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
54 JH/01 Rework error string handling in TLS interface so that the caller in
55 more cases is responsible for logging. This permits library-sourced
56 string to be attached to addresses during delivery, and collapses
57 pairs of long lines into single ones.
59 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
60 during configuration. Wildcards are allowed and expanded.
62 JH/02 Rework error string handling in DKIM to pass more info back to callers.
63 This permits better logging.
65 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
66 do not close it down and have the child transport start it up again on
67 the passed-on TCP connection. Instead, proxy the child (and any
68 subsequent ones) for TLS via a unix-domain socket channel. Logging is
69 affected: the continued delivery log lines do not have any DNSSEC, TLS
70 Certificate or OCSP information. TLS cipher information is still logged.
72 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
73 identical IP addresses on different listening ports. Will also affect
76 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
77 add noisy ifdef guards to special-case this sillyness.
78 Patch from Bernd Kuhls.
80 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
81 than 255 are no longer allowed.
83 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
84 Disable the session-cache too, which might reduce our load. Since we
85 currrectly use a new context for every connection, both as server and
86 client, there is no benefit for these.
87 GnuTLS appears to not support tickets server-side by default (we don't
88 call gnutls_session_ticket_enable_server()) but client side is enabled
89 by default on recent versions (3.1.3 +) unless the PFS priority string
92 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
93 <https://reproducible-builds.org/specs/source-date-epoch/>.
95 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
96 the check for any unsuccessful recipients did not notice the limit, and
97 erroneously found still-pending ones.
99 JH/08 Pipeline CHUNKING command and data together, on kernels that support
100 MSG_MORE. Only in-clear (not on TLS connections).
102 JH/09 Avoid using a temporary file during transport using dkim. Unless a
103 transport-filter is involved we can buffer the headers in memory for
104 creating the signature, and read the spool data file once for the
105 signature and again for transmission.
107 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
108 4.77 as the kernel support then wasn't solid, having issues in 64bit
109 mode. Now, it's been long enough. Add support for FreeBSD also.
111 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
112 case where the routing stage had gathered several addresses to send to
113 a host before calling the transport for the first, we previously failed
114 to close down TLS in the old transport process before passing the TCP
115 connection to the new process. The new one sent a STARTTLS command
116 which naturally failed, giving a failed delivery and bloating the retry
117 database. Investigation and fix prototype from Wolfgang Breyha.
119 JH/12 Fix check on SMTP command input synchronisation. Previously there were
120 false-negatives in the check that the sender had not preempted a response
121 or prompt from Exim (running as a server), due to that code's lack of
122 awareness of the SMTP input buffering.
124 PP/04 Add commandline_checks_require_admin option.
125 Exim drops privileges sanely, various checks such as -be aren't a
126 security problem, as long as you trust local users with access to their
127 own account. When invoked by services which pass untrusted data to
128 Exim, this might be an issue. Set this option in main configuration
129 AND make fixes to the calling application, such as using `--` to stop
132 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
133 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
134 responses to those, into a single TLS record each way (this usually means
135 a single packet). As a side issue, smtp_enforce_sync now works on TLS
138 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
139 affects you only if you're dancing at the edge of the param size limits.
140 If you are, and this message makes sense to you, then: raise the
141 configured limit or use OpenSSL 1.1. Nothing we can do for older
144 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
145 cmdline element to get the documented default one. Previously it was
148 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
149 are unsuable for non-message-specific reasons - eg. connection timeout,
150 banner-time rejection.
152 JH/16 Fix logging of delivery remote port, when specified by router, under
155 PP/06 Repair manualroute's ability to take options in any order, even if one
156 is the name of a transport.
159 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
161 JH/17 Change the list-building routines interface to use the expanding-string
162 triplet model, for better allocation and copying behaviour.
164 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
165 Previously it was constructed the first time a possibly-matching string
166 was met in the configuration file input during startup; now it is done
169 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
170 compatible one, to avoid the (poorly documented) possibility of a config
171 file in the working directory redirecting the DB files, possibly correpting
172 some existing file. CVE-2017-10140 assigned for BDB.
174 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
175 cache-hot. Previously, although the result was properly cached, the
176 initial verify call returned a defer.
178 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
179 the main verify for receipient in uncached-mode.
181 JH/22 Retire historical build files to an "unsupported" subdir. These are
182 defined as "ones for which we have no current evidence of testing".
184 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
185 if present. Previously it was ignored.
187 JH/24 Start using specified-initialisers in C structure init coding. This is
188 a C99 feature (it's 2017, so now considered safe).
190 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
191 if was a fixed-sized field and bitmask ops via macros; it is now more
194 PP/07 GitHub PR 56: Apply MariaDB build fix.
195 Patch provided by Jaroslav Škarvada.
197 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
198 during Coverity cleanups [4.87 JH/47]
199 Diagnosis and fix provided by Michael Fischer v. Mollard.
201 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
202 the right size to place the terminating semicolon on its own folded
203 line, the header hash was calculated to an incorrect value thanks to
204 the (relaxed) space the fold became.
206 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
207 and confused the parent.
209 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
210 which could crash as a result. This could lead to undeliverable messages.
212 JH/28 Logging: "next input sent too soon" now shows where input was truncated
215 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
216 matters on fast-turnover and PID-randomising systems, which were getting
217 out-of-order delivery.
219 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
220 a possibly-overlapping copy. The symptom was that "Remote host closed
221 connection in response to HELO" was logged instead of the actual 4xx
224 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
225 Previously only that bufferd was discarded, resulting in SYMTP command
228 JH/32 DKIM: when a message has multiple signatures matching an identity given
229 in dkim_verify_signers, run the dkim acl once for each. Previously only
230 one run was done. Bug 2189.
232 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
233 "panic the current process" to "deliberately defer". The panic log is
234 still written with the problem list name; the mail and reject logs now
235 get a temp-reject line for the message that was being handled, saying
236 something like "domains check lookup or other defer". The SMTP 451
237 message is still "Temporary local problem".
239 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
240 A crafted sequence of BDAT commands could result in in-use memory beeing
241 freed. CVE-2017-16943.
243 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
244 from SMTP input. Previously it was always done; now only done for DATA
245 and not BDAT commands. CVE-2017-16944.
247 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
248 to the message (such as an overlong header line). Previously this was
249 not done and we did not exit BDAT mode. Followon from the previous item
250 though a different problem.
256 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
257 than -2003 did; needs libidn2 in addition to libidn.
259 JH/02 The path option on a pipe transport is now expanded before use.
261 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
262 Patch provided by "Björn", documentation fix added too.
264 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
265 missing a wire-to-host endian conversion.
267 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
268 close after a BDAT command line could be taken as a following command,
269 giving a synch failure. Fix by only checking for synch immediately
270 before acknowledging the chunk.
272 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
273 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
274 macro. Patches provided by Josh Soref.
276 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
277 Previously we did not; the RFC seems ambiguous and VRFY is not listed
278 by IANA as a service extension. However, John Klensin suggests that we
281 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
282 the dkim code may be unix-mode line endings rather than smtp wire-format
283 CRLF, so prepend a CR to any bare LF.
285 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
286 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
288 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
289 the first were themselves being wrongly included in the feed into dkim
290 processing; with most chunk sizes in use this resulted in an incorrect
291 body hash calculated value.
293 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
294 DKIM signature block, for verification. Although advised against by
295 standards it is specifically not ruled illegal.
297 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
299 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
300 missing a body hash (the bh= tag).
302 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
303 It seems that HAProxy sends the Proxy Protocol information in clear and
304 only then does a TLS startup, so do the same.
306 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
307 TCP connections (such as for Spamd) unless the daemon successfully set
308 Fast Open mode on its listening sockets. This fixes breakage seen on
309 too-old kernels or those not configured for Fast Open, at the cost of
310 requiring both directions being enabled for TFO, and TFO never being used
311 by non-daemon-related Exim processes.
313 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
314 endings, at least on the first header line. Try to canonify any that get
315 past that check, despite the cost.
317 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
318 now limited to an arbitrary five deep, while parsing addresses with the
319 strip_excess_angle_brackets option enabled.
321 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
322 instead leave the unprompted TLS handshake in socket buffer for the
323 TLS library to consume.
325 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
327 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
329 JH/16 Drop variables when they go out of scope. Memory management drops a whole
330 region in one operation, for speed, and this leaves assigned pointers
331 dangling. Add checks run only under the testsuite which checks all
332 variables at a store-reset and panics on a dangling pointer; add code
333 explicitly nulling out all the variables discovered. Fixes one known
334 bug: a transport crash, where a dangling pointer for $sending_ip_address
335 originally assigned in a verify callout, is re-used.
337 PP/06 Drop '.' from @INC in various Perl scripts.
339 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
341 PP/08 Reduce a number of compilation warnings under clang; building with
342 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
343 should be warning-free.
345 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
347 HS/01 Fix portability problems introduced by PP/08 for platforms where
348 realloc(NULL) is not equivalent to malloc() [SunOS et al].
350 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
351 chunk. This allows us to accept broken chunked messages. We need a more
352 general solution here.
354 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
355 already-broken messages in the queue.
357 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
359 JH/19 Fix reference counting bug in routing-generated-address tracking.
365 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
366 supports it and a size is available (ie. the sending peer gave us one).
368 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
369 years of being deprecated). The replacements are the ACLs
370 acl_smtp_mime and acl_not_smtp_mime.
372 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
373 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
374 or even an in-clear connection were permitted. Now, if the host lookup
375 was dnssec and dane was requested then the host is only used if the
376 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
377 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
378 if one fails this test.
379 This means that a poorly-configured remote DNS will make it incommunicado;
380 but it protects against a DNS-interception attack on it.
382 JH/04 Bug 1810: make continued-use of an open smtp transport connection
383 non-noisy when a race steals the message being considered.
385 JH/05 If main configuration option tls_certificate is unset, generate a
386 self-signed certificate for inbound TLS connections.
388 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
389 in rewrites and routers.
391 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
392 and logged a warning sing 4.83; now they are a configuration file error.
394 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
395 (lacking @domain). Apply the same qualification processing as RCPT.
397 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
399 JH/10 Support ${sha256:} applied to a string (as well as the previous
402 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
403 a cutthrough deliver is pending, as we always want to make a connection.
404 This also avoids re-routing the message when later placing the cutthrough
405 connection after a verify cache hit.
406 Do not update it with the verify result either.
408 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
409 when routing results in more than one destination address.
411 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
412 signing (which inhibits the cutthrough capability). Previously only
413 the presence of an option was tested; now an expansion evaluating as
414 empty is permissible (obviously it should depend only on data available
415 when the cutthrough connection is made).
417 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
418 the relevant preceding SMTP command did not note the pipelining mode.
420 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
421 Previously they were not counted.
423 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
424 as one having no matching records. Previously we deferred the message
425 that needed the lookup.
427 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
428 distinguished as "(=".
430 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
431 for missing MX records. Previously it only worked for missing A records.
433 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
435 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
436 after the data-go-ahead and data-ack. Patch from Jason Betts.
438 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
439 even for a "none" policy. Patch from Tony Meyer.
441 JH/22 Fix continued use of a connection for further deliveries. If a port was
442 specified by a router, it must also match for the delivery to be
445 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
446 When one of the recipients of a message was unsuitable for the connection
447 (has no matching addresses), we lost track of needing to mark it
448 deferred. As a result mail would be lost.
450 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
452 JH/25 Decoding ACL controls is now done using a binary search; the source code
453 takes up less space and should be simpler to maintain. Merge the ACL
454 condition decode tables also, with similar effect.
456 JH/26 Fix problem with one_time used on a redirect router which returned the
457 parent address unchanged. A retry would see the parent address marked as
458 delivered, so not attempt the (identical) child. As a result mail would
461 JH/27 Fix a possible security hole, wherein a process operating with the Exim
462 UID can gain a root shell. Credit to http://www.halfdog.net/ for
463 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
466 JH/28 Enable {spool,log} filesystem space and inode checks as default.
467 Main config options check_{log,spool}_{inodes,space} are now
468 100 inodes, 10MB unless set otherwise in the configuration.
470 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
471 Previously it only applied to the main-section connection policy
474 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
476 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
477 by me. Added RFC7919 DH primes as an alternative.
479 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
480 are not in the system include path.
482 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
483 GnuTLS, when a session startup failed (eg because the client disconnected)
484 Exim did stdio operations after fclose. This was exposed by a recent
485 change which nulled out the file handle after the fclose.
487 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
488 signed directly by the cert-signing cert, rather than an intermediate
489 OCSP-signing cert. This is the model used by LetsEncrypt.
491 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
493 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
494 an incoming connection.
496 HS/02 Bug 1802: Do not half-close the connection after sending a request
499 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
500 fallback to "prime256v1".
502 JH/34 SECURITY: Use proper copy of DATA command in error message.
503 Could leak key material. Remotely exploitable. CVE-2016-9963.
509 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
510 and 3.4.4 - once the server is enabled to respond to an OCSP request
511 it does even when not requested, resulting in a stapling non-aware
512 client dropping the TLS connection.
514 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
515 support variable-length bit vectors. No functional change.
517 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
518 The I= interface field on outgoing lines is now after the H= remote
519 host field, same as incoming lines. There is a separate
520 outgoing_interface log selector which allows you to disable the
523 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
524 If not running log_selector +smtp_connection the mainlog would be held
525 open indefinitely after a "too many connections" event, including to a
526 deleted file after a log rotate. Leave the per net connection logging
527 leaving it open for efficiency as that will be quickly detected by the
528 check on the next write.
530 HS/01 Bug 1671: Fix post transport crash.
531 Processing the wait-<transport> messages could crash the delivery
532 process if the message IDs didn't exist for some reason. When
533 using 'split_spool_directory=yes' the construction of the spool
534 file name failed already, exposing the same netto behaviour.
536 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
537 mime_regex ACL conditions.
539 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
540 to DSN fail messages (bounces): remote IP, remote greeting, remote response
541 to HELO, local diagnostic string.
543 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
544 log line to debug. Even when configured with a tls authenticator many
545 client connections are expected to not authenticate in this way, so
546 an authenticate fail is not an error.
548 HS/02 Add the Exim version string to the process info. This way exiwhat
549 gives some more detail about the running daemon.
551 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
552 matter for fast-change records such as DNSBLs.
554 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
555 retry record, even if constant. There may be multiple transports with
556 different interface settings and the retry behaviour needs to be kept
559 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
561 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
563 JH/10 Bug 840: fix log_defer_output option of pipe transport
565 JH/11 Bug 830: use same host for all RCPTS of a message, even under
566 hosts_randomize. This matters a lot when combined with mua_wrapper.
568 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
569 ${quote_pgsql:<string>} operator.
571 JH/13 Bug 1708: avoid misaligned access in cached lookup.
573 JH/14 Change header file name for freeradius-client. Relevant if compiling
574 with Radius support; from the Gentoo tree and checked under Fedora.
576 JH/15 Bug 1712: Introduce $prdr_requested flag variable
578 JH/16 Bug 1714: Permit an empty string as expansion result for transport
579 option transport_filter, meaning no filtering.
581 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
583 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
584 defaults to "*" (all hosts). The variable is now available when not built
585 with TLS, default unset, mainly to enable keeping the testsuite sane.
586 If a server certificate is not supplied (via tls_certificate) an error is
587 logged, and clients will find TLS connections fail on startup. Presumably
588 they will retry in-clear.
589 Packagers of Exim are strongly encouraged to create a server certificate
590 at installation time.
592 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
593 with the $config_file variable.
595 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
596 in transport context, after the attempt, and per-recipient. The latter type
597 is per host attempted. The event data is the error message, and the errno
598 information encodes the lookup type (A vs. MX) used for the (first) host,
599 and the trailing two digits of the smtp 4xx response.
601 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
602 to write to mainlog (or rejectlog, paniclog) in the window between file
603 creation and permissions/ownership being changed. Particularly affects
604 installations where exicyclog is run as root, rather than exim user;
605 result is that the running daemon panics and dies.
607 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
609 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
610 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
611 "pri" and "weight". Note that the previous implicit priority given by the
612 list order is no longer honoured.
614 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
617 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
618 by defining SUPPORT_SOCKS.
620 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
621 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
622 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
623 variables are renamed to proxy_{local,external}_{address,port}.
625 JH/27 Move Internationalisation support from Experimental to mainline, enabled
626 for a build by defining SUPPORT_I18N
628 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
629 of the query string, and make ${quote_redis:} do that quoting.
631 JH/29 Move Events support from Experimental to mainline, enabled by default
632 and removable for a build by defining DISABLE_EVENT.
634 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
636 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
637 cached by the daemon.
639 JH/32 Move Redis support from Experimental to mainline, enabled for a build
640 by defining LOOKUP_REDIS. The libhiredis library is required.
642 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
643 keys are given for lookup.
645 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
646 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
647 only supported when built with TLS support. The PolarSSL SHA routines
648 are still used when the TLS library is too old for convenient support.
650 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
651 openssl_options), for security. OpenSSL forces this from version 1.1.0
652 server-side so match that on older versions.
654 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
655 allocation for $value could be released as the expansion processing
656 concluded, but leaving the global pointer active for it.
658 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
659 and to use the domains and local_parts ACL conditions.
661 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
662 incorrectly not doubled on cutthrough transmission, hence seen as a
663 body-termination at the receiving system - resulting in truncated mails.
664 Commonly the sender saw a TCP-level error, and retransmitted the message
665 via the normal store-and-forward channel. This could result in duplicates
666 received - but deduplicating mailstores were liable to retain only the
667 initial truncated version.
669 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
671 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
673 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
674 we're in there, support oversigning also; bug 1309.
676 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
678 HS/04 Add support for keep_environment and add_environment options.
680 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
681 either intentional arithmetic overflow during PRNG, or testing config-
684 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
685 delivery resulted in actual delivery. Cancel cutthrough before DATA
688 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
689 rejects a recipient: pass the reject to the originator.
691 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
692 Many were false-positives and ignorable, but it's worth fixing the
695 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
696 for the new environment-manipulation done at startup. Move the routines
697 from being local to tls.c to being global via the os.c file.
699 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
700 an extract embedded as result-arg for a map, the first arg for extract
701 is unavailable so we cannot tell if this is a numbered or keyed
702 extraction. Accept either.
708 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
711 JH/02 The smtp transport option "multi_domain" is now expanded.
713 JH/03 The smtp transport now requests PRDR by default, if the server offers
716 JH/04 Certificate name checking on server certificates, when exim is a client,
717 is now done by default. The transport option tls_verify_cert_hostnames
718 can be used to disable this per-host. The build option
719 EXPERIMENTAL_CERTNAMES is withdrawn.
721 JH/05 The value of the tls_verify_certificates smtp transport and main options
722 default to the word "system" to access the system default CA bundle.
723 For GnuTLS, only version 3.0.20 or later.
725 JH/06 Verification of the server certificate for a TLS connection is now tried
726 (but not required) by default. The verification status is now logged by
727 default, for both outbound TLS and client-certificate supplying inbound
730 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
733 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
734 Status Notification (bounce) messages are now MIME format per RFC 3464.
735 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
736 under the control of the dsn_advertise_hosts option, and routers may
737 have a dsn_lasthop option.
739 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
740 default, modifiable by a malware= option. The list separator for
741 the options can now be changed in the usual way. Bug 68.
743 JH/10 The smtp_receive_timeout main option is now expanded before use.
745 JH/11 The incoming_interface log option now also enables logging of the
746 local interface on delivery outgoing connections.
748 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
749 if the interface and destination host and port all match.
751 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
754 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
755 Patch from Andrew Lewis.
757 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
758 now supports optional time-restrictions, weighting, and priority
759 modifiers per server. Patch originally by <rommer@active.by>.
761 JH/16 The spamd_address main option now supports a mixed list of local
762 and remote servers. Remote servers can be IPv6 addresses, and
763 specify a port-range.
765 JH/17 Bug 68: The spamd_address main option now supports an optional
766 timeout value per server.
768 JH/18 Bug 1581: Router and transport options headers_add/remove can
769 now have the list separator specified.
771 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
774 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
777 JH/21 Support for the A6 type of dns record is withdrawn.
779 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
780 rather than the verbs used.
782 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
783 from 255 to 1024 chars.
785 JH/24 Verification callouts now attempt to use TLS by default.
787 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
788 are generic router options now. The defaults didn't change.
790 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
791 Original patch from Alexander Shikoff, worked over by JH.
793 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
794 files if ZCAT_COMMAND is not executable.
796 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
798 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
800 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
801 Normally benign, it bites when the pair was led to by a CNAME;
802 modern usage is to not canonicalize the domain to a CNAME target
803 (and we were inconsistent anyway for A-only vs AAAA+A).
805 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
807 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
808 when evaluating $sender_host_dnssec.
810 JH/31 Check the HELO verification lookup for DNSSEC, adding new
811 $sender_helo_dnssec variable.
813 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
815 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
817 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
819 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
820 documented as working, but never had. Support all but $spam_report.
822 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
823 added for tls authenticator.
825 HS/03 Add perl_taintmode main config option
831 TL/01 When running the test suite, the README says that variables such as
832 no_msglog_check are global and can be placed anywhere in a specific
833 test's script, however it was observed that placement needed to be near
834 the beginning for it to behave that way. Changed the runtest perl
835 script to read through the entire script once to detect and set these
836 variables, reset to the beginning of the script, and then run through
837 the script parsing/test process like normal.
839 TL/02 The BSD's have an arc4random API. One of the functions to induce
840 adding randomness was arc4random_stir(), but it has been removed in
841 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
842 function when detected.
844 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
845 cause callback expansion.
847 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
848 syntax errors in an expansion can be treated as a string instead of
849 logging or causing an error, due to the internal use of bool_lax
850 instead of bool when processing it.
852 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
853 server certificates when making smtp deliveries.
855 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
857 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
859 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
861 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
862 Merged patch from Sebastian Wiedenroth.
864 JH/05 Fix results-pipe from transport process. Several recipients, combined
865 with certificate use, exposed issues where response data items split
866 over buffer boundaries were not parsed properly. This eventually
867 resulted in duplicates being sent. This issue only became common enough
868 to notice due to the introduction of connection certificate information,
869 the item size being so much larger. Found and fixed by Wolfgang Breyha.
871 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
872 size buffer was used, resulting in syntax errors when an expansion
875 JH/07 Add support for directories of certificates when compiled with a GnuTLS
876 version 3.3.6 or later.
878 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
879 is EXPERIMENTAL_EVENT, the main-configuration and transport options
880 both become "event_action", the variables become $event_name, $event_data
881 and $event_defer_errno. There is a new variable $verify_mode, usable in
882 routers, transports and related events. The tls:cert event is now also
883 raised for inbound connections, if the main configuration event_action
886 TL/06 In test suite, disable OCSP for old versions of openssl which contained
887 early OCSP support, but no stapling (appears to be less than 1.0.0).
889 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
890 server certificate names available under the smtp transport option
891 "tls_verify_cert_hostname" now do not permit multi-component wildcard
894 JH/10 Time-related extraction expansions from certificates now use the main
895 option "timezone" setting for output formatting, and are consistent
896 between OpenSSL and GnuTLS compilations. Bug 1541.
898 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
899 encoded parameter in the incoming message. Bug 1558.
901 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
902 include certificate info, eximon was claiming there were spoolfile
905 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
907 JH/14 Log delivery-related information more consistently, using the sequence
908 "H=<name> [<ip>]" wherever possible.
910 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
911 are problematic for Debian distribution, omit them from the release
914 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
916 JH/16 Fix string representation of time values on 64bit time_t architectures.
919 JH/17 Fix a null-indirection in certextract expansions when a nondefault
920 output list separator was used.
925 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
926 checkers that were complaining about end of non-void function with no
929 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
930 This was a regression introduced in 4.83 by another bugfix.
932 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
934 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
935 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
941 TF/01 Correctly close the server side of TLS when forking for delivery.
943 When a message was received over SMTP with TLS, Exim failed to clear up
944 the incoming connection properly after forking off the child process to
945 deliver the message. In some situations the subsequent outgoing
946 delivery connection happened to have the same fd number as the incoming
947 connection previously had. Exim would try to use TLS and fail, logging
948 a "Bad file descriptor" error.
950 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
951 utilities have not been installed.
953 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
954 temporary space as the ACL may create new global variables.
956 TL/01 LDAP support uses per connection or global context settings, depending
957 upon the detected version of the libraries at build time.
959 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
960 to extract and use the src ip:port in logging and expansions as if it
961 were a direct connection from the outside internet. PPv2 support was
962 updated based on HAProxy spec change in May 2014.
964 JH/02 Add ${listextract {number}{list}{success}{fail}}.
966 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
967 Properly escape header and check for NULL return.
969 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
972 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
974 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
975 characters in header names, implemented as a verify condition.
976 Contributed by Michael Fischer v. Mollard.
978 TL/05 Rename SPF condition results err_perm and err_temp to standardized
979 results permerror and temperror. Previous values are deprecated but
980 still accepted. In a future release, err_perm and err_temp will be
981 completely removed, which will be a backward incompatibility if the
982 ACL tests for either of these two old results. Patch contributed by
983 user bes-internal on the mailing list.
985 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
987 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
988 selectors, in both main and reject logs.
990 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
993 JH/07 Add malware type "sock" for talking to simple daemon.
995 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
997 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
998 routers/transports under cutthrough routing.
1000 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1001 numbers. Touch up "bool" conditional to keep the same definition.
1003 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1005 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1007 JH/12 Expand items in router/transport headers_add or headers_remove lists
1008 individually rather than the list as a whole. Bug 1452.
1010 Required for reasonable handling of multiple headers_ options when
1011 they may be empty; requires that headers_remove items with embedded
1012 colons must have them doubled (or the list-separator changed).
1014 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1015 view the policy declared in the DMARC record. Currently, $dmarc_status
1016 is a combined value of both the record presence and the result of the
1019 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1021 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1022 dnslookup router and the smtp transport (applying to the forward
1025 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1026 of ldap servers used for a specific lookup. Patch provided by Heiko
1029 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1030 New variable $lookup_dnssec_authenticated for observability.
1032 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1033 Patch submitted by Lars Timman.
1035 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1037 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1038 Requires trusted mode and valid format message id, aborts otherwise.
1039 Patch contributed by Heiko Schlichting.
1041 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1042 certextract with support for various fields. Bug 1358.
1044 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1045 is requested by default, modifiable by smtp transport option
1048 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1049 operate on certificate variables to give certificate fingerprints
1050 Also new ${sha256:cert_variable}.
1052 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1054 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1055 Christian Aistleitner.
1057 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1059 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1060 file. Patch from Wolfgang Breyha.
1062 JH/25 Expand the coverage of the delivery $host and $host_address to
1063 client authenticators run in verify callout. Bug 1476.
1065 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1066 align with daemon_smtp_ports. Bug 72.
1068 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1069 support and error reporting did not work properly.
1071 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1072 and is readable. Patch from Andrew Colin Kissa.
1074 TL/14 Enhance documentation of ${run expansion and how it parses the
1075 commandline after expansion, particularly in the case when an
1076 unquoted variable expansion results in an empty value.
1078 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1080 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1081 by renaming away from it. Addresses GH issue 15, reported by
1084 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1086 TL/15 SECURITY: prevent double expansion in math comparison functions
1087 (can expand unsanitized data). Not remotely exploitable.
1094 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1096 PP/02 Make -n do something, by making it not do something.
1097 When combined with -bP, the name of an option is not output.
1099 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1102 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1103 $sender_host_name and config options to manage this, and basic check
1106 PP/05 DSCP support for outbound connections and control modifier for inbound.
1108 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1109 (Only plugin which currently uses this is kerberos4, which nobody should
1110 be using, but we should make it available and other future plugins might
1111 conceivably use it, even though it would break NAT; stuff *should* be
1112 using channel bindings instead).
1114 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1115 name; added for Sendmail compatibility; requires admin caller.
1116 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1117 just ignore it); requires trusted caller.
1118 Also parse but ignore: -Ac -Am -X<logfile>
1121 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1123 TL/02 Add +smtp_confirmation as a default logging option.
1125 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1126 Patch by Magnus Holmgren from 2007-02-20.
1128 TL/04 Bugzilla 1281 - Spec typo.
1129 Bugzilla 1283 - Spec typo.
1130 Bugzilla 1290 - Spec grammar fixes.
1132 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1134 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1136 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1137 mailing list by by Dmitry Isaikin.
1139 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1141 JH/02 Support "G" suffix to numbers in ${if comparisons.
1143 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1145 NM/01 Bugzilla 1197 - Spec typo
1146 Bugzilla 1196 - Spec examples corrections
1148 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1150 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1151 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1154 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1155 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1157 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1158 "acl {{name}{arg}...}", and optional args on acl condition
1161 JH/05 Permit multiple router/transport headers_add/remove lines.
1163 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1165 JH/07 Avoid using a waiting database for a single-message-only transport.
1166 Performance patch from Paul Fisher. Bugzilla 1262.
1168 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1171 JH/09 Add $headers_added variable, with content from use of ACL modifier
1172 add_header (but not yet added to the message). Bugzilla 199.
1174 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1175 Pulled from Bugzilla 817 by Wolfgang Breyha.
1177 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1179 (nb: this is the same fix as in Exim 4.80.1)
1181 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1184 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1187 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1189 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1190 advertise SMTP AUTH mechanism to us, instead of a generic
1191 protocol violation error. Also, make Exim more robust to bad
1192 data from the Dovecot auth socket.
1194 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1196 When a queue runner is handling a message, Exim first routes the
1197 recipient addresses, during which it prunes them based on the retry
1198 hints database. After that it attempts to deliver the message to
1199 any remaining recipients. It then updates the hints database using
1202 So if a recipient address works intermittently, it can get repeatedly
1203 deferred at routing time. The retry hints record remains fresh so the
1204 address never reaches the final cutoff time.
1206 This is a fairly common occurrence when a user is bumping up against
1207 their storage quota. Exim had some logic in its local delivery code
1208 to deal with this. However it did not apply to per-recipient defers
1209 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1211 This change adds a proper retry rule check during routing so that the
1212 final cutoff time is checked against the message's age. We only do
1213 this check if there is an address retry record and there is not a
1214 domain retry record; this implies that previous attempts to handle
1215 the address had the retry_use_local_parts option turned on. We use
1216 this as an approximation for the destination being like a local
1217 delivery, as in LMTP.
1219 I suspect this new check makes the old local delivery cutoff check
1220 redundant, but I have not verified this so I left the code in place.
1222 TF/02 Correct gecos expansion when From: is a prefix of the username.
1224 Test 0254 submits a message to Exim with the header
1228 When I ran the test suite under the user fanf2, Exim expanded
1229 the header to contain my full name, whereas it should have added
1230 a Resent-Sender: header. It erroneously treats any prefix of the
1231 username as equal to the username.
1233 This change corrects that bug.
1235 GF/01 DCC debug and logging tidyup
1236 Error conditions log to paniclog rather than rejectlog.
1237 Debug lines prefixed by "DCC: " to remove any ambiguity.
1239 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1241 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1242 Bug spotted by Jeremy Harris; was flawed since initial commit.
1243 Would have resulted in OCSP responses post-SNI triggering an Exim
1244 NULL dereference and crash.
1246 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1248 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1249 Bug detection, analysis and fix by Samuel Thibault.
1250 Bugzilla 1331, Debian bug #698092.
1252 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1254 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1255 Server implementation by Todd Lyons, client by JH.
1256 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1257 config variable "prdr_enable" controls whether the server
1258 advertises the facility. If the client requests PRDR a new
1259 acl_data_smtp_prdr ACL is called once for each recipient, after
1260 the body content is received and before the acl_smtp_data ACL.
1261 The client is controlled by both of: a hosts_try_prdr option
1262 on the smtp transport, and the server advertisement.
1263 Default client logging of deliveries and rejections involving
1264 PRDR are flagged with the string "PRDR".
1266 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1267 fclose(). Diagnosis by Todd Lyons.
1269 PP/17 Update configure.default to handle IPv6 localhost better.
1270 Patch by Alain Williams (plus minor tweaks).
1273 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1274 This is now consistent with GnuTLS, and is now documented: the
1275 previous undocumented portable approach to treating the option as
1276 unset was to force an expansion failure. That still works, and
1277 an empty string is now equivalent.
1279 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1280 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1281 not performing validation itself.
1283 PP/20 Added force_command boolean option to pipe transport.
1284 Patch from Nick Koston, of cPanel Inc.
1286 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1289 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1291 PP/21 Fix eximon continuous updating with timestamped log-files.
1292 Broken in a format-string cleanup in 4.80, missed when I repaired the
1293 other false fix of the same issue.
1294 Report and fix from Heiko Schlichting.
1297 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1298 Report from Prashanth Katuri.
1300 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1301 It's SecureTransport, so affects any MacOS clients which use the
1302 system-integrated TLS libraries, including email clients.
1304 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1305 using a MIME ACL for non-SMTP local injection.
1306 Report and assistance in diagnosis by Warren Baker.
1308 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1310 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1312 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1313 last id that failed so it may be referenced in subsequent ACL's.
1315 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1318 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1319 ldap library initialization, allowing self-signed CA's to be
1320 used. Also properly sets require_cert option later in code by
1321 using NULL (global ldap config) instead of ldap handle (per
1322 session). Bug diagnosis and testing by alxgomz.
1324 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1325 the src/util/ subdirectory.
1327 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1328 renamed to Transport Post Delivery Action by Jeremy Harris, as
1331 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1332 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1333 redis_servers = needs to be configured which will be used by the redis
1334 lookup. Patch from Warren Baker, of The Packet Hub.
1336 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1338 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1339 hostname or reverse DNS when processing a host list. Used suggestions
1340 from multiple comments on this bug.
1342 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1344 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1345 Missed a few lines, added it to make the runtest require no keyboard
1348 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1349 contains upper case chars. Make router use caseful_local_part.
1351 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1352 support when GnuTLS has been built with p11-kit.
1358 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1360 This, or similar/improved, will also be change PP/11 of 4.82.
1366 PP/01 Handle short writes when writing local log-files.
1367 In practice, only affects FreeBSD (8 onwards).
1368 Bugzilla 1053, with thanks to Dmitry Isaikin.
1370 NM/01 Bugzilla 949 - Documentation tweak
1372 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1375 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1377 PP/02 Implemented gsasl authenticator.
1379 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1381 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1382 `pkg-config foo` for cflags/libs.
1384 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1385 with rest of GSASL and with heimdal_gssapi.
1387 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1388 `pkg-config foo` for cflags/libs for the TLS implementation.
1390 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1391 properties get this fed in as external SSF. A number of robustness
1392 and debugging improvements to the cyrus_sasl authenticator.
1394 PP/08 cyrus_sasl server now expands the server_realm option.
1396 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1397 Patch by Jeremy Harris.
1399 PP/10 Added dbmjz lookup type.
1401 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1403 PP/12 MAIL args handles TAB as well as SP, for better interop with
1404 non-compliant senders.
1405 Analysis and variant patch by Todd Lyons.
1407 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1408 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1409 Patch from Dirk Mueller <dmueller@suse.com>
1411 PP/13 tls_peerdn now print-escaped for spool files.
1412 Observed some $tls_peerdn in wild which contained \n, which resulted
1413 in spool file corruption.
1415 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1416 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1417 or write after TLS renegotiation, which otherwise led to messages
1420 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1421 as a tracking header (ie: a signed header comes before the signature).
1422 Patch from Wolfgang Breyha.
1424 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1425 comma-sep list; embedded commas doubled.
1427 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1429 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1431 Report and patch from Dmitry Banschikov.
1433 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1434 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1435 failures. We appear to now support TLS1.1+ with Exim.
1437 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1438 lets Exim select keys and certificates based upon TLS SNI from client.
1439 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1440 before an outbound SMTP session. New log_selector, +tls_sni.
1442 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1443 NULL dereference. Report and patch from Alun Jones.
1445 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1446 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1447 Not seeing resolver debug output on NetBSD, but suspect this is a
1448 resolver implementation change.
1450 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1451 Left warnings. Added "eximon gdb" invocation mode.
1453 PP/21 Defaulting "accept_8bitmime" to true, not false.
1455 PP/22 Added -bw for inetd wait mode support.
1457 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1458 locate the relevant includes and libraries. Made this the default.
1460 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1461 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1463 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1464 This may cause build issues on older platforms.
1466 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1467 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1468 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1469 Added SNI support via GnuTLS too.
1470 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1472 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1474 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1475 Applied second patch from Janne, implementing suggestion to default
1476 multiple-strings-in-record handling to match SPF spec.
1478 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1480 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1481 read-only, out of scope).
1482 Patch from Wolfgang Breyha, report from Stuart Northfield.
1484 PP/29 Fix three issues highlighted by clang analyser static analysis.
1485 Only crash-plausible issue would require the Cambridge-specific
1486 iplookup router and a misconfiguration.
1487 Report from Marcin Mirosław.
1489 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1491 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1492 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1493 As part of this, removing so much warning spew let me fix some minor
1494 real issues in debug logging.
1496 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1497 assignment on my part. Fixed.
1499 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1500 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1501 Janne Snabb (who went above and beyond: thank you).
1503 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1504 string otherwise requires a connection and a bunch more work and it's
1505 relatively easy to get wrong. Should also expose TLS library linkage
1508 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1509 64-bit ${eval} (JH/03).
1511 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1512 GNU libc to support some of the 64-bit stuff, should not lead to
1513 conflicts. Defined before os.h is pulled in, so if a given platform
1514 needs to override this, it can.
1516 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1517 protection layer was required, which is not implemented.
1518 Bugzilla 1254, patch from Wolfgang Breyha.
1520 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1521 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1522 tls_dhparam take prime identifiers. Also unbreak combination of
1523 OpenSSL+DH_params+TLSSNI.
1525 PP/39 Disable SSLv2 by default in OpenSSL support.
1531 PP/01 Solaris build fix for Oracle's LDAP libraries.
1532 Bugzilla 1109, patch from Stephen Usher.
1534 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1536 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1539 TF/02 Fix a couple more cases where we did not log the error message
1540 when unlink() failed. See also change 4.74-TF/03.
1542 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1543 lock up or crash if it happened to be inside a call to libc when it
1544 got a SIGUSR1 from exiwhat.
1546 The SIGUSR1 handler appends the current process status to the process
1547 log which is later printed by exiwhat. It used to use the general
1548 purpose logging code to do this, but several functions it calls are
1549 not safe for signals.
1551 The new output code in the SIGUSR1 handler is specific to the process
1552 log, and simple enough that it's easy to inspect for signal safety.
1553 Removing some special cases also simplifies the general logging code.
1554 Removing the spurious timestamps from the process log simplifies
1557 TF/04 Improved ratelimit ACL condition.
1559 The /noupdate option has been deprecated in favour of /readonly which
1560 has clearer semantics. The /leaky, /strict, and /readonly update modes
1561 are mutually exclusive. The update mode is no longer included in the
1562 database key; it just determines when the database is updated. (This
1563 means that when you upgrade Exim will forget old rate measurements.)
1565 Exim now checks that the per_* options are used with an update mode that
1566 makes sense for the current ACL. For example, when Exim is processing a
1567 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1568 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1569 must specify per_mail/readonly. If you omit the update mode it defaults to
1570 /leaky where that makes sense (as before) or /readonly where required.
1572 The /noupdate option is now undocumented but still supported for
1573 backwards compatibility. It is equivalent to /readonly except that in
1574 ACLs where /readonly is required you may specify /leaky/noupdate or
1575 /strict/noupdate which are treated the same as /readonly.
1577 A useful new feature is the /count= option. This is a generalization
1578 of the per_byte option, so that you can measure the throughput of other
1579 aggregate values. For example, the per_byte option is now equivalent
1580 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1582 The per_rcpt option has been generalized using the /count= mechanism
1583 (though it's more complicated than the per_byte equivalence). When it is
1584 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1585 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1586 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1587 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1588 Note that using per_rcpt with a non-readonly update mode in more than
1589 one ACL will cause the recipients to be double-counted. (The per_mail
1590 and per_byte options don't have this problem.)
1592 The handling of very low rates has changed slightly. If the computed rate
1593 is less than the event's count (usually one) then this event is the first
1594 after a long gap. In this case the rate is set to the same as this event's
1595 count, so that the first message of a spam run is counted properly.
1597 The major new feature is a mechanism for counting the rate of unique
1598 events. The new per_addr option counts the number of different
1599 recipients that someone has sent messages to in the last time period. It
1600 behaves like per_rcpt if all the recipient addresses are different, but
1601 duplicate recipient addresses do not increase the measured rate. Like
1602 the /count= option this is a general mechanism, so the per_addr option
1603 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1604 example, measure the rate that a client uses different sender addresses
1605 with the options per_mail/unique=$sender_address. There are further
1606 details in the main documentation.
1608 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1610 TF/06 Removed a few PCRE remnants.
1612 TF/07 Automatically extract Exim's version number from tags in the git
1613 repository when doing development or release builds.
1615 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1616 Bugzilla 879. Patch from Paul Fisher.
1618 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1619 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1622 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1624 PP/05 Variable $av_failed, true if the AV scanner deferred.
1625 Bugzilla 1078. Patch from John Horne.
1627 PP/06 Stop make process more reliably on build failure.
1628 Bugzilla 1087. Patch from Heiko Schlittermann.
1630 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1631 Bugzilla 1089. Patch from Heiko Schlittermann.
1633 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1634 Bugzilla 1131. Patch from Holger Weiß.
1636 PP/09 Handle IPv6 addresses with SPF.
1637 Bugzilla 860. Patch from Wolfgang Breyha.
1639 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1641 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1644 PP/11 match_* no longer expand right-hand-side by default.
1645 New compile-time build option, EXPAND_LISTMATCH_RHS.
1646 New expansion conditions, "inlist", "inlisti".
1648 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1650 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1652 PP/14 fix log_write() format string regression from TF/03.
1653 Bugzilla 1152. Patch from Dmitry Isaikin.
1659 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1661 PP/02 Harmonised TLS library version reporting; only show if debugging.
1662 Layout now matches that introduced for other libraries in 4.74 PP/03.
1664 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1666 PP/04 New "dns_use_edns0" global option.
1668 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1671 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1672 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1674 TK/01 Updated PolarSSL code to 0.14.2.
1675 Bugzilla 1097. Patch from Andreas Metzler.
1677 PP/07 Catch divide-by-zero in ${eval:...}.
1678 Fixes bugzilla 1102.
1680 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1683 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1684 format-string attack -- SECURITY: remote arbitrary code execution.
1686 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1687 time unintentionally subject to list matching rules, letting the header
1688 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1689 arbitrary string expansion). This allowed for information disclosure.
1691 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1692 INT_MIN/-1 -- value coerced to INT_MAX.
1698 NM/01 Workaround for PCRE version dependency in version reporting
1701 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1702 This fixes portability to compilers other than gcc, notably
1703 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1705 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1706 makefiles for portability to HP-UX and POSIX correctness.
1708 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1709 Also via indented variable definition in the Makefile.
1710 (Debugging by Oliver Heesakkers).
1712 PP/02 Restore caching of spamd results with expanded spamd_address.
1713 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1715 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1716 Improves build reliability. Fix from: Frank Elsner
1718 NM/02 Fix wide character breakage in the rfc2047 coding
1719 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1721 NM/03 Allow underscore in dnslist lookups
1722 Fixes bug 1026. Patch from Graeme Fowler
1724 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1725 Code patches from Adam Ciarcinski of NetBSD.
1727 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1730 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1731 is logged, to avoid truncation. Patch from John Horne.
1733 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1734 Patch from Jakob Hirsch.
1736 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1737 SQL string expansion failure details.
1738 Patch from Andrey Oktyabrski.
1740 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1741 Patch from Simon Arlott.
1743 PP/09 New lookups functionality failed to compile on old gcc which rejects
1744 extern declarations in function scope.
1745 Patch from Oliver Fleischmann
1747 PP/10 Use sig_atomic_t for flags set from signal handlers.
1748 Check getgroups() return and improve debugging.
1749 Fixed developed for diagnosis in bug 927 (which turned out to be
1752 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1753 Patch from Mark Zealey.
1755 PP/12 Bugzilla 1056: Improved spamd server selection.
1756 Patch from Mark Zealey.
1758 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1759 Based on patch from Heiko Schlittermann.
1761 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1762 Patch from Uwe Doering, sign-off by Michael Haardt.
1764 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1765 variable declaration deep within a block. Bug and patch from
1768 PP/15 lookups-Makefile IRIX compatibility coercion.
1770 PP/16 Make DISABLE_DKIM build knob functional.
1772 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1773 Patch by Simon Arlott
1775 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1776 variable argument macros. Our copy now differs from upstream.
1782 TF/01 Failure to get a lock on a hints database can have serious
1783 consequences so log it to the panic log.
1785 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1786 controlled using the smtp_confirmation log selector.
1788 TF/03 Include the error message when we fail to unlink a spool file.
1790 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1791 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1792 for maintaining out-of-tree patches for some time.
1794 PP/01 Bugzilla 139: Documentation and portability issues.
1795 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1796 Handle per-OS dynamic-module compilation flags.
1798 PP/02 Let /dev/null have normal permissions.
1799 The 4.73 fixes were a little too stringent and complained about the
1800 permissions on /dev/null. Exempt it from some checks.
1801 Reported by Andreas M. Kirchwitz.
1803 PP/03 Report version information for many libraries, including
1804 Exim version information for dynamically loaded libraries. Created
1805 version.h, now support a version extension string for distributors
1806 who patch heavily. Dynamic module ABI change.
1808 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1809 privilege escalation vulnerability whereby the Exim run-time user
1810 can cause root to append content of the attacker's choosing to
1813 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1816 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1817 If dropping privileges for untrusted macros, we disabled normal logging
1818 on the basis that it would fail; for the Exim run-time user, this is not
1819 the case, and it resulted in successful deliveries going unlogged.
1820 Fixed. Reported by Andreas Metzler.
1826 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1827 only prepend for the Resent-* case. Fixes regression introduced in
1828 Exim 4.70 by NM/22 for Bugzilla 607.
1830 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1831 increasing numbers of administrators be bitten by this.
1833 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1835 PP/03 Bugzilla 994: added openssl_options main configuration option.
1837 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1839 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1841 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1843 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1844 without a peer certificate, leading to a segfault because of an
1845 assumption that peers always have certificates. Be a little more
1846 paranoid. Problem reported by Martin Tscholak.
1848 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1849 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1850 NB: ClamAV planning to remove STREAM in "middle of 2010".
1851 CL also introduces -bmalware, various -d+acl logging additions and
1852 more caution in buffer sizes.
1854 PP/09 Implemented reverse_ip expansion operator.
1856 PP/10 Bugzilla 937: provide a "debug" ACL control.
1858 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1860 PP/12 Bugzilla 973: Implement --version.
1862 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1864 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1866 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1868 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1869 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1870 ignore trailing whitespace.
1872 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1874 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1877 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1878 Notification from Dr Andrew Aitchison.
1880 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1881 ExtendedDetectionInfo response format.
1882 Notification from John Horne.
1884 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1887 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1888 XSL and documented dependency on system catalogs, with examples of how
1891 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1894 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1895 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1896 configuration file which is writeable by the Exim user or group.
1898 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1899 of configuration files to cover files specified with the -C option if
1900 they are going to be used with root privileges, not just the default
1903 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1904 option (effectively making it always true).
1906 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1907 files to be used while preserving root privileges.
1909 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1910 that rogue child processes cannot use them.
1912 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1913 run-time user, instead of root.
1915 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1916 Exim run-time user without dropping privileges.
1918 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1919 result string, instead of calling string_vformat() twice with the same
1922 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1923 for other users. Others should always drop root privileges if they use
1924 -C on the command line, even for a whitelisted configure file.
1926 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1928 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1934 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1935 $data_path, and $header_path variables; fixed documentation bugs and
1938 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1939 exipick to access non-standard spools, including the "frozen" queue
1942 NM/01 Bugzilla 965: Support mysql stored procedures.
1943 Patch from Alain Williams
1945 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1947 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1948 Patch from Andreas Metzler
1950 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1951 Patch from Kirill Miazine
1953 NM/05 Bugzilla 671: Added umask to procmail example.
1955 JJ/03 installed exipick 20100323.0, fixing doc bug
1957 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1958 directory. Notification and patch from Dan Rosenberg.
1960 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1962 TK/02 Improve log output when DKIM signing operation fails.
1964 MH/01 Treat the transport option dkim_domain as a colon separated
1965 list, not as a single string, and sign the message with each element,
1966 omitting multiple occurences of the same signer.
1968 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1969 Bugzilla 985, 986. Patch by Simon Arlott
1971 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1972 Patch by Simon Arlott
1974 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1975 MBX locking. Notification from Dan Rosenberg.
1981 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1983 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1985 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1987 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1989 NM/04 Bugzilla 915: Flag broken perl installation during build.
1995 TK/01 Added patch by Johannes Berg that expands the main option
1996 "spamd_address" if it starts with a dollar sign.
1998 TK/02 Write list of recipients to X-Envelope-Sender header when building
1999 the mbox-format spool file for content scanning (suggested by Jakob
2002 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2003 (http://www.dcc-servers.net/) support via dccifd. Activated by
2004 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2006 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2007 by Mark Daniel Reidel <mr@df.eu>.
2009 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2010 When building exim an external PCRE library is now needed -
2011 PCRE is a system library on the majority of modern systems.
2012 See entry on PCRE_LIBS in EDITME file.
2014 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2015 conversation. Added nologin parameter to request.
2016 Patch contributed by Kirill Miazine.
2018 TF/01 Do not log submission mode rewrites if they do not change the address.
2020 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2022 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2023 log files in place. Contributed by Roberto Lima.
2025 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2027 TF/03 Bugzilla 615: When checking the local_parts router precondition
2028 after a local_part_suffix or local_part_prefix option, Exim now
2029 does not use the address's named list lookup cache, since this
2030 contains cached lookups for the whole local part.
2032 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2033 Robert Millan. Documentation is in experimental-spec.txt.
2035 TF/04 Bugzilla 668: Fix parallel build (make -j).
2037 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2039 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2040 Patch provided by Jan Srzednicki.
2042 TF/05 Leading white space used to be stripped from $spam_report which
2043 wrecked the formatting. Now it is preserved.
2045 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2046 that they are available at delivery time.
2048 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2050 TF/08 TLS error reporting now respects the incoming_interface and
2051 incoming_port log selectors.
2053 TF/09 Produce a more useful error message if an SMTP transport's hosts
2054 setting expands to an empty string.
2056 NM/06 Bugzilla 744: EXPN did not work under TLS.
2057 Patch provided by Phil Pennock.
2059 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2060 Patch provided by Richard Godbee.
2062 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2063 acl_smtp_notquit, added index entry.
2065 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2066 Patch provided by Eugene Bujak.
2068 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2069 accept(). Patch provided by Maxim Dounin.
2071 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2072 Patch provided by Phil Pennock.
2074 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2076 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2077 Patch provided by Brad "anomie" Jorsch.
2079 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2080 Patch provided by Dean Brooks.
2082 TK/05 Add native DKIM support (does not depend on external libraries).
2084 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2085 Patch provided by Graeme Fowler.
2087 NM/16 Bugzilla 851: Documentation example syntax fix.
2089 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2091 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2094 NM/19 Bugzilla 745: TLS version reporting.
2095 Patch provided by Phil Pennock.
2097 NM/20 Bugzilla 167: bool: condition support.
2098 Patch provided by Phil Pennock.
2100 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2101 clients. Patch provided by Phil Pennock.
2103 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2104 Patch provided by Brad "anomie" Jorsch.
2106 NM/23 Bugzilla 687: Fix misparses in eximstats.
2107 Patch provided by Heiko Schlittermann.
2109 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2110 Patch provided by Heiko Schlittermann.
2112 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2113 plus update to original patch.
2115 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2117 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2118 Patch provided by David Brownlee.
2120 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2122 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2124 NM/30 Bugzilla 888: TLS documentation bugfixes.
2126 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2128 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2129 Unlike the original bugzilla I have changed all shell scripts in src tree.
2131 NM/33 Bugzilla 898: Transport filter timeout fix.
2132 Patch by Todd Rinaldo.
2134 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2135 Patch by Serge Demonchaux.
2137 NM/35 Bugzilla 39: Base64 decode bug fixes.
2138 Patch by Jakob Hirsch.
2140 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2142 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2144 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2146 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2152 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2153 ALT-N's libdkim that I have put here:
2154 http://duncanthrax.net/exim-experimental/
2156 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2157 were called 'true' and it seems that C99 defines that as a reserved
2158 keyword to be used with 'bool' variable types. That means you could
2159 not include C99-style headers which use bools without triggering
2160 build errors in sieve.c.
2162 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2163 as mailq or other aliases. Changed the --help handling significantly
2164 to do whats expected. exim_usage() emits usage/help information.
2166 SC/01 Added the -bylocaldomain option to eximstats.
2168 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2170 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2172 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2178 PH/01 Another patch from the Sieve maintainer.
2180 PH/02 When an IPv6 address is converted to a string for single-key lookup
2181 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2182 dots are used instead of colons so that keys in lsearch files need not
2183 contain colons. This was done some time before quoting was made available
2184 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2185 (notated using the quote facility) so as to distinguish them from IPv4
2186 keys. This meant that lookups for IP addresses in host lists did not work
2187 for iplsearch lookups.
2189 This has been fixed by arranging for IPv6 addresses to be expressed with
2190 colons if the lookup type is iplsearch. This is not incompatible, because
2191 previously such lookups could never work.
2193 The situation is now rather anomalous, since one *can* have colons in
2194 ordinary lsearch keys. However, making the change in all cases is
2195 incompatible and would probably break a number of configurations.
2197 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2200 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2201 conversion specification without a maximum field width, thereby enabling
2202 a rogue spamd server to cause a buffer overflow. While nobody in their
2203 right mind would setup Exim to query an untrusted spamd server, an
2204 attacker that gains access to a server running spamd could potentially
2205 exploit this vulnerability to run arbitrary code as the Exim user.
2207 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2208 $primary_hostname instead of what libspf2 thinks the hosts name is.
2210 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2211 a directory entry by the name of the lookup key. Previously, if a
2212 symlink pointed to a non-existing file or a file in a directory that
2213 Exim lacked permissions to read, a lookup for a key matching that
2214 symlink would fail. Now it is enough that a matching directory entry
2215 exists, symlink or not. (Bugzilla 503.)
2217 PH/03 The body_linecount and body_zerocount variables are now exported in the
2220 PH/04 Added the $dnslist_matched variable.
2222 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2223 This means they are set thereafter only if the connection becomes
2226 PH/06 Added the client_condition to authenticators so that some can be skipped
2227 by clients under certain conditions.
2229 PH/07 The error message for a badly-placed control=no_multiline_responses left
2230 "_responses" off the end of the name.
2232 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2234 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2235 (without spaces) instead of just copying the configuration text.
2237 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2239 PH/11 Added $max_received_linelength.
2241 PH/12 Added +ignore_defer and +include_defer to host lists.
2243 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2244 way in which PCRE > 7.0 is built.
2246 PH/14 Implemented queue_only_load_latch.
2248 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2249 MAIL command. The effect was to mangle the value on 64-bit systems.
2251 PH/16 Another patch from the Sieve maintainer.
2253 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2255 PH/18 If a system quota error occurred while trying to create the file for
2256 a maildir delivery, the message "Mailbox is full" was not appended to the
2257 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2258 only to a quota excession during the actual writing of the file.
2260 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2261 characters?) which causes problems in log lines. The DN values are now
2262 passed through string_printing() before being added to log lines.
2264 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2265 and InterBase are left for another time.)
2267 PH/21 Added message_body_newlines option.
2269 PH/22 Guard against possible overflow in moan_check_errorcopy().
2271 PH/23 POSIX allows open() to be a macro; guard against that.
2273 PH/24 If the recipient of an error message contained an @ in the local part
2274 (suitably quoted, of course), incorrect values were put in $domain and
2275 $local_part during the evaluation of errors_copy.
2281 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2282 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2285 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2286 issue a MAIL command.
2288 PH/02 In an ACL statement such as
2290 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2292 if a client was not listed at all, or was listed with a value other than
2293 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2294 the condition was not true (as it should be), so access was not denied.
2295 The bug was that the ! inversion was incorrectly passed on to the second
2296 item. This has been fixed.
2298 PH/03 Added additional dnslists conditions == and =& which are different from
2299 = and & when the dns lookup returns more than one IP address.
2301 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2302 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2304 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2305 FSYNC, which compiles an option called disable_fsync that allows for
2306 bypassing fsync(). The documentation is heavily laced with warnings.
2308 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2310 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2311 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2312 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2313 including adding "make clean"; (3) Added -fPIC when compiling the test
2314 dynamically loaded module, to get rid of a warning.
2316 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2317 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2318 = 0s. The bug is otherwise harmless.
2320 PH/07 There was a bug in the dovecot authenticator such that the value of
2321 $auth1 could be overwritten, and so not correctly preserved, after a
2322 successful authentication. This usually meant that the value preserved by
2323 the server_setid option was incorrect.
2325 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2327 PH/09 Installed PCRE release 7.0.
2329 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2330 run for batched SMTP input. It is now run at the start of every message
2331 in the batch. While fixing this I discovered that the process information
2332 (output by running exiwhat) was not always getting set for -bs and -bS
2333 input. This is fixed, and it now also says "batched" for BSMTP.
2335 PH/11 Added control=no_pipelining.
2337 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2338 patch, slightly modified), and move the expansion of helo_data till after
2339 the connection is made in the smtp transport (so it can use these
2342 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2344 PH/14 Added log_selector = +pid.
2346 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2348 PH/16 Add ${if forany and ${if forall.
2350 PH/17 Added dsn_from option to vary the From: line in DSNs.
2352 PH/18 Flush SMTP output before performing a callout, unless control =
2353 no_callout_flush is set.
2355 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2356 was true (the default) a successful delivery failed to delete the retry
2357 item, thus causing premature timeout of the address. The bug is now
2360 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2362 PH/21 Long custom messages for fakedefer and fakereject are now split up
2363 into multiline responses in the same way that messages for "deny" and
2364 other ACL rejections are.
2366 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2367 with slight modification.
2369 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2370 draft, changing the syntax and factoring some duplicate code".
2372 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2373 for deliveries of the second and subsequent messages over the same SMTP
2376 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2377 ${reduce, with only minor "tidies".
2379 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2381 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2382 expansion side effects.
2384 PH/27 When a message times out after an over-quota error from an Exim-imposed
2385 quota, the bounce message says "mailbox is full". This message was not
2386 being given when it was a system quota that was exceeded. It now should
2389 MH/03 Made $recipients available in local_scan(). local_scan() already has
2390 better access to the recipient list through recipients_list[], but
2391 $recipients can be useful in postmaster-provided expansion strings.
2393 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2394 in the case of a MAIL command with additional options following the
2395 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2396 were accidentally chopped off.
2398 PH/29 SMTP synchronization checks are implemented when a command is read -
2399 there is a check that no more input is waiting when there shouldn't be
2400 any. However, for some commands, a delay in an ACL can mean that it is
2401 some time before the response is written. In this time, more input might
2402 arrive, invalidly. So now there are extra checks after an ACL has run for
2403 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2404 pipelining has not been advertised.
2406 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2408 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2409 returned in the SMTP response when smtp_return_error_details was false.
2410 This has been fixed.
2412 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2413 instead of the C I/O that was originally supplied, because problems were
2414 reported on Solaris.
2416 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2417 Exim which did not show up earlier: it was assuming that a call to
2418 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2419 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2420 was a macro that became an assignment, so it seemed to work. This has
2421 changed to a proper function call with a void return, hence the compile
2422 error. Exim's code has been fixed.
2424 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2427 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2429 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2431 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2432 because it turns out to be too big or there is a timeout) there is no
2433 "Completed" line in the log. When some messages of this type were
2434 selected by exigrep, they were listed as "not completed". Others were
2435 picked up by some special patterns. I have improved the selection
2436 criteria to be more general.
2438 PH/38 The host_find_failed option in the manualroute router can now be set
2439 to "ignore", to completely ignore a host whose IP address cannot be
2440 found. If all hosts are ignored, the behaviour is controlled by the new
2441 host_all_ignored option.
2443 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2444 homing or because of multiple MX records with /mx) generated more than
2445 one IP address, and the following item turned out to be the local host,
2446 all the secondary addresses of the first item were incorrectly removed
2447 from the list, along with the local host and any following hosts (which
2448 is what is supposed to happen).
2450 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2451 whoever called Exim into the -H file. In the case of the daemon it was
2452 behaving confusingly. When first started, it used values for whoever
2453 started the daemon, but after a SIGHUP it used the Exim user (because it
2454 calls itself on a restart). I have changed the code so that it now always
2457 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2458 message are rejected with the same error (e.g. no authentication or bad
2459 sender address), and a DATA command is nevertheless sent (as can happen
2460 with PIPELINING or a stupid MUA), the error message that was given to the
2461 RCPT commands is included in the rejection of the DATA command. This is
2462 intended to be helpful for MUAs that show only the final error to their
2465 PH/42 Another patch from the Sieve maintainer.
2467 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2468 Eximstats - Fixed some broken HTML links and added missing column headers
2470 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2471 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2473 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2475 PH/43 Yet another patch from the Sieve maintainer.
2477 PH/44 I found a way to check for a TCP/IP connection going away before sending
2478 the response to the final '.' that terminates a message, but only in the
2479 case where the client has not sent further data following the '.'
2480 (unfortunately, this is allowed). However, in many cases there won't be
2481 any further data because there won't be any more messages to send. A call
2482 to select() can be used: if it shows that the input is "ready", there is
2483 either input waiting, or the socket has been closed. An attempt to read
2484 the next input character can distinguish the two cases. Previously, Exim
2485 would have sent an OK response which the client would never have see.
2486 This could lead to message repetition. This fix should cure that, at
2487 least in a lot of common cases.
2489 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2490 advertised in response to EHLO.
2496 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2497 fixed by 4.65/MH/01 (is this a record?) are fixed:
2499 (i) An empty string was always treated as zero by the numeric comparison
2500 operators. This behaviour has been restored.
2502 (ii) It is documented that the numeric comparison operators always treat
2503 their arguments as decimal numbers. This was broken in that numbers
2504 starting with 0 were being interpreted as octal.
2506 While fixing these problems I realized that there was another issue that
2507 hadn't been noticed. Values of message_size_limit (both the global option
2508 and the transport option) were treated as octal if they started with 0.
2509 The documentation was vague. These values are now always treated as
2510 decimal, and I will make that clear in the documentation.
2516 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2517 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2520 MH/01 Don't check that the operands of numeric comparison operators are
2521 integers when their expansion is in "skipping" mode (fixes bug
2522 introduced by 4.64-PH/07).
2524 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2525 child addresses, Exim now panics and dies. Previously, because the count
2526 is held in a short int, deliveries were likely to be lost. As such a
2527 large number of recipients for a single message is ridiculous
2528 (performance will be very, very poor), I have chosen to impose a limit
2529 rather than extend the field.
2535 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2536 leftover -K file (the existence of which was triggered by #402).
2537 While we were at it, introduced process PID as part of the -K
2538 filename. This should rule out race conditions when creating
2541 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2542 processing considerably. Previous code took too long for large mails,
2543 triggering a timeout which in turn triggers #401.
2545 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2546 in the DK code in transports.c. sendfile() is not really portable,
2547 hence the _LINUX specificness.
2549 TF/01 In the add_headers option to the mail command in an Exim filter,
2550 there was a bug that Exim would claim a syntax error in any
2551 header after the first one which had an odd number of characters
2554 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2555 callout verification, Exim cached a "reject" for the entire domain. This
2556 is correct for most verifications, but it is not correct for a recipient
2557 verification with use_sender or use_postmaster set, because in that case
2558 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2559 case of MAIL FROM:<> rejection from other early rejections (e.g.
2560 rejection of HELO). When verifying a recipient using a non-null MAIL
2561 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2562 Whatever the result of the callout, the value of the domain cache is
2563 left unchanged (for any other kind of callout, getting as far as trying
2564 RCPT means that the domain itself is ok).
2566 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2569 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2570 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2571 session, the daemon ignores ECONNECT errors and logs others; it now
2572 ignores EPIPE as well.
2574 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2575 (quoted-printable decoding).
2577 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2578 later the small subsequent patch to fix an introduced bug.
2580 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2582 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2584 PH/08 An error is now given if message_size_limit is specified negative.
2586 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2587 to be given (somewhat) arbitrary names.
2589 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2592 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2593 miscellaneous code fixes
2595 PH/10 Added the log_reject_target ACL modifier to specify where to log
2598 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2599 hostname. This is wrong, because it relates to the incoming message (and
2600 probably the interface on which it is arriving) and not to the outgoing
2601 callout (which could be using a different interface). This has been
2602 changed to use the value of the helo_data option from the smtp transport
2603 instead - this is what is used when a message is actually being sent. If
2604 there is no remote transport (possible with a router that sets up host
2605 addresses), $smtp_active_hostname is used.
2607 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2608 tweaks were necessary in order to get it to work (see also 21 below):
2609 (a) The code assumed that strncpy() returns a negative number on buffer
2610 overflow, which isn't the case. Replaced with Exim's string_format()
2612 (b) There were several signed/unsigned issues. I just did the minimum
2613 hacking in of casts. There is scope for a larger refactoring.
2614 (c) The code used strcasecmp() which is not a standard C function.
2615 Replaced with Exim's strcmpic() function.
2616 (d) The code set only $1; it now sets $auth1 as well.
2617 (e) A simple test gave the error "authentication client didn't specify
2618 service in request". It would seem that Dovecot has changed its
2619 interface. Fortunately there's a specification; I followed it and
2620 changed what the client sends and it appears to be working now.
2622 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2625 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2626 address A is aliased to B and C, where B exists and C does not. Without
2627 -v the output is "A verified" because verification stops after a
2628 successful redirection if more than one address is generated. However,
2629 with -v the child addresses are also verified. Exim was outputting "A
2630 failed to verify" and then showing the successful verification for C,
2631 with its parentage. It now outputs "B failed to verify", showing B's
2632 parentage before showing the successful verification of C.
2634 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2635 look up a TXT record in a specific list after matching in a combined
2638 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2639 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2640 they consult the DNS. I had assumed they would set it the way they
2641 wanted; and indeed my experiments on Linux seem to show that in some
2642 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2643 To be on the safe side, however, I have now made the interface to
2644 host_find_byname() similar to host_find_bydns(), with an argument
2645 containing the DNS resolver options. The host_find_byname() function now
2646 sets these options at its start, just as host_find_bydns() does. The smtp
2647 transport options dns_qualify_single and dns_search_parents are passed to
2648 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2649 of host_find_byname() use the default settings of RES_DEFNAMES
2650 (qualify_single) but not RES_DNSRCH (search_parents).
2652 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2653 spool_read_header() do less string testing, by means of a preliminary
2654 switch on the second character of optional "-foo" lines. (This is
2655 overdue, caused by the large number of possibilities that now exist.
2656 Originally there were few.) While I was there, I also converted the
2657 str(n)cmp tests so they don't re-test the leading "-" and the first
2658 character, in the hope this might squeeze out yet more improvement.
2660 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2661 flag allowing group syntax was set by the header_syntax check but not
2662 turned off, possible causing trouble later; (2) The flag was not being
2663 set at all for the header_verify test, causing "group"-style headers to
2664 be rejected. I have now set it in this case, and also caused header_
2665 verify to ignore an empty address taken from a group. While doing this, I
2666 came across some other cases where the code for allowing group syntax
2667 while scanning a header line wasn't quite right (mostly, not resetting
2668 the flag correctly in the right place). These bugs could have caused
2669 trouble for malformed header lines. I hope it is now all correct.
2671 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2672 with the "reply" argument non-NULL. The code, however (which originally
2673 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2674 but it didn't always do it. This confused somebody who was copying the
2675 code for some other use. I have removed all the tests.
2677 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2678 feature that was used to support insecure browsers during the U.S. crypto
2679 embargo. It requires special client support, and Exim is probably the
2680 only MTA that supported it -- and would never use it because real RSA is
2681 always available. This code has been removed, because it had the bad
2682 effect of slowing Exim down by computing (never used) parameters for the
2683 RSA_EXPORT functionality.
2685 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2686 authenticator to fail if there's a tab character in the incoming data
2687 (there should never be unless someone is messing about, as it's supposed
2688 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2689 if the connection is using TLS or if the remote IP is the same as the
2690 local IP, and the "valid-client-cert option" if a client certificate has
2693 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2694 authenticators. This can be used for authorization after authentication
2695 succeeds. (In the case of plaintext, it servers for both authentication
2698 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2699 if any retry times were supplied.
2701 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2702 connection, where there is no client IP address to check. In this
2703 situation, the verify now always succeeds.
2705 PH/25 Applied John Jetmore's -Mset patch.
2707 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2709 PH/27 In a string expansion for a processed (not raw) header when multiple
2710 headers of the same name were present, leading whitespace was being
2711 removed from all of them, but trailing whitespace was being removed only
2712 from the last one. Now trailing whitespace is removed from each header
2713 before concatenation. Completely empty headers in a concatenation (as
2714 before) are ignored.
2716 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2717 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2719 PH/29 [Removed. This was a change that I later backed out, and forgot to
2720 correct the ChangeLog entry (that I had efficiently created) before
2721 committing the later change.]
2723 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2724 address errors (4xx response to RCPT) over the same connection as other
2725 messages routed to the same hosts. Such deliveries are always "forced",
2726 so retry times are not inspected. This resulted in far too many retries
2727 for the affected addresses. The effect occurred only when there were more
2728 hosts than the hosts_max_try setting in the smtp transport when it had
2729 the 4xx errors. Those hosts that it had tried were not added to the list
2730 of hosts for which the message was waiting, so if all were tried, there
2731 was no problem. Two fixes have been applied:
2733 (i) If there are any address or message errors in an SMTP delivery, none
2734 of the hosts (tried or untried) are now added to the list of hosts
2735 for which the message is waiting, so the message should not be a
2736 candidate for sending over the same connection that was used for a
2737 successful delivery of some other message. This seems entirely
2738 reasonable: after all the message is NOT "waiting for some host".
2739 This is so "obvious" that I'm not sure why it wasn't done
2740 previously. Hope I haven't missed anything, but it can't do any
2741 harm, as the worst effect is to miss an optimization.
2743 (ii) If, despite (i), such a delivery is accidentally attempted, the
2744 routing retry time is respected, so at least it doesn't keep
2745 hammering the server.
2747 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2748 in ${readsocket because some servers need this prod.
2750 PH/32 Added some extra debug output when updating a wait-xxx database.
2752 PH/33 The hint "could be header name not terminated by colon", which has been
2753 given for certain expansion errors for a long time, was not being given
2754 for the ${if def:h_colon_omitted{... case.
2756 PH/34 The spec says: "With one important exception, whenever a domain list is
2757 being scanned, $domain contains the subject domain." There was at least
2758 one case where this was not true.
2760 PH/35 The error "getsockname() failed: connection reset by peer" was being
2761 written to the panic log as well as the main log, but it isn't really
2762 panic-worthy as it just means the connection died rather early on. I have
2763 removed the panic log writing for the ECONNRESET error when getsockname()
2766 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2767 runs only) independently of the message's sender address. This meant
2768 that, if the 4xx error was in fact related to the sender, a different
2769 message to the same recipient with a different sender could confuse
2770 things. In particular, this can happen when sending to a greylisting
2771 server, but other circumstances could also provoke similar problems.
2772 I have changed the default so that the retry time for these errors is now
2773 based a combination of the sender and recipient addresses. This change
2774 can be overridden by setting address_retry_include_sender=false in the
2777 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2778 remote server are returned as part of bounce messages. This was not
2779 happening for LMTP over a pipe (the lmtp transport), but now it is the
2780 same for both kinds of LMTP.
2782 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2783 in header lines that were in fact CNAMEs. This is no longer the case.
2785 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2786 and queue runs started by the daemon processed all messages. This has
2787 been fixed so that -R and -S can now usefully be given with -q<time>.
2789 PH/40 Import PCRE release 6.7 (fixes some bugs).
2791 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2793 PH/42 Give an error if -q is specified more than once.
2795 PH/43 Renamed the variables $interface_address and $interface_port as
2796 $received_ip_address and $received_port, to make it clear that these
2797 values apply to message reception, and not to the outgoing interface when
2798 a message is delivered. (The old names remain recognized, of course.)
2800 PH/44 There was no timeout on the connect() call when using a Unix domain
2801 socket in the ${readsocket expansion. There now is.
2803 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2804 be meaningful with "accept".
2806 SC/01 Eximstats V1.43
2807 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2809 SC/02 Eximstats V1.44
2810 Use a glob alias rather than an array ref in the generated
2811 parser. This improves both readability and performance.
2813 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2814 Collect SpamAssassin and rejection statistics.
2815 Don't display local sender or destination tables unless
2816 there is data to show.
2817 Added average volumes into the top table text output.
2819 SC/04 Eximstats V1.46
2820 Collect data on the number of addresses (recipients)
2821 as well as the number of messages.
2823 SC/05 Eximstats V1.47
2824 Added 'Message too big' to the list of mail rejection
2825 reasons (thanks to Marco Gaiarin).
2827 SC/06 Eximstats V1.48
2828 Mainlog lines which have GMT offsets and are too short to
2829 have a flag are now skipped.
2831 SC/07 Eximstats V1.49 (Alain Williams)
2832 Added the -emptyok flag.
2834 SC/08 Eximstats V1.50
2835 Fixes for obtaining the IP address from reject messages.
2837 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2838 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2839 whitespace changes from 4.64-PH/27
2841 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2844 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2845 are found, allow negative numbers in numeric criteria)
2847 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2849 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2850 to match changes made in 4.64-PH/43
2852 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2854 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2855 "domain" is actually an IP address, and force a failure. This locks out
2856 those revolvers/nameservers that support "A-for-A" lookups, in
2857 contravention of the specifications.
2859 PH/48 When a host name was looked up from an IP address, and the subsequent
2860 forward lookup of the name timed out, the host name was left in
2861 $sender_host_name, contrary to the specification.
2863 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2864 restricted to single-key lookups, Exim was not diagnosing an error if
2865 * or *@ was used with a query-style lookup.
2867 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2869 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2870 long ago, but noone interested enough thought of it. Let's just say that
2871 the "1.1" means that there are some new functions that weren't there at
2872 some point in the past.
2874 PH/51 Error processing for expansion failure of helo_data from an smtp
2875 transport during callout processing was broken.
2877 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2878 tested/used via the -bh/-bhc/-bs options.
2880 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2881 bug, fixed in subsequent PCRE releases).
2883 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2884 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2886 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2892 SC/01 Use a glob alias rather than an array ref in eximstats generated
2893 parser. This improves both readability and performance.
2895 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2896 Don't display local sender or destination tables in eximstats unless
2897 there is data to show.
2898 Added average volumes into the eximstats top table text output.
2900 SC/03 Collect data on the number of addresses (recipients) as well
2901 as the number of messages in eximstats.
2903 TF/01 Correct an error in the documentation for the redirect router. Exim
2904 does (usually) call initgroups() when daemonizing.
2906 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2907 with consistent privilege compared to when running as a daemon.
2909 TF/03 Note in the spec that $authenticated_id is not set for local
2910 submissions from trusted users.
2912 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2913 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2915 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2916 by adding some example configuration directives to the default
2917 configuration file. A little bit of work is required to uncomment the
2918 directives and define how usernames and passwords are checked, but
2919 there is now a framework to start from.
2921 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2922 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2923 without this. I don't know how relevant this is to other LDAP libraries.
2925 PH/02 Add the verb name to the "unknown ACL verb" error.
2927 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2929 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2931 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2932 directory not expanded when it should be if an expanded home directory
2933 was set for the address (which is overridden by the transport).
2935 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2938 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2939 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2940 because it is too late at that time, and has no effect.
2942 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2943 security issue with \' (bugzilla #107). I could not use the
2944 PQescapeStringConn() function, because it needs a PGconn value as one of
2947 PH/08 When testing addresses using -bt, indicate those final addresses that
2948 are duplicates that would not cause an additional delivery. At least one
2949 person was confused, thinking that -bt output corresponded to deliveries.
2950 (Suppressing duplicates isn't a good idea as you lose the information
2951 about possibly different redirections that led to the duplicates.)
2953 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2954 systems where poll() doesn't work, in particular OS X.
2956 PH/10 Added more information to debugging output for retry time not reached.
2958 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2959 operations in malware.c.
2961 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2964 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2965 syslog_duplication set false, log lines that would normally be written
2966 both the the main log and to the reject log were not written to syslog at
2969 PH/14 In the default configuration, change the use of "message" in ACL warn
2970 statements to "add_header".
2972 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2973 not followed by a command (e.g. "seen endif").
2975 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2976 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2979 PH/17 Added extra conditions to the default value of delay_warning_condition
2983 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2984 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2985 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2988 The Auto-Submitted: and various List- headers are standardised, whereas I
2989 don't think Precedence: ever was.
2991 PH/18 Refactored debugging code in route_finduser() to show more information,
2992 in particular, the error code if getpwnam() issues one.
2994 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2995 This is apparently needed in addition to the PH/07 change above to avoid
2996 any possible encoding problems.
2998 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2999 but not after initializing Perl.
3001 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3002 output them only if debugging. By default they are written stderr,
3003 apparently, which is not desirable.
3005 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3008 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3011 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3013 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3014 authenticated or an ident call has been made. Suppress the default
3015 values for $authenticated_id and $authenticated_sender (but permit -oMai
3016 and -oMas) when testing with -bh.
3018 PH/24 Re-jigged the order of the tests in the default configuration so that the
3019 tests for valid domains and recipients precede the DNS black list and CSA
3020 tests, on the grounds that those ones are more expensive.
3022 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3023 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3024 This bug exists in every version of Exim that I still have, right back to
3027 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3028 However, an attempt to turn on case-sensitivity in a regex key by
3029 including (?-i) didn't work because the subject string was already
3030 lowercased, and the effects were non-intuitive. It turns out that a
3031 one-line patch can be used to allow (?-i) to work as expected.
3037 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3038 other effects) broke the use of negated acl sub-conditions.
3040 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3043 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3044 "Deny" causes Exim to reject the incoming connection with a 554 error.
3045 Unfortunately, if there is a major crisis, such as a disk failure,
3046 tcp-wrappers gives "deny", whereas what one would like would be some
3047 kind of temporary error. A kludge has been added to help with this.
3048 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3049 554 error is used if errno is still zero or contains ENOENT (which occurs
3050 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3053 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3055 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3056 errors. Otherwise a message that provokes a temporary error (when other
3057 messages do not) can cause a whole host to time out.
3059 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3060 addresses were routed directly to files or pipes from a redirect router.
3061 File deliveries just didn't batch; pipe deliveries might have suffered
3064 PH/06 A failure to get a lock for a hints database would erroneously always say
3065 "Failed to get write lock", even when it was really a read lock.
3067 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3068 of 0600. This has been changed to use the value of the lockfile_mode
3069 option (which defaults to 0600).
3071 PH/08 Applied small patch from the Sieve maintainer.
3073 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3074 folder from quota calculations, a direct delivery into this folder messed
3075 up the contents of the maildirsize file. This was because the regex was
3076 used only to exclude .Trash (or whatever) when the size of the mailbox
3077 was calculated. There was no check that a delivery was happening into an
3078 excluded directory. This bug has been fixed by ignoring all quota
3079 processing for deliveries into excluded directories.
3081 PH/10 Added the maildirfolder_create_regex option to appendfile.
3087 PH/01 The code for finding all the local interface addresses on a FreeBSD
3088 system running IPv6 was broken. This may well have applied to all BSD
3089 systems, as well as to others that have similar system calls. The broken
3090 code found IPv4 interfaces correctly, but gave incorrect values for the
3091 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3092 that it would not match correctly against @[] and not recognize the IPv6
3095 PH/02 The ipliteral router was not recognizing addresses of the form user@
3096 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3098 PH/03 Added disable_ipv6.
3100 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3101 decoded form, because it is most often used to construct To: headers
3102 lines in autoreplies, and the decoded form may well be syntactically
3103 invalid. However, $reply_address has leading white space removed, and all
3104 newlines turned into spaces so that the autoreply transport does not
3107 PH/05 If group was specified without a user on a router, and no group or user
3108 was specified on a transport, the group from the router was ignored.
3110 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3111 for visible compile-time settings that can be used to change these
3112 numbers, for those that want even more. Backwards compatibility with old
3113 spool files has been maintained. However, going back to a previous Exim
3114 release will lost any variables that are in spool files.
3116 PH/07 Two small changes when running in the test harness: increase delay when
3117 passing a TCP/IP connection to a new process, in case the original
3118 process has to generate a bounce, and remove special handling of
3119 127.0.0.2 (sic), which is no longer necessary.
3121 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3122 be the same on different OS.
3124 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3127 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3128 whether --show-vars was specified or not
3130 JJ/02 exipick: Added support for new ACL variable spool format introduced
3133 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3134 syntactically invalid From: or Reply-to: line, and a filter used this to
3135 generate an autoreply, and therefore failed to obtain an address for the
3136 autoreply, Exim could try to deliver to a non-existent relative file
3137 name, causing unrelated and misleading errors. What now happens is that
3138 it logs this as a hard delivery error, but does not attempt to create a
3141 PH/11 The exinext utility has a -C option for testing purposes, but although
3142 the given file was scanned by exinext itself; it wasn't being passed on
3143 when Exim was called.
3145 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3146 an end-of-file indication when reading a command response.
3148 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3149 compiled. In many other places in Exim, IPv6 addresses are always
3150 recognized, so I have changed this. It also means that IPv4 domain
3151 literals of the form [IPV4:n.n.n.n] are now always recognized.
3153 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3154 used if the router is not running as root, for example, when verifying at
3155 ACL time, or when using -bh. The debugging output from this situation was
3156 non-existent - all you got was a failure to exec. I have made two
3159 (a) Failures to set uid/gid, the current directory, or a process leader
3160 in a subprocess such as that created by queryprogram now generate
3161 suitable debugging output when -d is set.
3163 (b) The queryprogram router detects when it is not running as root,
3164 outputs suitable debugging information if -d is set, and then runs
3165 the subprocess without attempting to change uid/gid.
3167 PH/15 Minor change to Makefile for building test_host (undocumented testing
3170 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3171 additional section of a DNS packet that returns MX or SRV records.
3172 Instead, it always explicitly searches for A/AAAA records. This avoids
3173 major problems that occur when a DNS server includes only records of one
3174 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3175 fixed another bug: if SRV records were looked up and the corresponding
3176 address records were *not* found in the additional section, the port
3177 values from the SRV records were lost.
3179 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3180 using the correct key (the original address) when searching the retry
3181 rules in order to find which one to use for generating the retry hint.
3183 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3184 adding the default one. Similarly, if it contains a Reply-To: header, the
3185 errors_reply_to option, if set, is not used.
3187 PH/19 When calculating a retry time, Exim used to measure the "time since
3188 failure" by looking at the "first failed" field in the retry record. Now
3189 it does not use this if it is later than than the arrival time of the
3190 message. Instead it uses the arrival time. This makes for better
3191 behaviour in cases where some deliveries succeed, thus re-setting the
3192 "first failed" field. An example is a quota failure for a huge message
3193 when small messages continue to be delivered. Without this change, the
3194 "time since failure" will always be short, possible causing more frequent
3195 delivery attempts for the huge message than are intended.
3196 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3198 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3199 $1, $2, $3) because the numerical variables can be reset during some
3200 expansion items (e.g. "match"), thereby losing the authentication data.
3202 PH/21 Make -bV show the size of off_t variables so that the test suite can
3203 decide whether to run tests for quotas > 2G.
3205 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3206 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3207 filecount value is greater than 2G or if a quota value is greater than 2G
3208 on a system where the size of off_t is not greater than 4, a panic error
3211 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3212 never match. The debug and -bh output now contains an explicit error
3213 message indicating a malformed IPv4 address or mask.
3215 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3216 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3217 PH/23 above applies.
3219 PH/25 Do not write to syslog when running in the test harness. The only
3220 occasion when this arises is a failure to open the main or panic logs
3221 (for which there is an explicit test).
3223 PH/26 Added the /no_tell option to "control=freeze".
3225 PH/27 If a host name lookup failed very early in a connection, for example, if
3226 the IP address matched host_lookup and the reverse lookup yielded a name
3227 that did not have a forward lookup, an error message of the form "no IP
3228 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3229 could be logged. Now it outputs the IP address instead of "NULL".
3231 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3232 allows the sender and the authenticated sender to be set when
3233 submitting a message from within Exim. Since child_open_exim() is
3234 documented for local_scan(), the new function should be too.
3236 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3237 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3238 results in an empty string is now treated as unset.
3240 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3242 PH/31 Added sender_verify_fail logging option.
3244 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3245 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3246 tidied the source and removed it altogether.
3248 PH/33 When a queue run was abandoned because the load average was too high, a
3249 log line was always written; now it is written only if the queue_run log
3250 selector is set. In addition, the log line for abandonment now contains
3251 information about the queue run such as the pid. This is always present
3252 in "start" and "stop" lines but was omitted from the "abandon" line.
3254 PH/34 Omit spaces between a header name and the colon in the error message that
3255 is given when verify = headers_syntax fails (if there are lots of them,
3256 the message gets confusing).
3258 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3259 names, as there are now some PTR records that contain slashes. This check
3260 is only to protect against broken name servers that fall over on strange
3261 characters, so the fact that it applies to all lookups doesn't matter.
3263 PH/36 Now that the new test suite is complete, we can remove some of the
3264 special code in Exim that was needed for the old test suite. For example,
3265 sorting DNS records because real resolvers return them in an arbitrary
3266 order. The new test suite's fake resolver always returns records in the
3269 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3270 bounces) except when queue_only is set, to avoid logging races between
3271 the different processes.
3273 PH/38 Panic-die if .include specifies a non-absolute path.
3275 PH/39 A tweak to the "H" retry rule from its user.
3277 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3278 a label. They prevented compilation on older perls.
3280 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3281 a warning to be raised on newish perls.
3283 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3284 on queue. Changes to match documented behaviour of showing count of
3285 messages matching specified criteria.
3287 PH/40 Changed the default ident timeout from 30s to 5s.
3289 PH/41 Added support for the use of login_cap features, on those BSD systems
3290 that have them, for controlling the resources used by pipe deliveries.
3292 PH/42 The content-scanning code uses fopen() to create files in which to put
3293 message data. Previously it was not paying any attention to the mode of
3294 the files. Exim runs with umask(0) because the rest of the code creates
3295 files with open(), and sets the required mode explicitly. Thus, these
3296 files were ending up world-writeable. This was not a big issue, because,
3297 being within the spool directory, they were not world-accessible. I have
3298 created a function called modefopen, which takes an additional mode
3299 argument. It sets umask(777), creates the file, chmods it to the required
3300 mode, then resets the umask. All the relevant calls to fopen() in the
3301 content scanning code have been changed to use this function.
3303 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3304 to 24 hours. This avoids potential overflow problems when processing G
3305 and H retry rules. I suspect nobody ever tinkers with this value.
3307 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3309 PH/45 When the plaintext authenticator is running as a client, the server's
3310 challenges are checked to ensure they are valid base64 strings. By
3311 default, the authentication attempt is cancelled if an invalid string is
3312 received. Setting client_ignore_invalid_base64 true ignores these errors.
3313 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3314 they are received. Thus, the responses can be made to depend on the
3315 challenges. If an invalid string is ignored, an empty string is placed in
3318 PH/46 Messages that are created by the autoreply transport now contains a
3319 References: header, in accordance with RFCs 2822 and 3834.
3321 PH/47 Added authenticated_sender_force to the smtp transport.
3323 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3325 PH/49 Installed latest patch from the Sieve maintainer.
3327 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3328 was also set, the appendfile transport was unnecessarily scanning a
3329 directory of message files (e.g. for maildir delivery) to find the count
3330 of files (along with the size), even though it did not need this
3331 information. It now does the scan only if it needs to find either the
3332 size of the count of files.
3334 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3336 PH/52 Two bugs concerned with error handling when the smtp transport is
3339 (i) Exim was not creating retry information for temporary errors given
3340 for individual recipients after the DATA command when the smtp transport
3341 was used in LMTP mode. This meant that they could be retried too
3342 frequently, and not timed out correctly.
3344 (ii) Exim was setting the flag that allows error details to be returned
3345 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3346 recipients that were returned after the DATA command.
3348 PH/53 This is related to PH/52, but is more general: for any failing address,
3349 when detailed error information was permitted to be returned to the
3350 sender, but the error was temporary, then after the final timeout, only
3351 "retry timeout exceeded" was returned. Now it returns the full error as
3352 well as "retry timeout exceeded".
3354 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3355 do this, and (what is worse) MTAs that accept it.
3357 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3358 will now be deprecated.
3360 PH/56 New os.c-cygwin from the Cygwin maintainer.
3362 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3363 formats (previously only available in exim formats via -bpr, -bpru,
3364 and -bpra. Now also available in native and exiqgrep formats)
3366 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3367 with very large, slow to parse queues
3369 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3371 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3373 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3374 responses to authentication challenges, though it was showing the
3375 challenges; (ii) I've removed the CR characters from the debug output for
3378 PH/58 Allow for the insertion of a newline as well as a space when a string
3379 is turned into more than one encoded-word during RFC 2047 encoding. The
3380 Sieve code now uses this.
3382 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3383 data_4xx, lost_connection, tls_required.
3385 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3386 message was being sent as an SMTP response.
3388 PH/61 Add -l and -k options to exicyclog.
3390 PH/62 When verifying, if an address was redirected to one new address, so that
3391 verification continued, and the new address failed or deferred after
3392 having set something in $address_data, the value of $address_data was not
3393 passed back to the ACL. This was different to the case when no
3394 redirection occurred. The value is now passed back in both cases.
3396 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3397 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3398 use login_cap.h, so on its own it isn't the distinguishing feature. The
3399 new name refers directly to the setclassresources() function.
3401 PH/65 Added configuration files for NetBSD3.
3403 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3405 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3406 is preferred over IPv4.
3408 PH/68 The bounce_return_message and bounce_return_body options were not being
3409 honoured for bounces generated during the reception of non-SMTP messages.
3410 In particular, this applied to messages rejected by the ACL. This bug has
3411 been fixed. However, if bounce_return_message is true and bounce_return_
3412 body is false, the headers that are returned for a non-SMTP message
3413 include only those that have been read before the error was detected.
3414 (In the case of an ACL rejection, they have all been read.)
3416 PH/69 The HTML version of the specification is now built in a directory called
3417 spec_html instead of spec.html, because the latter looks like a path with
3418 a MIME-type, and this confuses some software.
3420 PH/70 Catch two compiler warnings in sieve.c.
3422 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3423 function verify_get_ident() calls ip_connect() to connect a socket, but
3424 if the "connect()" function timed out, ip_connect() used to close the
3425 socket. However, verify_get_ident() also closes the socket later, and in
3426 between Exim writes to the log, which may get opened at this point. When
3427 the socket was closed in ip_connect(), the log could get the same file
3428 descriptor number as the socket. This naturally causes chaos. The fix is
3429 not to close the socket in ip_connect(); the socket should be closed by
3430 the function that creates it. There was only one place in the code where
3431 this was missing, in the iplookup router, which I don't think anybody now
3432 uses, but I've fixed it anyway.
3434 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3435 well as to direct DNS lookups. Otherwise the handling of names in host
3436 lists is inconsistent and therefore confusing.
3442 PH/01 Two changes to the default runtime configuration:
3444 (1) Move the checks for relay_from_hosts and authenticated clients from
3445 after to before the (commented out) DNS black list checks.
3447 (2) Add control=submission to the relay_from_hosts and authenticated
3448 clients checks, on the grounds that messages accepted by these
3449 statements are most likely to be submissions.
3451 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3453 (1) Generate an error if the third argument for the ${prvs expansion is
3456 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3459 (3) Reset the variables that are obtained from the first argument of
3460 ${prvscheck and used in the second argument before leaving the code,
3461 because their memory is reclaimed, so using them afterwards may do
3464 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3465 one (it's much easier than Tom thought :-).
3467 (5) Because of (4), we can now allow for the use of $prvscheck_result
3468 inside the third argument.
3470 PH/03 For some reason, the default setting of PATH when running a command from
3471 a pipe transport was just "/usr/bin". I have changed it to
3474 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3475 anything to be listed in the output from -bV.
3477 PH/05 When a filter generated an autoreply, the entire To: header line was
3478 quoted in the delivery log line, like this:
3480 => >A.N.Other <ano@some.domain> <original@ddress> ...
3482 This has been changed so that it extracts the operative address. There
3483 may be more than one such address. If so, they are comma-separated, like
3486 => >ano@some.domain,ona@other.domain <original@ddress> ...
3488 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3489 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3490 not being looked up in the rDNS to get a host name, Exim was showing the
3491 IP address twice in Received: lines, even though the IP addresses were
3492 identical. For example:
3494 Received: from [1.2.3.4] (helo=[1.2.3.4])
3496 However, if the real host name was known, it was omitting the HELO data
3497 if it matched the actual IP address. This has been tidied up so that it
3498 doesn't show the same IP address twice.
3500 PH/07 When both +timestamp and +memory debugging was on, the value given by
3501 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3502 called by the debug printing, thereby overwriting the timestamp buffer.
3503 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3505 PH/08 When the original message was included in an autoreply transport, it
3506 always said "this is a copy of the message, including all the headers",
3507 even if body_only or headers_only was set. It now gives an appropriate
3510 PH/09 Applied a patch from the Sieve maintainer which:
3512 o fixes some comments
3513 o adds the (disabled) notify extension core
3514 o adds some debug output for the result of if/elsif tests
3515 o points to the current vacation draft in the documentation
3516 and documents the missing references header update
3520 o fixes a bug in processing the envelope test (when testing
3521 multiple envelope elements, the last element determined the
3524 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3525 Electronic Mail") by including:
3527 Auto-submitted: auto-generated
3529 in the messages that it generates (bounce messages and others, such as
3530 warnings). In the case of bounce messages for non-SMTP messages, there was
3531 also a typo: it was using "Auto_submitted" (underscore instead of
3532 hyphen). Since every message generated by Exim is necessarily in response
3533 to another message, thes have all been changed to:
3535 Auto-Submitted: auto-replied
3537 in accordance with these statements in the RFC:
3539 The auto-replied keyword:
3541 - SHOULD be used on messages sent in direct response to another
3542 message by an automatic process,
3544 - MUST NOT be used on manually-generated messages,
3546 - MAY be used on Delivery Status Notifications (DSNs) and Message
3547 Disposition Notifications (MDNs),
3549 - MUST NOT be used on messages generated by automatic or periodic
3550 processes, except for messages which are automatic responses to
3553 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3554 to the default Received: header definition.
3556 PH/12 Added log selector acl_warn_skipped (default on).
3558 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3559 variables because (a) they are in the wrong storage pool and (b) even if
3560 they were copied, it wouldn't work properly because of the caching.
3562 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3563 checking when decoding. Apparently there are clients that generate
3564 overlong encoded strings. Why am I not surprised?
3566 PH/15 If the first argument of "${if match_address" was not empty, but did not
3567 contain an "@" character, Exim crashed. Now it writes a panic log message
3568 and treats the condition as false.
3570 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3572 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3573 extension "envelope-auth". The code is finished and in agreement with
3574 other implementations, but there is no documentation so far and in fact,
3575 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3576 not changing the active code.
3578 Print executed "if" and "elsif" statements when debugging is used. This
3579 helps a great deal to understand what a filter does.
3581 Document more things not specified clearly in RFC3028. I had all this
3582 sorted out, when out of a sudden new issues came to my mind. Oops."
3584 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3587 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3588 canonical form (as documented). However, after a host name lookup from
3589 the IP address, check_host() was doing a simple string comparison with
3590 addresses acquired from the DNS when checking that the found name did
3591 have the original IP as one of its addresses. Since any found IPv6
3592 addresses are likely to be in abbreviated form, the comparison could
3593 fail. Luckily, there already exists a function for doing the comparison
3594 by converting both addresses to binary, so now that is used instead of
3595 the text comparison.
3597 PH/20 There was another similar case to PH/19, when a complete host name was
3598 given in a host list; looking up its IP address could give an abbreviated
3599 form, whereas the current host's name might or might not be abbreviated.
3600 The same fix has been applied.
3606 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3607 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3610 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3611 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3613 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3615 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3616 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3617 Also, for "Auto-Submitted", treat anything other than "no" as
3618 non-personal, in accordance with RFC 3834. (Previously it treated
3619 anything starting "auto-" as non-personal.)
3621 TF/01 The control=submission/name=... option had a problem with syntax
3622 errors if the name included a slash character. The /name= option
3623 now slurps the rest of the string, so it can include any characters
3624 but it must come last in the list of options (after /sender_retain
3627 PH/05 Some modifications to the interface to the fake nameserver for the new
3635 TK/01 Added the "success_on_redirect" address verification option. See
3636 NewStuff for rationale and an example.
3638 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3640 PH/02 Patch to exigrep to allow it to work on syslog lines.
3642 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3643 fread() to skip over the body file's header line, because in Cygwin the
3644 header line is locked and is inaccessible.
3646 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3647 co-exist for some time) to make it clear that it is the Exim ID that is
3648 referenced, not the Message-ID: header line.
3650 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3651 string_format() function, because snprintf() does not exist on all
3654 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3655 ${stat: expansion item.
3657 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3658 protocol synchronization error", to keep the pedants happy.
3660 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3661 well as for IRIX systems, when gcc is being used. See the host.c source
3664 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3666 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3669 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3670 transport and to the smtp transport in LMTP mode.
3672 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3674 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3675 run a filter in a subprocess. This could lead to confusion in subsequent
3676 lookups in the parent process. There should also be a search_tidyup() at
3677 the end of the subprocess.
3679 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3680 only if the host matched helo_try_verify_hosts, which caused the
3681 verification to occur when the EHLO/HELO command was issued. The ACL just
3682 tested the remembered result. Now, if a previous verification attempt has
3683 not happened, "verify = helo" does it there and then.
3685 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3687 TK/03 Fix log output including CR from clamd.
3689 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3690 exist provoked a memory error which could cause a segfault.
3692 PH/15 Installed PCRE 6.2
3694 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3696 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3697 of the problem. Specifically, suggested +O2 rather than +O1 for the
3700 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3702 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3703 record, the port was not correctly logged when the outgoing_port log
3704 selector was set (it logged the transort's default port).
3706 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3707 fallback_hosts, and "hosts" in the smtp transport.
3709 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3710 host errors such as "Connection refused".
3712 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3713 authentication with radiusclient 0.4.9:
3715 - Error returned from rc_read_config was caught wrongly
3716 - Username/password not passed on to radius server due to wrong length.
3718 The presumption is that some radiusclient API changes for 4.51/PH/17
3719 were not taken care of correctly. The code is still untested by me (my
3720 Linux distribution still has 0.3.2 of radiusclient), but it was
3721 contributed by a Radius user.
3723 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3724 expanding the "port" option of the smtp transport.
3726 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3727 while reading a MIME header. Thanks to Tom Hughes for a patch.
3729 PH/24 Include config.h inside local_scan.h so that configuration settings are
3732 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3733 that in an ACL for RCPT (for example), you can examine exactly what was
3736 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3737 commands, but it was not correctly comparing the address with the actual
3738 client host address. Thus, it would show the EHLO address in Received:
3739 header lines when this was not necessary.
3741 PH/27 Added the % operator to ${eval:}.
3743 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3744 it should be ignoring failures (because with -C, for example, it has lost
3745 privilege). It wasn't ignoring creation failures other than "already
3748 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3751 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3752 because an input line was too long, either on its own, or by virtue of
3753 too many continuations, the temporary file was not being removed, and the
3754 return code was incorrect.
3756 PH/31 Missing "BOOL" in function definition in filtertest.c.
3758 PH/32 Applied Sieve patches from the maintainer.
3760 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3762 PH/33 Added "verify = not_blind".
3764 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3765 Local/Makefile (with some defaults set). These are used in built scripts
3766 such as exicyclog, but they have never been used in the exim_install
3767 script (though there are many overriding facilities there). I have
3768 arranged that the exim_install script now takes note of these two
3771 PH/35 Installed configuration files for Dragonfly.
3773 PH/36 When a locally submitted message by a trusted user did not contain a
3774 From: header, and the sender address was obtained from -f or from an SMTP
3775 MAIL command, and the trusted user did not use -F to supply a sender
3776 name, $originator_name was incorrectly used when constructing a From:
3777 header. Furthermore, $originator_name was used for submission mode
3778 messages from external hosts without From: headers in a similar way,
3779 which is clearly wrong.
3781 PH/37 Added control=suppress_local_fixups.
3783 PH/38 When log_selector = +received_sender was set, and the addition of the
3784 sender made the log line's construction buffer exactly full, or one byte
3785 less than full, an overflow happened when the terminating "\n" was
3788 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3789 when the result of a list match is failure because a DNS lookup failed.
3791 PH/40 RM_COMMAND is now used in the building process.
3793 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3794 the "build-* directories that it finds.
3796 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3797 address in a domain literal was a prefix of an interface address.
3799 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3800 when verifying a sender address, unless rewrite_headers is false.
3802 PH/44 Wrote a long comment about why errors_to addresses are verified as
3803 recipients, not senders.
3805 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3806 the ratelimit ACL was added.
3808 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3810 PH/46 Added extra information about PostgreSQL errors to the error string.
3812 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3813 avoiding the need to install special zones in a real server. This is
3814 backwards compatible; if it can't find the fake resolver, it drops back.
3815 Thus, both old and new test suites can be run.
3817 TF/02 Added util/ratelimit.pl
3819 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3820 clock is set back in time.
3822 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3823 Candler <B.Candler@pobox.com>.
3825 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3826 OK for addresses that are the result of redirections.
3828 PH/48 A number of further additions for the benefit of the new test suite,
3829 including a fake gethostbyname() that interfaces to the fake DNS resolver
3832 TF/06 The fix for widen_domains has also been applied to qualify_single and
3833 search_parents which are the other dnslookup options that can cause
3836 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3839 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3841 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3842 service" if it encountered an MX record with an empty target hostname.
3843 The message is now "an MX or SRV record indicated no SMTP service".
3845 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3846 if the DNS of the sending site is misconfigured. This is quite a
3847 common situation. This change restores the behaviour of treating a
3848 helo verification defer as a failure.
3850 PH/51 If self=fail was set on a router, the bounce message did not include the
3851 actual error message.
3857 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3859 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3860 command itself ended in error, the underlying message about the transport
3861 filter timeout was being overwritten with the pipe command error. Now the
3862 underlying error message should be appended to the second error message.
3864 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3866 PH/02 Reduce the amount of output that "make" produces by default. Full output
3867 can still be requested.
3869 PH/03 The warning log line about a condition test deferring for a "warn" verb
3870 was being output only once per connection, rather than after each
3871 occurrence (because it was using the same function as for successful
3872 "warn" verbs). This seems wrong, so I have changed it.
3874 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3875 it should not have, which might have caused a crash in the right
3876 circumstances, but probably never did.
3878 PH/04 Installed a modified version of Tony Finch's patch to make submission
3879 mode fix the return path as well as the Sender: header line, and to
3880 add a /name= option so that you can make the user's friendly name appear
3883 TF/03 Added the control = fakedefer ACL modifier.
3885 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3886 Mark Lowes for thorough testing.
3888 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3890 TK/03 Merged latest SRS patch from Miles Wilton.
3892 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3893 with the definition in sysexits.h (which is #included earlier).
3894 Fortunately, Exim does not actually use EX_OK. The code used to try to
3895 preserve the sysexits.h value, by assuming that macro definitions were
3896 scanned for macro replacements. I have been disabused of this notion,
3897 so now the code just undefines EX_OK before #including unistd.h.
3899 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3900 in the smtp transport. When a block could not be written in a single
3901 write() function, the timeout was being re-applied to each part-write.
3902 This seems wrong - if the receiver was accepting one byte at a time it
3903 would take for ever. The timeout is now adjusted when this happens. It
3904 doesn't have to be particularly precise.
3906 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3907 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3909 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3910 without a domain if the check to <postmaster@domain> fails.
3912 SC/01 Eximstats: added -xls and the ability to specify output files
3913 (patch written by Frank Heydlauf).
3915 SC/02 Eximstats: use FileHandles for outputting results.
3917 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3919 SC/04 Eximstats: fixed display of large numbers with -nvr option
3921 SC/05 Eximstats: fixed merging of reports with empty tables.
3923 SC/06 Eximstats: added the -include_original_destination flag
3925 SC/07 Eximstats: removed tabs and trailing whitespace.
3927 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3929 TK/06 MBOX spool code: Add real "From " MBOX separator line
3930 so the .eml file is really in mbox format (even though
3931 most programs do not really care). Patch from Alex Miller.
3933 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3934 The latter is generated from $received_to and is only set if the
3935 message has one envelope recipient. SA can use these headers,
3936 obviously out-of-the-box. Patch from Alex Miller.
3938 PH/08 The ${def test on a variable was returning false if the variable's
3939 value was "0", contrary to what the specification has always said!
3940 The result should be true unless the variable is empty.
3942 PH/09 The syntax error of a character other than { following "${if
3943 def:variable_name" (after optional whitespace) was not being diagnosed.
3944 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3945 accidental colon was present, for example, could give incorrect results.
3947 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3948 result is used (not including appendfile, where other changes are about
3951 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3952 This involved changing a lot of size variables from int to off_t. It
3953 should work with maildirs and everything.
3955 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3956 spamd dying while we are connected to it.
3958 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3961 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3962 function for BDB 4.3.
3964 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3966 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3967 the sources for PCRE have been re-arranged and more files are now
3970 PH/15 The code I had for printing potentially long long variables in PH/11
3971 above was not the best (it lost precision). The length of off_t variables
3972 is now inspected at build time, and an appropriate printing format (%ld
3973 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3974 to be "long long int" or "long int". This is needed for the internal
3975 formatting function string_vformat().
3977 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3978 the configuration file to be ":syslog", then the script "guesses" where
3979 the logs files are, rather than using the compiled in default. In our
3980 case the guess is not the same as the compiled default, so the script
3981 suddenly stopped working when I started to use syslog. The patch checks
3982 to see if log_file_path is "". If so, it attempts to read it from exim
3983 with no configuration file to get the compiled in version, before it
3984 falls back to the previous guessing code."
3986 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3987 implementing BATV in an Exim configuration. See NewStuff for the gory
3990 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3991 Makefile that are specific to HP-UX.
3993 PH/18 If the "use_postmaster" option was set for a recipient callout together
3994 with the "random" option, the postmaster address was used as the MAIL
3995 FROM address for the random test, but not for the subsequent recipient
3996 test. It is now used for both.
3998 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3999 patch removes a few documentation additions to RFC 3028, because the
4000 latest draft now contains them. It adds the new en;ascii-case comparator
4001 and a new error check for 8bit text in MIME parts. Comparator and
4002 require names are now matched exactly. I enabled the subaddress
4003 extension, but it is not well tested yet (read: it works for me)."
4005 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4006 rework some of the code of TK/09 above to avoid the hardwired use of
4007 "%lld" and "long long". Replaced the call to snprintf() with a call to
4010 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4011 records point to non-existent hosts", "retry timeout exceeded", and
4012 "retry time not reached for any host after a long failure period".
4014 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4015 experimental DomainKeys support:
4017 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4018 (2) On an error such as an illegally used "control", the wrong name for
4019 the control was given.
4021 These problems did NOT occur unless DomainKeys support was compiled.
4023 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4025 PH/24 Added ${if match_ip condition.
4027 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4028 databases so that it will be absolutely obvious if a crash occurs in the
4029 DB library. This is a regular occurrence (often caused by mis-matched
4032 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4033 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4034 picked up on a user's system that detects such things. There doesn't seem
4035 to be a gcc warning option for this - only an attribute that has to be
4036 put on the function's prototype. It seems that in Fedora Core 4 they have
4037 set this on a number of new functions. No doubt there will be more in due
4040 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4041 specify a transport. However, if an address that was verified by such a
4042 router was the subject of a callout, Exim crashed because it tried to
4043 read the rcpt_include_affixes from the non-existent transport. Now it
4044 just assumes that the setting of that option is false. This bug was
4045 introduced by 4.51/PH/31.
4047 PH/28 Changed -d+all to exclude +memory, because that information is very
4048 rarely of interest, but it makes the output a lot bigger. People tend to
4049 do -d+all out of habit.
4051 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4052 code in os-type was giving problems when libc.so lives in lib64, like on
4055 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4056 aren't the modern standard, and it seems that some systems' include files
4057 don't always have them. Exim was already checking for some of the newer
4058 ones like T_AAAA, and defining it itself. I've added checks for all the
4059 record types that Exim uses.
4061 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4062 not automatically generating a new one, as it is supposed to. This
4063 prevented TLS from working. If the file did exist, but contained invalid
4064 data, a new version was generated, as expected. It was only the case of a
4065 non-existent file that was broken.
4067 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4068 with a change in libdomainkeys > 0.64.
4070 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4071 from DNS. If the selector record carries the flag, it now has
4072 precedence over the domain-wide flag.
4074 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4076 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4077 the use of a port name that isn't defined in /etc/services) occurred, the
4078 message was deferred as in a normal delivery, and thus remained on the
4079 spool, instead of being failed because of the mua_wrapper setting. This
4080 is now fixed, and I tidied up some of the mua_wrapper messages at the
4083 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4084 the messages in a hash of arrays rather than using individual hashes.
4085 This is a bit cleaner and results in dramatic memory savings, albeit
4086 at a slight CPU cost.
4088 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4089 as requested by Marc Sherman.
4091 SC/10 Eximstats: added histograms for user specified patterns as requested
4094 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4096 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4097 fopen() in the content-scanning modules that did not already have it.
4103 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4104 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4106 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4108 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4110 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4111 to test Sieve filters that use "vacation".
4113 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4114 that changes the way the GnuTLS parameters are stored in the cache file.
4115 The new format can be generated externally. For backward compatibility,
4116 if the data in the cache doesn't make sense, Exim assumes it has read an
4117 old-format file, and it generates new data and writes a new file. This
4118 means that you can't go back to an older release without removing the
4121 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4122 work if there are any delivery delays because "one_time" forces the
4123 parent to be marked "delivered", so its unseen clone is never tried
4124 again. For this reason, Exim now forbids the simultaneous setting of
4127 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4128 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4129 entry if you want to know the details. The fix, however, neglected to
4130 consider the case when local delivery batching is involved. The test for
4131 "previously delivered" was not happening when checking to see if an
4132 address could be batched with a previous (undelivered) one; under
4133 certain circumstances this could lead to multiple deliveries to the same
4136 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4137 in its include files, and this causes problems building Exim.
4139 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4140 header_syntax) but Exim was just ignoring anything given after a slash.
4141 In particular, this caused confusion with an attempt to use "verify =
4142 reverse_host_lookup/defer_ok". An error is now given when options are
4143 supplied for verify items that do not have them. (Maybe reverse_host_
4144 lookup should have a defer_ok option, but that's a different point.)
4146 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4147 defined by RFC 821) to 2048, because there were problems with some AUTH
4148 commands, and RFC 1869 says the size should be increased for extended
4149 SMTP commands that take arguments.
4151 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4154 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4155 "unknown" error; now it says that the functionality isn't in the binary.
4157 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4158 an address' error message when a string expansion fails (syntax or
4159 whatever). Otherwise the password may appear in the log. Following change
4160 PH/42 below, there is no longer a chance of it appearing in a bounce
4163 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4165 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4166 of its addresses was ever tried. (Bugzilla bug #2.)
4168 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4169 the result incorrectly in the debug output. (It correctly added a newline
4170 to what was transported.)
4172 TF/01 Added $received_time.
4174 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4175 commented out examples of how to interface to a virus scanner and to
4176 SpamAssassin. Also added commented examples of av_scanner and
4177 spamd_address settings.
4179 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4180 and controls are allowed in which ACLs. There were a couple of minor
4181 errors. Some of the entries in the conditions table (which is a table of
4182 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4183 negation of where the condition IS allowed.
4185 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4187 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4188 header file does not have a version number, so I've had to invent a new
4189 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4190 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4191 radiusclient), but it was contributed by a Radius user.
4193 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4194 files or directories, for OpenSSL.
4196 PH/19 When an Exim process that is running as root has to create an Exim log
4197 file, it does so in a subprocess that runs as exim:exim so as to get the
4198 ownership right at creation (otherwise, other Exim processes might see
4199 the file with the wrong ownership). There was no test for failure of this
4200 fork() call, which would lead to the process getting stuck as it waited
4201 for a non-existent subprocess. Forks do occasionally fail when resources
4202 run out. I reviewed all the other calls to fork(); they all seem to check
4205 PH/20 When checking for unexpected SMTP input at connect time (before writing
4206 the banner), Exim was not dealing correctly with a non-positive return
4207 from the read() function. If the client had disconnected by this time,
4208 the result was a log entry for a synchronization error with an empty
4209 string after "input=" when read() returned zero. If read() returned -1
4210 (an event I could not check), uninitialized data bytes were printed.
4211 There were reports of junk text (parts of files, etc) appearing after
4214 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4216 PH/22 Added support for macro redefinition, and (re)definition in between
4217 driver and ACL definitions.
4219 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4220 forgetting to use the resulting value; it was using the unexpanded value.
4222 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4223 hadn't been configured. The fix is from Juergen Kreileder, who
4224 understands it better than I do:
4226 "Here's what I see happening with three configured cyrus_sasl
4227 authenticators configured (plain, login, cram-md5):
4229 On startup auth_cyrus_sasl_init() gets called for each of these.
4230 This means three calls to sasl_listmech() without a specified mech_list.
4231 => SASL tests which mechs of all available mechs actually work
4232 => three warnings about OTP not working
4233 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4235 With the patch, sasl_listmech() also gets called three times. But now
4236 SASL's mech_list option is set to the server_mech specified in the the
4237 authenticator. Or in other words, the answer from sasl_listmech()
4238 gets limited to just the mech you're testing for (which is different
4240 => the return list contains just 'plain' or 'login', 'cram-md5' or
4241 nothing depending on the value of ob->server_mech.
4243 I've just tested the patch: Authentication still works fine,
4244 unavailable mechs specified in the exim configuration are still
4245 caught, and the auth.log warnings about OTP are gone."
4247 PH/25 When debugging is enabled, the contents of the command line are added
4248 to the debugging output, even when log_selector=+arguments is not
4251 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4252 answer is "GNU", and only if the return is "GNU/something" is the answer
4255 PH/27 $acl_verify_message is now set immediately after the failure of a
4256 verification in an ACL, and so is available in subsequent modifiers. In
4257 particular, the message can be preserved by coding like this:
4259 warn !verify = sender
4260 set acl_m0 = $acl_verify_message
4262 Previously, $acl_verify_message was set only while expanding "message"
4263 and "log_message" when a very denied access.
4265 PH/28 Modified OS/os.c-Linux with
4267 -#ifndef OS_LOAD_AVERAGE
4268 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4270 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4271 nomenclature these days.)
4273 PH/29 Installed patch from the Sieve maintainer that adds the options
4274 sieve_useraddress and sieve_subaddress to the redirect router.
4276 PH/30 In these circumstances:
4277 . Two addresses routed to the same list of hosts;
4278 . First host does not offer TLS;
4279 . First host accepts first address;
4280 . First host gives temporary error to second address;
4281 . Second host offers TLS and a TLS session is established;
4282 . Second host accepts second address.
4283 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4284 and peerdn, if requested) that were in fact used only for the second
4287 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4288 attention to any local part prefix or suffix that was matched by the
4289 router that accepted the address. It now behaves in the same way as it
4290 does for delivery: the affixes are removed from the local part unless
4291 rcpt_include_affixes is set on the transport.
4293 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4294 timeout during the DATA phase of an incoming message.
4296 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4297 applied a patch sanctioned by the Sieve maintainer.
4299 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4300 the uid or gid is negative. A case of a negative gid caused this to be
4301 noticed. The fix allows for either to be negative.
4303 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4304 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4307 PH/36 The change PH/12 above was broken. Fixed it.
4309 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4310 the grounds that routing the same address twice would always produce the
4311 same answer. This might have been true once, but it is certainly no
4312 longer true now. Routing a child address may depend on the previous
4313 routing that produced that child. Some complicated redirection strategies
4314 went wrong when messages had multiple recipients, and made Exim's
4315 behaviour dependent on the order in which the addresses were given.
4317 I have moved the duplicate checking until after the routing is complete.
4318 Exim scans the addresses that are assigned to local and remote
4319 transports, and removes any duplicates. This means that more work will be
4320 done, as duplicates will always all be routed, but duplicates are
4321 presumably rare, so I don't expect this is of any significance.
4323 For deliveries to pipes, files, and autoreplies, the duplicate checking
4324 still happens during the routing process, since they are not going to be
4327 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4328 It corrects a timeout issue with spamd. This is Ian's comment: "The
4329 background is that sometimes spamd either never reads data from a
4330 connection it has accepted, or it never writes response data. The exiscan
4331 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4332 blindly assumes that writes won't block so it may never time out."
4334 PH/39 Allow G after quota size as well as K and M.
4336 PH/40 The value set for $authenticated_id in an authenticator may not contain
4337 binary zeroes or newlines because the value is written to log lines and
4338 to spool files. There was no check on this. Now the value is run through
4339 the string_printing() function so that such characters are converted to
4340 printable escape sequences.
4342 PH/41 $message_linecount is a new variable that contains the total number of
4343 lines in the message. Compare $body_linecount, which is the count for the
4346 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4347 bounce and delay warning messages, except in certain special cases, which
4350 (a) An SMTP error message from a remote host;
4351 (b) A message specified in a :fail: redirection;
4352 (c) A message specified in a "fail" command in a system filter;
4353 (d) A message specified in a FAIL return from the queryprogram router;
4354 (e) A message specified by the cannot_route_message router option.
4356 In these cases only, Exim does include the error details in bounce and
4357 warning messages. There are also a few cases where bland messages such
4358 as "unrouteable address" or "local delivery error" are given.
4360 PH/43 $value is now also set for the "else" part of a ${run expansion.
4362 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4363 being worked on, but at least Exim now implements the latest version to
4366 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4367 process to complete was treated as a delivery failure, a timeout while
4368 writing the message to the pipe was logged, but erroneously treated as a
4369 successful delivery. Such timeouts include transport filter timeouts. For
4370 consistency with the overall process timeout, these timeouts are now
4371 treated as errors, giving rise to delivery failures by default. However,
4372 there is now a new Boolean option for the pipe transport called
4373 timeout_defer, which, if set TRUE, converts the failures into defers for
4374 both kinds of timeout. A transport filter timeout is now identified in
4377 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4378 systems where "make" and "gmake" are different, calling "gmake" at top
4379 level broke things. I've arranged for the value of $(MAKE) to be passed
4380 from the Makefile to this script so that it can call the same version of
4384 A note about Exim versions 4.44 and 4.50
4385 ----------------------------------------
4387 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4388 changes of various kinds. As a consequence, a big documentation update was
4389 needed. This delayed the release for rather longer than seemed good, especially
4390 in the light of a couple of (minor) security issues. Therefore, the changes
4391 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4392 release. So 4.44 and 4.50 are in effect two different branches that both start
4395 I have left the 4.50 change log unchanged; it contains all the changes since
4396 4.43. The change log for 4.44 is below; many of its items are identical to
4397 those for 4.50. This seems to be the most sensible way to preserve the
4398 historical information.
4404 1. Minor wording change to the doc/README.SIEVE file.
4406 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4407 computation of the current number of files was incorrect.
4409 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4410 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4413 4. Give more explanation in the error message when the command for a transport
4414 filter fails to execute.
4416 5. There are several places where Exim runs a non-Exim command in a
4417 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4418 was being done only for the command run by the queryprogram router. It is
4419 now done for all such subprocesses. The other cases are: ${run, transport
4420 filters, and the commands run by the lmtp and pipe transports.
4422 6. Added CONFIGURE_GROUP build-time option.
4424 7. Some older OS have a limit of 256 on the maximum number of file
4425 descriptors. Exim was using setrlimit() to set 1000 as a large value
4426 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4427 systems. I've change it so that if it can't get 1000, it tries for 256.
4429 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4430 was an oversight, and furthermore, ever since the addition of extra
4431 controls (e.g. 4.43/32), the checks on when to allow different forms of
4432 "control" were broken. There should now be diagnostics for all cases when a
4433 control that does not make sense is encountered.
4435 9. Added the /retain_sender option to "control=submission".
4437 10. $recipients is now available in the predata ACL (oversight).
4439 11. Tidy the search cache before the fork to do a delivery from a message
4440 received from the command line. Otherwise the child will trigger a lookup
4441 failure and thereby defer the delivery if it tries to use (for example) a
4442 cached ldap connection that the parent has called unbind on.
4444 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4445 of $address_data from the recipient verification was clobbered by the
4446 sender verification.
4448 13. The value of address_data from a sender verification is now available in
4449 $sender_address_data in subsequent conditions in the ACL statement.
4451 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4453 15. Added a new option "connect=<time>" to callout options, to set a different
4456 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4457 was its contents. (It was OK if the option was not defined at all.)
4459 17. A "Completed" log line is now written for messages that are removed from
4460 the spool by the -Mrm option.
4462 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4463 information about exactly what failed.
4465 19. Added -dd to debug only the daemon process.
4467 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4468 handles timeouts, both on the server side and network timeouts. Renamed the
4469 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4471 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4472 It is now set to "smtps".
4474 22. $host_address is now set to the target address during the checking of
4475 ignore_target_hosts.
4477 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4478 being passed; this would have caused $sender_host_name to have been used if
4479 matching the list had actually called for a host name (not very likely,
4480 since this list is usually IP addresses). A host name is now passed as
4483 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4484 code that specifies a non-restarting handler (typically sigaction() in
4485 modern systems) in an attempt to fix a rare and obscure crash bug.
4487 25. Narrowed the window for a race in the daemon that could cause it to ignore
4488 SIGCHLD signals. This is not a major problem, because they are used only to
4489 wake it up if nothing else does.
4491 26. A malformed maildirsize file could cause Exim to calculate negative values
4492 for the mailbox size or file count. Odd effects could occur as a result.
4493 The maildirsize information is now recalculated if the size or filecount
4496 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4497 support for a long time. Removed HAVE_SYS_VFS_H.
4499 28. Installed the latest version of exipick from John Jetmore.
4501 29. In an address list, if the pattern was not a regular expression, an empty
4502 subject address (from a bounce message) matched only if the pattern was an
4503 empty string. Non-empty patterns were not even tested. This was the wrong
4504 because it is perfectly reasonable to use an empty address as part of a
4505 database query. An empty address is now tested by patterns that are
4506 lookups. However, all the other forms of pattern expect the subject to
4507 contain a local part and a domain, and therefore, for them, an empty
4508 address still always fails if the pattern is not itself empty.
4510 30. Exim went into a mad DNS loop when attempting to do a callout where the
4511 host was specified on an smtp transport, and looking it up yielded more
4512 than one IP address.
4514 31. Re-factored the code for checking spool and log partition space into a
4515 function that finds that data and another that does the check. The former
4516 is then used to implement four new variables: $spool_space, $log_space,
4517 $spool_inodes, and $log_inodes.
4519 32. The RFC2047 encoding function was originally intended for short strings
4520 such as real names; it was not keeping to the 75-character limit for
4521 encoded words that the RFC imposes. It now respects the limit, and
4522 generates multiple encoded words if necessary. To be on the safe side, I
4523 have increased the buffer size for the ${rfc2047: expansion operator from
4526 33. It is now permitted to omit both strings after an "if" condition; if the
4527 condition is true, the result is "true". As before, when the second string
4528 is omitted, a false condition yields an empty string. This makes it less
4529 cumbersome to write custom ACL and router conditions.
4531 34. Failure to deliver a bounce message always caused it to be frozen, even if
4532 there was an errors_to setting on the router. The errors_to setting is now
4535 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4536 canonical form (fully expanded) before being placed in
4537 $sender_host_address.
4539 36. The table in the code that translates DNS record types into text (T_A to
4540 "A" for instance) was missing entries for NS and CNAME. It is just possible
4541 that this could have caused confusion if both these types were looked up
4542 for the same domain, because the text type is used as part of Exim's
4543 per-process caching. But the chance of anyone hitting this buglet seems
4546 37. The dnsdb lookup has been extended in a number of ways.
4548 (1) There is a new type, "zns", which walks up the domain tree until it
4549 finds some nameserver records. It should be used with care.
4551 (2) There is a new type, "mxh", which is like "mx" except that it returns
4552 just the host names, not the priorities.
4554 (3) It is now possible to give a list of domains (or IP addresses) to be
4555 looked up. The behaviour when one of the lookups defers can be
4556 controlled by a keyword.
4558 (4) It is now possible to specify the separator character for use when
4559 multiple records are returned.
4561 38. The dnslists ACL condition has been extended: it is now possible to supply
4562 a list of IP addresses and/or domains to be looked up in a particular DNS
4565 39. Added log_selector=+queue_time_overall.
4567 40. When running the queue in the test harness, wait just a tad after forking a
4568 delivery process, to get repeatability of debugging output.
4570 41. Include certificate and key file names in error message when GnuTLS fails
4571 to set them up, because the GnuTLS error message doesn't include the name
4572 of the failing file when there is a problem reading it.
4574 42. Allow both -bf and -bF in the same test run.
4576 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4578 44. The "Exiscan patch" is now merged into the mainline Exim source.
4580 45. Sometimes the final signoff response after QUIT could fail to get
4581 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4582 < 0 before doing a fflush(). This bug looks as though it goes back to the
4583 introduction of TLS in release 3.20, but "sometimes" must have been rare
4584 because the tests only now provoked it.
4586 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4587 (this can affect the format of dates).
4589 47. exim_tidydb, when checking for the continued existence of a message for
4590 which it has found a message-specific retry record, was not finding
4591 messages that were in split spool directories. Consequently, it was
4592 deleting retry records that should have stayed in existence.
4594 48. Steve fixed some bugs in eximstats.
4596 49. The SPA authentication driver was not abandoning authentication and moving
4597 on to the next authenticator when an expansion was forced to fail,
4598 contradicting the general specification for all authenticators. Instead it
4599 was generating a temporary error. It now behaves as specified.
4601 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4602 (the order specifies the preference for clients). The order is now AES256,
4603 AES128, 3DES, ARCFOUR128.
4605 51. Small patch to Sieve code - explicitly set From: when generating an
4608 52. Exim crashed if a remote delivery caused a very long error message to be
4609 recorded - for instance if somebody sent an entire SpamAssassin report back
4610 as a large number of 550 error lines. This bug was coincidentally fixed by
4611 increasing the size of one of Exim's internal buffers (big_buffer) that
4612 happened as part of the Exiscan merge. However, to be on the safe side, I
4613 have made the code more robust (and fixed the comments that describe what
4616 53. Now that there can be additional text after "Completed" in log lines (if
4617 the queue_time_overall log selector is set), a one-byte patch to exigrep
4618 was needed to allow it to recognize "Completed" as not the last thing in
4621 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4622 patch that reportedly fixes this has been added. I am not expert enough to
4623 create a test for it. This is what the patch creator wrote:
4625 "I found a little strange behaviour of ldap code when working with
4626 Windows 2003 AD Domain, where users was placed in more than one
4627 Organization Units. When I tried to give exim partial DN, the exit code
4628 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4629 But simultaneously result of request was absolutely normal ldap result,
4630 so I produce this patch..."
4632 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4633 REFERENCE, so I have modified the code to exclude the patch when that macro
4636 55. Some experimental protocols are using DNS PTR records for new purposes. The
4637 keys for these records are domain names, not reversed IP addresses. The
4638 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4639 leaves it alone. Component reversal etc. now happens only for IP addresses.
4642 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4644 57. Double the size of the debug message buffer (to 2048) so that more of very
4645 long debug lines gets shown.
4647 58. The exicyclog utility now does better if the number of log files to keep
4648 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4650 59. Two changes related to the smtp_active_hostname option:
4652 (1) $smtp_active_hostname is now available as a variable.
4653 (2) The default for smtp_banner uses $smtp_active_hostname instead
4654 of $primary_hostname.
4656 60. The host_aton() function is supposed to be passed a string that is known
4657 to be a valid IP address. However, in the case of IPv6 addresses, it was
4658 not checking this. This is a hostage to fortune. Exim now panics and dies
4659 if the condition is not met. A case was found where this could be provoked
4660 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4661 components; fortuitously, this particular loophole had already been fixed
4662 by change 4.50/55 above.
4664 If there are any other similar loopholes, the new check in host_aton()
4665 itself should stop them being exploited. The report I received stated that
4666 data on the command line could provoke the exploit when Exim was running as
4667 exim, but did not say which command line option was involved. All I could
4668 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4669 running as the user.
4672 61. There was a buffer overflow vulnerability in the SPA authentication code
4673 (which came originally from the Samba project). I have added a test to the
4674 spa_base64_to_bits() function which I hope fixes it.
4677 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4678 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4680 63. The daemon start-up calls getloadavg() while still root for those OS that
4681 need the first call to be done as root, but it missed one case: when
4682 deliver_queue_load_max is set with deliver_drop_privilege. This is
4683 necessary for the benefit of the queue runner, because there is no re-exec
4684 when deliver_drop_privilege is set.
4686 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4687 This has been fixed.
4689 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4690 was in use, was not putting the data itself into the right store pool;
4691 consequently, it could be overwritten for a subsequent message in the same
4692 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4695 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4697 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4698 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4699 the function were treating the return as a boolean value, which happened to
4700 work because 0=false and not-0=true, but is not correct code.
4702 68. The host_aton() function was not handling scoped IPv6 addresses (those
4703 with, for example, "%eth0" on the end) correctly.
4705 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4706 negated items (that is, ~something) in unsigned ints. Some compilers
4707 apparently mutter when there is no cast.
4709 70. If an address verification called from an ACL failed, and did not produce a
4710 user-specific message (i.e. there was only a "system" message), nothing was
4711 put in $acl_verify_message. In this situation, it now puts the system
4714 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4715 session; change 4.31/43 added the unwanted input to the log line - except
4716 that it did not do this in the start of session case. It now does.
4718 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4719 This is wrong and can cause the other end to generate a synchronization
4720 error if it is another Exim or anything else that does the synchronization
4721 check. A QUIT command is no longer sent after a timeout.
4723 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4724 during host lookups.
4726 74. The defer_ok option of callout verification was not working if it was used
4727 when verifying addresses in header lines, that is, for this case:
4729 verify = header_sender/callout=defer_ok
4731 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4732 those file descriptors could be used for SMTP connections. If anything
4733 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4734 could be sent to the SMTP client, causing chaos. The daemon now opens
4735 stdin, stdout, and stderr to /dev/null when it puts itself into the
4738 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4739 log by default. The user can override this with suitable Perl magic.
4741 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4742 the log message when discard triggers, was not working for the DATA ACL or
4743 for the non-SMTP ACL.
4745 78. Error message wording change in sieve.c.
4747 79. If smtp_accept_max_per_host was set, the number of connections could be
4748 restricted to fewer than expected, because the daemon was trying to set up
4749 a new connection before checking whether the processes handling previous
4750 connections had finished. The check for completed processes is now done
4751 earlier. On busy systems, this bug wouldn't be noticed because something
4752 else would have woken the daemon, and it would have reaped the completed
4755 80. If a message was submitted locally by a user whose login name contained one
4756 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4757 It caused a spool format error. I have fixed the spool reading code. A
4758 related problem was that the "from" clause in the Received: line became
4759 illegal because of the space(s). It is now covered by ${quote_local_part.
4761 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4764 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4766 83. Patch from Sieve maintainer to fix unterminated string problem in
4767 "vacation" handling.
4769 84. Some minor changes to the Linux configuration files to help with other
4770 OS variants using glibc.
4772 85. One more patch for Sieve to update vacation handling to latest spec.
4775 ----------------------------------------------------
4776 See the note above about the 4.44 and 4.50 releases.
4777 ----------------------------------------------------
4783 1. Change 4.43/35 introduced a bug that caused file counts to be
4784 incorrectly computed when quota_filecount was set in an appendfile
4787 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4788 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4791 3. Give more explanation in the error message when the command for a transport
4792 filter fails to execute.
4794 4. There are several places where Exim runs a non-Exim command in a
4795 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4796 was being done only for the command run by the queryprogram router. It is
4797 now done for all such subprocesses. The other cases are: ${run, transport
4798 filters, and the commands run by the lmtp and pipe transports.
4800 5. Some older OS have a limit of 256 on the maximum number of file
4801 descriptors. Exim was using setrlimit() to set 1000 as a large value
4802 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4803 systems. I've change it so that if it can't get 1000, it tries for 256.
4805 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4806 was an oversight, and furthermore, ever since the addition of extra
4807 controls (e.g. 4.43/32), the checks on when to allow different forms of
4808 "control" were broken. There should now be diagnostics for all cases when a
4809 control that does not make sense is encountered.
4811 7. $recipients is now available in the predata ACL (oversight).
4813 8. Tidy the search cache before the fork to do a delivery from a message
4814 received from the command line. Otherwise the child will trigger a lookup
4815 failure and thereby defer the delivery if it tries to use (for example) a
4816 cached ldap connection that the parent has called unbind on.
4818 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4819 of $address_data from the recipient verification was clobbered by the
4820 sender verification.
4822 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4823 was its contents. (It was OK if the option was not defined at all.)
4825 11. A "Completed" log line is now written for messages that are removed from
4826 the spool by the -Mrm option.
4828 12. $host_address is now set to the target address during the checking of
4829 ignore_target_hosts.
4831 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4832 being passed; this would have caused $sender_host_name to have been used if
4833 matching the list had actually called for a host name (not very likely,
4834 since this list is usually IP addresses). A host name is now passed as
4837 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4838 code that specifies a non-restarting handler (typically sigaction() in
4839 modern systems) in an attempt to fix a rare and obscure crash bug.
4841 15. Narrowed the window for a race in the daemon that could cause it to ignore
4842 SIGCHLD signals. This is not a major problem, because they are used only to
4843 wake it up if nothing else does.
4845 16. A malformed maildirsize file could cause Exim to calculate negative values
4846 for the mailbox size or file count. Odd effects could occur as a result.
4847 The maildirsize information is now recalculated if the size or filecount
4850 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4851 support for a long time. Removed HAVE_SYS_VFS_H.
4853 18. Updated exipick to current release from John Jetmore.
4855 19. Allow an empty sender to be matched against a lookup in an address list.
4856 Previously the only cases considered were a regular expression, or an
4859 20. Exim went into a mad DNS lookup loop when doing a callout where the
4860 host was specified on the transport, if the DNS lookup yielded more than
4863 21. The RFC2047 encoding function was originally intended for short strings
4864 such as real names; it was not keeping to the 75-character limit for
4865 encoded words that the RFC imposes. It now respects the limit, and
4866 generates multiple encoded words if necessary. To be on the safe side, I
4867 have increased the buffer size for the ${rfc2047: expansion operator from
4870 22. Failure to deliver a bounce message always caused it to be frozen, even if
4871 there was an errors_to setting on the router. The errors_to setting is now
4874 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4875 canonical form (fully expanded) before being placed in
4876 $sender_host_address.
4878 24. Updated eximstats to version 1.33
4880 25. Include certificate and key file names in error message when GnuTLS fails
4881 to set them up, because the GnuTLS error message doesn't include the name
4882 of the failing file when there is a problem reading it.
4884 26. Expand error message when OpenSSL has problems setting up cert/key files.
4887 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4888 (this can affect the format of dates).
4890 28. exim_tidydb, when checking for the continued existence of a message for
4891 which it has found a message-specific retry record, was not finding
4892 messages that were in split spool directories. Consequently, it was
4893 deleting retry records that should have stayed in existence.
4895 29. eximstats updated to version 1.35
4896 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4897 1.35 - bugfix such that pie charts by volume are generated correctly
4899 30. The SPA authentication driver was not abandoning authentication and moving
4900 on to the next authenticator when an expansion was forced to fail,
4901 contradicting the general specification for all authenticators. Instead it
4902 was generating a temporary error. It now behaves as specified.
4904 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4905 (the order specifies the preference for clients). The order is now AES256,
4906 AES128, 3DES, ARCFOUR128.
4908 31. Small patch to Sieve code - explicitly set From: when generating an
4911 32. Exim crashed if a remote delivery caused a very long error message to be
4912 recorded - for instance if somebody sent an entire SpamAssassin report back
4913 as a large number of 550 error lines. This bug was coincidentally fixed by
4914 increasing the size of one of Exim's internal buffers (big_buffer) that
4915 happened as part of the Exiscan merge. However, to be on the safe side, I
4916 have made the code more robust (and fixed the comments that describe what
4919 33. Some experimental protocols are using DNS PTR records for new purposes. The
4920 keys for these records are domain names, not reversed IP addresses. The
4921 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4922 leaves it alone. Component reversal etc. now happens only for IP addresses.
4925 34. The host_aton() function is supposed to be passed a string that is known
4926 to be a valid IP address. However, in the case of IPv6 addresses, it was
4927 not checking this. This is a hostage to fortune. Exim now panics and dies
4928 if the condition is not met. A case was found where this could be provoked
4929 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4930 components; fortuitously, this particular loophole had already been fixed
4931 by change 4.50/55 or 4.44/33 above.
4933 If there are any other similar loopholes, the new check in host_aton()
4934 itself should stop them being exploited. The report I received stated that
4935 data on the command line could provoke the exploit when Exim was running as
4936 exim, but did not say which command line option was involved. All I could
4937 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4938 running as the user.
4941 35. There was a buffer overflow vulnerability in the SPA authentication code
4942 (which came originally from the Samba project). I have added a test to the
4943 spa_base64_to_bits() function which I hope fixes it.
4946 36. The daemon start-up calls getloadavg() while still root for those OS that
4947 need the first call to be done as root, but it missed one case: when
4948 deliver_queue_load_max is set with deliver_drop_privilege. This is
4949 necessary for the benefit of the queue runner, because there is no re-exec
4950 when deliver_drop_privilege is set.
4952 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4953 was in use, was not putting the data itself into the right store pool;
4954 consequently, it could be overwritten for a subsequent message in the same
4955 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4958 38. Sometimes the final signoff response after QUIT could fail to get
4959 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4960 < 0 before doing a fflush(). This bug looks as though it goes back to the
4961 introduction of TLS in release 3.20, but "sometimes" must have been rare
4962 because the tests only now provoked it.
4968 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4969 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4970 Now it returns an integer. A number of calls were still expecting a T/F
4971 return. Fortuitously, in all cases, the tests worked in OK situations,
4972 which is the norm. However, things would have gone wrong on any write
4973 failures on the smtp file descriptor. This function is used when sending
4974 messages over SMTP and also when doing verify callouts.
4976 2. When Exim is called to do synchronous delivery of a locally submitted
4977 message (the -odf or -odi options), it no longer closes stderr before doing
4980 3. Implemented the mua_wrapper option.
4982 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4984 5. Implemented the functions header_remove(), header_testname(),
4985 header_add_at_position(), and receive_remove_recipient(), and exported them
4988 6. If an ACL "warn" statement specified the addition of headers, Exim already
4989 inserted X-ACL-Warn: at the start if there was no header name. However, it
4990 was not making this test for the second and subsequent header lines if
4991 there were newlines in the string. This meant that an invalid header could
4992 be inserted if Exim was badly configured.
4994 7. Allow an ACL "warn" statement to add header lines at the start or after all
4995 the Received: headers, as well as at the end.
4997 8. Added the rcpt_4xx retry error code.
4999 9. Added postmaster_mailfrom=xxx to callout verification option.
5001 10. Added mailfrom=xxxx to the callout verification option, for verify=
5004 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5005 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5007 12. Inserted some casts to stop certain compilers warning when using pointer
5008 differences as field lengths or precisions in printf-type calls (mostly
5009 affecting debugging statements).
5011 13. Added optional readline() support for -be (dynamically loaded).
5013 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5014 same clock tick as a message's arrival, so that its received time was the
5015 same as the "first fail" time on the retry record, and that message
5016 remained on the queue past the ultimate address timeout, every queue runner
5017 would try a delivery (because it was past the ultimate address timeout) but
5018 after another failure, the ultimate address timeout, which should have then
5019 bounced the address, did not kick in. This was a "< instead of <=" error;
5020 in most cases the first failure would have been in the next clock tick
5021 after the received time, and all would be well.
5023 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5024 being recognized when the domain list was tested by the match_domain
5025 condition in an expansion string.
5027 16. Added the ${str2b64: operator.
5029 17. Exim was always calling setrlimit() to set a large limit for the number of
5030 processes, without checking whether the existing limit was already
5031 adequate. (It did check for the limit on file descriptors.) Furthermore,
5032 errors from getrlimit() and setrlimit() were being ignored. Now they are
5033 logged to the main and panic logs, but Exim does carry on, to try to do its
5034 job under whatever limits there are.
5036 18. Imported PCRE 5.0.
5038 19. Trivial typo in log message " temporarily refused connection" (the leading
5041 20. If the log selector return_path_on_delivery was set and an address was
5042 redirected to /dev/null, the delivery process crashed because it assumed
5043 that a return path would always be set for a "successful" delivery. In this
5044 case, the whole delivery is bypassed as an optimization, and therefore no
5047 21. Internal re-arrangement: the function for sending a challenge and reading
5048 a response while authentication was assuming a zero-terminated challenge
5049 string. It's now changed to take a pointer and a length, to allow for
5050 binary data in such strings.
5052 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5054 23. Exim was not respecting finduser_retries when seeking the login of the
5055 uid under which it was called; it was always trying 10 times. (The default
5056 setting of finduser_retries is zero.) Also, it was sleeping after the final
5057 failure, which is pointless.
5059 24. Implemented tls_on_connect_ports.
5061 25. Implemented acl_smtp_predata.
5063 26. If the domain in control=submission is set empty, Exim assumes that the
5064 authenticated id is a complete email address when it generates From: or
5065 Sender: header lines.
5067 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5068 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5069 chown and chgrp in /bin and hostname in /usr/bin.
5071 28. Exim was keeping the "process log" file open after each use, just as it
5072 does for the main log. This opens the possibility of it remaining open for
5073 long periods when the USR1 signal hits a daemon. Occasional processlog
5074 errors were reported, that could have been caused by this. Anyway, it seems
5075 much more sensible not to leave this file open at all, so that is what now
5078 29. The long-running daemon process does not normally write to the log once it
5079 has entered its main loop, and it closes the log before doing so. This is
5080 so that log files can straightforwardly be renamed and moved. However,
5081 there are a couple of unusual error situations where the daemon does write
5082 log entries, and I had neglected to close the log afterwards.
5084 30. The text of an SMTP error response that was received during a remote
5085 delivery was being truncated at 512 bytes. This is too short for some of
5086 the long messages that one sometimes sees. I've increased the limit to
5089 31. It is now possible to make retry rules that apply only when a message has a
5090 specific sender, in particular, an empty sender.
5092 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5093 it possible to be selective about when SMTP synchronization is enforced.
5095 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5097 32. Implemented hosts_connection_nolog.
5099 33. Added an ACL for QUIT.
5101 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5104 35. Added mailbox_size and mailbox_filecount to appendfile.
5106 36. Added control = no_multiline_responses to ACLs.
5108 37. There was a bug in the logic of the code that waits for the clock to tick
5109 in the case where the clock went backwards by a substantial amount such
5110 that the microsecond fraction of "now" was more than the microsecond
5111 fraction of "then" (but the whole seconds number was less).
5113 38. Added support for the libradius Radius client library this is found on
5114 FreeBSD (previously only the radiusclient library was supported).
5120 1. When certain lookups returned multiple values in the form name=value, the
5121 quoting of the values was not always being done properly. Specifically:
5122 (a) If the value started with a double quote, but contained no whitespace,
5124 (b) If the value contained whitespace other than a space character (i.e.
5125 tabs or newlines or carriage returns) it was not quoted.
5126 This fix has been applied to the mysql and pgsql lookups by writing a
5127 separate quoting function and calling it from the lookup code. The fix
5128 should probably also be applied to nisplus, ibase and oracle lookups, but
5129 since I cannot test any of those, I have not disturbed their existing code.
5131 2. A hit in the callout cache for a specific address caused a log line with no
5132 reason for rejecting RCPT. Now it says "Previous (cached) callout
5133 verification failure".
5135 3. There was an off-by-one bug in the queryprogram router. An over-long
5136 return line was truncated at 256 instead of 255 characters, thereby
5137 overflowing its buffer with the terminating zero. As well as fixing this, I
5138 have increased the buffer size to 1024 (and made a note to document this).
5140 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5141 when Exim is waiting for an SMTP response from a remote server, Exim
5142 restarts its select() call on the socket, thereby resetting its timeout.
5143 This is not a problem when such interrupts are rare. Somebody set up a cron
5144 job to run exiwhat every 2 minutes, which is less than the normal select()
5145 timeout (5 or 10 minutes). This meant that the select() timeout never
5146 kicked in because it was always reset. I have fixed this by comparing the
5147 time when an interrupt arrives with the time at the start of the first call
5148 to select(). If more time than the timeout has elapsed, the interrupt is
5149 treated as a timeout.
5151 5. Some internal re-factoring in preparation for the addition of Sieve
5152 extensions (by MH). In particular, the "personal" test is moved to a
5153 separate function, and given an option for scanning Cc: and Bcc: (which is
5154 not set for Exim filters).
5156 6. When Exim created an email address using the login of the caller as the
5157 local part (e.g. when creating a From: or Sender: header line), it was not
5158 quoting the local part when it contained special characters such as @.
5160 7. Installed new OpenBSD configuration files.
5162 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5163 try to make them clearer.
5165 9. Callout options, other than the timeout value, were being ignored when
5166 verifying sender addresses in header lines. For example, when using
5168 verify = header_sender/callout=no_cache
5170 the cache was (incorrectly) being used.
5172 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5173 only the Cygwin environment.
5175 11. When return_path_on_delivery was set as a log selector, if different remote
5176 addresses in the same message used different return paths and parallel
5177 remote delivery occurred, the wrong values would sometimes be logged.
5178 (Whenever a remote delivery process finished, the return path value from
5179 the most recently started remote delivery process was logged.)
5181 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5182 lines when AUTH and/or TLS are in use. This is the "received protocol"
5183 field. Exim used to use "asmtp" for authenticated SMTP, without any
5184 indication (in the protocol name) for TLS use. Now it follows the RFC and
5185 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5186 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5187 names appear in log lines as well as in Received: header lines.
5189 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5190 extensions, and comparison tests, and to fix some bugs.
5192 14. Changes to the "personal" filter test:
5194 (1) The test was buggy in that it was just doing the equivalent of
5195 "contains" tests on header lines. For example, if a user's address was
5196 anne@some.where, the "personal" test would incorrectly be true for
5198 To: susanne@some.where
5200 This test is now done by extracting each address from the header in turn,
5201 and checking the entire address. Other tests that are part of "personal"
5202 are now done using regular expressions (for example, to check local parts
5203 of addresses in From: header lines).
5205 (2) The list of non-personal local parts in From: addresses has been
5206 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5207 taken from the Sieve specification recommendations.
5209 (3) If the message contains any header line starting with "List-" it is
5210 treated as non-personal.
5212 (4) The test for "circular" in the Subject: header line has been removed
5213 because it now seems ill-conceived.
5215 15. Minor typos in src/EDITME comments corrected.
5217 16. Installed latest exipick from John Jetmore.
5219 17. If headers_add on a router specified a text string that was too long for
5220 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5221 of string_sprintf() is now avoided.
5223 18. $message_body_size was not set (it was always zero) when running the DATA
5224 ACL and the local_scan() function.
5226 19. For the "mail" command in an Exim filter, no default was being set for
5227 the once_repeat time, causing a random time value to be used if "once" was
5228 specified. (If the value happened to be <= 0, no repeat happened.) The
5229 default is now 0s, meaning "never repeat". The "vacation" command was OK
5230 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5231 (I found it when inspecting the code).
5233 20. There is now an overall timeout for performing a callout verification. It
5234 defaults to 4 times the callout timeout, which applies to individual SMTP
5235 commands during the callout. The overall timeout applies when there is more
5236 than one host that can be tried. The timeout is checked before trying the
5237 next host. This prevents very long delays if there are a large number of
5238 hosts and all are timing out (e.g. when the network connections are timing
5239 out). The value of the overall timeout can be changed by specifying an
5240 additional sub-option for "callout", called "maxwait". For example:
5242 verify = sender/callout=5s,maxwait=20s
5244 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5245 to the end before writing, but this should make it even safer).
5247 22. Exim was forgetting that it had advertised PIPELINING for the second and
5248 subsequent messages on an SMTP connection. It was also not resetting its
5249 memory on STARTTLS and an internal HELO.
5251 23. When Exim logs an SMTP synchronization error within a session, it now
5252 records whether PIPELINING has been advertised or not.
5254 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5255 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5256 rather than long int.
5258 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5260 26. Added the never_mail option to autoreply.
5266 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5267 crash if the getsockname() call failed; this can happen if a connection is
5268 closed very soon after it is established. The problem was simply in the
5269 order in which certain operations were done, causing Exim to try to write
5270 to the SMTP stream before it had set up the file descriptor. The bug has
5271 been fixed by making things happen in the correct order.
5277 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5278 before the connection was closed, thus losing the rejection response.
5280 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5281 some early Solaris releases, but causes trouble in current releases where
5282 socklen_t is defined.
5284 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5287 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5290 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5291 that releasing the top of it at the end releases what was used for sub-
5292 expansions (unless the block got too big). However, discard this block if
5293 the first thing is a variable or header, so that we can use its block when
5294 it is dynamic (useful for very large $message_headers, for example).
5296 6. Lookups now cache *every* query, not just the most recent. A new, separate
5297 store pool is used for this. It can be recovered when all lookup caches are
5298 flushed. Lookups now release memory at the end of their result strings.
5299 This has involved some general refactoring of the lookup sources.
5301 7. Some code has been added to the store_xxx() functions to reduce the amount
5302 of flapping under certain conditions.
5304 8. log_incoming_interface used to affect only the <= reception log lines. Now
5305 it causes the local interface and port to be added to several more SMTP log
5306 lines, for example "SMTP connection from", and rejection lines.
5308 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5310 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5312 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5313 could be overwritten at the end of the current message (or the start of a
5314 new message if it was set in a HELO ACL). The value is now preserved for
5315 the duration of the SMTP connection.
5317 12. If a transport had a headers_rewrite setting, and a matching header line
5318 contained an unqualified address, that address was qualified, even if it
5319 did not match any rewriting rules. The underlying bug was that the values
5320 of the flags that permit the existence of unqualified sender and recipient
5321 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5322 non-local messages, and by -bnq for local messages) were not being
5323 preserved with the message after it was received.
5325 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5326 "next input=" as part of the text comprising the host identity instead of
5327 the correct text. The code was using the same buffer for two different
5328 strings. However, depending on which order the printing function evaluated
5329 its arguments, the bug did not always show up. Under Linux, for example, my
5330 test suite worked just fine.
5332 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5333 doesn't work with some older versions of Perl. It has been changed to "my",
5334 which in any case is probably the better facility to use.
5336 15. A really picky compiler found some instances of statements for creating
5337 error messages that either had too many or two few arguments for the format
5340 16. The size of the buffer for calls to the DNS resolver has been increased
5341 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5342 for addresses that have a lot of PTR records. This alleviates a problem; it
5343 does not fully solve it.
5345 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5346 buffer now truncates the list and logs the incident, which is the same
5347 action as happens when Exim is looking up a host name and its aliases.
5348 Previously in this situation something unpredictable would happen;
5349 sometimes it was "internal error: store_reset failed".
5351 18. If a server dropped the connection unexpectedly when an Exim client was
5352 using GnuTLS and trying to read a response, the client delivery process
5353 crashed while trying to generate an error log message.
5355 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5356 string, for example:
5358 warn message = H1: something\nH2: something
5360 the text was added as a single header line from Exim's point of view
5361 though it ended up OK in the delivered message. However, searching for the
5362 second and subsequent header lines using $h_h2: did not work. This has been
5363 fixed. Similarly, if a system filter added multiple headers in this way,
5364 the routers could not see them.
5366 20. Expanded the error message when iplsearch is called with an invalid key to
5367 suggest using net-iplsearch in a host list.
5369 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5370 ACLs are no longer actually imposed (and a message to that effect is
5373 22. If a "gecos" field in a passwd entry contained escaped characters, in
5374 particular, if it contained a \" sequence, Exim got it wrong when building
5375 a From: or a Sender: header from that name. A second bug also caused
5376 incorrect handling when an unquoted " was present following a character
5377 that needed quoting.
5379 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5380 was not being matched caselessly.
5382 24. Arranged for all hyphens in the exim.8 source to be escaped with
5385 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5386 properly. Recipient callout cache records were still being keyed to include
5387 the sender, even when use_sender was set false. This led to far more
5388 callouts that were necessary. The sender is no longer included in the key
5389 when use_sender is false.
5391 26. Added "control = submission" modifier to ACLs.
5393 27. Added the ${base62d: operator to decode base 62 numbers.
5395 28. dnsdb lookups can now access SRV records.
5397 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5398 the configuration file.
5400 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5401 (-v) mode. This makes the output for a verbose queue run more intelligible.
5403 31. Added a use_postmaster feature to recipient callouts.
5405 32. Added the $body_zerocount variable, containing the number of binary zero
5406 bytes in the message body.
5408 33. The time of last modification of the "new" subdirectory is now used as the
5409 "mailbox time last read" when there is a quota error for a maildir
5412 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5414 35. Added +ignore_unknown as a special item in host lists.
5416 36. Code for decoding IPv6 addresses in host lists is now included, even if
5417 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5418 address was recognized as an IP address, but was then not correctly decoded
5419 into binary, causing unexpected and incorrect effects when compared with
5426 1. Very minor rewording of debugging text in manualroute to say "list of
5427 hosts" instead of "hostlist".
5429 2. If verify=header_syntax was set, and a header line with an unqualified
5430 address (no domain) and a large number of spaces between the end of the
5431 name and the colon was received, the reception process suffered a buffer
5432 overflow, and (when I tested it) crashed. This was caused by some obsolete
5433 code that should have been removed. The fix is to remove it!
5435 3. When running in the test harness, delay a bit after writing a bounce
5436 message to get a bit more predictability in the log output.
5438 4. Added a call to search_tidyup() just before forking a reception process. In
5439 theory, someone could use a lookup in the expansion of smtp_accept_max_
5440 per_host which, without the tidyup, could leave open a database connection.
5442 5. Added the variables $recipient_data and $sender_data which get set from a
5443 lookup success in an ACL "recipients" or "senders" condition, or a router
5444 "senders" option, similar to $domain_data and $local_part_data.
5446 6. Moved the writing of debug_print from before to after the "senders" test
5449 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5450 problems for message scanning, either using a data ACL, or using
5451 local_scan() because the Received: header was not generated till after they
5452 were called (in order to set the time as the time of reception completion).
5453 I have revised the way this works. The header is now generated after the
5454 body is received, but before the ACL or local_scan() are called. After they
5455 are run, the timestamp in the header is updated.
5461 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5462 before starting a queue runner without re-exec. This happened only when
5463 deliver_drop_privilege was set or when the Exim user was set to root. The
5464 effect of the bug was that timeouts during subsequent deliveries caused
5465 crashes instead of being properly handled. The handler is now left at its
5466 default (and expected) setting.
5468 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5469 message, again when deliver_drop_privilege is set or Exim is run as root.
5470 The bug described in (1) was not present in this case, but the tidying up
5471 of the other signals was missing. I have made the two cases consistent.
5473 3. The ignore_target_hosts setting on a manualroute router was being ignored
5474 for hosts that were looked up using the /MX notation.
5476 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5479 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5480 operated on the sender address. After changing the $sender_address to <>
5481 for the sender address verify, Exim was re-instated it as the original
5482 (before rewriting) address, but remembering that it had rewritten it, so it
5483 wasn't rewriting it again. This bug also had the effect of breaking the
5484 sender address verification caching when the sender address was rewritten.
5486 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5487 This has been changed so that if the ip literal address matches
5488 ignore_target_hosts, the router declines.
5490 7. Added expansion conditions match_domain, match_address, and match_local_
5491 part (NOT match_host).
5493 8. The placeholder for the Received: header didn't have a length field set.
5495 9. Added code to Exim itself and to exim_lock to test for a specific race
5496 condition that could lead to file corruption when using MBX delivery. The
5497 issue is with the lockfile that is created in /tmp. If this file is removed
5498 after a process has opened it but before that process has acquired a lock,
5499 there is the potential for a second process to recreate the file and also
5500 acquire a lock. This could lead to two Exim processes writing to the file
5501 at the same time. The added code performs the same test as UW imapd; it
5502 checks after acquiring the lock that its file descriptor still refers to
5503 the same named file.
5505 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5506 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5509 11. Added the smtp_active_hostname option. If used, this will typically be made
5510 to depend on the incoming interface address. Because $interface_address is
5511 not set up until the daemon has forked a reception process, error responses
5512 that can happen earlier (such as "too many connections") no longer contain
5515 12. If an expansion in a condition on a "warn" statement fails because a lookup
5516 defers, the "warn" statement is abandoned, and the next ACL statement is
5517 processed. Previously this caused the whole ACL to be aborted.
5519 13. Added the iplsearch lookup type.
5521 14. Added ident_timeout as a log selector.
5523 15. Added tls_certificate_verified as a log selector.
5525 16. Added a global option tls_require_ciphers (compare the smtp transport
5526 option of the same name). This controls incoming TLS connections.
5528 17. I finally figured out how to make tls_require_ciphers do a similar thing
5529 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5530 before starting the TLS session.
5532 18. Tabs are now shown as \t in -bP output.
5534 19. If the log selector return_path_on_delivery was set, Exim crashed when
5535 bouncing a message because it had too many Received: header lines.
5537 20. If two routers both had headers_remove settings, and the first one included
5538 a superfluous trailing colon, the final name in the first list and the
5539 first name in the second list were incorrectly joined into one item (with a
5540 colon in the middle).
5546 1. Added -C and -D options to the exinext utility, mainly to make it easier
5547 to include in the automated testing, but these could be helpful when
5548 multiple configurations are in use.
5550 2. The exinext utility was not formatting the output nicely when there was
5551 an alternate port involved in the retry record key, nor when there was a
5552 message id as well (for retries that were specific to a specific message
5553 and a specific host). It was also confused by IPv6 addresses, because of
5554 the additional colons they contain. I have fixed the IPv4 problem, and
5555 patched it up to do a reasonable job for IPv6.
5557 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5558 delivery, the log line now contains "pipelined" if PIPELINING was used.
5560 4. An SMTP transport process used to panic and die if the bind() call to set
5561 an explicit outgoing interface failed. This has been changed; it is now
5562 treated in the same way as a connect() failure.
5564 5. A reference to $sender_host_name in the part of a conditional expansion
5565 that was being skipped was still causing a DNS lookup. This no longer
5568 6. The def: expansion condition was not recognizing references to header lines
5569 that used bh_ and bheader_.
5571 7. Added the _cache feature to named lists.
5573 8. The code for checking quota_filecount in the appendfile transport was
5574 allowing one more file than it should have been.
5576 9. For compatibility with Sendmail, the command line option
5584 and sets the incoming protocol and host name (for trusted callers). The
5585 host name and its colon can be omitted when only the protocol is to be set.
5586 Note the Exim already has two private options, -pd and -ps, that refer to
5587 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5588 "s", but I don't think that's a major issue.
5590 10. A number of refactoring changes to the code, none of which should affect
5593 (a) The number of logging options was getting close to filling up the
5594 32-bit word that was used as a bit map. I have split them into two classes:
5595 those that are passed in the argument to log_write(), and those that are
5596 only ever tested independently outside of that function. These are now in
5597 separate 32-bit words, so there is plenty of room for expansion again.
5598 There is no change in the user interface or the logging behaviour.
5600 (b) When building, for example, log lines, the code previously used a
5601 macro that called string_cat() twice, in order to add two strings. This is
5602 not really sufficiently general. Furthermore, there was one instance where
5603 it was actually wrong because one of the argument was used twice, and in
5604 one call a function was used. (As it happened, calling the function twice
5605 did not affect the overall behaviour.) The macro has been replaced by a
5606 function that can join an arbitrary number of extra strings onto a growing
5609 (c) The code for expansion conditions now uses a table and a binary chop
5610 instead of a serial search (which was left over from when there were very
5611 few conditions). Also, it now recognizes conditions like "pam" even when
5612 the relevant support is not compiled in: a suitably worded error message is
5613 given if an attempt is made to use such a condition.
5615 11. Added ${time_interval:xxxxx}.
5617 12. A bug was causing one of the ddress fields not to be passed back correctly
5618 from remote delivery subprocesses. The field in question was not being
5619 subsequently used, so this caused to problems in practice.
5621 13. Added new log selectors queue_time and deliver_time.
5623 14. Might have fixed a bug in maildirsizefile handling that threw up
5624 "unexpected character" debug warnings, and recalculated the data
5625 unnecessarily. In any case, I expanded the warning message to give more
5628 15. Added the message "Restricted characters in address" to the statements in
5629 the default ACL that block characters like @ and % in local parts.
5631 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5632 Three changes have been made:
5634 (a) There was a serious bug; a negative response to MAIL caused the whole
5635 recipient domain to be cached as invalid, thereby blocking all messages
5636 to all local parts at the same domain, from all senders. This bug has
5637 been fixed. The domain is no longer cached after a negative response to
5638 MAIL if the sender used is not empty.
5640 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5643 (c) A new callout option, "use_sender" has been added for people who want
5644 the modified behaviour.
5650 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5653 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5654 indeed breaks things for older releases.
5656 3. Added additional logging to the case where there is a problem reading data
5657 from a filter that is running in a subprocess using a pipe, in order to
5658 try to track down a specific problem.
5660 4. Testing facility fudge: when running in the test harness and attempting
5661 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5662 getting "No route to host". Convert this to a timeout.
5664 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5667 6. Some OS don't have socklen_t but use size_t instead. This affects the
5668 fifth argument of getsockopt() amongst other things. This is now
5669 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5670 can be set for individual OS. I have set it for SunOS5, OSF1, and
5671 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5672 some earlier ones do not.
5674 7. Change 4.30/15 was not doing the test caselessly.
5676 8. The standard form for an IPv6 address literal was being rejected by address
5677 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5678 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5679 this, as well as the form without the "IPv6" on the front (but only when
5680 address literals are enabled, of course).
5682 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5684 10. Exim crashed if a message with an empty sender address specified by -f
5685 encountered a router with an errors_to setting. This could be provoked only
5686 by a command such as
5690 where an empty string was supplied; "<>" did not hit this bug.
5692 11. Installed PCRE release 4.5.
5694 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5695 remained set. It is now erased.
5697 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5698 times from message ids (which are base 36 rather than the normal 62).
5700 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5701 were being counted as actual protocol errors, and logged if the log
5702 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5703 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5704 and DATA following a set of rejected RCPTs do not count as protocol errors.
5705 In other words, Exim assumes they were pipelined, though this may not
5706 actually be the case. Of course, in all cases the client gets an
5707 appropriate error code.
5709 15. If a lookup fails in an ACL condition, a message about the failure may
5710 be available; it is used if testing the ACL cannot continue, because most
5711 such messages specify what the cause of the deferral is. However, some
5712 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5713 that caused an old message to be retained and used if a later statement
5714 caused a defer, replacing the real cause of the deferral.
5716 16. If an IP address had so many PTR records that the DNS lookup buffer
5717 was not large enough to hold them, Exim could crash while trying to process
5718 the truncated data. It now detects and logs this case.
5720 17. Further to 4.21/58, another change has been made: if (and only if) the
5721 first line of a message (the first header line) ends with CRLF, a bare LF
5722 in a subsequent header line has a space inserted after it, so as not to
5723 terminate the header.
5725 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5726 unnecessarily, used atoi() instead of strtol(), and didn't check the
5727 termination when getting file sizes from file names by regex.
5729 19. Completely re-implemented the support for maildirsize files, in the light
5730 of a number of problems with the previous contributed implementation
5731 (4.30/29). In particular:
5733 . If the quota is zero, the maildirsize file is maintained, but no quota is
5736 . If the maildir directory does not exist, it is created before any attempt
5737 to write a maildirsize file.
5739 . The quota value in the file is just a cache; if the quota is changed in
5740 the transport, the new value overrides.
5742 . A regular expression is available for excluding directories from the
5745 20. The autoreply transport checks the characters in options that define the
5746 message's headers; it allows continued headers, but it was checking with
5747 isspace() after an embedded newline instead of explicitly looking for a
5750 21. If all the "regular" hosts to which an address was routed had passed their
5751 expiry times, and had not reached their retry times, the address was
5752 bounced, even if fallback hosts were defined. Now Exim should go on to try
5755 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5756 equivalent code in the SMTP transport. Some hosts send humungous responses
5757 to HELO/EHLO, more than 1024 it seems.
5759 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5760 gives compiler warnings in some environments. I've now done it "properly",
5763 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5764 (because of problems with the built-in one) was declared to return uschar *
5765 instead of char *, causing compiler failure.
5767 25. Fixed a file descriptor leak when processing alias/forward files.
5769 26. Fixed a minor format string issue in dbfn.c.
5771 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5773 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5774 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5777 29. When a sender address is verified, it is cached, to save repeating the test
5778 when there is more than one recipient in a message. However, when the
5779 verification involves a callout, it is possible for different callout
5780 options to be set for different recipients. It is too complicated to keep
5781 track of this in the cache, so now Exim always runs a verification when a
5782 callout is required, relying on the callout cache for the optimization.
5783 The overhead is duplication of the address routing, but this should not be
5786 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5787 to be verified with callout=postmaster, and the main callout worked but the
5788 postmaster check failed, the verification correctly failed. However, if a
5789 subsequent RCPT command asked for sender verification *without* the
5790 postmaster check, incorrect caching caused this verification also to fail,
5793 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5794 it was not caching the DNS options (qualify_single, search_parents) that
5795 were used when the lookup failed. A subsequent lookup with different
5796 options therefore always gave the same answer, though there were cases
5797 where it should not have. (Example: a "domains = !$mx_any" option on a
5798 dnslookup router: the "domains" option is always processed without any
5799 widening, but the router might have qualify_single set.) Now Exim uses the
5800 cached value only when the same options are set.
5802 32. Added John Jetmore's "exipick" utility to the distribution.
5804 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5805 than a timeout (e.g. a certificate is required, and is not provided), an
5806 Exim server now closes the connection immediately. Previously it waited for
5807 the client to close - but if the client is SSL, it seems that they each
5808 wait for each other, leading to a delay before one of them times out.
5810 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5811 maintained 0.8.x compatibility because I don't think many are using it, and
5812 it is clearly obsolete.
5814 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5817 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5818 client certificate was expired. A simple patch fixes this, though I don't
5819 understand the full logic of why the verify callback is called multiple
5822 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5823 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5824 which causes problems with some clients (such as the Certicom SSL Plus
5825 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5826 disables the coutermeasure allowing Eudora to connect."
5828 38. Exim was not checking that a write() to a log file succeeded. This could
5829 lead to Bad Things if a log got too big, in particular if it hit a file
5830 size limit. Exim now panics and dies if it cannot write to a log file, just
5831 as it does if it cannot open a log file.
5833 39. Modified OS/Makefile-Linux so that it now contains
5835 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5837 The two -D definitions ensure that Exim is compiled with large file
5838 support, which makes it possible to handle log files that are bigger than
5841 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5842 instance) a domain was checked against a named list that involved a lookup,
5843 causing $domain_data to be set, then another domain was checked against the
5844 same list, then the first domain was re-checked, the value of $domain_data
5845 after the final check could be wrong. In particular, if the second check
5846 failed, it could be set empty. This bug probably also applied to
5849 41. The strip_trailing_dot option was not being applied to the address given
5850 with the -f command-line option.
5852 42. The code for reading a message's header from the spool was incrementing
5853 $received_count, but never initializing it. This meant that the value was
5854 incorrect (doubled) while delivering a message in the same process in which
5855 it was received. In the most common configuration of Exim, this never
5856 happens - a fresh exec is done - but it can happen when
5857 deliver_drop_privilege is set.
5859 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5860 it now includes up to 150 characters of the unexpected data in the log
5863 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5864 and building data strings. The size of both of these buffers was 10 000
5865 bytes - far larger than anybody would *ever* want, thought I. Needless to
5866 say, somebody hit the limit. I have increased the maximum line length to
5867 20 000 and the maximum data length of concatenated lines to 100 000. I have
5868 also fixed two bugs, because there was no checking on these buffers. Tsk,
5869 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5870 buffer is too small.
5872 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5873 lsearch lookups. Now it does.
5875 46. When parsing a route_list item in a manualroute router, a fixed-length
5876 buffer was used for the list of hosts. I made this 1024 bytes long,
5877 thinking that nobody would ever have a list of hosts that long. Wrong.
5878 Somebody had a whole pile of complicated expansion conditions, and the
5879 string was silently truncated, leading to an expansion error. It turns out
5880 that it is easier to change to an unlimited length (owing to other changes
5881 that have happened since this code was originally written) than to build
5882 structure for giving a limitation error. The length of the item that
5883 expands into the list of hosts is now unlimited.
5885 47. The lsearch lookup could not handle data where the length of text line was
5886 more than 4095 characters. Such lines were truncated, leading to shortened
5887 data being returned. It should now handle lines of any length.
5889 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5890 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5893 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5894 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5895 no longer happens in comment lines. A list of replacements is now placed
5896 at the head of all of the source files, except those whose only change is
5897 to replace PERL_COMMAND in the very first #! line.
5899 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5900 messages on the queue, with a bottom-up merge sort, using code contributed
5901 by Michael Haardt. This should make operations like -bp somewhat faster on
5902 large queues. It won't affect queue runners, except when queue_run_in_order
5905 51. Installed eximstats 1.31 in the distribution.
5907 52. Added support for SRV lookups to the dnslookup router.
5909 53. If an ACL referred to $message_body or $message_body_end, the value was not
5910 reset for any messages that followed in the same SMTP session.
5912 54. The store-handling optimization for building very long strings was not
5913 differentiating between the different store pools. I don't think this
5914 actually made any difference in practice, but I've tidied it.
5916 55. While running the routers to verify a sender address, $sender_address
5917 was still set to the sender address. This is wrong, because when routing to
5918 send a bounce to the sender, it would be empty. Therefore, I have changed
5919 it so that, while verifying a sender address, $sender_address is set to <>.
5920 (There is no change to what happens when verifying a recipient address.)
5922 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5923 target A or AAAA records (if not already returned) without resetting the
5924 qualify_single or search_parents options of the DNS resolver. These are
5925 inappropriate in this case because the targets of MX and SRV records must
5926 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5927 target that, when qualified, matched something in the local domain. These
5928 two options are now turned off when doing these lookups.
5930 57. It seems that at least some releases of Reiserfs (which does not have the
5931 concept of a fixed number of inodes) returns zero and not -1 for the
5932 number of available inodes. This interacted badly with check_spool_inodes,
5933 which assumed that -1 was the "no such thing" setting. What I have done is
5934 to check that the total number of inodes is greater than zero before doing
5935 the test of how many are available.
5937 58. When a "warn" ACL statement has a log_message modifier, the message is
5938 remembered, and not repeated. This is to avoid a lot of repetition when a
5939 message has many recipients that cause the same warning to be written.
5940 However, Exim was preserving the list of already written lines for an
5941 entire SMTP session, which doesn't seem right. The memory is now reset if a
5942 new message is started.
5944 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5945 debugging output unless log_rewrite was also set.
5947 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5948 of (int)(handle) when we know that handle contains (void *)(-1).
5950 61. The Exim daemon panic-logs an error return when it closes the incoming
5951 connection. However "connection reset by peer" seems to be common, and
5952 isn't really an error worthy of noting specially, so that particular error
5955 62. When Exim is trying to find all the local interfaces, it used to panic and
5956 die if the ioctl to get the interface flags failed. However, it seems that
5957 on at least one OS (Solaris 9) it is possible to have an interface that is
5958 included in the list of interfaces, but for which you get a failure error
5959 for this call. This happens when the interface is not "plumbed" into a
5960 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5961 failure of the "get flags" call assumes that the interface is down.
5963 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5964 makes life easier for people who are doing arithmetic on fields extracted
5965 from dates, where you often get leading zeros that should not be
5966 interpreted as octal.
5968 64. Added qualify_domain to the redirect router, to override the global
5971 65. If a pathologically long header line contained very many addresses (the
5972 report of this problem mentioned 10 000) and each of them was rewritten,
5973 Exim could use up a very large amount of memory. (It kept on making new
5974 copies of the header line as it rewrote, and never released the old ones.)
5975 At the expense of a bit more processing, the header rewriting function has
5976 been changed so that it no longer eats memory in this way.
5978 66. The generation of the Received: header has been moved from the time that a
5979 message starts to be received, to the time that it finishes. The timestamp
5980 in the Received: header should now be very close to that of the <= log
5981 line. There are two side-effects of this change:
5983 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5984 logged header lines no longer include the local Received: line, because
5985 it has not yet been created. The same applies to a copy of the message
5986 that is returned to a non-SMTP sender when a message is rejected.
5988 (b) When a filter file is tested using -bf, no additional Received: header
5989 is added to the test message. After some thought, I decided that this
5992 This change does not affect the value of $received_for. It is still set
5993 after address rewriting, but before local_scan() is called.
5995 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5997 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5998 gave an unhelpful panic error message, and a defer error. I have managed to
5999 change this behaviour so that it now rejects any supplied certificate,
6000 which seems right, as the list of acceptable certificates is empty.
6002 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6003 gave an unhelpful defer error. I have not managed to make this reject any
6004 supplied certificates, but the error message it gives is "no certificate
6005 supplied", which is not helpful.
6007 70. exigrep's output now also includes lines that are not associated with any
6008 message, but which match the given pattern. Implemented by a patch from
6009 Martin Sluka, which also tidied up the Perl a bit.
6011 71. Recipient callout verification, like sender verification, was using <> in
6012 the MAIL FROM command. This isn't really the right thing, since the actual
6013 sender may affect whether the remote host accepts the recipient or not. I
6014 have changed it to use the actual sender in the callout; this means that
6015 the cache record is now keyed on a recipient/sender pair, not just the
6016 recipient address. There doesn't seem to be a real danger of callout loops,
6017 since a callout by the remote host to check the sender would use <>.
6018 [SEE ABOVE: changed after hitting problems.]
6020 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6021 temporary errors. However, in the case of such a code being given after
6022 the end of a data transmission (i.e. after ".") Exim was failing to write
6023 a retry record for the message. (Yes, there was some broken host that was
6024 actually sending 8xx at this point.)
6026 73. An unknown lookup type in a host list could cause Exim to panic-die when
6027 the list was checked. (An example that provoked this was putting <; in the
6028 middle of a list instead of at the start.) If this happened during a DATA
6029 ACL check, a -D file could be left lying around. This kind of configuration
6030 error no longer causes Exim to die; instead it causes a defer error. The
6031 incident is still logged to the main and panic logs.
6033 74. Buglet left over from Exim 3 conversion. The message "too many messages
6034 in one connection" was written to the rejectlog but not the mainlog, except
6035 when address rewriting (yes!) was being logged.
6037 75. Added write_rejectlog option.
6039 76. When a system filter was run not as root (that is, when system_filter_user
6040 was set), the values of the $n variables were not being returned to the
6041 main process; thus, they were not subsequently available in the $sn
6044 77. Added +return_path_on_delivery log selector.
6046 78. A connection timeout was being treated differently from recipients deferred
6047 when testing hosts_max_try with a message that was older than the host's
6048 retry timeout. (The host should not be counted, thus allowing all hosts to
6049 be tried at least once before bouncing.) This may have been the cause of an
6050 occasionally reported bug whereby a message would remain on the queue
6051 longer than the retry timeout, but would be bounced if a delivery was
6052 forced. I say "may" because I never totally pinned down the problem;
6053 setting up timeout/retry tests is difficult. See also the next item.
6055 79. The ultimate address timeout was not being applied to errors that involved
6056 a combination of host plus message (for example, a timeout on a MAIL
6057 command). When an address resolved to a number of possible hosts, and they
6058 were not all tried for each delivery (e.g. because of hosts_max_try), a
6059 message could remain on the queue longer than the retry timeout.
6061 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6064 81. Fixed an obscure SMTP outgoing bug which required at least the following
6065 conditions: (a) there was another message waiting for the same server;
6066 (b) the server returned 5xx to all RCPT commands in the first message so
6067 that the message was not completed; (c) the server dropped the connection
6068 or gave a negative response to the RSET that Exim sends to abort the
6069 transaction. The observed case was a dropped connection after DATA that had
6070 been sent in pipelining mode. That is, the server had advertised PIPELINING
6071 but was not implementing it correctly. The effect of the bug was incorrect
6072 behaviour, such as trying another host, and this could lead to a crash.
6078 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6079 and daemon.c were passed as pointers to ints; they should have been
6080 pointers to socklen_t variables (which are typically unsigned ints).
6082 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6085 3. Fixed a really odd bug that affected only the testing scheme; patching a
6086 certain fixed string in the binary changed the value of another string that
6087 happened to be identical to the end of the original first string.
6089 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6090 name", it returns that address as the IP address. On some operating
6091 systems (e.g. Solaris), it also passes back the IP address string as the
6092 "host name". However, on others (e.g. Linux), it passes back an empty
6093 string. Exim wasn't checking for this, and was changing the host name to an
6094 empty string, assuming it had been canonicalized.
6096 5. Although rare, it is permitted to have more than one PTR record for a given
6097 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6098 all the names associated with an address, because they do in Solaris.
6099 However, it seems that they do not in Linux for data that comes from the
6100 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6101 I found this out when I moved to a new Linux workstation and tried to run
6102 the Exim test suite.
6104 To get round this problem I have changed the code so that it now does its
6105 own call to the DNS to look up PTR records when searching for a host name.
6106 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6107 addresses that are only in /etc/hosts are still found.
6109 This behaviour is, however, controlled by an option called host_lookup_
6110 order, which defaults to "bydns:byaddr". If people want to use the other
6111 order, or indeed, just use one or the other means of lookup, they can
6112 specify it in this variable.
6114 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6115 some operating systems, this comes back from gethostbyaddr() as an empty
6116 string, and this is what Exim used to test for. However, it seems that in
6117 other systems, "." is yielded. Exim now tests for this case too.
6119 7. The values of check_spool_space and check_log_space are now held internally
6120 as a number of kilobytes instead of an absolute number of bytes. If a
6121 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6122 kilobyte. This means that much larger values can be stored.
6124 8. Exim monitor: an attempt to get the action menu when not actually pointing
6125 at a message produces an empty menu entitled "No message selected". This
6126 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6127 no entries in it ("Shell widget menu has zero width and/or height"). So I
6128 have added a single, blank menu entry in this case.
6130 9. Added ${quote_local_part.
6132 10. MIME decoding is now applied to the contents of Subject: header lines when
6135 11. Now that a reference to $sender_host_address automatically causes a reverse
6136 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6137 host lookup before query-style lookups in lists that might use this
6138 variable. This has therefore been abolished, and the "net-" prefix is no
6139 longer necessary for query-style lookups.
6141 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6142 have been a typo for LFLAGS, so it has been changed.
6144 13. The install script calls Exim with "-C /dev/null" in order to find the
6145 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6146 to be output. However, since Exim outputs its version number before the
6147 error, it didn't break the script. It just looked ugly. I fixed this by
6148 always allowing "-C /dev/null" if the caller is root.
6150 14. Ignore overlarge ACL variable number when reading spool file - insurance
6151 against a later release with more variables having written the file.
6153 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6154 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6155 this, as well as the form without the "IPv6" on the front.
6157 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6158 OS/Makefile-Darwin file.
6160 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6161 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6163 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6164 variable. However, some OS use ints for inodes. I've added cast to long int
6165 to get rid of the compiler warning.
6167 19. I had forgotten to lock out "/../" in configuration file names when
6168 ALT_CONFIG_PREFIX was set.
6170 20. Routers used for verification do not need to specify transports. However,
6171 if such a router generated a host list, and callout was configured, Exim
6172 crashed, because it could not find a port number from the (non-existent)
6173 transport. It now assumes port 25 in this circumstance.
6175 21. Added the -t option to exigrep.
6177 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6178 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6179 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6180 them is set, LOOKUP_LSEARCH is forced.
6182 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6184 24. Added sender and host information to the "rejected by local_scan()" log
6185 line; previously there was no indication of these.
6187 25. Added .include_if_exists.
6189 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6190 Linux. It turns out that not all file systems support this. Apparently some
6191 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6192 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6193 not supported on the file descriptor, is now ignored when Exim is trying to
6194 sync a directory. This applies only to Linux.
6196 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6198 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6199 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6202 29. Added support for maildirsize files from supplied patch (modified a bit).
6204 30. The use of :fail: followed by an empty string could lead Exim to respond to
6205 sender verification failures with (e.g.):
6207 550 Verification failed for <xxx>
6208 550 Sender verify failed
6210 where the first response line was missing the '-' that indicates it is not
6211 the final line of the response.
6213 31. The loop for finding the name of the user that called Exim had a hardwired
6214 limit of 10; it now uses the value of finduser_retries, which is used for
6215 all other user lookups.
6217 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6220 33. Exim was neglecting to zero errno before one call of strtol() when
6221 expanding a string and expecting an integer value. On some systems this
6222 resulted in spurious "integer overflow" errors. Also, it was casting the
6223 result into an int without checking.
6225 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6226 did not work. The code looks as if it has *never* worked, though it appears
6227 to have been documented since at least release 1.62. I have made it work.
6229 35. The "timeout_DNS" error in retry rules, also documented since at least
6230 1.62, also never worked. As it isn't clear exactly what this means, and
6231 clearly it isn't a major issue, I have abolished the feature by treating it
6232 as "timeout", and writing a warning to the main and panic logs.
6234 36. The display of retry rules for -brt wasn't always showing the error code
6237 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6238 timeout_connect_A, timeout_connect_MX.
6240 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6241 to the empty sender.
6243 39. The daemon was not analysing the content of -oX till after it had closed
6244 stderr and disconnected from the controlling terminal. This meant that any
6245 syntax errors were only noted on the panic log, and the return code from
6246 the command was 0. By re-arranging the code a little, I've made the
6247 decoding happen first, so such errors now appear on stderr, and the return
6248 code is 1. However, the actual setting up of the sockets still happens in
6249 the disconnected process, so errors there are still only recorded on the
6252 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6253 connections (as happens on some IP stacks) was logged at start up time as
6254 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6255 it from "IPv6 and IPv4", which means that two separate sockets are being
6258 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6259 says whether AF_INET or AF_INET6 was passed as an argument.
6261 42. Exiwhat output was messed up when time zones were included in log
6264 43. Exiwhat now gives more information about the daemon's listening ports,
6265 and whether -tls-on-connect was used.
6267 44. The "port" option of the smtp transport is now expanded.
6269 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6270 silently ignored. Now an error message is written to the main and panic
6273 46. There's a new ACL modifier called "logwrite" which writes to a log file
6274 as soon as it is encountered.
6276 47. Added $local_user_uid and $local_user_gid at routing time.
6278 48. Exim crashed when trying to verify a sender address that was being
6281 49. Exim was recognizing only a space character after ".include". It now also
6282 recognizes a tab character.
6284 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6285 extracting the relevant information from the specification. The man page no
6286 longer contains scrambled data for the -d option, and I've added a section
6287 at the front about calling Exim under different names.
6289 51. Added "extra_headers" argument to the "mail" command in filter files.
6291 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6294 53. Installed eximstats 1.29.
6296 54. Added transport_filter_timeout as a generic transport option.
6298 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6299 Cc: header lines. This was required by RFC 822, but it not required by RFC
6302 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6303 incoming messages that did not have them. Now it does so only if the
6304 message originates locally, that is, if there is no associated remote host
6305 address. When Resent- header lines are present, this applies to the Resent-
6306 lines rather than the non-Resent- lines.
6308 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6309 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6311 58. Messages for configuration errors now include the name of the main
6312 configuration file - useful now that there may be more than one file in a
6313 list (.included file names were always shown).
6315 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6316 for those rare installations that do not start the daemon as root or run it
6317 setuid root. I've cut out the call to initgroups() if the daemon is not
6320 60. The Exim user and group can now be bound into the binary as text strings
6321 that are looked up at the start of Exim's processing.
6323 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6325 62. Added $mailstore_basename variable.
6327 63. Installed patch to sieve.c from Michael Haardt.
6329 64. When Exim failed to open the panic log after failing to open the main log,
6330 the original message it was trying to log was written to stderr and debug
6331 output, but if they were not available (the usual case in production), it
6332 was lost. Now it is written to syslog before the two lines that record the
6333 failures to open the logs.
6335 65. Users' Exim filters run in subprocesses under the user's uid. It is
6336 possible for a "deliver" command or an alias in a "personal" command to
6337 provoke an address rewrite. If logging of address rewriting is configured,
6338 this fails because the process is not running as root or exim. There may be
6339 a better way of dealing with this, but for the moment (because 4.30 needs
6340 to be released), I have disabled address rewrite logging when running a
6341 filter in a non-root, non-exim process.
6347 1. The buildconfig auxiliary program wasn't quoting the value set for
6348 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6349 not defined. This bug was masked in 4.22 by the effect that was fixed in
6352 2. Some messages that were rejected after a message id was allocated were
6353 shown as "incomplete" by exigrep. It no longer does this for messages that
6354 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6356 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6357 have allow_domain_literals set, the ID did not get logged in the <= line.
6358 Domain literals are now always recognized in Message-ID: header lines.
6360 4. The first argument for a ${extract expansion item is the key name or field
6361 number. Leading and trailing spaces in this item were not being ignored,
6362 causing some misleading effects.
6364 5. When deliver_drop_privilege was set, single queue runner processes started
6365 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6366 same command in the process it spins off) were not dropping privilege.
6368 6. When the daemon running as "exim" started a queue runner, it always
6369 re-executed Exim in the spun-off process. This is a waste of effort when
6370 deliver_drop_privilege is set. The new process now just calls the
6371 queue-runner function directly.
6377 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6380 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6381 ignored. Though the use of -oP was forcing the writing of a pid file, it
6382 was always written to the default place.
6384 3. If the message "no IP address found for host xxxx" is generated during
6385 incoming verification, it is now followed by identification of the incoming
6386 connection (so you can more easily find what provoked it).
6388 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6390 5. Added some features to "harden" Exim a bit more against certain attacks:
6392 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6393 be put in Local/Makefile. This is like the never_users runtime option,
6394 but it cannot be overridden. The default setting is "root".
6396 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6397 prefix string with which any file named in a -C command line option
6400 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6401 is retained for -C and -D only if the caller of Exim is root. Without
6402 it, the exim user may also use -C and -D and retain privilege.
6404 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6405 command line option is disabled.
6407 6. Macro names set by the -D option must start with an upper case letter, just
6408 like macro names defined in the configuration file.
6410 7. Added "dereference=" facility to LDAP.
6412 8. Two instances of the typo "uknown" in the source files are fixed.
6414 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6415 the Configure-Makefile script screwed up while processing it.
6417 10. Incorporated PCRE 4.4.
6419 11. The SMTP synchronization check was not operating right at the start of an
6420 SMTP session. For example, it could not catch a HELO sent before the client
6421 waited for the greeting. There is now a check for outstanding input at the
6422 point when the greeting is written. Because of the duplex, asynchronous
6423 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6424 way, but not yet received, when the check is performed.
6426 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6427 on TCP/IP sockets, because this apparently causes some broken clients to
6430 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6431 unchanged) from the Cygwin maintainer.
6433 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6434 was supported instead of testing for mbx. Effectively a typo.
6436 15. The spa authenticator server code was not checking that the input it
6437 received was valid base64.
6439 16. The debug output line for the "set" modifier in ACLs was not showing the
6440 name of the variable that was being set.
6442 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6444 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6445 lookup on its own. Something else was needed to trigger the lookup. For
6446 example, a match in host_lookup or the need for a host name in a host list.
6447 Now, if $sender_host_name is referenced and the host name has not yet been
6448 looked up, a lookup is performed. If the lookup fails, the variable remains
6449 empty, and $host_lookup_failed is set to "1".
6451 19. Added "eqi" as a case-independent comparison operator.
6453 20. The saslauthd authentication condition could segfault if neither service
6454 nor realm was specified.
6456 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6457 error message that was logged was misleading, and incoming SMTP
6458 connections were dropped. The message is now more accurate, and temporary
6459 errors are given to SMTP connections.
6461 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6462 (except RSET) with a 421 error, until QUIT is received. However, it was
6463 failing to send a response to QUIT.
6465 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6466 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6467 ACL is now run after the helo_try_verify_hosts code.
6469 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6470 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6471 case-independent, so other case variants are also recognized). Apparently
6472 some systems use these upper case variants.
6474 25. If more than two messages were waiting for the same host, and a transport
6475 filter was specified for the transport, Exim sent two messages over the
6476 same TCP/IP connection, and then failed with "socket operation on non-
6477 socket" when it tried to send the third.
6479 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6481 27. The extern definition of crypt16() in expand.c was not being excluded when
6482 the OS had its own crypt16() function.
6484 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6485 as a preferred synonym for return_size_limit, both as an option and as an
6488 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6490 30. Changed the default configuration ACL to relax the local part checking rule
6491 for addresses that are not in any local domains. For these addresses,
6492 slashes and pipe symbols are allowed within local parts, but the sequence
6493 /../ is explicitly forbidden.
6495 31. SPA server authentication was not clearing the challenge buffer before
6498 32. log_message in a "warn" ACL statement was writing to the reject log as
6499 well as to the main log, which contradicts the documentation and doesn't
6500 seem right (because no rejection is happening). So I have stopped it.
6502 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6503 However, I am unable to do any testing of this.
6505 34. Fixed an infelicity in the appendfile transport. When checking directories
6506 for a mailbox, to see if any needed to be created, it was accidentally
6507 using path names with one or more superfluous leading slashes; tracing
6508 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6510 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6511 contents are added to the log line that is written for every discarded
6512 recipient. (Previously a log_message setting was ignored.)
6514 36. The ${quote: operator now quotes the string if it is empty.
6516 37. The install script runs exim in order to find its version number. If for
6517 some reason other than non-existence or emptiness, which it checks, it
6518 could not run './exim', it was installing it with an empty version number,
6519 i.e. as "exim-". This error state is now caught, and the installation is
6522 38. An argument was missing from the function that creates an error message
6523 when Exim fails to connect to the socket for saslauthd authentication.
6524 This could cause Exim to crash, or give a corrupted message.
6526 39. Added isip, isip4, and isip6 to ${if conditions.
6528 40. The ACL variables $acl_xx are now saved with the message, and can be
6529 accessed later in routers, transports, and filters.
6531 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6532 strings in the file are not string-expanded.
6534 42. If a MAIL command specified a SIZE value that was too large to fit into an
6535 int variable, the check against message_size_limit failed. Such values are
6536 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6537 day this will have to be increased, but I don't think I want to be around
6538 when emails are that large.
6545 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6546 iconv() is not standard in FreeBSD.
6548 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6549 IPv6 enabled. The observed symptom was a segmentation fault on return from
6550 the function os_common_find_running_interfaces() in src/os.c.
6552 3. In the check_special_case() function in daemon.c I had used "errno" as an
6553 argument name, which causes warnings on some systems. This was basically a
6554 typo, since it was named "eno" in the comments!
6556 4. The code that waits for the clock to tick (at a resolution of some fraction
6557 of a second) so as to ensure message-id uniqueness was always waiting for
6558 at least one whole tick, when it could have waited for less. [This is
6559 almost certainly not relevant at current processor speeds, where it is
6560 unlikely to ever wait at all. But we try to future-proof.]
6562 5. The function that sleeps for a time interval that includes fractions of a
6563 second contained a race. It did not block SIGALRM between setting the
6564 timer, and suspending (a couple of lines later). If the interval was short
6565 and the sigsuspend() was delayed until after it had expired, the suspension
6566 never ended. On busy systems this could lead to processes getting stuck for
6569 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6570 process, before it forks any delivery processes. The open lookup caching
6571 mechanism meant that the open file or database connection was passed into
6572 the delivery process. The problem was that delivery processes always tidy
6573 up cached lookup data. This could cause a problem for the next delivery
6574 process started by the queue runner, because the external queue runner
6575 process does not know about the closure. So the next delivery process
6576 still has data in the lookup cache. In the case of a file lookup, there was
6577 no problem because closing a file descriptor in a subprocess doesn't affect
6578 the parent. However, if the lookup was caching a connection to a database,
6579 the connection was closed, and the second delivery process was likely to
6580 see errors such as "PGSQL: query failed: server closed the connection
6581 unexpectedly". The problem has been fixed by closing all cached lookups
6582 in a queue runner before running a delivery process.
6584 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6585 seem to have the "const" qualifier which it has on other OS. I've
6588 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6589 *of the same type* (client or server) with the same public name that an
6590 error should be diagnosed.
6592 9. When Exim looked up a host name for an IP address, but failed to find the
6593 original IP address when looking up the host name (a safety check), it
6594 output the message "<ip address> does not match any IP for NULL", which was
6595 confusing, to say the least. The bug was that the host name should have
6596 appeared instead of "NULL".
6598 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6599 user that is built into the binary, and the -C or -D options is used, root
6600 privilege is dropped before the configuration file is read. In addition,
6601 logging is switched to stderr instead of the normal log files. If the
6602 configuration then re-defines the Exim user, the unprivileged environment
6603 is probably not what is expected, so Exim logs a panic warning message (but
6606 However, if deliver_drop_privilege is set, the unprivileged state may well
6607 be exactly what is intended, so the warning has been cut out in that case,
6608 and Exim is allowed to try to write to its normal log files.
6614 1. smtp_return_error_details was not giving details for temporary sender
6615 or receiver verification errors.
6617 2. Diagnose a configuration error if two authenticators have the same public
6620 3. Exim used not to create the message log file for a message until the first
6621 delivery attempt. This could be confusing when incoming messages were held
6622 for policy or load reasons. The message log file is now created at the time
6623 the message is received, and an initial "Received" line is written to it.
6625 4. The automatically generated man page for command line options had a minor
6626 bug that caused no ill effects; however, a more serious problem was that
6627 the procedure for building the man page automatically didn't always
6628 operate. Consequently, release 4.20 contains an out-of-date version. This
6629 shouldn't happen again.
6631 5. When building Exim with embedded Perl support, the script that builds the
6632 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6633 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6635 6. The freeze_tell option was not being used for messages that were frozen on
6636 arrival, either by an ACL or by local_scan().
6638 7. Added the smtp_incomplete_transaction log selector.
6640 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6641 was accepting AUTH without a new EHLO.
6643 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6644 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6647 10. Logging of TCP/IP connections (when configured) now happens in the main
6648 daemon process instead of the child process, so that the TCP/IP connection
6649 count is more accurate (but it can never be perfect).
6651 11. The use of "drop" in a nested ACL was not being handled correctly in the
6652 outer ACL. Now, if condition failure induced by the nested "drop" causes
6653 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6654 or "require"), the connection is dropped.
6656 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6657 that yield "discard" can now be used with an "accept" or a "discard" verb,
6658 but an error is generated for any others (because I can't see a useful way
6659 to define what should happen).
6661 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6662 are now processed in the same way as lines in the Exim configuration file.
6663 In particular, continuation lines are supported.
6665 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6667 15. Added -ti meaning -t -i.
6669 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6670 domains, and warn by logging if others are found.
6672 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6673 returned from GIFCONF when Exim is trying to find the list of interfaces on
6674 a host. The code in os.c has been modified to copy each ifreq to an aligned
6675 structure in all cases.
6677 Also, in some cases, the returned ifreq's were being copied to a 'struct
6678 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6679 means the last couple of bytes of an IPv6 address could be chopped if the
6680 ifreq contained only a normal sockaddr (14 bytes storage).
6682 18. Named domain lists were not supported in the hosts_treat_as_local option.
6683 An entry such as +xxxx was not recognized, and was treated as a literal
6686 19. Ensure that header lines added by a DATA ACL are included in the reject log
6687 if the ACL subsequently rejects the message.
6689 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6690 MD5 (which is deprecated).
6692 21. When testing a filter file using -bf, Exim was writing a message when it
6693 took the sender from a "From " line in the message, but it was not doing so
6694 when it took $return_path from a Return-Path: header line. It now does.
6696 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6697 with a valid header line field name (a series of printing characters
6698 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6700 23. Changed "disc" in the source to "disk" to conform to the documentation and
6701 the book and for uniformity.
6703 24. Ignore Sendmail's -Ooption=value command line item.
6705 25. When execve() failed while trying to run a command in a pipe transport,
6706 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6707 could be confused with a return value of 69 from the command itself. This
6708 has been changed to 127, the value the shell returns if it is asked to run
6709 a non-existent command. The wording for the related log line suggests a
6710 non-existent command as the problem.
6712 26. If received_header_text expands to an empty string, do not add a Received:
6713 header line to the message. (Well, it adds a token one on the spool, but
6714 marks it "old" so that it doesn't get used or transmitted.)
6716 27. Installed eximstats 1.28 (addition of -nt option).
6718 28. There was no check for failure on the call to getsockname() in the daemon
6719 code. This can fail if there is a shortage of resources on the system, with
6720 ENOMEM, for example. A temporary error is now given on failure.
6722 29. Contrary to the C standard, it seems that in some environments, the
6723 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6724 program. Exim now does this explicitly; it affects the formatting of
6725 timestamps using strftime().
6727 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6728 complaints. I've now initialized all the variables, to avoid this.
6730 32. Header lines added by a system filter were not being "seen" during
6731 transport-time rewrites.
6733 33. The info_callback() function passed to OpenSSL is set up with type void
6734 (*)(SSL *, int, int), as described somewhere. However, when calling the
6735 function (actually a macro) that sets it up, the type void(*)() is
6736 expected. I've put in a cast to prevent warnings from picky compilers.
6738 34. If a DNS black list lookup found a CNAME record, but there were no A
6739 records associated with the domain it pointed at, Exim crashed.
6741 35. If a DNS black list lookup returned more than one A record, Exim ignored
6742 all but the first. It now scans all returned addresses if a particular IP
6743 value is being sought. In this situation, the contents of the
6744 $dnslist_value variable are a list of all the addresses, separated by a
6747 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6748 to accept a host name and all its aliases if the forward lookup for any of
6749 them yielded the IP address of the incoming connection. Now it accepts only
6750 those names whose forward lookup yields the correct IP address. Any other
6751 names are discarded. This closes a loophole whereby a rogue DNS
6752 administrator could create reverse DNS records to break through a
6753 wildcarded host restriction in an ACL.
6755 37. If a user filter or a system filter that ran in a subprocess used any of
6756 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6757 the wrong values were passed to the pipe command ($thisaddress had the
6758 value of $0, $0 had the value of $1, etc). This bug was introduced by
6759 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6761 38. Improved the line breaking for long SMTP error messages from ACLs.
6762 Previously, if there was no break point between 40 and 75 characters, Exim
6763 left the rest of the message alone. Two changes have been made: (a) I've
6764 reduced the minimum length to 35 characters; (b) if it can't find a break
6765 point between 35 and 75 characters, it looks ahead and uses the first one
6766 that it finds. This may give the occasional overlong line, but at least the
6767 remaining text gets split now.
6769 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6770 file descriptors might be low, and that setting 1000 would always raise it.
6771 It turns out that in some environments, the limit is already over 1000 and
6772 that lowering it causes trouble. So now Exim takes care not to decrease it.
6774 40. When delivering a message, the value of $return_path is set to $sender_
6775 address at the start of routing (routers may change the value). By an
6776 oversight, this default was not being set up when an address was tested by
6777 -bt or -bv, which affected the outcome if any router or filter referred to
6780 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6781 log only when "message" or "log_message" are set. However, if one of the
6782 conditions was an address verification, or a call to a nested ACL, the
6783 messages generated by the underlying test were being passed through. This
6784 no longer happens. The underlying message is available in $acl_verify_
6785 message for both "message" and "log_message" expansions, so it can be
6786 passed through if needed.
6788 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6789 new expansion $bh_ to give the encoded byte string without charset
6790 translation. Translation happens only if iconv() is available; HAVE_ICONV
6791 indicates this at build time. HEADERS_CHARSET gives the charset to
6792 translate to; headers_charset can change it in the configuration, and
6793 "headers charset" can change it in an individual filter file.
6795 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6796 that creates RFC 2047 encoded "words" labels them as that charset instead
6797 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6798 expansion operator; (ii) when Exim creates a From: line for a local
6799 message; (iii) when a header line is rewritten to include a "phrase" part.
6801 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6802 buggy, causing it to skip the first lines of messages whose message ID
6803 ended in 'D'. This would not have bitten before Exim release 4.14, because
6804 message IDs were unlikely to end in 'D' before then. The effect was to have
6805 incorrect size information for certain domains.
6807 45. #include "config.h" was missing at the start of the crypt16.c module. This
6808 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6811 46. If there was a timeout during a "random" callout check, Exim treated it as
6812 a failure of the random address, and carried on sending RSET and the real
6813 address. If the delay was just some slowness somewhere, the response to the
6814 original RCPT would be taken as a response to RSET and so on, causing
6815 mayhem of various kinds.
6817 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6818 when I implemented it. It didn't allow for the fact that some option values
6819 may legitimately be negative (e.g. size_addition), and it didn't even do
6820 the right test for positive values.
6822 48. Domain names in DNS records are case-independent. Exim always looks them up
6823 in lower case. Some resolvers return domain names in exactly the case they
6824 appear in the zone file, that is, they may contain uppercase letters. Not
6825 all resolvers do this - some return always lower case. Exim was treating a
6826 change of case by a resolver as a change of domain, similar to a widening
6827 of a domain abbreviation. This triggered its re-routing code and so it was
6828 trying to route what was effectively the same domain again. This normally
6829 caused routing to fail (because the router wouldn't handle the domain
6830 twice). Now Exim checks for this case specially, and just changes the
6831 casing of the domain that it ultimately uses when it transmits the message
6834 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6837 50. If a filter generated a file delivery with a non-absolute name (possible if
6838 no home directory exists for the router), the forbid_file option was not
6841 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6842 the existing equality matching.
6844 52. Exim was using ints instead of ino_t variables in some places where it was
6845 dealing with inode numbers.
6847 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6848 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6849 called TMPDIR, and if it finds it is different, it changes its value.
6851 54. The smtp_printf() function is now made available to local_scan() so
6852 additional output lines can be written before returning. There is also an
6853 smtp_fflush() function to enable the detection of a dropped connection.
6854 The variables smtp_input and smtp_batched_input are exported to
6857 55. Changed the default runtime configuration: the message "Unknown user"
6858 has been removed from the ACL, and instead placed on the localuser router,
6859 using the cannot_route_message feature. This means that any verification
6860 failures that generate their own messages won't get overridden. Similarly,
6861 the "Unrouteable address" message that was in the ACL for unverifiable
6862 relay addresses has also been removed.
6864 56. Added hosts_avoid_esmtp to the smtp transport.
6866 57. The exicyclog script was not checking for the esoteric option
6867 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6868 will work only if exicyclog is run under the appropriate euid.
6870 58. Following a discussion on the list, the rules by which Exim recognises line
6871 endings on incoming messages have been changed. The -dropcr and drop_cr
6872 options are now no-ops, retained only for backwards compatibility. The
6873 following line terminators are recognized: LF CRLF CR. However, special
6874 processing applies to CR:
6876 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6877 nor a local message in the state where . is a terminator.
6879 (ii) If a bare CR is encountered in a header line, an extra space is added
6880 after the line terminator so as not to end the header. The reasoning
6881 behind this is that bare CRs in header lines are most likely either
6882 to be mistakes, or people trying to play silly games.
6884 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6885 was being incorrectly given as 18 bytes larger than it should have been.
6886 This is a VOB (very old bug).
6888 60. This may never have affected anything current, but just in case it has:
6889 When the local host is found other than at the start of a list of hosts,
6890 the local host, those with the same MX, and any that follow, are discarded.
6891 When the list in question was part of a longer list of hosts, the following
6892 hosts (not currently being processed) were also being discarded. This no
6893 longer happens. I'm not sure if this situation could ever has previously
6896 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6899 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6900 header. This is something that is recommended in a new Internet Draft, and
6901 is something that is documented as being done by Sendmail. There are two
6902 possible values. For messages generated by the autoreply transport, Exim
6905 Auto-Submitted: auto-replied
6907 whereas for all other generated messages (e.g. bounces) it adds
6909 Auto-Submitted: auto-generated
6911 63. The "personal" condition in filters now includes a test for the
6912 Auto-Submitted: header. If it contains the string "auto-" the message it
6913 not considered personal.
6915 64. Added rcpt_include_affixes as a generic transport option.
6917 65. Added queue_only_override (default true).
6919 66. Added the syslog_duplication option.
6921 67. If what should have been the first header line of a message consisted of
6922 a space followed by a colon, Exim was mis-interpreting it as a header line.
6923 It isn't of course - it is syntactically invalid and should therefore be
6924 treated as the start of the message body. The misbehaviour could have
6925 caused a number of strange effects, including loss of data in subsequent
6926 header lines, and spool format errors.
6928 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6929 client host had authenticated. This control can now be exercised by an ACL
6930 for more flexibility.
6932 69. By default, callouts do not happen when testing with -bh. There is now a
6933 variant, -bhc, which does actually run the callout code, including
6934 consulting and updating the callout cache.
6936 70. Added support for saslauthd authentication, courtesy of Alexander
6939 71. If statvfs() failed on the spool or log directories while checking their
6940 size for availability, Exim confusingly gave the error "space shortage".
6941 Furthermore, in debugging mode it crashed with a floating point exception.
6942 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6943 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6944 serious problem, Exim now writes to the main and panic logs when this
6945 happens, with details of the failure. It then refuses to accept the
6946 incoming message, giving the message "spool directory problem" or "log
6947 directory problem" with a 421 code for SMTP messages.
6949 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6950 0, 1, and 2 exist, because some OS complain for execs without them (see
6951 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6952 descriptors. However, the code omitted to check that the open succeeded,
6953 causing mysterious errors if for some reason the permissions on /dev/null
6954 got screwed. Now Exim writes a message to the main and panic logs, and
6955 bombs out if it can't open /dev/null.
6957 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6958 interact so that it is all more flexible. It is supposed to remain
6959 backwards compatible. Also added extra_local_interfaces.
6961 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6962 to bomb out with an assertion failure - to the client this appears as a
6963 connection drop. This problem occurs in the part of the code that was taken
6964 from the Samba project. Fortunately, the assertion is in a very simple
6965 function, so I have fixed this by reproducing the function inline in the
6966 one place where it is called, and arranging for authentication to fail
6967 instead of killing the process with assert().
6969 75. The SPA client code was not working when the server requested OEM rather
6970 than Unicode encoding.
6972 76. Added code to make require_files with a specific uid setting more usable in
6973 the case where statting the file as root fails - usually a non-root-mounted
6974 NFS file system. When this happens and the failure is EACCES, Exim now
6975 forks a subprocess and does the per-uid checking as the relevant uid.
6977 77. Added process_log_path.
6979 78. If log_file_path was not explicitly set, a setting of check_log_space or
6980 check_log_inodes was ignored.
6982 79. If a space check for the spool or log partitions fails, the incident is now
6983 logged. Of course, in the latter case the data may get lost...
6985 80. Added the %p formatting code to string_format() so that it can be used to
6986 print addresses in debug_print(). Adjusted all the address printing in the
6987 debugging in store.c to use %p rather than %d.
6989 81. There was a concern that a line of code in smtp_in.c could overflow a
6990 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6991 initially expressed, the concern was not well-founded, because trailing
6992 spaces are removed early. However, if the trailing spaces were followed by
6993 a NULL, they did not get removed, so the overflow was possible. Two fixes
6996 (a) I re-wrote the offending code in a cleaner fashion.
6997 (b) If an incoming SMTP command contains a NULL character, it is rejected
7000 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7001 runs initgroups(), so that if the Exim user is in any additional groups,
7002 they will be used during message reception.
7008 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob