket.
1202 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1204 When a queue runner is handling a message, Exim first routes the
1205 recipient addresses, during which it prunes them based on the retry
1206 hints database. After that it attempts to deliver the message to
1207 any remaining recipients. It then updates the hints database using
1210 So if a recipient address works intermittently, it can get repeatedly
1211 deferred at routing time. The retry hints record remains fresh so the
1212 address never reaches the final cutoff time.
1214 This is a fairly common occurrence when a user is bumping up against
1215 their storage quota. Exim had some logic in its local delivery code
1216 to deal with this. However it did not apply to per-recipient defers
1217 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1219 This change adds a proper retry rule check during routing so that the
1220 final cutoff time is checked against the message's age. We only do
1221 this check if there is an address retry record and there is not a
1222 domain retry record; this implies that previous attempts to handle
1223 the address had the retry_use_local_parts option turned on. We use
1224 this as an approximation for the destination being like a local
1225 delivery, as in LMTP.
1227 I suspect this new check makes the old local delivery cutoff check
1228 redundant, but I have not verified this so I left the code in place.
1230 TF/02 Correct gecos expansion when From: is a prefix of the username.
1232 Test 0254 submits a message to Exim with the header
1236 When I ran the test suite under the user fanf2, Exim expanded
1237 the header to contain my full name, whereas it should have added
1238 a Resent-Sender: header. It erroneously treats any prefix of the
1239 username as equal to the username.
1241 This change corrects that bug.
1243 GF/01 DCC debug and logging tidyup
1244 Error conditions log to paniclog rather than rejectlog.
1245 Debug lines prefixed by "DCC: " to remove any ambiguity.
1247 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1249 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1250 Bug spotted by Jeremy Harris; was flawed since initial commit.
1251 Would have resulted in OCSP responses post-SNI triggering an Exim
1252 NULL dereference and crash.
1254 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1256 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1257 Bug detection, analysis and fix by Samuel Thibault.
1258 Bugzilla 1331, Debian bug #698092.
1260 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1262 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1263 Server implementation by Todd Lyons, client by JH.
1264 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1265 config variable "prdr_enable" controls whether the server
1266 advertises the facility. If the client requests PRDR a new
1267 acl_data_smtp_prdr ACL is called once for each recipient, after
1268 the body content is received and before the acl_smtp_data ACL.
1269 The client is controlled by both of: a hosts_try_prdr option
1270 on the smtp transport, and the server advertisement.
1271 Default client logging of deliveries and rejections involving
1272 PRDR are flagged with the string "PRDR".
1274 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1275 fclose(). Diagnosis by Todd Lyons.
1277 PP/17 Update configure.default to handle IPv6 localhost better.
1278 Patch by Alain Williams (plus minor tweaks).
1281 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1282 This is now consistent with GnuTLS, and is now documented: the
1283 previous undocumented portable approach to treating the option as
1284 unset was to force an expansion failure. That still works, and
1285 an empty string is now equivalent.
1287 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1288 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1289 not performing validation itself.
1291 PP/20 Added force_command boolean option to pipe transport.
1292 Patch from Nick Koston, of cPanel Inc.
1294 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1297 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1299 PP/21 Fix eximon continuous updating with timestamped log-files.
1300 Broken in a format-string cleanup in 4.80, missed when I repaired the
1301 other false fix of the same issue.
1302 Report and fix from Heiko Schlichting.
1305 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1306 Report from Prashanth Katuri.
1308 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1309 It's SecureTransport, so affects any MacOS clients which use the
1310 system-integrated TLS libraries, including email clients.
1312 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1313 using a MIME ACL for non-SMTP local injection.
1314 Report and assistance in diagnosis by Warren Baker.
1316 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1318 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1320 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1321 last id that failed so it may be referenced in subsequent ACL's.
1323 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1326 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1327 ldap library initialization, allowing self-signed CA's to be
1328 used. Also properly sets require_cert option later in code by
1329 using NULL (global ldap config) instead of ldap handle (per
1330 session). Bug diagnosis and testing by alxgomz.
1332 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1333 the src/util/ subdirectory.
1335 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1336 renamed to Transport Post Delivery Action by Jeremy Harris, as
1339 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1340 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1341 redis_servers = needs to be configured which will be used by the redis
1342 lookup. Patch from Warren Baker, of The Packet Hub.
1344 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1346 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1347 hostname or reverse DNS when processing a host list. Used suggestions
1348 from multiple comments on this bug.
1350 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1352 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1353 Missed a few lines, added it to make the runtest require no keyboard
1356 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1357 contains upper case chars. Make router use caseful_local_part.
1359 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1360 support when GnuTLS has been built with p11-kit.
1366 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1368 This, or similar/improved, will also be change PP/11 of 4.82.
1374 PP/01 Handle short writes when writing local log-files.
1375 In practice, only affects FreeBSD (8 onwards).
1376 Bugzilla 1053, with thanks to Dmitry Isaikin.
1378 NM/01 Bugzilla 949 - Documentation tweak
1380 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1383 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1385 PP/02 Implemented gsasl authenticator.
1387 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1389 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1390 `pkg-config foo` for cflags/libs.
1392 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1393 with rest of GSASL and with heimdal_gssapi.
1395 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1396 `pkg-config foo` for cflags/libs for the TLS implementation.
1398 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1399 properties get this fed in as external SSF. A number of robustness
1400 and debugging improvements to the cyrus_sasl authenticator.
1402 PP/08 cyrus_sasl server now expands the server_realm option.
1404 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1405 Patch by Jeremy Harris.
1407 PP/10 Added dbmjz lookup type.
1409 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1411 PP/12 MAIL args handles TAB as well as SP, for better interop with
1412 non-compliant senders.
1413 Analysis and variant patch by Todd Lyons.
1415 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1416 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1417 Patch from Dirk Mueller <dmueller@suse.com>
1419 PP/13 tls_peerdn now print-escaped for spool files.
1420 Observed some $tls_peerdn in wild which contained \n, which resulted
1421 in spool file corruption.
1423 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1424 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1425 or write after TLS renegotiation, which otherwise led to messages
1428 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1429 as a tracking header (ie: a signed header comes before the signature).
1430 Patch from Wolfgang Breyha.
1432 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1433 comma-sep list; embedded commas doubled.
1435 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1437 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1439 Report and patch from Dmitry Banschikov.
1441 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1442 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1443 failures. We appear to now support TLS1.1+ with Exim.
1445 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1446 lets Exim select keys and certificates based upon TLS SNI from client.
1447 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1448 before an outbound SMTP session. New log_selector, +tls_sni.
1450 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1451 NULL dereference. Report and patch from Alun Jones.
1453 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1454 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1455 Not seeing resolver debug output on NetBSD, but suspect this is a
1456 resolver implementation change.
1458 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1459 Left warnings. Added "eximon gdb" invocation mode.
1461 PP/21 Defaulting "accept_8bitmime" to true, not false.
1463 PP/22 Added -bw for inetd wait mode support.
1465 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1466 locate the relevant includes and libraries. Made this the default.
1468 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1469 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1471 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1472 This may cause build issues on older platforms.
1474 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1475 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1476 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1477 Added SNI support via GnuTLS too.
1478 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1480 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1482 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1483 Applied second patch from Janne, implementing suggestion to default
1484 multiple-strings-in-record handling to match SPF spec.
1486 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1488 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1489 read-only, out of scope).
1490 Patch from Wolfgang Breyha, report from Stuart Northfield.
1492 PP/29 Fix three issues highlighted by clang analyser static analysis.
1493 Only crash-plausible issue would require the Cambridge-specific
1494 iplookup router and a misconfiguration.
1495 Report from Marcin Mirosław.
1497 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1499 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1500 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1501 As part of this, removing so much warning spew let me fix some minor
1502 real issues in debug logging.
1504 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1505 assignment on my part. Fixed.
1507 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1508 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1509 Janne Snabb (who went above and beyond: thank you).
1511 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1512 string otherwise requires a connection and a bunch more work and it's
1513 relatively easy to get wrong. Should also expose TLS library linkage
1516 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1517 64-bit ${eval} (JH/03).
1519 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1520 GNU libc to support some of the 64-bit stuff, should not lead to
1521 conflicts. Defined before os.h is pulled in, so if a given platform
1522 needs to override this, it can.
1524 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1525 protection layer was required, which is not implemented.
1526 Bugzilla 1254, patch from Wolfgang Breyha.
1528 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1529 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1530 tls_dhparam take prime identifiers. Also unbreak combination of
1531 OpenSSL+DH_params+TLSSNI.
1533 PP/39 Disable SSLv2 by default in OpenSSL support.
1539 PP/01 Solaris build fix for Oracle's LDAP libraries.
1540 Bugzilla 1109, patch from Stephen Usher.
1542 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1544 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1547 TF/02 Fix a couple more cases where we did not log the error message
1548 when unlink() failed. See also change 4.74-TF/03.
1550 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1551 lock up or crash if it happened to be inside a call to libc when it
1552 got a SIGUSR1 from exiwhat.
1554 The SIGUSR1 handler appends the current process status to the process
1555 log which is later printed by exiwhat. It used to use the general
1556 purpose logging code to do this, but several functions it calls are
1557 not safe for signals.
1559 The new output code in the SIGUSR1 handler is specific to the process
1560 log, and simple enough that it's easy to inspect for signal safety.
1561 Removing some special cases also simplifies the general logging code.
1562 Removing the spurious timestamps from the process log simplifies
1565 TF/04 Improved ratelimit ACL condition.
1567 The /noupdate option has been deprecated in favour of /readonly which
1568 has clearer semantics. The /leaky, /strict, and /readonly update modes
1569 are mutually exclusive. The update mode is no longer included in the
1570 database key; it just determines when the database is updated. (This
1571 means that when you upgrade Exim will forget old rate measurements.)
1573 Exim now checks that the per_* options are used with an update mode that
1574 makes sense for the current ACL. For example, when Exim is processing a
1575 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1576 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1577 must specify per_mail/readonly. If you omit the update mode it defaults to
1578 /leaky where that makes sense (as before) or /readonly where required.
1580 The /noupdate option is now undocumented but still supported for
1581 backwards compatibility. It is equivalent to /readonly except that in
1582 ACLs where /readonly is required you may specify /leaky/noupdate or
1583 /strict/noupdate which are treated the same as /readonly.
1585 A useful new feature is the /count= option. This is a generalization
1586 of the per_byte option, so that you can measure the throughput of other
1587 aggregate values. For example, the per_byte option is now equivalent
1588 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1590 The per_rcpt option has been generalized using the /count= mechanism
1591 (though it's more complicated than the per_byte equivalence). When it is
1592 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1593 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1594 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1595 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1596 Note that using per_rcpt with a non-readonly update mode in more than
1597 one ACL will cause the recipients to be double-counted. (The per_mail
1598 and per_byte options don't have this problem.)
1600 The handling of very low rates has changed slightly. If the computed rate
1601 is less than the event's count (usually one) then this event is the first
1602 after a long gap. In this case the rate is set to the same as this event's
1603 count, so that the first message of a spam run is counted properly.
1605 The major new feature is a mechanism for counting the rate of unique
1606 events. The new per_addr option counts the number of different
1607 recipients that someone has sent messages to in the last time period. It
1608 behaves like per_rcpt if all the recipient addresses are different, but
1609 duplicate recipient addresses do not increase the measured rate. Like
1610 the /count= option this is a general mechanism, so the per_addr option
1611 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1612 example, measure the rate that a client uses different sender addresses
1613 with the options per_mail/unique=$sender_address. There are further
1614 details in the main documentation.
1616 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1618 TF/06 Removed a few PCRE remnants.
1620 TF/07 Automatically extract Exim's version number from tags in the git
1621 repository when doing development or release builds.
1623 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1624 Bugzilla 879. Patch from Paul Fisher.
1626 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1627 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1630 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1632 PP/05 Variable $av_failed, true if the AV scanner deferred.
1633 Bugzilla 1078. Patch from John Horne.
1635 PP/06 Stop make process more reliably on build failure.
1636 Bugzilla 1087. Patch from Heiko Schlittermann.
1638 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1639 Bugzilla 1089. Patch from Heiko Schlittermann.
1641 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1642 Bugzilla 1131. Patch from Holger Weiß.
1644 PP/09 Handle IPv6 addresses with SPF.
1645 Bugzilla 860. Patch from Wolfgang Breyha.
1647 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1649 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1652 PP/11 match_* no longer expand right-hand-side by default.
1653 New compile-time build option, EXPAND_LISTMATCH_RHS.
1654 New expansion conditions, "inlist", "inlisti".
1656 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1658 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1660 PP/14 fix log_write() format string regression from TF/03.
1661 Bugzilla 1152. Patch from Dmitry Isaikin.
1667 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1669 PP/02 Harmonised TLS library version reporting; only show if debugging.
1670 Layout now matches that introduced for other libraries in 4.74 PP/03.
1672 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1674 PP/04 New "dns_use_edns0" global option.
1676 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1679 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1680 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1682 TK/01 Updated PolarSSL code to 0.14.2.
1683 Bugzilla 1097. Patch from Andreas Metzler.
1685 PP/07 Catch divide-by-zero in ${eval:...}.
1686 Fixes bugzilla 1102.
1688 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1691 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1692 format-string attack -- SECURITY: remote arbitrary code execution.
1694 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1695 time unintentionally subject to list matching rules, letting the header
1696 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1697 arbitrary string expansion). This allowed for information disclosure.
1699 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1700 INT_MIN/-1 -- value coerced to INT_MAX.
1706 NM/01 Workaround for PCRE version dependency in version reporting
1709 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1710 This fixes portability to compilers other than gcc, notably
1711 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1713 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1714 makefiles for portability to HP-UX and POSIX correctness.
1716 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1717 Also via indented variable definition in the Makefile.
1718 (Debugging by Oliver Heesakkers).
1720 PP/02 Restore caching of spamd results with expanded spamd_address.
1721 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1723 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1724 Improves build reliability. Fix from: Frank Elsner
1726 NM/02 Fix wide character breakage in the rfc2047 coding
1727 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1729 NM/03 Allow underscore in dnslist lookups
1730 Fixes bug 1026. Patch from Graeme Fowler
1732 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1733 Code patches from Adam Ciarcinski of NetBSD.
1735 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1738 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1739 is logged, to avoid truncation. Patch from John Horne.
1741 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1742 Patch from Jakob Hirsch.
1744 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1745 SQL string expansion failure details.
1746 Patch from Andrey Oktyabrski.
1748 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1749 Patch from Simon Arlott.
1751 PP/09 New lookups functionality failed to compile on old gcc which rejects
1752 extern declarations in function scope.
1753 Patch from Oliver Fleischmann
1755 PP/10 Use sig_atomic_t for flags set from signal handlers.
1756 Check getgroups() return and improve debugging.
1757 Fixed developed for diagnosis in bug 927 (which turned out to be
1760 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1761 Patch from Mark Zealey.
1763 PP/12 Bugzilla 1056: Improved spamd server selection.
1764 Patch from Mark Zealey.
1766 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1767 Based on patch from Heiko Schlittermann.
1769 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1770 Patch from Uwe Doering, sign-off by Michael Haardt.
1772 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1773 variable declaration deep within a block. Bug and patch from
1776 PP/15 lookups-Makefile IRIX compatibility coercion.
1778 PP/16 Make DISABLE_DKIM build knob functional.
1780 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1781 Patch by Simon Arlott
1783 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1784 variable argument macros. Our copy now differs from upstream.
1790 TF/01 Failure to get a lock on a hints database can have serious
1791 consequences so log it to the panic log.
1793 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1794 controlled using the smtp_confirmation log selector.
1796 TF/03 Include the error message when we fail to unlink a spool file.
1798 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1799 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1800 for maintaining out-of-tree patches for some time.
1802 PP/01 Bugzilla 139: Documentation and portability issues.
1803 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1804 Handle per-OS dynamic-module compilation flags.
1806 PP/02 Let /dev/null have normal permissions.
1807 The 4.73 fixes were a little too stringent and complained about the
1808 permissions on /dev/null. Exempt it from some checks.
1809 Reported by Andreas M. Kirchwitz.
1811 PP/03 Report version information for many libraries, including
1812 Exim version information for dynamically loaded libraries. Created
1813 version.h, now support a version extension string for distributors
1814 who patch heavily. Dynamic module ABI change.
1816 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1817 privilege escalation vulnerability whereby the Exim run-time user
1818 can cause root to append content of the attacker's choosing to
1821 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1824 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1825 If dropping privileges for untrusted macros, we disabled normal logging
1826 on the basis that it would fail; for the Exim run-time user, this is not
1827 the case, and it resulted in successful deliveries going unlogged.
1828 Fixed. Reported by Andreas Metzler.
1834 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1835 only prepend for the Resent-* case. Fixes regression introduced in
1836 Exim 4.70 by NM/22 for Bugzilla 607.
1838 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1839 increasing numbers of administrators be bitten by this.
1841 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1843 PP/03 Bugzilla 994: added openssl_options main configuration option.
1845 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1847 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1849 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1851 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1852 without a peer certificate, leading to a segfault because of an
1853 assumption that peers always have certificates. Be a little more
1854 paranoid. Problem reported by Martin Tscholak.
1856 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1857 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1858 NB: ClamAV planning to remove STREAM in "middle of 2010".
1859 CL also introduces -bmalware, various -d+acl logging additions and
1860 more caution in buffer sizes.
1862 PP/09 Implemented reverse_ip expansion operator.
1864 PP/10 Bugzilla 937: provide a "debug" ACL control.
1866 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1868 PP/12 Bugzilla 973: Implement --version.
1870 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1872 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1874 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1876 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1877 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1878 ignore trailing whitespace.
1880 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1882 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1885 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1886 Notification from Dr Andrew Aitchison.
1888 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1889 ExtendedDetectionInfo response format.
1890 Notification from John Horne.
1892 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1895 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1896 XSL and documented dependency on system catalogs, with examples of how
1899 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1902 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1903 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1904 configuration file which is writeable by the Exim user or group.
1906 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1907 of configuration files to cover files specified with the -C option if
1908 they are going to be used with root privileges, not just the default
1911 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1912 option (effectively making it always true).
1914 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1915 files to be used while preserving root privileges.
1917 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1918 that rogue child processes cannot use them.
1920 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1921 run-time user, instead of root.
1923 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1924 Exim run-time user without dropping privileges.
1926 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1927 result string, instead of calling string_vformat() twice with the same
1930 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1931 for other users. Others should always drop root privileges if they use
1932 -C on the command line, even for a whitelisted configure file.
1934 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1936 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1942 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1943 $data_path, and $header_path variables; fixed documentation bugs and
1946 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1947 exipick to access non-standard spools, including the "frozen" queue
1950 NM/01 Bugzilla 965: Support mysql stored procedures.
1951 Patch from Alain Williams
1953 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1955 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1956 Patch from Andreas Metzler
1958 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1959 Patch from Kirill Miazine
1961 NM/05 Bugzilla 671: Added umask to procmail example.
1963 JJ/03 installed exipick 20100323.0, fixing doc bug
1965 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1966 directory. Notification and patch from Dan Rosenberg.
1968 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1970 TK/02 Improve log output when DKIM signing operation fails.
1972 MH/01 Treat the transport option dkim_domain as a colon separated
1973 list, not as a single string, and sign the message with each element,
1974 omitting multiple occurences of the same signer.
1976 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1977 Bugzilla 985, 986. Patch by Simon Arlott
1979 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1980 Patch by Simon Arlott
1982 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1983 MBX locking. Notification from Dan Rosenberg.
1989 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1991 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1993 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1995 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1997 NM/04 Bugzilla 915: Flag broken perl installation during build.
2003 TK/01 Added patch by Johannes Berg that expands the main option
2004 "spamd_address" if it starts with a dollar sign.
2006 TK/02 Write list of recipients to X-Envelope-Sender header when building
2007 the mbox-format spool file for content scanning (suggested by Jakob
2010 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2011 (http://www.dcc-servers.net/) support via dccifd. Activated by
2012 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2014 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2015 by Mark Daniel Reidel <mr@df.eu>.
2017 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2018 When building exim an external PCRE library is now needed -
2019 PCRE is a system library on the majority of modern systems.
2020 See entry on PCRE_LIBS in EDITME file.
2022 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2023 conversation. Added nologin parameter to request.
2024 Patch contributed by Kirill Miazine.
2026 TF/01 Do not log submission mode rewrites if they do not change the address.
2028 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2030 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2031 log files in place. Contributed by Roberto Lima.
2033 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2035 TF/03 Bugzilla 615: When checking the local_parts router precondition
2036 after a local_part_suffix or local_part_prefix option, Exim now
2037 does not use the address's named list lookup cache, since this
2038 contains cached lookups for the whole local part.
2040 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2041 Robert Millan. Documentation is in experimental-spec.txt.
2043 TF/04 Bugzilla 668: Fix parallel build (make -j).
2045 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2047 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2048 Patch provided by Jan Srzednicki.
2050 TF/05 Leading white space used to be stripped from $spam_report which
2051 wrecked the formatting. Now it is preserved.
2053 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2054 that they are available at delivery time.
2056 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2058 TF/08 TLS error reporting now respects the incoming_interface and
2059 incoming_port log selectors.
2061 TF/09 Produce a more useful error message if an SMTP transport's hosts
2062 setting expands to an empty string.
2064 NM/06 Bugzilla 744: EXPN did not work under TLS.
2065 Patch provided by Phil Pennock.
2067 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2068 Patch provided by Richard Godbee.
2070 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2071 acl_smtp_notquit, added index entry.
2073 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2074 Patch provided by Eugene Bujak.
2076 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2077 accept(). Patch provided by Maxim Dounin.
2079 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2080 Patch provided by Phil Pennock.
2082 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2084 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2085 Patch provided by Brad "anomie" Jorsch.
2087 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2088 Patch provided by Dean Brooks.
2090 TK/05 Add native DKIM support (does not depend on external libraries).
2092 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2093 Patch provided by Graeme Fowler.
2095 NM/16 Bugzilla 851: Documentation example syntax fix.
2097 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2099 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2102 NM/19 Bugzilla 745: TLS version reporting.
2103 Patch provided by Phil Pennock.
2105 NM/20 Bugzilla 167: bool: condition support.
2106 Patch provided by Phil Pennock.
2108 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2109 clients. Patch provided by Phil Pennock.
2111 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2112 Patch provided by Brad "anomie" Jorsch.
2114 NM/23 Bugzilla 687: Fix misparses in eximstats.
2115 Patch provided by Heiko Schlittermann.
2117 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2118 Patch provided by Heiko Schlittermann.
2120 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2121 plus update to original patch.
2123 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2125 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2126 Patch provided by David Brownlee.
2128 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2130 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2132 NM/30 Bugzilla 888: TLS documentation bugfixes.
2134 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2136 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2137 Unlike the original bugzilla I have changed all shell scripts in src tree.
2139 NM/33 Bugzilla 898: Transport filter timeout fix.
2140 Patch by Todd Rinaldo.
2142 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2143 Patch by Serge Demonchaux.
2145 NM/35 Bugzilla 39: Base64 decode bug fixes.
2146 Patch by Jakob Hirsch.
2148 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2150 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2152 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2154 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2160 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2161 ALT-N's libdkim that I have put here:
2162 http://duncanthrax.net/exim-experimental/
2164 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2165 were called 'true' and it seems that C99 defines that as a reserved
2166 keyword to be used with 'bool' variable types. That means you could
2167 not include C99-style headers which use bools without triggering
2168 build errors in sieve.c.
2170 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2171 as mailq or other aliases. Changed the --help handling significantly
2172 to do whats expected. exim_usage() emits usage/help information.
2174 SC/01 Added the -bylocaldomain option to eximstats.
2176 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2178 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2180 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2186 PH/01 Another patch from the Sieve maintainer.
2188 PH/02 When an IPv6 address is converted to a string for single-key lookup
2189 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2190 dots are used instead of colons so that keys in lsearch files need not
2191 contain colons. This was done some time before quoting was made available
2192 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2193 (notated using the quote facility) so as to distinguish them from IPv4
2194 keys. This meant that lookups for IP addresses in host lists did not work
2195 for iplsearch lookups.
2197 This has been fixed by arranging for IPv6 addresses to be expressed with
2198 colons if the lookup type is iplsearch. This is not incompatible, because
2199 previously such lookups could never work.
2201 The situation is now rather anomalous, since one *can* have colons in
2202 ordinary lsearch keys. However, making the change in all cases is
2203 incompatible and would probably break a number of configurations.
2205 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2208 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2209 conversion specification without a maximum field width, thereby enabling
2210 a rogue spamd server to cause a buffer overflow. While nobody in their
2211 right mind would setup Exim to query an untrusted spamd server, an
2212 attacker that gains access to a server running spamd could potentially
2213 exploit this vulnerability to run arbitrary code as the Exim user.
2215 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2216 $primary_hostname instead of what libspf2 thinks the hosts name is.
2218 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2219 a directory entry by the name of the lookup key. Previously, if a
2220 symlink pointed to a non-existing file or a file in a directory that
2221 Exim lacked permissions to read, a lookup for a key matching that
2222 symlink would fail. Now it is enough that a matching directory entry
2223 exists, symlink or not. (Bugzilla 503.)
2225 PH/03 The body_linecount and body_zerocount variables are now exported in the
2228 PH/04 Added the $dnslist_matched variable.
2230 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2231 This means they are set thereafter only if the connection becomes
2234 PH/06 Added the client_condition to authenticators so that some can be skipped
2235 by clients under certain conditions.
2237 PH/07 The error message for a badly-placed control=no_multiline_responses left
2238 "_responses" off the end of the name.
2240 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2242 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2243 (without spaces) instead of just copying the configuration text.
2245 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2247 PH/11 Added $max_received_linelength.
2249 PH/12 Added +ignore_defer and +include_defer to host lists.
2251 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2252 way in which PCRE > 7.0 is built.
2254 PH/14 Implemented queue_only_load_latch.
2256 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2257 MAIL command. The effect was to mangle the value on 64-bit systems.
2259 PH/16 Another patch from the Sieve maintainer.
2261 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2263 PH/18 If a system quota error occurred while trying to create the file for
2264 a maildir delivery, the message "Mailbox is full" was not appended to the
2265 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2266 only to a quota excession during the actual writing of the file.
2268 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2269 characters?) which causes problems in log lines. The DN values are now
2270 passed through string_printing() before being added to log lines.
2272 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2273 and InterBase are left for another time.)
2275 PH/21 Added message_body_newlines option.
2277 PH/22 Guard against possible overflow in moan_check_errorcopy().
2279 PH/23 POSIX allows open() to be a macro; guard against that.
2281 PH/24 If the recipient of an error message contained an @ in the local part
2282 (suitably quoted, of course), incorrect values were put in $domain and
2283 $local_part during the evaluation of errors_copy.
2289 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2290 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2293 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2294 issue a MAIL command.
2296 PH/02 In an ACL statement such as
2298 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2300 if a client was not listed at all, or was listed with a value other than
2301 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2302 the condition was not true (as it should be), so access was not denied.
2303 The bug was that the ! inversion was incorrectly passed on to the second
2304 item. This has been fixed.
2306 PH/03 Added additional dnslists conditions == and =& which are different from
2307 = and & when the dns lookup returns more than one IP address.
2309 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2310 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2312 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2313 FSYNC, which compiles an option called disable_fsync that allows for
2314 bypassing fsync(). The documentation is heavily laced with warnings.
2316 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2318 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2319 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2320 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2321 including adding "make clean"; (3) Added -fPIC when compiling the test
2322 dynamically loaded module, to get rid of a warning.
2324 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2325 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2326 = 0s. The bug is otherwise harmless.
2328 PH/07 There was a bug in the dovecot authenticator such that the value of
2329 $auth1 could be overwritten, and so not correctly preserved, after a
2330 successful authentication. This usually meant that the value preserved by
2331 the server_setid option was incorrect.
2333 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2335 PH/09 Installed PCRE release 7.0.
2337 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2338 run for batched SMTP input. It is now run at the start of every message
2339 in the batch. While fixing this I discovered that the process information
2340 (output by running exiwhat) was not always getting set for -bs and -bS
2341 input. This is fixed, and it now also says "batched" for BSMTP.
2343 PH/11 Added control=no_pipelining.
2345 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2346 patch, slightly modified), and move the expansion of helo_data till after
2347 the connection is made in the smtp transport (so it can use these
2350 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2352 PH/14 Added log_selector = +pid.
2354 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2356 PH/16 Add ${if forany and ${if forall.
2358 PH/17 Added dsn_from option to vary the From: line in DSNs.
2360 PH/18 Flush SMTP output before performing a callout, unless control =
2361 no_callout_flush is set.
2363 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2364 was true (the default) a successful delivery failed to delete the retry
2365 item, thus causing premature timeout of the address. The bug is now
2368 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2370 PH/21 Long custom messages for fakedefer and fakereject are now split up
2371 into multiline responses in the same way that messages for "deny" and
2372 other ACL rejections are.
2374 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2375 with slight modification.
2377 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2378 draft, changing the syntax and factoring some duplicate code".
2380 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2381 for deliveries of the second and subsequent messages over the same SMTP
2384 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2385 ${reduce, with only minor "tidies".
2387 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2389 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2390 expansion side effects.
2392 PH/27 When a message times out after an over-quota error from an Exim-imposed
2393 quota, the bounce message says "mailbox is full". This message was not
2394 being given when it was a system quota that was exceeded. It now should
2397 MH/03 Made $recipients available in local_scan(). local_scan() already has
2398 better access to the recipient list through recipients_list[], but
2399 $recipients can be useful in postmaster-provided expansion strings.
2401 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2402 in the case of a MAIL command with additional options following the
2403 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2404 were accidentally chopped off.
2406 PH/29 SMTP synchronization checks are implemented when a command is read -
2407 there is a check that no more input is waiting when there shouldn't be
2408 any. However, for some commands, a delay in an ACL can mean that it is
2409 some time before the response is written. In this time, more input might
2410 arrive, invalidly. So now there are extra checks after an ACL has run for
2411 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2412 pipelining has not been advertised.
2414 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2416 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2417 returned in the SMTP response when smtp_return_error_details was false.
2418 This has been fixed.
2420 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2421 instead of the C I/O that was originally supplied, because problems were
2422 reported on Solaris.
2424 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2425 Exim which did not show up earlier: it was assuming that a call to
2426 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2427 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2428 was a macro that became an assignment, so it seemed to work. This has
2429 changed to a proper function call with a void return, hence the compile
2430 error. Exim's code has been fixed.
2432 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2435 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2437 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2439 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2440 because it turns out to be too big or there is a timeout) there is no
2441 "Completed" line in the log. When some messages of this type were
2442 selected by exigrep, they were listed as "not completed". Others were
2443 picked up by some special patterns. I have improved the selection
2444 criteria to be more general.
2446 PH/38 The host_find_failed option in the manualroute router can now be set
2447 to "ignore", to completely ignore a host whose IP address cannot be
2448 found. If all hosts are ignored, the behaviour is controlled by the new
2449 host_all_ignored option.
2451 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2452 homing or because of multiple MX records with /mx) generated more than
2453 one IP address, and the following item turned out to be the local host,
2454 all the secondary addresses of the first item were incorrectly removed
2455 from the list, along with the local host and any following hosts (which
2456 is what is supposed to happen).
2458 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2459 whoever called Exim into the -H file. In the case of the daemon it was
2460 behaving confusingly. When first started, it used values for whoever
2461 started the daemon, but after a SIGHUP it used the Exim user (because it
2462 calls itself on a restart). I have changed the code so that it now always
2465 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2466 message are rejected with the same error (e.g. no authentication or bad
2467 sender address), and a DATA command is nevertheless sent (as can happen
2468 with PIPELINING or a stupid MUA), the error message that was given to the
2469 RCPT commands is included in the rejection of the DATA command. This is
2470 intended to be helpful for MUAs that show only the final error to their
2473 PH/42 Another patch from the Sieve maintainer.
2475 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2476 Eximstats - Fixed some broken HTML links and added missing column headers
2478 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2479 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2481 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2483 PH/43 Yet another patch from the Sieve maintainer.
2485 PH/44 I found a way to check for a TCP/IP connection going away before sending
2486 the response to the final '.' that terminates a message, but only in the
2487 case where the client has not sent further data following the '.'
2488 (unfortunately, this is allowed). However, in many cases there won't be
2489 any further data because there won't be any more messages to send. A call
2490 to select() can be used: if it shows that the input is "ready", there is
2491 either input waiting, or the socket has been closed. An attempt to read
2492 the next input character can distinguish the two cases. Previously, Exim
2493 would have sent an OK response which the client would never have see.
2494 This could lead to message repetition. This fix should cure that, at
2495 least in a lot of common cases.
2497 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2498 advertised in response to EHLO.
2504 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2505 fixed by 4.65/MH/01 (is this a record?) are fixed:
2507 (i) An empty string was always treated as zero by the numeric comparison
2508 operators. This behaviour has been restored.
2510 (ii) It is documented that the numeric comparison operators always treat
2511 their arguments as decimal numbers. This was broken in that numbers
2512 starting with 0 were being interpreted as octal.
2514 While fixing these problems I realized that there was another issue that
2515 hadn't been noticed. Values of message_size_limit (both the global option
2516 and the transport option) were treated as octal if they started with 0.
2517 The documentation was vague. These values are now always treated as
2518 decimal, and I will make that clear in the documentation.
2524 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2525 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2528 MH/01 Don't check that the operands of numeric comparison operators are
2529 integers when their expansion is in "skipping" mode (fixes bug
2530 introduced by 4.64-PH/07).
2532 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2533 child addresses, Exim now panics and dies. Previously, because the count
2534 is held in a short int, deliveries were likely to be lost. As such a
2535 large number of recipients for a single message is ridiculous
2536 (performance will be very, very poor), I have chosen to impose a limit
2537 rather than extend the field.
2543 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2544 leftover -K file (the existence of which was triggered by #402).
2545 While we were at it, introduced process PID as part of the -K
2546 filename. This should rule out race conditions when creating
2549 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2550 processing considerably. Previous code took too long for large mails,
2551 triggering a timeout which in turn triggers #401.
2553 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2554 in the DK code in transports.c. sendfile() is not really portable,
2555 hence the _LINUX specificness.
2557 TF/01 In the add_headers option to the mail command in an Exim filter,
2558 there was a bug that Exim would claim a syntax error in any
2559 header after the first one which had an odd number of characters
2562 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2563 callout verification, Exim cached a "reject" for the entire domain. This
2564 is correct for most verifications, but it is not correct for a recipient
2565 verification with use_sender or use_postmaster set, because in that case
2566 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2567 case of MAIL FROM:<> rejection from other early rejections (e.g.
2568 rejection of HELO). When verifying a recipient using a non-null MAIL
2569 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2570 Whatever the result of the callout, the value of the domain cache is
2571 left unchanged (for any other kind of callout, getting as far as trying
2572 RCPT means that the domain itself is ok).
2574 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2577 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2578 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2579 session, the daemon ignores ECONNECT errors and logs others; it now
2580 ignores EPIPE as well.
2582 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2583 (quoted-printable decoding).
2585 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2586 later the small subsequent patch to fix an introduced bug.
2588 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2590 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2592 PH/08 An error is now given if message_size_limit is specified negative.
2594 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2595 to be given (somewhat) arbitrary names.
2597 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2600 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2601 miscellaneous code fixes
2603 PH/10 Added the log_reject_target ACL modifier to specify where to log
2606 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2607 hostname. This is wrong, because it relates to the incoming message (and
2608 probably the interface on which it is arriving) and not to the outgoing
2609 callout (which could be using a different interface). This has been
2610 changed to use the value of the helo_data option from the smtp transport
2611 instead - this is what is used when a message is actually being sent. If
2612 there is no remote transport (possible with a router that sets up host
2613 addresses), $smtp_active_hostname is used.
2615 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2616 tweaks were necessary in order to get it to work (see also 21 below):
2617 (a) The code assumed that strncpy() returns a negative number on buffer
2618 overflow, which isn't the case. Replaced with Exim's string_format()
2620 (b) There were several signed/unsigned issues. I just did the minimum
2621 hacking in of casts. There is scope for a larger refactoring.
2622 (c) The code used strcasecmp() which is not a standard C function.
2623 Replaced with Exim's strcmpic() function.
2624 (d) The code set only $1; it now sets $auth1 as well.
2625 (e) A simple test gave the error "authentication client didn't specify
2626 service in request". It would seem that Dovecot has changed its
2627 interface. Fortunately there's a specification; I followed it and
2628 changed what the client sends and it appears to be working now.
2630 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2633 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2634 address A is aliased to B and C, where B exists and C does not. Without
2635 -v the output is "A verified" because verification stops after a
2636 successful redirection if more than one address is generated. However,
2637 with -v the child addresses are also verified. Exim was outputting "A
2638 failed to verify" and then showing the successful verification for C,
2639 with its parentage. It now outputs "B failed to verify", showing B's
2640 parentage before showing the successful verification of C.
2642 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2643 look up a TXT record in a specific list after matching in a combined
2646 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2647 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2648 they consult the DNS. I had assumed they would set it the way they
2649 wanted; and indeed my experiments on Linux seem to show that in some
2650 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2651 To be on the safe side, however, I have now made the interface to
2652 host_find_byname() similar to host_find_bydns(), with an argument
2653 containing the DNS resolver options. The host_find_byname() function now
2654 sets these options at its start, just as host_find_bydns() does. The smtp
2655 transport options dns_qualify_single and dns_search_parents are passed to
2656 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2657 of host_find_byname() use the default settings of RES_DEFNAMES
2658 (qualify_single) but not RES_DNSRCH (search_parents).
2660 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2661 spool_read_header() do less string testing, by means of a preliminary
2662 switch on the second character of optional "-foo" lines. (This is
2663 overdue, caused by the large number of possibilities that now exist.
2664 Originally there were few.) While I was there, I also converted the
2665 str(n)cmp tests so they don't re-test the leading "-" and the first
2666 character, in the hope this might squeeze out yet more improvement.
2668 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2669 flag allowing group syntax was set by the header_syntax check but not
2670 turned off, possible causing trouble later; (2) The flag was not being
2671 set at all for the header_verify test, causing "group"-style headers to
2672 be rejected. I have now set it in this case, and also caused header_
2673 verify to ignore an empty address taken from a group. While doing this, I
2674 came across some other cases where the code for allowing group syntax
2675 while scanning a header line wasn't quite right (mostly, not resetting
2676 the flag correctly in the right place). These bugs could have caused
2677 trouble for malformed header lines. I hope it is now all correct.
2679 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2680 with the "reply" argument non-NULL. The code, however (which originally
2681 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2682 but it didn't always do it. This confused somebody who was copying the
2683 code for some other use. I have removed all the tests.
2685 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2686 feature that was used to support insecure browsers during the U.S. crypto
2687 embargo. It requires special client support, and Exim is probably the
2688 only MTA that supported it -- and would never use it because real RSA is
2689 always available. This code has been removed, because it had the bad
2690 effect of slowing Exim down by computing (never used) parameters for the
2691 RSA_EXPORT functionality.
2693 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2694 authenticator to fail if there's a tab character in the incoming data
2695 (there should never be unless someone is messing about, as it's supposed
2696 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2697 if the connection is using TLS or if the remote IP is the same as the
2698 local IP, and the "valid-client-cert option" if a client certificate has
2701 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2702 authenticators. This can be used for authorization after authentication
2703 succeeds. (In the case of plaintext, it servers for both authentication
2706 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2707 if any retry times were supplied.
2709 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2710 connection, where there is no client IP address to check. In this
2711 situation, the verify now always succeeds.
2713 PH/25 Applied John Jetmore's -Mset patch.
2715 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2717 PH/27 In a string expansion for a processed (not raw) header when multiple
2718 headers of the same name were present, leading whitespace was being
2719 removed from all of them, but trailing whitespace was being removed only
2720 from the last one. Now trailing whitespace is removed from each header
2721 before concatenation. Completely empty headers in a concatenation (as
2722 before) are ignored.
2724 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2725 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2727 PH/29 [Removed. This was a change that I later backed out, and forgot to
2728 correct the ChangeLog entry (that I had efficiently created) before
2729 committing the later change.]
2731 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2732 address errors (4xx response to RCPT) over the same connection as other
2733 messages routed to the same hosts. Such deliveries are always "forced",
2734 so retry times are not inspected. This resulted in far too many retries
2735 for the affected addresses. The effect occurred only when there were more
2736 hosts than the hosts_max_try setting in the smtp transport when it had
2737 the 4xx errors. Those hosts that it had tried were not added to the list
2738 of hosts for which the message was waiting, so if all were tried, there
2739 was no problem. Two fixes have been applied:
2741 (i) If there are any address or message errors in an SMTP delivery, none
2742 of the hosts (tried or untried) are now added to the list of hosts
2743 for which the message is waiting, so the message should not be a
2744 candidate for sending over the same connection that was used for a
2745 successful delivery of some other message. This seems entirely
2746 reasonable: after all the message is NOT "waiting for some host".
2747 This is so "obvious" that I'm not sure why it wasn't done
2748 previously. Hope I haven't missed anything, but it can't do any
2749 harm, as the worst effect is to miss an optimization.
2751 (ii) If, despite (i), such a delivery is accidentally attempted, the
2752 routing retry time is respected, so at least it doesn't keep
2753 hammering the server.
2755 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2756 in ${readsocket because some servers need this prod.
2758 PH/32 Added some extra debug output when updating a wait-xxx database.
2760 PH/33 The hint "could be header name not terminated by colon", which has been
2761 given for certain expansion errors for a long time, was not being given
2762 for the ${if def:h_colon_omitted{... case.
2764 PH/34 The spec says: "With one important exception, whenever a domain list is
2765 being scanned, $domain contains the subject domain." There was at least
2766 one case where this was not true.
2768 PH/35 The error "getsockname() failed: connection reset by peer" was being
2769 written to the panic log as well as the main log, but it isn't really
2770 panic-worthy as it just means the connection died rather early on. I have
2771 removed the panic log writing for the ECONNRESET error when getsockname()
2774 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2775 runs only) independently of the message's sender address. This meant
2776 that, if the 4xx error was in fact related to the sender, a different
2777 message to the same recipient with a different sender could confuse
2778 things. In particular, this can happen when sending to a greylisting
2779 server, but other circumstances could also provoke similar problems.
2780 I have changed the default so that the retry time for these errors is now
2781 based a combination of the sender and recipient addresses. This change
2782 can be overridden by setting address_retry_include_sender=false in the
2785 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2786 remote server are returned as part of bounce messages. This was not
2787 happening for LMTP over a pipe (the lmtp transport), but now it is the
2788 same for both kinds of LMTP.
2790 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2791 in header lines that were in fact CNAMEs. This is no longer the case.
2793 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2794 and queue runs started by the daemon processed all messages. This has
2795 been fixed so that -R and -S can now usefully be given with -q<time>.
2797 PH/40 Import PCRE release 6.7 (fixes some bugs).
2799 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2801 PH/42 Give an error if -q is specified more than once.
2803 PH/43 Renamed the variables $interface_address and $interface_port as
2804 $received_ip_address and $received_port, to make it clear that these
2805 values apply to message reception, and not to the outgoing interface when
2806 a message is delivered. (The old names remain recognized, of course.)
2808 PH/44 There was no timeout on the connect() call when using a Unix domain
2809 socket in the ${readsocket expansion. There now is.
2811 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2812 be meaningful with "accept".
2814 SC/01 Eximstats V1.43
2815 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2817 SC/02 Eximstats V1.44
2818 Use a glob alias rather than an array ref in the generated
2819 parser. This improves both readability and performance.
2821 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2822 Collect SpamAssassin and rejection statistics.
2823 Don't display local sender or destination tables unless
2824 there is data to show.
2825 Added average volumes into the top table text output.
2827 SC/04 Eximstats V1.46
2828 Collect data on the number of addresses (recipients)
2829 as well as the number of messages.
2831 SC/05 Eximstats V1.47
2832 Added 'Message too big' to the list of mail rejection
2833 reasons (thanks to Marco Gaiarin).
2835 SC/06 Eximstats V1.48
2836 Mainlog lines which have GMT offsets and are too short to
2837 have a flag are now skipped.
2839 SC/07 Eximstats V1.49 (Alain Williams)
2840 Added the -emptyok flag.
2842 SC/08 Eximstats V1.50
2843 Fixes for obtaining the IP address from reject messages.
2845 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2846 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2847 whitespace changes from 4.64-PH/27
2849 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2852 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2853 are found, allow negative numbers in numeric criteria)
2855 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2857 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2858 to match changes made in 4.64-PH/43
2860 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2862 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2863 "domain" is actually an IP address, and force a failure. This locks out
2864 those revolvers/nameservers that support "A-for-A" lookups, in
2865 contravention of the specifications.
2867 PH/48 When a host name was looked up from an IP address, and the subsequent
2868 forward lookup of the name timed out, the host name was left in
2869 $sender_host_name, contrary to the specification.
2871 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2872 restricted to single-key lookups, Exim was not diagnosing an error if
2873 * or *@ was used with a query-style lookup.
2875 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2877 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2878 long ago, but noone interested enough thought of it. Let's just say that
2879 the "1.1" means that there are some new functions that weren't there at
2880 some point in the past.
2882 PH/51 Error processing for expansion failure of helo_data from an smtp
2883 transport during callout processing was broken.
2885 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2886 tested/used via the -bh/-bhc/-bs options.
2888 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2889 bug, fixed in subsequent PCRE releases).
2891 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2892 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2894 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2900 SC/01 Use a glob alias rather than an array ref in eximstats generated
2901 parser. This improves both readability and performance.
2903 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2904 Don't display local sender or destination tables in eximstats unless
2905 there is data to show.
2906 Added average volumes into the eximstats top table text output.
2908 SC/03 Collect data on the number of addresses (recipients) as well
2909 as the number of messages in eximstats.
2911 TF/01 Correct an error in the documentation for the redirect router. Exim
2912 does (usually) call initgroups() when daemonizing.
2914 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2915 with consistent privilege compared to when running as a daemon.
2917 TF/03 Note in the spec that $authenticated_id is not set for local
2918 submissions from trusted users.
2920 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2921 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2923 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2924 by adding some example configuration directives to the default
2925 configuration file. A little bit of work is required to uncomment the
2926 directives and define how usernames and passwords are checked, but
2927 there is now a framework to start from.
2929 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2930 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2931 without this. I don't know how relevant this is to other LDAP libraries.
2933 PH/02 Add the verb name to the "unknown ACL verb" error.
2935 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2937 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2939 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2940 directory not expanded when it should be if an expanded home directory
2941 was set for the address (which is overridden by the transport).
2943 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2946 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2947 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2948 because it is too late at that time, and has no effect.
2950 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2951 security issue with \' (bugzilla #107). I could not use the
2952 PQescapeStringConn() function, because it needs a PGconn value as one of
2955 PH/08 When testing addresses using -bt, indicate those final addresses that
2956 are duplicates that would not cause an additional delivery. At least one
2957 person was confused, thinking that -bt output corresponded to deliveries.
2958 (Suppressing duplicates isn't a good idea as you lose the information
2959 about possibly different redirections that led to the duplicates.)
2961 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2962 systems where poll() doesn't work, in particular OS X.
2964 PH/10 Added more information to debugging output for retry time not reached.
2966 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2967 operations in malware.c.
2969 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2972 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2973 syslog_duplication set false, log lines that would normally be written
2974 both the the main log and to the reject log were not written to syslog at
2977 PH/14 In the default configuration, change the use of "message" in ACL warn
2978 statements to "add_header".
2980 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2981 not followed by a command (e.g. "seen endif").
2983 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2984 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2987 PH/17 Added extra conditions to the default value of delay_warning_condition
2991 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2992 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2993 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2996 The Auto-Submitted: and various List- headers are standardised, whereas I
2997 don't think Precedence: ever was.
2999 PH/18 Refactored debugging code in route_finduser() to show more information,
3000 in particular, the error code if getpwnam() issues one.
3002 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3003 This is apparently needed in addition to the PH/07 change above to avoid
3004 any possible encoding problems.
3006 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3007 but not after initializing Perl.
3009 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3010 output them only if debugging. By default they are written stderr,
3011 apparently, which is not desirable.
3013 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3016 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3019 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3021 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3022 authenticated or an ident call has been made. Suppress the default
3023 values for $authenticated_id and $authenticated_sender (but permit -oMai
3024 and -oMas) when testing with -bh.
3026 PH/24 Re-jigged the order of the tests in the default configuration so that the
3027 tests for valid domains and recipients precede the DNS black list and CSA
3028 tests, on the grounds that those ones are more expensive.
3030 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3031 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3032 This bug exists in every version of Exim that I still have, right back to
3035 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3036 However, an attempt to turn on case-sensitivity in a regex key by
3037 including (?-i) didn't work because the subject string was already
3038 lowercased, and the effects were non-intuitive. It turns out that a
3039 one-line patch can be used to allow (?-i) to work as expected.
3045 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3046 other effects) broke the use of negated acl sub-conditions.
3048 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3051 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3052 "Deny" causes Exim to reject the incoming connection with a 554 error.
3053 Unfortunately, if there is a major crisis, such as a disk failure,
3054 tcp-wrappers gives "deny", whereas what one would like would be some
3055 kind of temporary error. A kludge has been added to help with this.
3056 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3057 554 error is used if errno is still zero or contains ENOENT (which occurs
3058 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3061 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3063 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3064 errors. Otherwise a message that provokes a temporary error (when other
3065 messages do not) can cause a whole host to time out.
3067 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3068 addresses were routed directly to files or pipes from a redirect router.
3069 File deliveries just didn't batch; pipe deliveries might have suffered
3072 PH/06 A failure to get a lock for a hints database would erroneously always say
3073 "Failed to get write lock", even when it was really a read lock.
3075 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3076 of 0600. This has been changed to use the value of the lockfile_mode
3077 option (which defaults to 0600).
3079 PH/08 Applied small patch from the Sieve maintainer.
3081 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3082 folder from quota calculations, a direct delivery into this folder messed
3083 up the contents of the maildirsize file. This was because the regex was
3084 used only to exclude .Trash (or whatever) when the size of the mailbox
3085 was calculated. There was no check that a delivery was happening into an
3086 excluded directory. This bug has been fixed by ignoring all quota
3087 processing for deliveries into excluded directories.
3089 PH/10 Added the maildirfolder_create_regex option to appendfile.
3095 PH/01 The code for finding all the local interface addresses on a FreeBSD
3096 system running IPv6 was broken. This may well have applied to all BSD
3097 systems, as well as to others that have similar system calls. The broken
3098 code found IPv4 interfaces correctly, but gave incorrect values for the
3099 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3100 that it would not match correctly against @[] and not recognize the IPv6
3103 PH/02 The ipliteral router was not recognizing addresses of the form user@
3104 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3106 PH/03 Added disable_ipv6.
3108 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3109 decoded form, because it is most often used to construct To: headers
3110 lines in autoreplies, and the decoded form may well be syntactically
3111 invalid. However, $reply_address has leading white space removed, and all
3112 newlines turned into spaces so that the autoreply transport does not
3115 PH/05 If group was specified without a user on a router, and no group or user
3116 was specified on a transport, the group from the router was ignored.
3118 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3119 for visible compile-time settings that can be used to change these
3120 numbers, for those that want even more. Backwards compatibility with old
3121 spool files has been maintained. However, going back to a previous Exim
3122 release will lost any variables that are in spool files.
3124 PH/07 Two small changes when running in the test harness: increase delay when
3125 passing a TCP/IP connection to a new process, in case the original
3126 process has to generate a bounce, and remove special handling of
3127 127.0.0.2 (sic), which is no longer necessary.
3129 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3130 be the same on different OS.
3132 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3135 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3136 whether --show-vars was specified or not
3138 JJ/02 exipick: Added support for new ACL variable spool format introduced
3141 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3142 syntactically invalid From: or Reply-to: line, and a filter used this to
3143 generate an autoreply, and therefore failed to obtain an address for the
3144 autoreply, Exim could try to deliver to a non-existent relative file
3145 name, causing unrelated and misleading errors. What now happens is that
3146 it logs this as a hard delivery error, but does not attempt to create a
3149 PH/11 The exinext utility has a -C option for testing purposes, but although
3150 the given file was scanned by exinext itself; it wasn't being passed on
3151 when Exim was called.
3153 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3154 an end-of-file indication when reading a command response.
3156 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3157 compiled. In many other places in Exim, IPv6 addresses are always
3158 recognized, so I have changed this. It also means that IPv4 domain
3159 literals of the form [IPV4:n.n.n.n] are now always recognized.
3161 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3162 used if the router is not running as root, for example, when verifying at
3163 ACL time, or when using -bh. The debugging output from this situation was
3164 non-existent - all you got was a failure to exec. I have made two
3167 (a) Failures to set uid/gid, the current directory, or a process leader
3168 in a subprocess such as that created by queryprogram now generate
3169 suitable debugging output when -d is set.
3171 (b) The queryprogram router detects when it is not running as root,
3172 outputs suitable debugging information if -d is set, and then runs
3173 the subprocess without attempting to change uid/gid.
3175 PH/15 Minor change to Makefile for building test_host (undocumented testing
3178 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3179 additional section of a DNS packet that returns MX or SRV records.
3180 Instead, it always explicitly searches for A/AAAA records. This avoids
3181 major problems that occur when a DNS server includes only records of one
3182 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3183 fixed another bug: if SRV records were looked up and the corresponding
3184 address records were *not* found in the additional section, the port
3185 values from the SRV records were lost.
3187 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3188 using the correct key (the original address) when searching the retry
3189 rules in order to find which one to use for generating the retry hint.
3191 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3192 adding the default one. Similarly, if it contains a Reply-To: header, the
3193 errors_reply_to option, if set, is not used.
3195 PH/19 When calculating a retry time, Exim used to measure the "time since
3196 failure" by looking at the "first failed" field in the retry record. Now
3197 it does not use this if it is later than than the arrival time of the
3198 message. Instead it uses the arrival time. This makes for better
3199 behaviour in cases where some deliveries succeed, thus re-setting the
3200 "first failed" field. An example is a quota failure for a huge message
3201 when small messages continue to be delivered. Without this change, the
3202 "time since failure" will always be short, possible causing more frequent
3203 delivery attempts for the huge message than are intended.
3204 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3206 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3207 $1, $2, $3) because the numerical variables can be reset during some
3208 expansion items (e.g. "match"), thereby losing the authentication data.
3210 PH/21 Make -bV show the size of off_t variables so that the test suite can
3211 decide whether to run tests for quotas > 2G.
3213 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3214 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3215 filecount value is greater than 2G or if a quota value is greater than 2G
3216 on a system where the size of off_t is not greater than 4, a panic error
3219 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3220 never match. The debug and -bh output now contains an explicit error
3221 message indicating a malformed IPv4 address or mask.
3223 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3224 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3225 PH/23 above applies.
3227 PH/25 Do not write to syslog when running in the test harness. The only
3228 occasion when this arises is a failure to open the main or panic logs
3229 (for which there is an explicit test).
3231 PH/26 Added the /no_tell option to "control=freeze".
3233 PH/27 If a host name lookup failed very early in a connection, for example, if
3234 the IP address matched host_lookup and the reverse lookup yielded a name
3235 that did not have a forward lookup, an error message of the form "no IP
3236 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3237 could be logged. Now it outputs the IP address instead of "NULL".
3239 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3240 allows the sender and the authenticated sender to be set when
3241 submitting a message from within Exim. Since child_open_exim() is
3242 documented for local_scan(), the new function should be too.
3244 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3245 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3246 results in an empty string is now treated as unset.
3248 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3250 PH/31 Added sender_verify_fail logging option.
3252 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3253 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3254 tidied the source and removed it altogether.
3256 PH/33 When a queue run was abandoned because the load average was too high, a
3257 log line was always written; now it is written only if the queue_run log
3258 selector is set. In addition, the log line for abandonment now contains
3259 information about the queue run such as the pid. This is always present
3260 in "start" and "stop" lines but was omitted from the "abandon" line.
3262 PH/34 Omit spaces between a header name and the colon in the error message that
3263 is given when verify = headers_syntax fails (if there are lots of them,
3264 the message gets confusing).
3266 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3267 names, as there are now some PTR records that contain slashes. This check
3268 is only to protect against broken name servers that fall over on strange
3269 characters, so the fact that it applies to all lookups doesn't matter.
3271 PH/36 Now that the new test suite is complete, we can remove some of the
3272 special code in Exim that was needed for the old test suite. For example,
3273 sorting DNS records because real resolvers return them in an arbitrary
3274 order. The new test suite's fake resolver always returns records in the
3277 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3278 bounces) except when queue_only is set, to avoid logging races between
3279 the different processes.
3281 PH/38 Panic-die if .include specifies a non-absolute path.
3283 PH/39 A tweak to the "H" retry rule from its user.
3285 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3286 a label. They prevented compilation on older perls.
3288 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3289 a warning to be raised on newish perls.
3291 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3292 on queue. Changes to match documented behaviour of showing count of
3293 messages matching specified criteria.
3295 PH/40 Changed the default ident timeout from 30s to 5s.
3297 PH/41 Added support for the use of login_cap features, on those BSD systems
3298 that have them, for controlling the resources used by pipe deliveries.
3300 PH/42 The content-scanning code uses fopen() to create files in which to put
3301 message data. Previously it was not paying any attention to the mode of
3302 the files. Exim runs with umask(0) because the rest of the code creates
3303 files with open(), and sets the required mode explicitly. Thus, these
3304 files were ending up world-writeable. This was not a big issue, because,
3305 being within the spool directory, they were not world-accessible. I have
3306 created a function called modefopen, which takes an additional mode
3307 argument. It sets umask(777), creates the file, chmods it to the required
3308 mode, then resets the umask. All the relevant calls to fopen() in the
3309 content scanning code have been changed to use this function.
3311 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3312 to 24 hours. This avoids potential overflow problems when processing G
3313 and H retry rules. I suspect nobody ever tinkers with this value.
3315 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3317 PH/45 When the plaintext authenticator is running as a client, the server's
3318 challenges are checked to ensure they are valid base64 strings. By
3319 default, the authentication attempt is cancelled if an invalid string is
3320 received. Setting client_ignore_invalid_base64 true ignores these errors.
3321 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3322 they are received. Thus, the responses can be made to depend on the
3323 challenges. If an invalid string is ignored, an empty string is placed in
3326 PH/46 Messages that are created by the autoreply transport now contains a
3327 References: header, in accordance with RFCs 2822 and 3834.
3329 PH/47 Added authenticated_sender_force to the smtp transport.
3331 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3333 PH/49 Installed latest patch from the Sieve maintainer.
3335 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3336 was also set, the appendfile transport was unnecessarily scanning a
3337 directory of message files (e.g. for maildir delivery) to find the count
3338 of files (along with the size), even though it did not need this
3339 information. It now does the scan only if it needs to find either the
3340 size of the count of files.
3342 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3344 PH/52 Two bugs concerned with error handling when the smtp transport is
3347 (i) Exim was not creating retry information for temporary errors given
3348 for individual recipients after the DATA command when the smtp transport
3349 was used in LMTP mode. This meant that they could be retried too
3350 frequently, and not timed out correctly.
3352 (ii) Exim was setting the flag that allows error details to be returned
3353 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3354 recipients that were returned after the DATA command.
3356 PH/53 This is related to PH/52, but is more general: for any failing address,
3357 when detailed error information was permitted to be returned to the
3358 sender, but the error was temporary, then after the final timeout, only
3359 "retry timeout exceeded" was returned. Now it returns the full error as
3360 well as "retry timeout exceeded".
3362 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3363 do this, and (what is worse) MTAs that accept it.
3365 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3366 will now be deprecated.
3368 PH/56 New os.c-cygwin from the Cygwin maintainer.
3370 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3371 formats (previously only available in exim formats via -bpr, -bpru,
3372 and -bpra. Now also available in native and exiqgrep formats)
3374 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3375 with very large, slow to parse queues
3377 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3379 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3381 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3382 responses to authentication challenges, though it was showing the
3383 challenges; (ii) I've removed the CR characters from the debug output for
3386 PH/58 Allow for the insertion of a newline as well as a space when a string
3387 is turned into more than one encoded-word during RFC 2047 encoding. The
3388 Sieve code now uses this.
3390 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3391 data_4xx, lost_connection, tls_required.
3393 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3394 message was being sent as an SMTP response.
3396 PH/61 Add -l and -k options to exicyclog.
3398 PH/62 When verifying, if an address was redirected to one new address, so that
3399 verification continued, and the new address failed or deferred after
3400 having set something in $address_data, the value of $address_data was not
3401 passed back to the ACL. This was different to the case when no
3402 redirection occurred. The value is now passed back in both cases.
3404 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3405 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3406 use login_cap.h, so on its own it isn't the distinguishing feature. The
3407 new name refers directly to the setclassresources() function.
3409 PH/65 Added configuration files for NetBSD3.
3411 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3413 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3414 is preferred over IPv4.
3416 PH/68 The bounce_return_message and bounce_return_body options were not being
3417 honoured for bounces generated during the reception of non-SMTP messages.
3418 In particular, this applied to messages rejected by the ACL. This bug has
3419 been fixed. However, if bounce_return_message is true and bounce_return_
3420 body is false, the headers that are returned for a non-SMTP message
3421 include only those that have been read before the error was detected.
3422 (In the case of an ACL rejection, they have all been read.)
3424 PH/69 The HTML version of the specification is now built in a directory called
3425 spec_html instead of spec.html, because the latter looks like a path with
3426 a MIME-type, and this confuses some software.
3428 PH/70 Catch two compiler warnings in sieve.c.
3430 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3431 function verify_get_ident() calls ip_connect() to connect a socket, but
3432 if the "connect()" function timed out, ip_connect() used to close the
3433 socket. However, verify_get_ident() also closes the socket later, and in
3434 between Exim writes to the log, which may get opened at this point. When
3435 the socket was closed in ip_connect(), the log could get the same file
3436 descriptor number as the socket. This naturally causes chaos. The fix is
3437 not to close the socket in ip_connect(); the socket should be closed by
3438 the function that creates it. There was only one place in the code where
3439 this was missing, in the iplookup router, which I don't think anybody now
3440 uses, but I've fixed it anyway.
3442 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3443 well as to direct DNS lookups. Otherwise the handling of names in host
3444 lists is inconsistent and therefore confusing.
3450 PH/01 Two changes to the default runtime configuration:
3452 (1) Move the checks for relay_from_hosts and authenticated clients from
3453 after to before the (commented out) DNS black list checks.
3455 (2) Add control=submission to the relay_from_hosts and authenticated
3456 clients checks, on the grounds that messages accepted by these
3457 statements are most likely to be submissions.
3459 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3461 (1) Generate an error if the third argument for the ${prvs expansion is
3464 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3467 (3) Reset the variables that are obtained from the first argument of
3468 ${prvscheck and used in the second argument before leaving the code,
3469 because their memory is reclaimed, so using them afterwards may do
3472 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3473 one (it's much easier than Tom thought :-).
3475 (5) Because of (4), we can now allow for the use of $prvscheck_result
3476 inside the third argument.
3478 PH/03 For some reason, the default setting of PATH when running a command from
3479 a pipe transport was just "/usr/bin". I have changed it to
3482 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3483 anything to be listed in the output from -bV.
3485 PH/05 When a filter generated an autoreply, the entire To: header line was
3486 quoted in the delivery log line, like this:
3488 => >A.N.Other <ano@some.domain> <original@ddress> ...
3490 This has been changed so that it extracts the operative address. There
3491 may be more than one such address. If so, they are comma-separated, like
3494 => >ano@some.domain,ona@other.domain <original@ddress> ...
3496 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3497 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3498 not being looked up in the rDNS to get a host name, Exim was showing the
3499 IP address twice in Received: lines, even though the IP addresses were
3500 identical. For example:
3502 Received: from [1.2.3.4] (helo=[1.2.3.4])
3504 However, if the real host name was known, it was omitting the HELO data
3505 if it matched the actual IP address. This has been tidied up so that it
3506 doesn't show the same IP address twice.
3508 PH/07 When both +timestamp and +memory debugging was on, the value given by
3509 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3510 called by the debug printing, thereby overwriting the timestamp buffer.
3511 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3513 PH/08 When the original message was included in an autoreply transport, it
3514 always said "this is a copy of the message, including all the headers",
3515 even if body_only or headers_only was set. It now gives an appropriate
3518 PH/09 Applied a patch from the Sieve maintainer which:
3520 o fixes some comments
3521 o adds the (disabled) notify extension core
3522 o adds some debug output for the result of if/elsif tests
3523 o points to the current vacation draft in the documentation
3524 and documents the missing references header update
3528 o fixes a bug in processing the envelope test (when testing
3529 multiple envelope elements, the last element determined the
3532 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3533 Electronic Mail") by including:
3535 Auto-submitted: auto-generated
3537 in the messages that it generates (bounce messages and others, such as
3538 warnings). In the case of bounce messages for non-SMTP messages, there was
3539 also a typo: it was using "Auto_submitted" (underscore instead of
3540 hyphen). Since every message generated by Exim is necessarily in response
3541 to another message, thes have all been changed to:
3543 Auto-Submitted: auto-replied
3545 in accordance with these statements in the RFC:
3547 The auto-replied keyword:
3549 - SHOULD be used on messages sent in direct response to another
3550 message by an automatic process,
3552 - MUST NOT be used on manually-generated messages,
3554 - MAY be used on Delivery Status Notifications (DSNs) and Message
3555 Disposition Notifications (MDNs),
3557 - MUST NOT be used on messages generated by automatic or periodic
3558 processes, except for messages which are automatic responses to
3561 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3562 to the default Received: header definition.
3564 PH/12 Added log selector acl_warn_skipped (default on).
3566 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3567 variables because (a) they are in the wrong storage pool and (b) even if
3568 they were copied, it wouldn't work properly because of the caching.
3570 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3571 checking when decoding. Apparently there are clients that generate
3572 overlong encoded strings. Why am I not surprised?
3574 PH/15 If the first argument of "${if match_address" was not empty, but did not
3575 contain an "@" character, Exim crashed. Now it writes a panic log message
3576 and treats the condition as false.
3578 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3580 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3581 extension "envelope-auth". The code is finished and in agreement with
3582 other implementations, but there is no documentation so far and in fact,
3583 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3584 not changing the active code.
3586 Print executed "if" and "elsif" statements when debugging is used. This
3587 helps a great deal to understand what a filter does.
3589 Document more things not specified clearly in RFC3028. I had all this
3590 sorted out, when out of a sudden new issues came to my mind. Oops."
3592 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3595 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3596 canonical form (as documented). However, after a host name lookup from
3597 the IP address, check_host() was doing a simple string comparison with
3598 addresses acquired from the DNS when checking that the found name did
3599 have the original IP as one of its addresses. Since any found IPv6
3600 addresses are likely to be in abbreviated form, the comparison could
3601 fail. Luckily, there already exists a function for doing the comparison
3602 by converting both addresses to binary, so now that is used instead of
3603 the text comparison.
3605 PH/20 There was another similar case to PH/19, when a complete host name was
3606 given in a host list; looking up its IP address could give an abbreviated
3607 form, whereas the current host's name might or might not be abbreviated.
3608 The same fix has been applied.
3614 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3615 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3618 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3619 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3621 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3623 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3624 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3625 Also, for "Auto-Submitted", treat anything other than "no" as
3626 non-personal, in accordance with RFC 3834. (Previously it treated
3627 anything starting "auto-" as non-personal.)
3629 TF/01 The control=submission/name=... option had a problem with syntax
3630 errors if the name included a slash character. The /name= option
3631 now slurps the rest of the string, so it can include any characters
3632 but it must come last in the list of options (after /sender_retain
3635 PH/05 Some modifications to the interface to the fake nameserver for the new
3643 TK/01 Added the "success_on_redirect" address verification option. See
3644 NewStuff for rationale and an example.
3646 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3648 PH/02 Patch to exigrep to allow it to work on syslog lines.
3650 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3651 fread() to skip over the body file's header line, because in Cygwin the
3652 header line is locked and is inaccessible.
3654 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3655 co-exist for some time) to make it clear that it is the Exim ID that is
3656 referenced, not the Message-ID: header line.
3658 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3659 string_format() function, because snprintf() does not exist on all
3662 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3663 ${stat: expansion item.
3665 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3666 protocol synchronization error", to keep the pedants happy.
3668 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3669 well as for IRIX systems, when gcc is being used. See the host.c source
3672 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3674 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3677 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3678 transport and to the smtp transport in LMTP mode.
3680 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3682 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3683 run a filter in a subprocess. This could lead to confusion in subsequent
3684 lookups in the parent process. There should also be a search_tidyup() at
3685 the end of the subprocess.
3687 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3688 only if the host matched helo_try_verify_hosts, which caused the
3689 verification to occur when the EHLO/HELO command was issued. The ACL just
3690 tested the remembered result. Now, if a previous verification attempt has
3691 not happened, "verify = helo" does it there and then.
3693 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3695 TK/03 Fix log output including CR from clamd.
3697 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3698 exist provoked a memory error which could cause a segfault.
3700 PH/15 Installed PCRE 6.2
3702 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3704 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3705 of the problem. Specifically, suggested +O2 rather than +O1 for the
3708 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3710 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3711 record, the port was not correctly logged when the outgoing_port log
3712 selector was set (it logged the transort's default port).
3714 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3715 fallback_hosts, and "hosts" in the smtp transport.
3717 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3718 host errors such as "Connection refused".
3720 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3721 authentication with radiusclient 0.4.9:
3723 - Error returned from rc_read_config was caught wrongly
3724 - Username/password not passed on to radius server due to wrong length.
3726 The presumption is that some radiusclient API changes for 4.51/PH/17
3727 were not taken care of correctly. The code is still untested by me (my
3728 Linux distribution still has 0.3.2 of radiusclient), but it was
3729 contributed by a Radius user.
3731 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3732 expanding the "port" option of the smtp transport.
3734 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3735 while reading a MIME header. Thanks to Tom Hughes for a patch.
3737 PH/24 Include config.h inside local_scan.h so that configuration settings are
3740 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3741 that in an ACL for RCPT (for example), you can examine exactly what was
3744 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3745 commands, but it was not correctly comparing the address with the actual
3746 client host address. Thus, it would show the EHLO address in Received:
3747 header lines when this was not necessary.
3749 PH/27 Added the % operator to ${eval:}.
3751 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3752 it should be ignoring failures (because with -C, for example, it has lost
3753 privilege). It wasn't ignoring creation failures other than "already
3756 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3759 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3760 because an input line was too long, either on its own, or by virtue of
3761 too many continuations, the temporary file was not being removed, and the
3762 return code was incorrect.
3764 PH/31 Missing "BOOL" in function definition in filtertest.c.
3766 PH/32 Applied Sieve patches from the maintainer.
3768 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3770 PH/33 Added "verify = not_blind".
3772 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3773 Local/Makefile (with some defaults set). These are used in built scripts
3774 such as exicyclog, but they have never been used in the exim_install
3775 script (though there are many overriding facilities there). I have
3776 arranged that the exim_install script now takes note of these two
3779 PH/35 Installed configuration files for Dragonfly.
3781 PH/36 When a locally submitted message by a trusted user did not contain a
3782 From: header, and the sender address was obtained from -f or from an SMTP
3783 MAIL command, and the trusted user did not use -F to supply a sender
3784 name, $originator_name was incorrectly used when constructing a From:
3785 header. Furthermore, $originator_name was used for submission mode
3786 messages from external hosts without From: headers in a similar way,
3787 which is clearly wrong.
3789 PH/37 Added control=suppress_local_fixups.
3791 PH/38 When log_selector = +received_sender was set, and the addition of the
3792 sender made the log line's construction buffer exactly full, or one byte
3793 less than full, an overflow happened when the terminating "\n" was
3796 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3797 when the result of a list match is failure because a DNS lookup failed.
3799 PH/40 RM_COMMAND is now used in the building process.
3801 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3802 the "build-* directories that it finds.
3804 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3805 address in a domain literal was a prefix of an interface address.
3807 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3808 when verifying a sender address, unless rewrite_headers is false.
3810 PH/44 Wrote a long comment about why errors_to addresses are verified as
3811 recipients, not senders.
3813 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3814 the ratelimit ACL was added.
3816 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3818 PH/46 Added extra information about PostgreSQL errors to the error string.
3820 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3821 avoiding the need to install special zones in a real server. This is
3822 backwards compatible; if it can't find the fake resolver, it drops back.
3823 Thus, both old and new test suites can be run.
3825 TF/02 Added util/ratelimit.pl
3827 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3828 clock is set back in time.
3830 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3831 Candler <B.Candler@pobox.com>.
3833 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3834 OK for addresses that are the result of redirections.
3836 PH/48 A number of further additions for the benefit of the new test suite,
3837 including a fake gethostbyname() that interfaces to the fake DNS resolver
3840 TF/06 The fix for widen_domains has also been applied to qualify_single and
3841 search_parents which are the other dnslookup options that can cause
3844 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3847 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3849 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3850 service" if it encountered an MX record with an empty target hostname.
3851 The message is now "an MX or SRV record indicated no SMTP service".
3853 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3854 if the DNS of the sending site is misconfigured. This is quite a
3855 common situation. This change restores the behaviour of treating a
3856 helo verification defer as a failure.
3858 PH/51 If self=fail was set on a router, the bounce message did not include the
3859 actual error message.
3865 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3867 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3868 command itself ended in error, the underlying message about the transport
3869 filter timeout was being overwritten with the pipe command error. Now the
3870 underlying error message should be appended to the second error message.
3872 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3874 PH/02 Reduce the amount of output that "make" produces by default. Full output
3875 can still be requested.
3877 PH/03 The warning log line about a condition test deferring for a "warn" verb
3878 was being output only once per connection, rather than after each
3879 occurrence (because it was using the same function as for successful
3880 "warn" verbs). This seems wrong, so I have changed it.
3882 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3883 it should not have, which might have caused a crash in the right
3884 circumstances, but probably never did.
3886 PH/04 Installed a modified version of Tony Finch's patch to make submission
3887 mode fix the return path as well as the Sender: header line, and to
3888 add a /name= option so that you can make the user's friendly name appear
3891 TF/03 Added the control = fakedefer ACL modifier.
3893 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3894 Mark Lowes for thorough testing.
3896 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3898 TK/03 Merged latest SRS patch from Miles Wilton.
3900 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3901 with the definition in sysexits.h (which is #included earlier).
3902 Fortunately, Exim does not actually use EX_OK. The code used to try to
3903 preserve the sysexits.h value, by assuming that macro definitions were
3904 scanned for macro replacements. I have been disabused of this notion,
3905 so now the code just undefines EX_OK before #including unistd.h.
3907 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3908 in the smtp transport. When a block could not be written in a single
3909 write() function, the timeout was being re-applied to each part-write.
3910 This seems wrong - if the receiver was accepting one byte at a time it
3911 would take for ever. The timeout is now adjusted when this happens. It
3912 doesn't have to be particularly precise.
3914 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3915 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3917 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3918 without a domain if the check to <postmaster@domain> fails.
3920 SC/01 Eximstats: added -xls and the ability to specify output files
3921 (patch written by Frank Heydlauf).
3923 SC/02 Eximstats: use FileHandles for outputting results.
3925 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3927 SC/04 Eximstats: fixed display of large numbers with -nvr option
3929 SC/05 Eximstats: fixed merging of reports with empty tables.
3931 SC/06 Eximstats: added the -include_original_destination flag
3933 SC/07 Eximstats: removed tabs and trailing whitespace.
3935 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3937 TK/06 MBOX spool code: Add real "From " MBOX separator line
3938 so the .eml file is really in mbox format (even though
3939 most programs do not really care). Patch from Alex Miller.
3941 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3942 The latter is generated from $received_to and is only set if the
3943 message has one envelope recipient. SA can use these headers,
3944 obviously out-of-the-box. Patch from Alex Miller.
3946 PH/08 The ${def test on a variable was returning false if the variable's
3947 value was "0", contrary to what the specification has always said!
3948 The result should be true unless the variable is empty.
3950 PH/09 The syntax error of a character other than { following "${if
3951 def:variable_name" (after optional whitespace) was not being diagnosed.
3952 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3953 accidental colon was present, for example, could give incorrect results.
3955 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3956 result is used (not including appendfile, where other changes are about
3959 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3960 This involved changing a lot of size variables from int to off_t. It
3961 should work with maildirs and everything.
3963 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3964 spamd dying while we are connected to it.
3966 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3969 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3970 function for BDB 4.3.
3972 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3974 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3975 the sources for PCRE have been re-arranged and more files are now
3978 PH/15 The code I had for printing potentially long long variables in PH/11
3979 above was not the best (it lost precision). The length of off_t variables
3980 is now inspected at build time, and an appropriate printing format (%ld
3981 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3982 to be "long long int" or "long int". This is needed for the internal
3983 formatting function string_vformat().
3985 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3986 the configuration file to be ":syslog", then the script "guesses" where
3987 the logs files are, rather than using the compiled in default. In our
3988 case the guess is not the same as the compiled default, so the script
3989 suddenly stopped working when I started to use syslog. The patch checks
3990 to see if log_file_path is "". If so, it attempts to read it from exim
3991 with no configuration file to get the compiled in version, before it
3992 falls back to the previous guessing code."
3994 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3995 implementing BATV in an Exim configuration. See NewStuff for the gory
3998 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3999 Makefile that are specific to HP-UX.
4001 PH/18 If the "use_postmaster" option was set for a recipient callout together
4002 with the "random" option, the postmaster address was used as the MAIL
4003 FROM address for the random test, but not for the subsequent recipient
4004 test. It is now used for both.
4006 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4007 patch removes a few documentation additions to RFC 3028, because the
4008 latest draft now contains them. It adds the new en;ascii-case comparator
4009 and a new error check for 8bit text in MIME parts. Comparator and
4010 require names are now matched exactly. I enabled the subaddress
4011 extension, but it is not well tested yet (read: it works for me)."
4013 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4014 rework some of the code of TK/09 above to avoid the hardwired use of
4015 "%lld" and "long long". Replaced the call to snprintf() with a call to
4018 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4019 records point to non-existent hosts", "retry timeout exceeded", and
4020 "retry time not reached for any host after a long failure period".
4022 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4023 experimental DomainKeys support:
4025 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4026 (2) On an error such as an illegally used "control", the wrong name for
4027 the control was given.
4029 These problems did NOT occur unless DomainKeys support was compiled.
4031 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4033 PH/24 Added ${if match_ip condition.
4035 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4036 databases so that it will be absolutely obvious if a crash occurs in the
4037 DB library. This is a regular occurrence (often caused by mis-matched
4040 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4041 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4042 picked up on a user's system that detects such things. There doesn't seem
4043 to be a gcc warning option for this - only an attribute that has to be
4044 put on the function's prototype. It seems that in Fedora Core 4 they have
4045 set this on a number of new functions. No doubt there will be more in due
4048 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4049 specify a transport. However, if an address that was verified by such a
4050 router was the subject of a callout, Exim crashed because it tried to
4051 read the rcpt_include_affixes from the non-existent transport. Now it
4052 just assumes that the setting of that option is false. This bug was
4053 introduced by 4.51/PH/31.
4055 PH/28 Changed -d+all to exclude +memory, because that information is very
4056 rarely of interest, but it makes the output a lot bigger. People tend to
4057 do -d+all out of habit.
4059 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4060 code in os-type was giving problems when libc.so lives in lib64, like on
4063 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4064 aren't the modern standard, and it seems that some systems' include files
4065 don't always have them. Exim was already checking for some of the newer
4066 ones like T_AAAA, and defining it itself. I've added checks for all the
4067 record types that Exim uses.
4069 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4070 not automatically generating a new one, as it is supposed to. This
4071 prevented TLS from working. If the file did exist, but contained invalid
4072 data, a new version was generated, as expected. It was only the case of a
4073 non-existent file that was broken.
4075 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4076 with a change in libdomainkeys > 0.64.
4078 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4079 from DNS. If the selector record carries the flag, it now has
4080 precedence over the domain-wide flag.
4082 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4084 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4085 the use of a port name that isn't defined in /etc/services) occurred, the
4086 message was deferred as in a normal delivery, and thus remained on the
4087 spool, instead of being failed because of the mua_wrapper setting. This
4088 is now fixed, and I tidied up some of the mua_wrapper messages at the
4091 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4092 the messages in a hash of arrays rather than using individual hashes.
4093 This is a bit cleaner and results in dramatic memory savings, albeit
4094 at a slight CPU cost.
4096 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4097 as requested by Marc Sherman.
4099 SC/10 Eximstats: added histograms for user specified patterns as requested
4102 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4104 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4105 fopen() in the content-scanning modules that did not already have it.
4111 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4112 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4114 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4116 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4118 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4119 to test Sieve filters that use "vacation".
4121 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4122 that changes the way the GnuTLS parameters are stored in the cache file.
4123 The new format can be generated externally. For backward compatibility,
4124 if the data in the cache doesn't make sense, Exim assumes it has read an
4125 old-format file, and it generates new data and writes a new file. This
4126 means that you can't go back to an older release without removing the
4129 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4130 work if there are any delivery delays because "one_time" forces the
4131 parent to be marked "delivered", so its unseen clone is never tried
4132 again. For this reason, Exim now forbids the simultaneous setting of
4135 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4136 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4137 entry if you want to know the details. The fix, however, neglected to
4138 consider the case when local delivery batching is involved. The test for
4139 "previously delivered" was not happening when checking to see if an
4140 address could be batched with a previous (undelivered) one; under
4141 certain circumstances this could lead to multiple deliveries to the same
4144 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4145 in its include files, and this causes problems building Exim.
4147 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4148 header_syntax) but Exim was just ignoring anything given after a slash.
4149 In particular, this caused confusion with an attempt to use "verify =
4150 reverse_host_lookup/defer_ok". An error is now given when options are
4151 supplied for verify items that do not have them. (Maybe reverse_host_
4152 lookup should have a defer_ok option, but that's a different point.)
4154 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4155 defined by RFC 821) to 2048, because there were problems with some AUTH
4156 commands, and RFC 1869 says the size should be increased for extended
4157 SMTP commands that take arguments.
4159 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4162 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4163 "unknown" error; now it says that the functionality isn't in the binary.
4165 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4166 an address' error message when a string expansion fails (syntax or
4167 whatever). Otherwise the password may appear in the log. Following change
4168 PH/42 below, there is no longer a chance of it appearing in a bounce
4171 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4173 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4174 of its addresses was ever tried. (Bugzilla bug #2.)
4176 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4177 the result incorrectly in the debug output. (It correctly added a newline
4178 to what was transported.)
4180 TF/01 Added $received_time.
4182 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4183 commented out examples of how to interface to a virus scanner and to
4184 SpamAssassin. Also added commented examples of av_scanner and
4185 spamd_address settings.
4187 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4188 and controls are allowed in which ACLs. There were a couple of minor
4189 errors. Some of the entries in the conditions table (which is a table of
4190 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4191 negation of where the condition IS allowed.
4193 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4195 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4196 header file does not have a version number, so I've had to invent a new
4197 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4198 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4199 radiusclient), but it was contributed by a Radius user.
4201 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4202 files or directories, for OpenSSL.
4204 PH/19 When an Exim process that is running as root has to create an Exim log
4205 file, it does so in a subprocess that runs as exim:exim so as to get the
4206 ownership right at creation (otherwise, other Exim processes might see
4207 the file with the wrong ownership). There was no test for failure of this
4208 fork() call, which would lead to the process getting stuck as it waited
4209 for a non-existent subprocess. Forks do occasionally fail when resources
4210 run out. I reviewed all the other calls to fork(); they all seem to check
4213 PH/20 When checking for unexpected SMTP input at connect time (before writing
4214 the banner), Exim was not dealing correctly with a non-positive return
4215 from the read() function. If the client had disconnected by this time,
4216 the result was a log entry for a synchronization error with an empty
4217 string after "input=" when read() returned zero. If read() returned -1
4218 (an event I could not check), uninitialized data bytes were printed.
4219 There were reports of junk text (parts of files, etc) appearing after
4222 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4224 PH/22 Added support for macro redefinition, and (re)definition in between
4225 driver and ACL definitions.
4227 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4228 forgetting to use the resulting value; it was using the unexpanded value.
4230 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4231 hadn't been configured. The fix is from Juergen Kreileder, who
4232 understands it better than I do:
4234 "Here's what I see happening with three configured cyrus_sasl
4235 authenticators configured (plain, login, cram-md5):
4237 On startup auth_cyrus_sasl_init() gets called for each of these.
4238 This means three calls to sasl_listmech() without a specified mech_list.
4239 => SASL tests which mechs of all available mechs actually work
4240 => three warnings about OTP not working
4241 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4243 With the patch, sasl_listmech() also gets called three times. But now
4244 SASL's mech_list option is set to the server_mech specified in the the
4245 authenticator. Or in other words, the answer from sasl_listmech()
4246 gets limited to just the mech you're testing for (which is different
4248 => the return list contains just 'plain' or 'login', 'cram-md5' or
4249 nothing depending on the value of ob->server_mech.
4251 I've just tested the patch: Authentication still works fine,
4252 unavailable mechs specified in the exim configuration are still
4253 caught, and the auth.log warnings about OTP are gone."
4255 PH/25 When debugging is enabled, the contents of the command line are added
4256 to the debugging output, even when log_selector=+arguments is not
4259 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4260 answer is "GNU", and only if the return is "GNU/something" is the answer
4263 PH/27 $acl_verify_message is now set immediately after the failure of a
4264 verification in an ACL, and so is available in subsequent modifiers. In
4265 particular, the message can be preserved by coding like this:
4267 warn !verify = sender
4268 set acl_m0 = $acl_verify_message
4270 Previously, $acl_verify_message was set only while expanding "message"
4271 and "log_message" when a very denied access.
4273 PH/28 Modified OS/os.c-Linux with
4275 -#ifndef OS_LOAD_AVERAGE
4276 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4278 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4279 nomenclature these days.)
4281 PH/29 Installed patch from the Sieve maintainer that adds the options
4282 sieve_useraddress and sieve_subaddress to the redirect router.
4284 PH/30 In these circumstances:
4285 . Two addresses routed to the same list of hosts;
4286 . First host does not offer TLS;
4287 . First host accepts first address;
4288 . First host gives temporary error to second address;
4289 . Second host offers TLS and a TLS session is established;
4290 . Second host accepts second address.
4291 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4292 and peerdn, if requested) that were in fact used only for the second
4295 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4296 attention to any local part prefix or suffix that was matched by the
4297 router that accepted the address. It now behaves in the same way as it
4298 does for delivery: the affixes are removed from the local part unless
4299 rcpt_include_affixes is set on the transport.
4301 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4302 timeout during the DATA phase of an incoming message.
4304 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4305 applied a patch sanctioned by the Sieve maintainer.
4307 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4308 the uid or gid is negative. A case of a negative gid caused this to be
4309 noticed. The fix allows for either to be negative.
4311 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4312 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4315 PH/36 The change PH/12 above was broken. Fixed it.
4317 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4318 the grounds that routing the same address twice would always produce the
4319 same answer. This might have been true once, but it is certainly no
4320 longer true now. Routing a child address may depend on the previous
4321 routing that produced that child. Some complicated redirection strategies
4322 went wrong when messages had multiple recipients, and made Exim's
4323 behaviour dependent on the order in which the addresses were given.
4325 I have moved the duplicate checking until after the routing is complete.
4326 Exim scans the addresses that are assigned to local and remote
4327 transports, and removes any duplicates. This means that more work will be
4328 done, as duplicates will always all be routed, but duplicates are
4329 presumably rare, so I don't expect this is of any significance.
4331 For deliveries to pipes, files, and autoreplies, the duplicate checking
4332 still happens during the routing process, since they are not going to be
4335 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4336 It corrects a timeout issue with spamd. This is Ian's comment: "The
4337 background is that sometimes spamd either never reads data from a
4338 connection it has accepted, or it never writes response data. The exiscan
4339 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4340 blindly assumes that writes won't block so it may never time out."
4342 PH/39 Allow G after quota size as well as K and M.
4344 PH/40 The value set for $authenticated_id in an authenticator may not contain
4345 binary zeroes or newlines because the value is written to log lines and
4346 to spool files. There was no check on this. Now the value is run through
4347 the string_printing() function so that such characters are converted to
4348 printable escape sequences.
4350 PH/41 $message_linecount is a new variable that contains the total number of
4351 lines in the message. Compare $body_linecount, which is the count for the
4354 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4355 bounce and delay warning messages, except in certain special cases, which
4358 (a) An SMTP error message from a remote host;
4359 (b) A message specified in a :fail: redirection;
4360 (c) A message specified in a "fail" command in a system filter;
4361 (d) A message specified in a FAIL return from the queryprogram router;
4362 (e) A message specified by the cannot_route_message router option.
4364 In these cases only, Exim does include the error details in bounce and
4365 warning messages. There are also a few cases where bland messages such
4366 as "unrouteable address" or "local delivery error" are given.
4368 PH/43 $value is now also set for the "else" part of a ${run expansion.
4370 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4371 being worked on, but at least Exim now implements the latest version to
4374 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4375 process to complete was treated as a delivery failure, a timeout while
4376 writing the message to the pipe was logged, but erroneously treated as a
4377 successful delivery. Such timeouts include transport filter timeouts. For
4378 consistency with the overall process timeout, these timeouts are now
4379 treated as errors, giving rise to delivery failures by default. However,
4380 there is now a new Boolean option for the pipe transport called
4381 timeout_defer, which, if set TRUE, converts the failures into defers for
4382 both kinds of timeout. A transport filter timeout is now identified in
4385 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4386 systems where "make" and "gmake" are different, calling "gmake" at top
4387 level broke things. I've arranged for the value of $(MAKE) to be passed
4388 from the Makefile to this script so that it can call the same version of
4392 A note about Exim versions 4.44 and 4.50
4393 ----------------------------------------
4395 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4396 changes of various kinds. As a consequence, a big documentation update was
4397 needed. This delayed the release for rather longer than seemed good, especially
4398 in the light of a couple of (minor) security issues. Therefore, the changes
4399 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4400 release. So 4.44 and 4.50 are in effect two different branches that both start
4403 I have left the 4.50 change log unchanged; it contains all the changes since
4404 4.43. The change log for 4.44 is below; many of its items are identical to
4405 those for 4.50. This seems to be the most sensible way to preserve the
4406 historical information.
4412 1. Minor wording change to the doc/README.SIEVE file.
4414 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4415 computation of the current number of files was incorrect.
4417 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4418 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4421 4. Give more explanation in the error message when the command for a transport
4422 filter fails to execute.
4424 5. There are several places where Exim runs a non-Exim command in a
4425 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4426 was being done only for the command run by the queryprogram router. It is
4427 now done for all such subprocesses. The other cases are: ${run, transport
4428 filters, and the commands run by the lmtp and pipe transports.
4430 6. Added CONFIGURE_GROUP build-time option.
4432 7. Some older OS have a limit of 256 on the maximum number of file
4433 descriptors. Exim was using setrlimit() to set 1000 as a large value
4434 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4435 systems. I've change it so that if it can't get 1000, it tries for 256.
4437 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4438 was an oversight, and furthermore, ever since the addition of extra
4439 controls (e.g. 4.43/32), the checks on when to allow different forms of
4440 "control" were broken. There should now be diagnostics for all cases when a
4441 control that does not make sense is encountered.
4443 9. Added the /retain_sender option to "control=submission".
4445 10. $recipients is now available in the predata ACL (oversight).
4447 11. Tidy the search cache before the fork to do a delivery from a message
4448 received from the command line. Otherwise the child will trigger a lookup
4449 failure and thereby defer the delivery if it tries to use (for example) a
4450 cached ldap connection that the parent has called unbind on.
4452 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4453 of $address_data from the recipient verification was clobbered by the
4454 sender verification.
4456 13. The value of address_data from a sender verification is now available in
4457 $sender_address_data in subsequent conditions in the ACL statement.
4459 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4461 15. Added a new option "connect=<time>" to callout options, to set a different
4464 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4465 was its contents. (It was OK if the option was not defined at all.)
4467 17. A "Completed" log line is now written for messages that are removed from
4468 the spool by the -Mrm option.
4470 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4471 information about exactly what failed.
4473 19. Added -dd to debug only the daemon process.
4475 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4476 handles timeouts, both on the server side and network timeouts. Renamed the
4477 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4479 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4480 It is now set to "smtps".
4482 22. $host_address is now set to the target address during the checking of
4483 ignore_target_hosts.
4485 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4486 being passed; this would have caused $sender_host_name to have been used if
4487 matching the list had actually called for a host name (not very likely,
4488 since this list is usually IP addresses). A host name is now passed as
4491 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4492 code that specifies a non-restarting handler (typically sigaction() in
4493 modern systems) in an attempt to fix a rare and obscure crash bug.
4495 25. Narrowed the window for a race in the daemon that could cause it to ignore
4496 SIGCHLD signals. This is not a major problem, because they are used only to
4497 wake it up if nothing else does.
4499 26. A malformed maildirsize file could cause Exim to calculate negative values
4500 for the mailbox size or file count. Odd effects could occur as a result.
4501 The maildirsize information is now recalculated if the size or filecount
4504 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4505 support for a long time. Removed HAVE_SYS_VFS_H.
4507 28. Installed the latest version of exipick from John Jetmore.
4509 29. In an address list, if the pattern was not a regular expression, an empty
4510 subject address (from a bounce message) matched only if the pattern was an
4511 empty string. Non-empty patterns were not even tested. This was the wrong
4512 because it is perfectly reasonable to use an empty address as part of a
4513 database query. An empty address is now tested by patterns that are
4514 lookups. However, all the other forms of pattern expect the subject to
4515 contain a local part and a domain, and therefore, for them, an empty
4516 address still always fails if the pattern is not itself empty.
4518 30. Exim went into a mad DNS loop when attempting to do a callout where the
4519 host was specified on an smtp transport, and looking it up yielded more
4520 than one IP address.
4522 31. Re-factored the code for checking spool and log partition space into a
4523 function that finds that data and another that does the check. The former
4524 is then used to implement four new variables: $spool_space, $log_space,
4525 $spool_inodes, and $log_inodes.
4527 32. The RFC2047 encoding function was originally intended for short strings
4528 such as real names; it was not keeping to the 75-character limit for
4529 encoded words that the RFC imposes. It now respects the limit, and
4530 generates multiple encoded words if necessary. To be on the safe side, I
4531 have increased the buffer size for the ${rfc2047: expansion operator from
4534 33. It is now permitted to omit both strings after an "if" condition; if the
4535 condition is true, the result is "true". As before, when the second string
4536 is omitted, a false condition yields an empty string. This makes it less
4537 cumbersome to write custom ACL and router conditions.
4539 34. Failure to deliver a bounce message always caused it to be frozen, even if
4540 there was an errors_to setting on the router. The errors_to setting is now
4543 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4544 canonical form (fully expanded) before being placed in
4545 $sender_host_address.
4547 36. The table in the code that translates DNS record types into text (T_A to
4548 "A" for instance) was missing entries for NS and CNAME. It is just possible
4549 that this could have caused confusion if both these types were looked up
4550 for the same domain, because the text type is used as part of Exim's
4551 per-process caching. But the chance of anyone hitting this buglet seems
4554 37. The dnsdb lookup has been extended in a number of ways.
4556 (1) There is a new type, "zns", which walks up the domain tree until it
4557 finds some nameserver records. It should be used with care.
4559 (2) There is a new type, "mxh", which is like "mx" except that it returns
4560 just the host names, not the priorities.
4562 (3) It is now possible to give a list of domains (or IP addresses) to be
4563 looked up. The behaviour when one of the lookups defers can be
4564 controlled by a keyword.
4566 (4) It is now possible to specify the separator character for use when
4567 multiple records are returned.
4569 38. The dnslists ACL condition has been extended: it is now possible to supply
4570 a list of IP addresses and/or domains to be looked up in a particular DNS
4573 39. Added log_selector=+queue_time_overall.
4575 40. When running the queue in the test harness, wait just a tad after forking a
4576 delivery process, to get repeatability of debugging output.
4578 41. Include certificate and key file names in error message when GnuTLS fails
4579 to set them up, because the GnuTLS error message doesn't include the name
4580 of the failing file when there is a problem reading it.
4582 42. Allow both -bf and -bF in the same test run.
4584 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4586 44. The "Exiscan patch" is now merged into the mainline Exim source.
4588 45. Sometimes the final signoff response after QUIT could fail to get
4589 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4590 < 0 before doing a fflush(). This bug looks as though it goes back to the
4591 introduction of TLS in release 3.20, but "sometimes" must have been rare
4592 because the tests only now provoked it.
4594 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4595 (this can affect the format of dates).
4597 47. exim_tidydb, when checking for the continued existence of a message for
4598 which it has found a message-specific retry record, was not finding
4599 messages that were in split spool directories. Consequently, it was
4600 deleting retry records that should have stayed in existence.
4602 48. Steve fixed some bugs in eximstats.
4604 49. The SPA authentication driver was not abandoning authentication and moving
4605 on to the next authenticator when an expansion was forced to fail,
4606 contradicting the general specification for all authenticators. Instead it
4607 was generating a temporary error. It now behaves as specified.
4609 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4610 (the order specifies the preference for clients). The order is now AES256,
4611 AES128, 3DES, ARCFOUR128.
4613 51. Small patch to Sieve code - explicitly set From: when generating an
4616 52. Exim crashed if a remote delivery caused a very long error message to be
4617 recorded - for instance if somebody sent an entire SpamAssassin report back
4618 as a large number of 550 error lines. This bug was coincidentally fixed by
4619 increasing the size of one of Exim's internal buffers (big_buffer) that
4620 happened as part of the Exiscan merge. However, to be on the safe side, I
4621 have made the code more robust (and fixed the comments that describe what
4624 53. Now that there can be additional text after "Completed" in log lines (if
4625 the queue_time_overall log selector is set), a one-byte patch to exigrep
4626 was needed to allow it to recognize "Completed" as not the last thing in
4629 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4630 patch that reportedly fixes this has been added. I am not expert enough to
4631 create a test for it. This is what the patch creator wrote:
4633 "I found a little strange behaviour of ldap code when working with
4634 Windows 2003 AD Domain, where users was placed in more than one
4635 Organization Units. When I tried to give exim partial DN, the exit code
4636 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4637 But simultaneously result of request was absolutely normal ldap result,
4638 so I produce this patch..."
4640 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4641 REFERENCE, so I have modified the code to exclude the patch when that macro
4644 55. Some experimental protocols are using DNS PTR records for new purposes. The
4645 keys for these records are domain names, not reversed IP addresses. The
4646 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4647 leaves it alone. Component reversal etc. now happens only for IP addresses.
4650 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4652 57. Double the size of the debug message buffer (to 2048) so that more of very
4653 long debug lines gets shown.
4655 58. The exicyclog utility now does better if the number of log files to keep
4656 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4658 59. Two changes related to the smtp_active_hostname option:
4660 (1) $smtp_active_hostname is now available as a variable.
4661 (2) The default for smtp_banner uses $smtp_active_hostname instead
4662 of $primary_hostname.
4664 60. The host_aton() function is supposed to be passed a string that is known
4665 to be a valid IP address. However, in the case of IPv6 addresses, it was
4666 not checking this. This is a hostage to fortune. Exim now panics and dies
4667 if the condition is not met. A case was found where this could be provoked
4668 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4669 components; fortuitously, this particular loophole had already been fixed
4670 by change 4.50/55 above.
4672 If there are any other similar loopholes, the new check in host_aton()
4673 itself should stop them being exploited. The report I received stated that
4674 data on the command line could provoke the exploit when Exim was running as
4675 exim, but did not say which command line option was involved. All I could
4676 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4677 running as the user.
4680 61. There was a buffer overflow vulnerability in the SPA authentication code
4681 (which came originally from the Samba project). I have added a test to the
4682 spa_base64_to_bits() function which I hope fixes it.
4685 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4686 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4688 63. The daemon start-up calls getloadavg() while still root for those OS that
4689 need the first call to be done as root, but it missed one case: when
4690 deliver_queue_load_max is set with deliver_drop_privilege. This is
4691 necessary for the benefit of the queue runner, because there is no re-exec
4692 when deliver_drop_privilege is set.
4694 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4695 This has been fixed.
4697 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4698 was in use, was not putting the data itself into the right store pool;
4699 consequently, it could be overwritten for a subsequent message in the same
4700 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4703 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4705 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4706 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4707 the function were treating the return as a boolean value, which happened to
4708 work because 0=false and not-0=true, but is not correct code.
4710 68. The host_aton() function was not handling scoped IPv6 addresses (those
4711 with, for example, "%eth0" on the end) correctly.
4713 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4714 negated items (that is, ~something) in unsigned ints. Some compilers
4715 apparently mutter when there is no cast.
4717 70. If an address verification called from an ACL failed, and did not produce a
4718 user-specific message (i.e. there was only a "system" message), nothing was
4719 put in $acl_verify_message. In this situation, it now puts the system
4722 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4723 session; change 4.31/43 added the unwanted input to the log line - except
4724 that it did not do this in the start of session case. It now does.
4726 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4727 This is wrong and can cause the other end to generate a synchronization
4728 error if it is another Exim or anything else that does the synchronization
4729 check. A QUIT command is no longer sent after a timeout.
4731 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4732 during host lookups.
4734 74. The defer_ok option of callout verification was not working if it was used
4735 when verifying addresses in header lines, that is, for this case:
4737 verify = header_sender/callout=defer_ok
4739 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4740 those file descriptors could be used for SMTP connections. If anything
4741 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4742 could be sent to the SMTP client, causing chaos. The daemon now opens
4743 stdin, stdout, and stderr to /dev/null when it puts itself into the
4746 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4747 log by default. The user can override this with suitable Perl magic.
4749 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4750 the log message when discard triggers, was not working for the DATA ACL or
4751 for the non-SMTP ACL.
4753 78. Error message wording change in sieve.c.
4755 79. If smtp_accept_max_per_host was set, the number of connections could be
4756 restricted to fewer than expected, because the daemon was trying to set up
4757 a new connection before checking whether the processes handling previous
4758 connections had finished. The check for completed processes is now done
4759 earlier. On busy systems, this bug wouldn't be noticed because something
4760 else would have woken the daemon, and it would have reaped the completed
4763 80. If a message was submitted locally by a user whose login name contained one
4764 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4765 It caused a spool format error. I have fixed the spool reading code. A
4766 related problem was that the "from" clause in the Received: line became
4767 illegal because of the space(s). It is now covered by ${quote_local_part.
4769 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4772 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4774 83. Patch from Sieve maintainer to fix unterminated string problem in
4775 "vacation" handling.
4777 84. Some minor changes to the Linux configuration files to help with other
4778 OS variants using glibc.
4780 85. One more patch for Sieve to update vacation handling to latest spec.
4783 ----------------------------------------------------
4784 See the note above about the 4.44 and 4.50 releases.
4785 ----------------------------------------------------
4791 1. Change 4.43/35 introduced a bug that caused file counts to be
4792 incorrectly computed when quota_filecount was set in an appendfile
4795 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4796 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4799 3. Give more explanation in the error message when the command for a transport
4800 filter fails to execute.
4802 4. There are several places where Exim runs a non-Exim command in a
4803 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4804 was being done only for the command run by the queryprogram router. It is
4805 now done for all such subprocesses. The other cases are: ${run, transport
4806 filters, and the commands run by the lmtp and pipe transports.
4808 5. Some older OS have a limit of 256 on the maximum number of file
4809 descriptors. Exim was using setrlimit() to set 1000 as a large value
4810 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4811 systems. I've change it so that if it can't get 1000, it tries for 256.
4813 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4814 was an oversight, and furthermore, ever since the addition of extra
4815 controls (e.g. 4.43/32), the checks on when to allow different forms of
4816 "control" were broken. There should now be diagnostics for all cases when a
4817 control that does not make sense is encountered.
4819 7. $recipients is now available in the predata ACL (oversight).
4821 8. Tidy the search cache before the fork to do a delivery from a message
4822 received from the command line. Otherwise the child will trigger a lookup
4823 failure and thereby defer the delivery if it tries to use (for example) a
4824 cached ldap connection that the parent has called unbind on.
4826 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4827 of $address_data from the recipient verification was clobbered by the
4828 sender verification.
4830 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4831 was its contents. (It was OK if the option was not defined at all.)
4833 11. A "Completed" log line is now written for messages that are removed from
4834 the spool by the -Mrm option.
4836 12. $host_address is now set to the target address during the checking of
4837 ignore_target_hosts.
4839 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4840 being passed; this would have caused $sender_host_name to have been used if
4841 matching the list had actually called for a host name (not very likely,
4842 since this list is usually IP addresses). A host name is now passed as
4845 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4846 code that specifies a non-restarting handler (typically sigaction() in
4847 modern systems) in an attempt to fix a rare and obscure crash bug.
4849 15. Narrowed the window for a race in the daemon that could cause it to ignore
4850 SIGCHLD signals. This is not a major problem, because they are used only to
4851 wake it up if nothing else does.
4853 16. A malformed maildirsize file could cause Exim to calculate negative values
4854 for the mailbox size or file count. Odd effects could occur as a result.
4855 The maildirsize information is now recalculated if the size or filecount
4858 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4859 support for a long time. Removed HAVE_SYS_VFS_H.
4861 18. Updated exipick to current release from John Jetmore.
4863 19. Allow an empty sender to be matched against a lookup in an address list.
4864 Previously the only cases considered were a regular expression, or an
4867 20. Exim went into a mad DNS lookup loop when doing a callout where the
4868 host was specified on the transport, if the DNS lookup yielded more than
4871 21. The RFC2047 encoding function was originally intended for short strings
4872 such as real names; it was not keeping to the 75-character limit for
4873 encoded words that the RFC imposes. It now respects the limit, and
4874 generates multiple encoded words if necessary. To be on the safe side, I
4875 have increased the buffer size for the ${rfc2047: expansion operator from
4878 22. Failure to deliver a bounce message always caused it to be frozen, even if
4879 there was an errors_to setting on the router. The errors_to setting is now
4882 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4883 canonical form (fully expanded) before being placed in
4884 $sender_host_address.
4886 24. Updated eximstats to version 1.33
4888 25. Include certificate and key file names in error message when GnuTLS fails
4889 to set them up, because the GnuTLS error message doesn't include the name
4890 of the failing file when there is a problem reading it.
4892 26. Expand error message when OpenSSL has problems setting up cert/key files.
4895 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4896 (this can affect the format of dates).
4898 28. exim_tidydb, when checking for the continued existence of a message for
4899 which it has found a message-specific retry record, was not finding
4900 messages that were in split spool directories. Consequently, it was
4901 deleting retry records that should have stayed in existence.
4903 29. eximstats updated to version 1.35
4904 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4905 1.35 - bugfix such that pie charts by volume are generated correctly
4907 30. The SPA authentication driver was not abandoning authentication and moving
4908 on to the next authenticator when an expansion was forced to fail,
4909 contradicting the general specification for all authenticators. Instead it
4910 was generating a temporary error. It now behaves as specified.
4912 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4913 (the order specifies the preference for clients). The order is now AES256,
4914 AES128, 3DES, ARCFOUR128.
4916 31. Small patch to Sieve code - explicitly set From: when generating an
4919 32. Exim crashed if a remote delivery caused a very long error message to be
4920 recorded - for instance if somebody sent an entire SpamAssassin report back
4921 as a large number of 550 error lines. This bug was coincidentally fixed by
4922 increasing the size of one of Exim's internal buffers (big_buffer) that
4923 happened as part of the Exiscan merge. However, to be on the safe side, I
4924 have made the code more robust (and fixed the comments that describe what
4927 33. Some experimental protocols are using DNS PTR records for new purposes. The
4928 keys for these records are domain names, not reversed IP addresses. The
4929 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4930 leaves it alone. Component reversal etc. now happens only for IP addresses.
4933 34. The host_aton() function is supposed to be passed a string that is known
4934 to be a valid IP address. However, in the case of IPv6 addresses, it was
4935 not checking this. This is a hostage to fortune. Exim now panics and dies
4936 if the condition is not met. A case was found where this could be provoked
4937 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4938 components; fortuitously, this particular loophole had already been fixed
4939 by change 4.50/55 or 4.44/33 above.
4941 If there are any other similar loopholes, the new check in host_aton()
4942 itself should stop them being exploited. The report I received stated that
4943 data on the command line could provoke the exploit when Exim was running as
4944 exim, but did not say which command line option was involved. All I could
4945 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4946 running as the user.
4949 35. There was a buffer overflow vulnerability in the SPA authentication code
4950 (which came originally from the Samba project). I have added a test to the
4951 spa_base64_to_bits() function which I hope fixes it.
4954 36. The daemon start-up calls getloadavg() while still root for those OS that
4955 need the first call to be done as root, but it missed one case: when
4956 deliver_queue_load_max is set with deliver_drop_privilege. This is
4957 necessary for the benefit of the queue runner, because there is no re-exec
4958 when deliver_drop_privilege is set.
4960 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4961 was in use, was not putting the data itself into the right store pool;
4962 consequently, it could be overwritten for a subsequent message in the same
4963 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4966 38. Sometimes the final signoff response after QUIT could fail to get
4967 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4968 < 0 before doing a fflush(). This bug looks as though it goes back to the
4969 introduction of TLS in release 3.20, but "sometimes" must have been rare
4970 because the tests only now provoked it.
4976 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4977 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4978 Now it returns an integer. A number of calls were still expecting a T/F
4979 return. Fortuitously, in all cases, the tests worked in OK situations,
4980 which is the norm. However, things would have gone wrong on any write
4981 failures on the smtp file descriptor. This function is used when sending
4982 messages over SMTP and also when doing verify callouts.
4984 2. When Exim is called to do synchronous delivery of a locally submitted
4985 message (the -odf or -odi options), it no longer closes stderr before doing
4988 3. Implemented the mua_wrapper option.
4990 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4992 5. Implemented the functions header_remove(), header_testname(),
4993 header_add_at_position(), and receive_remove_recipient(), and exported them
4996 6. If an ACL "warn" statement specified the addition of headers, Exim already
4997 inserted X-ACL-Warn: at the start if there was no header name. However, it
4998 was not making this test for the second and subsequent header lines if
4999 there were newlines in the string. This meant that an invalid header could
5000 be inserted if Exim was badly configured.
5002 7. Allow an ACL "warn" statement to add header lines at the start or after all
5003 the Received: headers, as well as at the end.
5005 8. Added the rcpt_4xx retry error code.
5007 9. Added postmaster_mailfrom=xxx to callout verification option.
5009 10. Added mailfrom=xxxx to the callout verification option, for verify=
5012 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5013 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5015 12. Inserted some casts to stop certain compilers warning when using pointer
5016 differences as field lengths or precisions in printf-type calls (mostly
5017 affecting debugging statements).
5019 13. Added optional readline() support for -be (dynamically loaded).
5021 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5022 same clock tick as a message's arrival, so that its received time was the
5023 same as the "first fail" time on the retry record, and that message
5024 remained on the queue past the ultimate address timeout, every queue runner
5025 would try a delivery (because it was past the ultimate address timeout) but
5026 after another failure, the ultimate address timeout, which should have then
5027 bounced the address, did not kick in. This was a "< instead of <=" error;
5028 in most cases the first failure would have been in the next clock tick
5029 after the received time, and all would be well.
5031 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5032 being recognized when the domain list was tested by the match_domain
5033 condition in an expansion string.
5035 16. Added the ${str2b64: operator.
5037 17. Exim was always calling setrlimit() to set a large limit for the number of
5038 processes, without checking whether the existing limit was already
5039 adequate. (It did check for the limit on file descriptors.) Furthermore,
5040 errors from getrlimit() and setrlimit() were being ignored. Now they are
5041 logged to the main and panic logs, but Exim does carry on, to try to do its
5042 job under whatever limits there are.
5044 18. Imported PCRE 5.0.
5046 19. Trivial typo in log message " temporarily refused connection" (the leading
5049 20. If the log selector return_path_on_delivery was set and an address was
5050 redirected to /dev/null, the delivery process crashed because it assumed
5051 that a return path would always be set for a "successful" delivery. In this
5052 case, the whole delivery is bypassed as an optimization, and therefore no
5055 21. Internal re-arrangement: the function for sending a challenge and reading
5056 a response while authentication was assuming a zero-terminated challenge
5057 string. It's now changed to take a pointer and a length, to allow for
5058 binary data in such strings.
5060 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5062 23. Exim was not respecting finduser_retries when seeking the login of the
5063 uid under which it was called; it was always trying 10 times. (The default
5064 setting of finduser_retries is zero.) Also, it was sleeping after the final
5065 failure, which is pointless.
5067 24. Implemented tls_on_connect_ports.
5069 25. Implemented acl_smtp_predata.
5071 26. If the domain in control=submission is set empty, Exim assumes that the
5072 authenticated id is a complete email address when it generates From: or
5073 Sender: header lines.
5075 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5076 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5077 chown and chgrp in /bin and hostname in /usr/bin.
5079 28. Exim was keeping the "process log" file open after each use, just as it
5080 does for the main log. This opens the possibility of it remaining open for
5081 long periods when the USR1 signal hits a daemon. Occasional processlog
5082 errors were reported, that could have been caused by this. Anyway, it seems
5083 much more sensible not to leave this file open at all, so that is what now
5086 29. The long-running daemon process does not normally write to the log once it
5087 has entered its main loop, and it closes the log before doing so. This is
5088 so that log files can straightforwardly be renamed and moved. However,
5089 there are a couple of unusual error situations where the daemon does write
5090 log entries, and I had neglected to close the log afterwards.
5092 30. The text of an SMTP error response that was received during a remote
5093 delivery was being truncated at 512 bytes. This is too short for some of
5094 the long messages that one sometimes sees. I've increased the limit to
5097 31. It is now possible to make retry rules that apply only when a message has a
5098 specific sender, in particular, an empty sender.
5100 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5101 it possible to be selective about when SMTP synchronization is enforced.
5103 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5105 32. Implemented hosts_connection_nolog.
5107 33. Added an ACL for QUIT.
5109 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5112 35. Added mailbox_size and mailbox_filecount to appendfile.
5114 36. Added control = no_multiline_responses to ACLs.
5116 37. There was a bug in the logic of the code that waits for the clock to tick
5117 in the case where the clock went backwards by a substantial amount such
5118 that the microsecond fraction of "now" was more than the microsecond
5119 fraction of "then" (but the whole seconds number was less).
5121 38. Added support for the libradius Radius client library this is found on
5122 FreeBSD (previously only the radiusclient library was supported).
5128 1. When certain lookups returned multiple values in the form name=value, the
5129 quoting of the values was not always being done properly. Specifically:
5130 (a) If the value started with a double quote, but contained no whitespace,
5132 (b) If the value contained whitespace other than a space character (i.e.
5133 tabs or newlines or carriage returns) it was not quoted.
5134 This fix has been applied to the mysql and pgsql lookups by writing a
5135 separate quoting function and calling it from the lookup code. The fix
5136 should probably also be applied to nisplus, ibase and oracle lookups, but
5137 since I cannot test any of those, I have not disturbed their existing code.
5139 2. A hit in the callout cache for a specific address caused a log line with no
5140 reason for rejecting RCPT. Now it says "Previous (cached) callout
5141 verification failure".
5143 3. There was an off-by-one bug in the queryprogram router. An over-long
5144 return line was truncated at 256 instead of 255 characters, thereby
5145 overflowing its buffer with the terminating zero. As well as fixing this, I
5146 have increased the buffer size to 1024 (and made a note to document this).
5148 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5149 when Exim is waiting for an SMTP response from a remote server, Exim
5150 restarts its select() call on the socket, thereby resetting its timeout.
5151 This is not a problem when such interrupts are rare. Somebody set up a cron
5152 job to run exiwhat every 2 minutes, which is less than the normal select()
5153 timeout (5 or 10 minutes). This meant that the select() timeout never
5154 kicked in because it was always reset. I have fixed this by comparing the
5155 time when an interrupt arrives with the time at the start of the first call
5156 to select(). If more time than the timeout has elapsed, the interrupt is
5157 treated as a timeout.
5159 5. Some internal re-factoring in preparation for the addition of Sieve
5160 extensions (by MH). In particular, the "personal" test is moved to a
5161 separate function, and given an option for scanning Cc: and Bcc: (which is
5162 not set for Exim filters).
5164 6. When Exim created an email address using the login of the caller as the
5165 local part (e.g. when creating a From: or Sender: header line), it was not
5166 quoting the local part when it contained special characters such as @.
5168 7. Installed new OpenBSD configuration files.
5170 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5171 try to make them clearer.
5173 9. Callout options, other than the timeout value, were being ignored when
5174 verifying sender addresses in header lines. For example, when using
5176 verify = header_sender/callout=no_cache
5178 the cache was (incorrectly) being used.
5180 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5181 only the Cygwin environment.
5183 11. When return_path_on_delivery was set as a log selector, if different remote
5184 addresses in the same message used different return paths and parallel
5185 remote delivery occurred, the wrong values would sometimes be logged.
5186 (Whenever a remote delivery process finished, the return path value from
5187 the most recently started remote delivery process was logged.)
5189 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5190 lines when AUTH and/or TLS are in use. This is the "received protocol"
5191 field. Exim used to use "asmtp" for authenticated SMTP, without any
5192 indication (in the protocol name) for TLS use. Now it follows the RFC and
5193 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5194 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5195 names appear in log lines as well as in Received: header lines.
5197 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5198 extensions, and comparison tests, and to fix some bugs.
5200 14. Changes to the "personal" filter test:
5202 (1) The test was buggy in that it was just doing the equivalent of
5203 "contains" tests on header lines. For example, if a user's address was
5204 anne@some.where, the "personal" test would incorrectly be true for
5206 To: susanne@some.where
5208 This test is now done by extracting each address from the header in turn,
5209 and checking the entire address. Other tests that are part of "personal"
5210 are now done using regular expressions (for example, to check local parts
5211 of addresses in From: header lines).
5213 (2) The list of non-personal local parts in From: addresses has been
5214 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5215 taken from the Sieve specification recommendations.
5217 (3) If the message contains any header line starting with "List-" it is
5218 treated as non-personal.
5220 (4) The test for "circular" in the Subject: header line has been removed
5221 because it now seems ill-conceived.
5223 15. Minor typos in src/EDITME comments corrected.
5225 16. Installed latest exipick from John Jetmore.
5227 17. If headers_add on a router specified a text string that was too long for
5228 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5229 of string_sprintf() is now avoided.
5231 18. $message_body_size was not set (it was always zero) when running the DATA
5232 ACL and the local_scan() function.
5234 19. For the "mail" command in an Exim filter, no default was being set for
5235 the once_repeat time, causing a random time value to be used if "once" was
5236 specified. (If the value happened to be <= 0, no repeat happened.) The
5237 default is now 0s, meaning "never repeat". The "vacation" command was OK
5238 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5239 (I found it when inspecting the code).
5241 20. There is now an overall timeout for performing a callout verification. It
5242 defaults to 4 times the callout timeout, which applies to individual SMTP
5243 commands during the callout. The overall timeout applies when there is more
5244 than one host that can be tried. The timeout is checked before trying the
5245 next host. This prevents very long delays if there are a large number of
5246 hosts and all are timing out (e.g. when the network connections are timing
5247 out). The value of the overall timeout can be changed by specifying an
5248 additional sub-option for "callout", called "maxwait". For example:
5250 verify = sender/callout=5s,maxwait=20s
5252 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5253 to the end before writing, but this should make it even safer).
5255 22. Exim was forgetting that it had advertised PIPELINING for the second and
5256 subsequent messages on an SMTP connection. It was also not resetting its
5257 memory on STARTTLS and an internal HELO.
5259 23. When Exim logs an SMTP synchronization error within a session, it now
5260 records whether PIPELINING has been advertised or not.
5262 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5263 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5264 rather than long int.
5266 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5268 26. Added the never_mail option to autoreply.
5274 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5275 crash if the getsockname() call failed; this can happen if a connection is
5276 closed very soon after it is established. The problem was simply in the
5277 order in which certain operations were done, causing Exim to try to write
5278 to the SMTP stream before it had set up the file descriptor. The bug has
5279 been fixed by making things happen in the correct order.
5285 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5286 before the connection was closed, thus losing the rejection response.
5288 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5289 some early Solaris releases, but causes trouble in current releases where
5290 socklen_t is defined.
5292 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5295 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5298 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5299 that releasing the top of it at the end releases what was used for sub-
5300 expansions (unless the block got too big). However, discard this block if
5301 the first thing is a variable or header, so that we can use its block when
5302 it is dynamic (useful for very large $message_headers, for example).
5304 6. Lookups now cache *every* query, not just the most recent. A new, separate
5305 store pool is used for this. It can be recovered when all lookup caches are
5306 flushed. Lookups now release memory at the end of their result strings.
5307 This has involved some general refactoring of the lookup sources.
5309 7. Some code has been added to the store_xxx() functions to reduce the amount
5310 of flapping under certain conditions.
5312 8. log_incoming_interface used to affect only the <= reception log lines. Now
5313 it causes the local interface and port to be added to several more SMTP log
5314 lines, for example "SMTP connection from", and rejection lines.
5316 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5318 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5320 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5321 could be overwritten at the end of the current message (or the start of a
5322 new message if it was set in a HELO ACL). The value is now preserved for
5323 the duration of the SMTP connection.
5325 12. If a transport had a headers_rewrite setting, and a matching header line
5326 contained an unqualified address, that address was qualified, even if it
5327 did not match any rewriting rules. The underlying bug was that the values
5328 of the flags that permit the existence of unqualified sender and recipient
5329 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5330 non-local messages, and by -bnq for local messages) were not being
5331 preserved with the message after it was received.
5333 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5334 "next input=" as part of the text comprising the host identity instead of
5335 the correct text. The code was using the same buffer for two different
5336 strings. However, depending on which order the printing function evaluated
5337 its arguments, the bug did not always show up. Under Linux, for example, my
5338 test suite worked just fine.
5340 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5341 doesn't work with some older versions of Perl. It has been changed to "my",
5342 which in any case is probably the better facility to use.
5344 15. A really picky compiler found some instances of statements for creating
5345 error messages that either had too many or two few arguments for the format
5348 16. The size of the buffer for calls to the DNS resolver has been increased
5349 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5350 for addresses that have a lot of PTR records. This alleviates a problem; it
5351 does not fully solve it.
5353 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5354 buffer now truncates the list and logs the incident, which is the same
5355 action as happens when Exim is looking up a host name and its aliases.
5356 Previously in this situation something unpredictable would happen;
5357 sometimes it was "internal error: store_reset failed".
5359 18. If a server dropped the connection unexpectedly when an Exim client was
5360 using GnuTLS and trying to read a response, the client delivery process
5361 crashed while trying to generate an error log message.
5363 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5364 string, for example:
5366 warn message = H1: something\nH2: something
5368 the text was added as a single header line from Exim's point of view
5369 though it ended up OK in the delivered message. However, searching for the
5370 second and subsequent header lines using $h_h2: did not work. This has been
5371 fixed. Similarly, if a system filter added multiple headers in this way,
5372 the routers could not see them.
5374 20. Expanded the error message when iplsearch is called with an invalid key to
5375 suggest using net-iplsearch in a host list.
5377 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5378 ACLs are no longer actually imposed (and a message to that effect is
5381 22. If a "gecos" field in a passwd entry contained escaped characters, in
5382 particular, if it contained a \" sequence, Exim got it wrong when building
5383 a From: or a Sender: header from that name. A second bug also caused
5384 incorrect handling when an unquoted " was present following a character
5385 that needed quoting.
5387 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5388 was not being matched caselessly.
5390 24. Arranged for all hyphens in the exim.8 source to be escaped with
5393 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5394 properly. Recipient callout cache records were still being keyed to include
5395 the sender, even when use_sender was set false. This led to far more
5396 callouts that were necessary. The sender is no longer included in the key
5397 when use_sender is false.
5399 26. Added "control = submission" modifier to ACLs.
5401 27. Added the ${base62d: operator to decode base 62 numbers.
5403 28. dnsdb lookups can now access SRV records.
5405 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5406 the configuration file.
5408 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5409 (-v) mode. This makes the output for a verbose queue run more intelligible.
5411 31. Added a use_postmaster feature to recipient callouts.
5413 32. Added the $body_zerocount variable, containing the number of binary zero
5414 bytes in the message body.
5416 33. The time of last modification of the "new" subdirectory is now used as the
5417 "mailbox time last read" when there is a quota error for a maildir
5420 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5422 35. Added +ignore_unknown as a special item in host lists.
5424 36. Code for decoding IPv6 addresses in host lists is now included, even if
5425 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5426 address was recognized as an IP address, but was then not correctly decoded
5427 into binary, causing unexpected and incorrect effects when compared with
5434 1. Very minor rewording of debugging text in manualroute to say "list of
5435 hosts" instead of "hostlist".
5437 2. If verify=header_syntax was set, and a header line with an unqualified
5438 address (no domain) and a large number of spaces between the end of the
5439 name and the colon was received, the reception process suffered a buffer
5440 overflow, and (when I tested it) crashed. This was caused by some obsolete
5441 code that should have been removed. The fix is to remove it!
5443 3. When running in the test harness, delay a bit after writing a bounce
5444 message to get a bit more predictability in the log output.
5446 4. Added a call to search_tidyup() just before forking a reception process. In
5447 theory, someone could use a lookup in the expansion of smtp_accept_max_
5448 per_host which, without the tidyup, could leave open a database connection.
5450 5. Added the variables $recipient_data and $sender_data which get set from a
5451 lookup success in an ACL "recipients" or "senders" condition, or a router
5452 "senders" option, similar to $domain_data and $local_part_data.
5454 6. Moved the writing of debug_print from before to after the "senders" test
5457 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5458 problems for message scanning, either using a data ACL, or using
5459 local_scan() because the Received: header was not generated till after they
5460 were called (in order to set the time as the time of reception completion).
5461 I have revised the way this works. The header is now generated after the
5462 body is received, but before the ACL or local_scan() are called. After they
5463 are run, the timestamp in the header is updated.
5469 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5470 before starting a queue runner without re-exec. This happened only when
5471 deliver_drop_privilege was set or when the Exim user was set to root. The
5472 effect of the bug was that timeouts during subsequent deliveries caused
5473 crashes instead of being properly handled. The handler is now left at its
5474 default (and expected) setting.
5476 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5477 message, again when deliver_drop_privilege is set or Exim is run as root.
5478 The bug described in (1) was not present in this case, but the tidying up
5479 of the other signals was missing. I have made the two cases consistent.
5481 3. The ignore_target_hosts setting on a manualroute router was being ignored
5482 for hosts that were looked up using the /MX notation.
5484 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5487 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5488 operated on the sender address. After changing the $sender_address to <>
5489 for the sender address verify, Exim was re-instated it as the original
5490 (before rewriting) address, but remembering that it had rewritten it, so it
5491 wasn't rewriting it again. This bug also had the effect of breaking the
5492 sender address verification caching when the sender address was rewritten.
5494 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5495 This has been changed so that if the ip literal address matches
5496 ignore_target_hosts, the router declines.
5498 7. Added expansion conditions match_domain, match_address, and match_local_
5499 part (NOT match_host).
5501 8. The placeholder for the Received: header didn't have a length field set.
5503 9. Added code to Exim itself and to exim_lock to test for a specific race
5504 condition that could lead to file corruption when using MBX delivery. The
5505 issue is with the lockfile that is created in /tmp. If this file is removed
5506 after a process has opened it but before that process has acquired a lock,
5507 there is the potential for a second process to recreate the file and also
5508 acquire a lock. This could lead to two Exim processes writing to the file
5509 at the same time. The added code performs the same test as UW imapd; it
5510 checks after acquiring the lock that its file descriptor still refers to
5511 the same named file.
5513 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5514 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5517 11. Added the smtp_active_hostname option. If used, this will typically be made
5518 to depend on the incoming interface address. Because $interface_address is
5519 not set up until the daemon has forked a reception process, error responses
5520 that can happen earlier (such as "too many connections") no longer contain
5523 12. If an expansion in a condition on a "warn" statement fails because a lookup
5524 defers, the "warn" statement is abandoned, and the next ACL statement is
5525 processed. Previously this caused the whole ACL to be aborted.
5527 13. Added the iplsearch lookup type.
5529 14. Added ident_timeout as a log selector.
5531 15. Added tls_certificate_verified as a log selector.
5533 16. Added a global option tls_require_ciphers (compare the smtp transport
5534 option of the same name). This controls incoming TLS connections.
5536 17. I finally figured out how to make tls_require_ciphers do a similar thing
5537 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5538 before starting the TLS session.
5540 18. Tabs are now shown as \t in -bP output.
5542 19. If the log selector return_path_on_delivery was set, Exim crashed when
5543 bouncing a message because it had too many Received: header lines.
5545 20. If two routers both had headers_remove settings, and the first one included
5546 a superfluous trailing colon, the final name in the first list and the
5547 first name in the second list were incorrectly joined into one item (with a
5548 colon in the middle).
5554 1. Added -C and -D options to the exinext utility, mainly to make it easier
5555 to include in the automated testing, but these could be helpful when
5556 multiple configurations are in use.
5558 2. The exinext utility was not formatting the output nicely when there was
5559 an alternate port involved in the retry record key, nor when there was a
5560 message id as well (for retries that were specific to a specific message
5561 and a specific host). It was also confused by IPv6 addresses, because of
5562 the additional colons they contain. I have fixed the IPv4 problem, and
5563 patched it up to do a reasonable job for IPv6.
5565 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5566 delivery, the log line now contains "pipelined" if PIPELINING was used.
5568 4. An SMTP transport process used to panic and die if the bind() call to set
5569 an explicit outgoing interface failed. This has been changed; it is now
5570 treated in the same way as a connect() failure.
5572 5. A reference to $sender_host_name in the part of a conditional expansion
5573 that was being skipped was still causing a DNS lookup. This no longer
5576 6. The def: expansion condition was not recognizing references to header lines
5577 that used bh_ and bheader_.
5579 7. Added the _cache feature to named lists.
5581 8. The code for checking quota_filecount in the appendfile transport was
5582 allowing one more file than it should have been.
5584 9. For compatibility with Sendmail, the command line option
5592 and sets the incoming protocol and host name (for trusted callers). The
5593 host name and its colon can be omitted when only the protocol is to be set.
5594 Note the Exim already has two private options, -pd and -ps, that refer to
5595 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5596 "s", but I don't think that's a major issue.
5598 10. A number of refactoring changes to the code, none of which should affect
5601 (a) The number of logging options was getting close to filling up the
5602 32-bit word that was used as a bit map. I have split them into two classes:
5603 those that are passed in the argument to log_write(), and those that are
5604 only ever tested independently outside of that function. These are now in
5605 separate 32-bit words, so there is plenty of room for expansion again.
5606 There is no change in the user interface or the logging behaviour.
5608 (b) When building, for example, log lines, the code previously used a
5609 macro that called string_cat() twice, in order to add two strings. This is
5610 not really sufficiently general. Furthermore, there was one instance where
5611 it was actually wrong because one of the argument was used twice, and in
5612 one call a function was used. (As it happened, calling the function twice
5613 did not affect the overall behaviour.) The macro has been replaced by a
5614 function that can join an arbitrary number of extra strings onto a growing
5617 (c) The code for expansion conditions now uses a table and a binary chop
5618 instead of a serial search (which was left over from when there were very
5619 few conditions). Also, it now recognizes conditions like "pam" even when
5620 the relevant support is not compiled in: a suitably worded error message is
5621 given if an attempt is made to use such a condition.
5623 11. Added ${time_interval:xxxxx}.
5625 12. A bug was causing one of the ddress fields not to be passed back correctly
5626 from remote delivery subprocesses. The field in question was not being
5627 subsequently used, so this caused to problems in practice.
5629 13. Added new log selectors queue_time and deliver_time.
5631 14. Might have fixed a bug in maildirsizefile handling that threw up
5632 "unexpected character" debug warnings, and recalculated the data
5633 unnecessarily. In any case, I expanded the warning message to give more
5636 15. Added the message "Restricted characters in address" to the statements in
5637 the default ACL that block characters like @ and % in local parts.
5639 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5640 Three changes have been made:
5642 (a) There was a serious bug; a negative response to MAIL caused the whole
5643 recipient domain to be cached as invalid, thereby blocking all messages
5644 to all local parts at the same domain, from all senders. This bug has
5645 been fixed. The domain is no longer cached after a negative response to
5646 MAIL if the sender used is not empty.
5648 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5651 (c) A new callout option, "use_sender" has been added for people who want
5652 the modified behaviour.
5658 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5661 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5662 indeed breaks things for older releases.
5664 3. Added additional logging to the case where there is a problem reading data
5665 from a filter that is running in a subprocess using a pipe, in order to
5666 try to track down a specific problem.
5668 4. Testing facility fudge: when running in the test harness and attempting
5669 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5670 getting "No route to host". Convert this to a timeout.
5672 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5675 6. Some OS don't have socklen_t but use size_t instead. This affects the
5676 fifth argument of getsockopt() amongst other things. This is now
5677 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5678 can be set for individual OS. I have set it for SunOS5, OSF1, and
5679 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5680 some earlier ones do not.
5682 7. Change 4.30/15 was not doing the test caselessly.
5684 8. The standard form for an IPv6 address literal was being rejected by address
5685 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5686 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5687 this, as well as the form without the "IPv6" on the front (but only when
5688 address literals are enabled, of course).
5690 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5692 10. Exim crashed if a message with an empty sender address specified by -f
5693 encountered a router with an errors_to setting. This could be provoked only
5694 by a command such as
5698 where an empty string was supplied; "<>" did not hit this bug.
5700 11. Installed PCRE release 4.5.
5702 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5703 remained set. It is now erased.
5705 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5706 times from message ids (which are base 36 rather than the normal 62).
5708 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5709 were being counted as actual protocol errors, and logged if the log
5710 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5711 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5712 and DATA following a set of rejected RCPTs do not count as protocol errors.
5713 In other words, Exim assumes they were pipelined, though this may not
5714 actually be the case. Of course, in all cases the client gets an
5715 appropriate error code.
5717 15. If a lookup fails in an ACL condition, a message about the failure may
5718 be available; it is used if testing the ACL cannot continue, because most
5719 such messages specify what the cause of the deferral is. However, some
5720 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5721 that caused an old message to be retained and used if a later statement
5722 caused a defer, replacing the real cause of the deferral.
5724 16. If an IP address had so many PTR records that the DNS lookup buffer
5725 was not large enough to hold them, Exim could crash while trying to process
5726 the truncated data. It now detects and logs this case.
5728 17. Further to 4.21/58, another change has been made: if (and only if) the
5729 first line of a message (the first header line) ends with CRLF, a bare LF
5730 in a subsequent header line has a space inserted after it, so as not to
5731 terminate the header.
5733 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5734 unnecessarily, used atoi() instead of strtol(), and didn't check the
5735 termination when getting file sizes from file names by regex.
5737 19. Completely re-implemented the support for maildirsize files, in the light
5738 of a number of problems with the previous contributed implementation
5739 (4.30/29). In particular:
5741 . If the quota is zero, the maildirsize file is maintained, but no quota is
5744 . If the maildir directory does not exist, it is created before any attempt
5745 to write a maildirsize file.
5747 . The quota value in the file is just a cache; if the quota is changed in
5748 the transport, the new value overrides.
5750 . A regular expression is available for excluding directories from the
5753 20. The autoreply transport checks the characters in options that define the
5754 message's headers; it allows continued headers, but it was checking with
5755 isspace() after an embedded newline instead of explicitly looking for a
5758 21. If all the "regular" hosts to which an address was routed had passed their
5759 expiry times, and had not reached their retry times, the address was
5760 bounced, even if fallback hosts were defined. Now Exim should go on to try
5763 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5764 equivalent code in the SMTP transport. Some hosts send humungous responses
5765 to HELO/EHLO, more than 1024 it seems.
5767 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5768 gives compiler warnings in some environments. I've now done it "properly",
5771 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5772 (because of problems with the built-in one) was declared to return uschar *
5773 instead of char *, causing compiler failure.
5775 25. Fixed a file descriptor leak when processing alias/forward files.
5777 26. Fixed a minor format string issue in dbfn.c.
5779 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5781 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5782 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5785 29. When a sender address is verified, it is cached, to save repeating the test
5786 when there is more than one recipient in a message. However, when the
5787 verification involves a callout, it is possible for different callout
5788 options to be set for different recipients. It is too complicated to keep
5789 track of this in the cache, so now Exim always runs a verification when a
5790 callout is required, relying on the callout cache for the optimization.
5791 The overhead is duplication of the address routing, but this should not be
5794 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5795 to be verified with callout=postmaster, and the main callout worked but the
5796 postmaster check failed, the verification correctly failed. However, if a
5797 subsequent RCPT command asked for sender verification *without* the
5798 postmaster check, incorrect caching caused this verification also to fail,
5801 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5802 it was not caching the DNS options (qualify_single, search_parents) that
5803 were used when the lookup failed. A subsequent lookup with different
5804 options therefore always gave the same answer, though there were cases
5805 where it should not have. (Example: a "domains = !$mx_any" option on a
5806 dnslookup router: the "domains" option is always processed without any
5807 widening, but the router might have qualify_single set.) Now Exim uses the
5808 cached value only when the same options are set.
5810 32. Added John Jetmore's "exipick" utility to the distribution.
5812 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5813 than a timeout (e.g. a certificate is required, and is not provided), an
5814 Exim server now closes the connection immediately. Previously it waited for
5815 the client to close - but if the client is SSL, it seems that they each
5816 wait for each other, leading to a delay before one of them times out.
5818 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5819 maintained 0.8.x compatibility because I don't think many are using it, and
5820 it is clearly obsolete.
5822 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5825 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5826 client certificate was expired. A simple patch fixes this, though I don't
5827 understand the full logic of why the verify callback is called multiple
5830 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5831 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5832 which causes problems with some clients (such as the Certicom SSL Plus
5833 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5834 disables the coutermeasure allowing Eudora to connect."
5836 38. Exim was not checking that a write() to a log file succeeded. This could
5837 lead to Bad Things if a log got too big, in particular if it hit a file
5838 size limit. Exim now panics and dies if it cannot write to a log file, just
5839 as it does if it cannot open a log file.
5841 39. Modified OS/Makefile-Linux so that it now contains
5843 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5845 The two -D definitions ensure that Exim is compiled with large file
5846 support, which makes it possible to handle log files that are bigger than
5849 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5850 instance) a domain was checked against a named list that involved a lookup,
5851 causing $domain_data to be set, then another domain was checked against the
5852 same list, then the first domain was re-checked, the value of $domain_data
5853 after the final check could be wrong. In particular, if the second check
5854 failed, it could be set empty. This bug probably also applied to
5857 41. The strip_trailing_dot option was not being applied to the address given
5858 with the -f command-line option.
5860 42. The code for reading a message's header from the spool was incrementing
5861 $received_count, but never initializing it. This meant that the value was
5862 incorrect (doubled) while delivering a message in the same process in which
5863 it was received. In the most common configuration of Exim, this never
5864 happens - a fresh exec is done - but it can happen when
5865 deliver_drop_privilege is set.
5867 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5868 it now includes up to 150 characters of the unexpected data in the log
5871 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5872 and building data strings. The size of both of these buffers was 10 000
5873 bytes - far larger than anybody would *ever* want, thought I. Needless to
5874 say, somebody hit the limit. I have increased the maximum line length to
5875 20 000 and the maximum data length of concatenated lines to 100 000. I have
5876 also fixed two bugs, because there was no checking on these buffers. Tsk,
5877 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5878 buffer is too small.
5880 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5881 lsearch lookups. Now it does.
5883 46. When parsing a route_list item in a manualroute router, a fixed-length
5884 buffer was used for the list of hosts. I made this 1024 bytes long,
5885 thinking that nobody would ever have a list of hosts that long. Wrong.
5886 Somebody had a whole pile of complicated expansion conditions, and the
5887 string was silently truncated, leading to an expansion error. It turns out
5888 that it is easier to change to an unlimited length (owing to other changes
5889 that have happened since this code was originally written) than to build
5890 structure for giving a limitation error. The length of the item that
5891 expands into the list of hosts is now unlimited.
5893 47. The lsearch lookup could not handle data where the length of text line was
5894 more than 4095 characters. Such lines were truncated, leading to shortened
5895 data being returned. It should now handle lines of any length.
5897 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5898 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5901 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5902 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5903 no longer happens in comment lines. A list of replacements is now placed
5904 at the head of all of the source files, except those whose only change is
5905 to replace PERL_COMMAND in the very first #! line.
5907 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5908 messages on the queue, with a bottom-up merge sort, using code contributed
5909 by Michael Haardt. This should make operations like -bp somewhat faster on
5910 large queues. It won't affect queue runners, except when queue_run_in_order
5913 51. Installed eximstats 1.31 in the distribution.
5915 52. Added support for SRV lookups to the dnslookup router.
5917 53. If an ACL referred to $message_body or $message_body_end, the value was not
5918 reset for any messages that followed in the same SMTP session.
5920 54. The store-handling optimization for building very long strings was not
5921 differentiating between the different store pools. I don't think this
5922 actually made any difference in practice, but I've tidied it.
5924 55. While running the routers to verify a sender address, $sender_address
5925 was still set to the sender address. This is wrong, because when routing to
5926 send a bounce to the sender, it would be empty. Therefore, I have changed
5927 it so that, while verifying a sender address, $sender_address is set to <>.
5928 (There is no change to what happens when verifying a recipient address.)
5930 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5931 target A or AAAA records (if not already returned) without resetting the
5932 qualify_single or search_parents options of the DNS resolver. These are
5933 inappropriate in this case because the targets of MX and SRV records must
5934 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5935 target that, when qualified, matched something in the local domain. These
5936 two options are now turned off when doing these lookups.
5938 57. It seems that at least some releases of Reiserfs (which does not have the
5939 concept of a fixed number of inodes) returns zero and not -1 for the
5940 number of available inodes. This interacted badly with check_spool_inodes,
5941 which assumed that -1 was the "no such thing" setting. What I have done is
5942 to check that the total number of inodes is greater than zero before doing
5943 the test of how many are available.
5945 58. When a "warn" ACL statement has a log_message modifier, the message is
5946 remembered, and not repeated. This is to avoid a lot of repetition when a
5947 message has many recipients that cause the same warning to be written.
5948 However, Exim was preserving the list of already written lines for an
5949 entire SMTP session, which doesn't seem right. The memory is now reset if a
5950 new message is started.
5952 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5953 debugging output unless log_rewrite was also set.
5955 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5956 of (int)(handle) when we know that handle contains (void *)(-1).
5958 61. The Exim daemon panic-logs an error return when it closes the incoming
5959 connection. However "connection reset by peer" seems to be common, and
5960 isn't really an error worthy of noting specially, so that particular error
5963 62. When Exim is trying to find all the local interfaces, it used to panic and
5964 die if the ioctl to get the interface flags failed. However, it seems that
5965 on at least one OS (Solaris 9) it is possible to have an interface that is
5966 included in the list of interfaces, but for which you get a failure error
5967 for this call. This happens when the interface is not "plumbed" into a
5968 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5969 failure of the "get flags" call assumes that the interface is down.
5971 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5972 makes life easier for people who are doing arithmetic on fields extracted
5973 from dates, where you often get leading zeros that should not be
5974 interpreted as octal.
5976 64. Added qualify_domain to the redirect router, to override the global
5979 65. If a pathologically long header line contained very many addresses (the
5980 report of this problem mentioned 10 000) and each of them was rewritten,
5981 Exim could use up a very large amount of memory. (It kept on making new
5982 copies of the header line as it rewrote, and never released the old ones.)
5983 At the expense of a bit more processing, the header rewriting function has
5984 been changed so that it no longer eats memory in this way.
5986 66. The generation of the Received: header has been moved from the time that a
5987 message starts to be received, to the time that it finishes. The timestamp
5988 in the Received: header should now be very close to that of the <= log
5989 line. There are two side-effects of this change:
5991 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5992 logged header lines no longer include the local Received: line, because
5993 it has not yet been created. The same applies to a copy of the message
5994 that is returned to a non-SMTP sender when a message is rejected.
5996 (b) When a filter file is tested using -bf, no additional Received: header
5997 is added to the test message. After some thought, I decided that this
6000 This change does not affect the value of $received_for. It is still set
6001 after address rewriting, but before local_scan() is called.
6003 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6005 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6006 gave an unhelpful panic error message, and a defer error. I have managed to
6007 change this behaviour so that it now rejects any supplied certificate,
6008 which seems right, as the list of acceptable certificates is empty.
6010 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6011 gave an unhelpful defer error. I have not managed to make this reject any
6012 supplied certificates, but the error message it gives is "no certificate
6013 supplied", which is not helpful.
6015 70. exigrep's output now also includes lines that are not associated with any
6016 message, but which match the given pattern. Implemented by a patch from
6017 Martin Sluka, which also tidied up the Perl a bit.
6019 71. Recipient callout verification, like sender verification, was using <> in
6020 the MAIL FROM command. This isn't really the right thing, since the actual
6021 sender may affect whether the remote host accepts the recipient or not. I
6022 have changed it to use the actual sender in the callout; this means that
6023 the cache record is now keyed on a recipient/sender pair, not just the
6024 recipient address. There doesn't seem to be a real danger of callout loops,
6025 since a callout by the remote host to check the sender would use <>.
6026 [SEE ABOVE: changed after hitting problems.]
6028 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6029 temporary errors. However, in the case of such a code being given after
6030 the end of a data transmission (i.e. after ".") Exim was failing to write
6031 a retry record for the message. (Yes, there was some broken host that was
6032 actually sending 8xx at this point.)
6034 73. An unknown lookup type in a host list could cause Exim to panic-die when
6035 the list was checked. (An example that provoked this was putting <; in the
6036 middle of a list instead of at the start.) If this happened during a DATA
6037 ACL check, a -D file could be left lying around. This kind of configuration
6038 error no longer causes Exim to die; instead it causes a defer error. The
6039 incident is still logged to the main and panic logs.
6041 74. Buglet left over from Exim 3 conversion. The message "too many messages
6042 in one connection" was written to the rejectlog but not the mainlog, except
6043 when address rewriting (yes!) was being logged.
6045 75. Added write_rejectlog option.
6047 76. When a system filter was run not as root (that is, when system_filter_user
6048 was set), the values of the $n variables were not being returned to the
6049 main process; thus, they were not subsequently available in the $sn
6052 77. Added +return_path_on_delivery log selector.
6054 78. A connection timeout was being treated differently from recipients deferred
6055 when testing hosts_max_try with a message that was older than the host's
6056 retry timeout. (The host should not be counted, thus allowing all hosts to
6057 be tried at least once before bouncing.) This may have been the cause of an
6058 occasionally reported bug whereby a message would remain on the queue
6059 longer than the retry timeout, but would be bounced if a delivery was
6060 forced. I say "may" because I never totally pinned down the problem;
6061 setting up timeout/retry tests is difficult. See also the next item.
6063 79. The ultimate address timeout was not being applied to errors that involved
6064 a combination of host plus message (for example, a timeout on a MAIL
6065 command). When an address resolved to a number of possible hosts, and they
6066 were not all tried for each delivery (e.g. because of hosts_max_try), a
6067 message could remain on the queue longer than the retry timeout.
6069 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6072 81. Fixed an obscure SMTP outgoing bug which required at least the following
6073 conditions: (a) there was another message waiting for the same server;
6074 (b) the server returned 5xx to all RCPT commands in the first message so
6075 that the message was not completed; (c) the server dropped the connection
6076 or gave a negative response to the RSET that Exim sends to abort the
6077 transaction. The observed case was a dropped connection after DATA that had
6078 been sent in pipelining mode. That is, the server had advertised PIPELINING
6079 but was not implementing it correctly. The effect of the bug was incorrect
6080 behaviour, such as trying another host, and this could lead to a crash.
6086 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6087 and daemon.c were passed as pointers to ints; they should have been
6088 pointers to socklen_t variables (which are typically unsigned ints).
6090 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6093 3. Fixed a really odd bug that affected only the testing scheme; patching a
6094 certain fixed string in the binary changed the value of another string that
6095 happened to be identical to the end of the original first string.
6097 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6098 name", it returns that address as the IP address. On some operating
6099 systems (e.g. Solaris), it also passes back the IP address string as the
6100 "host name". However, on others (e.g. Linux), it passes back an empty
6101 string. Exim wasn't checking for this, and was changing the host name to an
6102 empty string, assuming it had been canonicalized.
6104 5. Although rare, it is permitted to have more than one PTR record for a given
6105 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6106 all the names associated with an address, because they do in Solaris.
6107 However, it seems that they do not in Linux for data that comes from the
6108 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6109 I found this out when I moved to a new Linux workstation and tried to run
6110 the Exim test suite.
6112 To get round this problem I have changed the code so that it now does its
6113 own call to the DNS to look up PTR records when searching for a host name.
6114 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6115 addresses that are only in /etc/hosts are still found.
6117 This behaviour is, however, controlled by an option called host_lookup_
6118 order, which defaults to "bydns:byaddr". If people want to use the other
6119 order, or indeed, just use one or the other means of lookup, they can
6120 specify it in this variable.
6122 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6123 some operating systems, this comes back from gethostbyaddr() as an empty
6124 string, and this is what Exim used to test for. However, it seems that in
6125 other systems, "." is yielded. Exim now tests for this case too.
6127 7. The values of check_spool_space and check_log_space are now held internally
6128 as a number of kilobytes instead of an absolute number of bytes. If a
6129 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6130 kilobyte. This means that much larger values can be stored.
6132 8. Exim monitor: an attempt to get the action menu when not actually pointing
6133 at a message produces an empty menu entitled "No message selected". This
6134 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6135 no entries in it ("Shell widget menu has zero width and/or height"). So I
6136 have added a single, blank menu entry in this case.
6138 9. Added ${quote_local_part.
6140 10. MIME decoding is now applied to the contents of Subject: header lines when
6143 11. Now that a reference to $sender_host_address automatically causes a reverse
6144 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6145 host lookup before query-style lookups in lists that might use this
6146 variable. This has therefore been abolished, and the "net-" prefix is no
6147 longer necessary for query-style lookups.
6149 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6150 have been a typo for LFLAGS, so it has been changed.
6152 13. The install script calls Exim with "-C /dev/null" in order to find the
6153 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6154 to be output. However, since Exim outputs its version number before the
6155 error, it didn't break the script. It just looked ugly. I fixed this by
6156 always allowing "-C /dev/null" if the caller is root.
6158 14. Ignore overlarge ACL variable number when reading spool file - insurance
6159 against a later release with more variables having written the file.
6161 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6162 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6163 this, as well as the form without the "IPv6" on the front.
6165 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6166 OS/Makefile-Darwin file.
6168 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6169 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6171 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6172 variable. However, some OS use ints for inodes. I've added cast to long int
6173 to get rid of the compiler warning.
6175 19. I had forgotten to lock out "/../" in configuration file names when
6176 ALT_CONFIG_PREFIX was set.
6178 20. Routers used for verification do not need to specify transports. However,
6179 if such a router generated a host list, and callout was configured, Exim
6180 crashed, because it could not find a port number from the (non-existent)
6181 transport. It now assumes port 25 in this circumstance.
6183 21. Added the -t option to exigrep.
6185 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6186 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6187 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6188 them is set, LOOKUP_LSEARCH is forced.
6190 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6192 24. Added sender and host information to the "rejected by local_scan()" log
6193 line; previously there was no indication of these.
6195 25. Added .include_if_exists.
6197 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6198 Linux. It turns out that not all file systems support this. Apparently some
6199 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6200 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6201 not supported on the file descriptor, is now ignored when Exim is trying to
6202 sync a directory. This applies only to Linux.
6204 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6206 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6207 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6210 29. Added support for maildirsize files from supplied patch (modified a bit).
6212 30. The use of :fail: followed by an empty string could lead Exim to respond to
6213 sender verification failures with (e.g.):
6215 550 Verification failed for <xxx>
6216 550 Sender verify failed
6218 where the first response line was missing the '-' that indicates it is not
6219 the final line of the response.
6221 31. The loop for finding the name of the user that called Exim had a hardwired
6222 limit of 10; it now uses the value of finduser_retries, which is used for
6223 all other user lookups.
6225 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6228 33. Exim was neglecting to zero errno before one call of strtol() when
6229 expanding a string and expecting an integer value. On some systems this
6230 resulted in spurious "integer overflow" errors. Also, it was casting the
6231 result into an int without checking.
6233 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6234 did not work. The code looks as if it has *never* worked, though it appears
6235 to have been documented since at least release 1.62. I have made it work.
6237 35. The "timeout_DNS" error in retry rules, also documented since at least
6238 1.62, also never worked. As it isn't clear exactly what this means, and
6239 clearly it isn't a major issue, I have abolished the feature by treating it
6240 as "timeout", and writing a warning to the main and panic logs.
6242 36. The display of retry rules for -brt wasn't always showing the error code
6245 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6246 timeout_connect_A, timeout_connect_MX.
6248 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6249 to the empty sender.
6251 39. The daemon was not analysing the content of -oX till after it had closed
6252 stderr and disconnected from the controlling terminal. This meant that any
6253 syntax errors were only noted on the panic log, and the return code from
6254 the command was 0. By re-arranging the code a little, I've made the
6255 decoding happen first, so such errors now appear on stderr, and the return
6256 code is 1. However, the actual setting up of the sockets still happens in
6257 the disconnected process, so errors there are still only recorded on the
6260 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6261 connections (as happens on some IP stacks) was logged at start up time as
6262 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6263 it from "IPv6 and IPv4", which means that two separate sockets are being
6266 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6267 says whether AF_INET or AF_INET6 was passed as an argument.
6269 42. Exiwhat output was messed up when time zones were included in log
6272 43. Exiwhat now gives more information about the daemon's listening ports,
6273 and whether -tls-on-connect was used.
6275 44. The "port" option of the smtp transport is now expanded.
6277 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6278 silently ignored. Now an error message is written to the main and panic
6281 46. There's a new ACL modifier called "logwrite" which writes to a log file
6282 as soon as it is encountered.
6284 47. Added $local_user_uid and $local_user_gid at routing time.
6286 48. Exim crashed when trying to verify a sender address that was being
6289 49. Exim was recognizing only a space character after ".include". It now also
6290 recognizes a tab character.
6292 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6293 extracting the relevant information from the specification. The man page no
6294 longer contains scrambled data for the -d option, and I've added a section
6295 at the front about calling Exim under different names.
6297 51. Added "extra_headers" argument to the "mail" command in filter files.
6299 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6302 53. Installed eximstats 1.29.
6304 54. Added transport_filter_timeout as a generic transport option.
6306 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6307 Cc: header lines. This was required by RFC 822, but it not required by RFC
6310 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6311 incoming messages that did not have them. Now it does so only if the
6312 message originates locally, that is, if there is no associated remote host
6313 address. When Resent- header lines are present, this applies to the Resent-
6314 lines rather than the non-Resent- lines.
6316 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6317 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6319 58. Messages for configuration errors now include the name of the main
6320 configuration file - useful now that there may be more than one file in a
6321 list (.included file names were always shown).
6323 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6324 for those rare installations that do not start the daemon as root or run it
6325 setuid root. I've cut out the call to initgroups() if the daemon is not
6328 60. The Exim user and group can now be bound into the binary as text strings
6329 that are looked up at the start of Exim's processing.
6331 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6333 62. Added $mailstore_basename variable.
6335 63. Installed patch to sieve.c from Michael Haardt.
6337 64. When Exim failed to open the panic log after failing to open the main log,
6338 the original message it was trying to log was written to stderr and debug
6339 output, but if they were not available (the usual case in production), it
6340 was lost. Now it is written to syslog before the two lines that record the
6341 failures to open the logs.
6343 65. Users' Exim filters run in subprocesses under the user's uid. It is
6344 possible for a "deliver" command or an alias in a "personal" command to
6345 provoke an address rewrite. If logging of address rewriting is configured,
6346 this fails because the process is not running as root or exim. There may be
6347 a better way of dealing with this, but for the moment (because 4.30 needs
6348 to be released), I have disabled address rewrite logging when running a
6349 filter in a non-root, non-exim process.
6355 1. The buildconfig auxiliary program wasn't quoting the value set for
6356 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6357 not defined. This bug was masked in 4.22 by the effect that was fixed in
6360 2. Some messages that were rejected after a message id was allocated were
6361 shown as "incomplete" by exigrep. It no longer does this for messages that
6362 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6364 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6365 have allow_domain_literals set, the ID did not get logged in the <= line.
6366 Domain literals are now always recognized in Message-ID: header lines.
6368 4. The first argument for a ${extract expansion item is the key name or field
6369 number. Leading and trailing spaces in this item were not being ignored,
6370 causing some misleading effects.
6372 5. When deliver_drop_privilege was set, single queue runner processes started
6373 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6374 same command in the process it spins off) were not dropping privilege.
6376 6. When the daemon running as "exim" started a queue runner, it always
6377 re-executed Exim in the spun-off process. This is a waste of effort when
6378 deliver_drop_privilege is set. The new process now just calls the
6379 queue-runner function directly.
6385 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6388 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6389 ignored. Though the use of -oP was forcing the writing of a pid file, it
6390 was always written to the default place.
6392 3. If the message "no IP address found for host xxxx" is generated during
6393 incoming verification, it is now followed by identification of the incoming
6394 connection (so you can more easily find what provoked it).
6396 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6398 5. Added some features to "harden" Exim a bit more against certain attacks:
6400 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6401 be put in Local/Makefile. This is like the never_users runtime option,
6402 but it cannot be overridden. The default setting is "root".
6404 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6405 prefix string with which any file named in a -C command line option
6408 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6409 is retained for -C and -D only if the caller of Exim is root. Without
6410 it, the exim user may also use -C and -D and retain privilege.
6412 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6413 command line option is disabled.
6415 6. Macro names set by the -D option must start with an upper case letter, just
6416 like macro names defined in the configuration file.
6418 7. Added "dereference=" facility to LDAP.
6420 8. Two instances of the typo "uknown" in the source files are fixed.
6422 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6423 the Configure-Makefile script screwed up while processing it.
6425 10. Incorporated PCRE 4.4.
6427 11. The SMTP synchronization check was not operating right at the start of an
6428 SMTP session. For example, it could not catch a HELO sent before the client
6429 waited for the greeting. There is now a check for outstanding input at the
6430 point when the greeting is written. Because of the duplex, asynchronous
6431 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6432 way, but not yet received, when the check is performed.
6434 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6435 on TCP/IP sockets, because this apparently causes some broken clients to
6438 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6439 unchanged) from the Cygwin maintainer.
6441 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6442 was supported instead of testing for mbx. Effectively a typo.
6444 15. The spa authenticator server code was not checking that the input it
6445 received was valid base64.
6447 16. The debug output line for the "set" modifier in ACLs was not showing the
6448 name of the variable that was being set.
6450 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6452 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6453 lookup on its own. Something else was needed to trigger the lookup. For
6454 example, a match in host_lookup or the need for a host name in a host list.
6455 Now, if $sender_host_name is referenced and the host name has not yet been
6456 looked up, a lookup is performed. If the lookup fails, the variable remains
6457 empty, and $host_lookup_failed is set to "1".
6459 19. Added "eqi" as a case-independent comparison operator.
6461 20. The saslauthd authentication condition could segfault if neither service
6462 nor realm was specified.
6464 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6465 error message that was logged was misleading, and incoming SMTP
6466 connections were dropped. The message is now more accurate, and temporary
6467 errors are given to SMTP connections.
6469 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6470 (except RSET) with a 421 error, until QUIT is received. However, it was
6471 failing to send a response to QUIT.
6473 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6474 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6475 ACL is now run after the helo_try_verify_hosts code.
6477 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6478 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6479 case-independent, so other case variants are also recognized). Apparently
6480 some systems use these upper case variants.
6482 25. If more than two messages were waiting for the same host, and a transport
6483 filter was specified for the transport, Exim sent two messages over the
6484 same TCP/IP connection, and then failed with "socket operation on non-
6485 socket" when it tried to send the third.
6487 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6489 27. The extern definition of crypt16() in expand.c was not being excluded when
6490 the OS had its own crypt16() function.
6492 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6493 as a preferred synonym for return_size_limit, both as an option and as an
6496 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6498 30. Changed the default configuration ACL to relax the local part checking rule
6499 for addresses that are not in any local domains. For these addresses,
6500 slashes and pipe symbols are allowed within local parts, but the sequence
6501 /../ is explicitly forbidden.
6503 31. SPA server authentication was not clearing the challenge buffer before
6506 32. log_message in a "warn" ACL statement was writing to the reject log as
6507 well as to the main log, which contradicts the documentation and doesn't
6508 seem right (because no rejection is happening). So I have stopped it.
6510 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6511 However, I am unable to do any testing of this.
6513 34. Fixed an infelicity in the appendfile transport. When checking directories
6514 for a mailbox, to see if any needed to be created, it was accidentally
6515 using path names with one or more superfluous leading slashes; tracing
6516 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6518 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6519 contents are added to the log line that is written for every discarded
6520 recipient. (Previously a log_message setting was ignored.)
6522 36. The ${quote: operator now quotes the string if it is empty.
6524 37. The install script runs exim in order to find its version number. If for
6525 some reason other than non-existence or emptiness, which it checks, it
6526 could not run './exim', it was installing it with an empty version number,
6527 i.e. as "exim-". This error state is now caught, and the installation is
6530 38. An argument was missing from the function that creates an error message
6531 when Exim fails to connect to the socket for saslauthd authentication.
6532 This could cause Exim to crash, or give a corrupted message.
6534 39. Added isip, isip4, and isip6 to ${if conditions.
6536 40. The ACL variables $acl_xx are now saved with the message, and can be
6537 accessed later in routers, transports, and filters.
6539 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6540 strings in the file are not string-expanded.
6542 42. If a MAIL command specified a SIZE value that was too large to fit into an
6543 int variable, the check against message_size_limit failed. Such values are
6544 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6545 day this will have to be increased, but I don't think I want to be around
6546 when emails are that large.
6553 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6554 iconv() is not standard in FreeBSD.
6556 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6557 IPv6 enabled. The observed symptom was a segmentation fault on return from
6558 the function os_common_find_running_interfaces() in src/os.c.
6560 3. In the check_special_case() function in daemon.c I had used "errno" as an
6561 argument name, which causes warnings on some systems. This was basically a
6562 typo, since it was named "eno" in the comments!
6564 4. The code that waits for the clock to tick (at a resolution of some fraction
6565 of a second) so as to ensure message-id uniqueness was always waiting for
6566 at least one whole tick, when it could have waited for less. [This is
6567 almost certainly not relevant at current processor speeds, where it is
6568 unlikely to ever wait at all. But we try to future-proof.]
6570 5. The function that sleeps for a time interval that includes fractions of a
6571 second contained a race. It did not block SIGALRM between setting the
6572 timer, and suspending (a couple of lines later). If the interval was short
6573 and the sigsuspend() was delayed until after it had expired, the suspension
6574 never ended. On busy systems this could lead to processes getting stuck for
6577 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6578 process, before it forks any delivery processes. The open lookup caching
6579 mechanism meant that the open file or database connection was passed into
6580 the delivery process. The problem was that delivery processes always tidy
6581 up cached lookup data. This could cause a problem for the next delivery
6582 process started by the queue runner, because the external queue runner
6583 process does not know about the closure. So the next delivery process
6584 still has data in the lookup cache. In the case of a file lookup, there was
6585 no problem because closing a file descriptor in a subprocess doesn't affect
6586 the parent. However, if the lookup was caching a connection to a database,
6587 the connection was closed, and the second delivery process was likely to
6588 see errors such as "PGSQL: query failed: server closed the connection
6589 unexpectedly". The problem has been fixed by closing all cached lookups
6590 in a queue runner before running a delivery process.
6592 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6593 seem to have the "const" qualifier which it has on other OS. I've
6596 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6597 *of the same type* (client or server) with the same public name that an
6598 error should be diagnosed.
6600 9. When Exim looked up a host name for an IP address, but failed to find the
6601 original IP address when looking up the host name (a safety check), it
6602 output the message "<ip address> does not match any IP for NULL", which was
6603 confusing, to say the least. The bug was that the host name should have
6604 appeared instead of "NULL".
6606 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6607 user that is built into the binary, and the -C or -D options is used, root
6608 privilege is dropped before the configuration file is read. In addition,
6609 logging is switched to stderr instead of the normal log files. If the
6610 configuration then re-defines the Exim user, the unprivileged environment
6611 is probably not what is expected, so Exim logs a panic warning message (but
6614 However, if deliver_drop_privilege is set, the unprivileged state may well
6615 be exactly what is intended, so the warning has been cut out in that case,
6616 and Exim is allowed to try to write to its normal log files.
6622 1. smtp_return_error_details was not giving details for temporary sender
6623 or receiver verification errors.
6625 2. Diagnose a configuration error if two authenticators have the same public
6628 3. Exim used not to create the message log file for a message until the first
6629 delivery attempt. This could be confusing when incoming messages were held
6630 for policy or load reasons. The message log file is now created at the time
6631 the message is received, and an initial "Received" line is written to it.
6633 4. The automatically generated man page for command line options had a minor
6634 bug that caused no ill effects; however, a more serious problem was that
6635 the procedure for building the man page automatically didn't always
6636 operate. Consequently, release 4.20 contains an out-of-date version. This
6637 shouldn't happen again.
6639 5. When building Exim with embedded Perl support, the script that builds the
6640 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6641 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6643 6. The freeze_tell option was not being used for messages that were frozen on
6644 arrival, either by an ACL or by local_scan().
6646 7. Added the smtp_incomplete_transaction log selector.
6648 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6649 was accepting AUTH without a new EHLO.
6651 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6652 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6655 10. Logging of TCP/IP connections (when configured) now happens in the main
6656 daemon process instead of the child process, so that the TCP/IP connection
6657 count is more accurate (but it can never be perfect).
6659 11. The use of "drop" in a nested ACL was not being handled correctly in the
6660 outer ACL. Now, if condition failure induced by the nested "drop" causes
6661 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6662 or "require"), the connection is dropped.
6664 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6665 that yield "discard" can now be used with an "accept" or a "discard" verb,
6666 but an error is generated for any others (because I can't see a useful way
6667 to define what should happen).
6669 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6670 are now processed in the same way as lines in the Exim configuration file.
6671 In particular, continuation lines are supported.
6673 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6675 15. Added -ti meaning -t -i.
6677 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6678 domains, and warn by logging if others are found.
6680 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6681 returned from GIFCONF when Exim is trying to find the list of interfaces on
6682 a host. The code in os.c has been modified to copy each ifreq to an aligned
6683 structure in all cases.
6685 Also, in some cases, the returned ifreq's were being copied to a 'struct
6686 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6687 means the last couple of bytes of an IPv6 address could be chopped if the
6688 ifreq contained only a normal sockaddr (14 bytes storage).
6690 18. Named domain lists were not supported in the hosts_treat_as_local option.
6691 An entry such as +xxxx was not recognized, and was treated as a literal
6694 19. Ensure that header lines added by a DATA ACL are included in the reject log
6695 if the ACL subsequently rejects the message.
6697 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6698 MD5 (which is deprecated).
6700 21. When testing a filter file using -bf, Exim was writing a message when it
6701 took the sender from a "From " line in the message, but it was not doing so
6702 when it took $return_path from a Return-Path: header line. It now does.
6704 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6705 with a valid header line field name (a series of printing characters
6706 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6708 23. Changed "disc" in the source to "disk" to conform to the documentation and
6709 the book and for uniformity.
6711 24. Ignore Sendmail's -Ooption=value command line item.
6713 25. When execve() failed while trying to run a command in a pipe transport,
6714 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6715 could be confused with a return value of 69 from the command itself. This
6716 has been changed to 127, the value the shell returns if it is asked to run
6717 a non-existent command. The wording for the related log line suggests a
6718 non-existent command as the problem.
6720 26. If received_header_text expands to an empty string, do not add a Received:
6721 header line to the message. (Well, it adds a token one on the spool, but
6722 marks it "old" so that it doesn't get used or transmitted.)
6724 27. Installed eximstats 1.28 (addition of -nt option).
6726 28. There was no check for failure on the call to getsockname() in the daemon
6727 code. This can fail if there is a shortage of resources on the system, with
6728 ENOMEM, for example. A temporary error is now given on failure.
6730 29. Contrary to the C standard, it seems that in some environments, the
6731 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6732 program. Exim now does this explicitly; it affects the formatting of
6733 timestamps using strftime().
6735 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6736 complaints. I've now initialized all the variables, to avoid this.
6738 32. Header lines added by a system filter were not being "seen" during
6739 transport-time rewrites.
6741 33. The info_callback() function passed to OpenSSL is set up with type void
6742 (*)(SSL *, int, int), as described somewhere. However, when calling the
6743 function (actually a macro) that sets it up, the type void(*)() is
6744 expected. I've put in a cast to prevent warnings from picky compilers.
6746 34. If a DNS black list lookup found a CNAME record, but there were no A
6747 records associated with the domain it pointed at, Exim crashed.
6749 35. If a DNS black list lookup returned more than one A record, Exim ignored
6750 all but the first. It now scans all returned addresses if a particular IP
6751 value is being sought. In this situation, the contents of the
6752 $dnslist_value variable are a list of all the addresses, separated by a
6755 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6756 to accept a host name and all its aliases if the forward lookup for any of
6757 them yielded the IP address of the incoming connection. Now it accepts only
6758 those names whose forward lookup yields the correct IP address. Any other
6759 names are discarded. This closes a loophole whereby a rogue DNS
6760 administrator could create reverse DNS records to break through a
6761 wildcarded host restriction in an ACL.
6763 37. If a user filter or a system filter that ran in a subprocess used any of
6764 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6765 the wrong values were passed to the pipe command ($thisaddress had the
6766 value of $0, $0 had the value of $1, etc). This bug was introduced by
6767 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6769 38. Improved the line breaking for long SMTP error messages from ACLs.
6770 Previously, if there was no break point between 40 and 75 characters, Exim
6771 left the rest of the message alone. Two changes have been made: (a) I've
6772 reduced the minimum length to 35 characters; (b) if it can't find a break
6773 point between 35 and 75 characters, it looks ahead and uses the first one
6774 that it finds. This may give the occasional overlong line, but at least the
6775 remaining text gets split now.
6777 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6778 file descriptors might be low, and that setting 1000 would always raise it.
6779 It turns out that in some environments, the limit is already over 1000 and
6780 that lowering it causes trouble. So now Exim takes care not to decrease it.
6782 40. When delivering a message, the value of $return_path is set to $sender_
6783 address at the start of routing (routers may change the value). By an
6784 oversight, this default was not being set up when an address was tested by
6785 -bt or -bv, which affected the outcome if any router or filter referred to
6788 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6789 log only when "message" or "log_message" are set. However, if one of the
6790 conditions was an address verification, or a call to a nested ACL, the
6791 messages generated by the underlying test were being passed through. This
6792 no longer happens. The underlying message is available in $acl_verify_
6793 message for both "message" and "log_message" expansions, so it can be
6794 passed through if needed.
6796 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6797 new expansion $bh_ to give the encoded byte string without charset
6798 translation. Translation happens only if iconv() is available; HAVE_ICONV
6799 indicates this at build time. HEADERS_CHARSET gives the charset to
6800 translate to; headers_charset can change it in the configuration, and
6801 "headers charset" can change it in an individual filter file.
6803 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6804 that creates RFC 2047 encoded "words" labels them as that charset instead
6805 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6806 expansion operator; (ii) when Exim creates a From: line for a local
6807 message; (iii) when a header line is rewritten to include a "phrase" part.
6809 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6810 buggy, causing it to skip the first lines of messages whose message ID
6811 ended in 'D'. This would not have bitten before Exim release 4.14, because
6812 message IDs were unlikely to end in 'D' before then. The effect was to have
6813 incorrect size information for certain domains.
6815 45. #include "config.h" was missing at the start of the crypt16.c module. This
6816 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6819 46. If there was a timeout during a "random" callout check, Exim treated it as
6820 a failure of the random address, and carried on sending RSET and the real
6821 address. If the delay was just some slowness somewhere, the response to the
6822 original RCPT would be taken as a response to RSET and so on, causing
6823 mayhem of various kinds.
6825 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6826 when I implemented it. It didn't allow for the fact that some option values
6827 may legitimately be negative (e.g. size_addition), and it didn't even do
6828 the right test for positive values.
6830 48. Domain names in DNS records are case-independent. Exim always looks them up
6831 in lower case. Some resolvers return domain names in exactly the case they
6832 appear in the zone file, that is, they may contain uppercase letters. Not
6833 all resolvers do this - some return always lower case. Exim was treating a
6834 change of case by a resolver as a change of domain, similar to a widening
6835 of a domain abbreviation. This triggered its re-routing code and so it was
6836 trying to route what was effectively the same domain again. This normally
6837 caused routing to fail (because the router wouldn't handle the domain
6838 twice). Now Exim checks for this case specially, and just changes the
6839 casing of the domain that it ultimately uses when it transmits the message
6842 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6845 50. If a filter generated a file delivery with a non-absolute name (possible if
6846 no home directory exists for the router), the forbid_file option was not
6849 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6850 the existing equality matching.
6852 52. Exim was using ints instead of ino_t variables in some places where it was
6853 dealing with inode numbers.
6855 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6856 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6857 called TMPDIR, and if it finds it is different, it changes its value.
6859 54. The smtp_printf() function is now made available to local_scan() so
6860 additional output lines can be written before returning. There is also an
6861 smtp_fflush() function to enable the detection of a dropped connection.
6862 The variables smtp_input and smtp_batched_input are exported to
6865 55. Changed the default runtime configuration: the message "Unknown user"
6866 has been removed from the ACL, and instead placed on the localuser router,
6867 using the cannot_route_message feature. This means that any verification
6868 failures that generate their own messages won't get overridden. Similarly,
6869 the "Unrouteable address" message that was in the ACL for unverifiable
6870 relay addresses has also been removed.
6872 56. Added hosts_avoid_esmtp to the smtp transport.
6874 57. The exicyclog script was not checking for the esoteric option
6875 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6876 will work only if exicyclog is run under the appropriate euid.
6878 58. Following a discussion on the list, the rules by which Exim recognises line
6879 endings on incoming messages have been changed. The -dropcr and drop_cr
6880 options are now no-ops, retained only for backwards compatibility. The
6881 following line terminators are recognized: LF CRLF CR. However, special
6882 processing applies to CR:
6884 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6885 nor a local message in the state where . is a terminator.
6887 (ii) If a bare CR is encountered in a header line, an extra space is added
6888 after the line terminator so as not to end the header. The reasoning
6889 behind this is that bare CRs in header lines are most likely either
6890 to be mistakes, or people trying to play silly games.
6892 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6893 was being incorrectly given as 18 bytes larger than it should have been.
6894 This is a VOB (very old bug).
6896 60. This may never have affected anything current, but just in case it has:
6897 When the local host is found other than at the start of a list of hosts,
6898 the local host, those with the same MX, and any that follow, are discarded.
6899 When the list in question was part of a longer list of hosts, the following
6900 hosts (not currently being processed) were also being discarded. This no
6901 longer happens. I'm not sure if this situation could ever has previously
6904 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6907 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6908 header. This is something that is recommended in a new Internet Draft, and
6909 is something that is documented as being done by Sendmail. There are two
6910 possible values. For messages generated by the autoreply transport, Exim
6913 Auto-Submitted: auto-replied
6915 whereas for all other generated messages (e.g. bounces) it adds
6917 Auto-Submitted: auto-generated
6919 63. The "personal" condition in filters now includes a test for the
6920 Auto-Submitted: header. If it contains the string "auto-" the message it
6921 not considered personal.
6923 64. Added rcpt_include_affixes as a generic transport option.
6925 65. Added queue_only_override (default true).
6927 66. Added the syslog_duplication option.
6929 67. If what should have been the first header line of a message consisted of
6930 a space followed by a colon, Exim was mis-interpreting it as a header line.
6931 It isn't of course - it is syntactically invalid and should therefore be
6932 treated as the start of the message body. The misbehaviour could have
6933 caused a number of strange effects, including loss of data in subsequent
6934 header lines, and spool format errors.
6936 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6937 client host had authenticated. This control can now be exercised by an ACL
6938 for more flexibility.
6940 69. By default, callouts do not happen when testing with -bh. There is now a
6941 variant, -bhc, which does actually run the callout code, including
6942 consulting and updating the callout cache.
6944 70. Added support for saslauthd authentication, courtesy of Alexander
6947 71. If statvfs() failed on the spool or log directories while checking their
6948 size for availability, Exim confusingly gave the error "space shortage".
6949 Furthermore, in debugging mode it crashed with a floating point exception.
6950 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6951 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6952 serious problem, Exim now writes to the main and panic logs when this
6953 happens, with details of the failure. It then refuses to accept the
6954 incoming message, giving the message "spool directory problem" or "log
6955 directory problem" with a 421 code for SMTP messages.
6957 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6958 0, 1, and 2 exist, because some OS complain for execs without them (see
6959 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6960 descriptors. However, the code omitted to check that the open succeeded,
6961 causing mysterious errors if for some reason the permissions on /dev/null
6962 got screwed. Now Exim writes a message to the main and panic logs, and
6963 bombs out if it can't open /dev/null.
6965 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6966 interact so that it is all more flexible. It is supposed to remain
6967 backwards compatible. Also added extra_local_interfaces.
6969 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6970 to bomb out with an assertion failure - to the client this appears as a
6971 connection drop. This problem occurs in the part of the code that was taken
6972 from the Samba project. Fortunately, the assertion is in a very simple
6973 function, so I have fixed this by reproducing the function inline in the
6974 one place where it is called, and arranging for authentication to fail
6975 instead of killing the process with assert().
6977 75. The SPA client code was not working when the server requested OEM rather
6978 than Unicode encoding.
6980 76. Added code to make require_files with a specific uid setting more usable in
6981 the case where statting the file as root fails - usually a non-root-mounted
6982 NFS file system. When this happens and the failure is EACCES, Exim now
6983 forks a subprocess and does the per-uid checking as the relevant uid.
6985 77. Added process_log_path.
6987 78. If log_file_path was not explicitly set, a setting of check_log_space or
6988 check_log_inodes was ignored.
6990 79. If a space check for the spool or log partitions fails, the incident is now
6991 logged. Of course, in the latter case the data may get lost...
6993 80. Added the %p formatting code to string_format() so that it can be used to
6994 print addresses in debug_print(). Adjusted all the address printing in the
6995 debugging in store.c to use %p rather than %d.
6997 81. There was a concern that a line of code in smtp_in.c could overflow a
6998 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6999 initially expressed, the concern was not well-founded, because trailing
7000 spaces are removed early. However, if the trailing spaces were followed by
7001 a NULL, they did not get removed, so the overflow was possible. Two fixes
7004 (a) I re-wrote the offending code in a cleaner fashion.
7005 (b) If an incoming SMTP command contains a NULL character, it is rejected
7008 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7009 runs initgroups(), so that if the Exim user is in any additional groups,
7010 they will be used during message reception.
7016 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob