1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Replace the store_release() internal interface with store_newblock(),
12 which internalises the check required to safely use the old one, plus
13 the allocate and data copy operations duplicated in both (!) of the
16 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
17 modifier. This matches the restriction on the commandline.
19 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
20 Previously only the last row was returned.
22 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
23 we assumed that tags in the header were well-formed, and parsed the
24 element content after inspecting only the first char of the tag.
25 Assumptions at that stage could crash the receive process on malformed
28 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
29 While running the DKIM ACL we operate on the Permanent memory pool so that
30 variables created with "set" persist to the DATA ACL. Also (at any time)
31 DNS lookups that fail create cache records using the Permanent pool. But
32 expansions release any allocations made on the current pool - so a dnsdb
33 lookup expansion done in the DKIM ACL releases the memory used for the
34 DNS negative-cache, and bad things result. Solution is to switch to the
35 Main pool for expansions.
36 While we're in that code, add checks on the DNS cache during store_reset,
37 active in the testsuite.
38 Problem spotted, and debugging aided, by Wolfgang Breyha.
40 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
41 When none of the hosts presented to a transport match an already-open
42 connection, close it and proceed with the list. Previously we would
43 queue the message. Spotted by Lena with Yahoo, probably involving
46 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
47 Previously a spurious "250 OK id=" response was appended to the proper
50 JH/08 The "support for" informational output now, which built with Content
51 Scanning support, has a line for the malware scanner interfaces compiled
52 in. Interface can be individually included or not at build time.
54 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
55 by the template makefile "src/EDITME". The "STREAM" support for an older
56 ClamAV interface method is removed.
58 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
59 rows affected is given instead).
65 JH/01 Rework error string handling in TLS interface so that the caller in
66 more cases is responsible for logging. This permits library-sourced
67 string to be attached to addresses during delivery, and collapses
68 pairs of long lines into single ones.
70 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
71 during configuration. Wildcards are allowed and expanded.
73 JH/02 Rework error string handling in DKIM to pass more info back to callers.
74 This permits better logging.
76 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
77 do not close it down and have the child transport start it up again on
78 the passed-on TCP connection. Instead, proxy the child (and any
79 subsequent ones) for TLS via a unix-domain socket channel. Logging is
80 affected: the continued delivery log lines do not have any DNSSEC, TLS
81 Certificate or OCSP information. TLS cipher information is still logged.
83 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
84 identical IP addresses on different listening ports. Will also affect
87 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
88 add noisy ifdef guards to special-case this sillyness.
89 Patch from Bernd Kuhls.
91 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
92 than 255 are no longer allowed.
94 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
95 Disable the session-cache too, which might reduce our load. Since we
96 currrectly use a new context for every connection, both as server and
97 client, there is no benefit for these.
98 GnuTLS appears to not support tickets server-side by default (we don't
99 call gnutls_session_ticket_enable_server()) but client side is enabled
100 by default on recent versions (3.1.3 +) unless the PFS priority string
103 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
104 <https://reproducible-builds.org/specs/source-date-epoch/>.
106 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
107 the check for any unsuccessful recipients did not notice the limit, and
108 erroneously found still-pending ones.
110 JH/08 Pipeline CHUNKING command and data together, on kernels that support
111 MSG_MORE. Only in-clear (not on TLS connections).
113 JH/09 Avoid using a temporary file during transport using dkim. Unless a
114 transport-filter is involved we can buffer the headers in memory for
115 creating the signature, and read the spool data file once for the
116 signature and again for transmission.
118 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
119 4.77 as the kernel support then wasn't solid, having issues in 64bit
120 mode. Now, it's been long enough. Add support for FreeBSD also.
122 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
123 case where the routing stage had gathered several addresses to send to
124 a host before calling the transport for the first, we previously failed
125 to close down TLS in the old transport process before passing the TCP
126 connection to the new process. The new one sent a STARTTLS command
127 which naturally failed, giving a failed delivery and bloating the retry
128 database. Investigation and fix prototype from Wolfgang Breyha.
130 JH/12 Fix check on SMTP command input synchronisation. Previously there were
131 false-negatives in the check that the sender had not preempted a response
132 or prompt from Exim (running as a server), due to that code's lack of
133 awareness of the SMTP input buffering.
135 PP/04 Add commandline_checks_require_admin option.
136 Exim drops privileges sanely, various checks such as -be aren't a
137 security problem, as long as you trust local users with access to their
138 own account. When invoked by services which pass untrusted data to
139 Exim, this might be an issue. Set this option in main configuration
140 AND make fixes to the calling application, such as using `--` to stop
143 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
144 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
145 responses to those, into a single TLS record each way (this usually means
146 a single packet). As a side issue, smtp_enforce_sync now works on TLS
149 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
150 affects you only if you're dancing at the edge of the param size limits.
151 If you are, and this message makes sense to you, then: raise the
152 configured limit or use OpenSSL 1.1. Nothing we can do for older
155 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
156 cmdline element to get the documented default one. Previously it was
159 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
160 are unsuable for non-message-specific reasons - eg. connection timeout,
161 banner-time rejection.
163 JH/16 Fix logging of delivery remote port, when specified by router, under
166 PP/06 Repair manualroute's ability to take options in any order, even if one
167 is the name of a transport.
170 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
172 JH/17 Change the list-building routines interface to use the expanding-string
173 triplet model, for better allocation and copying behaviour.
175 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
176 Previously it was constructed the first time a possibly-matching string
177 was met in the configuration file input during startup; now it is done
180 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
181 compatible one, to avoid the (poorly documented) possibility of a config
182 file in the working directory redirecting the DB files, possibly correpting
183 some existing file. CVE-2017-10140 assigned for BDB.
185 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
186 cache-hot. Previously, although the result was properly cached, the
187 initial verify call returned a defer.
189 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
190 the main verify for receipient in uncached-mode.
192 JH/22 Retire historical build files to an "unsupported" subdir. These are
193 defined as "ones for which we have no current evidence of testing".
195 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
196 if present. Previously it was ignored.
198 JH/24 Start using specified-initialisers in C structure init coding. This is
199 a C99 feature (it's 2017, so now considered safe).
201 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
202 if was a fixed-sized field and bitmask ops via macros; it is now more
205 PP/07 GitHub PR 56: Apply MariaDB build fix.
206 Patch provided by Jaroslav Škarvada.
208 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
209 during Coverity cleanups [4.87 JH/47]
210 Diagnosis and fix provided by Michael Fischer v. Mollard.
212 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
213 the right size to place the terminating semicolon on its own folded
214 line, the header hash was calculated to an incorrect value thanks to
215 the (relaxed) space the fold became.
217 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
218 and confused the parent.
220 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
221 which could crash as a result. This could lead to undeliverable messages.
223 JH/28 Logging: "next input sent too soon" now shows where input was truncated
226 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
227 matters on fast-turnover and PID-randomising systems, which were getting
228 out-of-order delivery.
230 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
231 a possibly-overlapping copy. The symptom was that "Remote host closed
232 connection in response to HELO" was logged instead of the actual 4xx
235 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
236 Previously only that bufferd was discarded, resulting in SYMTP command
239 JH/32 DKIM: when a message has multiple signatures matching an identity given
240 in dkim_verify_signers, run the dkim acl once for each. Previously only
241 one run was done. Bug 2189.
243 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
244 "panic the current process" to "deliberately defer". The panic log is
245 still written with the problem list name; the mail and reject logs now
246 get a temp-reject line for the message that was being handled, saying
247 something like "domains check lookup or other defer". The SMTP 451
248 message is still "Temporary local problem".
250 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
251 A crafted sequence of BDAT commands could result in in-use memory beeing
252 freed. CVE-2017-16943.
254 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
255 from SMTP input. Previously it was always done; now only done for DATA
256 and not BDAT commands. CVE-2017-16944.
258 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
259 to the message (such as an overlong header line). Previously this was
260 not done and we did not exit BDAT mode. Followon from the previous item
261 though a different problem.
267 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
268 than -2003 did; needs libidn2 in addition to libidn.
270 JH/02 The path option on a pipe transport is now expanded before use.
272 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
273 Patch provided by "Björn", documentation fix added too.
275 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
276 missing a wire-to-host endian conversion.
278 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
279 close after a BDAT command line could be taken as a following command,
280 giving a synch failure. Fix by only checking for synch immediately
281 before acknowledging the chunk.
283 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
284 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
285 macro. Patches provided by Josh Soref.
287 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
288 Previously we did not; the RFC seems ambiguous and VRFY is not listed
289 by IANA as a service extension. However, John Klensin suggests that we
292 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
293 the dkim code may be unix-mode line endings rather than smtp wire-format
294 CRLF, so prepend a CR to any bare LF.
296 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
297 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
299 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
300 the first were themselves being wrongly included in the feed into dkim
301 processing; with most chunk sizes in use this resulted in an incorrect
302 body hash calculated value.
304 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
305 DKIM signature block, for verification. Although advised against by
306 standards it is specifically not ruled illegal.
308 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
310 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
311 missing a body hash (the bh= tag).
313 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
314 It seems that HAProxy sends the Proxy Protocol information in clear and
315 only then does a TLS startup, so do the same.
317 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
318 TCP connections (such as for Spamd) unless the daemon successfully set
319 Fast Open mode on its listening sockets. This fixes breakage seen on
320 too-old kernels or those not configured for Fast Open, at the cost of
321 requiring both directions being enabled for TFO, and TFO never being used
322 by non-daemon-related Exim processes.
324 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
325 endings, at least on the first header line. Try to canonify any that get
326 past that check, despite the cost.
328 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
329 now limited to an arbitrary five deep, while parsing addresses with the
330 strip_excess_angle_brackets option enabled.
332 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
333 instead leave the unprompted TLS handshake in socket buffer for the
334 TLS library to consume.
336 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
338 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
340 JH/16 Drop variables when they go out of scope. Memory management drops a whole
341 region in one operation, for speed, and this leaves assigned pointers
342 dangling. Add checks run only under the testsuite which checks all
343 variables at a store-reset and panics on a dangling pointer; add code
344 explicitly nulling out all the variables discovered. Fixes one known
345 bug: a transport crash, where a dangling pointer for $sending_ip_address
346 originally assigned in a verify callout, is re-used.
348 PP/06 Drop '.' from @INC in various Perl scripts.
350 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
352 PP/08 Reduce a number of compilation warnings under clang; building with
353 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
354 should be warning-free.
356 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
358 HS/01 Fix portability problems introduced by PP/08 for platforms where
359 realloc(NULL) is not equivalent to malloc() [SunOS et al].
361 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
362 chunk. This allows us to accept broken chunked messages. We need a more
363 general solution here.
365 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
366 already-broken messages in the queue.
368 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
370 JH/19 Fix reference counting bug in routing-generated-address tracking.
376 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
377 supports it and a size is available (ie. the sending peer gave us one).
379 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
380 years of being deprecated). The replacements are the ACLs
381 acl_smtp_mime and acl_not_smtp_mime.
383 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
384 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
385 or even an in-clear connection were permitted. Now, if the host lookup
386 was dnssec and dane was requested then the host is only used if the
387 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
388 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
389 if one fails this test.
390 This means that a poorly-configured remote DNS will make it incommunicado;
391 but it protects against a DNS-interception attack on it.
393 JH/04 Bug 1810: make continued-use of an open smtp transport connection
394 non-noisy when a race steals the message being considered.
396 JH/05 If main configuration option tls_certificate is unset, generate a
397 self-signed certificate for inbound TLS connections.
399 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
400 in rewrites and routers.
402 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
403 and logged a warning sing 4.83; now they are a configuration file error.
405 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
406 (lacking @domain). Apply the same qualification processing as RCPT.
408 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
410 JH/10 Support ${sha256:} applied to a string (as well as the previous
413 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
414 a cutthrough deliver is pending, as we always want to make a connection.
415 This also avoids re-routing the message when later placing the cutthrough
416 connection after a verify cache hit.
417 Do not update it with the verify result either.
419 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
420 when routing results in more than one destination address.
422 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
423 signing (which inhibits the cutthrough capability). Previously only
424 the presence of an option was tested; now an expansion evaluating as
425 empty is permissible (obviously it should depend only on data available
426 when the cutthrough connection is made).
428 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
429 the relevant preceding SMTP command did not note the pipelining mode.
431 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
432 Previously they were not counted.
434 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
435 as one having no matching records. Previously we deferred the message
436 that needed the lookup.
438 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
439 distinguished as "(=".
441 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
442 for missing MX records. Previously it only worked for missing A records.
444 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
446 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
447 after the data-go-ahead and data-ack. Patch from Jason Betts.
449 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
450 even for a "none" policy. Patch from Tony Meyer.
452 JH/22 Fix continued use of a connection for further deliveries. If a port was
453 specified by a router, it must also match for the delivery to be
456 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
457 When one of the recipients of a message was unsuitable for the connection
458 (has no matching addresses), we lost track of needing to mark it
459 deferred. As a result mail would be lost.
461 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
463 JH/25 Decoding ACL controls is now done using a binary search; the source code
464 takes up less space and should be simpler to maintain. Merge the ACL
465 condition decode tables also, with similar effect.
467 JH/26 Fix problem with one_time used on a redirect router which returned the
468 parent address unchanged. A retry would see the parent address marked as
469 delivered, so not attempt the (identical) child. As a result mail would
472 JH/27 Fix a possible security hole, wherein a process operating with the Exim
473 UID can gain a root shell. Credit to http://www.halfdog.net/ for
474 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
477 JH/28 Enable {spool,log} filesystem space and inode checks as default.
478 Main config options check_{log,spool}_{inodes,space} are now
479 100 inodes, 10MB unless set otherwise in the configuration.
481 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
482 Previously it only applied to the main-section connection policy
485 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
487 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
488 by me. Added RFC7919 DH primes as an alternative.
490 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
491 are not in the system include path.
493 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
494 GnuTLS, when a session startup failed (eg because the client disconnected)
495 Exim did stdio operations after fclose. This was exposed by a recent
496 change which nulled out the file handle after the fclose.
498 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
499 signed directly by the cert-signing cert, rather than an intermediate
500 OCSP-signing cert. This is the model used by LetsEncrypt.
502 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
504 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
505 an incoming connection.
507 HS/02 Bug 1802: Do not half-close the connection after sending a request
510 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
511 fallback to "prime256v1".
513 JH/34 SECURITY: Use proper copy of DATA command in error message.
514 Could leak key material. Remotely exploitable. CVE-2016-9963.
520 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
521 and 3.4.4 - once the server is enabled to respond to an OCSP request
522 it does even when not requested, resulting in a stapling non-aware
523 client dropping the TLS connection.
525 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
526 support variable-length bit vectors. No functional change.
528 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
529 The I= interface field on outgoing lines is now after the H= remote
530 host field, same as incoming lines. There is a separate
531 outgoing_interface log selector which allows you to disable the
534 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
535 If not running log_selector +smtp_connection the mainlog would be held
536 open indefinitely after a "too many connections" event, including to a
537 deleted file after a log rotate. Leave the per net connection logging
538 leaving it open for efficiency as that will be quickly detected by the
539 check on the next write.
541 HS/01 Bug 1671: Fix post transport crash.
542 Processing the wait-<transport> messages could crash the delivery
543 process if the message IDs didn't exist for some reason. When
544 using 'split_spool_directory=yes' the construction of the spool
545 file name failed already, exposing the same netto behaviour.
547 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
548 mime_regex ACL conditions.
550 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
551 to DSN fail messages (bounces): remote IP, remote greeting, remote response
552 to HELO, local diagnostic string.
554 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
555 log line to debug. Even when configured with a tls authenticator many
556 client connections are expected to not authenticate in this way, so
557 an authenticate fail is not an error.
559 HS/02 Add the Exim version string to the process info. This way exiwhat
560 gives some more detail about the running daemon.
562 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
563 matter for fast-change records such as DNSBLs.
565 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
566 retry record, even if constant. There may be multiple transports with
567 different interface settings and the retry behaviour needs to be kept
570 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
572 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
574 JH/10 Bug 840: fix log_defer_output option of pipe transport
576 JH/11 Bug 830: use same host for all RCPTS of a message, even under
577 hosts_randomize. This matters a lot when combined with mua_wrapper.
579 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
580 ${quote_pgsql:<string>} operator.
582 JH/13 Bug 1708: avoid misaligned access in cached lookup.
584 JH/14 Change header file name for freeradius-client. Relevant if compiling
585 with Radius support; from the Gentoo tree and checked under Fedora.
587 JH/15 Bug 1712: Introduce $prdr_requested flag variable
589 JH/16 Bug 1714: Permit an empty string as expansion result for transport
590 option transport_filter, meaning no filtering.
592 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
594 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
595 defaults to "*" (all hosts). The variable is now available when not built
596 with TLS, default unset, mainly to enable keeping the testsuite sane.
597 If a server certificate is not supplied (via tls_certificate) an error is
598 logged, and clients will find TLS connections fail on startup. Presumably
599 they will retry in-clear.
600 Packagers of Exim are strongly encouraged to create a server certificate
601 at installation time.
603 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
604 with the $config_file variable.
606 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
607 in transport context, after the attempt, and per-recipient. The latter type
608 is per host attempted. The event data is the error message, and the errno
609 information encodes the lookup type (A vs. MX) used for the (first) host,
610 and the trailing two digits of the smtp 4xx response.
612 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
613 to write to mainlog (or rejectlog, paniclog) in the window between file
614 creation and permissions/ownership being changed. Particularly affects
615 installations where exicyclog is run as root, rather than exim user;
616 result is that the running daemon panics and dies.
618 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
620 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
621 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
622 "pri" and "weight". Note that the previous implicit priority given by the
623 list order is no longer honoured.
625 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
628 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
629 by defining SUPPORT_SOCKS.
631 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
632 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
633 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
634 variables are renamed to proxy_{local,external}_{address,port}.
636 JH/27 Move Internationalisation support from Experimental to mainline, enabled
637 for a build by defining SUPPORT_I18N
639 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
640 of the query string, and make ${quote_redis:} do that quoting.
642 JH/29 Move Events support from Experimental to mainline, enabled by default
643 and removable for a build by defining DISABLE_EVENT.
645 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
647 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
648 cached by the daemon.
650 JH/32 Move Redis support from Experimental to mainline, enabled for a build
651 by defining LOOKUP_REDIS. The libhiredis library is required.
653 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
654 keys are given for lookup.
656 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
657 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
658 only supported when built with TLS support. The PolarSSL SHA routines
659 are still used when the TLS library is too old for convenient support.
661 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
662 openssl_options), for security. OpenSSL forces this from version 1.1.0
663 server-side so match that on older versions.
665 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
666 allocation for $value could be released as the expansion processing
667 concluded, but leaving the global pointer active for it.
669 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
670 and to use the domains and local_parts ACL conditions.
672 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
673 incorrectly not doubled on cutthrough transmission, hence seen as a
674 body-termination at the receiving system - resulting in truncated mails.
675 Commonly the sender saw a TCP-level error, and retransmitted the message
676 via the normal store-and-forward channel. This could result in duplicates
677 received - but deduplicating mailstores were liable to retain only the
678 initial truncated version.
680 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
682 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
684 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
685 we're in there, support oversigning also; bug 1309.
687 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
689 HS/04 Add support for keep_environment and add_environment options.
691 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
692 either intentional arithmetic overflow during PRNG, or testing config-
695 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
696 delivery resulted in actual delivery. Cancel cutthrough before DATA
699 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
700 rejects a recipient: pass the reject to the originator.
702 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
703 Many were false-positives and ignorable, but it's worth fixing the
706 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
707 for the new environment-manipulation done at startup. Move the routines
708 from being local to tls.c to being global via the os.c file.
710 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
711 an extract embedded as result-arg for a map, the first arg for extract
712 is unavailable so we cannot tell if this is a numbered or keyed
713 extraction. Accept either.
719 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
722 JH/02 The smtp transport option "multi_domain" is now expanded.
724 JH/03 The smtp transport now requests PRDR by default, if the server offers
727 JH/04 Certificate name checking on server certificates, when exim is a client,
728 is now done by default. The transport option tls_verify_cert_hostnames
729 can be used to disable this per-host. The build option
730 EXPERIMENTAL_CERTNAMES is withdrawn.
732 JH/05 The value of the tls_verify_certificates smtp transport and main options
733 default to the word "system" to access the system default CA bundle.
734 For GnuTLS, only version 3.0.20 or later.
736 JH/06 Verification of the server certificate for a TLS connection is now tried
737 (but not required) by default. The verification status is now logged by
738 default, for both outbound TLS and client-certificate supplying inbound
741 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
744 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
745 Status Notification (bounce) messages are now MIME format per RFC 3464.
746 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
747 under the control of the dsn_advertise_hosts option, and routers may
748 have a dsn_lasthop option.
750 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
751 default, modifiable by a malware= option. The list separator for
752 the options can now be changed in the usual way. Bug 68.
754 JH/10 The smtp_receive_timeout main option is now expanded before use.
756 JH/11 The incoming_interface log option now also enables logging of the
757 local interface on delivery outgoing connections.
759 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
760 if the interface and destination host and port all match.
762 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
765 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
766 Patch from Andrew Lewis.
768 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
769 now supports optional time-restrictions, weighting, and priority
770 modifiers per server. Patch originally by <rommer@active.by>.
772 JH/16 The spamd_address main option now supports a mixed list of local
773 and remote servers. Remote servers can be IPv6 addresses, and
774 specify a port-range.
776 JH/17 Bug 68: The spamd_address main option now supports an optional
777 timeout value per server.
779 JH/18 Bug 1581: Router and transport options headers_add/remove can
780 now have the list separator specified.
782 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
785 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
788 JH/21 Support for the A6 type of dns record is withdrawn.
790 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
791 rather than the verbs used.
793 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
794 from 255 to 1024 chars.
796 JH/24 Verification callouts now attempt to use TLS by default.
798 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
799 are generic router options now. The defaults didn't change.
801 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
802 Original patch from Alexander Shikoff, worked over by JH.
804 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
805 files if ZCAT_COMMAND is not executable.
807 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
809 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
811 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
812 Normally benign, it bites when the pair was led to by a CNAME;
813 modern usage is to not canonicalize the domain to a CNAME target
814 (and we were inconsistent anyway for A-only vs AAAA+A).
816 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
818 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
819 when evaluating $sender_host_dnssec.
821 JH/31 Check the HELO verification lookup for DNSSEC, adding new
822 $sender_helo_dnssec variable.
824 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
826 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
828 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
830 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
831 documented as working, but never had. Support all but $spam_report.
833 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
834 added for tls authenticator.
836 HS/03 Add perl_taintmode main config option
842 TL/01 When running the test suite, the README says that variables such as
843 no_msglog_check are global and can be placed anywhere in a specific
844 test's script, however it was observed that placement needed to be near
845 the beginning for it to behave that way. Changed the runtest perl
846 script to read through the entire script once to detect and set these
847 variables, reset to the beginning of the script, and then run through
848 the script parsing/test process like normal.
850 TL/02 The BSD's have an arc4random API. One of the functions to induce
851 adding randomness was arc4random_stir(), but it has been removed in
852 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
853 function when detected.
855 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
856 cause callback expansion.
858 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
859 syntax errors in an expansion can be treated as a string instead of
860 logging or causing an error, due to the internal use of bool_lax
861 instead of bool when processing it.
863 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
864 server certificates when making smtp deliveries.
866 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
868 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
870 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
872 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
873 Merged patch from Sebastian Wiedenroth.
875 JH/05 Fix results-pipe from transport process. Several recipients, combined
876 with certificate use, exposed issues where response data items split
877 over buffer boundaries were not parsed properly. This eventually
878 resulted in duplicates being sent. This issue only became common enough
879 to notice due to the introduction of connection certificate information,
880 the item size being so much larger. Found and fixed by Wolfgang Breyha.
882 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
883 size buffer was used, resulting in syntax errors when an expansion
886 JH/07 Add support for directories of certificates when compiled with a GnuTLS
887 version 3.3.6 or later.
889 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
890 is EXPERIMENTAL_EVENT, the main-configuration and transport options
891 both become "event_action", the variables become $event_name, $event_data
892 and $event_defer_errno. There is a new variable $verify_mode, usable in
893 routers, transports and related events. The tls:cert event is now also
894 raised for inbound connections, if the main configuration event_action
897 TL/06 In test suite, disable OCSP for old versions of openssl which contained
898 early OCSP support, but no stapling (appears to be less than 1.0.0).
900 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
901 server certificate names available under the smtp transport option
902 "tls_verify_cert_hostname" now do not permit multi-component wildcard
905 JH/10 Time-related extraction expansions from certificates now use the main
906 option "timezone" setting for output formatting, and are consistent
907 between OpenSSL and GnuTLS compilations. Bug 1541.
909 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
910 encoded parameter in the incoming message. Bug 1558.
912 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
913 include certificate info, eximon was claiming there were spoolfile
916 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
918 JH/14 Log delivery-related information more consistently, using the sequence
919 "H=<name> [<ip>]" wherever possible.
921 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
922 are problematic for Debian distribution, omit them from the release
925 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
927 JH/16 Fix string representation of time values on 64bit time_t architectures.
930 JH/17 Fix a null-indirection in certextract expansions when a nondefault
931 output list separator was used.
936 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
937 checkers that were complaining about end of non-void function with no
940 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
941 This was a regression introduced in 4.83 by another bugfix.
943 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
945 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
946 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
952 TF/01 Correctly close the server side of TLS when forking for delivery.
954 When a message was received over SMTP with TLS, Exim failed to clear up
955 the incoming connection properly after forking off the child process to
956 deliver the message. In some situations the subsequent outgoing
957 delivery connection happened to have the same fd number as the incoming
958 connection previously had. Exim would try to use TLS and fail, logging
959 a "Bad file descriptor" error.
961 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
962 utilities have not been installed.
964 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
965 temporary space as the ACL may create new global variables.
967 TL/01 LDAP support uses per connection or global context settings, depending
968 upon the detected version of the libraries at build time.
970 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
971 to extract and use the src ip:port in logging and expansions as if it
972 were a direct connection from the outside internet. PPv2 support was
973 updated based on HAProxy spec change in May 2014.
975 JH/02 Add ${listextract {number}{list}{success}{fail}}.
977 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
978 Properly escape header and check for NULL return.
980 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
983 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
985 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
986 characters in header names, implemented as a verify condition.
987 Contributed by Michael Fischer v. Mollard.
989 TL/05 Rename SPF condition results err_perm and err_temp to standardized
990 results permerror and temperror. Previous values are deprecated but
991 still accepted. In a future release, err_perm and err_temp will be
992 completely removed, which will be a backward incompatibility if the
993 ACL tests for either of these two old results. Patch contributed by
994 user bes-internal on the mailing list.
996 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
998 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
999 selectors, in both main and reject logs.
1001 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1004 JH/07 Add malware type "sock" for talking to simple daemon.
1006 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1008 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1009 routers/transports under cutthrough routing.
1011 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1012 numbers. Touch up "bool" conditional to keep the same definition.
1014 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1016 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1018 JH/12 Expand items in router/transport headers_add or headers_remove lists
1019 individually rather than the list as a whole. Bug 1452.
1021 Required for reasonable handling of multiple headers_ options when
1022 they may be empty; requires that headers_remove items with embedded
1023 colons must have them doubled (or the list-separator changed).
1025 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1026 view the policy declared in the DMARC record. Currently, $dmarc_status
1027 is a combined value of both the record presence and the result of the
1030 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1032 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1033 dnslookup router and the smtp transport (applying to the forward
1036 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1037 of ldap servers used for a specific lookup. Patch provided by Heiko
1040 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1041 New variable $lookup_dnssec_authenticated for observability.
1043 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1044 Patch submitted by Lars Timman.
1046 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1048 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1049 Requires trusted mode and valid format message id, aborts otherwise.
1050 Patch contributed by Heiko Schlichting.
1052 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1053 certextract with support for various fields. Bug 1358.
1055 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1056 is requested by default, modifiable by smtp transport option
1059 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1060 operate on certificate variables to give certificate fingerprints
1061 Also new ${sha256:cert_variable}.
1063 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1065 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1066 Christian Aistleitner.
1068 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1070 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1071 file. Patch from Wolfgang Breyha.
1073 JH/25 Expand the coverage of the delivery $host and $host_address to
1074 client authenticators run in verify callout. Bug 1476.
1076 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1077 align with daemon_smtp_ports. Bug 72.
1079 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1080 support and error reporting did not work properly.
1082 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1083 and is readable. Patch from Andrew Colin Kissa.
1085 TL/14 Enhance documentation of ${run expansion and how it parses the
1086 commandline after expansion, particularly in the case when an
1087 unquoted variable expansion results in an empty value.
1089 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1091 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1092 by renaming away from it. Addresses GH issue 15, reported by
1095 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1097 TL/15 SECURITY: prevent double expansion in math comparison functions
1098 (can expand unsanitized data). Not remotely exploitable.
1105 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1107 PP/02 Make -n do something, by making it not do something.
1108 When combined with -bP, the name of an option is not output.
1110 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1113 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1114 $sender_host_name and config options to manage this, and basic check
1117 PP/05 DSCP support for outbound connections and control modifier for inbound.
1119 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1120 (Only plugin which currently uses this is kerberos4, which nobody should
1121 be using, but we should make it available and other future plugins might
1122 conceivably use it, even though it would break NAT; stuff *should* be
1123 using channel bindings instead).
1125 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1126 name; added for Sendmail compatibility; requires admin caller.
1127 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1128 just ignore it); requires trusted caller.
1129 Also parse but ignore: -Ac -Am -X<logfile>
1132 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1134 TL/02 Add +smtp_confirmation as a default logging option.
1136 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1137 Patch by Magnus Holmgren from 2007-02-20.
1139 TL/04 Bugzilla 1281 - Spec typo.
1140 Bugzilla 1283 - Spec typo.
1141 Bugzilla 1290 - Spec grammar fixes.
1143 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1145 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1147 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1148 mailing list by by Dmitry Isaikin.
1150 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1152 JH/02 Support "G" suffix to numbers in ${if comparisons.
1154 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1156 NM/01 Bugzilla 1197 - Spec typo
1157 Bugzilla 1196 - Spec examples corrections
1159 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1161 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1162 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1165 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1166 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1168 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1169 "acl {{name}{arg}...}", and optional args on acl condition
1172 JH/05 Permit multiple router/transport headers_add/remove lines.
1174 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1176 JH/07 Avoid using a waiting database for a single-message-only transport.
1177 Performance patch from Paul Fisher. Bugzilla 1262.
1179 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1182 JH/09 Add $headers_added variable, with content from use of ACL modifier
1183 add_header (but not yet added to the message). Bugzilla 199.
1185 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1186 Pulled from Bugzilla 817 by Wolfgang Breyha.
1188 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1190 (nb: this is the same fix as in Exim 4.80.1)
1192 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1195 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1198 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1200 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1201 advertise SMTP AUTH mechanism to us, instead of a generic
1202 protocol violation error. Also, make Exim more robust to bad
1203 data from the Dovecot auth socket.
1205 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1207 When a queue runner is handling a message, Exim first routes the
1208 recipient addresses, during which it prunes them based on the retry
1209 hints database. After that it attempts to deliver the message to
1210 any remaining recipients. It then updates the hints database using
1213 So if a recipient address works intermittently, it can get repeatedly
1214 deferred at routing time. The retry hints record remains fresh so the
1215 address never reaches the final cutoff time.
1217 This is a fairly common occurrence when a user is bumping up against
1218 their storage quota. Exim had some logic in its local delivery code
1219 to deal with this. However it did not apply to per-recipient defers
1220 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1222 This change adds a proper retry rule check during routing so that the
1223 final cutoff time is checked against the message's age. We only do
1224 this check if there is an address retry record and there is not a
1225 domain retry record; this implies that previous attempts to handle
1226 the address had the retry_use_local_parts option turned on. We use
1227 this as an approximation for the destination being like a local
1228 delivery, as in LMTP.
1230 I suspect this new check makes the old local delivery cutoff check
1231 redundant, but I have not verified this so I left the code in place.
1233 TF/02 Correct gecos expansion when From: is a prefix of the username.
1235 Test 0254 submits a message to Exim with the header
1239 When I ran the test suite under the user fanf2, Exim expanded
1240 the header to contain my full name, whereas it should have added
1241 a Resent-Sender: header. It erroneously treats any prefix of the
1242 username as equal to the username.
1244 This change corrects that bug.
1246 GF/01 DCC debug and logging tidyup
1247 Error conditions log to paniclog rather than rejectlog.
1248 Debug lines prefixed by "DCC: " to remove any ambiguity.
1250 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1252 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1253 Bug spotted by Jeremy Harris; was flawed since initial commit.
1254 Would have resulted in OCSP responses post-SNI triggering an Exim
1255 NULL dereference and crash.
1257 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1259 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1260 Bug detection, analysis and fix by Samuel Thibault.
1261 Bugzilla 1331, Debian bug #698092.
1263 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1265 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1266 Server implementation by Todd Lyons, client by JH.
1267 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1268 config variable "prdr_enable" controls whether the server
1269 advertises the facility. If the client requests PRDR a new
1270 acl_data_smtp_prdr ACL is called once for each recipient, after
1271 the body content is received and before the acl_smtp_data ACL.
1272 The client is controlled by both of: a hosts_try_prdr option
1273 on the smtp transport, and the server advertisement.
1274 Default client logging of deliveries and rejections involving
1275 PRDR are flagged with the string "PRDR".
1277 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1278 fclose(). Diagnosis by Todd Lyons.
1280 PP/17 Update configure.default to handle IPv6 localhost better.
1281 Patch by Alain Williams (plus minor tweaks).
1284 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1285 This is now consistent with GnuTLS, and is now documented: the
1286 previous undocumented portable approach to treating the option as
1287 unset was to force an expansion failure. That still works, and
1288 an empty string is now equivalent.
1290 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1291 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1292 not performing validation itself.
1294 PP/20 Added force_command boolean option to pipe transport.
1295 Patch from Nick Koston, of cPanel Inc.
1297 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1300 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1302 PP/21 Fix eximon continuous updating with timestamped log-files.
1303 Broken in a format-string cleanup in 4.80, missed when I repaired the
1304 other false fix of the same issue.
1305 Report and fix from Heiko Schlichting.
1308 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1309 Report from Prashanth Katuri.
1311 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1312 It's SecureTransport, so affects any MacOS clients which use the
1313 system-integrated TLS libraries, including email clients.
1315 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1316 using a MIME ACL for non-SMTP local injection.
1317 Report and assistance in diagnosis by Warren Baker.
1319 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1321 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1323 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1324 last id that failed so it may be referenced in subsequent ACL's.
1326 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1329 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1330 ldap library initialization, allowing self-signed CA's to be
1331 used. Also properly sets require_cert option later in code by
1332 using NULL (global ldap config) instead of ldap handle (per
1333 session). Bug diagnosis and testing by alxgomz.
1335 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1336 the src/util/ subdirectory.
1338 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1339 renamed to Transport Post Delivery Action by Jeremy Harris, as
1342 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1343 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1344 redis_servers = needs to be configured which will be used by the redis
1345 lookup. Patch from Warren Baker, of The Packet Hub.
1347 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1349 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1350 hostname or reverse DNS when processing a host list. Used suggestions
1351 from multiple comments on this bug.
1353 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1355 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1356 Missed a few lines, added it to make the runtest require no keyboard
1359 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1360 contains upper case chars. Make router use caseful_local_part.
1362 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1363 support when GnuTLS has been built with p11-kit.
1369 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1371 This, or similar/improved, will also be change PP/11 of 4.82.
1377 PP/01 Handle short writes when writing local log-files.
1378 In practice, only affects FreeBSD (8 onwards).
1379 Bugzilla 1053, with thanks to Dmitry Isaikin.
1381 NM/01 Bugzilla 949 - Documentation tweak
1383 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1386 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1388 PP/02 Implemented gsasl authenticator.
1390 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1392 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1393 `pkg-config foo` for cflags/libs.
1395 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1396 with rest of GSASL and with heimdal_gssapi.
1398 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1399 `pkg-config foo` for cflags/libs for the TLS implementation.
1401 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1402 properties get this fed in as external SSF. A number of robustness
1403 and debugging improvements to the cyrus_sasl authenticator.
1405 PP/08 cyrus_sasl server now expands the server_realm option.
1407 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1408 Patch by Jeremy Harris.
1410 PP/10 Added dbmjz lookup type.
1412 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1414 PP/12 MAIL args handles TAB as well as SP, for better interop with
1415 non-compliant senders.
1416 Analysis and variant patch by Todd Lyons.
1418 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1419 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1420 Patch from Dirk Mueller <dmueller@suse.com>
1422 PP/13 tls_peerdn now print-escaped for spool files.
1423 Observed some $tls_peerdn in wild which contained \n, which resulted
1424 in spool file corruption.
1426 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1427 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1428 or write after TLS renegotiation, which otherwise led to messages
1431 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1432 as a tracking header (ie: a signed header comes before the signature).
1433 Patch from Wolfgang Breyha.
1435 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1436 comma-sep list; embedded commas doubled.
1438 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1440 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1442 Report and patch from Dmitry Banschikov.
1444 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1445 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1446 failures. We appear to now support TLS1.1+ with Exim.
1448 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1449 lets Exim select keys and certificates based upon TLS SNI from client.
1450 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1451 before an outbound SMTP session. New log_selector, +tls_sni.
1453 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1454 NULL dereference. Report and patch from Alun Jones.
1456 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1457 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1458 Not seeing resolver debug output on NetBSD, but suspect this is a
1459 resolver implementation change.
1461 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1462 Left warnings. Added "eximon gdb" invocation mode.
1464 PP/21 Defaulting "accept_8bitmime" to true, not false.
1466 PP/22 Added -bw for inetd wait mode support.
1468 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1469 locate the relevant includes and libraries. Made this the default.
1471 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1472 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1474 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1475 This may cause build issues on older platforms.
1477 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1478 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1479 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1480 Added SNI support via GnuTLS too.
1481 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1483 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1485 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1486 Applied second patch from Janne, implementing suggestion to default
1487 multiple-strings-in-record handling to match SPF spec.
1489 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1491 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1492 read-only, out of scope).
1493 Patch from Wolfgang Breyha, report from Stuart Northfield.
1495 PP/29 Fix three issues highlighted by clang analyser static analysis.
1496 Only crash-plausible issue would require the Cambridge-specific
1497 iplookup router and a misconfiguration.
1498 Report from Marcin Mirosław.
1500 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1502 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1503 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1504 As part of this, removing so much warning spew let me fix some minor
1505 real issues in debug logging.
1507 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1508 assignment on my part. Fixed.
1510 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1511 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1512 Janne Snabb (who went above and beyond: thank you).
1514 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1515 string otherwise requires a connection and a bunch more work and it's
1516 relatively easy to get wrong. Should also expose TLS library linkage
1519 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1520 64-bit ${eval} (JH/03).
1522 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1523 GNU libc to support some of the 64-bit stuff, should not lead to
1524 conflicts. Defined before os.h is pulled in, so if a given platform
1525 needs to override this, it can.
1527 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1528 protection layer was required, which is not implemented.
1529 Bugzilla 1254, patch from Wolfgang Breyha.
1531 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1532 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1533 tls_dhparam take prime identifiers. Also unbreak combination of
1534 OpenSSL+DH_params+TLSSNI.
1536 PP/39 Disable SSLv2 by default in OpenSSL support.
1542 PP/01 Solaris build fix for Oracle's LDAP libraries.
1543 Bugzilla 1109, patch from Stephen Usher.
1545 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1547 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1550 TF/02 Fix a couple more cases where we did not log the error message
1551 when unlink() failed. See also change 4.74-TF/03.
1553 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1554 lock up or crash if it happened to be inside a call to libc when it
1555 got a SIGUSR1 from exiwhat.
1557 The SIGUSR1 handler appends the current process status to the process
1558 log which is later printed by exiwhat. It used to use the general
1559 purpose logging code to do this, but several functions it calls are
1560 not safe for signals.
1562 The new output code in the SIGUSR1 handler is specific to the process
1563 log, and simple enough that it's easy to inspect for signal safety.
1564 Removing some special cases also simplifies the general logging code.
1565 Removing the spurious timestamps from the process log simplifies
1568 TF/04 Improved ratelimit ACL condition.
1570 The /noupdate option has been deprecated in favour of /readonly which
1571 has clearer semantics. The /leaky, /strict, and /readonly update modes
1572 are mutually exclusive. The update mode is no longer included in the
1573 database key; it just determines when the database is updated. (This
1574 means that when you upgrade Exim will forget old rate measurements.)
1576 Exim now checks that the per_* options are used with an update mode that
1577 makes sense for the current ACL. For example, when Exim is processing a
1578 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1579 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1580 must specify per_mail/readonly. If you omit the update mode it defaults to
1581 /leaky where that makes sense (as before) or /readonly where required.
1583 The /noupdate option is now undocumented but still supported for
1584 backwards compatibility. It is equivalent to /readonly except that in
1585 ACLs where /readonly is required you may specify /leaky/noupdate or
1586 /strict/noupdate which are treated the same as /readonly.
1588 A useful new feature is the /count= option. This is a generalization
1589 of the per_byte option, so that you can measure the throughput of other
1590 aggregate values. For example, the per_byte option is now equivalent
1591 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1593 The per_rcpt option has been generalized using the /count= mechanism
1594 (though it's more complicated than the per_byte equivalence). When it is
1595 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1596 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1597 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1598 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1599 Note that using per_rcpt with a non-readonly update mode in more than
1600 one ACL will cause the recipients to be double-counted. (The per_mail
1601 and per_byte options don't have this problem.)
1603 The handling of very low rates has changed slightly. If the computed rate
1604 is less than the event's count (usually one) then this event is the first
1605 after a long gap. In this case the rate is set to the same as this event's
1606 count, so that the first message of a spam run is counted properly.
1608 The major new feature is a mechanism for counting the rate of unique
1609 events. The new per_addr option counts the number of different
1610 recipients that someone has sent messages to in the last time period. It
1611 behaves like per_rcpt if all the recipient addresses are different, but
1612 duplicate recipient addresses do not increase the measured rate. Like
1613 the /count= option this is a general mechanism, so the per_addr option
1614 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1615 example, measure the rate that a client uses different sender addresses
1616 with the options per_mail/unique=$sender_address. There are further
1617 details in the main documentation.
1619 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1621 TF/06 Removed a few PCRE remnants.
1623 TF/07 Automatically extract Exim's version number from tags in the git
1624 repository when doing development or release builds.
1626 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1627 Bugzilla 879. Patch from Paul Fisher.
1629 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1630 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1633 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1635 PP/05 Variable $av_failed, true if the AV scanner deferred.
1636 Bugzilla 1078. Patch from John Horne.
1638 PP/06 Stop make process more reliably on build failure.
1639 Bugzilla 1087. Patch from Heiko Schlittermann.
1641 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1642 Bugzilla 1089. Patch from Heiko Schlittermann.
1644 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1645 Bugzilla 1131. Patch from Holger Weiß.
1647 PP/09 Handle IPv6 addresses with SPF.
1648 Bugzilla 860. Patch from Wolfgang Breyha.
1650 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1652 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1655 PP/11 match_* no longer expand right-hand-side by default.
1656 New compile-time build option, EXPAND_LISTMATCH_RHS.
1657 New expansion conditions, "inlist", "inlisti".
1659 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1661 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1663 PP/14 fix log_write() format string regression from TF/03.
1664 Bugzilla 1152. Patch from Dmitry Isaikin.
1670 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1672 PP/02 Harmonised TLS library version reporting; only show if debugging.
1673 Layout now matches that introduced for other libraries in 4.74 PP/03.
1675 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1677 PP/04 New "dns_use_edns0" global option.
1679 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1682 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1683 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1685 TK/01 Updated PolarSSL code to 0.14.2.
1686 Bugzilla 1097. Patch from Andreas Metzler.
1688 PP/07 Catch divide-by-zero in ${eval:...}.
1689 Fixes bugzilla 1102.
1691 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1694 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1695 format-string attack -- SECURITY: remote arbitrary code execution.
1697 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1698 time unintentionally subject to list matching rules, letting the header
1699 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1700 arbitrary string expansion). This allowed for information disclosure.
1702 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1703 INT_MIN/-1 -- value coerced to INT_MAX.
1709 NM/01 Workaround for PCRE version dependency in version reporting
1712 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1713 This fixes portability to compilers other than gcc, notably
1714 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1716 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1717 makefiles for portability to HP-UX and POSIX correctness.
1719 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1720 Also via indented variable definition in the Makefile.
1721 (Debugging by Oliver Heesakkers).
1723 PP/02 Restore caching of spamd results with expanded spamd_address.
1724 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1726 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1727 Improves build reliability. Fix from: Frank Elsner
1729 NM/02 Fix wide character breakage in the rfc2047 coding
1730 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1732 NM/03 Allow underscore in dnslist lookups
1733 Fixes bug 1026. Patch from Graeme Fowler
1735 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1736 Code patches from Adam Ciarcinski of NetBSD.
1738 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1741 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1742 is logged, to avoid truncation. Patch from John Horne.
1744 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1745 Patch from Jakob Hirsch.
1747 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1748 SQL string expansion failure details.
1749 Patch from Andrey Oktyabrski.
1751 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1752 Patch from Simon Arlott.
1754 PP/09 New lookups functionality failed to compile on old gcc which rejects
1755 extern declarations in function scope.
1756 Patch from Oliver Fleischmann
1758 PP/10 Use sig_atomic_t for flags set from signal handlers.
1759 Check getgroups() return and improve debugging.
1760 Fixed developed for diagnosis in bug 927 (which turned out to be
1763 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1764 Patch from Mark Zealey.
1766 PP/12 Bugzilla 1056: Improved spamd server selection.
1767 Patch from Mark Zealey.
1769 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1770 Based on patch from Heiko Schlittermann.
1772 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1773 Patch from Uwe Doering, sign-off by Michael Haardt.
1775 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1776 variable declaration deep within a block. Bug and patch from
1779 PP/15 lookups-Makefile IRIX compatibility coercion.
1781 PP/16 Make DISABLE_DKIM build knob functional.
1783 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1784 Patch by Simon Arlott
1786 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1787 variable argument macros. Our copy now differs from upstream.
1793 TF/01 Failure to get a lock on a hints database can have serious
1794 consequences so log it to the panic log.
1796 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1797 controlled using the smtp_confirmation log selector.
1799 TF/03 Include the error message when we fail to unlink a spool file.
1801 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1802 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1803 for maintaining out-of-tree patches for some time.
1805 PP/01 Bugzilla 139: Documentation and portability issues.
1806 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1807 Handle per-OS dynamic-module compilation flags.
1809 PP/02 Let /dev/null have normal permissions.
1810 The 4.73 fixes were a little too stringent and complained about the
1811 permissions on /dev/null. Exempt it from some checks.
1812 Reported by Andreas M. Kirchwitz.
1814 PP/03 Report version information for many libraries, including
1815 Exim version information for dynamically loaded libraries. Created
1816 version.h, now support a version extension string for distributors
1817 who patch heavily. Dynamic module ABI change.
1819 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1820 privilege escalation vulnerability whereby the Exim run-time user
1821 can cause root to append content of the attacker's choosing to
1824 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1827 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1828 If dropping privileges for untrusted macros, we disabled normal logging
1829 on the basis that it would fail; for the Exim run-time user, this is not
1830 the case, and it resulted in successful deliveries going unlogged.
1831 Fixed. Reported by Andreas Metzler.
1837 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1838 only prepend for the Resent-* case. Fixes regression introduced in
1839 Exim 4.70 by NM/22 for Bugzilla 607.
1841 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1842 increasing numbers of administrators be bitten by this.
1844 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1846 PP/03 Bugzilla 994: added openssl_options main configuration option.
1848 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1850 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1852 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1854 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1855 without a peer certificate, leading to a segfault because of an
1856 assumption that peers always have certificates. Be a little more
1857 paranoid. Problem reported by Martin Tscholak.
1859 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1860 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1861 NB: ClamAV planning to remove STREAM in "middle of 2010".
1862 CL also introduces -bmalware, various -d+acl logging additions and
1863 more caution in buffer sizes.
1865 PP/09 Implemented reverse_ip expansion operator.
1867 PP/10 Bugzilla 937: provide a "debug" ACL control.
1869 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1871 PP/12 Bugzilla 973: Implement --version.
1873 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1875 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1877 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1879 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1880 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1881 ignore trailing whitespace.
1883 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1885 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1888 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1889 Notification from Dr Andrew Aitchison.
1891 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1892 ExtendedDetectionInfo response format.
1893 Notification from John Horne.
1895 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1898 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1899 XSL and documented dependency on system catalogs, with examples of how
1902 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1905 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1906 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1907 configuration file which is writeable by the Exim user or group.
1909 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1910 of configuration files to cover files specified with the -C option if
1911 they are going to be used with root privileges, not just the default
1914 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1915 option (effectively making it always true).
1917 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1918 files to be used while preserving root privileges.
1920 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1921 that rogue child processes cannot use them.
1923 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1924 run-time user, instead of root.
1926 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1927 Exim run-time user without dropping privileges.
1929 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1930 result string, instead of calling string_vformat() twice with the same
1933 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1934 for other users. Others should always drop root privileges if they use
1935 -C on the command line, even for a whitelisted configure file.
1937 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1939 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1945 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1946 $data_path, and $header_path variables; fixed documentation bugs and
1949 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1950 exipick to access non-standard spools, including the "frozen" queue
1953 NM/01 Bugzilla 965: Support mysql stored procedures.
1954 Patch from Alain Williams
1956 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1958 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1959 Patch from Andreas Metzler
1961 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1962 Patch from Kirill Miazine
1964 NM/05 Bugzilla 671: Added umask to procmail example.
1966 JJ/03 installed exipick 20100323.0, fixing doc bug
1968 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1969 directory. Notification and patch from Dan Rosenberg.
1971 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1973 TK/02 Improve log output when DKIM signing operation fails.
1975 MH/01 Treat the transport option dkim_domain as a colon separated
1976 list, not as a single string, and sign the message with each element,
1977 omitting multiple occurences of the same signer.
1979 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1980 Bugzilla 985, 986. Patch by Simon Arlott
1982 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1983 Patch by Simon Arlott
1985 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1986 MBX locking. Notification from Dan Rosenberg.
1992 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1994 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1996 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1998 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2000 NM/04 Bugzilla 915: Flag broken perl installation during build.
2006 TK/01 Added patch by Johannes Berg that expands the main option
2007 "spamd_address" if it starts with a dollar sign.
2009 TK/02 Write list of recipients to X-Envelope-Sender header when building
2010 the mbox-format spool file for content scanning (suggested by Jakob
2013 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2014 (http://www.dcc-servers.net/) support via dccifd. Activated by
2015 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2017 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2018 by Mark Daniel Reidel <mr@df.eu>.
2020 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2021 When building exim an external PCRE library is now needed -
2022 PCRE is a system library on the majority of modern systems.
2023 See entry on PCRE_LIBS in EDITME file.
2025 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2026 conversation. Added nologin parameter to request.
2027 Patch contributed by Kirill Miazine.
2029 TF/01 Do not log submission mode rewrites if they do not change the address.
2031 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2033 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2034 log files in place. Contributed by Roberto Lima.
2036 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2038 TF/03 Bugzilla 615: When checking the local_parts router precondition
2039 after a local_part_suffix or local_part_prefix option, Exim now
2040 does not use the address's named list lookup cache, since this
2041 contains cached lookups for the whole local part.
2043 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2044 Robert Millan. Documentation is in experimental-spec.txt.
2046 TF/04 Bugzilla 668: Fix parallel build (make -j).
2048 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2050 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2051 Patch provided by Jan Srzednicki.
2053 TF/05 Leading white space used to be stripped from $spam_report which
2054 wrecked the formatting. Now it is preserved.
2056 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2057 that they are available at delivery time.
2059 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2061 TF/08 TLS error reporting now respects the incoming_interface and
2062 incoming_port log selectors.
2064 TF/09 Produce a more useful error message if an SMTP transport's hosts
2065 setting expands to an empty string.
2067 NM/06 Bugzilla 744: EXPN did not work under TLS.
2068 Patch provided by Phil Pennock.
2070 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2071 Patch provided by Richard Godbee.
2073 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2074 acl_smtp_notquit, added index entry.
2076 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2077 Patch provided by Eugene Bujak.
2079 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2080 accept(). Patch provided by Maxim Dounin.
2082 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2083 Patch provided by Phil Pennock.
2085 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2087 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2088 Patch provided by Brad "anomie" Jorsch.
2090 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2091 Patch provided by Dean Brooks.
2093 TK/05 Add native DKIM support (does not depend on external libraries).
2095 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2096 Patch provided by Graeme Fowler.
2098 NM/16 Bugzilla 851: Documentation example syntax fix.
2100 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2102 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2105 NM/19 Bugzilla 745: TLS version reporting.
2106 Patch provided by Phil Pennock.
2108 NM/20 Bugzilla 167: bool: condition support.
2109 Patch provided by Phil Pennock.
2111 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2112 clients. Patch provided by Phil Pennock.
2114 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2115 Patch provided by Brad "anomie" Jorsch.
2117 NM/23 Bugzilla 687: Fix misparses in eximstats.
2118 Patch provided by Heiko Schlittermann.
2120 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2121 Patch provided by Heiko Schlittermann.
2123 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2124 plus update to original patch.
2126 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2128 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2129 Patch provided by David Brownlee.
2131 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2133 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2135 NM/30 Bugzilla 888: TLS documentation bugfixes.
2137 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2139 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2140 Unlike the original bugzilla I have changed all shell scripts in src tree.
2142 NM/33 Bugzilla 898: Transport filter timeout fix.
2143 Patch by Todd Rinaldo.
2145 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2146 Patch by Serge Demonchaux.
2148 NM/35 Bugzilla 39: Base64 decode bug fixes.
2149 Patch by Jakob Hirsch.
2151 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2153 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2155 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2157 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2163 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2164 ALT-N's libdkim that I have put here:
2165 http://duncanthrax.net/exim-experimental/
2167 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2168 were called 'true' and it seems that C99 defines that as a reserved
2169 keyword to be used with 'bool' variable types. That means you could
2170 not include C99-style headers which use bools without triggering
2171 build errors in sieve.c.
2173 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2174 as mailq or other aliases. Changed the --help handling significantly
2175 to do whats expected. exim_usage() emits usage/help information.
2177 SC/01 Added the -bylocaldomain option to eximstats.
2179 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2181 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2183 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2189 PH/01 Another patch from the Sieve maintainer.
2191 PH/02 When an IPv6 address is converted to a string for single-key lookup
2192 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2193 dots are used instead of colons so that keys in lsearch files need not
2194 contain colons. This was done some time before quoting was made available
2195 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2196 (notated using the quote facility) so as to distinguish them from IPv4
2197 keys. This meant that lookups for IP addresses in host lists did not work
2198 for iplsearch lookups.
2200 This has been fixed by arranging for IPv6 addresses to be expressed with
2201 colons if the lookup type is iplsearch. This is not incompatible, because
2202 previously such lookups could never work.
2204 The situation is now rather anomalous, since one *can* have colons in
2205 ordinary lsearch keys. However, making the change in all cases is
2206 incompatible and would probably break a number of configurations.
2208 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2211 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2212 conversion specification without a maximum field width, thereby enabling
2213 a rogue spamd server to cause a buffer overflow. While nobody in their
2214 right mind would setup Exim to query an untrusted spamd server, an
2215 attacker that gains access to a server running spamd could potentially
2216 exploit this vulnerability to run arbitrary code as the Exim user.
2218 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2219 $primary_hostname instead of what libspf2 thinks the hosts name is.
2221 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2222 a directory entry by the name of the lookup key. Previously, if a
2223 symlink pointed to a non-existing file or a file in a directory that
2224 Exim lacked permissions to read, a lookup for a key matching that
2225 symlink would fail. Now it is enough that a matching directory entry
2226 exists, symlink or not. (Bugzilla 503.)
2228 PH/03 The body_linecount and body_zerocount variables are now exported in the
2231 PH/04 Added the $dnslist_matched variable.
2233 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2234 This means they are set thereafter only if the connection becomes
2237 PH/06 Added the client_condition to authenticators so that some can be skipped
2238 by clients under certain conditions.
2240 PH/07 The error message for a badly-placed control=no_multiline_responses left
2241 "_responses" off the end of the name.
2243 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2245 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2246 (without spaces) instead of just copying the configuration text.
2248 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2250 PH/11 Added $max_received_linelength.
2252 PH/12 Added +ignore_defer and +include_defer to host lists.
2254 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2255 way in which PCRE > 7.0 is built.
2257 PH/14 Implemented queue_only_load_latch.
2259 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2260 MAIL command. The effect was to mangle the value on 64-bit systems.
2262 PH/16 Another patch from the Sieve maintainer.
2264 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2266 PH/18 If a system quota error occurred while trying to create the file for
2267 a maildir delivery, the message "Mailbox is full" was not appended to the
2268 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2269 only to a quota excession during the actual writing of the file.
2271 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2272 characters?) which causes problems in log lines. The DN values are now
2273 passed through string_printing() before being added to log lines.
2275 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2276 and InterBase are left for another time.)
2278 PH/21 Added message_body_newlines option.
2280 PH/22 Guard against possible overflow in moan_check_errorcopy().
2282 PH/23 POSIX allows open() to be a macro; guard against that.
2284 PH/24 If the recipient of an error message contained an @ in the local part
2285 (suitably quoted, of course), incorrect values were put in $domain and
2286 $local_part during the evaluation of errors_copy.
2292 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2293 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2296 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2297 issue a MAIL command.
2299 PH/02 In an ACL statement such as
2301 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2303 if a client was not listed at all, or was listed with a value other than
2304 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2305 the condition was not true (as it should be), so access was not denied.
2306 The bug was that the ! inversion was incorrectly passed on to the second
2307 item. This has been fixed.
2309 PH/03 Added additional dnslists conditions == and =& which are different from
2310 = and & when the dns lookup returns more than one IP address.
2312 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2313 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2315 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2316 FSYNC, which compiles an option called disable_fsync that allows for
2317 bypassing fsync(). The documentation is heavily laced with warnings.
2319 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2321 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2322 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2323 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2324 including adding "make clean"; (3) Added -fPIC when compiling the test
2325 dynamically loaded module, to get rid of a warning.
2327 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2328 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2329 = 0s. The bug is otherwise harmless.
2331 PH/07 There was a bug in the dovecot authenticator such that the value of
2332 $auth1 could be overwritten, and so not correctly preserved, after a
2333 successful authentication. This usually meant that the value preserved by
2334 the server_setid option was incorrect.
2336 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2338 PH/09 Installed PCRE release 7.0.
2340 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2341 run for batched SMTP input. It is now run at the start of every message
2342 in the batch. While fixing this I discovered that the process information
2343 (output by running exiwhat) was not always getting set for -bs and -bS
2344 input. This is fixed, and it now also says "batched" for BSMTP.
2346 PH/11 Added control=no_pipelining.
2348 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2349 patch, slightly modified), and move the expansion of helo_data till after
2350 the connection is made in the smtp transport (so it can use these
2353 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2355 PH/14 Added log_selector = +pid.
2357 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2359 PH/16 Add ${if forany and ${if forall.
2361 PH/17 Added dsn_from option to vary the From: line in DSNs.
2363 PH/18 Flush SMTP output before performing a callout, unless control =
2364 no_callout_flush is set.
2366 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2367 was true (the default) a successful delivery failed to delete the retry
2368 item, thus causing premature timeout of the address. The bug is now
2371 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2373 PH/21 Long custom messages for fakedefer and fakereject are now split up
2374 into multiline responses in the same way that messages for "deny" and
2375 other ACL rejections are.
2377 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2378 with slight modification.
2380 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2381 draft, changing the syntax and factoring some duplicate code".
2383 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2384 for deliveries of the second and subsequent messages over the same SMTP
2387 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2388 ${reduce, with only minor "tidies".
2390 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2392 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2393 expansion side effects.
2395 PH/27 When a message times out after an over-quota error from an Exim-imposed
2396 quota, the bounce message says "mailbox is full". This message was not
2397 being given when it was a system quota that was exceeded. It now should
2400 MH/03 Made $recipients available in local_scan(). local_scan() already has
2401 better access to the recipient list through recipients_list[], but
2402 $recipients can be useful in postmaster-provided expansion strings.
2404 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2405 in the case of a MAIL command with additional options following the
2406 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2407 were accidentally chopped off.
2409 PH/29 SMTP synchronization checks are implemented when a command is read -
2410 there is a check that no more input is waiting when there shouldn't be
2411 any. However, for some commands, a delay in an ACL can mean that it is
2412 some time before the response is written. In this time, more input might
2413 arrive, invalidly. So now there are extra checks after an ACL has run for
2414 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2415 pipelining has not been advertised.
2417 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2419 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2420 returned in the SMTP response when smtp_return_error_details was false.
2421 This has been fixed.
2423 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2424 instead of the C I/O that was originally supplied, because problems were
2425 reported on Solaris.
2427 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2428 Exim which did not show up earlier: it was assuming that a call to
2429 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2430 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2431 was a macro that became an assignment, so it seemed to work. This has
2432 changed to a proper function call with a void return, hence the compile
2433 error. Exim's code has been fixed.
2435 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2438 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2440 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2442 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2443 because it turns out to be too big or there is a timeout) there is no
2444 "Completed" line in the log. When some messages of this type were
2445 selected by exigrep, they were listed as "not completed". Others were
2446 picked up by some special patterns. I have improved the selection
2447 criteria to be more general.
2449 PH/38 The host_find_failed option in the manualroute router can now be set
2450 to "ignore", to completely ignore a host whose IP address cannot be
2451 found. If all hosts are ignored, the behaviour is controlled by the new
2452 host_all_ignored option.
2454 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2455 homing or because of multiple MX records with /mx) generated more than
2456 one IP address, and the following item turned out to be the local host,
2457 all the secondary addresses of the first item were incorrectly removed
2458 from the list, along with the local host and any following hosts (which
2459 is what is supposed to happen).
2461 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2462 whoever called Exim into the -H file. In the case of the daemon it was
2463 behaving confusingly. When first started, it used values for whoever
2464 started the daemon, but after a SIGHUP it used the Exim user (because it
2465 calls itself on a restart). I have changed the code so that it now always
2468 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2469 message are rejected with the same error (e.g. no authentication or bad
2470 sender address), and a DATA command is nevertheless sent (as can happen
2471 with PIPELINING or a stupid MUA), the error message that was given to the
2472 RCPT commands is included in the rejection of the DATA command. This is
2473 intended to be helpful for MUAs that show only the final error to their
2476 PH/42 Another patch from the Sieve maintainer.
2478 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2479 Eximstats - Fixed some broken HTML links and added missing column headers
2481 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2482 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2484 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2486 PH/43 Yet another patch from the Sieve maintainer.
2488 PH/44 I found a way to check for a TCP/IP connection going away before sending
2489 the response to the final '.' that terminates a message, but only in the
2490 case where the client has not sent further data following the '.'
2491 (unfortunately, this is allowed). However, in many cases there won't be
2492 any further data because there won't be any more messages to send. A call
2493 to select() can be used: if it shows that the input is "ready", there is
2494 either input waiting, or the socket has been closed. An attempt to read
2495 the next input character can distinguish the two cases. Previously, Exim
2496 would have sent an OK response which the client would never have see.
2497 This could lead to message repetition. This fix should cure that, at
2498 least in a lot of common cases.
2500 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2501 advertised in response to EHLO.
2507 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2508 fixed by 4.65/MH/01 (is this a record?) are fixed:
2510 (i) An empty string was always treated as zero by the numeric comparison
2511 operators. This behaviour has been restored.
2513 (ii) It is documented that the numeric comparison operators always treat
2514 their arguments as decimal numbers. This was broken in that numbers
2515 starting with 0 were being interpreted as octal.
2517 While fixing these problems I realized that there was another issue that
2518 hadn't been noticed. Values of message_size_limit (both the global option
2519 and the transport option) were treated as octal if they started with 0.
2520 The documentation was vague. These values are now always treated as
2521 decimal, and I will make that clear in the documentation.
2527 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2528 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2531 MH/01 Don't check that the operands of numeric comparison operators are
2532 integers when their expansion is in "skipping" mode (fixes bug
2533 introduced by 4.64-PH/07).
2535 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2536 child addresses, Exim now panics and dies. Previously, because the count
2537 is held in a short int, deliveries were likely to be lost. As such a
2538 large number of recipients for a single message is ridiculous
2539 (performance will be very, very poor), I have chosen to impose a limit
2540 rather than extend the field.
2546 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2547 leftover -K file (the existence of which was triggered by #402).
2548 While we were at it, introduced process PID as part of the -K
2549 filename. This should rule out race conditions when creating
2552 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2553 processing considerably. Previous code took too long for large mails,
2554 triggering a timeout which in turn triggers #401.
2556 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2557 in the DK code in transports.c. sendfile() is not really portable,
2558 hence the _LINUX specificness.
2560 TF/01 In the add_headers option to the mail command in an Exim filter,
2561 there was a bug that Exim would claim a syntax error in any
2562 header after the first one which had an odd number of characters
2565 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2566 callout verification, Exim cached a "reject" for the entire domain. This
2567 is correct for most verifications, but it is not correct for a recipient
2568 verification with use_sender or use_postmaster set, because in that case
2569 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2570 case of MAIL FROM:<> rejection from other early rejections (e.g.
2571 rejection of HELO). When verifying a recipient using a non-null MAIL
2572 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2573 Whatever the result of the callout, the value of the domain cache is
2574 left unchanged (for any other kind of callout, getting as far as trying
2575 RCPT means that the domain itself is ok).
2577 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2580 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2581 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2582 session, the daemon ignores ECONNECT errors and logs others; it now
2583 ignores EPIPE as well.
2585 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2586 (quoted-printable decoding).
2588 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2589 later the small subsequent patch to fix an introduced bug.
2591 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2593 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2595 PH/08 An error is now given if message_size_limit is specified negative.
2597 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2598 to be given (somewhat) arbitrary names.
2600 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2603 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2604 miscellaneous code fixes
2606 PH/10 Added the log_reject_target ACL modifier to specify where to log
2609 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2610 hostname. This is wrong, because it relates to the incoming message (and
2611 probably the interface on which it is arriving) and not to the outgoing
2612 callout (which could be using a different interface). This has been
2613 changed to use the value of the helo_data option from the smtp transport
2614 instead - this is what is used when a message is actually being sent. If
2615 there is no remote transport (possible with a router that sets up host
2616 addresses), $smtp_active_hostname is used.
2618 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2619 tweaks were necessary in order to get it to work (see also 21 below):
2620 (a) The code assumed that strncpy() returns a negative number on buffer
2621 overflow, which isn't the case. Replaced with Exim's string_format()
2623 (b) There were several signed/unsigned issues. I just did the minimum
2624 hacking in of casts. There is scope for a larger refactoring.
2625 (c) The code used strcasecmp() which is not a standard C function.
2626 Replaced with Exim's strcmpic() function.
2627 (d) The code set only $1; it now sets $auth1 as well.
2628 (e) A simple test gave the error "authentication client didn't specify
2629 service in request". It would seem that Dovecot has changed its
2630 interface. Fortunately there's a specification; I followed it and
2631 changed what the client sends and it appears to be working now.
2633 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2636 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2637 address A is aliased to B and C, where B exists and C does not. Without
2638 -v the output is "A verified" because verification stops after a
2639 successful redirection if more than one address is generated. However,
2640 with -v the child addresses are also verified. Exim was outputting "A
2641 failed to verify" and then showing the successful verification for C,
2642 with its parentage. It now outputs "B failed to verify", showing B's
2643 parentage before showing the successful verification of C.
2645 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2646 look up a TXT record in a specific list after matching in a combined
2649 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2650 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2651 they consult the DNS. I had assumed they would set it the way they
2652 wanted; and indeed my experiments on Linux seem to show that in some
2653 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2654 To be on the safe side, however, I have now made the interface to
2655 host_find_byname() similar to host_find_bydns(), with an argument
2656 containing the DNS resolver options. The host_find_byname() function now
2657 sets these options at its start, just as host_find_bydns() does. The smtp
2658 transport options dns_qualify_single and dns_search_parents are passed to
2659 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2660 of host_find_byname() use the default settings of RES_DEFNAMES
2661 (qualify_single) but not RES_DNSRCH (search_parents).
2663 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2664 spool_read_header() do less string testing, by means of a preliminary
2665 switch on the second character of optional "-foo" lines. (This is
2666 overdue, caused by the large number of possibilities that now exist.
2667 Originally there were few.) While I was there, I also converted the
2668 str(n)cmp tests so they don't re-test the leading "-" and the first
2669 character, in the hope this might squeeze out yet more improvement.
2671 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2672 flag allowing group syntax was set by the header_syntax check but not
2673 turned off, possible causing trouble later; (2) The flag was not being
2674 set at all for the header_verify test, causing "group"-style headers to
2675 be rejected. I have now set it in this case, and also caused header_
2676 verify to ignore an empty address taken from a group. While doing this, I
2677 came across some other cases where the code for allowing group syntax
2678 while scanning a header line wasn't quite right (mostly, not resetting
2679 the flag correctly in the right place). These bugs could have caused
2680 trouble for malformed header lines. I hope it is now all correct.
2682 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2683 with the "reply" argument non-NULL. The code, however (which originally
2684 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2685 but it didn't always do it. This confused somebody who was copying the
2686 code for some other use. I have removed all the tests.
2688 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2689 feature that was used to support insecure browsers during the U.S. crypto
2690 embargo. It requires special client support, and Exim is probably the
2691 only MTA that supported it -- and would never use it because real RSA is
2692 always available. This code has been removed, because it had the bad
2693 effect of slowing Exim down by computing (never used) parameters for the
2694 RSA_EXPORT functionality.
2696 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2697 authenticator to fail if there's a tab character in the incoming data
2698 (there should never be unless someone is messing about, as it's supposed
2699 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2700 if the connection is using TLS or if the remote IP is the same as the
2701 local IP, and the "valid-client-cert option" if a client certificate has
2704 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2705 authenticators. This can be used for authorization after authentication
2706 succeeds. (In the case of plaintext, it servers for both authentication
2709 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2710 if any retry times were supplied.
2712 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2713 connection, where there is no client IP address to check. In this
2714 situation, the verify now always succeeds.
2716 PH/25 Applied John Jetmore's -Mset patch.
2718 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2720 PH/27 In a string expansion for a processed (not raw) header when multiple
2721 headers of the same name were present, leading whitespace was being
2722 removed from all of them, but trailing whitespace was being removed only
2723 from the last one. Now trailing whitespace is removed from each header
2724 before concatenation. Completely empty headers in a concatenation (as
2725 before) are ignored.
2727 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2728 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2730 PH/29 [Removed. This was a change that I later backed out, and forgot to
2731 correct the ChangeLog entry (that I had efficiently created) before
2732 committing the later change.]
2734 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2735 address errors (4xx response to RCPT) over the same connection as other
2736 messages routed to the same hosts. Such deliveries are always "forced",
2737 so retry times are not inspected. This resulted in far too many retries
2738 for the affected addresses. The effect occurred only when there were more
2739 hosts than the hosts_max_try setting in the smtp transport when it had
2740 the 4xx errors. Those hosts that it had tried were not added to the list
2741 of hosts for which the message was waiting, so if all were tried, there
2742 was no problem. Two fixes have been applied:
2744 (i) If there are any address or message errors in an SMTP delivery, none
2745 of the hosts (tried or untried) are now added to the list of hosts
2746 for which the message is waiting, so the message should not be a
2747 candidate for sending over the same connection that was used for a
2748 successful delivery of some other message. This seems entirely
2749 reasonable: after all the message is NOT "waiting for some host".
2750 This is so "obvious" that I'm not sure why it wasn't done
2751 previously. Hope I haven't missed anything, but it can't do any
2752 harm, as the worst effect is to miss an optimization.
2754 (ii) If, despite (i), such a delivery is accidentally attempted, the
2755 routing retry time is respected, so at least it doesn't keep
2756 hammering the server.
2758 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2759 in ${readsocket because some servers need this prod.
2761 PH/32 Added some extra debug output when updating a wait-xxx database.
2763 PH/33 The hint "could be header name not terminated by colon", which has been
2764 given for certain expansion errors for a long time, was not being given
2765 for the ${if def:h_colon_omitted{... case.
2767 PH/34 The spec says: "With one important exception, whenever a domain list is
2768 being scanned, $domain contains the subject domain." There was at least
2769 one case where this was not true.
2771 PH/35 The error "getsockname() failed: connection reset by peer" was being
2772 written to the panic log as well as the main log, but it isn't really
2773 panic-worthy as it just means the connection died rather early on. I have
2774 removed the panic log writing for the ECONNRESET error when getsockname()
2777 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2778 runs only) independently of the message's sender address. This meant
2779 that, if the 4xx error was in fact related to the sender, a different
2780 message to the same recipient with a different sender could confuse
2781 things. In particular, this can happen when sending to a greylisting
2782 server, but other circumstances could also provoke similar problems.
2783 I have changed the default so that the retry time for these errors is now
2784 based a combination of the sender and recipient addresses. This change
2785 can be overridden by setting address_retry_include_sender=false in the
2788 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2789 remote server are returned as part of bounce messages. This was not
2790 happening for LMTP over a pipe (the lmtp transport), but now it is the
2791 same for both kinds of LMTP.
2793 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2794 in header lines that were in fact CNAMEs. This is no longer the case.
2796 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2797 and queue runs started by the daemon processed all messages. This has
2798 been fixed so that -R and -S can now usefully be given with -q<time>.
2800 PH/40 Import PCRE release 6.7 (fixes some bugs).
2802 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2804 PH/42 Give an error if -q is specified more than once.
2806 PH/43 Renamed the variables $interface_address and $interface_port as
2807 $received_ip_address and $received_port, to make it clear that these
2808 values apply to message reception, and not to the outgoing interface when
2809 a message is delivered. (The old names remain recognized, of course.)
2811 PH/44 There was no timeout on the connect() call when using a Unix domain
2812 socket in the ${readsocket expansion. There now is.
2814 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2815 be meaningful with "accept".
2817 SC/01 Eximstats V1.43
2818 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2820 SC/02 Eximstats V1.44
2821 Use a glob alias rather than an array ref in the generated
2822 parser. This improves both readability and performance.
2824 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2825 Collect SpamAssassin and rejection statistics.
2826 Don't display local sender or destination tables unless
2827 there is data to show.
2828 Added average volumes into the top table text output.
2830 SC/04 Eximstats V1.46
2831 Collect data on the number of addresses (recipients)
2832 as well as the number of messages.
2834 SC/05 Eximstats V1.47
2835 Added 'Message too big' to the list of mail rejection
2836 reasons (thanks to Marco Gaiarin).
2838 SC/06 Eximstats V1.48
2839 Mainlog lines which have GMT offsets and are too short to
2840 have a flag are now skipped.
2842 SC/07 Eximstats V1.49 (Alain Williams)
2843 Added the -emptyok flag.
2845 SC/08 Eximstats V1.50
2846 Fixes for obtaining the IP address from reject messages.
2848 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2849 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2850 whitespace changes from 4.64-PH/27
2852 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2855 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2856 are found, allow negative numbers in numeric criteria)
2858 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2860 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2861 to match changes made in 4.64-PH/43
2863 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2865 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2866 "domain" is actually an IP address, and force a failure. This locks out
2867 those revolvers/nameservers that support "A-for-A" lookups, in
2868 contravention of the specifications.
2870 PH/48 When a host name was looked up from an IP address, and the subsequent
2871 forward lookup of the name timed out, the host name was left in
2872 $sender_host_name, contrary to the specification.
2874 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2875 restricted to single-key lookups, Exim was not diagnosing an error if
2876 * or *@ was used with a query-style lookup.
2878 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2880 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2881 long ago, but noone interested enough thought of it. Let's just say that
2882 the "1.1" means that there are some new functions that weren't there at
2883 some point in the past.
2885 PH/51 Error processing for expansion failure of helo_data from an smtp
2886 transport during callout processing was broken.
2888 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2889 tested/used via the -bh/-bhc/-bs options.
2891 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2892 bug, fixed in subsequent PCRE releases).
2894 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2895 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2897 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2903 SC/01 Use a glob alias rather than an array ref in eximstats generated
2904 parser. This improves both readability and performance.
2906 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2907 Don't display local sender or destination tables in eximstats unless
2908 there is data to show.
2909 Added average volumes into the eximstats top table text output.
2911 SC/03 Collect data on the number of addresses (recipients) as well
2912 as the number of messages in eximstats.
2914 TF/01 Correct an error in the documentation for the redirect router. Exim
2915 does (usually) call initgroups() when daemonizing.
2917 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2918 with consistent privilege compared to when running as a daemon.
2920 TF/03 Note in the spec that $authenticated_id is not set for local
2921 submissions from trusted users.
2923 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2924 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2926 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2927 by adding some example configuration directives to the default
2928 configuration file. A little bit of work is required to uncomment the
2929 directives and define how usernames and passwords are checked, but
2930 there is now a framework to start from.
2932 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2933 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2934 without this. I don't know how relevant this is to other LDAP libraries.
2936 PH/02 Add the verb name to the "unknown ACL verb" error.
2938 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2940 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2942 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2943 directory not expanded when it should be if an expanded home directory
2944 was set for the address (which is overridden by the transport).
2946 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2949 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2950 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2951 because it is too late at that time, and has no effect.
2953 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2954 security issue with \' (bugzilla #107). I could not use the
2955 PQescapeStringConn() function, because it needs a PGconn value as one of
2958 PH/08 When testing addresses using -bt, indicate those final addresses that
2959 are duplicates that would not cause an additional delivery. At least one
2960 person was confused, thinking that -bt output corresponded to deliveries.
2961 (Suppressing duplicates isn't a good idea as you lose the information
2962 about possibly different redirections that led to the duplicates.)
2964 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2965 systems where poll() doesn't work, in particular OS X.
2967 PH/10 Added more information to debugging output for retry time not reached.
2969 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2970 operations in malware.c.
2972 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2975 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2976 syslog_duplication set false, log lines that would normally be written
2977 both the the main log and to the reject log were not written to syslog at
2980 PH/14 In the default configuration, change the use of "message" in ACL warn
2981 statements to "add_header".
2983 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2984 not followed by a command (e.g. "seen endif").
2986 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2987 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2990 PH/17 Added extra conditions to the default value of delay_warning_condition
2994 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2995 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2996 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2999 The Auto-Submitted: and various List- headers are standardised, whereas I
3000 don't think Precedence: ever was.
3002 PH/18 Refactored debugging code in route_finduser() to show more information,
3003 in particular, the error code if getpwnam() issues one.
3005 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3006 This is apparently needed in addition to the PH/07 change above to avoid
3007 any possible encoding problems.
3009 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3010 but not after initializing Perl.
3012 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3013 output them only if debugging. By default they are written stderr,
3014 apparently, which is not desirable.
3016 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3019 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3022 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3024 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3025 authenticated or an ident call has been made. Suppress the default
3026 values for $authenticated_id and $authenticated_sender (but permit -oMai
3027 and -oMas) when testing with -bh.
3029 PH/24 Re-jigged the order of the tests in the default configuration so that the
3030 tests for valid domains and recipients precede the DNS black list and CSA
3031 tests, on the grounds that those ones are more expensive.
3033 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3034 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3035 This bug exists in every version of Exim that I still have, right back to
3038 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3039 However, an attempt to turn on case-sensitivity in a regex key by
3040 including (?-i) didn't work because the subject string was already
3041 lowercased, and the effects were non-intuitive. It turns out that a
3042 one-line patch can be used to allow (?-i) to work as expected.
3048 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3049 other effects) broke the use of negated acl sub-conditions.
3051 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3054 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3055 "Deny" causes Exim to reject the incoming connection with a 554 error.
3056 Unfortunately, if there is a major crisis, such as a disk failure,
3057 tcp-wrappers gives "deny", whereas what one would like would be some
3058 kind of temporary error. A kludge has been added to help with this.
3059 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3060 554 error is used if errno is still zero or contains ENOENT (which occurs
3061 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3064 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3066 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3067 errors. Otherwise a message that provokes a temporary error (when other
3068 messages do not) can cause a whole host to time out.
3070 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3071 addresses were routed directly to files or pipes from a redirect router.
3072 File deliveries just didn't batch; pipe deliveries might have suffered
3075 PH/06 A failure to get a lock for a hints database would erroneously always say
3076 "Failed to get write lock", even when it was really a read lock.
3078 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3079 of 0600. This has been changed to use the value of the lockfile_mode
3080 option (which defaults to 0600).
3082 PH/08 Applied small patch from the Sieve maintainer.
3084 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3085 folder from quota calculations, a direct delivery into this folder messed
3086 up the contents of the maildirsize file. This was because the regex was
3087 used only to exclude .Trash (or whatever) when the size of the mailbox
3088 was calculated. There was no check that a delivery was happening into an
3089 excluded directory. This bug has been fixed by ignoring all quota
3090 processing for deliveries into excluded directories.
3092 PH/10 Added the maildirfolder_create_regex option to appendfile.
3098 PH/01 The code for finding all the local interface addresses on a FreeBSD
3099 system running IPv6 was broken. This may well have applied to all BSD
3100 systems, as well as to others that have similar system calls. The broken
3101 code found IPv4 interfaces correctly, but gave incorrect values for the
3102 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3103 that it would not match correctly against @[] and not recognize the IPv6
3106 PH/02 The ipliteral router was not recognizing addresses of the form user@
3107 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3109 PH/03 Added disable_ipv6.
3111 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3112 decoded form, because it is most often used to construct To: headers
3113 lines in autoreplies, and the decoded form may well be syntactically
3114 invalid. However, $reply_address has leading white space removed, and all
3115 newlines turned into spaces so that the autoreply transport does not
3118 PH/05 If group was specified without a user on a router, and no group or user
3119 was specified on a transport, the group from the router was ignored.
3121 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3122 for visible compile-time settings that can be used to change these
3123 numbers, for those that want even more. Backwards compatibility with old
3124 spool files has been maintained. However, going back to a previous Exim
3125 release will lost any variables that are in spool files.
3127 PH/07 Two small changes when running in the test harness: increase delay when
3128 passing a TCP/IP connection to a new process, in case the original
3129 process has to generate a bounce, and remove special handling of
3130 127.0.0.2 (sic), which is no longer necessary.
3132 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3133 be the same on different OS.
3135 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3138 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3139 whether --show-vars was specified or not
3141 JJ/02 exipick: Added support for new ACL variable spool format introduced
3144 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3145 syntactically invalid From: or Reply-to: line, and a filter used this to
3146 generate an autoreply, and therefore failed to obtain an address for the
3147 autoreply, Exim could try to deliver to a non-existent relative file
3148 name, causing unrelated and misleading errors. What now happens is that
3149 it logs this as a hard delivery error, but does not attempt to create a
3152 PH/11 The exinext utility has a -C option for testing purposes, but although
3153 the given file was scanned by exinext itself; it wasn't being passed on
3154 when Exim was called.
3156 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3157 an end-of-file indication when reading a command response.
3159 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3160 compiled. In many other places in Exim, IPv6 addresses are always
3161 recognized, so I have changed this. It also means that IPv4 domain
3162 literals of the form [IPV4:n.n.n.n] are now always recognized.
3164 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3165 used if the router is not running as root, for example, when verifying at
3166 ACL time, or when using -bh. The debugging output from this situation was
3167 non-existent - all you got was a failure to exec. I have made two
3170 (a) Failures to set uid/gid, the current directory, or a process leader
3171 in a subprocess such as that created by queryprogram now generate
3172 suitable debugging output when -d is set.
3174 (b) The queryprogram router detects when it is not running as root,
3175 outputs suitable debugging information if -d is set, and then runs
3176 the subprocess without attempting to change uid/gid.
3178 PH/15 Minor change to Makefile for building test_host (undocumented testing
3181 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3182 additional section of a DNS packet that returns MX or SRV records.
3183 Instead, it always explicitly searches for A/AAAA records. This avoids
3184 major problems that occur when a DNS server includes only records of one
3185 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3186 fixed another bug: if SRV records were looked up and the corresponding
3187 address records were *not* found in the additional section, the port
3188 values from the SRV records were lost.
3190 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3191 using the correct key (the original address) when searching the retry
3192 rules in order to find which one to use for generating the retry hint.
3194 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3195 adding the default one. Similarly, if it contains a Reply-To: header, the
3196 errors_reply_to option, if set, is not used.
3198 PH/19 When calculating a retry time, Exim used to measure the "time since
3199 failure" by looking at the "first failed" field in the retry record. Now
3200 it does not use this if it is later than than the arrival time of the
3201 message. Instead it uses the arrival time. This makes for better
3202 behaviour in cases where some deliveries succeed, thus re-setting the
3203 "first failed" field. An example is a quota failure for a huge message
3204 when small messages continue to be delivered. Without this change, the
3205 "time since failure" will always be short, possible causing more frequent
3206 delivery attempts for the huge message than are intended.
3207 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3209 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3210 $1, $2, $3) because the numerical variables can be reset during some
3211 expansion items (e.g. "match"), thereby losing the authentication data.
3213 PH/21 Make -bV show the size of off_t variables so that the test suite can
3214 decide whether to run tests for quotas > 2G.
3216 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3217 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3218 filecount value is greater than 2G or if a quota value is greater than 2G
3219 on a system where the size of off_t is not greater than 4, a panic error
3222 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3223 never match. The debug and -bh output now contains an explicit error
3224 message indicating a malformed IPv4 address or mask.
3226 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3227 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3228 PH/23 above applies.
3230 PH/25 Do not write to syslog when running in the test harness. The only
3231 occasion when this arises is a failure to open the main or panic logs
3232 (for which there is an explicit test).
3234 PH/26 Added the /no_tell option to "control=freeze".
3236 PH/27 If a host name lookup failed very early in a connection, for example, if
3237 the IP address matched host_lookup and the reverse lookup yielded a name
3238 that did not have a forward lookup, an error message of the form "no IP
3239 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3240 could be logged. Now it outputs the IP address instead of "NULL".
3242 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3243 allows the sender and the authenticated sender to be set when
3244 submitting a message from within Exim. Since child_open_exim() is
3245 documented for local_scan(), the new function should be too.
3247 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3248 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3249 results in an empty string is now treated as unset.
3251 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3253 PH/31 Added sender_verify_fail logging option.
3255 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3256 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3257 tidied the source and removed it altogether.
3259 PH/33 When a queue run was abandoned because the load average was too high, a
3260 log line was always written; now it is written only if the queue_run log
3261 selector is set. In addition, the log line for abandonment now contains
3262 information about the queue run such as the pid. This is always present
3263 in "start" and "stop" lines but was omitted from the "abandon" line.
3265 PH/34 Omit spaces between a header name and the colon in the error message that
3266 is given when verify = headers_syntax fails (if there are lots of them,
3267 the message gets confusing).
3269 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3270 names, as there are now some PTR records that contain slashes. This check
3271 is only to protect against broken name servers that fall over on strange
3272 characters, so the fact that it applies to all lookups doesn't matter.
3274 PH/36 Now that the new test suite is complete, we can remove some of the
3275 special code in Exim that was needed for the old test suite. For example,
3276 sorting DNS records because real resolvers return them in an arbitrary
3277 order. The new test suite's fake resolver always returns records in the
3280 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3281 bounces) except when queue_only is set, to avoid logging races between
3282 the different processes.
3284 PH/38 Panic-die if .include specifies a non-absolute path.
3286 PH/39 A tweak to the "H" retry rule from its user.
3288 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3289 a label. They prevented compilation on older perls.
3291 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3292 a warning to be raised on newish perls.
3294 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3295 on queue. Changes to match documented behaviour of showing count of
3296 messages matching specified criteria.
3298 PH/40 Changed the default ident timeout from 30s to 5s.
3300 PH/41 Added support for the use of login_cap features, on those BSD systems
3301 that have them, for controlling the resources used by pipe deliveries.
3303 PH/42 The content-scanning code uses fopen() to create files in which to put
3304 message data. Previously it was not paying any attention to the mode of
3305 the files. Exim runs with umask(0) because the rest of the code creates
3306 files with open(), and sets the required mode explicitly. Thus, these
3307 files were ending up world-writeable. This was not a big issue, because,
3308 being within the spool directory, they were not world-accessible. I have
3309 created a function called modefopen, which takes an additional mode
3310 argument. It sets umask(777), creates the file, chmods it to the required
3311 mode, then resets the umask. All the relevant calls to fopen() in the
3312 content scanning code have been changed to use this function.
3314 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3315 to 24 hours. This avoids potential overflow problems when processing G
3316 and H retry rules. I suspect nobody ever tinkers with this value.
3318 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3320 PH/45 When the plaintext authenticator is running as a client, the server's
3321 challenges are checked to ensure they are valid base64 strings. By
3322 default, the authentication attempt is cancelled if an invalid string is
3323 received. Setting client_ignore_invalid_base64 true ignores these errors.
3324 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3325 they are received. Thus, the responses can be made to depend on the
3326 challenges. If an invalid string is ignored, an empty string is placed in
3329 PH/46 Messages that are created by the autoreply transport now contains a
3330 References: header, in accordance with RFCs 2822 and 3834.
3332 PH/47 Added authenticated_sender_force to the smtp transport.
3334 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3336 PH/49 Installed latest patch from the Sieve maintainer.
3338 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3339 was also set, the appendfile transport was unnecessarily scanning a
3340 directory of message files (e.g. for maildir delivery) to find the count
3341 of files (along with the size), even though it did not need this
3342 information. It now does the scan only if it needs to find either the
3343 size of the count of files.
3345 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3347 PH/52 Two bugs concerned with error handling when the smtp transport is
3350 (i) Exim was not creating retry information for temporary errors given
3351 for individual recipients after the DATA command when the smtp transport
3352 was used in LMTP mode. This meant that they could be retried too
3353 frequently, and not timed out correctly.
3355 (ii) Exim was setting the flag that allows error details to be returned
3356 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3357 recipients that were returned after the DATA command.
3359 PH/53 This is related to PH/52, but is more general: for any failing address,
3360 when detailed error information was permitted to be returned to the
3361 sender, but the error was temporary, then after the final timeout, only
3362 "retry timeout exceeded" was returned. Now it returns the full error as
3363 well as "retry timeout exceeded".
3365 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3366 do this, and (what is worse) MTAs that accept it.
3368 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3369 will now be deprecated.
3371 PH/56 New os.c-cygwin from the Cygwin maintainer.
3373 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3374 formats (previously only available in exim formats via -bpr, -bpru,
3375 and -bpra. Now also available in native and exiqgrep formats)
3377 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3378 with very large, slow to parse queues
3380 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3382 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3384 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3385 responses to authentication challenges, though it was showing the
3386 challenges; (ii) I've removed the CR characters from the debug output for
3389 PH/58 Allow for the insertion of a newline as well as a space when a string
3390 is turned into more than one encoded-word during RFC 2047 encoding. The
3391 Sieve code now uses this.
3393 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3394 data_4xx, lost_connection, tls_required.
3396 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3397 message was being sent as an SMTP response.
3399 PH/61 Add -l and -k options to exicyclog.
3401 PH/62 When verifying, if an address was redirected to one new address, so that
3402 verification continued, and the new address failed or deferred after
3403 having set something in $address_data, the value of $address_data was not
3404 passed back to the ACL. This was different to the case when no
3405 redirection occurred. The value is now passed back in both cases.
3407 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3408 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3409 use login_cap.h, so on its own it isn't the distinguishing feature. The
3410 new name refers directly to the setclassresources() function.
3412 PH/65 Added configuration files for NetBSD3.
3414 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3416 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3417 is preferred over IPv4.
3419 PH/68 The bounce_return_message and bounce_return_body options were not being
3420 honoured for bounces generated during the reception of non-SMTP messages.
3421 In particular, this applied to messages rejected by the ACL. This bug has
3422 been fixed. However, if bounce_return_message is true and bounce_return_
3423 body is false, the headers that are returned for a non-SMTP message
3424 include only those that have been read before the error was detected.
3425 (In the case of an ACL rejection, they have all been read.)
3427 PH/69 The HTML version of the specification is now built in a directory called
3428 spec_html instead of spec.html, because the latter looks like a path with
3429 a MIME-type, and this confuses some software.
3431 PH/70 Catch two compiler warnings in sieve.c.
3433 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3434 function verify_get_ident() calls ip_connect() to connect a socket, but
3435 if the "connect()" function timed out, ip_connect() used to close the
3436 socket. However, verify_get_ident() also closes the socket later, and in
3437 between Exim writes to the log, which may get opened at this point. When
3438 the socket was closed in ip_connect(), the log could get the same file
3439 descriptor number as the socket. This naturally causes chaos. The fix is
3440 not to close the socket in ip_connect(); the socket should be closed by
3441 the function that creates it. There was only one place in the code where
3442 this was missing, in the iplookup router, which I don't think anybody now
3443 uses, but I've fixed it anyway.
3445 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3446 well as to direct DNS lookups. Otherwise the handling of names in host
3447 lists is inconsistent and therefore confusing.
3453 PH/01 Two changes to the default runtime configuration:
3455 (1) Move the checks for relay_from_hosts and authenticated clients from
3456 after to before the (commented out) DNS black list checks.
3458 (2) Add control=submission to the relay_from_hosts and authenticated
3459 clients checks, on the grounds that messages accepted by these
3460 statements are most likely to be submissions.
3462 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3464 (1) Generate an error if the third argument for the ${prvs expansion is
3467 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3470 (3) Reset the variables that are obtained from the first argument of
3471 ${prvscheck and used in the second argument before leaving the code,
3472 because their memory is reclaimed, so using them afterwards may do
3475 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3476 one (it's much easier than Tom thought :-).
3478 (5) Because of (4), we can now allow for the use of $prvscheck_result
3479 inside the third argument.
3481 PH/03 For some reason, the default setting of PATH when running a command from
3482 a pipe transport was just "/usr/bin". I have changed it to
3485 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3486 anything to be listed in the output from -bV.
3488 PH/05 When a filter generated an autoreply, the entire To: header line was
3489 quoted in the delivery log line, like this:
3491 => >A.N.Other <ano@some.domain> <original@ddress> ...
3493 This has been changed so that it extracts the operative address. There
3494 may be more than one such address. If so, they are comma-separated, like
3497 => >ano@some.domain,ona@other.domain <original@ddress> ...
3499 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3500 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3501 not being looked up in the rDNS to get a host name, Exim was showing the
3502 IP address twice in Received: lines, even though the IP addresses were
3503 identical. For example:
3505 Received: from [1.2.3.4] (helo=[1.2.3.4])
3507 However, if the real host name was known, it was omitting the HELO data
3508 if it matched the actual IP address. This has been tidied up so that it
3509 doesn't show the same IP address twice.
3511 PH/07 When both +timestamp and +memory debugging was on, the value given by
3512 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3513 called by the debug printing, thereby overwriting the timestamp buffer.
3514 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3516 PH/08 When the original message was included in an autoreply transport, it
3517 always said "this is a copy of the message, including all the headers",
3518 even if body_only or headers_only was set. It now gives an appropriate
3521 PH/09 Applied a patch from the Sieve maintainer which:
3523 o fixes some comments
3524 o adds the (disabled) notify extension core
3525 o adds some debug output for the result of if/elsif tests
3526 o points to the current vacation draft in the documentation
3527 and documents the missing references header update
3531 o fixes a bug in processing the envelope test (when testing
3532 multiple envelope elements, the last element determined the
3535 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3536 Electronic Mail") by including:
3538 Auto-submitted: auto-generated
3540 in the messages that it generates (bounce messages and others, such as
3541 warnings). In the case of bounce messages for non-SMTP messages, there was
3542 also a typo: it was using "Auto_submitted" (underscore instead of
3543 hyphen). Since every message generated by Exim is necessarily in response
3544 to another message, thes have all been changed to:
3546 Auto-Submitted: auto-replied
3548 in accordance with these statements in the RFC:
3550 The auto-replied keyword:
3552 - SHOULD be used on messages sent in direct response to another
3553 message by an automatic process,
3555 - MUST NOT be used on manually-generated messages,
3557 - MAY be used on Delivery Status Notifications (DSNs) and Message
3558 Disposition Notifications (MDNs),
3560 - MUST NOT be used on messages generated by automatic or periodic
3561 processes, except for messages which are automatic responses to
3564 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3565 to the default Received: header definition.
3567 PH/12 Added log selector acl_warn_skipped (default on).
3569 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3570 variables because (a) they are in the wrong storage pool and (b) even if
3571 they were copied, it wouldn't work properly because of the caching.
3573 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3574 checking when decoding. Apparently there are clients that generate
3575 overlong encoded strings. Why am I not surprised?
3577 PH/15 If the first argument of "${if match_address" was not empty, but did not
3578 contain an "@" character, Exim crashed. Now it writes a panic log message
3579 and treats the condition as false.
3581 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3583 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3584 extension "envelope-auth". The code is finished and in agreement with
3585 other implementations, but there is no documentation so far and in fact,
3586 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3587 not changing the active code.
3589 Print executed "if" and "elsif" statements when debugging is used. This
3590 helps a great deal to understand what a filter does.
3592 Document more things not specified clearly in RFC3028. I had all this
3593 sorted out, when out of a sudden new issues came to my mind. Oops."
3595 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3598 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3599 canonical form (as documented). However, after a host name lookup from
3600 the IP address, check_host() was doing a simple string comparison with
3601 addresses acquired from the DNS when checking that the found name did
3602 have the original IP as one of its addresses. Since any found IPv6
3603 addresses are likely to be in abbreviated form, the comparison could
3604 fail. Luckily, there already exists a function for doing the comparison
3605 by converting both addresses to binary, so now that is used instead of
3606 the text comparison.
3608 PH/20 There was another similar case to PH/19, when a complete host name was
3609 given in a host list; looking up its IP address could give an abbreviated
3610 form, whereas the current host's name might or might not be abbreviated.
3611 The same fix has been applied.
3617 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3618 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3621 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3622 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3624 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3626 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3627 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3628 Also, for "Auto-Submitted", treat anything other than "no" as
3629 non-personal, in accordance with RFC 3834. (Previously it treated
3630 anything starting "auto-" as non-personal.)
3632 TF/01 The control=submission/name=... option had a problem with syntax
3633 errors if the name included a slash character. The /name= option
3634 now slurps the rest of the string, so it can include any characters
3635 but it must come last in the list of options (after /sender_retain
3638 PH/05 Some modifications to the interface to the fake nameserver for the new
3646 TK/01 Added the "success_on_redirect" address verification option. See
3647 NewStuff for rationale and an example.
3649 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3651 PH/02 Patch to exigrep to allow it to work on syslog lines.
3653 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3654 fread() to skip over the body file's header line, because in Cygwin the
3655 header line is locked and is inaccessible.
3657 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3658 co-exist for some time) to make it clear that it is the Exim ID that is
3659 referenced, not the Message-ID: header line.
3661 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3662 string_format() function, because snprintf() does not exist on all
3665 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3666 ${stat: expansion item.
3668 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3669 protocol synchronization error", to keep the pedants happy.
3671 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3672 well as for IRIX systems, when gcc is being used. See the host.c source
3675 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3677 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3680 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3681 transport and to the smtp transport in LMTP mode.
3683 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3685 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3686 run a filter in a subprocess. This could lead to confusion in subsequent
3687 lookups in the parent process. There should also be a search_tidyup() at
3688 the end of the subprocess.
3690 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3691 only if the host matched helo_try_verify_hosts, which caused the
3692 verification to occur when the EHLO/HELO command was issued. The ACL just
3693 tested the remembered result. Now, if a previous verification attempt has
3694 not happened, "verify = helo" does it there and then.
3696 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3698 TK/03 Fix log output including CR from clamd.
3700 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3701 exist provoked a memory error which could cause a segfault.
3703 PH/15 Installed PCRE 6.2
3705 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3707 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3708 of the problem. Specifically, suggested +O2 rather than +O1 for the
3711 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3713 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3714 record, the port was not correctly logged when the outgoing_port log
3715 selector was set (it logged the transort's default port).
3717 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3718 fallback_hosts, and "hosts" in the smtp transport.
3720 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3721 host errors such as "Connection refused".
3723 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3724 authentication with radiusclient 0.4.9:
3726 - Error returned from rc_read_config was caught wrongly
3727 - Username/password not passed on to radius server due to wrong length.
3729 The presumption is that some radiusclient API changes for 4.51/PH/17
3730 were not taken care of correctly. The code is still untested by me (my
3731 Linux distribution still has 0.3.2 of radiusclient), but it was
3732 contributed by a Radius user.
3734 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3735 expanding the "port" option of the smtp transport.
3737 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3738 while reading a MIME header. Thanks to Tom Hughes for a patch.
3740 PH/24 Include config.h inside local_scan.h so that configuration settings are
3743 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3744 that in an ACL for RCPT (for example), you can examine exactly what was
3747 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3748 commands, but it was not correctly comparing the address with the actual
3749 client host address. Thus, it would show the EHLO address in Received:
3750 header lines when this was not necessary.
3752 PH/27 Added the % operator to ${eval:}.
3754 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3755 it should be ignoring failures (because with -C, for example, it has lost
3756 privilege). It wasn't ignoring creation failures other than "already
3759 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3762 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3763 because an input line was too long, either on its own, or by virtue of
3764 too many continuations, the temporary file was not being removed, and the
3765 return code was incorrect.
3767 PH/31 Missing "BOOL" in function definition in filtertest.c.
3769 PH/32 Applied Sieve patches from the maintainer.
3771 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3773 PH/33 Added "verify = not_blind".
3775 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3776 Local/Makefile (with some defaults set). These are used in built scripts
3777 such as exicyclog, but they have never been used in the exim_install
3778 script (though there are many overriding facilities there). I have
3779 arranged that the exim_install script now takes note of these two
3782 PH/35 Installed configuration files for Dragonfly.
3784 PH/36 When a locally submitted message by a trusted user did not contain a
3785 From: header, and the sender address was obtained from -f or from an SMTP
3786 MAIL command, and the trusted user did not use -F to supply a sender
3787 name, $originator_name was incorrectly used when constructing a From:
3788 header. Furthermore, $originator_name was used for submission mode
3789 messages from external hosts without From: headers in a similar way,
3790 which is clearly wrong.
3792 PH/37 Added control=suppress_local_fixups.
3794 PH/38 When log_selector = +received_sender was set, and the addition of the
3795 sender made the log line's construction buffer exactly full, or one byte
3796 less than full, an overflow happened when the terminating "\n" was
3799 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3800 when the result of a list match is failure because a DNS lookup failed.
3802 PH/40 RM_COMMAND is now used in the building process.
3804 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3805 the "build-* directories that it finds.
3807 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3808 address in a domain literal was a prefix of an interface address.
3810 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3811 when verifying a sender address, unless rewrite_headers is false.
3813 PH/44 Wrote a long comment about why errors_to addresses are verified as
3814 recipients, not senders.
3816 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3817 the ratelimit ACL was added.
3819 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3821 PH/46 Added extra information about PostgreSQL errors to the error string.
3823 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3824 avoiding the need to install special zones in a real server. This is
3825 backwards compatible; if it can't find the fake resolver, it drops back.
3826 Thus, both old and new test suites can be run.
3828 TF/02 Added util/ratelimit.pl
3830 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3831 clock is set back in time.
3833 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3834 Candler <B.Candler@pobox.com>.
3836 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3837 OK for addresses that are the result of redirections.
3839 PH/48 A number of further additions for the benefit of the new test suite,
3840 including a fake gethostbyname() that interfaces to the fake DNS resolver
3843 TF/06 The fix for widen_domains has also been applied to qualify_single and
3844 search_parents which are the other dnslookup options that can cause
3847 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3850 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3852 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3853 service" if it encountered an MX record with an empty target hostname.
3854 The message is now "an MX or SRV record indicated no SMTP service".
3856 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3857 if the DNS of the sending site is misconfigured. This is quite a
3858 common situation. This change restores the behaviour of treating a
3859 helo verification defer as a failure.
3861 PH/51 If self=fail was set on a router, the bounce message did not include the
3862 actual error message.
3868 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3870 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3871 command itself ended in error, the underlying message about the transport
3872 filter timeout was being overwritten with the pipe command error. Now the
3873 underlying error message should be appended to the second error message.
3875 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3877 PH/02 Reduce the amount of output that "make" produces by default. Full output
3878 can still be requested.
3880 PH/03 The warning log line about a condition test deferring for a "warn" verb
3881 was being output only once per connection, rather than after each
3882 occurrence (because it was using the same function as for successful
3883 "warn" verbs). This seems wrong, so I have changed it.
3885 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3886 it should not have, which might have caused a crash in the right
3887 circumstances, but probably never did.
3889 PH/04 Installed a modified version of Tony Finch's patch to make submission
3890 mode fix the return path as well as the Sender: header line, and to
3891 add a /name= option so that you can make the user's friendly name appear
3894 TF/03 Added the control = fakedefer ACL modifier.
3896 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3897 Mark Lowes for thorough testing.
3899 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3901 TK/03 Merged latest SRS patch from Miles Wilton.
3903 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3904 with the definition in sysexits.h (which is #included earlier).
3905 Fortunately, Exim does not actually use EX_OK. The code used to try to
3906 preserve the sysexits.h value, by assuming that macro definitions were
3907 scanned for macro replacements. I have been disabused of this notion,
3908 so now the code just undefines EX_OK before #including unistd.h.
3910 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3911 in the smtp transport. When a block could not be written in a single
3912 write() function, the timeout was being re-applied to each part-write.
3913 This seems wrong - if the receiver was accepting one byte at a time it
3914 would take for ever. The timeout is now adjusted when this happens. It
3915 doesn't have to be particularly precise.
3917 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3918 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3920 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3921 without a domain if the check to <postmaster@domain> fails.
3923 SC/01 Eximstats: added -xls and the ability to specify output files
3924 (patch written by Frank Heydlauf).
3926 SC/02 Eximstats: use FileHandles for outputting results.
3928 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3930 SC/04 Eximstats: fixed display of large numbers with -nvr option
3932 SC/05 Eximstats: fixed merging of reports with empty tables.
3934 SC/06 Eximstats: added the -include_original_destination flag
3936 SC/07 Eximstats: removed tabs and trailing whitespace.
3938 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3940 TK/06 MBOX spool code: Add real "From " MBOX separator line
3941 so the .eml file is really in mbox format (even though
3942 most programs do not really care). Patch from Alex Miller.
3944 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3945 The latter is generated from $received_to and is only set if the
3946 message has one envelope recipient. SA can use these headers,
3947 obviously out-of-the-box. Patch from Alex Miller.
3949 PH/08 The ${def test on a variable was returning false if the variable's
3950 value was "0", contrary to what the specification has always said!
3951 The result should be true unless the variable is empty.
3953 PH/09 The syntax error of a character other than { following "${if
3954 def:variable_name" (after optional whitespace) was not being diagnosed.
3955 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3956 accidental colon was present, for example, could give incorrect results.
3958 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3959 result is used (not including appendfile, where other changes are about
3962 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3963 This involved changing a lot of size variables from int to off_t. It
3964 should work with maildirs and everything.
3966 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3967 spamd dying while we are connected to it.
3969 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3972 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3973 function for BDB 4.3.
3975 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3977 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3978 the sources for PCRE have been re-arranged and more files are now
3981 PH/15 The code I had for printing potentially long long variables in PH/11
3982 above was not the best (it lost precision). The length of off_t variables
3983 is now inspected at build time, and an appropriate printing format (%ld
3984 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3985 to be "long long int" or "long int". This is needed for the internal
3986 formatting function string_vformat().
3988 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3989 the configuration file to be ":syslog", then the script "guesses" where
3990 the logs files are, rather than using the compiled in default. In our
3991 case the guess is not the same as the compiled default, so the script
3992 suddenly stopped working when I started to use syslog. The patch checks
3993 to see if log_file_path is "". If so, it attempts to read it from exim
3994 with no configuration file to get the compiled in version, before it
3995 falls back to the previous guessing code."
3997 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3998 implementing BATV in an Exim configuration. See NewStuff for the gory
4001 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4002 Makefile that are specific to HP-UX.
4004 PH/18 If the "use_postmaster" option was set for a recipient callout together
4005 with the "random" option, the postmaster address was used as the MAIL
4006 FROM address for the random test, but not for the subsequent recipient
4007 test. It is now used for both.
4009 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4010 patch removes a few documentation additions to RFC 3028, because the
4011 latest draft now contains them. It adds the new en;ascii-case comparator
4012 and a new error check for 8bit text in MIME parts. Comparator and
4013 require names are now matched exactly. I enabled the subaddress
4014 extension, but it is not well tested yet (read: it works for me)."
4016 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4017 rework some of the code of TK/09 above to avoid the hardwired use of
4018 "%lld" and "long long". Replaced the call to snprintf() with a call to
4021 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4022 records point to non-existent hosts", "retry timeout exceeded", and
4023 "retry time not reached for any host after a long failure period".
4025 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4026 experimental DomainKeys support:
4028 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4029 (2) On an error such as an illegally used "control", the wrong name for
4030 the control was given.
4032 These problems did NOT occur unless DomainKeys support was compiled.
4034 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4036 PH/24 Added ${if match_ip condition.
4038 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4039 databases so that it will be absolutely obvious if a crash occurs in the
4040 DB library. This is a regular occurrence (often caused by mis-matched
4043 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4044 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4045 picked up on a user's system that detects such things. There doesn't seem
4046 to be a gcc warning option for this - only an attribute that has to be
4047 put on the function's prototype. It seems that in Fedora Core 4 they have
4048 set this on a number of new functions. No doubt there will be more in due
4051 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4052 specify a transport. However, if an address that was verified by such a
4053 router was the subject of a callout, Exim crashed because it tried to
4054 read the rcpt_include_affixes from the non-existent transport. Now it
4055 just assumes that the setting of that option is false. This bug was
4056 introduced by 4.51/PH/31.
4058 PH/28 Changed -d+all to exclude +memory, because that information is very
4059 rarely of interest, but it makes the output a lot bigger. People tend to
4060 do -d+all out of habit.
4062 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4063 code in os-type was giving problems when libc.so lives in lib64, like on
4066 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4067 aren't the modern standard, and it seems that some systems' include files
4068 don't always have them. Exim was already checking for some of the newer
4069 ones like T_AAAA, and defining it itself. I've added checks for all the
4070 record types that Exim uses.
4072 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4073 not automatically generating a new one, as it is supposed to. This
4074 prevented TLS from working. If the file did exist, but contained invalid
4075 data, a new version was generated, as expected. It was only the case of a
4076 non-existent file that was broken.
4078 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4079 with a change in libdomainkeys > 0.64.
4081 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4082 from DNS. If the selector record carries the flag, it now has
4083 precedence over the domain-wide flag.
4085 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4087 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4088 the use of a port name that isn't defined in /etc/services) occurred, the
4089 message was deferred as in a normal delivery, and thus remained on the
4090 spool, instead of being failed because of the mua_wrapper setting. This
4091 is now fixed, and I tidied up some of the mua_wrapper messages at the
4094 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4095 the messages in a hash of arrays rather than using individual hashes.
4096 This is a bit cleaner and results in dramatic memory savings, albeit
4097 at a slight CPU cost.
4099 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4100 as requested by Marc Sherman.
4102 SC/10 Eximstats: added histograms for user specified patterns as requested
4105 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4107 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4108 fopen() in the content-scanning modules that did not already have it.
4114 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4115 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4117 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4119 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4121 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4122 to test Sieve filters that use "vacation".
4124 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4125 that changes the way the GnuTLS parameters are stored in the cache file.
4126 The new format can be generated externally. For backward compatibility,
4127 if the data in the cache doesn't make sense, Exim assumes it has read an
4128 old-format file, and it generates new data and writes a new file. This
4129 means that you can't go back to an older release without removing the
4132 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4133 work if there are any delivery delays because "one_time" forces the
4134 parent to be marked "delivered", so its unseen clone is never tried
4135 again. For this reason, Exim now forbids the simultaneous setting of
4138 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4139 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4140 entry if you want to know the details. The fix, however, neglected to
4141 consider the case when local delivery batching is involved. The test for
4142 "previously delivered" was not happening when checking to see if an
4143 address could be batched with a previous (undelivered) one; under
4144 certain circumstances this could lead to multiple deliveries to the same
4147 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4148 in its include files, and this causes problems building Exim.
4150 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4151 header_syntax) but Exim was just ignoring anything given after a slash.
4152 In particular, this caused confusion with an attempt to use "verify =
4153 reverse_host_lookup/defer_ok". An error is now given when options are
4154 supplied for verify items that do not have them. (Maybe reverse_host_
4155 lookup should have a defer_ok option, but that's a different point.)
4157 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4158 defined by RFC 821) to 2048, because there were problems with some AUTH
4159 commands, and RFC 1869 says the size should be increased for extended
4160 SMTP commands that take arguments.
4162 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4165 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4166 "unknown" error; now it says that the functionality isn't in the binary.
4168 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4169 an address' error message when a string expansion fails (syntax or
4170 whatever). Otherwise the password may appear in the log. Following change
4171 PH/42 below, there is no longer a chance of it appearing in a bounce
4174 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4176 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4177 of its addresses was ever tried. (Bugzilla bug #2.)
4179 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4180 the result incorrectly in the debug output. (It correctly added a newline
4181 to what was transported.)
4183 TF/01 Added $received_time.
4185 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4186 commented out examples of how to interface to a virus scanner and to
4187 SpamAssassin. Also added commented examples of av_scanner and
4188 spamd_address settings.
4190 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4191 and controls are allowed in which ACLs. There were a couple of minor
4192 errors. Some of the entries in the conditions table (which is a table of
4193 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4194 negation of where the condition IS allowed.
4196 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4198 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4199 header file does not have a version number, so I've had to invent a new
4200 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4201 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4202 radiusclient), but it was contributed by a Radius user.
4204 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4205 files or directories, for OpenSSL.
4207 PH/19 When an Exim process that is running as root has to create an Exim log
4208 file, it does so in a subprocess that runs as exim:exim so as to get the
4209 ownership right at creation (otherwise, other Exim processes might see
4210 the file with the wrong ownership). There was no test for failure of this
4211 fork() call, which would lead to the process getting stuck as it waited
4212 for a non-existent subprocess. Forks do occasionally fail when resources
4213 run out. I reviewed all the other calls to fork(); they all seem to check
4216 PH/20 When checking for unexpected SMTP input at connect time (before writing
4217 the banner), Exim was not dealing correctly with a non-positive return
4218 from the read() function. If the client had disconnected by this time,
4219 the result was a log entry for a synchronization error with an empty
4220 string after "input=" when read() returned zero. If read() returned -1
4221 (an event I could not check), uninitialized data bytes were printed.
4222 There were reports of junk text (parts of files, etc) appearing after
4225 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4227 PH/22 Added support for macro redefinition, and (re)definition in between
4228 driver and ACL definitions.
4230 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4231 forgetting to use the resulting value; it was using the unexpanded value.
4233 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4234 hadn't been configured. The fix is from Juergen Kreileder, who
4235 understands it better than I do:
4237 "Here's what I see happening with three configured cyrus_sasl
4238 authenticators configured (plain, login, cram-md5):
4240 On startup auth_cyrus_sasl_init() gets called for each of these.
4241 This means three calls to sasl_listmech() without a specified mech_list.
4242 => SASL tests which mechs of all available mechs actually work
4243 => three warnings about OTP not working
4244 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4246 With the patch, sasl_listmech() also gets called three times. But now
4247 SASL's mech_list option is set to the server_mech specified in the the
4248 authenticator. Or in other words, the answer from sasl_listmech()
4249 gets limited to just the mech you're testing for (which is different
4251 => the return list contains just 'plain' or 'login', 'cram-md5' or
4252 nothing depending on the value of ob->server_mech.
4254 I've just tested the patch: Authentication still works fine,
4255 unavailable mechs specified in the exim configuration are still
4256 caught, and the auth.log warnings about OTP are gone."
4258 PH/25 When debugging is enabled, the contents of the command line are added
4259 to the debugging output, even when log_selector=+arguments is not
4262 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4263 answer is "GNU", and only if the return is "GNU/something" is the answer
4266 PH/27 $acl_verify_message is now set immediately after the failure of a
4267 verification in an ACL, and so is available in subsequent modifiers. In
4268 particular, the message can be preserved by coding like this:
4270 warn !verify = sender
4271 set acl_m0 = $acl_verify_message
4273 Previously, $acl_verify_message was set only while expanding "message"
4274 and "log_message" when a very denied access.
4276 PH/28 Modified OS/os.c-Linux with
4278 -#ifndef OS_LOAD_AVERAGE
4279 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4281 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4282 nomenclature these days.)
4284 PH/29 Installed patch from the Sieve maintainer that adds the options
4285 sieve_useraddress and sieve_subaddress to the redirect router.
4287 PH/30 In these circumstances:
4288 . Two addresses routed to the same list of hosts;
4289 . First host does not offer TLS;
4290 . First host accepts first address;
4291 . First host gives temporary error to second address;
4292 . Second host offers TLS and a TLS session is established;
4293 . Second host accepts second address.
4294 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4295 and peerdn, if requested) that were in fact used only for the second
4298 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4299 attention to any local part prefix or suffix that was matched by the
4300 router that accepted the address. It now behaves in the same way as it
4301 does for delivery: the affixes are removed from the local part unless
4302 rcpt_include_affixes is set on the transport.
4304 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4305 timeout during the DATA phase of an incoming message.
4307 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4308 applied a patch sanctioned by the Sieve maintainer.
4310 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4311 the uid or gid is negative. A case of a negative gid caused this to be
4312 noticed. The fix allows for either to be negative.
4314 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4315 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4318 PH/36 The change PH/12 above was broken. Fixed it.
4320 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4321 the grounds that routing the same address twice would always produce the
4322 same answer. This might have been true once, but it is certainly no
4323 longer true now. Routing a child address may depend on the previous
4324 routing that produced that child. Some complicated redirection strategies
4325 went wrong when messages had multiple recipients, and made Exim's
4326 behaviour dependent on the order in which the addresses were given.
4328 I have moved the duplicate checking until after the routing is complete.
4329 Exim scans the addresses that are assigned to local and remote
4330 transports, and removes any duplicates. This means that more work will be
4331 done, as duplicates will always all be routed, but duplicates are
4332 presumably rare, so I don't expect this is of any significance.
4334 For deliveries to pipes, files, and autoreplies, the duplicate checking
4335 still happens during the routing process, since they are not going to be
4338 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4339 It corrects a timeout issue with spamd. This is Ian's comment: "The
4340 background is that sometimes spamd either never reads data from a
4341 connection it has accepted, or it never writes response data. The exiscan
4342 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4343 blindly assumes that writes won't block so it may never time out."
4345 PH/39 Allow G after quota size as well as K and M.
4347 PH/40 The value set for $authenticated_id in an authenticator may not contain
4348 binary zeroes or newlines because the value is written to log lines and
4349 to spool files. There was no check on this. Now the value is run through
4350 the string_printing() function so that such characters are converted to
4351 printable escape sequences.
4353 PH/41 $message_linecount is a new variable that contains the total number of
4354 lines in the message. Compare $body_linecount, which is the count for the
4357 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4358 bounce and delay warning messages, except in certain special cases, which
4361 (a) An SMTP error message from a remote host;
4362 (b) A message specified in a :fail: redirection;
4363 (c) A message specified in a "fail" command in a system filter;
4364 (d) A message specified in a FAIL return from the queryprogram router;
4365 (e) A message specified by the cannot_route_message router option.
4367 In these cases only, Exim does include the error details in bounce and
4368 warning messages. There are also a few cases where bland messages such
4369 as "unrouteable address" or "local delivery error" are given.
4371 PH/43 $value is now also set for the "else" part of a ${run expansion.
4373 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4374 being worked on, but at least Exim now implements the latest version to
4377 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4378 process to complete was treated as a delivery failure, a timeout while
4379 writing the message to the pipe was logged, but erroneously treated as a
4380 successful delivery. Such timeouts include transport filter timeouts. For
4381 consistency with the overall process timeout, these timeouts are now
4382 treated as errors, giving rise to delivery failures by default. However,
4383 there is now a new Boolean option for the pipe transport called
4384 timeout_defer, which, if set TRUE, converts the failures into defers for
4385 both kinds of timeout. A transport filter timeout is now identified in
4388 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4389 systems where "make" and "gmake" are different, calling "gmake" at top
4390 level broke things. I've arranged for the value of $(MAKE) to be passed
4391 from the Makefile to this script so that it can call the same version of
4395 A note about Exim versions 4.44 and 4.50
4396 ----------------------------------------
4398 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4399 changes of various kinds. As a consequence, a big documentation update was
4400 needed. This delayed the release for rather longer than seemed good, especially
4401 in the light of a couple of (minor) security issues. Therefore, the changes
4402 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4403 release. So 4.44 and 4.50 are in effect two different branches that both start
4406 I have left the 4.50 change log unchanged; it contains all the changes since
4407 4.43. The change log for 4.44 is below; many of its items are identical to
4408 those for 4.50. This seems to be the most sensible way to preserve the
4409 historical information.
4415 1. Minor wording change to the doc/README.SIEVE file.
4417 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4418 computation of the current number of files was incorrect.
4420 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4421 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4424 4. Give more explanation in the error message when the command for a transport
4425 filter fails to execute.
4427 5. There are several places where Exim runs a non-Exim command in a
4428 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4429 was being done only for the command run by the queryprogram router. It is
4430 now done for all such subprocesses. The other cases are: ${run, transport
4431 filters, and the commands run by the lmtp and pipe transports.
4433 6. Added CONFIGURE_GROUP build-time option.
4435 7. Some older OS have a limit of 256 on the maximum number of file
4436 descriptors. Exim was using setrlimit() to set 1000 as a large value
4437 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4438 systems. I've change it so that if it can't get 1000, it tries for 256.
4440 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4441 was an oversight, and furthermore, ever since the addition of extra
4442 controls (e.g. 4.43/32), the checks on when to allow different forms of
4443 "control" were broken. There should now be diagnostics for all cases when a
4444 control that does not make sense is encountered.
4446 9. Added the /retain_sender option to "control=submission".
4448 10. $recipients is now available in the predata ACL (oversight).
4450 11. Tidy the search cache before the fork to do a delivery from a message
4451 received from the command line. Otherwise the child will trigger a lookup
4452 failure and thereby defer the delivery if it tries to use (for example) a
4453 cached ldap connection that the parent has called unbind on.
4455 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4456 of $address_data from the recipient verification was clobbered by the
4457 sender verification.
4459 13. The value of address_data from a sender verification is now available in
4460 $sender_address_data in subsequent conditions in the ACL statement.
4462 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4464 15. Added a new option "connect=<time>" to callout options, to set a different
4467 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4468 was its contents. (It was OK if the option was not defined at all.)
4470 17. A "Completed" log line is now written for messages that are removed from
4471 the spool by the -Mrm option.
4473 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4474 information about exactly what failed.
4476 19. Added -dd to debug only the daemon process.
4478 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4479 handles timeouts, both on the server side and network timeouts. Renamed the
4480 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4482 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4483 It is now set to "smtps".
4485 22. $host_address is now set to the target address during the checking of
4486 ignore_target_hosts.
4488 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4489 being passed; this would have caused $sender_host_name to have been used if
4490 matching the list had actually called for a host name (not very likely,
4491 since this list is usually IP addresses). A host name is now passed as
4494 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4495 code that specifies a non-restarting handler (typically sigaction() in
4496 modern systems) in an attempt to fix a rare and obscure crash bug.
4498 25. Narrowed the window for a race in the daemon that could cause it to ignore
4499 SIGCHLD signals. This is not a major problem, because they are used only to
4500 wake it up if nothing else does.
4502 26. A malformed maildirsize file could cause Exim to calculate negative values
4503 for the mailbox size or file count. Odd effects could occur as a result.
4504 The maildirsize information is now recalculated if the size or filecount
4507 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4508 support for a long time. Removed HAVE_SYS_VFS_H.
4510 28. Installed the latest version of exipick from John Jetmore.
4512 29. In an address list, if the pattern was not a regular expression, an empty
4513 subject address (from a bounce message) matched only if the pattern was an
4514 empty string. Non-empty patterns were not even tested. This was the wrong
4515 because it is perfectly reasonable to use an empty address as part of a
4516 database query. An empty address is now tested by patterns that are
4517 lookups. However, all the other forms of pattern expect the subject to
4518 contain a local part and a domain, and therefore, for them, an empty
4519 address still always fails if the pattern is not itself empty.
4521 30. Exim went into a mad DNS loop when attempting to do a callout where the
4522 host was specified on an smtp transport, and looking it up yielded more
4523 than one IP address.
4525 31. Re-factored the code for checking spool and log partition space into a
4526 function that finds that data and another that does the check. The former
4527 is then used to implement four new variables: $spool_space, $log_space,
4528 $spool_inodes, and $log_inodes.
4530 32. The RFC2047 encoding function was originally intended for short strings
4531 such as real names; it was not keeping to the 75-character limit for
4532 encoded words that the RFC imposes. It now respects the limit, and
4533 generates multiple encoded words if necessary. To be on the safe side, I
4534 have increased the buffer size for the ${rfc2047: expansion operator from
4537 33. It is now permitted to omit both strings after an "if" condition; if the
4538 condition is true, the result is "true". As before, when the second string
4539 is omitted, a false condition yields an empty string. This makes it less
4540 cumbersome to write custom ACL and router conditions.
4542 34. Failure to deliver a bounce message always caused it to be frozen, even if
4543 there was an errors_to setting on the router. The errors_to setting is now
4546 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4547 canonical form (fully expanded) before being placed in
4548 $sender_host_address.
4550 36. The table in the code that translates DNS record types into text (T_A to
4551 "A" for instance) was missing entries for NS and CNAME. It is just possible
4552 that this could have caused confusion if both these types were looked up
4553 for the same domain, because the text type is used as part of Exim's
4554 per-process caching. But the chance of anyone hitting this buglet seems
4557 37. The dnsdb lookup has been extended in a number of ways.
4559 (1) There is a new type, "zns", which walks up the domain tree until it
4560 finds some nameserver records. It should be used with care.
4562 (2) There is a new type, "mxh", which is like "mx" except that it returns
4563 just the host names, not the priorities.
4565 (3) It is now possible to give a list of domains (or IP addresses) to be
4566 looked up. The behaviour when one of the lookups defers can be
4567 controlled by a keyword.
4569 (4) It is now possible to specify the separator character for use when
4570 multiple records are returned.
4572 38. The dnslists ACL condition has been extended: it is now possible to supply
4573 a list of IP addresses and/or domains to be looked up in a particular DNS
4576 39. Added log_selector=+queue_time_overall.
4578 40. When running the queue in the test harness, wait just a tad after forking a
4579 delivery process, to get repeatability of debugging output.
4581 41. Include certificate and key file names in error message when GnuTLS fails
4582 to set them up, because the GnuTLS error message doesn't include the name
4583 of the failing file when there is a problem reading it.
4585 42. Allow both -bf and -bF in the same test run.
4587 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4589 44. The "Exiscan patch" is now merged into the mainline Exim source.
4591 45. Sometimes the final signoff response after QUIT could fail to get
4592 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4593 < 0 before doing a fflush(). This bug looks as though it goes back to the
4594 introduction of TLS in release 3.20, but "sometimes" must have been rare
4595 because the tests only now provoked it.
4597 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4598 (this can affect the format of dates).
4600 47. exim_tidydb, when checking for the continued existence of a message for
4601 which it has found a message-specific retry record, was not finding
4602 messages that were in split spool directories. Consequently, it was
4603 deleting retry records that should have stayed in existence.
4605 48. Steve fixed some bugs in eximstats.
4607 49. The SPA authentication driver was not abandoning authentication and moving
4608 on to the next authenticator when an expansion was forced to fail,
4609 contradicting the general specification for all authenticators. Instead it
4610 was generating a temporary error. It now behaves as specified.
4612 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4613 (the order specifies the preference for clients). The order is now AES256,
4614 AES128, 3DES, ARCFOUR128.
4616 51. Small patch to Sieve code - explicitly set From: when generating an
4619 52. Exim crashed if a remote delivery caused a very long error message to be
4620 recorded - for instance if somebody sent an entire SpamAssassin report back
4621 as a large number of 550 error lines. This bug was coincidentally fixed by
4622 increasing the size of one of Exim's internal buffers (big_buffer) that
4623 happened as part of the Exiscan merge. However, to be on the safe side, I
4624 have made the code more robust (and fixed the comments that describe what
4627 53. Now that there can be additional text after "Completed" in log lines (if
4628 the queue_time_overall log selector is set), a one-byte patch to exigrep
4629 was needed to allow it to recognize "Completed" as not the last thing in
4632 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4633 patch that reportedly fixes this has been added. I am not expert enough to
4634 create a test for it. This is what the patch creator wrote:
4636 "I found a little strange behaviour of ldap code when working with
4637 Windows 2003 AD Domain, where users was placed in more than one
4638 Organization Units. When I tried to give exim partial DN, the exit code
4639 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4640 But simultaneously result of request was absolutely normal ldap result,
4641 so I produce this patch..."
4643 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4644 REFERENCE, so I have modified the code to exclude the patch when that macro
4647 55. Some experimental protocols are using DNS PTR records for new purposes. The
4648 keys for these records are domain names, not reversed IP addresses. The
4649 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4650 leaves it alone. Component reversal etc. now happens only for IP addresses.
4653 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4655 57. Double the size of the debug message buffer (to 2048) so that more of very
4656 long debug lines gets shown.
4658 58. The exicyclog utility now does better if the number of log files to keep
4659 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4661 59. Two changes related to the smtp_active_hostname option:
4663 (1) $smtp_active_hostname is now available as a variable.
4664 (2) The default for smtp_banner uses $smtp_active_hostname instead
4665 of $primary_hostname.
4667 60. The host_aton() function is supposed to be passed a string that is known
4668 to be a valid IP address. However, in the case of IPv6 addresses, it was
4669 not checking this. This is a hostage to fortune. Exim now panics and dies
4670 if the condition is not met. A case was found where this could be provoked
4671 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4672 components; fortuitously, this particular loophole had already been fixed
4673 by change 4.50/55 above.
4675 If there are any other similar loopholes, the new check in host_aton()
4676 itself should stop them being exploited. The report I received stated that
4677 data on the command line could provoke the exploit when Exim was running as
4678 exim, but did not say which command line option was involved. All I could
4679 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4680 running as the user.
4683 61. There was a buffer overflow vulnerability in the SPA authentication code
4684 (which came originally from the Samba project). I have added a test to the
4685 spa_base64_to_bits() function which I hope fixes it.
4688 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4689 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4691 63. The daemon start-up calls getloadavg() while still root for those OS that
4692 need the first call to be done as root, but it missed one case: when
4693 deliver_queue_load_max is set with deliver_drop_privilege. This is
4694 necessary for the benefit of the queue runner, because there is no re-exec
4695 when deliver_drop_privilege is set.
4697 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4698 This has been fixed.
4700 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4701 was in use, was not putting the data itself into the right store pool;
4702 consequently, it could be overwritten for a subsequent message in the same
4703 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4706 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4708 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4709 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4710 the function were treating the return as a boolean value, which happened to
4711 work because 0=false and not-0=true, but is not correct code.
4713 68. The host_aton() function was not handling scoped IPv6 addresses (those
4714 with, for example, "%eth0" on the end) correctly.
4716 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4717 negated items (that is, ~something) in unsigned ints. Some compilers
4718 apparently mutter when there is no cast.
4720 70. If an address verification called from an ACL failed, and did not produce a
4721 user-specific message (i.e. there was only a "system" message), nothing was
4722 put in $acl_verify_message. In this situation, it now puts the system
4725 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4726 session; change 4.31/43 added the unwanted input to the log line - except
4727 that it did not do this in the start of session case. It now does.
4729 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4730 This is wrong and can cause the other end to generate a synchronization
4731 error if it is another Exim or anything else that does the synchronization
4732 check. A QUIT command is no longer sent after a timeout.
4734 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4735 during host lookups.
4737 74. The defer_ok option of callout verification was not working if it was used
4738 when verifying addresses in header lines, that is, for this case:
4740 verify = header_sender/callout=defer_ok
4742 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4743 those file descriptors could be used for SMTP connections. If anything
4744 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4745 could be sent to the SMTP client, causing chaos. The daemon now opens
4746 stdin, stdout, and stderr to /dev/null when it puts itself into the
4749 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4750 log by default. The user can override this with suitable Perl magic.
4752 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4753 the log message when discard triggers, was not working for the DATA ACL or
4754 for the non-SMTP ACL.
4756 78. Error message wording change in sieve.c.
4758 79. If smtp_accept_max_per_host was set, the number of connections could be
4759 restricted to fewer than expected, because the daemon was trying to set up
4760 a new connection before checking whether the processes handling previous
4761 connections had finished. The check for completed processes is now done
4762 earlier. On busy systems, this bug wouldn't be noticed because something
4763 else would have woken the daemon, and it would have reaped the completed
4766 80. If a message was submitted locally by a user whose login name contained one
4767 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4768 It caused a spool format error. I have fixed the spool reading code. A
4769 related problem was that the "from" clause in the Received: line became
4770 illegal because of the space(s). It is now covered by ${quote_local_part.
4772 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4775 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4777 83. Patch from Sieve maintainer to fix unterminated string problem in
4778 "vacation" handling.
4780 84. Some minor changes to the Linux configuration files to help with other
4781 OS variants using glibc.
4783 85. One more patch for Sieve to update vacation handling to latest spec.
4786 ----------------------------------------------------
4787 See the note above about the 4.44 and 4.50 releases.
4788 ----------------------------------------------------
4794 1. Change 4.43/35 introduced a bug that caused file counts to be
4795 incorrectly computed when quota_filecount was set in an appendfile
4798 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4799 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4802 3. Give more explanation in the error message when the command for a transport
4803 filter fails to execute.
4805 4. There are several places where Exim runs a non-Exim command in a
4806 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4807 was being done only for the command run by the queryprogram router. It is
4808 now done for all such subprocesses. The other cases are: ${run, transport
4809 filters, and the commands run by the lmtp and pipe transports.
4811 5. Some older OS have a limit of 256 on the maximum number of file
4812 descriptors. Exim was using setrlimit() to set 1000 as a large value
4813 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4814 systems. I've change it so that if it can't get 1000, it tries for 256.
4816 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4817 was an oversight, and furthermore, ever since the addition of extra
4818 controls (e.g. 4.43/32), the checks on when to allow different forms of
4819 "control" were broken. There should now be diagnostics for all cases when a
4820 control that does not make sense is encountered.
4822 7. $recipients is now available in the predata ACL (oversight).
4824 8. Tidy the search cache before the fork to do a delivery from a message
4825 received from the command line. Otherwise the child will trigger a lookup
4826 failure and thereby defer the delivery if it tries to use (for example) a
4827 cached ldap connection that the parent has called unbind on.
4829 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4830 of $address_data from the recipient verification was clobbered by the
4831 sender verification.
4833 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4834 was its contents. (It was OK if the option was not defined at all.)
4836 11. A "Completed" log line is now written for messages that are removed from
4837 the spool by the -Mrm option.
4839 12. $host_address is now set to the target address during the checking of
4840 ignore_target_hosts.
4842 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4843 being passed; this would have caused $sender_host_name to have been used if
4844 matching the list had actually called for a host name (not very likely,
4845 since this list is usually IP addresses). A host name is now passed as
4848 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4849 code that specifies a non-restarting handler (typically sigaction() in
4850 modern systems) in an attempt to fix a rare and obscure crash bug.
4852 15. Narrowed the window for a race in the daemon that could cause it to ignore
4853 SIGCHLD signals. This is not a major problem, because they are used only to
4854 wake it up if nothing else does.
4856 16. A malformed maildirsize file could cause Exim to calculate negative values
4857 for the mailbox size or file count. Odd effects could occur as a result.
4858 The maildirsize information is now recalculated if the size or filecount
4861 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4862 support for a long time. Removed HAVE_SYS_VFS_H.
4864 18. Updated exipick to current release from John Jetmore.
4866 19. Allow an empty sender to be matched against a lookup in an address list.
4867 Previously the only cases considered were a regular expression, or an
4870 20. Exim went into a mad DNS lookup loop when doing a callout where the
4871 host was specified on the transport, if the DNS lookup yielded more than
4874 21. The RFC2047 encoding function was originally intended for short strings
4875 such as real names; it was not keeping to the 75-character limit for
4876 encoded words that the RFC imposes. It now respects the limit, and
4877 generates multiple encoded words if necessary. To be on the safe side, I
4878 have increased the buffer size for the ${rfc2047: expansion operator from
4881 22. Failure to deliver a bounce message always caused it to be frozen, even if
4882 there was an errors_to setting on the router. The errors_to setting is now
4885 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4886 canonical form (fully expanded) before being placed in
4887 $sender_host_address.
4889 24. Updated eximstats to version 1.33
4891 25. Include certificate and key file names in error message when GnuTLS fails
4892 to set them up, because the GnuTLS error message doesn't include the name
4893 of the failing file when there is a problem reading it.
4895 26. Expand error message when OpenSSL has problems setting up cert/key files.
4898 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4899 (this can affect the format of dates).
4901 28. exim_tidydb, when checking for the continued existence of a message for
4902 which it has found a message-specific retry record, was not finding
4903 messages that were in split spool directories. Consequently, it was
4904 deleting retry records that should have stayed in existence.
4906 29. eximstats updated to version 1.35
4907 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4908 1.35 - bugfix such that pie charts by volume are generated correctly
4910 30. The SPA authentication driver was not abandoning authentication and moving
4911 on to the next authenticator when an expansion was forced to fail,
4912 contradicting the general specification for all authenticators. Instead it
4913 was generating a temporary error. It now behaves as specified.
4915 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4916 (the order specifies the preference for clients). The order is now AES256,
4917 AES128, 3DES, ARCFOUR128.
4919 31. Small patch to Sieve code - explicitly set From: when generating an
4922 32. Exim crashed if a remote delivery caused a very long error message to be
4923 recorded - for instance if somebody sent an entire SpamAssassin report back
4924 as a large number of 550 error lines. This bug was coincidentally fixed by
4925 increasing the size of one of Exim's internal buffers (big_buffer) that
4926 happened as part of the Exiscan merge. However, to be on the safe side, I
4927 have made the code more robust (and fixed the comments that describe what
4930 33. Some experimental protocols are using DNS PTR records for new purposes. The
4931 keys for these records are domain names, not reversed IP addresses. The
4932 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4933 leaves it alone. Component reversal etc. now happens only for IP addresses.
4936 34. The host_aton() function is supposed to be passed a string that is known
4937 to be a valid IP address. However, in the case of IPv6 addresses, it was
4938 not checking this. This is a hostage to fortune. Exim now panics and dies
4939 if the condition is not met. A case was found where this could be provoked
4940 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4941 components; fortuitously, this particular loophole had already been fixed
4942 by change 4.50/55 or 4.44/33 above.
4944 If there are any other similar loopholes, the new check in host_aton()
4945 itself should stop them being exploited. The report I received stated that
4946 data on the command line could provoke the exploit when Exim was running as
4947 exim, but did not say which command line option was involved. All I could
4948 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4949 running as the user.
4952 35. There was a buffer overflow vulnerability in the SPA authentication code
4953 (which came originally from the Samba project). I have added a test to the
4954 spa_base64_to_bits() function which I hope fixes it.
4957 36. The daemon start-up calls getloadavg() while still root for those OS that
4958 need the first call to be done as root, but it missed one case: when
4959 deliver_queue_load_max is set with deliver_drop_privilege. This is
4960 necessary for the benefit of the queue runner, because there is no re-exec
4961 when deliver_drop_privilege is set.
4963 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4964 was in use, was not putting the data itself into the right store pool;
4965 consequently, it could be overwritten for a subsequent message in the same
4966 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4969 38. Sometimes the final signoff response after QUIT could fail to get
4970 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4971 < 0 before doing a fflush(). This bug looks as though it goes back to the
4972 introduction of TLS in release 3.20, but "sometimes" must have been rare
4973 because the tests only now provoked it.
4979 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4980 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4981 Now it returns an integer. A number of calls were still expecting a T/F
4982 return. Fortuitously, in all cases, the tests worked in OK situations,
4983 which is the norm. However, things would have gone wrong on any write
4984 failures on the smtp file descriptor. This function is used when sending
4985 messages over SMTP and also when doing verify callouts.
4987 2. When Exim is called to do synchronous delivery of a locally submitted
4988 message (the -odf or -odi options), it no longer closes stderr before doing
4991 3. Implemented the mua_wrapper option.
4993 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4995 5. Implemented the functions header_remove(), header_testname(),
4996 header_add_at_position(), and receive_remove_recipient(), and exported them
4999 6. If an ACL "warn" statement specified the addition of headers, Exim already
5000 inserted X-ACL-Warn: at the start if there was no header name. However, it
5001 was not making this test for the second and subsequent header lines if
5002 there were newlines in the string. This meant that an invalid header could
5003 be inserted if Exim was badly configured.
5005 7. Allow an ACL "warn" statement to add header lines at the start or after all
5006 the Received: headers, as well as at the end.
5008 8. Added the rcpt_4xx retry error code.
5010 9. Added postmaster_mailfrom=xxx to callout verification option.
5012 10. Added mailfrom=xxxx to the callout verification option, for verify=
5015 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5016 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5018 12. Inserted some casts to stop certain compilers warning when using pointer
5019 differences as field lengths or precisions in printf-type calls (mostly
5020 affecting debugging statements).
5022 13. Added optional readline() support for -be (dynamically loaded).
5024 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5025 same clock tick as a message's arrival, so that its received time was the
5026 same as the "first fail" time on the retry record, and that message
5027 remained on the queue past the ultimate address timeout, every queue runner
5028 would try a delivery (because it was past the ultimate address timeout) but
5029 after another failure, the ultimate address timeout, which should have then
5030 bounced the address, did not kick in. This was a "< instead of <=" error;
5031 in most cases the first failure would have been in the next clock tick
5032 after the received time, and all would be well.
5034 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5035 being recognized when the domain list was tested by the match_domain
5036 condition in an expansion string.
5038 16. Added the ${str2b64: operator.
5040 17. Exim was always calling setrlimit() to set a large limit for the number of
5041 processes, without checking whether the existing limit was already
5042 adequate. (It did check for the limit on file descriptors.) Furthermore,
5043 errors from getrlimit() and setrlimit() were being ignored. Now they are
5044 logged to the main and panic logs, but Exim does carry on, to try to do its
5045 job under whatever limits there are.
5047 18. Imported PCRE 5.0.
5049 19. Trivial typo in log message " temporarily refused connection" (the leading
5052 20. If the log selector return_path_on_delivery was set and an address was
5053 redirected to /dev/null, the delivery process crashed because it assumed
5054 that a return path would always be set for a "successful" delivery. In this
5055 case, the whole delivery is bypassed as an optimization, and therefore no
5058 21. Internal re-arrangement: the function for sending a challenge and reading
5059 a response while authentication was assuming a zero-terminated challenge
5060 string. It's now changed to take a pointer and a length, to allow for
5061 binary data in such strings.
5063 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5065 23. Exim was not respecting finduser_retries when seeking the login of the
5066 uid under which it was called; it was always trying 10 times. (The default
5067 setting of finduser_retries is zero.) Also, it was sleeping after the final
5068 failure, which is pointless.
5070 24. Implemented tls_on_connect_ports.
5072 25. Implemented acl_smtp_predata.
5074 26. If the domain in control=submission is set empty, Exim assumes that the
5075 authenticated id is a complete email address when it generates From: or
5076 Sender: header lines.
5078 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5079 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5080 chown and chgrp in /bin and hostname in /usr/bin.
5082 28. Exim was keeping the "process log" file open after each use, just as it
5083 does for the main log. This opens the possibility of it remaining open for
5084 long periods when the USR1 signal hits a daemon. Occasional processlog
5085 errors were reported, that could have been caused by this. Anyway, it seems
5086 much more sensible not to leave this file open at all, so that is what now
5089 29. The long-running daemon process does not normally write to the log once it
5090 has entered its main loop, and it closes the log before doing so. This is
5091 so that log files can straightforwardly be renamed and moved. However,
5092 there are a couple of unusual error situations where the daemon does write
5093 log entries, and I had neglected to close the log afterwards.
5095 30. The text of an SMTP error response that was received during a remote
5096 delivery was being truncated at 512 bytes. This is too short for some of
5097 the long messages that one sometimes sees. I've increased the limit to
5100 31. It is now possible to make retry rules that apply only when a message has a
5101 specific sender, in particular, an empty sender.
5103 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5104 it possible to be selective about when SMTP synchronization is enforced.
5106 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5108 32. Implemented hosts_connection_nolog.
5110 33. Added an ACL for QUIT.
5112 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5115 35. Added mailbox_size and mailbox_filecount to appendfile.
5117 36. Added control = no_multiline_responses to ACLs.
5119 37. There was a bug in the logic of the code that waits for the clock to tick
5120 in the case where the clock went backwards by a substantial amount such
5121 that the microsecond fraction of "now" was more than the microsecond
5122 fraction of "then" (but the whole seconds number was less).
5124 38. Added support for the libradius Radius client library this is found on
5125 FreeBSD (previously only the radiusclient library was supported).
5131 1. When certain lookups returned multiple values in the form name=value, the
5132 quoting of the values was not always being done properly. Specifically:
5133 (a) If the value started with a double quote, but contained no whitespace,
5135 (b) If the value contained whitespace other than a space character (i.e.
5136 tabs or newlines or carriage returns) it was not quoted.
5137 This fix has been applied to the mysql and pgsql lookups by writing a
5138 separate quoting function and calling it from the lookup code. The fix
5139 should probably also be applied to nisplus, ibase and oracle lookups, but
5140 since I cannot test any of those, I have not disturbed their existing code.
5142 2. A hit in the callout cache for a specific address caused a log line with no
5143 reason for rejecting RCPT. Now it says "Previous (cached) callout
5144 verification failure".
5146 3. There was an off-by-one bug in the queryprogram router. An over-long
5147 return line was truncated at 256 instead of 255 characters, thereby
5148 overflowing its buffer with the terminating zero. As well as fixing this, I
5149 have increased the buffer size to 1024 (and made a note to document this).
5151 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5152 when Exim is waiting for an SMTP response from a remote server, Exim
5153 restarts its select() call on the socket, thereby resetting its timeout.
5154 This is not a problem when such interrupts are rare. Somebody set up a cron
5155 job to run exiwhat every 2 minutes, which is less than the normal select()
5156 timeout (5 or 10 minutes). This meant that the select() timeout never
5157 kicked in because it was always reset. I have fixed this by comparing the
5158 time when an interrupt arrives with the time at the start of the first call
5159 to select(). If more time than the timeout has elapsed, the interrupt is
5160 treated as a timeout.
5162 5. Some internal re-factoring in preparation for the addition of Sieve
5163 extensions (by MH). In particular, the "personal" test is moved to a
5164 separate function, and given an option for scanning Cc: and Bcc: (which is
5165 not set for Exim filters).
5167 6. When Exim created an email address using the login of the caller as the
5168 local part (e.g. when creating a From: or Sender: header line), it was not
5169 quoting the local part when it contained special characters such as @.
5171 7. Installed new OpenBSD configuration files.
5173 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5174 try to make them clearer.
5176 9. Callout options, other than the timeout value, were being ignored when
5177 verifying sender addresses in header lines. For example, when using
5179 verify = header_sender/callout=no_cache
5181 the cache was (incorrectly) being used.
5183 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5184 only the Cygwin environment.
5186 11. When return_path_on_delivery was set as a log selector, if different remote
5187 addresses in the same message used different return paths and parallel
5188 remote delivery occurred, the wrong values would sometimes be logged.
5189 (Whenever a remote delivery process finished, the return path value from
5190 the most recently started remote delivery process was logged.)
5192 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5193 lines when AUTH and/or TLS are in use. This is the "received protocol"
5194 field. Exim used to use "asmtp" for authenticated SMTP, without any
5195 indication (in the protocol name) for TLS use. Now it follows the RFC and
5196 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5197 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5198 names appear in log lines as well as in Received: header lines.
5200 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5201 extensions, and comparison tests, and to fix some bugs.
5203 14. Changes to the "personal" filter test:
5205 (1) The test was buggy in that it was just doing the equivalent of
5206 "contains" tests on header lines. For example, if a user's address was
5207 anne@some.where, the "personal" test would incorrectly be true for
5209 To: susanne@some.where
5211 This test is now done by extracting each address from the header in turn,
5212 and checking the entire address. Other tests that are part of "personal"
5213 are now done using regular expressions (for example, to check local parts
5214 of addresses in From: header lines).
5216 (2) The list of non-personal local parts in From: addresses has been
5217 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5218 taken from the Sieve specification recommendations.
5220 (3) If the message contains any header line starting with "List-" it is
5221 treated as non-personal.
5223 (4) The test for "circular" in the Subject: header line has been removed
5224 because it now seems ill-conceived.
5226 15. Minor typos in src/EDITME comments corrected.
5228 16. Installed latest exipick from John Jetmore.
5230 17. If headers_add on a router specified a text string that was too long for
5231 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5232 of string_sprintf() is now avoided.
5234 18. $message_body_size was not set (it was always zero) when running the DATA
5235 ACL and the local_scan() function.
5237 19. For the "mail" command in an Exim filter, no default was being set for
5238 the once_repeat time, causing a random time value to be used if "once" was
5239 specified. (If the value happened to be <= 0, no repeat happened.) The
5240 default is now 0s, meaning "never repeat". The "vacation" command was OK
5241 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5242 (I found it when inspecting the code).
5244 20. There is now an overall timeout for performing a callout verification. It
5245 defaults to 4 times the callout timeout, which applies to individual SMTP
5246 commands during the callout. The overall timeout applies when there is more
5247 than one host that can be tried. The timeout is checked before trying the
5248 next host. This prevents very long delays if there are a large number of
5249 hosts and all are timing out (e.g. when the network connections are timing
5250 out). The value of the overall timeout can be changed by specifying an
5251 additional sub-option for "callout", called "maxwait". For example:
5253 verify = sender/callout=5s,maxwait=20s
5255 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5256 to the end before writing, but this should make it even safer).
5258 22. Exim was forgetting that it had advertised PIPELINING for the second and
5259 subsequent messages on an SMTP connection. It was also not resetting its
5260 memory on STARTTLS and an internal HELO.
5262 23. When Exim logs an SMTP synchronization error within a session, it now
5263 records whether PIPELINING has been advertised or not.
5265 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5266 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5267 rather than long int.
5269 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5271 26. Added the never_mail option to autoreply.
5277 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5278 crash if the getsockname() call failed; this can happen if a connection is
5279 closed very soon after it is established. The problem was simply in the
5280 order in which certain operations were done, causing Exim to try to write
5281 to the SMTP stream before it had set up the file descriptor. The bug has
5282 been fixed by making things happen in the correct order.
5288 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5289 before the connection was closed, thus losing the rejection response.
5291 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5292 some early Solaris releases, but causes trouble in current releases where
5293 socklen_t is defined.
5295 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5298 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5301 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5302 that releasing the top of it at the end releases what was used for sub-
5303 expansions (unless the block got too big). However, discard this block if
5304 the first thing is a variable or header, so that we can use its block when
5305 it is dynamic (useful for very large $message_headers, for example).
5307 6. Lookups now cache *every* query, not just the most recent. A new, separate
5308 store pool is used for this. It can be recovered when all lookup caches are
5309 flushed. Lookups now release memory at the end of their result strings.
5310 This has involved some general refactoring of the lookup sources.
5312 7. Some code has been added to the store_xxx() functions to reduce the amount
5313 of flapping under certain conditions.
5315 8. log_incoming_interface used to affect only the <= reception log lines. Now
5316 it causes the local interface and port to be added to several more SMTP log
5317 lines, for example "SMTP connection from", and rejection lines.
5319 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5321 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5323 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5324 could be overwritten at the end of the current message (or the start of a
5325 new message if it was set in a HELO ACL). The value is now preserved for
5326 the duration of the SMTP connection.
5328 12. If a transport had a headers_rewrite setting, and a matching header line
5329 contained an unqualified address, that address was qualified, even if it
5330 did not match any rewriting rules. The underlying bug was that the values
5331 of the flags that permit the existence of unqualified sender and recipient
5332 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5333 non-local messages, and by -bnq for local messages) were not being
5334 preserved with the message after it was received.
5336 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5337 "next input=" as part of the text comprising the host identity instead of
5338 the correct text. The code was using the same buffer for two different
5339 strings. However, depending on which order the printing function evaluated
5340 its arguments, the bug did not always show up. Under Linux, for example, my
5341 test suite worked just fine.
5343 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5344 doesn't work with some older versions of Perl. It has been changed to "my",
5345 which in any case is probably the better facility to use.
5347 15. A really picky compiler found some instances of statements for creating
5348 error messages that either had too many or two few arguments for the format
5351 16. The size of the buffer for calls to the DNS resolver has been increased
5352 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5353 for addresses that have a lot of PTR records. This alleviates a problem; it
5354 does not fully solve it.
5356 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5357 buffer now truncates the list and logs the incident, which is the same
5358 action as happens when Exim is looking up a host name and its aliases.
5359 Previously in this situation something unpredictable would happen;
5360 sometimes it was "internal error: store_reset failed".
5362 18. If a server dropped the connection unexpectedly when an Exim client was
5363 using GnuTLS and trying to read a response, the client delivery process
5364 crashed while trying to generate an error log message.
5366 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5367 string, for example:
5369 warn message = H1: something\nH2: something
5371 the text was added as a single header line from Exim's point of view
5372 though it ended up OK in the delivered message. However, searching for the
5373 second and subsequent header lines using $h_h2: did not work. This has been
5374 fixed. Similarly, if a system filter added multiple headers in this way,
5375 the routers could not see them.
5377 20. Expanded the error message when iplsearch is called with an invalid key to
5378 suggest using net-iplsearch in a host list.
5380 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5381 ACLs are no longer actually imposed (and a message to that effect is
5384 22. If a "gecos" field in a passwd entry contained escaped characters, in
5385 particular, if it contained a \" sequence, Exim got it wrong when building
5386 a From: or a Sender: header from that name. A second bug also caused
5387 incorrect handling when an unquoted " was present following a character
5388 that needed quoting.
5390 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5391 was not being matched caselessly.
5393 24. Arranged for all hyphens in the exim.8 source to be escaped with
5396 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5397 properly. Recipient callout cache records were still being keyed to include
5398 the sender, even when use_sender was set false. This led to far more
5399 callouts that were necessary. The sender is no longer included in the key
5400 when use_sender is false.
5402 26. Added "control = submission" modifier to ACLs.
5404 27. Added the ${base62d: operator to decode base 62 numbers.
5406 28. dnsdb lookups can now access SRV records.
5408 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5409 the configuration file.
5411 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5412 (-v) mode. This makes the output for a verbose queue run more intelligible.
5414 31. Added a use_postmaster feature to recipient callouts.
5416 32. Added the $body_zerocount variable, containing the number of binary zero
5417 bytes in the message body.
5419 33. The time of last modification of the "new" subdirectory is now used as the
5420 "mailbox time last read" when there is a quota error for a maildir
5423 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5425 35. Added +ignore_unknown as a special item in host lists.
5427 36. Code for decoding IPv6 addresses in host lists is now included, even if
5428 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5429 address was recognized as an IP address, but was then not correctly decoded
5430 into binary, causing unexpected and incorrect effects when compared with
5437 1. Very minor rewording of debugging text in manualroute to say "list of
5438 hosts" instead of "hostlist".
5440 2. If verify=header_syntax was set, and a header line with an unqualified
5441 address (no domain) and a large number of spaces between the end of the
5442 name and the colon was received, the reception process suffered a buffer
5443 overflow, and (when I tested it) crashed. This was caused by some obsolete
5444 code that should have been removed. The fix is to remove it!
5446 3. When running in the test harness, delay a bit after writing a bounce
5447 message to get a bit more predictability in the log output.
5449 4. Added a call to search_tidyup() just before forking a reception process. In
5450 theory, someone could use a lookup in the expansion of smtp_accept_max_
5451 per_host which, without the tidyup, could leave open a database connection.
5453 5. Added the variables $recipient_data and $sender_data which get set from a
5454 lookup success in an ACL "recipients" or "senders" condition, or a router
5455 "senders" option, similar to $domain_data and $local_part_data.
5457 6. Moved the writing of debug_print from before to after the "senders" test
5460 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5461 problems for message scanning, either using a data ACL, or using
5462 local_scan() because the Received: header was not generated till after they
5463 were called (in order to set the time as the time of reception completion).
5464 I have revised the way this works. The header is now generated after the
5465 body is received, but before the ACL or local_scan() are called. After they
5466 are run, the timestamp in the header is updated.
5472 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5473 before starting a queue runner without re-exec. This happened only when
5474 deliver_drop_privilege was set or when the Exim user was set to root. The
5475 effect of the bug was that timeouts during subsequent deliveries caused
5476 crashes instead of being properly handled. The handler is now left at its
5477 default (and expected) setting.
5479 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5480 message, again when deliver_drop_privilege is set or Exim is run as root.
5481 The bug described in (1) was not present in this case, but the tidying up
5482 of the other signals was missing. I have made the two cases consistent.
5484 3. The ignore_target_hosts setting on a manualroute router was being ignored
5485 for hosts that were looked up using the /MX notation.
5487 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5490 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5491 operated on the sender address. After changing the $sender_address to <>
5492 for the sender address verify, Exim was re-instated it as the original
5493 (before rewriting) address, but remembering that it had rewritten it, so it
5494 wasn't rewriting it again. This bug also had the effect of breaking the
5495 sender address verification caching when the sender address was rewritten.
5497 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5498 This has been changed so that if the ip literal address matches
5499 ignore_target_hosts, the router declines.
5501 7. Added expansion conditions match_domain, match_address, and match_local_
5502 part (NOT match_host).
5504 8. The placeholder for the Received: header didn't have a length field set.
5506 9. Added code to Exim itself and to exim_lock to test for a specific race
5507 condition that could lead to file corruption when using MBX delivery. The
5508 issue is with the lockfile that is created in /tmp. If this file is removed
5509 after a process has opened it but before that process has acquired a lock,
5510 there is the potential for a second process to recreate the file and also
5511 acquire a lock. This could lead to two Exim processes writing to the file
5512 at the same time. The added code performs the same test as UW imapd; it
5513 checks after acquiring the lock that its file descriptor still refers to
5514 the same named file.
5516 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5517 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5520 11. Added the smtp_active_hostname option. If used, this will typically be made
5521 to depend on the incoming interface address. Because $interface_address is
5522 not set up until the daemon has forked a reception process, error responses
5523 that can happen earlier (such as "too many connections") no longer contain
5526 12. If an expansion in a condition on a "warn" statement fails because a lookup
5527 defers, the "warn" statement is abandoned, and the next ACL statement is
5528 processed. Previously this caused the whole ACL to be aborted.
5530 13. Added the iplsearch lookup type.
5532 14. Added ident_timeout as a log selector.
5534 15. Added tls_certificate_verified as a log selector.
5536 16. Added a global option tls_require_ciphers (compare the smtp transport
5537 option of the same name). This controls incoming TLS connections.
5539 17. I finally figured out how to make tls_require_ciphers do a similar thing
5540 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5541 before starting the TLS session.
5543 18. Tabs are now shown as \t in -bP output.
5545 19. If the log selector return_path_on_delivery was set, Exim crashed when
5546 bouncing a message because it had too many Received: header lines.
5548 20. If two routers both had headers_remove settings, and the first one included
5549 a superfluous trailing colon, the final name in the first list and the
5550 first name in the second list were incorrectly joined into one item (with a
5551 colon in the middle).
5557 1. Added -C and -D options to the exinext utility, mainly to make it easier
5558 to include in the automated testing, but these could be helpful when
5559 multiple configurations are in use.
5561 2. The exinext utility was not formatting the output nicely when there was
5562 an alternate port involved in the retry record key, nor when there was a
5563 message id as well (for retries that were specific to a specific message
5564 and a specific host). It was also confused by IPv6 addresses, because of
5565 the additional colons they contain. I have fixed the IPv4 problem, and
5566 patched it up to do a reasonable job for IPv6.
5568 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5569 delivery, the log line now contains "pipelined" if PIPELINING was used.
5571 4. An SMTP transport process used to panic and die if the bind() call to set
5572 an explicit outgoing interface failed. This has been changed; it is now
5573 treated in the same way as a connect() failure.
5575 5. A reference to $sender_host_name in the part of a conditional expansion
5576 that was being skipped was still causing a DNS lookup. This no longer
5579 6. The def: expansion condition was not recognizing references to header lines
5580 that used bh_ and bheader_.
5582 7. Added the _cache feature to named lists.
5584 8. The code for checking quota_filecount in the appendfile transport was
5585 allowing one more file than it should have been.
5587 9. For compatibility with Sendmail, the command line option
5595 and sets the incoming protocol and host name (for trusted callers). The
5596 host name and its colon can be omitted when only the protocol is to be set.
5597 Note the Exim already has two private options, -pd and -ps, that refer to
5598 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5599 "s", but I don't think that's a major issue.
5601 10. A number of refactoring changes to the code, none of which should affect
5604 (a) The number of logging options was getting close to filling up the
5605 32-bit word that was used as a bit map. I have split them into two classes:
5606 those that are passed in the argument to log_write(), and those that are
5607 only ever tested independently outside of that function. These are now in
5608 separate 32-bit words, so there is plenty of room for expansion again.
5609 There is no change in the user interface or the logging behaviour.
5611 (b) When building, for example, log lines, the code previously used a
5612 macro that called string_cat() twice, in order to add two strings. This is
5613 not really sufficiently general. Furthermore, there was one instance where
5614 it was actually wrong because one of the argument was used twice, and in
5615 one call a function was used. (As it happened, calling the function twice
5616 did not affect the overall behaviour.) The macro has been replaced by a
5617 function that can join an arbitrary number of extra strings onto a growing
5620 (c) The code for expansion conditions now uses a table and a binary chop
5621 instead of a serial search (which was left over from when there were very
5622 few conditions). Also, it now recognizes conditions like "pam" even when
5623 the relevant support is not compiled in: a suitably worded error message is
5624 given if an attempt is made to use such a condition.
5626 11. Added ${time_interval:xxxxx}.
5628 12. A bug was causing one of the ddress fields not to be passed back correctly
5629 from remote delivery subprocesses. The field in question was not being
5630 subsequently used, so this caused to problems in practice.
5632 13. Added new log selectors queue_time and deliver_time.
5634 14. Might have fixed a bug in maildirsizefile handling that threw up
5635 "unexpected character" debug warnings, and recalculated the data
5636 unnecessarily. In any case, I expanded the warning message to give more
5639 15. Added the message "Restricted characters in address" to the statements in
5640 the default ACL that block characters like @ and % in local parts.
5642 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5643 Three changes have been made:
5645 (a) There was a serious bug; a negative response to MAIL caused the whole
5646 recipient domain to be cached as invalid, thereby blocking all messages
5647 to all local parts at the same domain, from all senders. This bug has
5648 been fixed. The domain is no longer cached after a negative response to
5649 MAIL if the sender used is not empty.
5651 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5654 (c) A new callout option, "use_sender" has been added for people who want
5655 the modified behaviour.
5661 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5664 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5665 indeed breaks things for older releases.
5667 3. Added additional logging to the case where there is a problem reading data
5668 from a filter that is running in a subprocess using a pipe, in order to
5669 try to track down a specific problem.
5671 4. Testing facility fudge: when running in the test harness and attempting
5672 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5673 getting "No route to host". Convert this to a timeout.
5675 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5678 6. Some OS don't have socklen_t but use size_t instead. This affects the
5679 fifth argument of getsockopt() amongst other things. This is now
5680 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5681 can be set for individual OS. I have set it for SunOS5, OSF1, and
5682 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5683 some earlier ones do not.
5685 7. Change 4.30/15 was not doing the test caselessly.
5687 8. The standard form for an IPv6 address literal was being rejected by address
5688 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5689 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5690 this, as well as the form without the "IPv6" on the front (but only when
5691 address literals are enabled, of course).
5693 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5695 10. Exim crashed if a message with an empty sender address specified by -f
5696 encountered a router with an errors_to setting. This could be provoked only
5697 by a command such as
5701 where an empty string was supplied; "<>" did not hit this bug.
5703 11. Installed PCRE release 4.5.
5705 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5706 remained set. It is now erased.
5708 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5709 times from message ids (which are base 36 rather than the normal 62).
5711 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5712 were being counted as actual protocol errors, and logged if the log
5713 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5714 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5715 and DATA following a set of rejected RCPTs do not count as protocol errors.
5716 In other words, Exim assumes they were pipelined, though this may not
5717 actually be the case. Of course, in all cases the client gets an
5718 appropriate error code.
5720 15. If a lookup fails in an ACL condition, a message about the failure may
5721 be available; it is used if testing the ACL cannot continue, because most
5722 such messages specify what the cause of the deferral is. However, some
5723 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5724 that caused an old message to be retained and used if a later statement
5725 caused a defer, replacing the real cause of the deferral.
5727 16. If an IP address had so many PTR records that the DNS lookup buffer
5728 was not large enough to hold them, Exim could crash while trying to process
5729 the truncated data. It now detects and logs this case.
5731 17. Further to 4.21/58, another change has been made: if (and only if) the
5732 first line of a message (the first header line) ends with CRLF, a bare LF
5733 in a subsequent header line has a space inserted after it, so as not to
5734 terminate the header.
5736 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5737 unnecessarily, used atoi() instead of strtol(), and didn't check the
5738 termination when getting file sizes from file names by regex.
5740 19. Completely re-implemented the support for maildirsize files, in the light
5741 of a number of problems with the previous contributed implementation
5742 (4.30/29). In particular:
5744 . If the quota is zero, the maildirsize file is maintained, but no quota is
5747 . If the maildir directory does not exist, it is created before any attempt
5748 to write a maildirsize file.
5750 . The quota value in the file is just a cache; if the quota is changed in
5751 the transport, the new value overrides.
5753 . A regular expression is available for excluding directories from the
5756 20. The autoreply transport checks the characters in options that define the
5757 message's headers; it allows continued headers, but it was checking with
5758 isspace() after an embedded newline instead of explicitly looking for a
5761 21. If all the "regular" hosts to which an address was routed had passed their
5762 expiry times, and had not reached their retry times, the address was
5763 bounced, even if fallback hosts were defined. Now Exim should go on to try
5766 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5767 equivalent code in the SMTP transport. Some hosts send humungous responses
5768 to HELO/EHLO, more than 1024 it seems.
5770 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5771 gives compiler warnings in some environments. I've now done it "properly",
5774 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5775 (because of problems with the built-in one) was declared to return uschar *
5776 instead of char *, causing compiler failure.
5778 25. Fixed a file descriptor leak when processing alias/forward files.
5780 26. Fixed a minor format string issue in dbfn.c.
5782 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5784 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5785 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5788 29. When a sender address is verified, it is cached, to save repeating the test
5789 when there is more than one recipient in a message. However, when the
5790 verification involves a callout, it is possible for different callout
5791 options to be set for different recipients. It is too complicated to keep
5792 track of this in the cache, so now Exim always runs a verification when a
5793 callout is required, relying on the callout cache for the optimization.
5794 The overhead is duplication of the address routing, but this should not be
5797 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5798 to be verified with callout=postmaster, and the main callout worked but the
5799 postmaster check failed, the verification correctly failed. However, if a
5800 subsequent RCPT command asked for sender verification *without* the
5801 postmaster check, incorrect caching caused this verification also to fail,
5804 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5805 it was not caching the DNS options (qualify_single, search_parents) that
5806 were used when the lookup failed. A subsequent lookup with different
5807 options therefore always gave the same answer, though there were cases
5808 where it should not have. (Example: a "domains = !$mx_any" option on a
5809 dnslookup router: the "domains" option is always processed without any
5810 widening, but the router might have qualify_single set.) Now Exim uses the
5811 cached value only when the same options are set.
5813 32. Added John Jetmore's "exipick" utility to the distribution.
5815 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5816 than a timeout (e.g. a certificate is required, and is not provided), an
5817 Exim server now closes the connection immediately. Previously it waited for
5818 the client to close - but if the client is SSL, it seems that they each
5819 wait for each other, leading to a delay before one of them times out.
5821 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5822 maintained 0.8.x compatibility because I don't think many are using it, and
5823 it is clearly obsolete.
5825 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5828 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5829 client certificate was expired. A simple patch fixes this, though I don't
5830 understand the full logic of why the verify callback is called multiple
5833 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5834 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5835 which causes problems with some clients (such as the Certicom SSL Plus
5836 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5837 disables the coutermeasure allowing Eudora to connect."
5839 38. Exim was not checking that a write() to a log file succeeded. This could
5840 lead to Bad Things if a log got too big, in particular if it hit a file
5841 size limit. Exim now panics and dies if it cannot write to a log file, just
5842 as it does if it cannot open a log file.
5844 39. Modified OS/Makefile-Linux so that it now contains
5846 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5848 The two -D definitions ensure that Exim is compiled with large file
5849 support, which makes it possible to handle log files that are bigger than
5852 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5853 instance) a domain was checked against a named list that involved a lookup,
5854 causing $domain_data to be set, then another domain was checked against the
5855 same list, then the first domain was re-checked, the value of $domain_data
5856 after the final check could be wrong. In particular, if the second check
5857 failed, it could be set empty. This bug probably also applied to
5860 41. The strip_trailing_dot option was not being applied to the address given
5861 with the -f command-line option.
5863 42. The code for reading a message's header from the spool was incrementing
5864 $received_count, but never initializing it. This meant that the value was
5865 incorrect (doubled) while delivering a message in the same process in which
5866 it was received. In the most common configuration of Exim, this never
5867 happens - a fresh exec is done - but it can happen when
5868 deliver_drop_privilege is set.
5870 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5871 it now includes up to 150 characters of the unexpected data in the log
5874 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5875 and building data strings. The size of both of these buffers was 10 000
5876 bytes - far larger than anybody would *ever* want, thought I. Needless to
5877 say, somebody hit the limit. I have increased the maximum line length to
5878 20 000 and the maximum data length of concatenated lines to 100 000. I have
5879 also fixed two bugs, because there was no checking on these buffers. Tsk,
5880 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5881 buffer is too small.
5883 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5884 lsearch lookups. Now it does.
5886 46. When parsing a route_list item in a manualroute router, a fixed-length
5887 buffer was used for the list of hosts. I made this 1024 bytes long,
5888 thinking that nobody would ever have a list of hosts that long. Wrong.
5889 Somebody had a whole pile of complicated expansion conditions, and the
5890 string was silently truncated, leading to an expansion error. It turns out
5891 that it is easier to change to an unlimited length (owing to other changes
5892 that have happened since this code was originally written) than to build
5893 structure for giving a limitation error. The length of the item that
5894 expands into the list of hosts is now unlimited.
5896 47. The lsearch lookup could not handle data where the length of text line was
5897 more than 4095 characters. Such lines were truncated, leading to shortened
5898 data being returned. It should now handle lines of any length.
5900 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5901 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5904 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5905 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5906 no longer happens in comment lines. A list of replacements is now placed
5907 at the head of all of the source files, except those whose only change is
5908 to replace PERL_COMMAND in the very first #! line.
5910 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5911 messages on the queue, with a bottom-up merge sort, using code contributed
5912 by Michael Haardt. This should make operations like -bp somewhat faster on
5913 large queues. It won't affect queue runners, except when queue_run_in_order
5916 51. Installed eximstats 1.31 in the distribution.
5918 52. Added support for SRV lookups to the dnslookup router.
5920 53. If an ACL referred to $message_body or $message_body_end, the value was not
5921 reset for any messages that followed in the same SMTP session.
5923 54. The store-handling optimization for building very long strings was not
5924 differentiating between the different store pools. I don't think this
5925 actually made any difference in practice, but I've tidied it.
5927 55. While running the routers to verify a sender address, $sender_address
5928 was still set to the sender address. This is wrong, because when routing to
5929 send a bounce to the sender, it would be empty. Therefore, I have changed
5930 it so that, while verifying a sender address, $sender_address is set to <>.
5931 (There is no change to what happens when verifying a recipient address.)
5933 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5934 target A or AAAA records (if not already returned) without resetting the
5935 qualify_single or search_parents options of the DNS resolver. These are
5936 inappropriate in this case because the targets of MX and SRV records must
5937 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5938 target that, when qualified, matched something in the local domain. These
5939 two options are now turned off when doing these lookups.
5941 57. It seems that at least some releases of Reiserfs (which does not have the
5942 concept of a fixed number of inodes) returns zero and not -1 for the
5943 number of available inodes. This interacted badly with check_spool_inodes,
5944 which assumed that -1 was the "no such thing" setting. What I have done is
5945 to check that the total number of inodes is greater than zero before doing
5946 the test of how many are available.
5948 58. When a "warn" ACL statement has a log_message modifier, the message is
5949 remembered, and not repeated. This is to avoid a lot of repetition when a
5950 message has many recipients that cause the same warning to be written.
5951 However, Exim was preserving the list of already written lines for an
5952 entire SMTP session, which doesn't seem right. The memory is now reset if a
5953 new message is started.
5955 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5956 debugging output unless log_rewrite was also set.
5958 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5959 of (int)(handle) when we know that handle contains (void *)(-1).
5961 61. The Exim daemon panic-logs an error return when it closes the incoming
5962 connection. However "connection reset by peer" seems to be common, and
5963 isn't really an error worthy of noting specially, so that particular error
5966 62. When Exim is trying to find all the local interfaces, it used to panic and
5967 die if the ioctl to get the interface flags failed. However, it seems that
5968 on at least one OS (Solaris 9) it is possible to have an interface that is
5969 included in the list of interfaces, but for which you get a failure error
5970 for this call. This happens when the interface is not "plumbed" into a
5971 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5972 failure of the "get flags" call assumes that the interface is down.
5974 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5975 makes life easier for people who are doing arithmetic on fields extracted
5976 from dates, where you often get leading zeros that should not be
5977 interpreted as octal.
5979 64. Added qualify_domain to the redirect router, to override the global
5982 65. If a pathologically long header line contained very many addresses (the
5983 report of this problem mentioned 10 000) and each of them was rewritten,
5984 Exim could use up a very large amount of memory. (It kept on making new
5985 copies of the header line as it rewrote, and never released the old ones.)
5986 At the expense of a bit more processing, the header rewriting function has
5987 been changed so that it no longer eats memory in this way.
5989 66. The generation of the Received: header has been moved from the time that a
5990 message starts to be received, to the time that it finishes. The timestamp
5991 in the Received: header should now be very close to that of the <= log
5992 line. There are two side-effects of this change:
5994 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5995 logged header lines no longer include the local Received: line, because
5996 it has not yet been created. The same applies to a copy of the message
5997 that is returned to a non-SMTP sender when a message is rejected.
5999 (b) When a filter file is tested using -bf, no additional Received: header
6000 is added to the test message. After some thought, I decided that this
6003 This change does not affect the value of $received_for. It is still set
6004 after address rewriting, but before local_scan() is called.
6006 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6008 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6009 gave an unhelpful panic error message, and a defer error. I have managed to
6010 change this behaviour so that it now rejects any supplied certificate,
6011 which seems right, as the list of acceptable certificates is empty.
6013 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6014 gave an unhelpful defer error. I have not managed to make this reject any
6015 supplied certificates, but the error message it gives is "no certificate
6016 supplied", which is not helpful.
6018 70. exigrep's output now also includes lines that are not associated with any
6019 message, but which match the given pattern. Implemented by a patch from
6020 Martin Sluka, which also tidied up the Perl a bit.
6022 71. Recipient callout verification, like sender verification, was using <> in
6023 the MAIL FROM command. This isn't really the right thing, since the actual
6024 sender may affect whether the remote host accepts the recipient or not. I
6025 have changed it to use the actual sender in the callout; this means that
6026 the cache record is now keyed on a recipient/sender pair, not just the
6027 recipient address. There doesn't seem to be a real danger of callout loops,
6028 since a callout by the remote host to check the sender would use <>.
6029 [SEE ABOVE: changed after hitting problems.]
6031 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6032 temporary errors. However, in the case of such a code being given after
6033 the end of a data transmission (i.e. after ".") Exim was failing to write
6034 a retry record for the message. (Yes, there was some broken host that was
6035 actually sending 8xx at this point.)
6037 73. An unknown lookup type in a host list could cause Exim to panic-die when
6038 the list was checked. (An example that provoked this was putting <; in the
6039 middle of a list instead of at the start.) If this happened during a DATA
6040 ACL check, a -D file could be left lying around. This kind of configuration
6041 error no longer causes Exim to die; instead it causes a defer error. The
6042 incident is still logged to the main and panic logs.
6044 74. Buglet left over from Exim 3 conversion. The message "too many messages
6045 in one connection" was written to the rejectlog but not the mainlog, except
6046 when address rewriting (yes!) was being logged.
6048 75. Added write_rejectlog option.
6050 76. When a system filter was run not as root (that is, when system_filter_user
6051 was set), the values of the $n variables were not being returned to the
6052 main process; thus, they were not subsequently available in the $sn
6055 77. Added +return_path_on_delivery log selector.
6057 78. A connection timeout was being treated differently from recipients deferred
6058 when testing hosts_max_try with a message that was older than the host's
6059 retry timeout. (The host should not be counted, thus allowing all hosts to
6060 be tried at least once before bouncing.) This may have been the cause of an
6061 occasionally reported bug whereby a message would remain on the queue
6062 longer than the retry timeout, but would be bounced if a delivery was
6063 forced. I say "may" because I never totally pinned down the problem;
6064 setting up timeout/retry tests is difficult. See also the next item.
6066 79. The ultimate address timeout was not being applied to errors that involved
6067 a combination of host plus message (for example, a timeout on a MAIL
6068 command). When an address resolved to a number of possible hosts, and they
6069 were not all tried for each delivery (e.g. because of hosts_max_try), a
6070 message could remain on the queue longer than the retry timeout.
6072 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6075 81. Fixed an obscure SMTP outgoing bug which required at least the following
6076 conditions: (a) there was another message waiting for the same server;
6077 (b) the server returned 5xx to all RCPT commands in the first message so
6078 that the message was not completed; (c) the server dropped the connection
6079 or gave a negative response to the RSET that Exim sends to abort the
6080 transaction. The observed case was a dropped connection after DATA that had
6081 been sent in pipelining mode. That is, the server had advertised PIPELINING
6082 but was not implementing it correctly. The effect of the bug was incorrect
6083 behaviour, such as trying another host, and this could lead to a crash.
6089 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6090 and daemon.c were passed as pointers to ints; they should have been
6091 pointers to socklen_t variables (which are typically unsigned ints).
6093 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6096 3. Fixed a really odd bug that affected only the testing scheme; patching a
6097 certain fixed string in the binary changed the value of another string that
6098 happened to be identical to the end of the original first string.
6100 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6101 name", it returns that address as the IP address. On some operating
6102 systems (e.g. Solaris), it also passes back the IP address string as the
6103 "host name". However, on others (e.g. Linux), it passes back an empty
6104 string. Exim wasn't checking for this, and was changing the host name to an
6105 empty string, assuming it had been canonicalized.
6107 5. Although rare, it is permitted to have more than one PTR record for a given
6108 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6109 all the names associated with an address, because they do in Solaris.
6110 However, it seems that they do not in Linux for data that comes from the
6111 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6112 I found this out when I moved to a new Linux workstation and tried to run
6113 the Exim test suite.
6115 To get round this problem I have changed the code so that it now does its
6116 own call to the DNS to look up PTR records when searching for a host name.
6117 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6118 addresses that are only in /etc/hosts are still found.
6120 This behaviour is, however, controlled by an option called host_lookup_
6121 order, which defaults to "bydns:byaddr". If people want to use the other
6122 order, or indeed, just use one or the other means of lookup, they can
6123 specify it in this variable.
6125 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6126 some operating systems, this comes back from gethostbyaddr() as an empty
6127 string, and this is what Exim used to test for. However, it seems that in
6128 other systems, "." is yielded. Exim now tests for this case too.
6130 7. The values of check_spool_space and check_log_space are now held internally
6131 as a number of kilobytes instead of an absolute number of bytes. If a
6132 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6133 kilobyte. This means that much larger values can be stored.
6135 8. Exim monitor: an attempt to get the action menu when not actually pointing
6136 at a message produces an empty menu entitled "No message selected". This
6137 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6138 no entries in it ("Shell widget menu has zero width and/or height"). So I
6139 have added a single, blank menu entry in this case.
6141 9. Added ${quote_local_part.
6143 10. MIME decoding is now applied to the contents of Subject: header lines when
6146 11. Now that a reference to $sender_host_address automatically causes a reverse
6147 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6148 host lookup before query-style lookups in lists that might use this
6149 variable. This has therefore been abolished, and the "net-" prefix is no
6150 longer necessary for query-style lookups.
6152 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6153 have been a typo for LFLAGS, so it has been changed.
6155 13. The install script calls Exim with "-C /dev/null" in order to find the
6156 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6157 to be output. However, since Exim outputs its version number before the
6158 error, it didn't break the script. It just looked ugly. I fixed this by
6159 always allowing "-C /dev/null" if the caller is root.
6161 14. Ignore overlarge ACL variable number when reading spool file - insurance
6162 against a later release with more variables having written the file.
6164 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6165 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6166 this, as well as the form without the "IPv6" on the front.
6168 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6169 OS/Makefile-Darwin file.
6171 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6172 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6174 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6175 variable. However, some OS use ints for inodes. I've added cast to long int
6176 to get rid of the compiler warning.
6178 19. I had forgotten to lock out "/../" in configuration file names when
6179 ALT_CONFIG_PREFIX was set.
6181 20. Routers used for verification do not need to specify transports. However,
6182 if such a router generated a host list, and callout was configured, Exim
6183 crashed, because it could not find a port number from the (non-existent)
6184 transport. It now assumes port 25 in this circumstance.
6186 21. Added the -t option to exigrep.
6188 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6189 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6190 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6191 them is set, LOOKUP_LSEARCH is forced.
6193 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6195 24. Added sender and host information to the "rejected by local_scan()" log
6196 line; previously there was no indication of these.
6198 25. Added .include_if_exists.
6200 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6201 Linux. It turns out that not all file systems support this. Apparently some
6202 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6203 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6204 not supported on the file descriptor, is now ignored when Exim is trying to
6205 sync a directory. This applies only to Linux.
6207 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6209 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6210 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6213 29. Added support for maildirsize files from supplied patch (modified a bit).
6215 30. The use of :fail: followed by an empty string could lead Exim to respond to
6216 sender verification failures with (e.g.):
6218 550 Verification failed for <xxx>
6219 550 Sender verify failed
6221 where the first response line was missing the '-' that indicates it is not
6222 the final line of the response.
6224 31. The loop for finding the name of the user that called Exim had a hardwired
6225 limit of 10; it now uses the value of finduser_retries, which is used for
6226 all other user lookups.
6228 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6231 33. Exim was neglecting to zero errno before one call of strtol() when
6232 expanding a string and expecting an integer value. On some systems this
6233 resulted in spurious "integer overflow" errors. Also, it was casting the
6234 result into an int without checking.
6236 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6237 did not work. The code looks as if it has *never* worked, though it appears
6238 to have been documented since at least release 1.62. I have made it work.
6240 35. The "timeout_DNS" error in retry rules, also documented since at least
6241 1.62, also never worked. As it isn't clear exactly what this means, and
6242 clearly it isn't a major issue, I have abolished the feature by treating it
6243 as "timeout", and writing a warning to the main and panic logs.
6245 36. The display of retry rules for -brt wasn't always showing the error code
6248 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6249 timeout_connect_A, timeout_connect_MX.
6251 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6252 to the empty sender.
6254 39. The daemon was not analysing the content of -oX till after it had closed
6255 stderr and disconnected from the controlling terminal. This meant that any
6256 syntax errors were only noted on the panic log, and the return code from
6257 the command was 0. By re-arranging the code a little, I've made the
6258 decoding happen first, so such errors now appear on stderr, and the return
6259 code is 1. However, the actual setting up of the sockets still happens in
6260 the disconnected process, so errors there are still only recorded on the
6263 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6264 connections (as happens on some IP stacks) was logged at start up time as
6265 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6266 it from "IPv6 and IPv4", which means that two separate sockets are being
6269 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6270 says whether AF_INET or AF_INET6 was passed as an argument.
6272 42. Exiwhat output was messed up when time zones were included in log
6275 43. Exiwhat now gives more information about the daemon's listening ports,
6276 and whether -tls-on-connect was used.
6278 44. The "port" option of the smtp transport is now expanded.
6280 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6281 silently ignored. Now an error message is written to the main and panic
6284 46. There's a new ACL modifier called "logwrite" which writes to a log file
6285 as soon as it is encountered.
6287 47. Added $local_user_uid and $local_user_gid at routing time.
6289 48. Exim crashed when trying to verify a sender address that was being
6292 49. Exim was recognizing only a space character after ".include". It now also
6293 recognizes a tab character.
6295 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6296 extracting the relevant information from the specification. The man page no
6297 longer contains scrambled data for the -d option, and I've added a section
6298 at the front about calling Exim under different names.
6300 51. Added "extra_headers" argument to the "mail" command in filter files.
6302 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6305 53. Installed eximstats 1.29.
6307 54. Added transport_filter_timeout as a generic transport option.
6309 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6310 Cc: header lines. This was required by RFC 822, but it not required by RFC
6313 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6314 incoming messages that did not have them. Now it does so only if the
6315 message originates locally, that is, if there is no associated remote host
6316 address. When Resent- header lines are present, this applies to the Resent-
6317 lines rather than the non-Resent- lines.
6319 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6320 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6322 58. Messages for configuration errors now include the name of the main
6323 configuration file - useful now that there may be more than one file in a
6324 list (.included file names were always shown).
6326 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6327 for those rare installations that do not start the daemon as root or run it
6328 setuid root. I've cut out the call to initgroups() if the daemon is not
6331 60. The Exim user and group can now be bound into the binary as text strings
6332 that are looked up at the start of Exim's processing.
6334 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6336 62. Added $mailstore_basename variable.
6338 63. Installed patch to sieve.c from Michael Haardt.
6340 64. When Exim failed to open the panic log after failing to open the main log,
6341 the original message it was trying to log was written to stderr and debug
6342 output, but if they were not available (the usual case in production), it
6343 was lost. Now it is written to syslog before the two lines that record the
6344 failures to open the logs.
6346 65. Users' Exim filters run in subprocesses under the user's uid. It is
6347 possible for a "deliver" command or an alias in a "personal" command to
6348 provoke an address rewrite. If logging of address rewriting is configured,
6349 this fails because the process is not running as root or exim. There may be
6350 a better way of dealing with this, but for the moment (because 4.30 needs
6351 to be released), I have disabled address rewrite logging when running a
6352 filter in a non-root, non-exim process.
6358 1. The buildconfig auxiliary program wasn't quoting the value set for
6359 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6360 not defined. This bug was masked in 4.22 by the effect that was fixed in
6363 2. Some messages that were rejected after a message id was allocated were
6364 shown as "incomplete" by exigrep. It no longer does this for messages that
6365 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6367 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6368 have allow_domain_literals set, the ID did not get logged in the <= line.
6369 Domain literals are now always recognized in Message-ID: header lines.
6371 4. The first argument for a ${extract expansion item is the key name or field
6372 number. Leading and trailing spaces in this item were not being ignored,
6373 causing some misleading effects.
6375 5. When deliver_drop_privilege was set, single queue runner processes started
6376 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6377 same command in the process it spins off) were not dropping privilege.
6379 6. When the daemon running as "exim" started a queue runner, it always
6380 re-executed Exim in the spun-off process. This is a waste of effort when
6381 deliver_drop_privilege is set. The new process now just calls the
6382 queue-runner function directly.
6388 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6391 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6392 ignored. Though the use of -oP was forcing the writing of a pid file, it
6393 was always written to the default place.
6395 3. If the message "no IP address found for host xxxx" is generated during
6396 incoming verification, it is now followed by identification of the incoming
6397 connection (so you can more easily find what provoked it).
6399 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6401 5. Added some features to "harden" Exim a bit more against certain attacks:
6403 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6404 be put in Local/Makefile. This is like the never_users runtime option,
6405 but it cannot be overridden. The default setting is "root".
6407 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6408 prefix string with which any file named in a -C command line option
6411 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6412 is retained for -C and -D only if the caller of Exim is root. Without
6413 it, the exim user may also use -C and -D and retain privilege.
6415 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6416 command line option is disabled.
6418 6. Macro names set by the -D option must start with an upper case letter, just
6419 like macro names defined in the configuration file.
6421 7. Added "dereference=" facility to LDAP.
6423 8. Two instances of the typo "uknown" in the source files are fixed.
6425 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6426 the Configure-Makefile script screwed up while processing it.
6428 10. Incorporated PCRE 4.4.
6430 11. The SMTP synchronization check was not operating right at the start of an
6431 SMTP session. For example, it could not catch a HELO sent before the client
6432 waited for the greeting. There is now a check for outstanding input at the
6433 point when the greeting is written. Because of the duplex, asynchronous
6434 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6435 way, but not yet received, when the check is performed.
6437 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6438 on TCP/IP sockets, because this apparently causes some broken clients to
6441 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6442 unchanged) from the Cygwin maintainer.
6444 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6445 was supported instead of testing for mbx. Effectively a typo.
6447 15. The spa authenticator server code was not checking that the input it
6448 received was valid base64.
6450 16. The debug output line for the "set" modifier in ACLs was not showing the
6451 name of the variable that was being set.
6453 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6455 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6456 lookup on its own. Something else was needed to trigger the lookup. For
6457 example, a match in host_lookup or the need for a host name in a host list.
6458 Now, if $sender_host_name is referenced and the host name has not yet been
6459 looked up, a lookup is performed. If the lookup fails, the variable remains
6460 empty, and $host_lookup_failed is set to "1".
6462 19. Added "eqi" as a case-independent comparison operator.
6464 20. The saslauthd authentication condition could segfault if neither service
6465 nor realm was specified.
6467 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6468 error message that was logged was misleading, and incoming SMTP
6469 connections were dropped. The message is now more accurate, and temporary
6470 errors are given to SMTP connections.
6472 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6473 (except RSET) with a 421 error, until QUIT is received. However, it was
6474 failing to send a response to QUIT.
6476 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6477 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6478 ACL is now run after the helo_try_verify_hosts code.
6480 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6481 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6482 case-independent, so other case variants are also recognized). Apparently
6483 some systems use these upper case variants.
6485 25. If more than two messages were waiting for the same host, and a transport
6486 filter was specified for the transport, Exim sent two messages over the
6487 same TCP/IP connection, and then failed with "socket operation on non-
6488 socket" when it tried to send the third.
6490 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6492 27. The extern definition of crypt16() in expand.c was not being excluded when
6493 the OS had its own crypt16() function.
6495 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6496 as a preferred synonym for return_size_limit, both as an option and as an
6499 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6501 30. Changed the default configuration ACL to relax the local part checking rule
6502 for addresses that are not in any local domains. For these addresses,
6503 slashes and pipe symbols are allowed within local parts, but the sequence
6504 /../ is explicitly forbidden.
6506 31. SPA server authentication was not clearing the challenge buffer before
6509 32. log_message in a "warn" ACL statement was writing to the reject log as
6510 well as to the main log, which contradicts the documentation and doesn't
6511 seem right (because no rejection is happening). So I have stopped it.
6513 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6514 However, I am unable to do any testing of this.
6516 34. Fixed an infelicity in the appendfile transport. When checking directories
6517 for a mailbox, to see if any needed to be created, it was accidentally
6518 using path names with one or more superfluous leading slashes; tracing
6519 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6521 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6522 contents are added to the log line that is written for every discarded
6523 recipient. (Previously a log_message setting was ignored.)
6525 36. The ${quote: operator now quotes the string if it is empty.
6527 37. The install script runs exim in order to find its version number. If for
6528 some reason other than non-existence or emptiness, which it checks, it
6529 could not run './exim', it was installing it with an empty version number,
6530 i.e. as "exim-". This error state is now caught, and the installation is
6533 38. An argument was missing from the function that creates an error message
6534 when Exim fails to connect to the socket for saslauthd authentication.
6535 This could cause Exim to crash, or give a corrupted message.
6537 39. Added isip, isip4, and isip6 to ${if conditions.
6539 40. The ACL variables $acl_xx are now saved with the message, and can be
6540 accessed later in routers, transports, and filters.
6542 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6543 strings in the file are not string-expanded.
6545 42. If a MAIL command specified a SIZE value that was too large to fit into an
6546 int variable, the check against message_size_limit failed. Such values are
6547 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6548 day this will have to be increased, but I don't think I want to be around
6549 when emails are that large.
6556 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6557 iconv() is not standard in FreeBSD.
6559 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6560 IPv6 enabled. The observed symptom was a segmentation fault on return from
6561 the function os_common_find_running_interfaces() in src/os.c.
6563 3. In the check_special_case() function in daemon.c I had used "errno" as an
6564 argument name, which causes warnings on some systems. This was basically a
6565 typo, since it was named "eno" in the comments!
6567 4. The code that waits for the clock to tick (at a resolution of some fraction
6568 of a second) so as to ensure message-id uniqueness was always waiting for
6569 at least one whole tick, when it could have waited for less. [This is
6570 almost certainly not relevant at current processor speeds, where it is
6571 unlikely to ever wait at all. But we try to future-proof.]
6573 5. The function that sleeps for a time interval that includes fractions of a
6574 second contained a race. It did not block SIGALRM between setting the
6575 timer, and suspending (a couple of lines later). If the interval was short
6576 and the sigsuspend() was delayed until after it had expired, the suspension
6577 never ended. On busy systems this could lead to processes getting stuck for
6580 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6581 process, before it forks any delivery processes. The open lookup caching
6582 mechanism meant that the open file or database connection was passed into
6583 the delivery process. The problem was that delivery processes always tidy
6584 up cached lookup data. This could cause a problem for the next delivery
6585 process started by the queue runner, because the external queue runner
6586 process does not know about the closure. So the next delivery process
6587 still has data in the lookup cache. In the case of a file lookup, there was
6588 no problem because closing a file descriptor in a subprocess doesn't affect
6589 the parent. However, if the lookup was caching a connection to a database,
6590 the connection was closed, and the second delivery process was likely to
6591 see errors such as "PGSQL: query failed: server closed the connection
6592 unexpectedly". The problem has been fixed by closing all cached lookups
6593 in a queue runner before running a delivery process.
6595 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6596 seem to have the "const" qualifier which it has on other OS. I've
6599 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6600 *of the same type* (client or server) with the same public name that an
6601 error should be diagnosed.
6603 9. When Exim looked up a host name for an IP address, but failed to find the
6604 original IP address when looking up the host name (a safety check), it
6605 output the message "<ip address> does not match any IP for NULL", which was
6606 confusing, to say the least. The bug was that the host name should have
6607 appeared instead of "NULL".
6609 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6610 user that is built into the binary, and the -C or -D options is used, root
6611 privilege is dropped before the configuration file is read. In addition,
6612 logging is switched to stderr instead of the normal log files. If the
6613 configuration then re-defines the Exim user, the unprivileged environment
6614 is probably not what is expected, so Exim logs a panic warning message (but
6617 However, if deliver_drop_privilege is set, the unprivileged state may well
6618 be exactly what is intended, so the warning has been cut out in that case,
6619 and Exim is allowed to try to write to its normal log files.
6625 1. smtp_return_error_details was not giving details for temporary sender
6626 or receiver verification errors.
6628 2. Diagnose a configuration error if two authenticators have the same public
6631 3. Exim used not to create the message log file for a message until the first
6632 delivery attempt. This could be confusing when incoming messages were held
6633 for policy or load reasons. The message log file is now created at the time
6634 the message is received, and an initial "Received" line is written to it.
6636 4. The automatically generated man page for command line options had a minor
6637 bug that caused no ill effects; however, a more serious problem was that
6638 the procedure for building the man page automatically didn't always
6639 operate. Consequently, release 4.20 contains an out-of-date version. This
6640 shouldn't happen again.
6642 5. When building Exim with embedded Perl support, the script that builds the
6643 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6644 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6646 6. The freeze_tell option was not being used for messages that were frozen on
6647 arrival, either by an ACL or by local_scan().
6649 7. Added the smtp_incomplete_transaction log selector.
6651 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6652 was accepting AUTH without a new EHLO.
6654 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6655 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6658 10. Logging of TCP/IP connections (when configured) now happens in the main
6659 daemon process instead of the child process, so that the TCP/IP connection
6660 count is more accurate (but it can never be perfect).
6662 11. The use of "drop" in a nested ACL was not being handled correctly in the
6663 outer ACL. Now, if condition failure induced by the nested "drop" causes
6664 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6665 or "require"), the connection is dropped.
6667 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6668 that yield "discard" can now be used with an "accept" or a "discard" verb,
6669 but an error is generated for any others (because I can't see a useful way
6670 to define what should happen).
6672 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6673 are now processed in the same way as lines in the Exim configuration file.
6674 In particular, continuation lines are supported.
6676 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6678 15. Added -ti meaning -t -i.
6680 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6681 domains, and warn by logging if others are found.
6683 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6684 returned from GIFCONF when Exim is trying to find the list of interfaces on
6685 a host. The code in os.c has been modified to copy each ifreq to an aligned
6686 structure in all cases.
6688 Also, in some cases, the returned ifreq's were being copied to a 'struct
6689 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6690 means the last couple of bytes of an IPv6 address could be chopped if the
6691 ifreq contained only a normal sockaddr (14 bytes storage).
6693 18. Named domain lists were not supported in the hosts_treat_as_local option.
6694 An entry such as +xxxx was not recognized, and was treated as a literal
6697 19. Ensure that header lines added by a DATA ACL are included in the reject log
6698 if the ACL subsequently rejects the message.
6700 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6701 MD5 (which is deprecated).
6703 21. When testing a filter file using -bf, Exim was writing a message when it
6704 took the sender from a "From " line in the message, but it was not doing so
6705 when it took $return_path from a Return-Path: header line. It now does.
6707 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6708 with a valid header line field name (a series of printing characters
6709 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6711 23. Changed "disc" in the source to "disk" to conform to the documentation and
6712 the book and for uniformity.
6714 24. Ignore Sendmail's -Ooption=value command line item.
6716 25. When execve() failed while trying to run a command in a pipe transport,
6717 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6718 could be confused with a return value of 69 from the command itself. This
6719 has been changed to 127, the value the shell returns if it is asked to run
6720 a non-existent command. The wording for the related log line suggests a
6721 non-existent command as the problem.
6723 26. If received_header_text expands to an empty string, do not add a Received:
6724 header line to the message. (Well, it adds a token one on the spool, but
6725 marks it "old" so that it doesn't get used or transmitted.)
6727 27. Installed eximstats 1.28 (addition of -nt option).
6729 28. There was no check for failure on the call to getsockname() in the daemon
6730 code. This can fail if there is a shortage of resources on the system, with
6731 ENOMEM, for example. A temporary error is now given on failure.
6733 29. Contrary to the C standard, it seems that in some environments, the
6734 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6735 program. Exim now does this explicitly; it affects the formatting of
6736 timestamps using strftime().
6738 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6739 complaints. I've now initialized all the variables, to avoid this.
6741 32. Header lines added by a system filter were not being "seen" during
6742 transport-time rewrites.
6744 33. The info_callback() function passed to OpenSSL is set up with type void
6745 (*)(SSL *, int, int), as described somewhere. However, when calling the
6746 function (actually a macro) that sets it up, the type void(*)() is
6747 expected. I've put in a cast to prevent warnings from picky compilers.
6749 34. If a DNS black list lookup found a CNAME record, but there were no A
6750 records associated with the domain it pointed at, Exim crashed.
6752 35. If a DNS black list lookup returned more than one A record, Exim ignored
6753 all but the first. It now scans all returned addresses if a particular IP
6754 value is being sought. In this situation, the contents of the
6755 $dnslist_value variable are a list of all the addresses, separated by a
6758 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6759 to accept a host name and all its aliases if the forward lookup for any of
6760 them yielded the IP address of the incoming connection. Now it accepts only
6761 those names whose forward lookup yields the correct IP address. Any other
6762 names are discarded. This closes a loophole whereby a rogue DNS
6763 administrator could create reverse DNS records to break through a
6764 wildcarded host restriction in an ACL.
6766 37. If a user filter or a system filter that ran in a subprocess used any of
6767 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6768 the wrong values were passed to the pipe command ($thisaddress had the
6769 value of $0, $0 had the value of $1, etc). This bug was introduced by
6770 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6772 38. Improved the line breaking for long SMTP error messages from ACLs.
6773 Previously, if there was no break point between 40 and 75 characters, Exim
6774 left the rest of the message alone. Two changes have been made: (a) I've
6775 reduced the minimum length to 35 characters; (b) if it can't find a break
6776 point between 35 and 75 characters, it looks ahead and uses the first one
6777 that it finds. This may give the occasional overlong line, but at least the
6778 remaining text gets split now.
6780 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6781 file descriptors might be low, and that setting 1000 would always raise it.
6782 It turns out that in some environments, the limit is already over 1000 and
6783 that lowering it causes trouble. So now Exim takes care not to decrease it.
6785 40. When delivering a message, the value of $return_path is set to $sender_
6786 address at the start of routing (routers may change the value). By an
6787 oversight, this default was not being set up when an address was tested by
6788 -bt or -bv, which affected the outcome if any router or filter referred to
6791 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6792 log only when "message" or "log_message" are set. However, if one of the
6793 conditions was an address verification, or a call to a nested ACL, the
6794 messages generated by the underlying test were being passed through. This
6795 no longer happens. The underlying message is available in $acl_verify_
6796 message for both "message" and "log_message" expansions, so it can be
6797 passed through if needed.
6799 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6800 new expansion $bh_ to give the encoded byte string without charset
6801 translation. Translation happens only if iconv() is available; HAVE_ICONV
6802 indicates this at build time. HEADERS_CHARSET gives the charset to
6803 translate to; headers_charset can change it in the configuration, and
6804 "headers charset" can change it in an individual filter file.
6806 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6807 that creates RFC 2047 encoded "words" labels them as that charset instead
6808 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6809 expansion operator; (ii) when Exim creates a From: line for a local
6810 message; (iii) when a header line is rewritten to include a "phrase" part.
6812 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6813 buggy, causing it to skip the first lines of messages whose message ID
6814 ended in 'D'. This would not have bitten before Exim release 4.14, because
6815 message IDs were unlikely to end in 'D' before then. The effect was to have
6816 incorrect size information for certain domains.
6818 45. #include "config.h" was missing at the start of the crypt16.c module. This
6819 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6822 46. If there was a timeout during a "random" callout check, Exim treated it as
6823 a failure of the random address, and carried on sending RSET and the real
6824 address. If the delay was just some slowness somewhere, the response to the
6825 original RCPT would be taken as a response to RSET and so on, causing
6826 mayhem of various kinds.
6828 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6829 when I implemented it. It didn't allow for the fact that some option values
6830 may legitimately be negative (e.g. size_addition), and it didn't even do
6831 the right test for positive values.
6833 48. Domain names in DNS records are case-independent. Exim always looks them up
6834 in lower case. Some resolvers return domain names in exactly the case they
6835 appear in the zone file, that is, they may contain uppercase letters. Not
6836 all resolvers do this - some return always lower case. Exim was treating a
6837 change of case by a resolver as a change of domain, similar to a widening
6838 of a domain abbreviation. This triggered its re-routing code and so it was
6839 trying to route what was effectively the same domain again. This normally
6840 caused routing to fail (because the router wouldn't handle the domain
6841 twice). Now Exim checks for this case specially, and just changes the
6842 casing of the domain that it ultimately uses when it transmits the message
6845 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6848 50. If a filter generated a file delivery with a non-absolute name (possible if
6849 no home directory exists for the router), the forbid_file option was not
6852 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6853 the existing equality matching.
6855 52. Exim was using ints instead of ino_t variables in some places where it was
6856 dealing with inode numbers.
6858 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6859 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6860 called TMPDIR, and if it finds it is different, it changes its value.
6862 54. The smtp_printf() function is now made available to local_scan() so
6863 additional output lines can be written before returning. There is also an
6864 smtp_fflush() function to enable the detection of a dropped connection.
6865 The variables smtp_input and smtp_batched_input are exported to
6868 55. Changed the default runtime configuration: the message "Unknown user"
6869 has been removed from the ACL, and instead placed on the localuser router,
6870 using the cannot_route_message feature. This means that any verification
6871 failures that generate their own messages won't get overridden. Similarly,
6872 the "Unrouteable address" message that was in the ACL for unverifiable
6873 relay addresses has also been removed.
6875 56. Added hosts_avoid_esmtp to the smtp transport.
6877 57. The exicyclog script was not checking for the esoteric option
6878 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6879 will work only if exicyclog is run under the appropriate euid.
6881 58. Following a discussion on the list, the rules by which Exim recognises line
6882 endings on incoming messages have been changed. The -dropcr and drop_cr
6883 options are now no-ops, retained only for backwards compatibility. The
6884 following line terminators are recognized: LF CRLF CR. However, special
6885 processing applies to CR:
6887 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6888 nor a local message in the state where . is a terminator.
6890 (ii) If a bare CR is encountered in a header line, an extra space is added
6891 after the line terminator so as not to end the header. The reasoning
6892 behind this is that bare CRs in header lines are most likely either
6893 to be mistakes, or people trying to play silly games.
6895 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6896 was being incorrectly given as 18 bytes larger than it should have been.
6897 This is a VOB (very old bug).
6899 60. This may never have affected anything current, but just in case it has:
6900 When the local host is found other than at the start of a list of hosts,
6901 the local host, those with the same MX, and any that follow, are discarded.
6902 When the list in question was part of a longer list of hosts, the following
6903 hosts (not currently being processed) were also being discarded. This no
6904 longer happens. I'm not sure if this situation could ever has previously
6907 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6910 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6911 header. This is something that is recommended in a new Internet Draft, and
6912 is something that is documented as being done by Sendmail. There are two
6913 possible values. For messages generated by the autoreply transport, Exim
6916 Auto-Submitted: auto-replied
6918 whereas for all other generated messages (e.g. bounces) it adds
6920 Auto-Submitted: auto-generated
6922 63. The "personal" condition in filters now includes a test for the
6923 Auto-Submitted: header. If it contains the string "auto-" the message it
6924 not considered personal.
6926 64. Added rcpt_include_affixes as a generic transport option.
6928 65. Added queue_only_override (default true).
6930 66. Added the syslog_duplication option.
6932 67. If what should have been the first header line of a message consisted of
6933 a space followed by a colon, Exim was mis-interpreting it as a header line.
6934 It isn't of course - it is syntactically invalid and should therefore be
6935 treated as the start of the message body. The misbehaviour could have
6936 caused a number of strange effects, including loss of data in subsequent
6937 header lines, and spool format errors.
6939 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6940 client host had authenticated. This control can now be exercised by an ACL
6941 for more flexibility.
6943 69. By default, callouts do not happen when testing with -bh. There is now a
6944 variant, -bhc, which does actually run the callout code, including
6945 consulting and updating the callout cache.
6947 70. Added support for saslauthd authentication, courtesy of Alexander
6950 71. If statvfs() failed on the spool or log directories while checking their
6951 size for availability, Exim confusingly gave the error "space shortage".
6952 Furthermore, in debugging mode it crashed with a floating point exception.
6953 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6954 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6955 serious problem, Exim now writes to the main and panic logs when this
6956 happens, with details of the failure. It then refuses to accept the
6957 incoming message, giving the message "spool directory problem" or "log
6958 directory problem" with a 421 code for SMTP messages.
6960 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6961 0, 1, and 2 exist, because some OS complain for execs without them (see
6962 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6963 descriptors. However, the code omitted to check that the open succeeded,
6964 causing mysterious errors if for some reason the permissions on /dev/null
6965 got screwed. Now Exim writes a message to the main and panic logs, and
6966 bombs out if it can't open /dev/null.
6968 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6969 interact so that it is all more flexible. It is supposed to remain
6970 backwards compatible. Also added extra_local_interfaces.
6972 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6973 to bomb out with an assertion failure - to the client this appears as a
6974 connection drop. This problem occurs in the part of the code that was taken
6975 from the Samba project. Fortunately, the assertion is in a very simple
6976 function, so I have fixed this by reproducing the function inline in the
6977 one place where it is called, and arranging for authentication to fail
6978 instead of killing the process with assert().
6980 75. The SPA client code was not working when the server requested OEM rather
6981 than Unicode encoding.
6983 76. Added code to make require_files with a specific uid setting more usable in
6984 the case where statting the file as root fails - usually a non-root-mounted
6985 NFS file system. When this happens and the failure is EACCES, Exim now
6986 forks a subprocess and does the per-uid checking as the relevant uid.
6988 77. Added process_log_path.
6990 78. If log_file_path was not explicitly set, a setting of check_log_space or
6991 check_log_inodes was ignored.
6993 79. If a space check for the spool or log partitions fails, the incident is now
6994 logged. Of course, in the latter case the data may get lost...
6996 80. Added the %p formatting code to string_format() so that it can be used to
6997 print addresses in debug_print(). Adjusted all the address printing in the
6998 debugging in store.c to use %p rather than %d.
7000 81. There was a concern that a line of code in smtp_in.c could overflow a
7001 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7002 initially expressed, the concern was not well-founded, because trailing
7003 spaces are removed early. However, if the trailing spaces were followed by
7004 a NULL, they did not get removed, so the overflow was possible. Two fixes
7007 (a) I re-wrote the offending code in a cleaner fashion.
7008 (b) If an incoming SMTP command contains a NULL character, it is rejected
7011 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7012 runs initgroups(), so that if the Exim user is in any additional groups,
7013 they will be used during message reception.
7019 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob