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
19 JH/01 Rework error string handling in TLS interface so that the caller in
20 more cases is responsible for logging. This permits library-sourced
21 string to be attached to addresses during delivery, and collapses
22 pairs of long lines into single ones.
24 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
25 during configuration. Wildcards are allowed and expanded.
27 JH/02 Rework error string handling in DKIM to pass more info back to callers.
28 This permits better logging.
30 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
31 do not close it down and have the child transport start it up again on
32 the passed-on TCP connection. Instead, proxy the child (and any
33 subsequent ones) for TLS via a unix-domain socket channel. Logging is
34 affected: the continued delivery log lines do not have any DNSSEC, TLS
35 Certificate or OCSP information. TLS cipher information is still logged.
37 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
38 identical IP addresses on different listening ports. Will also affect
41 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
42 add noisy ifdef guards to special-case this sillyness.
43 Patch from Bernd Kuhls.
45 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
46 than 255 are no longer allowed.
48 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
49 Disable the session-cache too, which might reduce our load. Since we
50 currrectly use a new context for every connection, both as server and
51 client, there is no benefit for these.
52 GnuTLS appears to not support tickets server-side by default (we don't
53 call gnutls_session_ticket_enable_server()) but client side is enabled
54 by default on recent versions (3.1.3 +) unless the PFS priority string
57 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
58 <https://reproducible-builds.org/specs/source-date-epoch/>.
60 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
61 the check for any unsuccessful recipients did not notice the limit, and
62 erroneously found still-pending ones.
64 JH/08 Pipeline CHUNKING command and data together, on kernels that support
65 MSG_MORE. Only in-clear (not on TLS connections).
67 JH/09 Avoid using a temporary file during transport using dkim. Unless a
68 transport-filter is involved we can buffer the headers in memory for
69 creating the signature, and read the spool data file once for the
70 signature and again for transmission.
72 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
73 4.77 as the kernel support then wasn't solid, having issues in 64bit
74 mode. Now, it's been long enough. Add support for FreeBSD also.
76 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
77 case where the routing stage had gathered several addresses to send to
78 a host before calling the transport for the first, we previously failed
79 to close down TLS in the old transport process before passing the TCP
80 connection to the new process. The new one sent a STARTTLS command
81 which naturally failed, giving a failed delivery and bloating the retry
82 database. Investigation and fix prototype from Wolfgang Breyha.
84 JH/12 Fix check on SMTP command input synchronisation. Previously there were
85 false-negatives in the check that the sender had not preempted a response
86 or prompt from Exim (running as a server), due to that code's lack of
87 awareness of the SMTP input buffering.
89 PP/04 Add commandline_checks_require_admin option.
90 Exim drops privileges sanely, various checks such as -be aren't a
91 security problem, as long as you trust local users with access to their
92 own account. When invoked by services which pass untrusted data to
93 Exim, this might be an issue. Set this option in main configuration
94 AND make fixes to the calling application, such as using `--` to stop
97 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
98 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
99 responses to those, into a single TLS record each way (this usually means
100 a single packet). As a side issue, smtp_enforce_sync now works on TLS
103 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
104 affects you only if you're dancing at the edge of the param size limits.
105 If you are, and this message makes sense to you, then: raise the
106 configured limit or use OpenSSL 1.1. Nothing we can do for older
109 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
110 cmdline element to get the documented default one. Previously it was
113 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
114 are unsuable for non-message-specific reasons - eg. connection timeout,
115 banner-time rejection.
117 JH/16 Fix logging of delivery remote port, when specified by router, under
120 PP/06 Repair manualroute's ability to take options in any order, even if one
121 is the name of a transport.
124 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
126 JH/17 Change the list-building routines interface to use the expanding-string
127 triplet model, for better allocation and copying behaviour.
129 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
130 Previously it was constructed the first time a possibly-matching string
131 was met in the configuration file input during startup; now it is done
134 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
135 compatible one, to avoid the (poorly documented) possibility of a config
136 file in the working directory redirecting the DB files, possibly correpting
137 some existing file. CVE-2017-10140 assigned for BDB.
139 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
140 cache-hot. Previously, although the result was properly cached, the
141 initial verify call returned a defer.
143 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
144 the main verify for receipient in uncached-mode.
146 JH/22 Retire historical build files to an "unsupported" subdir. These are
147 defined as "ones for which we have no current evidence of testing".
149 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
150 if present. Previously it was ignored.
152 JH/24 Start using specified-initialisers in C structure init coding. This is
153 a C99 feature (it's 2017, so now considered safe).
155 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
156 if was a fixed-sized field and bitmask ops via macros; it is now more
159 PP/07 GitHub PR 56: Apply MariaDB build fix.
160 Patch provided by Jaroslav Škarvada.
162 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
163 during Coverity cleanups [4.87 JH/47]
164 Diagnosis and fix provided by Michael Fischer v. Mollard.
166 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
167 the right size to place the terminating semicolon on its own folded
168 line, the header hash was calculated to an incorrect value thanks to
169 the (relaxed) space the fold became.
171 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
172 and confused the parent.
174 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
175 which could crash as a result. This could lead to undeliverable messages.
177 JH/28 Logging: "next input sent too soon" now shows where input was truncated
180 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
181 matters on fast-turnover and PID-randomising systems, which were getting
182 out-of-order delivery.
184 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
185 a possibly-overlapping copy. The symptom was that "Remote host closed
186 connection in response to HELO" was logged instead of the actual 4xx
189 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
190 Previously only that bufferd was discarded, resulting in SYMTP command
193 JH/32 DKIM: when a message has multiple signatures matching an identity given
194 in dkim_verify_signers, run the dkim acl once for each. Previously only
195 one run was done. Bug 2189.
197 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
198 "panic the current process" to "deliberately defer". The panic log is
199 still written with the problem list name; the mail and reject logs now
200 get a temp-reject line for the message that was being handled, saying
201 something like "domains check lookup or other defer". The SMTP 451
202 message is still "Temporary local problem".
204 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
205 A crafted sequence of BDAT commands could result in in-use memory beeing
206 freed. CVE-2017-16943.
208 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
209 from SMTP input. Previously it was always done; now only done for DATA
210 and not BDAT commands. CVE-2017-16944.
212 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
213 to the message (such as an overlong header line). Previously this was
214 not done and we did not exit BDAT mode. Followon from the previous item
215 though a different problem.
221 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
222 than -2003 did; needs libidn2 in addition to libidn.
224 JH/02 The path option on a pipe transport is now expanded before use.
226 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
227 Patch provided by "Björn", documentation fix added too.
229 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
230 missing a wire-to-host endian conversion.
232 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
233 close after a BDAT command line could be taken as a following command,
234 giving a synch failure. Fix by only checking for synch immediately
235 before acknowledging the chunk.
237 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
238 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
239 macro. Patches provided by Josh Soref.
241 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
242 Previously we did not; the RFC seems ambiguous and VRFY is not listed
243 by IANA as a service extension. However, John Klensin suggests that we
246 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
247 the dkim code may be unix-mode line endings rather than smtp wire-format
248 CRLF, so prepend a CR to any bare LF.
250 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
251 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
253 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
254 the first were themselves being wrongly included in the feed into dkim
255 processing; with most chunk sizes in use this resulted in an incorrect
256 body hash calculated value.
258 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
259 DKIM signature block, for verification. Although advised against by
260 standards it is specifically not ruled illegal.
262 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
264 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
265 missing a body hash (the bh= tag).
267 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
268 It seems that HAProxy sends the Proxy Protocol information in clear and
269 only then does a TLS startup, so do the same.
271 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
272 TCP connections (such as for Spamd) unless the daemon successfully set
273 Fast Open mode on its listening sockets. This fixes breakage seen on
274 too-old kernels or those not configured for Fast Open, at the cost of
275 requiring both directions being enabled for TFO, and TFO never being used
276 by non-daemon-related Exim processes.
278 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
279 endings, at least on the first header line. Try to canonify any that get
280 past that check, despite the cost.
282 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
283 now limited to an arbitrary five deep, while parsing addresses with the
284 strip_excess_angle_brackets option enabled.
286 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
287 instead leave the unprompted TLS handshake in socket buffer for the
288 TLS library to consume.
290 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
292 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
294 JH/16 Drop variables when they go out of scope. Memory management drops a whole
295 region in one operation, for speed, and this leaves assigned pointers
296 dangling. Add checks run only under the testsuite which checks all
297 variables at a store-reset and panics on a dangling pointer; add code
298 explicitly nulling out all the variables discovered. Fixes one known
299 bug: a transport crash, where a dangling pointer for $sending_ip_address
300 originally assigned in a verify callout, is re-used.
302 PP/06 Drop '.' from @INC in various Perl scripts.
304 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
306 PP/08 Reduce a number of compilation warnings under clang; building with
307 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
308 should be warning-free.
310 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
312 HS/01 Fix portability problems introduced by PP/08 for platforms where
313 realloc(NULL) is not equivalent to malloc() [SunOS et al].
315 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
316 chunk. This allows us to accept broken chunked messages. We need a more
317 general solution here.
319 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
320 already-broken messages in the queue.
322 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
324 JH/19 Fix reference counting bug in routing-generated-address tracking.
330 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
331 supports it and a size is available (ie. the sending peer gave us one).
333 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
334 years of being deprecated). The replacements are the ACLs
335 acl_smtp_mime and acl_not_smtp_mime.
337 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
338 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
339 or even an in-clear connection were permitted. Now, if the host lookup
340 was dnssec and dane was requested then the host is only used if the
341 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
342 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
343 if one fails this test.
344 This means that a poorly-configured remote DNS will make it incommunicado;
345 but it protects against a DNS-interception attack on it.
347 JH/04 Bug 1810: make continued-use of an open smtp transport connection
348 non-noisy when a race steals the message being considered.
350 JH/05 If main configuration option tls_certificate is unset, generate a
351 self-signed certificate for inbound TLS connections.
353 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
354 in rewrites and routers.
356 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
357 and logged a warning sing 4.83; now they are a configuration file error.
359 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
360 (lacking @domain). Apply the same qualification processing as RCPT.
362 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
364 JH/10 Support ${sha256:} applied to a string (as well as the previous
367 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
368 a cutthrough deliver is pending, as we always want to make a connection.
369 This also avoids re-routing the message when later placing the cutthrough
370 connection after a verify cache hit.
371 Do not update it with the verify result either.
373 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
374 when routing results in more than one destination address.
376 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
377 signing (which inhibits the cutthrough capability). Previously only
378 the presence of an option was tested; now an expansion evaluating as
379 empty is permissible (obviously it should depend only on data available
380 when the cutthrough connection is made).
382 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
383 the relevant preceding SMTP command did not note the pipelining mode.
385 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
386 Previously they were not counted.
388 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
389 as one having no matching records. Previously we deferred the message
390 that needed the lookup.
392 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
393 distinguished as "(=".
395 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
396 for missing MX records. Previously it only worked for missing A records.
398 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
400 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
401 after the data-go-ahead and data-ack. Patch from Jason Betts.
403 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
404 even for a "none" policy. Patch from Tony Meyer.
406 JH/22 Fix continued use of a connection for further deliveries. If a port was
407 specified by a router, it must also match for the delivery to be
410 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
411 When one of the recipients of a message was unsuitable for the connection
412 (has no matching addresses), we lost track of needing to mark it
413 deferred. As a result mail would be lost.
415 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
417 JH/25 Decoding ACL controls is now done using a binary search; the source code
418 takes up less space and should be simpler to maintain. Merge the ACL
419 condition decode tables also, with similar effect.
421 JH/26 Fix problem with one_time used on a redirect router which returned the
422 parent address unchanged. A retry would see the parent address marked as
423 delivered, so not attempt the (identical) child. As a result mail would
426 JH/27 Fix a possible security hole, wherein a process operating with the Exim
427 UID can gain a root shell. Credit to http://www.halfdog.net/ for
428 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
431 JH/28 Enable {spool,log} filesystem space and inode checks as default.
432 Main config options check_{log,spool}_{inodes,space} are now
433 100 inodes, 10MB unless set otherwise in the configuration.
435 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
436 Previously it only applied to the main-section connection policy
439 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
441 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
442 by me. Added RFC7919 DH primes as an alternative.
444 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
445 are not in the system include path.
447 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
448 GnuTLS, when a session startup failed (eg because the client disconnected)
449 Exim did stdio operations after fclose. This was exposed by a recent
450 change which nulled out the file handle after the fclose.
452 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
453 signed directly by the cert-signing cert, rather than an intermediate
454 OCSP-signing cert. This is the model used by LetsEncrypt.
456 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
458 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
459 an incoming connection.
461 HS/02 Bug 1802: Do not half-close the connection after sending a request
464 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
465 fallback to "prime256v1".
467 JH/34 SECURITY: Use proper copy of DATA command in error message.
468 Could leak key material. Remotely exploitable. CVE-2016-9963.
474 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
475 and 3.4.4 - once the server is enabled to respond to an OCSP request
476 it does even when not requested, resulting in a stapling non-aware
477 client dropping the TLS connection.
479 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
480 support variable-length bit vectors. No functional change.
482 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
483 The I= interface field on outgoing lines is now after the H= remote
484 host field, same as incoming lines. There is a separate
485 outgoing_interface log selector which allows you to disable the
488 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
489 If not running log_selector +smtp_connection the mainlog would be held
490 open indefinitely after a "too many connections" event, including to a
491 deleted file after a log rotate. Leave the per net connection logging
492 leaving it open for efficiency as that will be quickly detected by the
493 check on the next write.
495 HS/01 Bug 1671: Fix post transport crash.
496 Processing the wait-<transport> messages could crash the delivery
497 process if the message IDs didn't exist for some reason. When
498 using 'split_spool_directory=yes' the construction of the spool
499 file name failed already, exposing the same netto behaviour.
501 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
502 mime_regex ACL conditions.
504 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
505 to DSN fail messages (bounces): remote IP, remote greeting, remote response
506 to HELO, local diagnostic string.
508 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
509 log line to debug. Even when configured with a tls authenticator many
510 client connections are expected to not authenticate in this way, so
511 an authenticate fail is not an error.
513 HS/02 Add the Exim version string to the process info. This way exiwhat
514 gives some more detail about the running daemon.
516 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
517 matter for fast-change records such as DNSBLs.
519 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
520 retry record, even if constant. There may be multiple transports with
521 different interface settings and the retry behaviour needs to be kept
524 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
526 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
528 JH/10 Bug 840: fix log_defer_output option of pipe transport
530 JH/11 Bug 830: use same host for all RCPTS of a message, even under
531 hosts_randomize. This matters a lot when combined with mua_wrapper.
533 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
534 ${quote_pgsql:<string>} operator.
536 JH/13 Bug 1708: avoid misaligned access in cached lookup.
538 JH/14 Change header file name for freeradius-client. Relevant if compiling
539 with Radius support; from the Gentoo tree and checked under Fedora.
541 JH/15 Bug 1712: Introduce $prdr_requested flag variable
543 JH/16 Bug 1714: Permit an empty string as expansion result for transport
544 option transport_filter, meaning no filtering.
546 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
548 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
549 defaults to "*" (all hosts). The variable is now available when not built
550 with TLS, default unset, mainly to enable keeping the testsuite sane.
551 If a server certificate is not supplied (via tls_certificate) an error is
552 logged, and clients will find TLS connections fail on startup. Presumably
553 they will retry in-clear.
554 Packagers of Exim are strongly encouraged to create a server certificate
555 at installation time.
557 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
558 with the $config_file variable.
560 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
561 in transport context, after the attempt, and per-recipient. The latter type
562 is per host attempted. The event data is the error message, and the errno
563 information encodes the lookup type (A vs. MX) used for the (first) host,
564 and the trailing two digits of the smtp 4xx response.
566 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
567 to write to mainlog (or rejectlog, paniclog) in the window between file
568 creation and permissions/ownership being changed. Particularly affects
569 installations where exicyclog is run as root, rather than exim user;
570 result is that the running daemon panics and dies.
572 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
574 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
575 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
576 "pri" and "weight". Note that the previous implicit priority given by the
577 list order is no longer honoured.
579 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
582 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
583 by defining SUPPORT_SOCKS.
585 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
586 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
587 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
588 variables are renamed to proxy_{local,external}_{address,port}.
590 JH/27 Move Internationalisation support from Experimental to mainline, enabled
591 for a build by defining SUPPORT_I18N
593 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
594 of the query string, and make ${quote_redis:} do that quoting.
596 JH/29 Move Events support from Experimental to mainline, enabled by default
597 and removable for a build by defining DISABLE_EVENT.
599 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
601 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
602 cached by the daemon.
604 JH/32 Move Redis support from Experimental to mainline, enabled for a build
605 by defining LOOKUP_REDIS. The libhiredis library is required.
607 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
608 keys are given for lookup.
610 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
611 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
612 only supported when built with TLS support. The PolarSSL SHA routines
613 are still used when the TLS library is too old for convenient support.
615 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
616 openssl_options), for security. OpenSSL forces this from version 1.1.0
617 server-side so match that on older versions.
619 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
620 allocation for $value could be released as the expansion processing
621 concluded, but leaving the global pointer active for it.
623 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
624 and to use the domains and local_parts ACL conditions.
626 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
627 incorrectly not doubled on cutthrough transmission, hence seen as a
628 body-termination at the receiving system - resulting in truncated mails.
629 Commonly the sender saw a TCP-level error, and retransmitted the message
630 via the normal store-and-forward channel. This could result in duplicates
631 received - but deduplicating mailstores were liable to retain only the
632 initial truncated version.
634 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
636 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
638 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
639 we're in there, support oversigning also; bug 1309.
641 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
643 HS/04 Add support for keep_environment and add_environment options.
645 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
646 either intentional arithmetic overflow during PRNG, or testing config-
649 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
650 delivery resulted in actual delivery. Cancel cutthrough before DATA
653 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
654 rejects a recipient: pass the reject to the originator.
656 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
657 Many were false-positives and ignorable, but it's worth fixing the
660 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
661 for the new environment-manipulation done at startup. Move the routines
662 from being local to tls.c to being global via the os.c file.
664 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
665 an extract embedded as result-arg for a map, the first arg for extract
666 is unavailable so we cannot tell if this is a numbered or keyed
667 extraction. Accept either.
673 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
676 JH/02 The smtp transport option "multi_domain" is now expanded.
678 JH/03 The smtp transport now requests PRDR by default, if the server offers
681 JH/04 Certificate name checking on server certificates, when exim is a client,
682 is now done by default. The transport option tls_verify_cert_hostnames
683 can be used to disable this per-host. The build option
684 EXPERIMENTAL_CERTNAMES is withdrawn.
686 JH/05 The value of the tls_verify_certificates smtp transport and main options
687 default to the word "system" to access the system default CA bundle.
688 For GnuTLS, only version 3.0.20 or later.
690 JH/06 Verification of the server certificate for a TLS connection is now tried
691 (but not required) by default. The verification status is now logged by
692 default, for both outbound TLS and client-certificate supplying inbound
695 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
698 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
699 Status Notification (bounce) messages are now MIME format per RFC 3464.
700 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
701 under the control of the dsn_advertise_hosts option, and routers may
702 have a dsn_lasthop option.
704 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
705 default, modifiable by a malware= option. The list separator for
706 the options can now be changed in the usual way. Bug 68.
708 JH/10 The smtp_receive_timeout main option is now expanded before use.
710 JH/11 The incoming_interface log option now also enables logging of the
711 local interface on delivery outgoing connections.
713 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
714 if the interface and destination host and port all match.
716 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
719 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
720 Patch from Andrew Lewis.
722 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
723 now supports optional time-restrictions, weighting, and priority
724 modifiers per server. Patch originally by <rommer@active.by>.
726 JH/16 The spamd_address main option now supports a mixed list of local
727 and remote servers. Remote servers can be IPv6 addresses, and
728 specify a port-range.
730 JH/17 Bug 68: The spamd_address main option now supports an optional
731 timeout value per server.
733 JH/18 Bug 1581: Router and transport options headers_add/remove can
734 now have the list separator specified.
736 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
739 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
742 JH/21 Support for the A6 type of dns record is withdrawn.
744 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
745 rather than the verbs used.
747 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
748 from 255 to 1024 chars.
750 JH/24 Verification callouts now attempt to use TLS by default.
752 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
753 are generic router options now. The defaults didn't change.
755 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
756 Original patch from Alexander Shikoff, worked over by JH.
758 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
759 files if ZCAT_COMMAND is not executable.
761 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
763 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
765 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
766 Normally benign, it bites when the pair was led to by a CNAME;
767 modern usage is to not canonicalize the domain to a CNAME target
768 (and we were inconsistent anyway for A-only vs AAAA+A).
770 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
772 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
773 when evaluating $sender_host_dnssec.
775 JH/31 Check the HELO verification lookup for DNSSEC, adding new
776 $sender_helo_dnssec variable.
778 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
780 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
782 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
784 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
785 documented as working, but never had. Support all but $spam_report.
787 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
788 added for tls authenticator.
790 HS/03 Add perl_taintmode main config option
796 TL/01 When running the test suite, the README says that variables such as
797 no_msglog_check are global and can be placed anywhere in a specific
798 test's script, however it was observed that placement needed to be near
799 the beginning for it to behave that way. Changed the runtest perl
800 script to read through the entire script once to detect and set these
801 variables, reset to the beginning of the script, and then run through
802 the script parsing/test process like normal.
804 TL/02 The BSD's have an arc4random API. One of the functions to induce
805 adding randomness was arc4random_stir(), but it has been removed in
806 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
807 function when detected.
809 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
810 cause callback expansion.
812 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
813 syntax errors in an expansion can be treated as a string instead of
814 logging or causing an error, due to the internal use of bool_lax
815 instead of bool when processing it.
817 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
818 server certificates when making smtp deliveries.
820 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
822 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
824 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
826 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
827 Merged patch from Sebastian Wiedenroth.
829 JH/05 Fix results-pipe from transport process. Several recipients, combined
830 with certificate use, exposed issues where response data items split
831 over buffer boundaries were not parsed properly. This eventually
832 resulted in duplicates being sent. This issue only became common enough
833 to notice due to the introduction of connection certificate information,
834 the item size being so much larger. Found and fixed by Wolfgang Breyha.
836 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
837 size buffer was used, resulting in syntax errors when an expansion
840 JH/07 Add support for directories of certificates when compiled with a GnuTLS
841 version 3.3.6 or later.
843 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
844 is EXPERIMENTAL_EVENT, the main-configuration and transport options
845 both become "event_action", the variables become $event_name, $event_data
846 and $event_defer_errno. There is a new variable $verify_mode, usable in
847 routers, transports and related events. The tls:cert event is now also
848 raised for inbound connections, if the main configuration event_action
851 TL/06 In test suite, disable OCSP for old versions of openssl which contained
852 early OCSP support, but no stapling (appears to be less than 1.0.0).
854 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
855 server certificate names available under the smtp transport option
856 "tls_verify_cert_hostname" now do not permit multi-component wildcard
859 JH/10 Time-related extraction expansions from certificates now use the main
860 option "timezone" setting for output formatting, and are consistent
861 between OpenSSL and GnuTLS compilations. Bug 1541.
863 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
864 encoded parameter in the incoming message. Bug 1558.
866 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
867 include certificate info, eximon was claiming there were spoolfile
870 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
872 JH/14 Log delivery-related information more consistently, using the sequence
873 "H=<name> [<ip>]" wherever possible.
875 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
876 are problematic for Debian distribution, omit them from the release
879 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
881 JH/16 Fix string representation of time values on 64bit time_t architectures.
884 JH/17 Fix a null-indirection in certextract expansions when a nondefault
885 output list separator was used.
890 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
891 checkers that were complaining about end of non-void function with no
894 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
895 This was a regression introduced in 4.83 by another bugfix.
897 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
899 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
900 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
906 TF/01 Correctly close the server side of TLS when forking for delivery.
908 When a message was received over SMTP with TLS, Exim failed to clear up
909 the incoming connection properly after forking off the child process to
910 deliver the message. In some situations the subsequent outgoing
911 delivery connection happened to have the same fd number as the incoming
912 connection previously had. Exim would try to use TLS and fail, logging
913 a "Bad file descriptor" error.
915 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
916 utilities have not been installed.
918 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
919 temporary space as the ACL may create new global variables.
921 TL/01 LDAP support uses per connection or global context settings, depending
922 upon the detected version of the libraries at build time.
924 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
925 to extract and use the src ip:port in logging and expansions as if it
926 were a direct connection from the outside internet. PPv2 support was
927 updated based on HAProxy spec change in May 2014.
929 JH/02 Add ${listextract {number}{list}{success}{fail}}.
931 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
932 Properly escape header and check for NULL return.
934 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
937 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
939 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
940 characters in header names, implemented as a verify condition.
941 Contributed by Michael Fischer v. Mollard.
943 TL/05 Rename SPF condition results err_perm and err_temp to standardized
944 results permerror and temperror. Previous values are deprecated but
945 still accepted. In a future release, err_perm and err_temp will be
946 completely removed, which will be a backward incompatibility if the
947 ACL tests for either of these two old results. Patch contributed by
948 user bes-internal on the mailing list.
950 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
952 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
953 selectors, in both main and reject logs.
955 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
958 JH/07 Add malware type "sock" for talking to simple daemon.
960 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
962 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
963 routers/transports under cutthrough routing.
965 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
966 numbers. Touch up "bool" conditional to keep the same definition.
968 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
970 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
972 JH/12 Expand items in router/transport headers_add or headers_remove lists
973 individually rather than the list as a whole. Bug 1452.
975 Required for reasonable handling of multiple headers_ options when
976 they may be empty; requires that headers_remove items with embedded
977 colons must have them doubled (or the list-separator changed).
979 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
980 view the policy declared in the DMARC record. Currently, $dmarc_status
981 is a combined value of both the record presence and the result of the
984 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
986 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
987 dnslookup router and the smtp transport (applying to the forward
990 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
991 of ldap servers used for a specific lookup. Patch provided by Heiko
994 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
995 New variable $lookup_dnssec_authenticated for observability.
997 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
998 Patch submitted by Lars Timman.
1000 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1002 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1003 Requires trusted mode and valid format message id, aborts otherwise.
1004 Patch contributed by Heiko Schlichting.
1006 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1007 certextract with support for various fields. Bug 1358.
1009 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1010 is requested by default, modifiable by smtp transport option
1013 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1014 operate on certificate variables to give certificate fingerprints
1015 Also new ${sha256:cert_variable}.
1017 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1019 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1020 Christian Aistleitner.
1022 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1024 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1025 file. Patch from Wolfgang Breyha.
1027 JH/25 Expand the coverage of the delivery $host and $host_address to
1028 client authenticators run in verify callout. Bug 1476.
1030 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1031 align with daemon_smtp_ports. Bug 72.
1033 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1034 support and error reporting did not work properly.
1036 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1037 and is readable. Patch from Andrew Colin Kissa.
1039 TL/14 Enhance documentation of ${run expansion and how it parses the
1040 commandline after expansion, particularly in the case when an
1041 unquoted variable expansion results in an empty value.
1043 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1045 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1046 by renaming away from it. Addresses GH issue 15, reported by
1049 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1051 TL/15 SECURITY: prevent double expansion in math comparison functions
1052 (can expand unsanitized data). Not remotely exploitable.
1059 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1061 PP/02 Make -n do something, by making it not do something.
1062 When combined with -bP, the name of an option is not output.
1064 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1067 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1068 $sender_host_name and config options to manage this, and basic check
1071 PP/05 DSCP support for outbound connections and control modifier for inbound.
1073 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1074 (Only plugin which currently uses this is kerberos4, which nobody should
1075 be using, but we should make it available and other future plugins might
1076 conceivably use it, even though it would break NAT; stuff *should* be
1077 using channel bindings instead).
1079 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1080 name; added for Sendmail compatibility; requires admin caller.
1081 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1082 just ignore it); requires trusted caller.
1083 Also parse but ignore: -Ac -Am -X<logfile>
1086 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1088 TL/02 Add +smtp_confirmation as a default logging option.
1090 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1091 Patch by Magnus Holmgren from 2007-02-20.
1093 TL/04 Bugzilla 1281 - Spec typo.
1094 Bugzilla 1283 - Spec typo.
1095 Bugzilla 1290 - Spec grammar fixes.
1097 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1099 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1101 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1102 mailing list by by Dmitry Isaikin.
1104 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1106 JH/02 Support "G" suffix to numbers in ${if comparisons.
1108 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1110 NM/01 Bugzilla 1197 - Spec typo
1111 Bugzilla 1196 - Spec examples corrections
1113 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1115 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1116 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1119 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1120 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1122 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1123 "acl {{name}{arg}...}", and optional args on acl condition
1126 JH/05 Permit multiple router/transport headers_add/remove lines.
1128 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1130 JH/07 Avoid using a waiting database for a single-message-only transport.
1131 Performance patch from Paul Fisher. Bugzilla 1262.
1133 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1136 JH/09 Add $headers_added variable, with content from use of ACL modifier
1137 add_header (but not yet added to the message). Bugzilla 199.
1139 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1140 Pulled from Bugzilla 817 by Wolfgang Breyha.
1142 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1144 (nb: this is the same fix as in Exim 4.80.1)
1146 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1149 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1152 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1154 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1155 advertise SMTP AUTH mechanism to us, instead of a generic
1156 protocol violation error. Also, make Exim more robust to bad
1157 data from the Dovecot auth socket.
1159 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1161 When a queue runner is handling a message, Exim first routes the
1162 recipient addresses, during which it prunes them based on the retry
1163 hints database. After that it attempts to deliver the message to
1164 any remaining recipients. It then updates the hints database using
1167 So if a recipient address works intermittently, it can get repeatedly
1168 deferred at routing time. The retry hints record remains fresh so the
1169 address never reaches the final cutoff time.
1171 This is a fairly common occurrence when a user is bumping up against
1172 their storage quota. Exim had some logic in its local delivery code
1173 to deal with this. However it did not apply to per-recipient defers
1174 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1176 This change adds a proper retry rule check during routing so that the
1177 final cutoff time is checked against the message's age. We only do
1178 this check if there is an address retry record and there is not a
1179 domain retry record; this implies that previous attempts to handle
1180 the address had the retry_use_local_parts option turned on. We use
1181 this as an approximation for the destination being like a local
1182 delivery, as in LMTP.
1184 I suspect this new check makes the old local delivery cutoff check
1185 redundant, but I have not verified this so I left the code in place.
1187 TF/02 Correct gecos expansion when From: is a prefix of the username.
1189 Test 0254 submits a message to Exim with the header
1193 When I ran the test suite under the user fanf2, Exim expanded
1194 the header to contain my full name, whereas it should have added
1195 a Resent-Sender: header. It erroneously treats any prefix of the
1196 username as equal to the username.
1198 This change corrects that bug.
1200 GF/01 DCC debug and logging tidyup
1201 Error conditions log to paniclog rather than rejectlog.
1202 Debug lines prefixed by "DCC: " to remove any ambiguity.
1204 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1206 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1207 Bug spotted by Jeremy Harris; was flawed since initial commit.
1208 Would have resulted in OCSP responses post-SNI triggering an Exim
1209 NULL dereference and crash.
1211 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1213 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1214 Bug detection, analysis and fix by Samuel Thibault.
1215 Bugzilla 1331, Debian bug #698092.
1217 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1219 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1220 Server implementation by Todd Lyons, client by JH.
1221 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1222 config variable "prdr_enable" controls whether the server
1223 advertises the facility. If the client requests PRDR a new
1224 acl_data_smtp_prdr ACL is called once for each recipient, after
1225 the body content is received and before the acl_smtp_data ACL.
1226 The client is controlled by both of: a hosts_try_prdr option
1227 on the smtp transport, and the server advertisement.
1228 Default client logging of deliveries and rejections involving
1229 PRDR are flagged with the string "PRDR".
1231 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1232 fclose(). Diagnosis by Todd Lyons.
1234 PP/17 Update configure.default to handle IPv6 localhost better.
1235 Patch by Alain Williams (plus minor tweaks).
1238 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1239 This is now consistent with GnuTLS, and is now documented: the
1240 previous undocumented portable approach to treating the option as
1241 unset was to force an expansion failure. That still works, and
1242 an empty string is now equivalent.
1244 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1245 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1246 not performing validation itself.
1248 PP/20 Added force_command boolean option to pipe transport.
1249 Patch from Nick Koston, of cPanel Inc.
1251 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1254 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1256 PP/21 Fix eximon continuous updating with timestamped log-files.
1257 Broken in a format-string cleanup in 4.80, missed when I repaired the
1258 other false fix of the same issue.
1259 Report and fix from Heiko Schlichting.
1262 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1263 Report from Prashanth Katuri.
1265 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1266 It's SecureTransport, so affects any MacOS clients which use the
1267 system-integrated TLS libraries, including email clients.
1269 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1270 using a MIME ACL for non-SMTP local injection.
1271 Report and assistance in diagnosis by Warren Baker.
1273 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1275 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1277 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1278 last id that failed so it may be referenced in subsequent ACL's.
1280 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1283 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1284 ldap library initialization, allowing self-signed CA's to be
1285 used. Also properly sets require_cert option later in code by
1286 using NULL (global ldap config) instead of ldap handle (per
1287 session). Bug diagnosis and testing by alxgomz.
1289 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1290 the src/util/ subdirectory.
1292 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1293 renamed to Transport Post Delivery Action by Jeremy Harris, as
1296 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1297 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1298 redis_servers = needs to be configured which will be used by the redis
1299 lookup. Patch from Warren Baker, of The Packet Hub.
1301 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1303 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1304 hostname or reverse DNS when processing a host list. Used suggestions
1305 from multiple comments on this bug.
1307 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1309 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1310 Missed a few lines, added it to make the runtest require no keyboard
1313 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1314 contains upper case chars. Make router use caseful_local_part.
1316 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1317 support when GnuTLS has been built with p11-kit.
1323 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1325 This, or similar/improved, will also be change PP/11 of 4.82.
1331 PP/01 Handle short writes when writing local log-files.
1332 In practice, only affects FreeBSD (8 onwards).
1333 Bugzilla 1053, with thanks to Dmitry Isaikin.
1335 NM/01 Bugzilla 949 - Documentation tweak
1337 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1340 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1342 PP/02 Implemented gsasl authenticator.
1344 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1346 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1347 `pkg-config foo` for cflags/libs.
1349 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1350 with rest of GSASL and with heimdal_gssapi.
1352 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1353 `pkg-config foo` for cflags/libs for the TLS implementation.
1355 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1356 properties get this fed in as external SSF. A number of robustness
1357 and debugging improvements to the cyrus_sasl authenticator.
1359 PP/08 cyrus_sasl server now expands the server_realm option.
1361 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1362 Patch by Jeremy Harris.
1364 PP/10 Added dbmjz lookup type.
1366 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1368 PP/12 MAIL args handles TAB as well as SP, for better interop with
1369 non-compliant senders.
1370 Analysis and variant patch by Todd Lyons.
1372 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1373 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1374 Patch from Dirk Mueller <dmueller@suse.com>
1376 PP/13 tls_peerdn now print-escaped for spool files.
1377 Observed some $tls_peerdn in wild which contained \n, which resulted
1378 in spool file corruption.
1380 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1381 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1382 or write after TLS renegotiation, which otherwise led to messages
1385 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1386 as a tracking header (ie: a signed header comes before the signature).
1387 Patch from Wolfgang Breyha.
1389 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1390 comma-sep list; embedded commas doubled.
1392 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1394 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1396 Report and patch from Dmitry Banschikov.
1398 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1399 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1400 failures. We appear to now support TLS1.1+ with Exim.
1402 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1403 lets Exim select keys and certificates based upon TLS SNI from client.
1404 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1405 before an outbound SMTP session. New log_selector, +tls_sni.
1407 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1408 NULL dereference. Report and patch from Alun Jones.
1410 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1411 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1412 Not seeing resolver debug output on NetBSD, but suspect this is a
1413 resolver implementation change.
1415 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1416 Left warnings. Added "eximon gdb" invocation mode.
1418 PP/21 Defaulting "accept_8bitmime" to true, not false.
1420 PP/22 Added -bw for inetd wait mode support.
1422 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1423 locate the relevant includes and libraries. Made this the default.
1425 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1426 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1428 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1429 This may cause build issues on older platforms.
1431 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1432 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1433 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1434 Added SNI support via GnuTLS too.
1435 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1437 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1439 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1440 Applied second patch from Janne, implementing suggestion to default
1441 multiple-strings-in-record handling to match SPF spec.
1443 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1445 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1446 read-only, out of scope).
1447 Patch from Wolfgang Breyha, report from Stuart Northfield.
1449 PP/29 Fix three issues highlighted by clang analyser static analysis.
1450 Only crash-plausible issue would require the Cambridge-specific
1451 iplookup router and a misconfiguration.
1452 Report from Marcin Mirosław.
1454 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1456 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1457 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1458 As part of this, removing so much warning spew let me fix some minor
1459 real issues in debug logging.
1461 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1462 assignment on my part. Fixed.
1464 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1465 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1466 Janne Snabb (who went above and beyond: thank you).
1468 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1469 string otherwise requires a connection and a bunch more work and it's
1470 relatively easy to get wrong. Should also expose TLS library linkage
1473 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1474 64-bit ${eval} (JH/03).
1476 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1477 GNU libc to support some of the 64-bit stuff, should not lead to
1478 conflicts. Defined before os.h is pulled in, so if a given platform
1479 needs to override this, it can.
1481 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1482 protection layer was required, which is not implemented.
1483 Bugzilla 1254, patch from Wolfgang Breyha.
1485 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1486 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1487 tls_dhparam take prime identifiers. Also unbreak combination of
1488 OpenSSL+DH_params+TLSSNI.
1490 PP/39 Disable SSLv2 by default in OpenSSL support.
1496 PP/01 Solaris build fix for Oracle's LDAP libraries.
1497 Bugzilla 1109, patch from Stephen Usher.
1499 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1501 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1504 TF/02 Fix a couple more cases where we did not log the error message
1505 when unlink() failed. See also change 4.74-TF/03.
1507 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1508 lock up or crash if it happened to be inside a call to libc when it
1509 got a SIGUSR1 from exiwhat.
1511 The SIGUSR1 handler appends the current process status to the process
1512 log which is later printed by exiwhat. It used to use the general
1513 purpose logging code to do this, but several functions it calls are
1514 not safe for signals.
1516 The new output code in the SIGUSR1 handler is specific to the process
1517 log, and simple enough that it's easy to inspect for signal safety.
1518 Removing some special cases also simplifies the general logging code.
1519 Removing the spurious timestamps from the process log simplifies
1522 TF/04 Improved ratelimit ACL condition.
1524 The /noupdate option has been deprecated in favour of /readonly which
1525 has clearer semantics. The /leaky, /strict, and /readonly update modes
1526 are mutually exclusive. The update mode is no longer included in the
1527 database key; it just determines when the database is updated. (This
1528 means that when you upgrade Exim will forget old rate measurements.)
1530 Exim now checks that the per_* options are used with an update mode that
1531 makes sense for the current ACL. For example, when Exim is processing a
1532 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1533 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1534 must specify per_mail/readonly. If you omit the update mode it defaults to
1535 /leaky where that makes sense (as before) or /readonly where required.
1537 The /noupdate option is now undocumented but still supported for
1538 backwards compatibility. It is equivalent to /readonly except that in
1539 ACLs where /readonly is required you may specify /leaky/noupdate or
1540 /strict/noupdate which are treated the same as /readonly.
1542 A useful new feature is the /count= option. This is a generalization
1543 of the per_byte option, so that you can measure the throughput of other
1544 aggregate values. For example, the per_byte option is now equivalent
1545 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1547 The per_rcpt option has been generalized using the /count= mechanism
1548 (though it's more complicated than the per_byte equivalence). When it is
1549 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1550 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1551 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1552 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1553 Note that using per_rcpt with a non-readonly update mode in more than
1554 one ACL will cause the recipients to be double-counted. (The per_mail
1555 and per_byte options don't have this problem.)
1557 The handling of very low rates has changed slightly. If the computed rate
1558 is less than the event's count (usually one) then this event is the first
1559 after a long gap. In this case the rate is set to the same as this event's
1560 count, so that the first message of a spam run is counted properly.
1562 The major new feature is a mechanism for counting the rate of unique
1563 events. The new per_addr option counts the number of different
1564 recipients that someone has sent messages to in the last time period. It
1565 behaves like per_rcpt if all the recipient addresses are different, but
1566 duplicate recipient addresses do not increase the measured rate. Like
1567 the /count= option this is a general mechanism, so the per_addr option
1568 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1569 example, measure the rate that a client uses different sender addresses
1570 with the options per_mail/unique=$sender_address. There are further
1571 details in the main documentation.
1573 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1575 TF/06 Removed a few PCRE remnants.
1577 TF/07 Automatically extract Exim's version number from tags in the git
1578 repository when doing development or release builds.
1580 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1581 Bugzilla 879. Patch from Paul Fisher.
1583 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1584 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1587 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1589 PP/05 Variable $av_failed, true if the AV scanner deferred.
1590 Bugzilla 1078. Patch from John Horne.
1592 PP/06 Stop make process more reliably on build failure.
1593 Bugzilla 1087. Patch from Heiko Schlittermann.
1595 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1596 Bugzilla 1089. Patch from Heiko Schlittermann.
1598 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1599 Bugzilla 1131. Patch from Holger Weiß.
1601 PP/09 Handle IPv6 addresses with SPF.
1602 Bugzilla 860. Patch from Wolfgang Breyha.
1604 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1606 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1609 PP/11 match_* no longer expand right-hand-side by default.
1610 New compile-time build option, EXPAND_LISTMATCH_RHS.
1611 New expansion conditions, "inlist", "inlisti".
1613 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1615 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1617 PP/14 fix log_write() format string regression from TF/03.
1618 Bugzilla 1152. Patch from Dmitry Isaikin.
1624 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1626 PP/02 Harmonised TLS library version reporting; only show if debugging.
1627 Layout now matches that introduced for other libraries in 4.74 PP/03.
1629 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1631 PP/04 New "dns_use_edns0" global option.
1633 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1636 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1637 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1639 TK/01 Updated PolarSSL code to 0.14.2.
1640 Bugzilla 1097. Patch from Andreas Metzler.
1642 PP/07 Catch divide-by-zero in ${eval:...}.
1643 Fixes bugzilla 1102.
1645 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1648 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1649 format-string attack -- SECURITY: remote arbitrary code execution.
1651 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1652 time unintentionally subject to list matching rules, letting the header
1653 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1654 arbitrary string expansion). This allowed for information disclosure.
1656 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1657 INT_MIN/-1 -- value coerced to INT_MAX.
1663 NM/01 Workaround for PCRE version dependency in version reporting
1666 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1667 This fixes portability to compilers other than gcc, notably
1668 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1670 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1671 makefiles for portability to HP-UX and POSIX correctness.
1673 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1674 Also via indented variable definition in the Makefile.
1675 (Debugging by Oliver Heesakkers).
1677 PP/02 Restore caching of spamd results with expanded spamd_address.
1678 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1680 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1681 Improves build reliability. Fix from: Frank Elsner
1683 NM/02 Fix wide character breakage in the rfc2047 coding
1684 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1686 NM/03 Allow underscore in dnslist lookups
1687 Fixes bug 1026. Patch from Graeme Fowler
1689 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1690 Code patches from Adam Ciarcinski of NetBSD.
1692 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1695 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1696 is logged, to avoid truncation. Patch from John Horne.
1698 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1699 Patch from Jakob Hirsch.
1701 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1702 SQL string expansion failure details.
1703 Patch from Andrey Oktyabrski.
1705 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1706 Patch from Simon Arlott.
1708 PP/09 New lookups functionality failed to compile on old gcc which rejects
1709 extern declarations in function scope.
1710 Patch from Oliver Fleischmann
1712 PP/10 Use sig_atomic_t for flags set from signal handlers.
1713 Check getgroups() return and improve debugging.
1714 Fixed developed for diagnosis in bug 927 (which turned out to be
1717 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1718 Patch from Mark Zealey.
1720 PP/12 Bugzilla 1056: Improved spamd server selection.
1721 Patch from Mark Zealey.
1723 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1724 Based on patch from Heiko Schlittermann.
1726 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1727 Patch from Uwe Doering, sign-off by Michael Haardt.
1729 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1730 variable declaration deep within a block. Bug and patch from
1733 PP/15 lookups-Makefile IRIX compatibility coercion.
1735 PP/16 Make DISABLE_DKIM build knob functional.
1737 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1738 Patch by Simon Arlott
1740 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1741 variable argument macros. Our copy now differs from upstream.
1747 TF/01 Failure to get a lock on a hints database can have serious
1748 consequences so log it to the panic log.
1750 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1751 controlled using the smtp_confirmation log selector.
1753 TF/03 Include the error message when we fail to unlink a spool file.
1755 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1756 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1757 for maintaining out-of-tree patches for some time.
1759 PP/01 Bugzilla 139: Documentation and portability issues.
1760 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1761 Handle per-OS dynamic-module compilation flags.
1763 PP/02 Let /dev/null have normal permissions.
1764 The 4.73 fixes were a little too stringent and complained about the
1765 permissions on /dev/null. Exempt it from some checks.
1766 Reported by Andreas M. Kirchwitz.
1768 PP/03 Report version information for many libraries, including
1769 Exim version information for dynamically loaded libraries. Created
1770 version.h, now support a version extension string for distributors
1771 who patch heavily. Dynamic module ABI change.
1773 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1774 privilege escalation vulnerability whereby the Exim run-time user
1775 can cause root to append content of the attacker's choosing to
1778 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1781 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1782 If dropping privileges for untrusted macros, we disabled normal logging
1783 on the basis that it would fail; for the Exim run-time user, this is not
1784 the case, and it resulted in successful deliveries going unlogged.
1785 Fixed. Reported by Andreas Metzler.
1791 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1792 only prepend for the Resent-* case. Fixes regression introduced in
1793 Exim 4.70 by NM/22 for Bugzilla 607.
1795 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1796 increasing numbers of administrators be bitten by this.
1798 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1800 PP/03 Bugzilla 994: added openssl_options main configuration option.
1802 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1804 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1806 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1808 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1809 without a peer certificate, leading to a segfault because of an
1810 assumption that peers always have certificates. Be a little more
1811 paranoid. Problem reported by Martin Tscholak.
1813 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1814 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1815 NB: ClamAV planning to remove STREAM in "middle of 2010".
1816 CL also introduces -bmalware, various -d+acl logging additions and
1817 more caution in buffer sizes.
1819 PP/09 Implemented reverse_ip expansion operator.
1821 PP/10 Bugzilla 937: provide a "debug" ACL control.
1823 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1825 PP/12 Bugzilla 973: Implement --version.
1827 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1829 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1831 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1833 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1834 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1835 ignore trailing whitespace.
1837 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1839 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1842 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1843 Notification from Dr Andrew Aitchison.
1845 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1846 ExtendedDetectionInfo response format.
1847 Notification from John Horne.
1849 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1852 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1853 XSL and documented dependency on system catalogs, with examples of how
1856 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1859 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1860 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1861 configuration file which is writeable by the Exim user or group.
1863 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1864 of configuration files to cover files specified with the -C option if
1865 they are going to be used with root privileges, not just the default
1868 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1869 option (effectively making it always true).
1871 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1872 files to be used while preserving root privileges.
1874 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1875 that rogue child processes cannot use them.
1877 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1878 run-time user, instead of root.
1880 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1881 Exim run-time user without dropping privileges.
1883 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1884 result string, instead of calling string_vformat() twice with the same
1887 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1888 for other users. Others should always drop root privileges if they use
1889 -C on the command line, even for a whitelisted configure file.
1891 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1893 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1899 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1900 $data_path, and $header_path variables; fixed documentation bugs and
1903 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1904 exipick to access non-standard spools, including the "frozen" queue
1907 NM/01 Bugzilla 965: Support mysql stored procedures.
1908 Patch from Alain Williams
1910 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1912 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1913 Patch from Andreas Metzler
1915 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1916 Patch from Kirill Miazine
1918 NM/05 Bugzilla 671: Added umask to procmail example.
1920 JJ/03 installed exipick 20100323.0, fixing doc bug
1922 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1923 directory. Notification and patch from Dan Rosenberg.
1925 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1927 TK/02 Improve log output when DKIM signing operation fails.
1929 MH/01 Treat the transport option dkim_domain as a colon separated
1930 list, not as a single string, and sign the message with each element,
1931 omitting multiple occurences of the same signer.
1933 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1934 Bugzilla 985, 986. Patch by Simon Arlott
1936 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1937 Patch by Simon Arlott
1939 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1940 MBX locking. Notification from Dan Rosenberg.
1946 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1948 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1950 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1952 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1954 NM/04 Bugzilla 915: Flag broken perl installation during build.
1960 TK/01 Added patch by Johannes Berg that expands the main option
1961 "spamd_address" if it starts with a dollar sign.
1963 TK/02 Write list of recipients to X-Envelope-Sender header when building
1964 the mbox-format spool file for content scanning (suggested by Jakob
1967 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1968 (http://www.dcc-servers.net/) support via dccifd. Activated by
1969 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1971 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1972 by Mark Daniel Reidel <mr@df.eu>.
1974 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1975 When building exim an external PCRE library is now needed -
1976 PCRE is a system library on the majority of modern systems.
1977 See entry on PCRE_LIBS in EDITME file.
1979 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1980 conversation. Added nologin parameter to request.
1981 Patch contributed by Kirill Miazine.
1983 TF/01 Do not log submission mode rewrites if they do not change the address.
1985 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1987 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1988 log files in place. Contributed by Roberto Lima.
1990 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1992 TF/03 Bugzilla 615: When checking the local_parts router precondition
1993 after a local_part_suffix or local_part_prefix option, Exim now
1994 does not use the address's named list lookup cache, since this
1995 contains cached lookups for the whole local part.
1997 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1998 Robert Millan. Documentation is in experimental-spec.txt.
2000 TF/04 Bugzilla 668: Fix parallel build (make -j).
2002 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2004 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2005 Patch provided by Jan Srzednicki.
2007 TF/05 Leading white space used to be stripped from $spam_report which
2008 wrecked the formatting. Now it is preserved.
2010 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2011 that they are available at delivery time.
2013 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2015 TF/08 TLS error reporting now respects the incoming_interface and
2016 incoming_port log selectors.
2018 TF/09 Produce a more useful error message if an SMTP transport's hosts
2019 setting expands to an empty string.
2021 NM/06 Bugzilla 744: EXPN did not work under TLS.
2022 Patch provided by Phil Pennock.
2024 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2025 Patch provided by Richard Godbee.
2027 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2028 acl_smtp_notquit, added index entry.
2030 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2031 Patch provided by Eugene Bujak.
2033 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2034 accept(). Patch provided by Maxim Dounin.
2036 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2037 Patch provided by Phil Pennock.
2039 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2041 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2042 Patch provided by Brad "anomie" Jorsch.
2044 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2045 Patch provided by Dean Brooks.
2047 TK/05 Add native DKIM support (does not depend on external libraries).
2049 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2050 Patch provided by Graeme Fowler.
2052 NM/16 Bugzilla 851: Documentation example syntax fix.
2054 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2056 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2059 NM/19 Bugzilla 745: TLS version reporting.
2060 Patch provided by Phil Pennock.
2062 NM/20 Bugzilla 167: bool: condition support.
2063 Patch provided by Phil Pennock.
2065 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2066 clients. Patch provided by Phil Pennock.
2068 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2069 Patch provided by Brad "anomie" Jorsch.
2071 NM/23 Bugzilla 687: Fix misparses in eximstats.
2072 Patch provided by Heiko Schlittermann.
2074 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2075 Patch provided by Heiko Schlittermann.
2077 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2078 plus update to original patch.
2080 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2082 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2083 Patch provided by David Brownlee.
2085 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2087 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2089 NM/30 Bugzilla 888: TLS documentation bugfixes.
2091 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2093 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2094 Unlike the original bugzilla I have changed all shell scripts in src tree.
2096 NM/33 Bugzilla 898: Transport filter timeout fix.
2097 Patch by Todd Rinaldo.
2099 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2100 Patch by Serge Demonchaux.
2102 NM/35 Bugzilla 39: Base64 decode bug fixes.
2103 Patch by Jakob Hirsch.
2105 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2107 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2109 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2111 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2117 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2118 ALT-N's libdkim that I have put here:
2119 http://duncanthrax.net/exim-experimental/
2121 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2122 were called 'true' and it seems that C99 defines that as a reserved
2123 keyword to be used with 'bool' variable types. That means you could
2124 not include C99-style headers which use bools without triggering
2125 build errors in sieve.c.
2127 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2128 as mailq or other aliases. Changed the --help handling significantly
2129 to do whats expected. exim_usage() emits usage/help information.
2131 SC/01 Added the -bylocaldomain option to eximstats.
2133 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2135 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2137 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2143 PH/01 Another patch from the Sieve maintainer.
2145 PH/02 When an IPv6 address is converted to a string for single-key lookup
2146 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2147 dots are used instead of colons so that keys in lsearch files need not
2148 contain colons. This was done some time before quoting was made available
2149 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2150 (notated using the quote facility) so as to distinguish them from IPv4
2151 keys. This meant that lookups for IP addresses in host lists did not work
2152 for iplsearch lookups.
2154 This has been fixed by arranging for IPv6 addresses to be expressed with
2155 colons if the lookup type is iplsearch. This is not incompatible, because
2156 previously such lookups could never work.
2158 The situation is now rather anomalous, since one *can* have colons in
2159 ordinary lsearch keys. However, making the change in all cases is
2160 incompatible and would probably break a number of configurations.
2162 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2165 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2166 conversion specification without a maximum field width, thereby enabling
2167 a rogue spamd server to cause a buffer overflow. While nobody in their
2168 right mind would setup Exim to query an untrusted spamd server, an
2169 attacker that gains access to a server running spamd could potentially
2170 exploit this vulnerability to run arbitrary code as the Exim user.
2172 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2173 $primary_hostname instead of what libspf2 thinks the hosts name is.
2175 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2176 a directory entry by the name of the lookup key. Previously, if a
2177 symlink pointed to a non-existing file or a file in a directory that
2178 Exim lacked permissions to read, a lookup for a key matching that
2179 symlink would fail. Now it is enough that a matching directory entry
2180 exists, symlink or not. (Bugzilla 503.)
2182 PH/03 The body_linecount and body_zerocount variables are now exported in the
2185 PH/04 Added the $dnslist_matched variable.
2187 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2188 This means they are set thereafter only if the connection becomes
2191 PH/06 Added the client_condition to authenticators so that some can be skipped
2192 by clients under certain conditions.
2194 PH/07 The error message for a badly-placed control=no_multiline_responses left
2195 "_responses" off the end of the name.
2197 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2199 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2200 (without spaces) instead of just copying the configuration text.
2202 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2204 PH/11 Added $max_received_linelength.
2206 PH/12 Added +ignore_defer and +include_defer to host lists.
2208 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2209 way in which PCRE > 7.0 is built.
2211 PH/14 Implemented queue_only_load_latch.
2213 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2214 MAIL command. The effect was to mangle the value on 64-bit systems.
2216 PH/16 Another patch from the Sieve maintainer.
2218 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2220 PH/18 If a system quota error occurred while trying to create the file for
2221 a maildir delivery, the message "Mailbox is full" was not appended to the
2222 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2223 only to a quota excession during the actual writing of the file.
2225 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2226 characters?) which causes problems in log lines. The DN values are now
2227 passed through string_printing() before being added to log lines.
2229 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2230 and InterBase are left for another time.)
2232 PH/21 Added message_body_newlines option.
2234 PH/22 Guard against possible overflow in moan_check_errorcopy().
2236 PH/23 POSIX allows open() to be a macro; guard against that.
2238 PH/24 If the recipient of an error message contained an @ in the local part
2239 (suitably quoted, of course), incorrect values were put in $domain and
2240 $local_part during the evaluation of errors_copy.
2246 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2247 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2250 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2251 issue a MAIL command.
2253 PH/02 In an ACL statement such as
2255 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2257 if a client was not listed at all, or was listed with a value other than
2258 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2259 the condition was not true (as it should be), so access was not denied.
2260 The bug was that the ! inversion was incorrectly passed on to the second
2261 item. This has been fixed.
2263 PH/03 Added additional dnslists conditions == and =& which are different from
2264 = and & when the dns lookup returns more than one IP address.
2266 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2267 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2269 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2270 FSYNC, which compiles an option called disable_fsync that allows for
2271 bypassing fsync(). The documentation is heavily laced with warnings.
2273 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2275 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2276 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2277 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2278 including adding "make clean"; (3) Added -fPIC when compiling the test
2279 dynamically loaded module, to get rid of a warning.
2281 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2282 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2283 = 0s. The bug is otherwise harmless.
2285 PH/07 There was a bug in the dovecot authenticator such that the value of
2286 $auth1 could be overwritten, and so not correctly preserved, after a
2287 successful authentication. This usually meant that the value preserved by
2288 the server_setid option was incorrect.
2290 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2292 PH/09 Installed PCRE release 7.0.
2294 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2295 run for batched SMTP input. It is now run at the start of every message
2296 in the batch. While fixing this I discovered that the process information
2297 (output by running exiwhat) was not always getting set for -bs and -bS
2298 input. This is fixed, and it now also says "batched" for BSMTP.
2300 PH/11 Added control=no_pipelining.
2302 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2303 patch, slightly modified), and move the expansion of helo_data till after
2304 the connection is made in the smtp transport (so it can use these
2307 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2309 PH/14 Added log_selector = +pid.
2311 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2313 PH/16 Add ${if forany and ${if forall.
2315 PH/17 Added dsn_from option to vary the From: line in DSNs.
2317 PH/18 Flush SMTP output before performing a callout, unless control =
2318 no_callout_flush is set.
2320 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2321 was true (the default) a successful delivery failed to delete the retry
2322 item, thus causing premature timeout of the address. The bug is now
2325 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2327 PH/21 Long custom messages for fakedefer and fakereject are now split up
2328 into multiline responses in the same way that messages for "deny" and
2329 other ACL rejections are.
2331 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2332 with slight modification.
2334 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2335 draft, changing the syntax and factoring some duplicate code".
2337 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2338 for deliveries of the second and subsequent messages over the same SMTP
2341 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2342 ${reduce, with only minor "tidies".
2344 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2346 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2347 expansion side effects.
2349 PH/27 When a message times out after an over-quota error from an Exim-imposed
2350 quota, the bounce message says "mailbox is full". This message was not
2351 being given when it was a system quota that was exceeded. It now should
2354 MH/03 Made $recipients available in local_scan(). local_scan() already has
2355 better access to the recipient list through recipients_list[], but
2356 $recipients can be useful in postmaster-provided expansion strings.
2358 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2359 in the case of a MAIL command with additional options following the
2360 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2361 were accidentally chopped off.
2363 PH/29 SMTP synchronization checks are implemented when a command is read -
2364 there is a check that no more input is waiting when there shouldn't be
2365 any. However, for some commands, a delay in an ACL can mean that it is
2366 some time before the response is written. In this time, more input might
2367 arrive, invalidly. So now there are extra checks after an ACL has run for
2368 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2369 pipelining has not been advertised.
2371 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2373 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2374 returned in the SMTP response when smtp_return_error_details was false.
2375 This has been fixed.
2377 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2378 instead of the C I/O that was originally supplied, because problems were
2379 reported on Solaris.
2381 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2382 Exim which did not show up earlier: it was assuming that a call to
2383 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2384 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2385 was a macro that became an assignment, so it seemed to work. This has
2386 changed to a proper function call with a void return, hence the compile
2387 error. Exim's code has been fixed.
2389 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2392 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2394 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2396 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2397 because it turns out to be too big or there is a timeout) there is no
2398 "Completed" line in the log. When some messages of this type were
2399 selected by exigrep, they were listed as "not completed". Others were
2400 picked up by some special patterns. I have improved the selection
2401 criteria to be more general.
2403 PH/38 The host_find_failed option in the manualroute router can now be set
2404 to "ignore", to completely ignore a host whose IP address cannot be
2405 found. If all hosts are ignored, the behaviour is controlled by the new
2406 host_all_ignored option.
2408 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2409 homing or because of multiple MX records with /mx) generated more than
2410 one IP address, and the following item turned out to be the local host,
2411 all the secondary addresses of the first item were incorrectly removed
2412 from the list, along with the local host and any following hosts (which
2413 is what is supposed to happen).
2415 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2416 whoever called Exim into the -H file. In the case of the daemon it was
2417 behaving confusingly. When first started, it used values for whoever
2418 started the daemon, but after a SIGHUP it used the Exim user (because it
2419 calls itself on a restart). I have changed the code so that it now always
2422 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2423 message are rejected with the same error (e.g. no authentication or bad
2424 sender address), and a DATA command is nevertheless sent (as can happen
2425 with PIPELINING or a stupid MUA), the error message that was given to the
2426 RCPT commands is included in the rejection of the DATA command. This is
2427 intended to be helpful for MUAs that show only the final error to their
2430 PH/42 Another patch from the Sieve maintainer.
2432 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2433 Eximstats - Fixed some broken HTML links and added missing column headers
2435 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2436 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2438 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2440 PH/43 Yet another patch from the Sieve maintainer.
2442 PH/44 I found a way to check for a TCP/IP connection going away before sending
2443 the response to the final '.' that terminates a message, but only in the
2444 case where the client has not sent further data following the '.'
2445 (unfortunately, this is allowed). However, in many cases there won't be
2446 any further data because there won't be any more messages to send. A call
2447 to select() can be used: if it shows that the input is "ready", there is
2448 either input waiting, or the socket has been closed. An attempt to read
2449 the next input character can distinguish the two cases. Previously, Exim
2450 would have sent an OK response which the client would never have see.
2451 This could lead to message repetition. This fix should cure that, at
2452 least in a lot of common cases.
2454 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2455 advertised in response to EHLO.
2461 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2462 fixed by 4.65/MH/01 (is this a record?) are fixed:
2464 (i) An empty string was always treated as zero by the numeric comparison
2465 operators. This behaviour has been restored.
2467 (ii) It is documented that the numeric comparison operators always treat
2468 their arguments as decimal numbers. This was broken in that numbers
2469 starting with 0 were being interpreted as octal.
2471 While fixing these problems I realized that there was another issue that
2472 hadn't been noticed. Values of message_size_limit (both the global option
2473 and the transport option) were treated as octal if they started with 0.
2474 The documentation was vague. These values are now always treated as
2475 decimal, and I will make that clear in the documentation.
2481 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2482 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2485 MH/01 Don't check that the operands of numeric comparison operators are
2486 integers when their expansion is in "skipping" mode (fixes bug
2487 introduced by 4.64-PH/07).
2489 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2490 child addresses, Exim now panics and dies. Previously, because the count
2491 is held in a short int, deliveries were likely to be lost. As such a
2492 large number of recipients for a single message is ridiculous
2493 (performance will be very, very poor), I have chosen to impose a limit
2494 rather than extend the field.
2500 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2501 leftover -K file (the existence of which was triggered by #402).
2502 While we were at it, introduced process PID as part of the -K
2503 filename. This should rule out race conditions when creating
2506 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2507 processing considerably. Previous code took too long for large mails,
2508 triggering a timeout which in turn triggers #401.
2510 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2511 in the DK code in transports.c. sendfile() is not really portable,
2512 hence the _LINUX specificness.
2514 TF/01 In the add_headers option to the mail command in an Exim filter,
2515 there was a bug that Exim would claim a syntax error in any
2516 header after the first one which had an odd number of characters
2519 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2520 callout verification, Exim cached a "reject" for the entire domain. This
2521 is correct for most verifications, but it is not correct for a recipient
2522 verification with use_sender or use_postmaster set, because in that case
2523 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2524 case of MAIL FROM:<> rejection from other early rejections (e.g.
2525 rejection of HELO). When verifying a recipient using a non-null MAIL
2526 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2527 Whatever the result of the callout, the value of the domain cache is
2528 left unchanged (for any other kind of callout, getting as far as trying
2529 RCPT means that the domain itself is ok).
2531 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2534 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2535 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2536 session, the daemon ignores ECONNECT errors and logs others; it now
2537 ignores EPIPE as well.
2539 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2540 (quoted-printable decoding).
2542 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2543 later the small subsequent patch to fix an introduced bug.
2545 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2547 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2549 PH/08 An error is now given if message_size_limit is specified negative.
2551 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2552 to be given (somewhat) arbitrary names.
2554 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2557 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2558 miscellaneous code fixes
2560 PH/10 Added the log_reject_target ACL modifier to specify where to log
2563 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2564 hostname. This is wrong, because it relates to the incoming message (and
2565 probably the interface on which it is arriving) and not to the outgoing
2566 callout (which could be using a different interface). This has been
2567 changed to use the value of the helo_data option from the smtp transport
2568 instead - this is what is used when a message is actually being sent. If
2569 there is no remote transport (possible with a router that sets up host
2570 addresses), $smtp_active_hostname is used.
2572 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2573 tweaks were necessary in order to get it to work (see also 21 below):
2574 (a) The code assumed that strncpy() returns a negative number on buffer
2575 overflow, which isn't the case. Replaced with Exim's string_format()
2577 (b) There were several signed/unsigned issues. I just did the minimum
2578 hacking in of casts. There is scope for a larger refactoring.
2579 (c) The code used strcasecmp() which is not a standard C function.
2580 Replaced with Exim's strcmpic() function.
2581 (d) The code set only $1; it now sets $auth1 as well.
2582 (e) A simple test gave the error "authentication client didn't specify
2583 service in request". It would seem that Dovecot has changed its
2584 interface. Fortunately there's a specification; I followed it and
2585 changed what the client sends and it appears to be working now.
2587 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2590 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2591 address A is aliased to B and C, where B exists and C does not. Without
2592 -v the output is "A verified" because verification stops after a
2593 successful redirection if more than one address is generated. However,
2594 with -v the child addresses are also verified. Exim was outputting "A
2595 failed to verify" and then showing the successful verification for C,
2596 with its parentage. It now outputs "B failed to verify", showing B's
2597 parentage before showing the successful verification of C.
2599 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2600 look up a TXT record in a specific list after matching in a combined
2603 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2604 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2605 they consult the DNS. I had assumed they would set it the way they
2606 wanted; and indeed my experiments on Linux seem to show that in some
2607 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2608 To be on the safe side, however, I have now made the interface to
2609 host_find_byname() similar to host_find_bydns(), with an argument
2610 containing the DNS resolver options. The host_find_byname() function now
2611 sets these options at its start, just as host_find_bydns() does. The smtp
2612 transport options dns_qualify_single and dns_search_parents are passed to
2613 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2614 of host_find_byname() use the default settings of RES_DEFNAMES
2615 (qualify_single) but not RES_DNSRCH (search_parents).
2617 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2618 spool_read_header() do less string testing, by means of a preliminary
2619 switch on the second character of optional "-foo" lines. (This is
2620 overdue, caused by the large number of possibilities that now exist.
2621 Originally there were few.) While I was there, I also converted the
2622 str(n)cmp tests so they don't re-test the leading "-" and the first
2623 character, in the hope this might squeeze out yet more improvement.
2625 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2626 flag allowing group syntax was set by the header_syntax check but not
2627 turned off, possible causing trouble later; (2) The flag was not being
2628 set at all for the header_verify test, causing "group"-style headers to
2629 be rejected. I have now set it in this case, and also caused header_
2630 verify to ignore an empty address taken from a group. While doing this, I
2631 came across some other cases where the code for allowing group syntax
2632 while scanning a header line wasn't quite right (mostly, not resetting
2633 the flag correctly in the right place). These bugs could have caused
2634 trouble for malformed header lines. I hope it is now all correct.
2636 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2637 with the "reply" argument non-NULL. The code, however (which originally
2638 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2639 but it didn't always do it. This confused somebody who was copying the
2640 code for some other use. I have removed all the tests.
2642 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2643 feature that was used to support insecure browsers during the U.S. crypto
2644 embargo. It requires special client support, and Exim is probably the
2645 only MTA that supported it -- and would never use it because real RSA is
2646 always available. This code has been removed, because it had the bad
2647 effect of slowing Exim down by computing (never used) parameters for the
2648 RSA_EXPORT functionality.
2650 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2651 authenticator to fail if there's a tab character in the incoming data
2652 (there should never be unless someone is messing about, as it's supposed
2653 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2654 if the connection is using TLS or if the remote IP is the same as the
2655 local IP, and the "valid-client-cert option" if a client certificate has
2658 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2659 authenticators. This can be used for authorization after authentication
2660 succeeds. (In the case of plaintext, it servers for both authentication
2663 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2664 if any retry times were supplied.
2666 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2667 connection, where there is no client IP address to check. In this
2668 situation, the verify now always succeeds.
2670 PH/25 Applied John Jetmore's -Mset patch.
2672 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2674 PH/27 In a string expansion for a processed (not raw) header when multiple
2675 headers of the same name were present, leading whitespace was being
2676 removed from all of them, but trailing whitespace was being removed only
2677 from the last one. Now trailing whitespace is removed from each header
2678 before concatenation. Completely empty headers in a concatenation (as
2679 before) are ignored.
2681 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2682 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2684 PH/29 [Removed. This was a change that I later backed out, and forgot to
2685 correct the ChangeLog entry (that I had efficiently created) before
2686 committing the later change.]
2688 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2689 address errors (4xx response to RCPT) over the same connection as other
2690 messages routed to the same hosts. Such deliveries are always "forced",
2691 so retry times are not inspected. This resulted in far too many retries
2692 for the affected addresses. The effect occurred only when there were more
2693 hosts than the hosts_max_try setting in the smtp transport when it had
2694 the 4xx errors. Those hosts that it had tried were not added to the list
2695 of hosts for which the message was waiting, so if all were tried, there
2696 was no problem. Two fixes have been applied:
2698 (i) If there are any address or message errors in an SMTP delivery, none
2699 of the hosts (tried or untried) are now added to the list of hosts
2700 for which the message is waiting, so the message should not be a
2701 candidate for sending over the same connection that was used for a
2702 successful delivery of some other message. This seems entirely
2703 reasonable: after all the message is NOT "waiting for some host".
2704 This is so "obvious" that I'm not sure why it wasn't done
2705 previously. Hope I haven't missed anything, but it can't do any
2706 harm, as the worst effect is to miss an optimization.
2708 (ii) If, despite (i), such a delivery is accidentally attempted, the
2709 routing retry time is respected, so at least it doesn't keep
2710 hammering the server.
2712 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2713 in ${readsocket because some servers need this prod.
2715 PH/32 Added some extra debug output when updating a wait-xxx database.
2717 PH/33 The hint "could be header name not terminated by colon", which has been
2718 given for certain expansion errors for a long time, was not being given
2719 for the ${if def:h_colon_omitted{... case.
2721 PH/34 The spec says: "With one important exception, whenever a domain list is
2722 being scanned, $domain contains the subject domain." There was at least
2723 one case where this was not true.
2725 PH/35 The error "getsockname() failed: connection reset by peer" was being
2726 written to the panic log as well as the main log, but it isn't really
2727 panic-worthy as it just means the connection died rather early on. I have
2728 removed the panic log writing for the ECONNRESET error when getsockname()
2731 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2732 runs only) independently of the message's sender address. This meant
2733 that, if the 4xx error was in fact related to the sender, a different
2734 message to the same recipient with a different sender could confuse
2735 things. In particular, this can happen when sending to a greylisting
2736 server, but other circumstances could also provoke similar problems.
2737 I have changed the default so that the retry time for these errors is now
2738 based a combination of the sender and recipient addresses. This change
2739 can be overridden by setting address_retry_include_sender=false in the
2742 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2743 remote server are returned as part of bounce messages. This was not
2744 happening for LMTP over a pipe (the lmtp transport), but now it is the
2745 same for both kinds of LMTP.
2747 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2748 in header lines that were in fact CNAMEs. This is no longer the case.
2750 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2751 and queue runs started by the daemon processed all messages. This has
2752 been fixed so that -R and -S can now usefully be given with -q<time>.
2754 PH/40 Import PCRE release 6.7 (fixes some bugs).
2756 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2758 PH/42 Give an error if -q is specified more than once.
2760 PH/43 Renamed the variables $interface_address and $interface_port as
2761 $received_ip_address and $received_port, to make it clear that these
2762 values apply to message reception, and not to the outgoing interface when
2763 a message is delivered. (The old names remain recognized, of course.)
2765 PH/44 There was no timeout on the connect() call when using a Unix domain
2766 socket in the ${readsocket expansion. There now is.
2768 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2769 be meaningful with "accept".
2771 SC/01 Eximstats V1.43
2772 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2774 SC/02 Eximstats V1.44
2775 Use a glob alias rather than an array ref in the generated
2776 parser. This improves both readability and performance.
2778 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2779 Collect SpamAssassin and rejection statistics.
2780 Don't display local sender or destination tables unless
2781 there is data to show.
2782 Added average volumes into the top table text output.
2784 SC/04 Eximstats V1.46
2785 Collect data on the number of addresses (recipients)
2786 as well as the number of messages.
2788 SC/05 Eximstats V1.47
2789 Added 'Message too big' to the list of mail rejection
2790 reasons (thanks to Marco Gaiarin).
2792 SC/06 Eximstats V1.48
2793 Mainlog lines which have GMT offsets and are too short to
2794 have a flag are now skipped.
2796 SC/07 Eximstats V1.49 (Alain Williams)
2797 Added the -emptyok flag.
2799 SC/08 Eximstats V1.50
2800 Fixes for obtaining the IP address from reject messages.
2802 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2803 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2804 whitespace changes from 4.64-PH/27
2806 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2809 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2810 are found, allow negative numbers in numeric criteria)
2812 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2814 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2815 to match changes made in 4.64-PH/43
2817 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2819 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2820 "domain" is actually an IP address, and force a failure. This locks out
2821 those revolvers/nameservers that support "A-for-A" lookups, in
2822 contravention of the specifications.
2824 PH/48 When a host name was looked up from an IP address, and the subsequent
2825 forward lookup of the name timed out, the host name was left in
2826 $sender_host_name, contrary to the specification.
2828 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2829 restricted to single-key lookups, Exim was not diagnosing an error if
2830 * or *@ was used with a query-style lookup.
2832 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2834 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2835 long ago, but noone interested enough thought of it. Let's just say that
2836 the "1.1" means that there are some new functions that weren't there at
2837 some point in the past.
2839 PH/51 Error processing for expansion failure of helo_data from an smtp
2840 transport during callout processing was broken.
2842 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2843 tested/used via the -bh/-bhc/-bs options.
2845 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2846 bug, fixed in subsequent PCRE releases).
2848 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2849 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2851 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2857 SC/01 Use a glob alias rather than an array ref in eximstats generated
2858 parser. This improves both readability and performance.
2860 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2861 Don't display local sender or destination tables in eximstats unless
2862 there is data to show.
2863 Added average volumes into the eximstats top table text output.
2865 SC/03 Collect data on the number of addresses (recipients) as well
2866 as the number of messages in eximstats.
2868 TF/01 Correct an error in the documentation for the redirect router. Exim
2869 does (usually) call initgroups() when daemonizing.
2871 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2872 with consistent privilege compared to when running as a daemon.
2874 TF/03 Note in the spec that $authenticated_id is not set for local
2875 submissions from trusted users.
2877 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2878 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2880 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2881 by adding some example configuration directives to the default
2882 configuration file. A little bit of work is required to uncomment the
2883 directives and define how usernames and passwords are checked, but
2884 there is now a framework to start from.
2886 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2887 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2888 without this. I don't know how relevant this is to other LDAP libraries.
2890 PH/02 Add the verb name to the "unknown ACL verb" error.
2892 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2894 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2896 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2897 directory not expanded when it should be if an expanded home directory
2898 was set for the address (which is overridden by the transport).
2900 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2903 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2904 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2905 because it is too late at that time, and has no effect.
2907 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2908 security issue with \' (bugzilla #107). I could not use the
2909 PQescapeStringConn() function, because it needs a PGconn value as one of
2912 PH/08 When testing addresses using -bt, indicate those final addresses that
2913 are duplicates that would not cause an additional delivery. At least one
2914 person was confused, thinking that -bt output corresponded to deliveries.
2915 (Suppressing duplicates isn't a good idea as you lose the information
2916 about possibly different redirections that led to the duplicates.)
2918 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2919 systems where poll() doesn't work, in particular OS X.
2921 PH/10 Added more information to debugging output for retry time not reached.
2923 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2924 operations in malware.c.
2926 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2929 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2930 syslog_duplication set false, log lines that would normally be written
2931 both the the main log and to the reject log were not written to syslog at
2934 PH/14 In the default configuration, change the use of "message" in ACL warn
2935 statements to "add_header".
2937 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2938 not followed by a command (e.g. "seen endif").
2940 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2941 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2944 PH/17 Added extra conditions to the default value of delay_warning_condition
2948 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2949 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2950 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2953 The Auto-Submitted: and various List- headers are standardised, whereas I
2954 don't think Precedence: ever was.
2956 PH/18 Refactored debugging code in route_finduser() to show more information,
2957 in particular, the error code if getpwnam() issues one.
2959 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2960 This is apparently needed in addition to the PH/07 change above to avoid
2961 any possible encoding problems.
2963 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2964 but not after initializing Perl.
2966 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2967 output them only if debugging. By default they are written stderr,
2968 apparently, which is not desirable.
2970 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2973 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2976 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2978 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2979 authenticated or an ident call has been made. Suppress the default
2980 values for $authenticated_id and $authenticated_sender (but permit -oMai
2981 and -oMas) when testing with -bh.
2983 PH/24 Re-jigged the order of the tests in the default configuration so that the
2984 tests for valid domains and recipients precede the DNS black list and CSA
2985 tests, on the grounds that those ones are more expensive.
2987 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2988 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2989 This bug exists in every version of Exim that I still have, right back to
2992 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2993 However, an attempt to turn on case-sensitivity in a regex key by
2994 including (?-i) didn't work because the subject string was already
2995 lowercased, and the effects were non-intuitive. It turns out that a
2996 one-line patch can be used to allow (?-i) to work as expected.
3002 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3003 other effects) broke the use of negated acl sub-conditions.
3005 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3008 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3009 "Deny" causes Exim to reject the incoming connection with a 554 error.
3010 Unfortunately, if there is a major crisis, such as a disk failure,
3011 tcp-wrappers gives "deny", whereas what one would like would be some
3012 kind of temporary error. A kludge has been added to help with this.
3013 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3014 554 error is used if errno is still zero or contains ENOENT (which occurs
3015 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3018 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3020 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3021 errors. Otherwise a message that provokes a temporary error (when other
3022 messages do not) can cause a whole host to time out.
3024 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3025 addresses were routed directly to files or pipes from a redirect router.
3026 File deliveries just didn't batch; pipe deliveries might have suffered
3029 PH/06 A failure to get a lock for a hints database would erroneously always say
3030 "Failed to get write lock", even when it was really a read lock.
3032 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3033 of 0600. This has been changed to use the value of the lockfile_mode
3034 option (which defaults to 0600).
3036 PH/08 Applied small patch from the Sieve maintainer.
3038 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3039 folder from quota calculations, a direct delivery into this folder messed
3040 up the contents of the maildirsize file. This was because the regex was
3041 used only to exclude .Trash (or whatever) when the size of the mailbox
3042 was calculated. There was no check that a delivery was happening into an
3043 excluded directory. This bug has been fixed by ignoring all quota
3044 processing for deliveries into excluded directories.
3046 PH/10 Added the maildirfolder_create_regex option to appendfile.
3052 PH/01 The code for finding all the local interface addresses on a FreeBSD
3053 system running IPv6 was broken. This may well have applied to all BSD
3054 systems, as well as to others that have similar system calls. The broken
3055 code found IPv4 interfaces correctly, but gave incorrect values for the
3056 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3057 that it would not match correctly against @[] and not recognize the IPv6
3060 PH/02 The ipliteral router was not recognizing addresses of the form user@
3061 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3063 PH/03 Added disable_ipv6.
3065 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3066 decoded form, because it is most often used to construct To: headers
3067 lines in autoreplies, and the decoded form may well be syntactically
3068 invalid. However, $reply_address has leading white space removed, and all
3069 newlines turned into spaces so that the autoreply transport does not
3072 PH/05 If group was specified without a user on a router, and no group or user
3073 was specified on a transport, the group from the router was ignored.
3075 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3076 for visible compile-time settings that can be used to change these
3077 numbers, for those that want even more. Backwards compatibility with old
3078 spool files has been maintained. However, going back to a previous Exim
3079 release will lost any variables that are in spool files.
3081 PH/07 Two small changes when running in the test harness: increase delay when
3082 passing a TCP/IP connection to a new process, in case the original
3083 process has to generate a bounce, and remove special handling of
3084 127.0.0.2 (sic), which is no longer necessary.
3086 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3087 be the same on different OS.
3089 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3092 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3093 whether --show-vars was specified or not
3095 JJ/02 exipick: Added support for new ACL variable spool format introduced
3098 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3099 syntactically invalid From: or Reply-to: line, and a filter used this to
3100 generate an autoreply, and therefore failed to obtain an address for the
3101 autoreply, Exim could try to deliver to a non-existent relative file
3102 name, causing unrelated and misleading errors. What now happens is that
3103 it logs this as a hard delivery error, but does not attempt to create a
3106 PH/11 The exinext utility has a -C option for testing purposes, but although
3107 the given file was scanned by exinext itself; it wasn't being passed on
3108 when Exim was called.
3110 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3111 an end-of-file indication when reading a command response.
3113 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3114 compiled. In many other places in Exim, IPv6 addresses are always
3115 recognized, so I have changed this. It also means that IPv4 domain
3116 literals of the form [IPV4:n.n.n.n] are now always recognized.
3118 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3119 used if the router is not running as root, for example, when verifying at
3120 ACL time, or when using -bh. The debugging output from this situation was
3121 non-existent - all you got was a failure to exec. I have made two
3124 (a) Failures to set uid/gid, the current directory, or a process leader
3125 in a subprocess such as that created by queryprogram now generate
3126 suitable debugging output when -d is set.
3128 (b) The queryprogram router detects when it is not running as root,
3129 outputs suitable debugging information if -d is set, and then runs
3130 the subprocess without attempting to change uid/gid.
3132 PH/15 Minor change to Makefile for building test_host (undocumented testing
3135 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3136 additional section of a DNS packet that returns MX or SRV records.
3137 Instead, it always explicitly searches for A/AAAA records. This avoids
3138 major problems that occur when a DNS server includes only records of one
3139 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3140 fixed another bug: if SRV records were looked up and the corresponding
3141 address records were *not* found in the additional section, the port
3142 values from the SRV records were lost.
3144 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3145 using the correct key (the original address) when searching the retry
3146 rules in order to find which one to use for generating the retry hint.
3148 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3149 adding the default one. Similarly, if it contains a Reply-To: header, the
3150 errors_reply_to option, if set, is not used.
3152 PH/19 When calculating a retry time, Exim used to measure the "time since
3153 failure" by looking at the "first failed" field in the retry record. Now
3154 it does not use this if it is later than than the arrival time of the
3155 message. Instead it uses the arrival time. This makes for better
3156 behaviour in cases where some deliveries succeed, thus re-setting the
3157 "first failed" field. An example is a quota failure for a huge message
3158 when small messages continue to be delivered. Without this change, the
3159 "time since failure" will always be short, possible causing more frequent
3160 delivery attempts for the huge message than are intended.
3161 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3163 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3164 $1, $2, $3) because the numerical variables can be reset during some
3165 expansion items (e.g. "match"), thereby losing the authentication data.
3167 PH/21 Make -bV show the size of off_t variables so that the test suite can
3168 decide whether to run tests for quotas > 2G.
3170 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3171 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3172 filecount value is greater than 2G or if a quota value is greater than 2G
3173 on a system where the size of off_t is not greater than 4, a panic error
3176 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3177 never match. The debug and -bh output now contains an explicit error
3178 message indicating a malformed IPv4 address or mask.
3180 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3181 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3182 PH/23 above applies.
3184 PH/25 Do not write to syslog when running in the test harness. The only
3185 occasion when this arises is a failure to open the main or panic logs
3186 (for which there is an explicit test).
3188 PH/26 Added the /no_tell option to "control=freeze".
3190 PH/27 If a host name lookup failed very early in a connection, for example, if
3191 the IP address matched host_lookup and the reverse lookup yielded a name
3192 that did not have a forward lookup, an error message of the form "no IP
3193 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3194 could be logged. Now it outputs the IP address instead of "NULL".
3196 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3197 allows the sender and the authenticated sender to be set when
3198 submitting a message from within Exim. Since child_open_exim() is
3199 documented for local_scan(), the new function should be too.
3201 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3202 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3203 results in an empty string is now treated as unset.
3205 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3207 PH/31 Added sender_verify_fail logging option.
3209 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3210 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3211 tidied the source and removed it altogether.
3213 PH/33 When a queue run was abandoned because the load average was too high, a
3214 log line was always written; now it is written only if the queue_run log
3215 selector is set. In addition, the log line for abandonment now contains
3216 information about the queue run such as the pid. This is always present
3217 in "start" and "stop" lines but was omitted from the "abandon" line.
3219 PH/34 Omit spaces between a header name and the colon in the error message that
3220 is given when verify = headers_syntax fails (if there are lots of them,
3221 the message gets confusing).
3223 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3224 names, as there are now some PTR records that contain slashes. This check
3225 is only to protect against broken name servers that fall over on strange
3226 characters, so the fact that it applies to all lookups doesn't matter.
3228 PH/36 Now that the new test suite is complete, we can remove some of the
3229 special code in Exim that was needed for the old test suite. For example,
3230 sorting DNS records because real resolvers return them in an arbitrary
3231 order. The new test suite's fake resolver always returns records in the
3234 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3235 bounces) except when queue_only is set, to avoid logging races between
3236 the different processes.
3238 PH/38 Panic-die if .include specifies a non-absolute path.
3240 PH/39 A tweak to the "H" retry rule from its user.
3242 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3243 a label. They prevented compilation on older perls.
3245 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3246 a warning to be raised on newish perls.
3248 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3249 on queue. Changes to match documented behaviour of showing count of
3250 messages matching specified criteria.
3252 PH/40 Changed the default ident timeout from 30s to 5s.
3254 PH/41 Added support for the use of login_cap features, on those BSD systems
3255 that have them, for controlling the resources used by pipe deliveries.
3257 PH/42 The content-scanning code uses fopen() to create files in which to put
3258 message data. Previously it was not paying any attention to the mode of
3259 the files. Exim runs with umask(0) because the rest of the code creates
3260 files with open(), and sets the required mode explicitly. Thus, these
3261 files were ending up world-writeable. This was not a big issue, because,
3262 being within the spool directory, they were not world-accessible. I have
3263 created a function called modefopen, which takes an additional mode
3264 argument. It sets umask(777), creates the file, chmods it to the required
3265 mode, then resets the umask. All the relevant calls to fopen() in the
3266 content scanning code have been changed to use this function.
3268 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3269 to 24 hours. This avoids potential overflow problems when processing G
3270 and H retry rules. I suspect nobody ever tinkers with this value.
3272 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3274 PH/45 When the plaintext authenticator is running as a client, the server's
3275 challenges are checked to ensure they are valid base64 strings. By
3276 default, the authentication attempt is cancelled if an invalid string is
3277 received. Setting client_ignore_invalid_base64 true ignores these errors.
3278 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3279 they are received. Thus, the responses can be made to depend on the
3280 challenges. If an invalid string is ignored, an empty string is placed in
3283 PH/46 Messages that are created by the autoreply transport now contains a
3284 References: header, in accordance with RFCs 2822 and 3834.
3286 PH/47 Added authenticated_sender_force to the smtp transport.
3288 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3290 PH/49 Installed latest patch from the Sieve maintainer.
3292 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3293 was also set, the appendfile transport was unnecessarily scanning a
3294 directory of message files (e.g. for maildir delivery) to find the count
3295 of files (along with the size), even though it did not need this
3296 information. It now does the scan only if it needs to find either the
3297 size of the count of files.
3299 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3301 PH/52 Two bugs concerned with error handling when the smtp transport is
3304 (i) Exim was not creating retry information for temporary errors given
3305 for individual recipients after the DATA command when the smtp transport
3306 was used in LMTP mode. This meant that they could be retried too
3307 frequently, and not timed out correctly.
3309 (ii) Exim was setting the flag that allows error details to be returned
3310 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3311 recipients that were returned after the DATA command.
3313 PH/53 This is related to PH/52, but is more general: for any failing address,
3314 when detailed error information was permitted to be returned to the
3315 sender, but the error was temporary, then after the final timeout, only
3316 "retry timeout exceeded" was returned. Now it returns the full error as
3317 well as "retry timeout exceeded".
3319 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3320 do this, and (what is worse) MTAs that accept it.
3322 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3323 will now be deprecated.
3325 PH/56 New os.c-cygwin from the Cygwin maintainer.
3327 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3328 formats (previously only available in exim formats via -bpr, -bpru,
3329 and -bpra. Now also available in native and exiqgrep formats)
3331 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3332 with very large, slow to parse queues
3334 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3336 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3338 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3339 responses to authentication challenges, though it was showing the
3340 challenges; (ii) I've removed the CR characters from the debug output for
3343 PH/58 Allow for the insertion of a newline as well as a space when a string
3344 is turned into more than one encoded-word during RFC 2047 encoding. The
3345 Sieve code now uses this.
3347 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3348 data_4xx, lost_connection, tls_required.
3350 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3351 message was being sent as an SMTP response.
3353 PH/61 Add -l and -k options to exicyclog.
3355 PH/62 When verifying, if an address was redirected to one new address, so that
3356 verification continued, and the new address failed or deferred after
3357 having set something in $address_data, the value of $address_data was not
3358 passed back to the ACL. This was different to the case when no
3359 redirection occurred. The value is now passed back in both cases.
3361 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3362 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3363 use login_cap.h, so on its own it isn't the distinguishing feature. The
3364 new name refers directly to the setclassresources() function.
3366 PH/65 Added configuration files for NetBSD3.
3368 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3370 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3371 is preferred over IPv4.
3373 PH/68 The bounce_return_message and bounce_return_body options were not being
3374 honoured for bounces generated during the reception of non-SMTP messages.
3375 In particular, this applied to messages rejected by the ACL. This bug has
3376 been fixed. However, if bounce_return_message is true and bounce_return_
3377 body is false, the headers that are returned for a non-SMTP message
3378 include only those that have been read before the error was detected.
3379 (In the case of an ACL rejection, they have all been read.)
3381 PH/69 The HTML version of the specification is now built in a directory called
3382 spec_html instead of spec.html, because the latter looks like a path with
3383 a MIME-type, and this confuses some software.
3385 PH/70 Catch two compiler warnings in sieve.c.
3387 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3388 function verify_get_ident() calls ip_connect() to connect a socket, but
3389 if the "connect()" function timed out, ip_connect() used to close the
3390 socket. However, verify_get_ident() also closes the socket later, and in
3391 between Exim writes to the log, which may get opened at this point. When
3392 the socket was closed in ip_connect(), the log could get the same file
3393 descriptor number as the socket. This naturally causes chaos. The fix is
3394 not to close the socket in ip_connect(); the socket should be closed by
3395 the function that creates it. There was only one place in the code where
3396 this was missing, in the iplookup router, which I don't think anybody now
3397 uses, but I've fixed it anyway.
3399 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3400 well as to direct DNS lookups. Otherwise the handling of names in host
3401 lists is inconsistent and therefore confusing.
3407 PH/01 Two changes to the default runtime configuration:
3409 (1) Move the checks for relay_from_hosts and authenticated clients from
3410 after to before the (commented out) DNS black list checks.
3412 (2) Add control=submission to the relay_from_hosts and authenticated
3413 clients checks, on the grounds that messages accepted by these
3414 statements are most likely to be submissions.
3416 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3418 (1) Generate an error if the third argument for the ${prvs expansion is
3421 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3424 (3) Reset the variables that are obtained from the first argument of
3425 ${prvscheck and used in the second argument before leaving the code,
3426 because their memory is reclaimed, so using them afterwards may do
3429 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3430 one (it's much easier than Tom thought :-).
3432 (5) Because of (4), we can now allow for the use of $prvscheck_result
3433 inside the third argument.
3435 PH/03 For some reason, the default setting of PATH when running a command from
3436 a pipe transport was just "/usr/bin". I have changed it to
3439 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3440 anything to be listed in the output from -bV.
3442 PH/05 When a filter generated an autoreply, the entire To: header line was
3443 quoted in the delivery log line, like this:
3445 => >A.N.Other <ano@some.domain> <original@ddress> ...
3447 This has been changed so that it extracts the operative address. There
3448 may be more than one such address. If so, they are comma-separated, like
3451 => >ano@some.domain,ona@other.domain <original@ddress> ...
3453 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3454 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3455 not being looked up in the rDNS to get a host name, Exim was showing the
3456 IP address twice in Received: lines, even though the IP addresses were
3457 identical. For example:
3459 Received: from [1.2.3.4] (helo=[1.2.3.4])
3461 However, if the real host name was known, it was omitting the HELO data
3462 if it matched the actual IP address. This has been tidied up so that it
3463 doesn't show the same IP address twice.
3465 PH/07 When both +timestamp and +memory debugging was on, the value given by
3466 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3467 called by the debug printing, thereby overwriting the timestamp buffer.
3468 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3470 PH/08 When the original message was included in an autoreply transport, it
3471 always said "this is a copy of the message, including all the headers",
3472 even if body_only or headers_only was set. It now gives an appropriate
3475 PH/09 Applied a patch from the Sieve maintainer which:
3477 o fixes some comments
3478 o adds the (disabled) notify extension core
3479 o adds some debug output for the result of if/elsif tests
3480 o points to the current vacation draft in the documentation
3481 and documents the missing references header update
3485 o fixes a bug in processing the envelope test (when testing
3486 multiple envelope elements, the last element determined the
3489 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3490 Electronic Mail") by including:
3492 Auto-submitted: auto-generated
3494 in the messages that it generates (bounce messages and others, such as
3495 warnings). In the case of bounce messages for non-SMTP messages, there was
3496 also a typo: it was using "Auto_submitted" (underscore instead of
3497 hyphen). Since every message generated by Exim is necessarily in response
3498 to another message, thes have all been changed to:
3500 Auto-Submitted: auto-replied
3502 in accordance with these statements in the RFC:
3504 The auto-replied keyword:
3506 - SHOULD be used on messages sent in direct response to another
3507 message by an automatic process,
3509 - MUST NOT be used on manually-generated messages,
3511 - MAY be used on Delivery Status Notifications (DSNs) and Message
3512 Disposition Notifications (MDNs),
3514 - MUST NOT be used on messages generated by automatic or periodic
3515 processes, except for messages which are automatic responses to
3518 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3519 to the default Received: header definition.
3521 PH/12 Added log selector acl_warn_skipped (default on).
3523 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3524 variables because (a) they are in the wrong storage pool and (b) even if
3525 they were copied, it wouldn't work properly because of the caching.
3527 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3528 checking when decoding. Apparently there are clients that generate
3529 overlong encoded strings. Why am I not surprised?
3531 PH/15 If the first argument of "${if match_address" was not empty, but did not
3532 contain an "@" character, Exim crashed. Now it writes a panic log message
3533 and treats the condition as false.
3535 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3537 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3538 extension "envelope-auth". The code is finished and in agreement with
3539 other implementations, but there is no documentation so far and in fact,
3540 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3541 not changing the active code.
3543 Print executed "if" and "elsif" statements when debugging is used. This
3544 helps a great deal to understand what a filter does.
3546 Document more things not specified clearly in RFC3028. I had all this
3547 sorted out, when out of a sudden new issues came to my mind. Oops."
3549 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3552 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3553 canonical form (as documented). However, after a host name lookup from
3554 the IP address, check_host() was doing a simple string comparison with
3555 addresses acquired from the DNS when checking that the found name did
3556 have the original IP as one of its addresses. Since any found IPv6
3557 addresses are likely to be in abbreviated form, the comparison could
3558 fail. Luckily, there already exists a function for doing the comparison
3559 by converting both addresses to binary, so now that is used instead of
3560 the text comparison.
3562 PH/20 There was another similar case to PH/19, when a complete host name was
3563 given in a host list; looking up its IP address could give an abbreviated
3564 form, whereas the current host's name might or might not be abbreviated.
3565 The same fix has been applied.
3571 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3572 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3575 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3576 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3578 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3580 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3581 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3582 Also, for "Auto-Submitted", treat anything other than "no" as
3583 non-personal, in accordance with RFC 3834. (Previously it treated
3584 anything starting "auto-" as non-personal.)
3586 TF/01 The control=submission/name=... option had a problem with syntax
3587 errors if the name included a slash character. The /name= option
3588 now slurps the rest of the string, so it can include any characters
3589 but it must come last in the list of options (after /sender_retain
3592 PH/05 Some modifications to the interface to the fake nameserver for the new
3600 TK/01 Added the "success_on_redirect" address verification option. See
3601 NewStuff for rationale and an example.
3603 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3605 PH/02 Patch to exigrep to allow it to work on syslog lines.
3607 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3608 fread() to skip over the body file's header line, because in Cygwin the
3609 header line is locked and is inaccessible.
3611 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3612 co-exist for some time) to make it clear that it is the Exim ID that is
3613 referenced, not the Message-ID: header line.
3615 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3616 string_format() function, because snprintf() does not exist on all
3619 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3620 ${stat: expansion item.
3622 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3623 protocol synchronization error", to keep the pedants happy.
3625 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3626 well as for IRIX systems, when gcc is being used. See the host.c source
3629 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3631 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3634 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3635 transport and to the smtp transport in LMTP mode.
3637 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3639 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3640 run a filter in a subprocess. This could lead to confusion in subsequent
3641 lookups in the parent process. There should also be a search_tidyup() at
3642 the end of the subprocess.
3644 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3645 only if the host matched helo_try_verify_hosts, which caused the
3646 verification to occur when the EHLO/HELO command was issued. The ACL just
3647 tested the remembered result. Now, if a previous verification attempt has
3648 not happened, "verify = helo" does it there and then.
3650 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3652 TK/03 Fix log output including CR from clamd.
3654 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3655 exist provoked a memory error which could cause a segfault.
3657 PH/15 Installed PCRE 6.2
3659 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3661 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3662 of the problem. Specifically, suggested +O2 rather than +O1 for the
3665 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3667 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3668 record, the port was not correctly logged when the outgoing_port log
3669 selector was set (it logged the transort's default port).
3671 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3672 fallback_hosts, and "hosts" in the smtp transport.
3674 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3675 host errors such as "Connection refused".
3677 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3678 authentication with radiusclient 0.4.9:
3680 - Error returned from rc_read_config was caught wrongly
3681 - Username/password not passed on to radius server due to wrong length.
3683 The presumption is that some radiusclient API changes for 4.51/PH/17
3684 were not taken care of correctly. The code is still untested by me (my
3685 Linux distribution still has 0.3.2 of radiusclient), but it was
3686 contributed by a Radius user.
3688 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3689 expanding the "port" option of the smtp transport.
3691 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3692 while reading a MIME header. Thanks to Tom Hughes for a patch.
3694 PH/24 Include config.h inside local_scan.h so that configuration settings are
3697 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3698 that in an ACL for RCPT (for example), you can examine exactly what was
3701 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3702 commands, but it was not correctly comparing the address with the actual
3703 client host address. Thus, it would show the EHLO address in Received:
3704 header lines when this was not necessary.
3706 PH/27 Added the % operator to ${eval:}.
3708 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3709 it should be ignoring failures (because with -C, for example, it has lost
3710 privilege). It wasn't ignoring creation failures other than "already
3713 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3716 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3717 because an input line was too long, either on its own, or by virtue of
3718 too many continuations, the temporary file was not being removed, and the
3719 return code was incorrect.
3721 PH/31 Missing "BOOL" in function definition in filtertest.c.
3723 PH/32 Applied Sieve patches from the maintainer.
3725 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3727 PH/33 Added "verify = not_blind".
3729 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3730 Local/Makefile (with some defaults set). These are used in built scripts
3731 such as exicyclog, but they have never been used in the exim_install
3732 script (though there are many overriding facilities there). I have
3733 arranged that the exim_install script now takes note of these two
3736 PH/35 Installed configuration files for Dragonfly.
3738 PH/36 When a locally submitted message by a trusted user did not contain a
3739 From: header, and the sender address was obtained from -f or from an SMTP
3740 MAIL command, and the trusted user did not use -F to supply a sender
3741 name, $originator_name was incorrectly used when constructing a From:
3742 header. Furthermore, $originator_name was used for submission mode
3743 messages from external hosts without From: headers in a similar way,
3744 which is clearly wrong.
3746 PH/37 Added control=suppress_local_fixups.
3748 PH/38 When log_selector = +received_sender was set, and the addition of the
3749 sender made the log line's construction buffer exactly full, or one byte
3750 less than full, an overflow happened when the terminating "\n" was
3753 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3754 when the result of a list match is failure because a DNS lookup failed.
3756 PH/40 RM_COMMAND is now used in the building process.
3758 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3759 the "build-* directories that it finds.
3761 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3762 address in a domain literal was a prefix of an interface address.
3764 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3765 when verifying a sender address, unless rewrite_headers is false.
3767 PH/44 Wrote a long comment about why errors_to addresses are verified as
3768 recipients, not senders.
3770 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3771 the ratelimit ACL was added.
3773 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3775 PH/46 Added extra information about PostgreSQL errors to the error string.
3777 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3778 avoiding the need to install special zones in a real server. This is
3779 backwards compatible; if it can't find the fake resolver, it drops back.
3780 Thus, both old and new test suites can be run.
3782 TF/02 Added util/ratelimit.pl
3784 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3785 clock is set back in time.
3787 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3788 Candler <B.Candler@pobox.com>.
3790 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3791 OK for addresses that are the result of redirections.
3793 PH/48 A number of further additions for the benefit of the new test suite,
3794 including a fake gethostbyname() that interfaces to the fake DNS resolver
3797 TF/06 The fix for widen_domains has also been applied to qualify_single and
3798 search_parents which are the other dnslookup options that can cause
3801 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3804 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3806 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3807 service" if it encountered an MX record with an empty target hostname.
3808 The message is now "an MX or SRV record indicated no SMTP service".
3810 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3811 if the DNS of the sending site is misconfigured. This is quite a
3812 common situation. This change restores the behaviour of treating a
3813 helo verification defer as a failure.
3815 PH/51 If self=fail was set on a router, the bounce message did not include the
3816 actual error message.
3822 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3824 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3825 command itself ended in error, the underlying message about the transport
3826 filter timeout was being overwritten with the pipe command error. Now the
3827 underlying error message should be appended to the second error message.
3829 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3831 PH/02 Reduce the amount of output that "make" produces by default. Full output
3832 can still be requested.
3834 PH/03 The warning log line about a condition test deferring for a "warn" verb
3835 was being output only once per connection, rather than after each
3836 occurrence (because it was using the same function as for successful
3837 "warn" verbs). This seems wrong, so I have changed it.
3839 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3840 it should not have, which might have caused a crash in the right
3841 circumstances, but probably never did.
3843 PH/04 Installed a modified version of Tony Finch's patch to make submission
3844 mode fix the return path as well as the Sender: header line, and to
3845 add a /name= option so that you can make the user's friendly name appear
3848 TF/03 Added the control = fakedefer ACL modifier.
3850 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3851 Mark Lowes for thorough testing.
3853 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3855 TK/03 Merged latest SRS patch from Miles Wilton.
3857 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3858 with the definition in sysexits.h (which is #included earlier).
3859 Fortunately, Exim does not actually use EX_OK. The code used to try to
3860 preserve the sysexits.h value, by assuming that macro definitions were
3861 scanned for macro replacements. I have been disabused of this notion,
3862 so now the code just undefines EX_OK before #including unistd.h.
3864 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3865 in the smtp transport. When a block could not be written in a single
3866 write() function, the timeout was being re-applied to each part-write.
3867 This seems wrong - if the receiver was accepting one byte at a time it
3868 would take for ever. The timeout is now adjusted when this happens. It
3869 doesn't have to be particularly precise.
3871 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3872 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3874 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3875 without a domain if the check to <postmaster@domain> fails.
3877 SC/01 Eximstats: added -xls and the ability to specify output files
3878 (patch written by Frank Heydlauf).
3880 SC/02 Eximstats: use FileHandles for outputting results.
3882 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3884 SC/04 Eximstats: fixed display of large numbers with -nvr option
3886 SC/05 Eximstats: fixed merging of reports with empty tables.
3888 SC/06 Eximstats: added the -include_original_destination flag
3890 SC/07 Eximstats: removed tabs and trailing whitespace.
3892 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3894 TK/06 MBOX spool code: Add real "From " MBOX separator line
3895 so the .eml file is really in mbox format (even though
3896 most programs do not really care). Patch from Alex Miller.
3898 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3899 The latter is generated from $received_to and is only set if the
3900 message has one envelope recipient. SA can use these headers,
3901 obviously out-of-the-box. Patch from Alex Miller.
3903 PH/08 The ${def test on a variable was returning false if the variable's
3904 value was "0", contrary to what the specification has always said!
3905 The result should be true unless the variable is empty.
3907 PH/09 The syntax error of a character other than { following "${if
3908 def:variable_name" (after optional whitespace) was not being diagnosed.
3909 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3910 accidental colon was present, for example, could give incorrect results.
3912 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3913 result is used (not including appendfile, where other changes are about
3916 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3917 This involved changing a lot of size variables from int to off_t. It
3918 should work with maildirs and everything.
3920 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3921 spamd dying while we are connected to it.
3923 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3926 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3927 function for BDB 4.3.
3929 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3931 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3932 the sources for PCRE have been re-arranged and more files are now
3935 PH/15 The code I had for printing potentially long long variables in PH/11
3936 above was not the best (it lost precision). The length of off_t variables
3937 is now inspected at build time, and an appropriate printing format (%ld
3938 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3939 to be "long long int" or "long int". This is needed for the internal
3940 formatting function string_vformat().
3942 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3943 the configuration file to be ":syslog", then the script "guesses" where
3944 the logs files are, rather than using the compiled in default. In our
3945 case the guess is not the same as the compiled default, so the script
3946 suddenly stopped working when I started to use syslog. The patch checks
3947 to see if log_file_path is "". If so, it attempts to read it from exim
3948 with no configuration file to get the compiled in version, before it
3949 falls back to the previous guessing code."
3951 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3952 implementing BATV in an Exim configuration. See NewStuff for the gory
3955 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3956 Makefile that are specific to HP-UX.
3958 PH/18 If the "use_postmaster" option was set for a recipient callout together
3959 with the "random" option, the postmaster address was used as the MAIL
3960 FROM address for the random test, but not for the subsequent recipient
3961 test. It is now used for both.
3963 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3964 patch removes a few documentation additions to RFC 3028, because the
3965 latest draft now contains them. It adds the new en;ascii-case comparator
3966 and a new error check for 8bit text in MIME parts. Comparator and
3967 require names are now matched exactly. I enabled the subaddress
3968 extension, but it is not well tested yet (read: it works for me)."
3970 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3971 rework some of the code of TK/09 above to avoid the hardwired use of
3972 "%lld" and "long long". Replaced the call to snprintf() with a call to
3975 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3976 records point to non-existent hosts", "retry timeout exceeded", and
3977 "retry time not reached for any host after a long failure period".
3979 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3980 experimental DomainKeys support:
3982 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3983 (2) On an error such as an illegally used "control", the wrong name for
3984 the control was given.
3986 These problems did NOT occur unless DomainKeys support was compiled.
3988 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3990 PH/24 Added ${if match_ip condition.
3992 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3993 databases so that it will be absolutely obvious if a crash occurs in the
3994 DB library. This is a regular occurrence (often caused by mis-matched
3997 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3998 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3999 picked up on a user's system that detects such things. There doesn't seem
4000 to be a gcc warning option for this - only an attribute that has to be
4001 put on the function's prototype. It seems that in Fedora Core 4 they have
4002 set this on a number of new functions. No doubt there will be more in due
4005 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4006 specify a transport. However, if an address that was verified by such a
4007 router was the subject of a callout, Exim crashed because it tried to
4008 read the rcpt_include_affixes from the non-existent transport. Now it
4009 just assumes that the setting of that option is false. This bug was
4010 introduced by 4.51/PH/31.
4012 PH/28 Changed -d+all to exclude +memory, because that information is very
4013 rarely of interest, but it makes the output a lot bigger. People tend to
4014 do -d+all out of habit.
4016 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4017 code in os-type was giving problems when libc.so lives in lib64, like on
4020 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4021 aren't the modern standard, and it seems that some systems' include files
4022 don't always have them. Exim was already checking for some of the newer
4023 ones like T_AAAA, and defining it itself. I've added checks for all the
4024 record types that Exim uses.
4026 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4027 not automatically generating a new one, as it is supposed to. This
4028 prevented TLS from working. If the file did exist, but contained invalid
4029 data, a new version was generated, as expected. It was only the case of a
4030 non-existent file that was broken.
4032 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4033 with a change in libdomainkeys > 0.64.
4035 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4036 from DNS. If the selector record carries the flag, it now has
4037 precedence over the domain-wide flag.
4039 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4041 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4042 the use of a port name that isn't defined in /etc/services) occurred, the
4043 message was deferred as in a normal delivery, and thus remained on the
4044 spool, instead of being failed because of the mua_wrapper setting. This
4045 is now fixed, and I tidied up some of the mua_wrapper messages at the
4048 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4049 the messages in a hash of arrays rather than using individual hashes.
4050 This is a bit cleaner and results in dramatic memory savings, albeit
4051 at a slight CPU cost.
4053 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4054 as requested by Marc Sherman.
4056 SC/10 Eximstats: added histograms for user specified patterns as requested
4059 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4061 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4062 fopen() in the content-scanning modules that did not already have it.
4068 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4069 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4071 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4073 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4075 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4076 to test Sieve filters that use "vacation".
4078 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4079 that changes the way the GnuTLS parameters are stored in the cache file.
4080 The new format can be generated externally. For backward compatibility,
4081 if the data in the cache doesn't make sense, Exim assumes it has read an
4082 old-format file, and it generates new data and writes a new file. This
4083 means that you can't go back to an older release without removing the
4086 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4087 work if there are any delivery delays because "one_time" forces the
4088 parent to be marked "delivered", so its unseen clone is never tried
4089 again. For this reason, Exim now forbids the simultaneous setting of
4092 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4093 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4094 entry if you want to know the details. The fix, however, neglected to
4095 consider the case when local delivery batching is involved. The test for
4096 "previously delivered" was not happening when checking to see if an
4097 address could be batched with a previous (undelivered) one; under
4098 certain circumstances this could lead to multiple deliveries to the same
4101 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4102 in its include files, and this causes problems building Exim.
4104 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4105 header_syntax) but Exim was just ignoring anything given after a slash.
4106 In particular, this caused confusion with an attempt to use "verify =
4107 reverse_host_lookup/defer_ok". An error is now given when options are
4108 supplied for verify items that do not have them. (Maybe reverse_host_
4109 lookup should have a defer_ok option, but that's a different point.)
4111 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4112 defined by RFC 821) to 2048, because there were problems with some AUTH
4113 commands, and RFC 1869 says the size should be increased for extended
4114 SMTP commands that take arguments.
4116 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4119 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4120 "unknown" error; now it says that the functionality isn't in the binary.
4122 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4123 an address' error message when a string expansion fails (syntax or
4124 whatever). Otherwise the password may appear in the log. Following change
4125 PH/42 below, there is no longer a chance of it appearing in a bounce
4128 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4130 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4131 of its addresses was ever tried. (Bugzilla bug #2.)
4133 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4134 the result incorrectly in the debug output. (It correctly added a newline
4135 to what was transported.)
4137 TF/01 Added $received_time.
4139 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4140 commented out examples of how to interface to a virus scanner and to
4141 SpamAssassin. Also added commented examples of av_scanner and
4142 spamd_address settings.
4144 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4145 and controls are allowed in which ACLs. There were a couple of minor
4146 errors. Some of the entries in the conditions table (which is a table of
4147 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4148 negation of where the condition IS allowed.
4150 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4152 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4153 header file does not have a version number, so I've had to invent a new
4154 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4155 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4156 radiusclient), but it was contributed by a Radius user.
4158 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4159 files or directories, for OpenSSL.
4161 PH/19 When an Exim process that is running as root has to create an Exim log
4162 file, it does so in a subprocess that runs as exim:exim so as to get the
4163 ownership right at creation (otherwise, other Exim processes might see
4164 the file with the wrong ownership). There was no test for failure of this
4165 fork() call, which would lead to the process getting stuck as it waited
4166 for a non-existent subprocess. Forks do occasionally fail when resources
4167 run out. I reviewed all the other calls to fork(); they all seem to check
4170 PH/20 When checking for unexpected SMTP input at connect time (before writing
4171 the banner), Exim was not dealing correctly with a non-positive return
4172 from the read() function. If the client had disconnected by this time,
4173 the result was a log entry for a synchronization error with an empty
4174 string after "input=" when read() returned zero. If read() returned -1
4175 (an event I could not check), uninitialized data bytes were printed.
4176 There were reports of junk text (parts of files, etc) appearing after
4179 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4181 PH/22 Added support for macro redefinition, and (re)definition in between
4182 driver and ACL definitions.
4184 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4185 forgetting to use the resulting value; it was using the unexpanded value.
4187 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4188 hadn't been configured. The fix is from Juergen Kreileder, who
4189 understands it better than I do:
4191 "Here's what I see happening with three configured cyrus_sasl
4192 authenticators configured (plain, login, cram-md5):
4194 On startup auth_cyrus_sasl_init() gets called for each of these.
4195 This means three calls to sasl_listmech() without a specified mech_list.
4196 => SASL tests which mechs of all available mechs actually work
4197 => three warnings about OTP not working
4198 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4200 With the patch, sasl_listmech() also gets called three times. But now
4201 SASL's mech_list option is set to the server_mech specified in the the
4202 authenticator. Or in other words, the answer from sasl_listmech()
4203 gets limited to just the mech you're testing for (which is different
4205 => the return list contains just 'plain' or 'login', 'cram-md5' or
4206 nothing depending on the value of ob->server_mech.
4208 I've just tested the patch: Authentication still works fine,
4209 unavailable mechs specified in the exim configuration are still
4210 caught, and the auth.log warnings about OTP are gone."
4212 PH/25 When debugging is enabled, the contents of the command line are added
4213 to the debugging output, even when log_selector=+arguments is not
4216 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4217 answer is "GNU", and only if the return is "GNU/something" is the answer
4220 PH/27 $acl_verify_message is now set immediately after the failure of a
4221 verification in an ACL, and so is available in subsequent modifiers. In
4222 particular, the message can be preserved by coding like this:
4224 warn !verify = sender
4225 set acl_m0 = $acl_verify_message
4227 Previously, $acl_verify_message was set only while expanding "message"
4228 and "log_message" when a very denied access.
4230 PH/28 Modified OS/os.c-Linux with
4232 -#ifndef OS_LOAD_AVERAGE
4233 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4235 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4236 nomenclature these days.)
4238 PH/29 Installed patch from the Sieve maintainer that adds the options
4239 sieve_useraddress and sieve_subaddress to the redirect router.
4241 PH/30 In these circumstances:
4242 . Two addresses routed to the same list of hosts;
4243 . First host does not offer TLS;
4244 . First host accepts first address;
4245 . First host gives temporary error to second address;
4246 . Second host offers TLS and a TLS session is established;
4247 . Second host accepts second address.
4248 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4249 and peerdn, if requested) that were in fact used only for the second
4252 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4253 attention to any local part prefix or suffix that was matched by the
4254 router that accepted the address. It now behaves in the same way as it
4255 does for delivery: the affixes are removed from the local part unless
4256 rcpt_include_affixes is set on the transport.
4258 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4259 timeout during the DATA phase of an incoming message.
4261 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4262 applied a patch sanctioned by the Sieve maintainer.
4264 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4265 the uid or gid is negative. A case of a negative gid caused this to be
4266 noticed. The fix allows for either to be negative.
4268 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4269 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4272 PH/36 The change PH/12 above was broken. Fixed it.
4274 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4275 the grounds that routing the same address twice would always produce the
4276 same answer. This might have been true once, but it is certainly no
4277 longer true now. Routing a child address may depend on the previous
4278 routing that produced that child. Some complicated redirection strategies
4279 went wrong when messages had multiple recipients, and made Exim's
4280 behaviour dependent on the order in which the addresses were given.
4282 I have moved the duplicate checking until after the routing is complete.
4283 Exim scans the addresses that are assigned to local and remote
4284 transports, and removes any duplicates. This means that more work will be
4285 done, as duplicates will always all be routed, but duplicates are
4286 presumably rare, so I don't expect this is of any significance.
4288 For deliveries to pipes, files, and autoreplies, the duplicate checking
4289 still happens during the routing process, since they are not going to be
4292 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4293 It corrects a timeout issue with spamd. This is Ian's comment: "The
4294 background is that sometimes spamd either never reads data from a
4295 connection it has accepted, or it never writes response data. The exiscan
4296 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4297 blindly assumes that writes won't block so it may never time out."
4299 PH/39 Allow G after quota size as well as K and M.
4301 PH/40 The value set for $authenticated_id in an authenticator may not contain
4302 binary zeroes or newlines because the value is written to log lines and
4303 to spool files. There was no check on this. Now the value is run through
4304 the string_printing() function so that such characters are converted to
4305 printable escape sequences.
4307 PH/41 $message_linecount is a new variable that contains the total number of
4308 lines in the message. Compare $body_linecount, which is the count for the
4311 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4312 bounce and delay warning messages, except in certain special cases, which
4315 (a) An SMTP error message from a remote host;
4316 (b) A message specified in a :fail: redirection;
4317 (c) A message specified in a "fail" command in a system filter;
4318 (d) A message specified in a FAIL return from the queryprogram router;
4319 (e) A message specified by the cannot_route_message router option.
4321 In these cases only, Exim does include the error details in bounce and
4322 warning messages. There are also a few cases where bland messages such
4323 as "unrouteable address" or "local delivery error" are given.
4325 PH/43 $value is now also set for the "else" part of a ${run expansion.
4327 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4328 being worked on, but at least Exim now implements the latest version to
4331 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4332 process to complete was treated as a delivery failure, a timeout while
4333 writing the message to the pipe was logged, but erroneously treated as a
4334 successful delivery. Such timeouts include transport filter timeouts. For
4335 consistency with the overall process timeout, these timeouts are now
4336 treated as errors, giving rise to delivery failures by default. However,
4337 there is now a new Boolean option for the pipe transport called
4338 timeout_defer, which, if set TRUE, converts the failures into defers for
4339 both kinds of timeout. A transport filter timeout is now identified in
4342 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4343 systems where "make" and "gmake" are different, calling "gmake" at top
4344 level broke things. I've arranged for the value of $(MAKE) to be passed
4345 from the Makefile to this script so that it can call the same version of
4349 A note about Exim versions 4.44 and 4.50
4350 ----------------------------------------
4352 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4353 changes of various kinds. As a consequence, a big documentation update was
4354 needed. This delayed the release for rather longer than seemed good, especially
4355 in the light of a couple of (minor) security issues. Therefore, the changes
4356 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4357 release. So 4.44 and 4.50 are in effect two different branches that both start
4360 I have left the 4.50 change log unchanged; it contains all the changes since
4361 4.43. The change log for 4.44 is below; many of its items are identical to
4362 those for 4.50. This seems to be the most sensible way to preserve the
4363 historical information.
4369 1. Minor wording change to the doc/README.SIEVE file.
4371 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4372 computation of the current number of files was incorrect.
4374 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4375 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4378 4. Give more explanation in the error message when the command for a transport
4379 filter fails to execute.
4381 5. There are several places where Exim runs a non-Exim command in a
4382 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4383 was being done only for the command run by the queryprogram router. It is
4384 now done for all such subprocesses. The other cases are: ${run, transport
4385 filters, and the commands run by the lmtp and pipe transports.
4387 6. Added CONFIGURE_GROUP build-time option.
4389 7. Some older OS have a limit of 256 on the maximum number of file
4390 descriptors. Exim was using setrlimit() to set 1000 as a large value
4391 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4392 systems. I've change it so that if it can't get 1000, it tries for 256.
4394 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4395 was an oversight, and furthermore, ever since the addition of extra
4396 controls (e.g. 4.43/32), the checks on when to allow different forms of
4397 "control" were broken. There should now be diagnostics for all cases when a
4398 control that does not make sense is encountered.
4400 9. Added the /retain_sender option to "control=submission".
4402 10. $recipients is now available in the predata ACL (oversight).
4404 11. Tidy the search cache before the fork to do a delivery from a message
4405 received from the command line. Otherwise the child will trigger a lookup
4406 failure and thereby defer the delivery if it tries to use (for example) a
4407 cached ldap connection that the parent has called unbind on.
4409 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4410 of $address_data from the recipient verification was clobbered by the
4411 sender verification.
4413 13. The value of address_data from a sender verification is now available in
4414 $sender_address_data in subsequent conditions in the ACL statement.
4416 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4418 15. Added a new option "connect=<time>" to callout options, to set a different
4421 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4422 was its contents. (It was OK if the option was not defined at all.)
4424 17. A "Completed" log line is now written for messages that are removed from
4425 the spool by the -Mrm option.
4427 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4428 information about exactly what failed.
4430 19. Added -dd to debug only the daemon process.
4432 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4433 handles timeouts, both on the server side and network timeouts. Renamed the
4434 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4436 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4437 It is now set to "smtps".
4439 22. $host_address is now set to the target address during the checking of
4440 ignore_target_hosts.
4442 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4443 being passed; this would have caused $sender_host_name to have been used if
4444 matching the list had actually called for a host name (not very likely,
4445 since this list is usually IP addresses). A host name is now passed as
4448 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4449 code that specifies a non-restarting handler (typically sigaction() in
4450 modern systems) in an attempt to fix a rare and obscure crash bug.
4452 25. Narrowed the window for a race in the daemon that could cause it to ignore
4453 SIGCHLD signals. This is not a major problem, because they are used only to
4454 wake it up if nothing else does.
4456 26. A malformed maildirsize file could cause Exim to calculate negative values
4457 for the mailbox size or file count. Odd effects could occur as a result.
4458 The maildirsize information is now recalculated if the size or filecount
4461 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4462 support for a long time. Removed HAVE_SYS_VFS_H.
4464 28. Installed the latest version of exipick from John Jetmore.
4466 29. In an address list, if the pattern was not a regular expression, an empty
4467 subject address (from a bounce message) matched only if the pattern was an
4468 empty string. Non-empty patterns were not even tested. This was the wrong
4469 because it is perfectly reasonable to use an empty address as part of a
4470 database query. An empty address is now tested by patterns that are
4471 lookups. However, all the other forms of pattern expect the subject to
4472 contain a local part and a domain, and therefore, for them, an empty
4473 address still always fails if the pattern is not itself empty.
4475 30. Exim went into a mad DNS loop when attempting to do a callout where the
4476 host was specified on an smtp transport, and looking it up yielded more
4477 than one IP address.
4479 31. Re-factored the code for checking spool and log partition space into a
4480 function that finds that data and another that does the check. The former
4481 is then used to implement four new variables: $spool_space, $log_space,
4482 $spool_inodes, and $log_inodes.
4484 32. The RFC2047 encoding function was originally intended for short strings
4485 such as real names; it was not keeping to the 75-character limit for
4486 encoded words that the RFC imposes. It now respects the limit, and
4487 generates multiple encoded words if necessary. To be on the safe side, I
4488 have increased the buffer size for the ${rfc2047: expansion operator from
4491 33. It is now permitted to omit both strings after an "if" condition; if the
4492 condition is true, the result is "true". As before, when the second string
4493 is omitted, a false condition yields an empty string. This makes it less
4494 cumbersome to write custom ACL and router conditions.
4496 34. Failure to deliver a bounce message always caused it to be frozen, even if
4497 there was an errors_to setting on the router. The errors_to setting is now
4500 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4501 canonical form (fully expanded) before being placed in
4502 $sender_host_address.
4504 36. The table in the code that translates DNS record types into text (T_A to
4505 "A" for instance) was missing entries for NS and CNAME. It is just possible
4506 that this could have caused confusion if both these types were looked up
4507 for the same domain, because the text type is used as part of Exim's
4508 per-process caching. But the chance of anyone hitting this buglet seems
4511 37. The dnsdb lookup has been extended in a number of ways.
4513 (1) There is a new type, "zns", which walks up the domain tree until it
4514 finds some nameserver records. It should be used with care.
4516 (2) There is a new type, "mxh", which is like "mx" except that it returns
4517 just the host names, not the priorities.
4519 (3) It is now possible to give a list of domains (or IP addresses) to be
4520 looked up. The behaviour when one of the lookups defers can be
4521 controlled by a keyword.
4523 (4) It is now possible to specify the separator character for use when
4524 multiple records are returned.
4526 38. The dnslists ACL condition has been extended: it is now possible to supply
4527 a list of IP addresses and/or domains to be looked up in a particular DNS
4530 39. Added log_selector=+queue_time_overall.
4532 40. When running the queue in the test harness, wait just a tad after forking a
4533 delivery process, to get repeatability of debugging output.
4535 41. Include certificate and key file names in error message when GnuTLS fails
4536 to set them up, because the GnuTLS error message doesn't include the name
4537 of the failing file when there is a problem reading it.
4539 42. Allow both -bf and -bF in the same test run.
4541 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4543 44. The "Exiscan patch" is now merged into the mainline Exim source.
4545 45. Sometimes the final signoff response after QUIT could fail to get
4546 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4547 < 0 before doing a fflush(). This bug looks as though it goes back to the
4548 introduction of TLS in release 3.20, but "sometimes" must have been rare
4549 because the tests only now provoked it.
4551 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4552 (this can affect the format of dates).
4554 47. exim_tidydb, when checking for the continued existence of a message for
4555 which it has found a message-specific retry record, was not finding
4556 messages that were in split spool directories. Consequently, it was
4557 deleting retry records that should have stayed in existence.
4559 48. Steve fixed some bugs in eximstats.
4561 49. The SPA authentication driver was not abandoning authentication and moving
4562 on to the next authenticator when an expansion was forced to fail,
4563 contradicting the general specification for all authenticators. Instead it
4564 was generating a temporary error. It now behaves as specified.
4566 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4567 (the order specifies the preference for clients). The order is now AES256,
4568 AES128, 3DES, ARCFOUR128.
4570 51. Small patch to Sieve code - explicitly set From: when generating an
4573 52. Exim crashed if a remote delivery caused a very long error message to be
4574 recorded - for instance if somebody sent an entire SpamAssassin report back
4575 as a large number of 550 error lines. This bug was coincidentally fixed by
4576 increasing the size of one of Exim's internal buffers (big_buffer) that
4577 happened as part of the Exiscan merge. However, to be on the safe side, I
4578 have made the code more robust (and fixed the comments that describe what
4581 53. Now that there can be additional text after "Completed" in log lines (if
4582 the queue_time_overall log selector is set), a one-byte patch to exigrep
4583 was needed to allow it to recognize "Completed" as not the last thing in
4586 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4587 patch that reportedly fixes this has been added. I am not expert enough to
4588 create a test for it. This is what the patch creator wrote:
4590 "I found a little strange behaviour of ldap code when working with
4591 Windows 2003 AD Domain, where users was placed in more than one
4592 Organization Units. When I tried to give exim partial DN, the exit code
4593 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4594 But simultaneously result of request was absolutely normal ldap result,
4595 so I produce this patch..."
4597 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4598 REFERENCE, so I have modified the code to exclude the patch when that macro
4601 55. Some experimental protocols are using DNS PTR records for new purposes. The
4602 keys for these records are domain names, not reversed IP addresses. The
4603 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4604 leaves it alone. Component reversal etc. now happens only for IP addresses.
4607 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4609 57. Double the size of the debug message buffer (to 2048) so that more of very
4610 long debug lines gets shown.
4612 58. The exicyclog utility now does better if the number of log files to keep
4613 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4615 59. Two changes related to the smtp_active_hostname option:
4617 (1) $smtp_active_hostname is now available as a variable.
4618 (2) The default for smtp_banner uses $smtp_active_hostname instead
4619 of $primary_hostname.
4621 60. The host_aton() function is supposed to be passed a string that is known
4622 to be a valid IP address. However, in the case of IPv6 addresses, it was
4623 not checking this. This is a hostage to fortune. Exim now panics and dies
4624 if the condition is not met. A case was found where this could be provoked
4625 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4626 components; fortuitously, this particular loophole had already been fixed
4627 by change 4.50/55 above.
4629 If there are any other similar loopholes, the new check in host_aton()
4630 itself should stop them being exploited. The report I received stated that
4631 data on the command line could provoke the exploit when Exim was running as
4632 exim, but did not say which command line option was involved. All I could
4633 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4634 running as the user.
4637 61. There was a buffer overflow vulnerability in the SPA authentication code
4638 (which came originally from the Samba project). I have added a test to the
4639 spa_base64_to_bits() function which I hope fixes it.
4642 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4643 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4645 63. The daemon start-up calls getloadavg() while still root for those OS that
4646 need the first call to be done as root, but it missed one case: when
4647 deliver_queue_load_max is set with deliver_drop_privilege. This is
4648 necessary for the benefit of the queue runner, because there is no re-exec
4649 when deliver_drop_privilege is set.
4651 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4652 This has been fixed.
4654 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4655 was in use, was not putting the data itself into the right store pool;
4656 consequently, it could be overwritten for a subsequent message in the same
4657 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4660 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4662 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4663 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4664 the function were treating the return as a boolean value, which happened to
4665 work because 0=false and not-0=true, but is not correct code.
4667 68. The host_aton() function was not handling scoped IPv6 addresses (those
4668 with, for example, "%eth0" on the end) correctly.
4670 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4671 negated items (that is, ~something) in unsigned ints. Some compilers
4672 apparently mutter when there is no cast.
4674 70. If an address verification called from an ACL failed, and did not produce a
4675 user-specific message (i.e. there was only a "system" message), nothing was
4676 put in $acl_verify_message. In this situation, it now puts the system
4679 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4680 session; change 4.31/43 added the unwanted input to the log line - except
4681 that it did not do this in the start of session case. It now does.
4683 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4684 This is wrong and can cause the other end to generate a synchronization
4685 error if it is another Exim or anything else that does the synchronization
4686 check. A QUIT command is no longer sent after a timeout.
4688 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4689 during host lookups.
4691 74. The defer_ok option of callout verification was not working if it was used
4692 when verifying addresses in header lines, that is, for this case:
4694 verify = header_sender/callout=defer_ok
4696 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4697 those file descriptors could be used for SMTP connections. If anything
4698 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4699 could be sent to the SMTP client, causing chaos. The daemon now opens
4700 stdin, stdout, and stderr to /dev/null when it puts itself into the
4703 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4704 log by default. The user can override this with suitable Perl magic.
4706 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4707 the log message when discard triggers, was not working for the DATA ACL or
4708 for the non-SMTP ACL.
4710 78. Error message wording change in sieve.c.
4712 79. If smtp_accept_max_per_host was set, the number of connections could be
4713 restricted to fewer than expected, because the daemon was trying to set up
4714 a new connection before checking whether the processes handling previous
4715 connections had finished. The check for completed processes is now done
4716 earlier. On busy systems, this bug wouldn't be noticed because something
4717 else would have woken the daemon, and it would have reaped the completed
4720 80. If a message was submitted locally by a user whose login name contained one
4721 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4722 It caused a spool format error. I have fixed the spool reading code. A
4723 related problem was that the "from" clause in the Received: line became
4724 illegal because of the space(s). It is now covered by ${quote_local_part.
4726 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4729 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4731 83. Patch from Sieve maintainer to fix unterminated string problem in
4732 "vacation" handling.
4734 84. Some minor changes to the Linux configuration files to help with other
4735 OS variants using glibc.
4737 85. One more patch for Sieve to update vacation handling to latest spec.
4740 ----------------------------------------------------
4741 See the note above about the 4.44 and 4.50 releases.
4742 ----------------------------------------------------
4748 1. Change 4.43/35 introduced a bug that caused file counts to be
4749 incorrectly computed when quota_filecount was set in an appendfile
4752 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4753 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4756 3. Give more explanation in the error message when the command for a transport
4757 filter fails to execute.
4759 4. There are several places where Exim runs a non-Exim command in a
4760 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4761 was being done only for the command run by the queryprogram router. It is
4762 now done for all such subprocesses. The other cases are: ${run, transport
4763 filters, and the commands run by the lmtp and pipe transports.
4765 5. Some older OS have a limit of 256 on the maximum number of file
4766 descriptors. Exim was using setrlimit() to set 1000 as a large value
4767 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4768 systems. I've change it so that if it can't get 1000, it tries for 256.
4770 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4771 was an oversight, and furthermore, ever since the addition of extra
4772 controls (e.g. 4.43/32), the checks on when to allow different forms of
4773 "control" were broken. There should now be diagnostics for all cases when a
4774 control that does not make sense is encountered.
4776 7. $recipients is now available in the predata ACL (oversight).
4778 8. Tidy the search cache before the fork to do a delivery from a message
4779 received from the command line. Otherwise the child will trigger a lookup
4780 failure and thereby defer the delivery if it tries to use (for example) a
4781 cached ldap connection that the parent has called unbind on.
4783 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4784 of $address_data from the recipient verification was clobbered by the
4785 sender verification.
4787 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4788 was its contents. (It was OK if the option was not defined at all.)
4790 11. A "Completed" log line is now written for messages that are removed from
4791 the spool by the -Mrm option.
4793 12. $host_address is now set to the target address during the checking of
4794 ignore_target_hosts.
4796 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4797 being passed; this would have caused $sender_host_name to have been used if
4798 matching the list had actually called for a host name (not very likely,
4799 since this list is usually IP addresses). A host name is now passed as
4802 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4803 code that specifies a non-restarting handler (typically sigaction() in
4804 modern systems) in an attempt to fix a rare and obscure crash bug.
4806 15. Narrowed the window for a race in the daemon that could cause it to ignore
4807 SIGCHLD signals. This is not a major problem, because they are used only to
4808 wake it up if nothing else does.
4810 16. A malformed maildirsize file could cause Exim to calculate negative values
4811 for the mailbox size or file count. Odd effects could occur as a result.
4812 The maildirsize information is now recalculated if the size or filecount
4815 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4816 support for a long time. Removed HAVE_SYS_VFS_H.
4818 18. Updated exipick to current release from John Jetmore.
4820 19. Allow an empty sender to be matched against a lookup in an address list.
4821 Previously the only cases considered were a regular expression, or an
4824 20. Exim went into a mad DNS lookup loop when doing a callout where the
4825 host was specified on the transport, if the DNS lookup yielded more than
4828 21. The RFC2047 encoding function was originally intended for short strings
4829 such as real names; it was not keeping to the 75-character limit for
4830 encoded words that the RFC imposes. It now respects the limit, and
4831 generates multiple encoded words if necessary. To be on the safe side, I
4832 have increased the buffer size for the ${rfc2047: expansion operator from
4835 22. Failure to deliver a bounce message always caused it to be frozen, even if
4836 there was an errors_to setting on the router. The errors_to setting is now
4839 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4840 canonical form (fully expanded) before being placed in
4841 $sender_host_address.
4843 24. Updated eximstats to version 1.33
4845 25. Include certificate and key file names in error message when GnuTLS fails
4846 to set them up, because the GnuTLS error message doesn't include the name
4847 of the failing file when there is a problem reading it.
4849 26. Expand error message when OpenSSL has problems setting up cert/key files.
4852 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4853 (this can affect the format of dates).
4855 28. exim_tidydb, when checking for the continued existence of a message for
4856 which it has found a message-specific retry record, was not finding
4857 messages that were in split spool directories. Consequently, it was
4858 deleting retry records that should have stayed in existence.
4860 29. eximstats updated to version 1.35
4861 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4862 1.35 - bugfix such that pie charts by volume are generated correctly
4864 30. The SPA authentication driver was not abandoning authentication and moving
4865 on to the next authenticator when an expansion was forced to fail,
4866 contradicting the general specification for all authenticators. Instead it
4867 was generating a temporary error. It now behaves as specified.
4869 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4870 (the order specifies the preference for clients). The order is now AES256,
4871 AES128, 3DES, ARCFOUR128.
4873 31. Small patch to Sieve code - explicitly set From: when generating an
4876 32. Exim crashed if a remote delivery caused a very long error message to be
4877 recorded - for instance if somebody sent an entire SpamAssassin report back
4878 as a large number of 550 error lines. This bug was coincidentally fixed by
4879 increasing the size of one of Exim's internal buffers (big_buffer) that
4880 happened as part of the Exiscan merge. However, to be on the safe side, I
4881 have made the code more robust (and fixed the comments that describe what
4884 33. Some experimental protocols are using DNS PTR records for new purposes. The
4885 keys for these records are domain names, not reversed IP addresses. The
4886 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4887 leaves it alone. Component reversal etc. now happens only for IP addresses.
4890 34. The host_aton() function is supposed to be passed a string that is known
4891 to be a valid IP address. However, in the case of IPv6 addresses, it was
4892 not checking this. This is a hostage to fortune. Exim now panics and dies
4893 if the condition is not met. A case was found where this could be provoked
4894 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4895 components; fortuitously, this particular loophole had already been fixed
4896 by change 4.50/55 or 4.44/33 above.
4898 If there are any other similar loopholes, the new check in host_aton()
4899 itself should stop them being exploited. The report I received stated that
4900 data on the command line could provoke the exploit when Exim was running as
4901 exim, but did not say which command line option was involved. All I could
4902 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4903 running as the user.
4906 35. There was a buffer overflow vulnerability in the SPA authentication code
4907 (which came originally from the Samba project). I have added a test to the
4908 spa_base64_to_bits() function which I hope fixes it.
4911 36. The daemon start-up calls getloadavg() while still root for those OS that
4912 need the first call to be done as root, but it missed one case: when
4913 deliver_queue_load_max is set with deliver_drop_privilege. This is
4914 necessary for the benefit of the queue runner, because there is no re-exec
4915 when deliver_drop_privilege is set.
4917 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4918 was in use, was not putting the data itself into the right store pool;
4919 consequently, it could be overwritten for a subsequent message in the same
4920 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4923 38. Sometimes the final signoff response after QUIT could fail to get
4924 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4925 < 0 before doing a fflush(). This bug looks as though it goes back to the
4926 introduction of TLS in release 3.20, but "sometimes" must have been rare
4927 because the tests only now provoked it.
4933 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4934 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4935 Now it returns an integer. A number of calls were still expecting a T/F
4936 return. Fortuitously, in all cases, the tests worked in OK situations,
4937 which is the norm. However, things would have gone wrong on any write
4938 failures on the smtp file descriptor. This function is used when sending
4939 messages over SMTP and also when doing verify callouts.
4941 2. When Exim is called to do synchronous delivery of a locally submitted
4942 message (the -odf or -odi options), it no longer closes stderr before doing
4945 3. Implemented the mua_wrapper option.
4947 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4949 5. Implemented the functions header_remove(), header_testname(),
4950 header_add_at_position(), and receive_remove_recipient(), and exported them
4953 6. If an ACL "warn" statement specified the addition of headers, Exim already
4954 inserted X-ACL-Warn: at the start if there was no header name. However, it
4955 was not making this test for the second and subsequent header lines if
4956 there were newlines in the string. This meant that an invalid header could
4957 be inserted if Exim was badly configured.
4959 7. Allow an ACL "warn" statement to add header lines at the start or after all
4960 the Received: headers, as well as at the end.
4962 8. Added the rcpt_4xx retry error code.
4964 9. Added postmaster_mailfrom=xxx to callout verification option.
4966 10. Added mailfrom=xxxx to the callout verification option, for verify=
4969 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4970 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4972 12. Inserted some casts to stop certain compilers warning when using pointer
4973 differences as field lengths or precisions in printf-type calls (mostly
4974 affecting debugging statements).
4976 13. Added optional readline() support for -be (dynamically loaded).
4978 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4979 same clock tick as a message's arrival, so that its received time was the
4980 same as the "first fail" time on the retry record, and that message
4981 remained on the queue past the ultimate address timeout, every queue runner
4982 would try a delivery (because it was past the ultimate address timeout) but
4983 after another failure, the ultimate address timeout, which should have then
4984 bounced the address, did not kick in. This was a "< instead of <=" error;
4985 in most cases the first failure would have been in the next clock tick
4986 after the received time, and all would be well.
4988 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4989 being recognized when the domain list was tested by the match_domain
4990 condition in an expansion string.
4992 16. Added the ${str2b64: operator.
4994 17. Exim was always calling setrlimit() to set a large limit for the number of
4995 processes, without checking whether the existing limit was already
4996 adequate. (It did check for the limit on file descriptors.) Furthermore,
4997 errors from getrlimit() and setrlimit() were being ignored. Now they are
4998 logged to the main and panic logs, but Exim does carry on, to try to do its
4999 job under whatever limits there are.
5001 18. Imported PCRE 5.0.
5003 19. Trivial typo in log message " temporarily refused connection" (the leading
5006 20. If the log selector return_path_on_delivery was set and an address was
5007 redirected to /dev/null, the delivery process crashed because it assumed
5008 that a return path would always be set for a "successful" delivery. In this
5009 case, the whole delivery is bypassed as an optimization, and therefore no
5012 21. Internal re-arrangement: the function for sending a challenge and reading
5013 a response while authentication was assuming a zero-terminated challenge
5014 string. It's now changed to take a pointer and a length, to allow for
5015 binary data in such strings.
5017 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5019 23. Exim was not respecting finduser_retries when seeking the login of the
5020 uid under which it was called; it was always trying 10 times. (The default
5021 setting of finduser_retries is zero.) Also, it was sleeping after the final
5022 failure, which is pointless.
5024 24. Implemented tls_on_connect_ports.
5026 25. Implemented acl_smtp_predata.
5028 26. If the domain in control=submission is set empty, Exim assumes that the
5029 authenticated id is a complete email address when it generates From: or
5030 Sender: header lines.
5032 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5033 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5034 chown and chgrp in /bin and hostname in /usr/bin.
5036 28. Exim was keeping the "process log" file open after each use, just as it
5037 does for the main log. This opens the possibility of it remaining open for
5038 long periods when the USR1 signal hits a daemon. Occasional processlog
5039 errors were reported, that could have been caused by this. Anyway, it seems
5040 much more sensible not to leave this file open at all, so that is what now
5043 29. The long-running daemon process does not normally write to the log once it
5044 has entered its main loop, and it closes the log before doing so. This is
5045 so that log files can straightforwardly be renamed and moved. However,
5046 there are a couple of unusual error situations where the daemon does write
5047 log entries, and I had neglected to close the log afterwards.
5049 30. The text of an SMTP error response that was received during a remote
5050 delivery was being truncated at 512 bytes. This is too short for some of
5051 the long messages that one sometimes sees. I've increased the limit to
5054 31. It is now possible to make retry rules that apply only when a message has a
5055 specific sender, in particular, an empty sender.
5057 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5058 it possible to be selective about when SMTP synchronization is enforced.
5060 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5062 32. Implemented hosts_connection_nolog.
5064 33. Added an ACL for QUIT.
5066 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5069 35. Added mailbox_size and mailbox_filecount to appendfile.
5071 36. Added control = no_multiline_responses to ACLs.
5073 37. There was a bug in the logic of the code that waits for the clock to tick
5074 in the case where the clock went backwards by a substantial amount such
5075 that the microsecond fraction of "now" was more than the microsecond
5076 fraction of "then" (but the whole seconds number was less).
5078 38. Added support for the libradius Radius client library this is found on
5079 FreeBSD (previously only the radiusclient library was supported).
5085 1. When certain lookups returned multiple values in the form name=value, the
5086 quoting of the values was not always being done properly. Specifically:
5087 (a) If the value started with a double quote, but contained no whitespace,
5089 (b) If the value contained whitespace other than a space character (i.e.
5090 tabs or newlines or carriage returns) it was not quoted.
5091 This fix has been applied to the mysql and pgsql lookups by writing a
5092 separate quoting function and calling it from the lookup code. The fix
5093 should probably also be applied to nisplus, ibase and oracle lookups, but
5094 since I cannot test any of those, I have not disturbed their existing code.
5096 2. A hit in the callout cache for a specific address caused a log line with no
5097 reason for rejecting RCPT. Now it says "Previous (cached) callout
5098 verification failure".
5100 3. There was an off-by-one bug in the queryprogram router. An over-long
5101 return line was truncated at 256 instead of 255 characters, thereby
5102 overflowing its buffer with the terminating zero. As well as fixing this, I
5103 have increased the buffer size to 1024 (and made a note to document this).
5105 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5106 when Exim is waiting for an SMTP response from a remote server, Exim
5107 restarts its select() call on the socket, thereby resetting its timeout.
5108 This is not a problem when such interrupts are rare. Somebody set up a cron
5109 job to run exiwhat every 2 minutes, which is less than the normal select()
5110 timeout (5 or 10 minutes). This meant that the select() timeout never
5111 kicked in because it was always reset. I have fixed this by comparing the
5112 time when an interrupt arrives with the time at the start of the first call
5113 to select(). If more time than the timeout has elapsed, the interrupt is
5114 treated as a timeout.
5116 5. Some internal re-factoring in preparation for the addition of Sieve
5117 extensions (by MH). In particular, the "personal" test is moved to a
5118 separate function, and given an option for scanning Cc: and Bcc: (which is
5119 not set for Exim filters).
5121 6. When Exim created an email address using the login of the caller as the
5122 local part (e.g. when creating a From: or Sender: header line), it was not
5123 quoting the local part when it contained special characters such as @.
5125 7. Installed new OpenBSD configuration files.
5127 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5128 try to make them clearer.
5130 9. Callout options, other than the timeout value, were being ignored when
5131 verifying sender addresses in header lines. For example, when using
5133 verify = header_sender/callout=no_cache
5135 the cache was (incorrectly) being used.
5137 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5138 only the Cygwin environment.
5140 11. When return_path_on_delivery was set as a log selector, if different remote
5141 addresses in the same message used different return paths and parallel
5142 remote delivery occurred, the wrong values would sometimes be logged.
5143 (Whenever a remote delivery process finished, the return path value from
5144 the most recently started remote delivery process was logged.)
5146 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5147 lines when AUTH and/or TLS are in use. This is the "received protocol"
5148 field. Exim used to use "asmtp" for authenticated SMTP, without any
5149 indication (in the protocol name) for TLS use. Now it follows the RFC and
5150 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5151 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5152 names appear in log lines as well as in Received: header lines.
5154 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5155 extensions, and comparison tests, and to fix some bugs.
5157 14. Changes to the "personal" filter test:
5159 (1) The test was buggy in that it was just doing the equivalent of
5160 "contains" tests on header lines. For example, if a user's address was
5161 anne@some.where, the "personal" test would incorrectly be true for
5163 To: susanne@some.where
5165 This test is now done by extracting each address from the header in turn,
5166 and checking the entire address. Other tests that are part of "personal"
5167 are now done using regular expressions (for example, to check local parts
5168 of addresses in From: header lines).
5170 (2) The list of non-personal local parts in From: addresses has been
5171 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5172 taken from the Sieve specification recommendations.
5174 (3) If the message contains any header line starting with "List-" it is
5175 treated as non-personal.
5177 (4) The test for "circular" in the Subject: header line has been removed
5178 because it now seems ill-conceived.
5180 15. Minor typos in src/EDITME comments corrected.
5182 16. Installed latest exipick from John Jetmore.
5184 17. If headers_add on a router specified a text string that was too long for
5185 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5186 of string_sprintf() is now avoided.
5188 18. $message_body_size was not set (it was always zero) when running the DATA
5189 ACL and the local_scan() function.
5191 19. For the "mail" command in an Exim filter, no default was being set for
5192 the once_repeat time, causing a random time value to be used if "once" was
5193 specified. (If the value happened to be <= 0, no repeat happened.) The
5194 default is now 0s, meaning "never repeat". The "vacation" command was OK
5195 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5196 (I found it when inspecting the code).
5198 20. There is now an overall timeout for performing a callout verification. It
5199 defaults to 4 times the callout timeout, which applies to individual SMTP
5200 commands during the callout. The overall timeout applies when there is more
5201 than one host that can be tried. The timeout is checked before trying the
5202 next host. This prevents very long delays if there are a large number of
5203 hosts and all are timing out (e.g. when the network connections are timing
5204 out). The value of the overall timeout can be changed by specifying an
5205 additional sub-option for "callout", called "maxwait". For example:
5207 verify = sender/callout=5s,maxwait=20s
5209 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5210 to the end before writing, but this should make it even safer).
5212 22. Exim was forgetting that it had advertised PIPELINING for the second and
5213 subsequent messages on an SMTP connection. It was also not resetting its
5214 memory on STARTTLS and an internal HELO.
5216 23. When Exim logs an SMTP synchronization error within a session, it now
5217 records whether PIPELINING has been advertised or not.
5219 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5220 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5221 rather than long int.
5223 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5225 26. Added the never_mail option to autoreply.
5231 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5232 crash if the getsockname() call failed; this can happen if a connection is
5233 closed very soon after it is established. The problem was simply in the
5234 order in which certain operations were done, causing Exim to try to write
5235 to the SMTP stream before it had set up the file descriptor. The bug has
5236 been fixed by making things happen in the correct order.
5242 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5243 before the connection was closed, thus losing the rejection response.
5245 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5246 some early Solaris releases, but causes trouble in current releases where
5247 socklen_t is defined.
5249 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5252 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5255 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5256 that releasing the top of it at the end releases what was used for sub-
5257 expansions (unless the block got too big). However, discard this block if
5258 the first thing is a variable or header, so that we can use its block when
5259 it is dynamic (useful for very large $message_headers, for example).
5261 6. Lookups now cache *every* query, not just the most recent. A new, separate
5262 store pool is used for this. It can be recovered when all lookup caches are
5263 flushed. Lookups now release memory at the end of their result strings.
5264 This has involved some general refactoring of the lookup sources.
5266 7. Some code has been added to the store_xxx() functions to reduce the amount
5267 of flapping under certain conditions.
5269 8. log_incoming_interface used to affect only the <= reception log lines. Now
5270 it causes the local interface and port to be added to several more SMTP log
5271 lines, for example "SMTP connection from", and rejection lines.
5273 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5275 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5277 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5278 could be overwritten at the end of the current message (or the start of a
5279 new message if it was set in a HELO ACL). The value is now preserved for
5280 the duration of the SMTP connection.
5282 12. If a transport had a headers_rewrite setting, and a matching header line
5283 contained an unqualified address, that address was qualified, even if it
5284 did not match any rewriting rules. The underlying bug was that the values
5285 of the flags that permit the existence of unqualified sender and recipient
5286 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5287 non-local messages, and by -bnq for local messages) were not being
5288 preserved with the message after it was received.
5290 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5291 "next input=" as part of the text comprising the host identity instead of
5292 the correct text. The code was using the same buffer for two different
5293 strings. However, depending on which order the printing function evaluated
5294 its arguments, the bug did not always show up. Under Linux, for example, my
5295 test suite worked just fine.
5297 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5298 doesn't work with some older versions of Perl. It has been changed to "my",
5299 which in any case is probably the better facility to use.
5301 15. A really picky compiler found some instances of statements for creating
5302 error messages that either had too many or two few arguments for the format
5305 16. The size of the buffer for calls to the DNS resolver has been increased
5306 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5307 for addresses that have a lot of PTR records. This alleviates a problem; it
5308 does not fully solve it.
5310 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5311 buffer now truncates the list and logs the incident, which is the same
5312 action as happens when Exim is looking up a host name and its aliases.
5313 Previously in this situation something unpredictable would happen;
5314 sometimes it was "internal error: store_reset failed".
5316 18. If a server dropped the connection unexpectedly when an Exim client was
5317 using GnuTLS and trying to read a response, the client delivery process
5318 crashed while trying to generate an error log message.
5320 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5321 string, for example:
5323 warn message = H1: something\nH2: something
5325 the text was added as a single header line from Exim's point of view
5326 though it ended up OK in the delivered message. However, searching for the
5327 second and subsequent header lines using $h_h2: did not work. This has been
5328 fixed. Similarly, if a system filter added multiple headers in this way,
5329 the routers could not see them.
5331 20. Expanded the error message when iplsearch is called with an invalid key to
5332 suggest using net-iplsearch in a host list.
5334 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5335 ACLs are no longer actually imposed (and a message to that effect is
5338 22. If a "gecos" field in a passwd entry contained escaped characters, in
5339 particular, if it contained a \" sequence, Exim got it wrong when building
5340 a From: or a Sender: header from that name. A second bug also caused
5341 incorrect handling when an unquoted " was present following a character
5342 that needed quoting.
5344 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5345 was not being matched caselessly.
5347 24. Arranged for all hyphens in the exim.8 source to be escaped with
5350 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5351 properly. Recipient callout cache records were still being keyed to include
5352 the sender, even when use_sender was set false. This led to far more
5353 callouts that were necessary. The sender is no longer included in the key
5354 when use_sender is false.
5356 26. Added "control = submission" modifier to ACLs.
5358 27. Added the ${base62d: operator to decode base 62 numbers.
5360 28. dnsdb lookups can now access SRV records.
5362 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5363 the configuration file.
5365 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5366 (-v) mode. This makes the output for a verbose queue run more intelligible.
5368 31. Added a use_postmaster feature to recipient callouts.
5370 32. Added the $body_zerocount variable, containing the number of binary zero
5371 bytes in the message body.
5373 33. The time of last modification of the "new" subdirectory is now used as the
5374 "mailbox time last read" when there is a quota error for a maildir
5377 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5379 35. Added +ignore_unknown as a special item in host lists.
5381 36. Code for decoding IPv6 addresses in host lists is now included, even if
5382 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5383 address was recognized as an IP address, but was then not correctly decoded
5384 into binary, causing unexpected and incorrect effects when compared with
5391 1. Very minor rewording of debugging text in manualroute to say "list of
5392 hosts" instead of "hostlist".
5394 2. If verify=header_syntax was set, and a header line with an unqualified
5395 address (no domain) and a large number of spaces between the end of the
5396 name and the colon was received, the reception process suffered a buffer
5397 overflow, and (when I tested it) crashed. This was caused by some obsolete
5398 code that should have been removed. The fix is to remove it!
5400 3. When running in the test harness, delay a bit after writing a bounce
5401 message to get a bit more predictability in the log output.
5403 4. Added a call to search_tidyup() just before forking a reception process. In
5404 theory, someone could use a lookup in the expansion of smtp_accept_max_
5405 per_host which, without the tidyup, could leave open a database connection.
5407 5. Added the variables $recipient_data and $sender_data which get set from a
5408 lookup success in an ACL "recipients" or "senders" condition, or a router
5409 "senders" option, similar to $domain_data and $local_part_data.
5411 6. Moved the writing of debug_print from before to after the "senders" test
5414 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5415 problems for message scanning, either using a data ACL, or using
5416 local_scan() because the Received: header was not generated till after they
5417 were called (in order to set the time as the time of reception completion).
5418 I have revised the way this works. The header is now generated after the
5419 body is received, but before the ACL or local_scan() are called. After they
5420 are run, the timestamp in the header is updated.
5426 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5427 before starting a queue runner without re-exec. This happened only when
5428 deliver_drop_privilege was set or when the Exim user was set to root. The
5429 effect of the bug was that timeouts during subsequent deliveries caused
5430 crashes instead of being properly handled. The handler is now left at its
5431 default (and expected) setting.
5433 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5434 message, again when deliver_drop_privilege is set or Exim is run as root.
5435 The bug described in (1) was not present in this case, but the tidying up
5436 of the other signals was missing. I have made the two cases consistent.
5438 3. The ignore_target_hosts setting on a manualroute router was being ignored
5439 for hosts that were looked up using the /MX notation.
5441 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5444 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5445 operated on the sender address. After changing the $sender_address to <>
5446 for the sender address verify, Exim was re-instated it as the original
5447 (before rewriting) address, but remembering that it had rewritten it, so it
5448 wasn't rewriting it again. This bug also had the effect of breaking the
5449 sender address verification caching when the sender address was rewritten.
5451 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5452 This has been changed so that if the ip literal address matches
5453 ignore_target_hosts, the router declines.
5455 7. Added expansion conditions match_domain, match_address, and match_local_
5456 part (NOT match_host).
5458 8. The placeholder for the Received: header didn't have a length field set.
5460 9. Added code to Exim itself and to exim_lock to test for a specific race
5461 condition that could lead to file corruption when using MBX delivery. The
5462 issue is with the lockfile that is created in /tmp. If this file is removed
5463 after a process has opened it but before that process has acquired a lock,
5464 there is the potential for a second process to recreate the file and also
5465 acquire a lock. This could lead to two Exim processes writing to the file
5466 at the same time. The added code performs the same test as UW imapd; it
5467 checks after acquiring the lock that its file descriptor still refers to
5468 the same named file.
5470 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5471 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5474 11. Added the smtp_active_hostname option. If used, this will typically be made
5475 to depend on the incoming interface address. Because $interface_address is
5476 not set up until the daemon has forked a reception process, error responses
5477 that can happen earlier (such as "too many connections") no longer contain
5480 12. If an expansion in a condition on a "warn" statement fails because a lookup
5481 defers, the "warn" statement is abandoned, and the next ACL statement is
5482 processed. Previously this caused the whole ACL to be aborted.
5484 13. Added the iplsearch lookup type.
5486 14. Added ident_timeout as a log selector.
5488 15. Added tls_certificate_verified as a log selector.
5490 16. Added a global option tls_require_ciphers (compare the smtp transport
5491 option of the same name). This controls incoming TLS connections.
5493 17. I finally figured out how to make tls_require_ciphers do a similar thing
5494 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5495 before starting the TLS session.
5497 18. Tabs are now shown as \t in -bP output.
5499 19. If the log selector return_path_on_delivery was set, Exim crashed when
5500 bouncing a message because it had too many Received: header lines.
5502 20. If two routers both had headers_remove settings, and the first one included
5503 a superfluous trailing colon, the final name in the first list and the
5504 first name in the second list were incorrectly joined into one item (with a
5505 colon in the middle).
5511 1. Added -C and -D options to the exinext utility, mainly to make it easier
5512 to include in the automated testing, but these could be helpful when
5513 multiple configurations are in use.
5515 2. The exinext utility was not formatting the output nicely when there was
5516 an alternate port involved in the retry record key, nor when there was a
5517 message id as well (for retries that were specific to a specific message
5518 and a specific host). It was also confused by IPv6 addresses, because of
5519 the additional colons they contain. I have fixed the IPv4 problem, and
5520 patched it up to do a reasonable job for IPv6.
5522 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5523 delivery, the log line now contains "pipelined" if PIPELINING was used.
5525 4. An SMTP transport process used to panic and die if the bind() call to set
5526 an explicit outgoing interface failed. This has been changed; it is now
5527 treated in the same way as a connect() failure.
5529 5. A reference to $sender_host_name in the part of a conditional expansion
5530 that was being skipped was still causing a DNS lookup. This no longer
5533 6. The def: expansion condition was not recognizing references to header lines
5534 that used bh_ and bheader_.
5536 7. Added the _cache feature to named lists.
5538 8. The code for checking quota_filecount in the appendfile transport was
5539 allowing one more file than it should have been.
5541 9. For compatibility with Sendmail, the command line option
5549 and sets the incoming protocol and host name (for trusted callers). The
5550 host name and its colon can be omitted when only the protocol is to be set.
5551 Note the Exim already has two private options, -pd and -ps, that refer to
5552 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5553 "s", but I don't think that's a major issue.
5555 10. A number of refactoring changes to the code, none of which should affect
5558 (a) The number of logging options was getting close to filling up the
5559 32-bit word that was used as a bit map. I have split them into two classes:
5560 those that are passed in the argument to log_write(), and those that are
5561 only ever tested independently outside of that function. These are now in
5562 separate 32-bit words, so there is plenty of room for expansion again.
5563 There is no change in the user interface or the logging behaviour.
5565 (b) When building, for example, log lines, the code previously used a
5566 macro that called string_cat() twice, in order to add two strings. This is
5567 not really sufficiently general. Furthermore, there was one instance where
5568 it was actually wrong because one of the argument was used twice, and in
5569 one call a function was used. (As it happened, calling the function twice
5570 did not affect the overall behaviour.) The macro has been replaced by a
5571 function that can join an arbitrary number of extra strings onto a growing
5574 (c) The code for expansion conditions now uses a table and a binary chop
5575 instead of a serial search (which was left over from when there were very
5576 few conditions). Also, it now recognizes conditions like "pam" even when
5577 the relevant support is not compiled in: a suitably worded error message is
5578 given if an attempt is made to use such a condition.
5580 11. Added ${time_interval:xxxxx}.
5582 12. A bug was causing one of the ddress fields not to be passed back correctly
5583 from remote delivery subprocesses. The field in question was not being
5584 subsequently used, so this caused to problems in practice.
5586 13. Added new log selectors queue_time and deliver_time.
5588 14. Might have fixed a bug in maildirsizefile handling that threw up
5589 "unexpected character" debug warnings, and recalculated the data
5590 unnecessarily. In any case, I expanded the warning message to give more
5593 15. Added the message "Restricted characters in address" to the statements in
5594 the default ACL that block characters like @ and % in local parts.
5596 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5597 Three changes have been made:
5599 (a) There was a serious bug; a negative response to MAIL caused the whole
5600 recipient domain to be cached as invalid, thereby blocking all messages
5601 to all local parts at the same domain, from all senders. This bug has
5602 been fixed. The domain is no longer cached after a negative response to
5603 MAIL if the sender used is not empty.
5605 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5608 (c) A new callout option, "use_sender" has been added for people who want
5609 the modified behaviour.
5615 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5618 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5619 indeed breaks things for older releases.
5621 3. Added additional logging to the case where there is a problem reading data
5622 from a filter that is running in a subprocess using a pipe, in order to
5623 try to track down a specific problem.
5625 4. Testing facility fudge: when running in the test harness and attempting
5626 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5627 getting "No route to host". Convert this to a timeout.
5629 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5632 6. Some OS don't have socklen_t but use size_t instead. This affects the
5633 fifth argument of getsockopt() amongst other things. This is now
5634 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5635 can be set for individual OS. I have set it for SunOS5, OSF1, and
5636 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5637 some earlier ones do not.
5639 7. Change 4.30/15 was not doing the test caselessly.
5641 8. The standard form for an IPv6 address literal was being rejected by address
5642 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5643 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5644 this, as well as the form without the "IPv6" on the front (but only when
5645 address literals are enabled, of course).
5647 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5649 10. Exim crashed if a message with an empty sender address specified by -f
5650 encountered a router with an errors_to setting. This could be provoked only
5651 by a command such as
5655 where an empty string was supplied; "<>" did not hit this bug.
5657 11. Installed PCRE release 4.5.
5659 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5660 remained set. It is now erased.
5662 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5663 times from message ids (which are base 36 rather than the normal 62).
5665 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5666 were being counted as actual protocol errors, and logged if the log
5667 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5668 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5669 and DATA following a set of rejected RCPTs do not count as protocol errors.
5670 In other words, Exim assumes they were pipelined, though this may not
5671 actually be the case. Of course, in all cases the client gets an
5672 appropriate error code.
5674 15. If a lookup fails in an ACL condition, a message about the failure may
5675 be available; it is used if testing the ACL cannot continue, because most
5676 such messages specify what the cause of the deferral is. However, some
5677 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5678 that caused an old message to be retained and used if a later statement
5679 caused a defer, replacing the real cause of the deferral.
5681 16. If an IP address had so many PTR records that the DNS lookup buffer
5682 was not large enough to hold them, Exim could crash while trying to process
5683 the truncated data. It now detects and logs this case.
5685 17. Further to 4.21/58, another change has been made: if (and only if) the
5686 first line of a message (the first header line) ends with CRLF, a bare LF
5687 in a subsequent header line has a space inserted after it, so as not to
5688 terminate the header.
5690 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5691 unnecessarily, used atoi() instead of strtol(), and didn't check the
5692 termination when getting file sizes from file names by regex.
5694 19. Completely re-implemented the support for maildirsize files, in the light
5695 of a number of problems with the previous contributed implementation
5696 (4.30/29). In particular:
5698 . If the quota is zero, the maildirsize file is maintained, but no quota is
5701 . If the maildir directory does not exist, it is created before any attempt
5702 to write a maildirsize file.
5704 . The quota value in the file is just a cache; if the quota is changed in
5705 the transport, the new value overrides.
5707 . A regular expression is available for excluding directories from the
5710 20. The autoreply transport checks the characters in options that define the
5711 message's headers; it allows continued headers, but it was checking with
5712 isspace() after an embedded newline instead of explicitly looking for a
5715 21. If all the "regular" hosts to which an address was routed had passed their
5716 expiry times, and had not reached their retry times, the address was
5717 bounced, even if fallback hosts were defined. Now Exim should go on to try
5720 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5721 equivalent code in the SMTP transport. Some hosts send humungous responses
5722 to HELO/EHLO, more than 1024 it seems.
5724 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5725 gives compiler warnings in some environments. I've now done it "properly",
5728 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5729 (because of problems with the built-in one) was declared to return uschar *
5730 instead of char *, causing compiler failure.
5732 25. Fixed a file descriptor leak when processing alias/forward files.
5734 26. Fixed a minor format string issue in dbfn.c.
5736 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5738 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5739 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5742 29. When a sender address is verified, it is cached, to save repeating the test
5743 when there is more than one recipient in a message. However, when the
5744 verification involves a callout, it is possible for different callout
5745 options to be set for different recipients. It is too complicated to keep
5746 track of this in the cache, so now Exim always runs a verification when a
5747 callout is required, relying on the callout cache for the optimization.
5748 The overhead is duplication of the address routing, but this should not be
5751 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5752 to be verified with callout=postmaster, and the main callout worked but the
5753 postmaster check failed, the verification correctly failed. However, if a
5754 subsequent RCPT command asked for sender verification *without* the
5755 postmaster check, incorrect caching caused this verification also to fail,
5758 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5759 it was not caching the DNS options (qualify_single, search_parents) that
5760 were used when the lookup failed. A subsequent lookup with different
5761 options therefore always gave the same answer, though there were cases
5762 where it should not have. (Example: a "domains = !$mx_any" option on a
5763 dnslookup router: the "domains" option is always processed without any
5764 widening, but the router might have qualify_single set.) Now Exim uses the
5765 cached value only when the same options are set.
5767 32. Added John Jetmore's "exipick" utility to the distribution.
5769 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5770 than a timeout (e.g. a certificate is required, and is not provided), an
5771 Exim server now closes the connection immediately. Previously it waited for
5772 the client to close - but if the client is SSL, it seems that they each
5773 wait for each other, leading to a delay before one of them times out.
5775 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5776 maintained 0.8.x compatibility because I don't think many are using it, and
5777 it is clearly obsolete.
5779 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5782 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5783 client certificate was expired. A simple patch fixes this, though I don't
5784 understand the full logic of why the verify callback is called multiple
5787 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5788 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5789 which causes problems with some clients (such as the Certicom SSL Plus
5790 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5791 disables the coutermeasure allowing Eudora to connect."
5793 38. Exim was not checking that a write() to a log file succeeded. This could
5794 lead to Bad Things if a log got too big, in particular if it hit a file
5795 size limit. Exim now panics and dies if it cannot write to a log file, just
5796 as it does if it cannot open a log file.
5798 39. Modified OS/Makefile-Linux so that it now contains
5800 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5802 The two -D definitions ensure that Exim is compiled with large file
5803 support, which makes it possible to handle log files that are bigger than
5806 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5807 instance) a domain was checked against a named list that involved a lookup,
5808 causing $domain_data to be set, then another domain was checked against the
5809 same list, then the first domain was re-checked, the value of $domain_data
5810 after the final check could be wrong. In particular, if the second check
5811 failed, it could be set empty. This bug probably also applied to
5814 41. The strip_trailing_dot option was not being applied to the address given
5815 with the -f command-line option.
5817 42. The code for reading a message's header from the spool was incrementing
5818 $received_count, but never initializing it. This meant that the value was
5819 incorrect (doubled) while delivering a message in the same process in which
5820 it was received. In the most common configuration of Exim, this never
5821 happens - a fresh exec is done - but it can happen when
5822 deliver_drop_privilege is set.
5824 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5825 it now includes up to 150 characters of the unexpected data in the log
5828 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5829 and building data strings. The size of both of these buffers was 10 000
5830 bytes - far larger than anybody would *ever* want, thought I. Needless to
5831 say, somebody hit the limit. I have increased the maximum line length to
5832 20 000 and the maximum data length of concatenated lines to 100 000. I have
5833 also fixed two bugs, because there was no checking on these buffers. Tsk,
5834 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5835 buffer is too small.
5837 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5838 lsearch lookups. Now it does.
5840 46. When parsing a route_list item in a manualroute router, a fixed-length
5841 buffer was used for the list of hosts. I made this 1024 bytes long,
5842 thinking that nobody would ever have a list of hosts that long. Wrong.
5843 Somebody had a whole pile of complicated expansion conditions, and the
5844 string was silently truncated, leading to an expansion error. It turns out
5845 that it is easier to change to an unlimited length (owing to other changes
5846 that have happened since this code was originally written) than to build
5847 structure for giving a limitation error. The length of the item that
5848 expands into the list of hosts is now unlimited.
5850 47. The lsearch lookup could not handle data where the length of text line was
5851 more than 4095 characters. Such lines were truncated, leading to shortened
5852 data being returned. It should now handle lines of any length.
5854 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5855 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5858 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5859 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5860 no longer happens in comment lines. A list of replacements is now placed
5861 at the head of all of the source files, except those whose only change is
5862 to replace PERL_COMMAND in the very first #! line.
5864 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5865 messages on the queue, with a bottom-up merge sort, using code contributed
5866 by Michael Haardt. This should make operations like -bp somewhat faster on
5867 large queues. It won't affect queue runners, except when queue_run_in_order
5870 51. Installed eximstats 1.31 in the distribution.
5872 52. Added support for SRV lookups to the dnslookup router.
5874 53. If an ACL referred to $message_body or $message_body_end, the value was not
5875 reset for any messages that followed in the same SMTP session.
5877 54. The store-handling optimization for building very long strings was not
5878 differentiating between the different store pools. I don't think this
5879 actually made any difference in practice, but I've tidied it.
5881 55. While running the routers to verify a sender address, $sender_address
5882 was still set to the sender address. This is wrong, because when routing to
5883 send a bounce to the sender, it would be empty. Therefore, I have changed
5884 it so that, while verifying a sender address, $sender_address is set to <>.
5885 (There is no change to what happens when verifying a recipient address.)
5887 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5888 target A or AAAA records (if not already returned) without resetting the
5889 qualify_single or search_parents options of the DNS resolver. These are
5890 inappropriate in this case because the targets of MX and SRV records must
5891 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5892 target that, when qualified, matched something in the local domain. These
5893 two options are now turned off when doing these lookups.
5895 57. It seems that at least some releases of Reiserfs (which does not have the
5896 concept of a fixed number of inodes) returns zero and not -1 for the
5897 number of available inodes. This interacted badly with check_spool_inodes,
5898 which assumed that -1 was the "no such thing" setting. What I have done is
5899 to check that the total number of inodes is greater than zero before doing
5900 the test of how many are available.
5902 58. When a "warn" ACL statement has a log_message modifier, the message is
5903 remembered, and not repeated. This is to avoid a lot of repetition when a
5904 message has many recipients that cause the same warning to be written.
5905 However, Exim was preserving the list of already written lines for an
5906 entire SMTP session, which doesn't seem right. The memory is now reset if a
5907 new message is started.
5909 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5910 debugging output unless log_rewrite was also set.
5912 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5913 of (int)(handle) when we know that handle contains (void *)(-1).
5915 61. The Exim daemon panic-logs an error return when it closes the incoming
5916 connection. However "connection reset by peer" seems to be common, and
5917 isn't really an error worthy of noting specially, so that particular error
5920 62. When Exim is trying to find all the local interfaces, it used to panic and
5921 die if the ioctl to get the interface flags failed. However, it seems that
5922 on at least one OS (Solaris 9) it is possible to have an interface that is
5923 included in the list of interfaces, but for which you get a failure error
5924 for this call. This happens when the interface is not "plumbed" into a
5925 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5926 failure of the "get flags" call assumes that the interface is down.
5928 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5929 makes life easier for people who are doing arithmetic on fields extracted
5930 from dates, where you often get leading zeros that should not be
5931 interpreted as octal.
5933 64. Added qualify_domain to the redirect router, to override the global
5936 65. If a pathologically long header line contained very many addresses (the
5937 report of this problem mentioned 10 000) and each of them was rewritten,
5938 Exim could use up a very large amount of memory. (It kept on making new
5939 copies of the header line as it rewrote, and never released the old ones.)
5940 At the expense of a bit more processing, the header rewriting function has
5941 been changed so that it no longer eats memory in this way.
5943 66. The generation of the Received: header has been moved from the time that a
5944 message starts to be received, to the time that it finishes. The timestamp
5945 in the Received: header should now be very close to that of the <= log
5946 line. There are two side-effects of this change:
5948 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5949 logged header lines no longer include the local Received: line, because
5950 it has not yet been created. The same applies to a copy of the message
5951 that is returned to a non-SMTP sender when a message is rejected.
5953 (b) When a filter file is tested using -bf, no additional Received: header
5954 is added to the test message. After some thought, I decided that this
5957 This change does not affect the value of $received_for. It is still set
5958 after address rewriting, but before local_scan() is called.
5960 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5962 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5963 gave an unhelpful panic error message, and a defer error. I have managed to
5964 change this behaviour so that it now rejects any supplied certificate,
5965 which seems right, as the list of acceptable certificates is empty.
5967 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5968 gave an unhelpful defer error. I have not managed to make this reject any
5969 supplied certificates, but the error message it gives is "no certificate
5970 supplied", which is not helpful.
5972 70. exigrep's output now also includes lines that are not associated with any
5973 message, but which match the given pattern. Implemented by a patch from
5974 Martin Sluka, which also tidied up the Perl a bit.
5976 71. Recipient callout verification, like sender verification, was using <> in
5977 the MAIL FROM command. This isn't really the right thing, since the actual
5978 sender may affect whether the remote host accepts the recipient or not. I
5979 have changed it to use the actual sender in the callout; this means that
5980 the cache record is now keyed on a recipient/sender pair, not just the
5981 recipient address. There doesn't seem to be a real danger of callout loops,
5982 since a callout by the remote host to check the sender would use <>.
5983 [SEE ABOVE: changed after hitting problems.]
5985 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5986 temporary errors. However, in the case of such a code being given after
5987 the end of a data transmission (i.e. after ".") Exim was failing to write
5988 a retry record for the message. (Yes, there was some broken host that was
5989 actually sending 8xx at this point.)
5991 73. An unknown lookup type in a host list could cause Exim to panic-die when
5992 the list was checked. (An example that provoked this was putting <; in the
5993 middle of a list instead of at the start.) If this happened during a DATA
5994 ACL check, a -D file could be left lying around. This kind of configuration
5995 error no longer causes Exim to die; instead it causes a defer error. The
5996 incident is still logged to the main and panic logs.
5998 74. Buglet left over from Exim 3 conversion. The message "too many messages
5999 in one connection" was written to the rejectlog but not the mainlog, except
6000 when address rewriting (yes!) was being logged.
6002 75. Added write_rejectlog option.
6004 76. When a system filter was run not as root (that is, when system_filter_user
6005 was set), the values of the $n variables were not being returned to the
6006 main process; thus, they were not subsequently available in the $sn
6009 77. Added +return_path_on_delivery log selector.
6011 78. A connection timeout was being treated differently from recipients deferred
6012 when testing hosts_max_try with a message that was older than the host's
6013 retry timeout. (The host should not be counted, thus allowing all hosts to
6014 be tried at least once before bouncing.) This may have been the cause of an
6015 occasionally reported bug whereby a message would remain on the queue
6016 longer than the retry timeout, but would be bounced if a delivery was
6017 forced. I say "may" because I never totally pinned down the problem;
6018 setting up timeout/retry tests is difficult. See also the next item.
6020 79. The ultimate address timeout was not being applied to errors that involved
6021 a combination of host plus message (for example, a timeout on a MAIL
6022 command). When an address resolved to a number of possible hosts, and they
6023 were not all tried for each delivery (e.g. because of hosts_max_try), a
6024 message could remain on the queue longer than the retry timeout.
6026 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6029 81. Fixed an obscure SMTP outgoing bug which required at least the following
6030 conditions: (a) there was another message waiting for the same server;
6031 (b) the server returned 5xx to all RCPT commands in the first message so
6032 that the message was not completed; (c) the server dropped the connection
6033 or gave a negative response to the RSET that Exim sends to abort the
6034 transaction. The observed case was a dropped connection after DATA that had
6035 been sent in pipelining mode. That is, the server had advertised PIPELINING
6036 but was not implementing it correctly. The effect of the bug was incorrect
6037 behaviour, such as trying another host, and this could lead to a crash.
6043 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6044 and daemon.c were passed as pointers to ints; they should have been
6045 pointers to socklen_t variables (which are typically unsigned ints).
6047 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6050 3. Fixed a really odd bug that affected only the testing scheme; patching a
6051 certain fixed string in the binary changed the value of another string that
6052 happened to be identical to the end of the original first string.
6054 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6055 name", it returns that address as the IP address. On some operating
6056 systems (e.g. Solaris), it also passes back the IP address string as the
6057 "host name". However, on others (e.g. Linux), it passes back an empty
6058 string. Exim wasn't checking for this, and was changing the host name to an
6059 empty string, assuming it had been canonicalized.
6061 5. Although rare, it is permitted to have more than one PTR record for a given
6062 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6063 all the names associated with an address, because they do in Solaris.
6064 However, it seems that they do not in Linux for data that comes from the
6065 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6066 I found this out when I moved to a new Linux workstation and tried to run
6067 the Exim test suite.
6069 To get round this problem I have changed the code so that it now does its
6070 own call to the DNS to look up PTR records when searching for a host name.
6071 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6072 addresses that are only in /etc/hosts are still found.
6074 This behaviour is, however, controlled by an option called host_lookup_
6075 order, which defaults to "bydns:byaddr". If people want to use the other
6076 order, or indeed, just use one or the other means of lookup, they can
6077 specify it in this variable.
6079 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6080 some operating systems, this comes back from gethostbyaddr() as an empty
6081 string, and this is what Exim used to test for. However, it seems that in
6082 other systems, "." is yielded. Exim now tests for this case too.
6084 7. The values of check_spool_space and check_log_space are now held internally
6085 as a number of kilobytes instead of an absolute number of bytes. If a
6086 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6087 kilobyte. This means that much larger values can be stored.
6089 8. Exim monitor: an attempt to get the action menu when not actually pointing
6090 at a message produces an empty menu entitled "No message selected". This
6091 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6092 no entries in it ("Shell widget menu has zero width and/or height"). So I
6093 have added a single, blank menu entry in this case.
6095 9. Added ${quote_local_part.
6097 10. MIME decoding is now applied to the contents of Subject: header lines when
6100 11. Now that a reference to $sender_host_address automatically causes a reverse
6101 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6102 host lookup before query-style lookups in lists that might use this
6103 variable. This has therefore been abolished, and the "net-" prefix is no
6104 longer necessary for query-style lookups.
6106 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6107 have been a typo for LFLAGS, so it has been changed.
6109 13. The install script calls Exim with "-C /dev/null" in order to find the
6110 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6111 to be output. However, since Exim outputs its version number before the
6112 error, it didn't break the script. It just looked ugly. I fixed this by
6113 always allowing "-C /dev/null" if the caller is root.
6115 14. Ignore overlarge ACL variable number when reading spool file - insurance
6116 against a later release with more variables having written the file.
6118 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6119 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6120 this, as well as the form without the "IPv6" on the front.
6122 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6123 OS/Makefile-Darwin file.
6125 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6126 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6128 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6129 variable. However, some OS use ints for inodes. I've added cast to long int
6130 to get rid of the compiler warning.
6132 19. I had forgotten to lock out "/../" in configuration file names when
6133 ALT_CONFIG_PREFIX was set.
6135 20. Routers used for verification do not need to specify transports. However,
6136 if such a router generated a host list, and callout was configured, Exim
6137 crashed, because it could not find a port number from the (non-existent)
6138 transport. It now assumes port 25 in this circumstance.
6140 21. Added the -t option to exigrep.
6142 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6143 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6144 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6145 them is set, LOOKUP_LSEARCH is forced.
6147 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6149 24. Added sender and host information to the "rejected by local_scan()" log
6150 line; previously there was no indication of these.
6152 25. Added .include_if_exists.
6154 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6155 Linux. It turns out that not all file systems support this. Apparently some
6156 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6157 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6158 not supported on the file descriptor, is now ignored when Exim is trying to
6159 sync a directory. This applies only to Linux.
6161 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6163 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6164 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6167 29. Added support for maildirsize files from supplied patch (modified a bit).
6169 30. The use of :fail: followed by an empty string could lead Exim to respond to
6170 sender verification failures with (e.g.):
6172 550 Verification failed for <xxx>
6173 550 Sender verify failed
6175 where the first response line was missing the '-' that indicates it is not
6176 the final line of the response.
6178 31. The loop for finding the name of the user that called Exim had a hardwired
6179 limit of 10; it now uses the value of finduser_retries, which is used for
6180 all other user lookups.
6182 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6185 33. Exim was neglecting to zero errno before one call of strtol() when
6186 expanding a string and expecting an integer value. On some systems this
6187 resulted in spurious "integer overflow" errors. Also, it was casting the
6188 result into an int without checking.
6190 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6191 did not work. The code looks as if it has *never* worked, though it appears
6192 to have been documented since at least release 1.62. I have made it work.
6194 35. The "timeout_DNS" error in retry rules, also documented since at least
6195 1.62, also never worked. As it isn't clear exactly what this means, and
6196 clearly it isn't a major issue, I have abolished the feature by treating it
6197 as "timeout", and writing a warning to the main and panic logs.
6199 36. The display of retry rules for -brt wasn't always showing the error code
6202 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6203 timeout_connect_A, timeout_connect_MX.
6205 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6206 to the empty sender.
6208 39. The daemon was not analysing the content of -oX till after it had closed
6209 stderr and disconnected from the controlling terminal. This meant that any
6210 syntax errors were only noted on the panic log, and the return code from
6211 the command was 0. By re-arranging the code a little, I've made the
6212 decoding happen first, so such errors now appear on stderr, and the return
6213 code is 1. However, the actual setting up of the sockets still happens in
6214 the disconnected process, so errors there are still only recorded on the
6217 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6218 connections (as happens on some IP stacks) was logged at start up time as
6219 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6220 it from "IPv6 and IPv4", which means that two separate sockets are being
6223 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6224 says whether AF_INET or AF_INET6 was passed as an argument.
6226 42. Exiwhat output was messed up when time zones were included in log
6229 43. Exiwhat now gives more information about the daemon's listening ports,
6230 and whether -tls-on-connect was used.
6232 44. The "port" option of the smtp transport is now expanded.
6234 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6235 silently ignored. Now an error message is written to the main and panic
6238 46. There's a new ACL modifier called "logwrite" which writes to a log file
6239 as soon as it is encountered.
6241 47. Added $local_user_uid and $local_user_gid at routing time.
6243 48. Exim crashed when trying to verify a sender address that was being
6246 49. Exim was recognizing only a space character after ".include". It now also
6247 recognizes a tab character.
6249 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6250 extracting the relevant information from the specification. The man page no
6251 longer contains scrambled data for the -d option, and I've added a section
6252 at the front about calling Exim under different names.
6254 51. Added "extra_headers" argument to the "mail" command in filter files.
6256 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6259 53. Installed eximstats 1.29.
6261 54. Added transport_filter_timeout as a generic transport option.
6263 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6264 Cc: header lines. This was required by RFC 822, but it not required by RFC
6267 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6268 incoming messages that did not have them. Now it does so only if the
6269 message originates locally, that is, if there is no associated remote host
6270 address. When Resent- header lines are present, this applies to the Resent-
6271 lines rather than the non-Resent- lines.
6273 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6274 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6276 58. Messages for configuration errors now include the name of the main
6277 configuration file - useful now that there may be more than one file in a
6278 list (.included file names were always shown).
6280 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6281 for those rare installations that do not start the daemon as root or run it
6282 setuid root. I've cut out the call to initgroups() if the daemon is not
6285 60. The Exim user and group can now be bound into the binary as text strings
6286 that are looked up at the start of Exim's processing.
6288 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6290 62. Added $mailstore_basename variable.
6292 63. Installed patch to sieve.c from Michael Haardt.
6294 64. When Exim failed to open the panic log after failing to open the main log,
6295 the original message it was trying to log was written to stderr and debug
6296 output, but if they were not available (the usual case in production), it
6297 was lost. Now it is written to syslog before the two lines that record the
6298 failures to open the logs.
6300 65. Users' Exim filters run in subprocesses under the user's uid. It is
6301 possible for a "deliver" command or an alias in a "personal" command to
6302 provoke an address rewrite. If logging of address rewriting is configured,
6303 this fails because the process is not running as root or exim. There may be
6304 a better way of dealing with this, but for the moment (because 4.30 needs
6305 to be released), I have disabled address rewrite logging when running a
6306 filter in a non-root, non-exim process.
6312 1. The buildconfig auxiliary program wasn't quoting the value set for
6313 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6314 not defined. This bug was masked in 4.22 by the effect that was fixed in
6317 2. Some messages that were rejected after a message id was allocated were
6318 shown as "incomplete" by exigrep. It no longer does this for messages that
6319 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6321 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6322 have allow_domain_literals set, the ID did not get logged in the <= line.
6323 Domain literals are now always recognized in Message-ID: header lines.
6325 4. The first argument for a ${extract expansion item is the key name or field
6326 number. Leading and trailing spaces in this item were not being ignored,
6327 causing some misleading effects.
6329 5. When deliver_drop_privilege was set, single queue runner processes started
6330 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6331 same command in the process it spins off) were not dropping privilege.
6333 6. When the daemon running as "exim" started a queue runner, it always
6334 re-executed Exim in the spun-off process. This is a waste of effort when
6335 deliver_drop_privilege is set. The new process now just calls the
6336 queue-runner function directly.
6342 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6345 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6346 ignored. Though the use of -oP was forcing the writing of a pid file, it
6347 was always written to the default place.
6349 3. If the message "no IP address found for host xxxx" is generated during
6350 incoming verification, it is now followed by identification of the incoming
6351 connection (so you can more easily find what provoked it).
6353 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6355 5. Added some features to "harden" Exim a bit more against certain attacks:
6357 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6358 be put in Local/Makefile. This is like the never_users runtime option,
6359 but it cannot be overridden. The default setting is "root".
6361 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6362 prefix string with which any file named in a -C command line option
6365 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6366 is retained for -C and -D only if the caller of Exim is root. Without
6367 it, the exim user may also use -C and -D and retain privilege.
6369 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6370 command line option is disabled.
6372 6. Macro names set by the -D option must start with an upper case letter, just
6373 like macro names defined in the configuration file.
6375 7. Added "dereference=" facility to LDAP.
6377 8. Two instances of the typo "uknown" in the source files are fixed.
6379 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6380 the Configure-Makefile script screwed up while processing it.
6382 10. Incorporated PCRE 4.4.
6384 11. The SMTP synchronization check was not operating right at the start of an
6385 SMTP session. For example, it could not catch a HELO sent before the client
6386 waited for the greeting. There is now a check for outstanding input at the
6387 point when the greeting is written. Because of the duplex, asynchronous
6388 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6389 way, but not yet received, when the check is performed.
6391 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6392 on TCP/IP sockets, because this apparently causes some broken clients to
6395 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6396 unchanged) from the Cygwin maintainer.
6398 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6399 was supported instead of testing for mbx. Effectively a typo.
6401 15. The spa authenticator server code was not checking that the input it
6402 received was valid base64.
6404 16. The debug output line for the "set" modifier in ACLs was not showing the
6405 name of the variable that was being set.
6407 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6409 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6410 lookup on its own. Something else was needed to trigger the lookup. For
6411 example, a match in host_lookup or the need for a host name in a host list.
6412 Now, if $sender_host_name is referenced and the host name has not yet been
6413 looked up, a lookup is performed. If the lookup fails, the variable remains
6414 empty, and $host_lookup_failed is set to "1".
6416 19. Added "eqi" as a case-independent comparison operator.
6418 20. The saslauthd authentication condition could segfault if neither service
6419 nor realm was specified.
6421 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6422 error message that was logged was misleading, and incoming SMTP
6423 connections were dropped. The message is now more accurate, and temporary
6424 errors are given to SMTP connections.
6426 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6427 (except RSET) with a 421 error, until QUIT is received. However, it was
6428 failing to send a response to QUIT.
6430 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6431 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6432 ACL is now run after the helo_try_verify_hosts code.
6434 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6435 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6436 case-independent, so other case variants are also recognized). Apparently
6437 some systems use these upper case variants.
6439 25. If more than two messages were waiting for the same host, and a transport
6440 filter was specified for the transport, Exim sent two messages over the
6441 same TCP/IP connection, and then failed with "socket operation on non-
6442 socket" when it tried to send the third.
6444 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6446 27. The extern definition of crypt16() in expand.c was not being excluded when
6447 the OS had its own crypt16() function.
6449 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6450 as a preferred synonym for return_size_limit, both as an option and as an
6453 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6455 30. Changed the default configuration ACL to relax the local part checking rule
6456 for addresses that are not in any local domains. For these addresses,
6457 slashes and pipe symbols are allowed within local parts, but the sequence
6458 /../ is explicitly forbidden.
6460 31. SPA server authentication was not clearing the challenge buffer before
6463 32. log_message in a "warn" ACL statement was writing to the reject log as
6464 well as to the main log, which contradicts the documentation and doesn't
6465 seem right (because no rejection is happening). So I have stopped it.
6467 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6468 However, I am unable to do any testing of this.
6470 34. Fixed an infelicity in the appendfile transport. When checking directories
6471 for a mailbox, to see if any needed to be created, it was accidentally
6472 using path names with one or more superfluous leading slashes; tracing
6473 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6475 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6476 contents are added to the log line that is written for every discarded
6477 recipient. (Previously a log_message setting was ignored.)
6479 36. The ${quote: operator now quotes the string if it is empty.
6481 37. The install script runs exim in order to find its version number. If for
6482 some reason other than non-existence or emptiness, which it checks, it
6483 could not run './exim', it was installing it with an empty version number,
6484 i.e. as "exim-". This error state is now caught, and the installation is
6487 38. An argument was missing from the function that creates an error message
6488 when Exim fails to connect to the socket for saslauthd authentication.
6489 This could cause Exim to crash, or give a corrupted message.
6491 39. Added isip, isip4, and isip6 to ${if conditions.
6493 40. The ACL variables $acl_xx are now saved with the message, and can be
6494 accessed later in routers, transports, and filters.
6496 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6497 strings in the file are not string-expanded.
6499 42. If a MAIL command specified a SIZE value that was too large to fit into an
6500 int variable, the check against message_size_limit failed. Such values are
6501 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6502 day this will have to be increased, but I don't think I want to be around
6503 when emails are that large.
6510 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6511 iconv() is not standard in FreeBSD.
6513 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6514 IPv6 enabled. The observed symptom was a segmentation fault on return from
6515 the function os_common_find_running_interfaces() in src/os.c.
6517 3. In the check_special_case() function in daemon.c I had used "errno" as an
6518 argument name, which causes warnings on some systems. This was basically a
6519 typo, since it was named "eno" in the comments!
6521 4. The code that waits for the clock to tick (at a resolution of some fraction
6522 of a second) so as to ensure message-id uniqueness was always waiting for
6523 at least one whole tick, when it could have waited for less. [This is
6524 almost certainly not relevant at current processor speeds, where it is
6525 unlikely to ever wait at all. But we try to future-proof.]
6527 5. The function that sleeps for a time interval that includes fractions of a
6528 second contained a race. It did not block SIGALRM between setting the
6529 timer, and suspending (a couple of lines later). If the interval was short
6530 and the sigsuspend() was delayed until after it had expired, the suspension
6531 never ended. On busy systems this could lead to processes getting stuck for
6534 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6535 process, before it forks any delivery processes. The open lookup caching
6536 mechanism meant that the open file or database connection was passed into
6537 the delivery process. The problem was that delivery processes always tidy
6538 up cached lookup data. This could cause a problem for the next delivery
6539 process started by the queue runner, because the external queue runner
6540 process does not know about the closure. So the next delivery process
6541 still has data in the lookup cache. In the case of a file lookup, there was
6542 no problem because closing a file descriptor in a subprocess doesn't affect
6543 the parent. However, if the lookup was caching a connection to a database,
6544 the connection was closed, and the second delivery process was likely to
6545 see errors such as "PGSQL: query failed: server closed the connection
6546 unexpectedly". The problem has been fixed by closing all cached lookups
6547 in a queue runner before running a delivery process.
6549 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6550 seem to have the "const" qualifier which it has on other OS. I've
6553 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6554 *of the same type* (client or server) with the same public name that an
6555 error should be diagnosed.
6557 9. When Exim looked up a host name for an IP address, but failed to find the
6558 original IP address when looking up the host name (a safety check), it
6559 output the message "<ip address> does not match any IP for NULL", which was
6560 confusing, to say the least. The bug was that the host name should have
6561 appeared instead of "NULL".
6563 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6564 user that is built into the binary, and the -C or -D options is used, root
6565 privilege is dropped before the configuration file is read. In addition,
6566 logging is switched to stderr instead of the normal log files. If the
6567 configuration then re-defines the Exim user, the unprivileged environment
6568 is probably not what is expected, so Exim logs a panic warning message (but
6571 However, if deliver_drop_privilege is set, the unprivileged state may well
6572 be exactly what is intended, so the warning has been cut out in that case,
6573 and Exim is allowed to try to write to its normal log files.
6579 1. smtp_return_error_details was not giving details for temporary sender
6580 or receiver verification errors.
6582 2. Diagnose a configuration error if two authenticators have the same public
6585 3. Exim used not to create the message log file for a message until the first
6586 delivery attempt. This could be confusing when incoming messages were held
6587 for policy or load reasons. The message log file is now created at the time
6588 the message is received, and an initial "Received" line is written to it.
6590 4. The automatically generated man page for command line options had a minor
6591 bug that caused no ill effects; however, a more serious problem was that
6592 the procedure for building the man page automatically didn't always
6593 operate. Consequently, release 4.20 contains an out-of-date version. This
6594 shouldn't happen again.
6596 5. When building Exim with embedded Perl support, the script that builds the
6597 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6598 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6600 6. The freeze_tell option was not being used for messages that were frozen on
6601 arrival, either by an ACL or by local_scan().
6603 7. Added the smtp_incomplete_transaction log selector.
6605 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6606 was accepting AUTH without a new EHLO.
6608 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6609 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6612 10. Logging of TCP/IP connections (when configured) now happens in the main
6613 daemon process instead of the child process, so that the TCP/IP connection
6614 count is more accurate (but it can never be perfect).
6616 11. The use of "drop" in a nested ACL was not being handled correctly in the
6617 outer ACL. Now, if condition failure induced by the nested "drop" causes
6618 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6619 or "require"), the connection is dropped.
6621 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6622 that yield "discard" can now be used with an "accept" or a "discard" verb,
6623 but an error is generated for any others (because I can't see a useful way
6624 to define what should happen).
6626 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6627 are now processed in the same way as lines in the Exim configuration file.
6628 In particular, continuation lines are supported.
6630 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6632 15. Added -ti meaning -t -i.
6634 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6635 domains, and warn by logging if others are found.
6637 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6638 returned from GIFCONF when Exim is trying to find the list of interfaces on
6639 a host. The code in os.c has been modified to copy each ifreq to an aligned
6640 structure in all cases.
6642 Also, in some cases, the returned ifreq's were being copied to a 'struct
6643 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6644 means the last couple of bytes of an IPv6 address could be chopped if the
6645 ifreq contained only a normal sockaddr (14 bytes storage).
6647 18. Named domain lists were not supported in the hosts_treat_as_local option.
6648 An entry such as +xxxx was not recognized, and was treated as a literal
6651 19. Ensure that header lines added by a DATA ACL are included in the reject log
6652 if the ACL subsequently rejects the message.
6654 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6655 MD5 (which is deprecated).
6657 21. When testing a filter file using -bf, Exim was writing a message when it
6658 took the sender from a "From " line in the message, but it was not doing so
6659 when it took $return_path from a Return-Path: header line. It now does.
6661 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6662 with a valid header line field name (a series of printing characters
6663 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6665 23. Changed "disc" in the source to "disk" to conform to the documentation and
6666 the book and for uniformity.
6668 24. Ignore Sendmail's -Ooption=value command line item.
6670 25. When execve() failed while trying to run a command in a pipe transport,
6671 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6672 could be confused with a return value of 69 from the command itself. This
6673 has been changed to 127, the value the shell returns if it is asked to run
6674 a non-existent command. The wording for the related log line suggests a
6675 non-existent command as the problem.
6677 26. If received_header_text expands to an empty string, do not add a Received:
6678 header line to the message. (Well, it adds a token one on the spool, but
6679 marks it "old" so that it doesn't get used or transmitted.)
6681 27. Installed eximstats 1.28 (addition of -nt option).
6683 28. There was no check for failure on the call to getsockname() in the daemon
6684 code. This can fail if there is a shortage of resources on the system, with
6685 ENOMEM, for example. A temporary error is now given on failure.
6687 29. Contrary to the C standard, it seems that in some environments, the
6688 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6689 program. Exim now does this explicitly; it affects the formatting of
6690 timestamps using strftime().
6692 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6693 complaints. I've now initialized all the variables, to avoid this.
6695 32. Header lines added by a system filter were not being "seen" during
6696 transport-time rewrites.
6698 33. The info_callback() function passed to OpenSSL is set up with type void
6699 (*)(SSL *, int, int), as described somewhere. However, when calling the
6700 function (actually a macro) that sets it up, the type void(*)() is
6701 expected. I've put in a cast to prevent warnings from picky compilers.
6703 34. If a DNS black list lookup found a CNAME record, but there were no A
6704 records associated with the domain it pointed at, Exim crashed.
6706 35. If a DNS black list lookup returned more than one A record, Exim ignored
6707 all but the first. It now scans all returned addresses if a particular IP
6708 value is being sought. In this situation, the contents of the
6709 $dnslist_value variable are a list of all the addresses, separated by a
6712 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6713 to accept a host name and all its aliases if the forward lookup for any of
6714 them yielded the IP address of the incoming connection. Now it accepts only
6715 those names whose forward lookup yields the correct IP address. Any other
6716 names are discarded. This closes a loophole whereby a rogue DNS
6717 administrator could create reverse DNS records to break through a
6718 wildcarded host restriction in an ACL.
6720 37. If a user filter or a system filter that ran in a subprocess used any of
6721 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6722 the wrong values were passed to the pipe command ($thisaddress had the
6723 value of $0, $0 had the value of $1, etc). This bug was introduced by
6724 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6726 38. Improved the line breaking for long SMTP error messages from ACLs.
6727 Previously, if there was no break point between 40 and 75 characters, Exim
6728 left the rest of the message alone. Two changes have been made: (a) I've
6729 reduced the minimum length to 35 characters; (b) if it can't find a break
6730 point between 35 and 75 characters, it looks ahead and uses the first one
6731 that it finds. This may give the occasional overlong line, but at least the
6732 remaining text gets split now.
6734 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6735 file descriptors might be low, and that setting 1000 would always raise it.
6736 It turns out that in some environments, the limit is already over 1000 and
6737 that lowering it causes trouble. So now Exim takes care not to decrease it.
6739 40. When delivering a message, the value of $return_path is set to $sender_
6740 address at the start of routing (routers may change the value). By an
6741 oversight, this default was not being set up when an address was tested by
6742 -bt or -bv, which affected the outcome if any router or filter referred to
6745 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6746 log only when "message" or "log_message" are set. However, if one of the
6747 conditions was an address verification, or a call to a nested ACL, the
6748 messages generated by the underlying test were being passed through. This
6749 no longer happens. The underlying message is available in $acl_verify_
6750 message for both "message" and "log_message" expansions, so it can be
6751 passed through if needed.
6753 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6754 new expansion $bh_ to give the encoded byte string without charset
6755 translation. Translation happens only if iconv() is available; HAVE_ICONV
6756 indicates this at build time. HEADERS_CHARSET gives the charset to
6757 translate to; headers_charset can change it in the configuration, and
6758 "headers charset" can change it in an individual filter file.
6760 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6761 that creates RFC 2047 encoded "words" labels them as that charset instead
6762 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6763 expansion operator; (ii) when Exim creates a From: line for a local
6764 message; (iii) when a header line is rewritten to include a "phrase" part.
6766 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6767 buggy, causing it to skip the first lines of messages whose message ID
6768 ended in 'D'. This would not have bitten before Exim release 4.14, because
6769 message IDs were unlikely to end in 'D' before then. The effect was to have
6770 incorrect size information for certain domains.
6772 45. #include "config.h" was missing at the start of the crypt16.c module. This
6773 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6776 46. If there was a timeout during a "random" callout check, Exim treated it as
6777 a failure of the random address, and carried on sending RSET and the real
6778 address. If the delay was just some slowness somewhere, the response to the
6779 original RCPT would be taken as a response to RSET and so on, causing
6780 mayhem of various kinds.
6782 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6783 when I implemented it. It didn't allow for the fact that some option values
6784 may legitimately be negative (e.g. size_addition), and it didn't even do
6785 the right test for positive values.
6787 48. Domain names in DNS records are case-independent. Exim always looks them up
6788 in lower case. Some resolvers return domain names in exactly the case they
6789 appear in the zone file, that is, they may contain uppercase letters. Not
6790 all resolvers do this - some return always lower case. Exim was treating a
6791 change of case by a resolver as a change of domain, similar to a widening
6792 of a domain abbreviation. This triggered its re-routing code and so it was
6793 trying to route what was effectively the same domain again. This normally
6794 caused routing to fail (because the router wouldn't handle the domain
6795 twice). Now Exim checks for this case specially, and just changes the
6796 casing of the domain that it ultimately uses when it transmits the message
6799 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6802 50. If a filter generated a file delivery with a non-absolute name (possible if
6803 no home directory exists for the router), the forbid_file option was not
6806 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6807 the existing equality matching.
6809 52. Exim was using ints instead of ino_t variables in some places where it was
6810 dealing with inode numbers.
6812 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6813 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6814 called TMPDIR, and if it finds it is different, it changes its value.
6816 54. The smtp_printf() function is now made available to local_scan() so
6817 additional output lines can be written before returning. There is also an
6818 smtp_fflush() function to enable the detection of a dropped connection.
6819 The variables smtp_input and smtp_batched_input are exported to
6822 55. Changed the default runtime configuration: the message "Unknown user"
6823 has been removed from the ACL, and instead placed on the localuser router,
6824 using the cannot_route_message feature. This means that any verification
6825 failures that generate their own messages won't get overridden. Similarly,
6826 the "Unrouteable address" message that was in the ACL for unverifiable
6827 relay addresses has also been removed.
6829 56. Added hosts_avoid_esmtp to the smtp transport.
6831 57. The exicyclog script was not checking for the esoteric option
6832 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6833 will work only if exicyclog is run under the appropriate euid.
6835 58. Following a discussion on the list, the rules by which Exim recognises line
6836 endings on incoming messages have been changed. The -dropcr and drop_cr
6837 options are now no-ops, retained only for backwards compatibility. The
6838 following line terminators are recognized: LF CRLF CR. However, special
6839 processing applies to CR:
6841 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6842 nor a local message in the state where . is a terminator.
6844 (ii) If a bare CR is encountered in a header line, an extra space is added
6845 after the line terminator so as not to end the header. The reasoning
6846 behind this is that bare CRs in header lines are most likely either
6847 to be mistakes, or people trying to play silly games.
6849 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6850 was being incorrectly given as 18 bytes larger than it should have been.
6851 This is a VOB (very old bug).
6853 60. This may never have affected anything current, but just in case it has:
6854 When the local host is found other than at the start of a list of hosts,
6855 the local host, those with the same MX, and any that follow, are discarded.
6856 When the list in question was part of a longer list of hosts, the following
6857 hosts (not currently being processed) were also being discarded. This no
6858 longer happens. I'm not sure if this situation could ever has previously
6861 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6864 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6865 header. This is something that is recommended in a new Internet Draft, and
6866 is something that is documented as being done by Sendmail. There are two
6867 possible values. For messages generated by the autoreply transport, Exim
6870 Auto-Submitted: auto-replied
6872 whereas for all other generated messages (e.g. bounces) it adds
6874 Auto-Submitted: auto-generated
6876 63. The "personal" condition in filters now includes a test for the
6877 Auto-Submitted: header. If it contains the string "auto-" the message it
6878 not considered personal.
6880 64. Added rcpt_include_affixes as a generic transport option.
6882 65. Added queue_only_override (default true).
6884 66. Added the syslog_duplication option.
6886 67. If what should have been the first header line of a message consisted of
6887 a space followed by a colon, Exim was mis-interpreting it as a header line.
6888 It isn't of course - it is syntactically invalid and should therefore be
6889 treated as the start of the message body. The misbehaviour could have
6890 caused a number of strange effects, including loss of data in subsequent
6891 header lines, and spool format errors.
6893 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6894 client host had authenticated. This control can now be exercised by an ACL
6895 for more flexibility.
6897 69. By default, callouts do not happen when testing with -bh. There is now a
6898 variant, -bhc, which does actually run the callout code, including
6899 consulting and updating the callout cache.
6901 70. Added support for saslauthd authentication, courtesy of Alexander
6904 71. If statvfs() failed on the spool or log directories while checking their
6905 size for availability, Exim confusingly gave the error "space shortage".
6906 Furthermore, in debugging mode it crashed with a floating point exception.
6907 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6908 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6909 serious problem, Exim now writes to the main and panic logs when this
6910 happens, with details of the failure. It then refuses to accept the
6911 incoming message, giving the message "spool directory problem" or "log
6912 directory problem" with a 421 code for SMTP messages.
6914 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6915 0, 1, and 2 exist, because some OS complain for execs without them (see
6916 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6917 descriptors. However, the code omitted to check that the open succeeded,
6918 causing mysterious errors if for some reason the permissions on /dev/null
6919 got screwed. Now Exim writes a message to the main and panic logs, and
6920 bombs out if it can't open /dev/null.
6922 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6923 interact so that it is all more flexible. It is supposed to remain
6924 backwards compatible. Also added extra_local_interfaces.
6926 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6927 to bomb out with an assertion failure - to the client this appears as a
6928 connection drop. This problem occurs in the part of the code that was taken
6929 from the Samba project. Fortunately, the assertion is in a very simple
6930 function, so I have fixed this by reproducing the function inline in the
6931 one place where it is called, and arranging for authentication to fail
6932 instead of killing the process with assert().
6934 75. The SPA client code was not working when the server requested OEM rather
6935 than Unicode encoding.
6937 76. Added code to make require_files with a specific uid setting more usable in
6938 the case where statting the file as root fails - usually a non-root-mounted
6939 NFS file system. When this happens and the failure is EACCES, Exim now
6940 forks a subprocess and does the per-uid checking as the relevant uid.
6942 77. Added process_log_path.
6944 78. If log_file_path was not explicitly set, a setting of check_log_space or
6945 check_log_inodes was ignored.
6947 79. If a space check for the spool or log partitions fails, the incident is now
6948 logged. Of course, in the latter case the data may get lost...
6950 80. Added the %p formatting code to string_format() so that it can be used to
6951 print addresses in debug_print(). Adjusted all the address printing in the
6952 debugging in store.c to use %p rather than %d.
6954 81. There was a concern that a line of code in smtp_in.c could overflow a
6955 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6956 initially expressed, the concern was not well-founded, because trailing
6957 spaces are removed early. However, if the trailing spaces were followed by
6958 a NULL, they did not get removed, so the overflow was possible. Two fixes
6961 (a) I re-wrote the offending code in a cleaner fashion.
6962 (b) If an incoming SMTP command contains a NULL character, it is rejected
6965 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6966 runs initgroups(), so that if the Exim user is in any additional groups,
6967 they will be used during message reception.
6973 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob