1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Replace the store_release() internal interface with store_newblock(),
12 which internalises the check required to safely use the old one, plus
13 the allocate and data copy operations duplicated in both (!) of the
16 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
17 modifier. This matches the restriction on the commandline.
19 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
20 Previously only the last row was returned.
22 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
23 we assumed that tags in the header were well-formed, and parsed the
24 element content after inspecting only the first char of the tag.
25 Assumptions at that stage could crash the receive process on malformed
32 JH/01 Rework error string handling in TLS interface so that the caller in
33 more cases is responsible for logging. This permits library-sourced
34 string to be attached to addresses during delivery, and collapses
35 pairs of long lines into single ones.
37 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
38 during configuration. Wildcards are allowed and expanded.
40 JH/02 Rework error string handling in DKIM to pass more info back to callers.
41 This permits better logging.
43 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
44 do not close it down and have the child transport start it up again on
45 the passed-on TCP connection. Instead, proxy the child (and any
46 subsequent ones) for TLS via a unix-domain socket channel. Logging is
47 affected: the continued delivery log lines do not have any DNSSEC, TLS
48 Certificate or OCSP information. TLS cipher information is still logged.
50 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
51 identical IP addresses on different listening ports. Will also affect
54 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
55 add noisy ifdef guards to special-case this sillyness.
56 Patch from Bernd Kuhls.
58 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
59 than 255 are no longer allowed.
61 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
62 Disable the session-cache too, which might reduce our load. Since we
63 currrectly use a new context for every connection, both as server and
64 client, there is no benefit for these.
65 GnuTLS appears to not support tickets server-side by default (we don't
66 call gnutls_session_ticket_enable_server()) but client side is enabled
67 by default on recent versions (3.1.3 +) unless the PFS priority string
70 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
71 <https://reproducible-builds.org/specs/source-date-epoch/>.
73 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
74 the check for any unsuccessful recipients did not notice the limit, and
75 erroneously found still-pending ones.
77 JH/08 Pipeline CHUNKING command and data together, on kernels that support
78 MSG_MORE. Only in-clear (not on TLS connections).
80 JH/09 Avoid using a temporary file during transport using dkim. Unless a
81 transport-filter is involved we can buffer the headers in memory for
82 creating the signature, and read the spool data file once for the
83 signature and again for transmission.
85 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
86 4.77 as the kernel support then wasn't solid, having issues in 64bit
87 mode. Now, it's been long enough. Add support for FreeBSD also.
89 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
90 case where the routing stage had gathered several addresses to send to
91 a host before calling the transport for the first, we previously failed
92 to close down TLS in the old transport process before passing the TCP
93 connection to the new process. The new one sent a STARTTLS command
94 which naturally failed, giving a failed delivery and bloating the retry
95 database. Investigation and fix prototype from Wolfgang Breyha.
97 JH/12 Fix check on SMTP command input synchronisation. Previously there were
98 false-negatives in the check that the sender had not preempted a response
99 or prompt from Exim (running as a server), due to that code's lack of
100 awareness of the SMTP input buffering.
102 PP/04 Add commandline_checks_require_admin option.
103 Exim drops privileges sanely, various checks such as -be aren't a
104 security problem, as long as you trust local users with access to their
105 own account. When invoked by services which pass untrusted data to
106 Exim, this might be an issue. Set this option in main configuration
107 AND make fixes to the calling application, such as using `--` to stop
110 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
111 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
112 responses to those, into a single TLS record each way (this usually means
113 a single packet). As a side issue, smtp_enforce_sync now works on TLS
116 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
117 affects you only if you're dancing at the edge of the param size limits.
118 If you are, and this message makes sense to you, then: raise the
119 configured limit or use OpenSSL 1.1. Nothing we can do for older
122 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
123 cmdline element to get the documented default one. Previously it was
126 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
127 are unsuable for non-message-specific reasons - eg. connection timeout,
128 banner-time rejection.
130 JH/16 Fix logging of delivery remote port, when specified by router, under
133 PP/06 Repair manualroute's ability to take options in any order, even if one
134 is the name of a transport.
137 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
139 JH/17 Change the list-building routines interface to use the expanding-string
140 triplet model, for better allocation and copying behaviour.
142 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
143 Previously it was constructed the first time a possibly-matching string
144 was met in the configuration file input during startup; now it is done
147 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
148 compatible one, to avoid the (poorly documented) possibility of a config
149 file in the working directory redirecting the DB files, possibly correpting
150 some existing file. CVE-2017-10140 assigned for BDB.
152 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
153 cache-hot. Previously, although the result was properly cached, the
154 initial verify call returned a defer.
156 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
157 the main verify for receipient in uncached-mode.
159 JH/22 Retire historical build files to an "unsupported" subdir. These are
160 defined as "ones for which we have no current evidence of testing".
162 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
163 if present. Previously it was ignored.
165 JH/24 Start using specified-initialisers in C structure init coding. This is
166 a C99 feature (it's 2017, so now considered safe).
168 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
169 if was a fixed-sized field and bitmask ops via macros; it is now more
172 PP/07 GitHub PR 56: Apply MariaDB build fix.
173 Patch provided by Jaroslav Škarvada.
175 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
176 during Coverity cleanups [4.87 JH/47]
177 Diagnosis and fix provided by Michael Fischer v. Mollard.
179 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
180 the right size to place the terminating semicolon on its own folded
181 line, the header hash was calculated to an incorrect value thanks to
182 the (relaxed) space the fold became.
184 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
185 and confused the parent.
187 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
188 which could crash as a result. This could lead to undeliverable messages.
190 JH/28 Logging: "next input sent too soon" now shows where input was truncated
193 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
194 matters on fast-turnover and PID-randomising systems, which were getting
195 out-of-order delivery.
197 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
198 a possibly-overlapping copy. The symptom was that "Remote host closed
199 connection in response to HELO" was logged instead of the actual 4xx
202 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
203 Previously only that bufferd was discarded, resulting in SYMTP command
206 JH/32 DKIM: when a message has multiple signatures matching an identity given
207 in dkim_verify_signers, run the dkim acl once for each. Previously only
208 one run was done. Bug 2189.
210 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
211 "panic the current process" to "deliberately defer". The panic log is
212 still written with the problem list name; the mail and reject logs now
213 get a temp-reject line for the message that was being handled, saying
214 something like "domains check lookup or other defer". The SMTP 451
215 message is still "Temporary local problem".
217 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
218 A crafted sequence of BDAT commands could result in in-use memory beeing
219 freed. CVE-2017-16943.
221 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
222 from SMTP input. Previously it was always done; now only done for DATA
223 and not BDAT commands. CVE-2017-16944.
225 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
226 to the message (such as an overlong header line). Previously this was
227 not done and we did not exit BDAT mode. Followon from the previous item
228 though a different problem.
234 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
235 than -2003 did; needs libidn2 in addition to libidn.
237 JH/02 The path option on a pipe transport is now expanded before use.
239 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
240 Patch provided by "Björn", documentation fix added too.
242 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
243 missing a wire-to-host endian conversion.
245 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
246 close after a BDAT command line could be taken as a following command,
247 giving a synch failure. Fix by only checking for synch immediately
248 before acknowledging the chunk.
250 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
251 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
252 macro. Patches provided by Josh Soref.
254 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
255 Previously we did not; the RFC seems ambiguous and VRFY is not listed
256 by IANA as a service extension. However, John Klensin suggests that we
259 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
260 the dkim code may be unix-mode line endings rather than smtp wire-format
261 CRLF, so prepend a CR to any bare LF.
263 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
264 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
266 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
267 the first were themselves being wrongly included in the feed into dkim
268 processing; with most chunk sizes in use this resulted in an incorrect
269 body hash calculated value.
271 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
272 DKIM signature block, for verification. Although advised against by
273 standards it is specifically not ruled illegal.
275 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
277 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
278 missing a body hash (the bh= tag).
280 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
281 It seems that HAProxy sends the Proxy Protocol information in clear and
282 only then does a TLS startup, so do the same.
284 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
285 TCP connections (such as for Spamd) unless the daemon successfully set
286 Fast Open mode on its listening sockets. This fixes breakage seen on
287 too-old kernels or those not configured for Fast Open, at the cost of
288 requiring both directions being enabled for TFO, and TFO never being used
289 by non-daemon-related Exim processes.
291 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
292 endings, at least on the first header line. Try to canonify any that get
293 past that check, despite the cost.
295 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
296 now limited to an arbitrary five deep, while parsing addresses with the
297 strip_excess_angle_brackets option enabled.
299 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
300 instead leave the unprompted TLS handshake in socket buffer for the
301 TLS library to consume.
303 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
305 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
307 JH/16 Drop variables when they go out of scope. Memory management drops a whole
308 region in one operation, for speed, and this leaves assigned pointers
309 dangling. Add checks run only under the testsuite which checks all
310 variables at a store-reset and panics on a dangling pointer; add code
311 explicitly nulling out all the variables discovered. Fixes one known
312 bug: a transport crash, where a dangling pointer for $sending_ip_address
313 originally assigned in a verify callout, is re-used.
315 PP/06 Drop '.' from @INC in various Perl scripts.
317 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
319 PP/08 Reduce a number of compilation warnings under clang; building with
320 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
321 should be warning-free.
323 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
325 HS/01 Fix portability problems introduced by PP/08 for platforms where
326 realloc(NULL) is not equivalent to malloc() [SunOS et al].
328 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
329 chunk. This allows us to accept broken chunked messages. We need a more
330 general solution here.
332 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
333 already-broken messages in the queue.
335 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
337 JH/19 Fix reference counting bug in routing-generated-address tracking.
343 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
344 supports it and a size is available (ie. the sending peer gave us one).
346 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
347 years of being deprecated). The replacements are the ACLs
348 acl_smtp_mime and acl_not_smtp_mime.
350 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
351 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
352 or even an in-clear connection were permitted. Now, if the host lookup
353 was dnssec and dane was requested then the host is only used if the
354 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
355 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
356 if one fails this test.
357 This means that a poorly-configured remote DNS will make it incommunicado;
358 but it protects against a DNS-interception attack on it.
360 JH/04 Bug 1810: make continued-use of an open smtp transport connection
361 non-noisy when a race steals the message being considered.
363 JH/05 If main configuration option tls_certificate is unset, generate a
364 self-signed certificate for inbound TLS connections.
366 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
367 in rewrites and routers.
369 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
370 and logged a warning sing 4.83; now they are a configuration file error.
372 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
373 (lacking @domain). Apply the same qualification processing as RCPT.
375 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
377 JH/10 Support ${sha256:} applied to a string (as well as the previous
380 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
381 a cutthrough deliver is pending, as we always want to make a connection.
382 This also avoids re-routing the message when later placing the cutthrough
383 connection after a verify cache hit.
384 Do not update it with the verify result either.
386 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
387 when routing results in more than one destination address.
389 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
390 signing (which inhibits the cutthrough capability). Previously only
391 the presence of an option was tested; now an expansion evaluating as
392 empty is permissible (obviously it should depend only on data available
393 when the cutthrough connection is made).
395 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
396 the relevant preceding SMTP command did not note the pipelining mode.
398 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
399 Previously they were not counted.
401 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
402 as one having no matching records. Previously we deferred the message
403 that needed the lookup.
405 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
406 distinguished as "(=".
408 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
409 for missing MX records. Previously it only worked for missing A records.
411 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
413 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
414 after the data-go-ahead and data-ack. Patch from Jason Betts.
416 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
417 even for a "none" policy. Patch from Tony Meyer.
419 JH/22 Fix continued use of a connection for further deliveries. If a port was
420 specified by a router, it must also match for the delivery to be
423 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
424 When one of the recipients of a message was unsuitable for the connection
425 (has no matching addresses), we lost track of needing to mark it
426 deferred. As a result mail would be lost.
428 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
430 JH/25 Decoding ACL controls is now done using a binary search; the source code
431 takes up less space and should be simpler to maintain. Merge the ACL
432 condition decode tables also, with similar effect.
434 JH/26 Fix problem with one_time used on a redirect router which returned the
435 parent address unchanged. A retry would see the parent address marked as
436 delivered, so not attempt the (identical) child. As a result mail would
439 JH/27 Fix a possible security hole, wherein a process operating with the Exim
440 UID can gain a root shell. Credit to http://www.halfdog.net/ for
441 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
444 JH/28 Enable {spool,log} filesystem space and inode checks as default.
445 Main config options check_{log,spool}_{inodes,space} are now
446 100 inodes, 10MB unless set otherwise in the configuration.
448 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
449 Previously it only applied to the main-section connection policy
452 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
454 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
455 by me. Added RFC7919 DH primes as an alternative.
457 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
458 are not in the system include path.
460 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
461 GnuTLS, when a session startup failed (eg because the client disconnected)
462 Exim did stdio operations after fclose. This was exposed by a recent
463 change which nulled out the file handle after the fclose.
465 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
466 signed directly by the cert-signing cert, rather than an intermediate
467 OCSP-signing cert. This is the model used by LetsEncrypt.
469 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
471 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
472 an incoming connection.
474 HS/02 Bug 1802: Do not half-close the connection after sending a request
477 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
478 fallback to "prime256v1".
480 JH/34 SECURITY: Use proper copy of DATA command in error message.
481 Could leak key material. Remotely exploitable. CVE-2016-9963.
487 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
488 and 3.4.4 - once the server is enabled to respond to an OCSP request
489 it does even when not requested, resulting in a stapling non-aware
490 client dropping the TLS connection.
492 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
493 support variable-length bit vectors. No functional change.
495 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
496 The I= interface field on outgoing lines is now after the H= remote
497 host field, same as incoming lines. There is a separate
498 outgoing_interface log selector which allows you to disable the
501 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
502 If not running log_selector +smtp_connection the mainlog would be held
503 open indefinitely after a "too many connections" event, including to a
504 deleted file after a log rotate. Leave the per net connection logging
505 leaving it open for efficiency as that will be quickly detected by the
506 check on the next write.
508 HS/01 Bug 1671: Fix post transport crash.
509 Processing the wait-<transport> messages could crash the delivery
510 process if the message IDs didn't exist for some reason. When
511 using 'split_spool_directory=yes' the construction of the spool
512 file name failed already, exposing the same netto behaviour.
514 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
515 mime_regex ACL conditions.
517 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
518 to DSN fail messages (bounces): remote IP, remote greeting, remote response
519 to HELO, local diagnostic string.
521 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
522 log line to debug. Even when configured with a tls authenticator many
523 client connections are expected to not authenticate in this way, so
524 an authenticate fail is not an error.
526 HS/02 Add the Exim version string to the process info. This way exiwhat
527 gives some more detail about the running daemon.
529 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
530 matter for fast-change records such as DNSBLs.
532 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
533 retry record, even if constant. There may be multiple transports with
534 different interface settings and the retry behaviour needs to be kept
537 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
539 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
541 JH/10 Bug 840: fix log_defer_output option of pipe transport
543 JH/11 Bug 830: use same host for all RCPTS of a message, even under
544 hosts_randomize. This matters a lot when combined with mua_wrapper.
546 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
547 ${quote_pgsql:<string>} operator.
549 JH/13 Bug 1708: avoid misaligned access in cached lookup.
551 JH/14 Change header file name for freeradius-client. Relevant if compiling
552 with Radius support; from the Gentoo tree and checked under Fedora.
554 JH/15 Bug 1712: Introduce $prdr_requested flag variable
556 JH/16 Bug 1714: Permit an empty string as expansion result for transport
557 option transport_filter, meaning no filtering.
559 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
561 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
562 defaults to "*" (all hosts). The variable is now available when not built
563 with TLS, default unset, mainly to enable keeping the testsuite sane.
564 If a server certificate is not supplied (via tls_certificate) an error is
565 logged, and clients will find TLS connections fail on startup. Presumably
566 they will retry in-clear.
567 Packagers of Exim are strongly encouraged to create a server certificate
568 at installation time.
570 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
571 with the $config_file variable.
573 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
574 in transport context, after the attempt, and per-recipient. The latter type
575 is per host attempted. The event data is the error message, and the errno
576 information encodes the lookup type (A vs. MX) used for the (first) host,
577 and the trailing two digits of the smtp 4xx response.
579 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
580 to write to mainlog (or rejectlog, paniclog) in the window between file
581 creation and permissions/ownership being changed. Particularly affects
582 installations where exicyclog is run as root, rather than exim user;
583 result is that the running daemon panics and dies.
585 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
587 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
588 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
589 "pri" and "weight". Note that the previous implicit priority given by the
590 list order is no longer honoured.
592 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
595 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
596 by defining SUPPORT_SOCKS.
598 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
599 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
600 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
601 variables are renamed to proxy_{local,external}_{address,port}.
603 JH/27 Move Internationalisation support from Experimental to mainline, enabled
604 for a build by defining SUPPORT_I18N
606 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
607 of the query string, and make ${quote_redis:} do that quoting.
609 JH/29 Move Events support from Experimental to mainline, enabled by default
610 and removable for a build by defining DISABLE_EVENT.
612 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
614 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
615 cached by the daemon.
617 JH/32 Move Redis support from Experimental to mainline, enabled for a build
618 by defining LOOKUP_REDIS. The libhiredis library is required.
620 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
621 keys are given for lookup.
623 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
624 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
625 only supported when built with TLS support. The PolarSSL SHA routines
626 are still used when the TLS library is too old for convenient support.
628 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
629 openssl_options), for security. OpenSSL forces this from version 1.1.0
630 server-side so match that on older versions.
632 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
633 allocation for $value could be released as the expansion processing
634 concluded, but leaving the global pointer active for it.
636 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
637 and to use the domains and local_parts ACL conditions.
639 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
640 incorrectly not doubled on cutthrough transmission, hence seen as a
641 body-termination at the receiving system - resulting in truncated mails.
642 Commonly the sender saw a TCP-level error, and retransmitted the message
643 via the normal store-and-forward channel. This could result in duplicates
644 received - but deduplicating mailstores were liable to retain only the
645 initial truncated version.
647 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
649 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
651 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
652 we're in there, support oversigning also; bug 1309.
654 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
656 HS/04 Add support for keep_environment and add_environment options.
658 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
659 either intentional arithmetic overflow during PRNG, or testing config-
662 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
663 delivery resulted in actual delivery. Cancel cutthrough before DATA
666 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
667 rejects a recipient: pass the reject to the originator.
669 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
670 Many were false-positives and ignorable, but it's worth fixing the
673 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
674 for the new environment-manipulation done at startup. Move the routines
675 from being local to tls.c to being global via the os.c file.
677 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
678 an extract embedded as result-arg for a map, the first arg for extract
679 is unavailable so we cannot tell if this is a numbered or keyed
680 extraction. Accept either.
686 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
689 JH/02 The smtp transport option "multi_domain" is now expanded.
691 JH/03 The smtp transport now requests PRDR by default, if the server offers
694 JH/04 Certificate name checking on server certificates, when exim is a client,
695 is now done by default. The transport option tls_verify_cert_hostnames
696 can be used to disable this per-host. The build option
697 EXPERIMENTAL_CERTNAMES is withdrawn.
699 JH/05 The value of the tls_verify_certificates smtp transport and main options
700 default to the word "system" to access the system default CA bundle.
701 For GnuTLS, only version 3.0.20 or later.
703 JH/06 Verification of the server certificate for a TLS connection is now tried
704 (but not required) by default. The verification status is now logged by
705 default, for both outbound TLS and client-certificate supplying inbound
708 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
711 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
712 Status Notification (bounce) messages are now MIME format per RFC 3464.
713 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
714 under the control of the dsn_advertise_hosts option, and routers may
715 have a dsn_lasthop option.
717 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
718 default, modifiable by a malware= option. The list separator for
719 the options can now be changed in the usual way. Bug 68.
721 JH/10 The smtp_receive_timeout main option is now expanded before use.
723 JH/11 The incoming_interface log option now also enables logging of the
724 local interface on delivery outgoing connections.
726 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
727 if the interface and destination host and port all match.
729 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
732 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
733 Patch from Andrew Lewis.
735 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
736 now supports optional time-restrictions, weighting, and priority
737 modifiers per server. Patch originally by <rommer@active.by>.
739 JH/16 The spamd_address main option now supports a mixed list of local
740 and remote servers. Remote servers can be IPv6 addresses, and
741 specify a port-range.
743 JH/17 Bug 68: The spamd_address main option now supports an optional
744 timeout value per server.
746 JH/18 Bug 1581: Router and transport options headers_add/remove can
747 now have the list separator specified.
749 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
752 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
755 JH/21 Support for the A6 type of dns record is withdrawn.
757 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
758 rather than the verbs used.
760 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
761 from 255 to 1024 chars.
763 JH/24 Verification callouts now attempt to use TLS by default.
765 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
766 are generic router options now. The defaults didn't change.
768 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
769 Original patch from Alexander Shikoff, worked over by JH.
771 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
772 files if ZCAT_COMMAND is not executable.
774 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
776 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
778 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
779 Normally benign, it bites when the pair was led to by a CNAME;
780 modern usage is to not canonicalize the domain to a CNAME target
781 (and we were inconsistent anyway for A-only vs AAAA+A).
783 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
785 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
786 when evaluating $sender_host_dnssec.
788 JH/31 Check the HELO verification lookup for DNSSEC, adding new
789 $sender_helo_dnssec variable.
791 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
793 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
795 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
797 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
798 documented as working, but never had. Support all but $spam_report.
800 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
801 added for tls authenticator.
803 HS/03 Add perl_taintmode main config option
809 TL/01 When running the test suite, the README says that variables such as
810 no_msglog_check are global and can be placed anywhere in a specific
811 test's script, however it was observed that placement needed to be near
812 the beginning for it to behave that way. Changed the runtest perl
813 script to read through the entire script once to detect and set these
814 variables, reset to the beginning of the script, and then run through
815 the script parsing/test process like normal.
817 TL/02 The BSD's have an arc4random API. One of the functions to induce
818 adding randomness was arc4random_stir(), but it has been removed in
819 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
820 function when detected.
822 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
823 cause callback expansion.
825 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
826 syntax errors in an expansion can be treated as a string instead of
827 logging or causing an error, due to the internal use of bool_lax
828 instead of bool when processing it.
830 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
831 server certificates when making smtp deliveries.
833 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
835 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
837 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
839 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
840 Merged patch from Sebastian Wiedenroth.
842 JH/05 Fix results-pipe from transport process. Several recipients, combined
843 with certificate use, exposed issues where response data items split
844 over buffer boundaries were not parsed properly. This eventually
845 resulted in duplicates being sent. This issue only became common enough
846 to notice due to the introduction of connection certificate information,
847 the item size being so much larger. Found and fixed by Wolfgang Breyha.
849 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
850 size buffer was used, resulting in syntax errors when an expansion
853 JH/07 Add support for directories of certificates when compiled with a GnuTLS
854 version 3.3.6 or later.
856 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
857 is EXPERIMENTAL_EVENT, the main-configuration and transport options
858 both become "event_action", the variables become $event_name, $event_data
859 and $event_defer_errno. There is a new variable $verify_mode, usable in
860 routers, transports and related events. The tls:cert event is now also
861 raised for inbound connections, if the main configuration event_action
864 TL/06 In test suite, disable OCSP for old versions of openssl which contained
865 early OCSP support, but no stapling (appears to be less than 1.0.0).
867 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
868 server certificate names available under the smtp transport option
869 "tls_verify_cert_hostname" now do not permit multi-component wildcard
872 JH/10 Time-related extraction expansions from certificates now use the main
873 option "timezone" setting for output formatting, and are consistent
874 between OpenSSL and GnuTLS compilations. Bug 1541.
876 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
877 encoded parameter in the incoming message. Bug 1558.
879 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
880 include certificate info, eximon was claiming there were spoolfile
883 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
885 JH/14 Log delivery-related information more consistently, using the sequence
886 "H=<name> [<ip>]" wherever possible.
888 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
889 are problematic for Debian distribution, omit them from the release
892 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
894 JH/16 Fix string representation of time values on 64bit time_t architectures.
897 JH/17 Fix a null-indirection in certextract expansions when a nondefault
898 output list separator was used.
903 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
904 checkers that were complaining about end of non-void function with no
907 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
908 This was a regression introduced in 4.83 by another bugfix.
910 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
912 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
913 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
919 TF/01 Correctly close the server side of TLS when forking for delivery.
921 When a message was received over SMTP with TLS, Exim failed to clear up
922 the incoming connection properly after forking off the child process to
923 deliver the message. In some situations the subsequent outgoing
924 delivery connection happened to have the same fd number as the incoming
925 connection previously had. Exim would try to use TLS and fail, logging
926 a "Bad file descriptor" error.
928 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
929 utilities have not been installed.
931 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
932 temporary space as the ACL may create new global variables.
934 TL/01 LDAP support uses per connection or global context settings, depending
935 upon the detected version of the libraries at build time.
937 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
938 to extract and use the src ip:port in logging and expansions as if it
939 were a direct connection from the outside internet. PPv2 support was
940 updated based on HAProxy spec change in May 2014.
942 JH/02 Add ${listextract {number}{list}{success}{fail}}.
944 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
945 Properly escape header and check for NULL return.
947 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
950 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
952 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
953 characters in header names, implemented as a verify condition.
954 Contributed by Michael Fischer v. Mollard.
956 TL/05 Rename SPF condition results err_perm and err_temp to standardized
957 results permerror and temperror. Previous values are deprecated but
958 still accepted. In a future release, err_perm and err_temp will be
959 completely removed, which will be a backward incompatibility if the
960 ACL tests for either of these two old results. Patch contributed by
961 user bes-internal on the mailing list.
963 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
965 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
966 selectors, in both main and reject logs.
968 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
971 JH/07 Add malware type "sock" for talking to simple daemon.
973 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
975 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
976 routers/transports under cutthrough routing.
978 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
979 numbers. Touch up "bool" conditional to keep the same definition.
981 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
983 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
985 JH/12 Expand items in router/transport headers_add or headers_remove lists
986 individually rather than the list as a whole. Bug 1452.
988 Required for reasonable handling of multiple headers_ options when
989 they may be empty; requires that headers_remove items with embedded
990 colons must have them doubled (or the list-separator changed).
992 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
993 view the policy declared in the DMARC record. Currently, $dmarc_status
994 is a combined value of both the record presence and the result of the
997 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
999 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1000 dnslookup router and the smtp transport (applying to the forward
1003 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1004 of ldap servers used for a specific lookup. Patch provided by Heiko
1007 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1008 New variable $lookup_dnssec_authenticated for observability.
1010 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1011 Patch submitted by Lars Timman.
1013 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1015 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1016 Requires trusted mode and valid format message id, aborts otherwise.
1017 Patch contributed by Heiko Schlichting.
1019 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1020 certextract with support for various fields. Bug 1358.
1022 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1023 is requested by default, modifiable by smtp transport option
1026 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1027 operate on certificate variables to give certificate fingerprints
1028 Also new ${sha256:cert_variable}.
1030 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1032 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1033 Christian Aistleitner.
1035 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1037 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1038 file. Patch from Wolfgang Breyha.
1040 JH/25 Expand the coverage of the delivery $host and $host_address to
1041 client authenticators run in verify callout. Bug 1476.
1043 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1044 align with daemon_smtp_ports. Bug 72.
1046 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1047 support and error reporting did not work properly.
1049 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1050 and is readable. Patch from Andrew Colin Kissa.
1052 TL/14 Enhance documentation of ${run expansion and how it parses the
1053 commandline after expansion, particularly in the case when an
1054 unquoted variable expansion results in an empty value.
1056 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1058 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1059 by renaming away from it. Addresses GH issue 15, reported by
1062 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1064 TL/15 SECURITY: prevent double expansion in math comparison functions
1065 (can expand unsanitized data). Not remotely exploitable.
1072 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1074 PP/02 Make -n do something, by making it not do something.
1075 When combined with -bP, the name of an option is not output.
1077 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1080 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1081 $sender_host_name and config options to manage this, and basic check
1084 PP/05 DSCP support for outbound connections and control modifier for inbound.
1086 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1087 (Only plugin which currently uses this is kerberos4, which nobody should
1088 be using, but we should make it available and other future plugins might
1089 conceivably use it, even though it would break NAT; stuff *should* be
1090 using channel bindings instead).
1092 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1093 name; added for Sendmail compatibility; requires admin caller.
1094 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1095 just ignore it); requires trusted caller.
1096 Also parse but ignore: -Ac -Am -X<logfile>
1099 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1101 TL/02 Add +smtp_confirmation as a default logging option.
1103 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1104 Patch by Magnus Holmgren from 2007-02-20.
1106 TL/04 Bugzilla 1281 - Spec typo.
1107 Bugzilla 1283 - Spec typo.
1108 Bugzilla 1290 - Spec grammar fixes.
1110 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1112 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1114 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1115 mailing list by by Dmitry Isaikin.
1117 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1119 JH/02 Support "G" suffix to numbers in ${if comparisons.
1121 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1123 NM/01 Bugzilla 1197 - Spec typo
1124 Bugzilla 1196 - Spec examples corrections
1126 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1128 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1129 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1132 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1133 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1135 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1136 "acl {{name}{arg}...}", and optional args on acl condition
1139 JH/05 Permit multiple router/transport headers_add/remove lines.
1141 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1143 JH/07 Avoid using a waiting database for a single-message-only transport.
1144 Performance patch from Paul Fisher. Bugzilla 1262.
1146 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1149 JH/09 Add $headers_added variable, with content from use of ACL modifier
1150 add_header (but not yet added to the message). Bugzilla 199.
1152 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1153 Pulled from Bugzilla 817 by Wolfgang Breyha.
1155 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1157 (nb: this is the same fix as in Exim 4.80.1)
1159 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1162 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1165 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1167 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1168 advertise SMTP AUTH mechanism to us, instead of a generic
1169 protocol violation error. Also, make Exim more robust to bad
1170 data from the Dovecot auth socket.
1172 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1174 When a queue runner is handling a message, Exim first routes the
1175 recipient addresses, during which it prunes them based on the retry
1176 hints database. After that it attempts to deliver the message to
1177 any remaining recipients. It then updates the hints database using
1180 So if a recipient address works intermittently, it can get repeatedly
1181 deferred at routing time. The retry hints record remains fresh so the
1182 address never reaches the final cutoff time.
1184 This is a fairly common occurrence when a user is bumping up against
1185 their storage quota. Exim had some logic in its local delivery code
1186 to deal with this. However it did not apply to per-recipient defers
1187 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1189 This change adds a proper retry rule check during routing so that the
1190 final cutoff time is checked against the message's age. We only do
1191 this check if there is an address retry record and there is not a
1192 domain retry record; this implies that previous attempts to handle
1193 the address had the retry_use_local_parts option turned on. We use
1194 this as an approximation for the destination being like a local
1195 delivery, as in LMTP.
1197 I suspect this new check makes the old local delivery cutoff check
1198 redundant, but I have not verified this so I left the code in place.
1200 TF/02 Correct gecos expansion when From: is a prefix of the username.
1202 Test 0254 submits a message to Exim with the header
1206 When I ran the test suite under the user fanf2, Exim expanded
1207 the header to contain my full name, whereas it should have added
1208 a Resent-Sender: header. It erroneously treats any prefix of the
1209 username as equal to the username.
1211 This change corrects that bug.
1213 GF/01 DCC debug and logging tidyup
1214 Error conditions log to paniclog rather than rejectlog.
1215 Debug lines prefixed by "DCC: " to remove any ambiguity.
1217 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1219 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1220 Bug spotted by Jeremy Harris; was flawed since initial commit.
1221 Would have resulted in OCSP responses post-SNI triggering an Exim
1222 NULL dereference and crash.
1224 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1226 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1227 Bug detection, analysis and fix by Samuel Thibault.
1228 Bugzilla 1331, Debian bug #698092.
1230 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1232 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1233 Server implementation by Todd Lyons, client by JH.
1234 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1235 config variable "prdr_enable" controls whether the server
1236 advertises the facility. If the client requests PRDR a new
1237 acl_data_smtp_prdr ACL is called once for each recipient, after
1238 the body content is received and before the acl_smtp_data ACL.
1239 The client is controlled by both of: a hosts_try_prdr option
1240 on the smtp transport, and the server advertisement.
1241 Default client logging of deliveries and rejections involving
1242 PRDR are flagged with the string "PRDR".
1244 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1245 fclose(). Diagnosis by Todd Lyons.
1247 PP/17 Update configure.default to handle IPv6 localhost better.
1248 Patch by Alain Williams (plus minor tweaks).
1251 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1252 This is now consistent with GnuTLS, and is now documented: the
1253 previous undocumented portable approach to treating the option as
1254 unset was to force an expansion failure. That still works, and
1255 an empty string is now equivalent.
1257 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1258 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1259 not performing validation itself.
1261 PP/20 Added force_command boolean option to pipe transport.
1262 Patch from Nick Koston, of cPanel Inc.
1264 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1267 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1269 PP/21 Fix eximon continuous updating with timestamped log-files.
1270 Broken in a format-string cleanup in 4.80, missed when I repaired the
1271 other false fix of the same issue.
1272 Report and fix from Heiko Schlichting.
1275 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1276 Report from Prashanth Katuri.
1278 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1279 It's SecureTransport, so affects any MacOS clients which use the
1280 system-integrated TLS libraries, including email clients.
1282 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1283 using a MIME ACL for non-SMTP local injection.
1284 Report and assistance in diagnosis by Warren Baker.
1286 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1288 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1290 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1291 last id that failed so it may be referenced in subsequent ACL's.
1293 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1296 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1297 ldap library initialization, allowing self-signed CA's to be
1298 used. Also properly sets require_cert option later in code by
1299 using NULL (global ldap config) instead of ldap handle (per
1300 session). Bug diagnosis and testing by alxgomz.
1302 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1303 the src/util/ subdirectory.
1305 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1306 renamed to Transport Post Delivery Action by Jeremy Harris, as
1309 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1310 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1311 redis_servers = needs to be configured which will be used by the redis
1312 lookup. Patch from Warren Baker, of The Packet Hub.
1314 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1316 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1317 hostname or reverse DNS when processing a host list. Used suggestions
1318 from multiple comments on this bug.
1320 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1322 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1323 Missed a few lines, added it to make the runtest require no keyboard
1326 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1327 contains upper case chars. Make router use caseful_local_part.
1329 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1330 support when GnuTLS has been built with p11-kit.
1336 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1338 This, or similar/improved, will also be change PP/11 of 4.82.
1344 PP/01 Handle short writes when writing local log-files.
1345 In practice, only affects FreeBSD (8 onwards).
1346 Bugzilla 1053, with thanks to Dmitry Isaikin.
1348 NM/01 Bugzilla 949 - Documentation tweak
1350 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1353 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1355 PP/02 Implemented gsasl authenticator.
1357 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1359 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1360 `pkg-config foo` for cflags/libs.
1362 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1363 with rest of GSASL and with heimdal_gssapi.
1365 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1366 `pkg-config foo` for cflags/libs for the TLS implementation.
1368 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1369 properties get this fed in as external SSF. A number of robustness
1370 and debugging improvements to the cyrus_sasl authenticator.
1372 PP/08 cyrus_sasl server now expands the server_realm option.
1374 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1375 Patch by Jeremy Harris.
1377 PP/10 Added dbmjz lookup type.
1379 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1381 PP/12 MAIL args handles TAB as well as SP, for better interop with
1382 non-compliant senders.
1383 Analysis and variant patch by Todd Lyons.
1385 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1386 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1387 Patch from Dirk Mueller <dmueller@suse.com>
1389 PP/13 tls_peerdn now print-escaped for spool files.
1390 Observed some $tls_peerdn in wild which contained \n, which resulted
1391 in spool file corruption.
1393 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1394 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1395 or write after TLS renegotiation, which otherwise led to messages
1398 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1399 as a tracking header (ie: a signed header comes before the signature).
1400 Patch from Wolfgang Breyha.
1402 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1403 comma-sep list; embedded commas doubled.
1405 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1407 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1409 Report and patch from Dmitry Banschikov.
1411 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1412 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1413 failures. We appear to now support TLS1.1+ with Exim.
1415 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1416 lets Exim select keys and certificates based upon TLS SNI from client.
1417 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1418 before an outbound SMTP session. New log_selector, +tls_sni.
1420 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1421 NULL dereference. Report and patch from Alun Jones.
1423 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1424 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1425 Not seeing resolver debug output on NetBSD, but suspect this is a
1426 resolver implementation change.
1428 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1429 Left warnings. Added "eximon gdb" invocation mode.
1431 PP/21 Defaulting "accept_8bitmime" to true, not false.
1433 PP/22 Added -bw for inetd wait mode support.
1435 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1436 locate the relevant includes and libraries. Made this the default.
1438 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1439 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1441 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1442 This may cause build issues on older platforms.
1444 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1445 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1446 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1447 Added SNI support via GnuTLS too.
1448 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1450 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1452 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1453 Applied second patch from Janne, implementing suggestion to default
1454 multiple-strings-in-record handling to match SPF spec.
1456 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1458 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1459 read-only, out of scope).
1460 Patch from Wolfgang Breyha, report from Stuart Northfield.
1462 PP/29 Fix three issues highlighted by clang analyser static analysis.
1463 Only crash-plausible issue would require the Cambridge-specific
1464 iplookup router and a misconfiguration.
1465 Report from Marcin Mirosław.
1467 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1469 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1470 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1471 As part of this, removing so much warning spew let me fix some minor
1472 real issues in debug logging.
1474 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1475 assignment on my part. Fixed.
1477 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1478 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1479 Janne Snabb (who went above and beyond: thank you).
1481 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1482 string otherwise requires a connection and a bunch more work and it's
1483 relatively easy to get wrong. Should also expose TLS library linkage
1486 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1487 64-bit ${eval} (JH/03).
1489 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1490 GNU libc to support some of the 64-bit stuff, should not lead to
1491 conflicts. Defined before os.h is pulled in, so if a given platform
1492 needs to override this, it can.
1494 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1495 protection layer was required, which is not implemented.
1496 Bugzilla 1254, patch from Wolfgang Breyha.
1498 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1499 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1500 tls_dhparam take prime identifiers. Also unbreak combination of
1501 OpenSSL+DH_params+TLSSNI.
1503 PP/39 Disable SSLv2 by default in OpenSSL support.
1509 PP/01 Solaris build fix for Oracle's LDAP libraries.
1510 Bugzilla 1109, patch from Stephen Usher.
1512 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1514 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1517 TF/02 Fix a couple more cases where we did not log the error message
1518 when unlink() failed. See also change 4.74-TF/03.
1520 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1521 lock up or crash if it happened to be inside a call to libc when it
1522 got a SIGUSR1 from exiwhat.
1524 The SIGUSR1 handler appends the current process status to the process
1525 log which is later printed by exiwhat. It used to use the general
1526 purpose logging code to do this, but several functions it calls are
1527 not safe for signals.
1529 The new output code in the SIGUSR1 handler is specific to the process
1530 log, and simple enough that it's easy to inspect for signal safety.
1531 Removing some special cases also simplifies the general logging code.
1532 Removing the spurious timestamps from the process log simplifies
1535 TF/04 Improved ratelimit ACL condition.
1537 The /noupdate option has been deprecated in favour of /readonly which
1538 has clearer semantics. The /leaky, /strict, and /readonly update modes
1539 are mutually exclusive. The update mode is no longer included in the
1540 database key; it just determines when the database is updated. (This
1541 means that when you upgrade Exim will forget old rate measurements.)
1543 Exim now checks that the per_* options are used with an update mode that
1544 makes sense for the current ACL. For example, when Exim is processing a
1545 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1546 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1547 must specify per_mail/readonly. If you omit the update mode it defaults to
1548 /leaky where that makes sense (as before) or /readonly where required.
1550 The /noupdate option is now undocumented but still supported for
1551 backwards compatibility. It is equivalent to /readonly except that in
1552 ACLs where /readonly is required you may specify /leaky/noupdate or
1553 /strict/noupdate which are treated the same as /readonly.
1555 A useful new feature is the /count= option. This is a generalization
1556 of the per_byte option, so that you can measure the throughput of other
1557 aggregate values. For example, the per_byte option is now equivalent
1558 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1560 The per_rcpt option has been generalized using the /count= mechanism
1561 (though it's more complicated than the per_byte equivalence). When it is
1562 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1563 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1564 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1565 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1566 Note that using per_rcpt with a non-readonly update mode in more than
1567 one ACL will cause the recipients to be double-counted. (The per_mail
1568 and per_byte options don't have this problem.)
1570 The handling of very low rates has changed slightly. If the computed rate
1571 is less than the event's count (usually one) then this event is the first
1572 after a long gap. In this case the rate is set to the same as this event's
1573 count, so that the first message of a spam run is counted properly.
1575 The major new feature is a mechanism for counting the rate of unique
1576 events. The new per_addr option counts the number of different
1577 recipients that someone has sent messages to in the last time period. It
1578 behaves like per_rcpt if all the recipient addresses are different, but
1579 duplicate recipient addresses do not increase the measured rate. Like
1580 the /count= option this is a general mechanism, so the per_addr option
1581 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1582 example, measure the rate that a client uses different sender addresses
1583 with the options per_mail/unique=$sender_address. There are further
1584 details in the main documentation.
1586 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1588 TF/06 Removed a few PCRE remnants.
1590 TF/07 Automatically extract Exim's version number from tags in the git
1591 repository when doing development or release builds.
1593 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1594 Bugzilla 879. Patch from Paul Fisher.
1596 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1597 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1600 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1602 PP/05 Variable $av_failed, true if the AV scanner deferred.
1603 Bugzilla 1078. Patch from John Horne.
1605 PP/06 Stop make process more reliably on build failure.
1606 Bugzilla 1087. Patch from Heiko Schlittermann.
1608 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1609 Bugzilla 1089. Patch from Heiko Schlittermann.
1611 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1612 Bugzilla 1131. Patch from Holger Weiß.
1614 PP/09 Handle IPv6 addresses with SPF.
1615 Bugzilla 860. Patch from Wolfgang Breyha.
1617 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1619 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1622 PP/11 match_* no longer expand right-hand-side by default.
1623 New compile-time build option, EXPAND_LISTMATCH_RHS.
1624 New expansion conditions, "inlist", "inlisti".
1626 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1628 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1630 PP/14 fix log_write() format string regression from TF/03.
1631 Bugzilla 1152. Patch from Dmitry Isaikin.
1637 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1639 PP/02 Harmonised TLS library version reporting; only show if debugging.
1640 Layout now matches that introduced for other libraries in 4.74 PP/03.
1642 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1644 PP/04 New "dns_use_edns0" global option.
1646 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1649 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1650 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1652 TK/01 Updated PolarSSL code to 0.14.2.
1653 Bugzilla 1097. Patch from Andreas Metzler.
1655 PP/07 Catch divide-by-zero in ${eval:...}.
1656 Fixes bugzilla 1102.
1658 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1661 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1662 format-string attack -- SECURITY: remote arbitrary code execution.
1664 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1665 time unintentionally subject to list matching rules, letting the header
1666 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1667 arbitrary string expansion). This allowed for information disclosure.
1669 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1670 INT_MIN/-1 -- value coerced to INT_MAX.
1676 NM/01 Workaround for PCRE version dependency in version reporting
1679 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1680 This fixes portability to compilers other than gcc, notably
1681 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1683 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1684 makefiles for portability to HP-UX and POSIX correctness.
1686 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1687 Also via indented variable definition in the Makefile.
1688 (Debugging by Oliver Heesakkers).
1690 PP/02 Restore caching of spamd results with expanded spamd_address.
1691 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1693 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1694 Improves build reliability. Fix from: Frank Elsner
1696 NM/02 Fix wide character breakage in the rfc2047 coding
1697 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1699 NM/03 Allow underscore in dnslist lookups
1700 Fixes bug 1026. Patch from Graeme Fowler
1702 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1703 Code patches from Adam Ciarcinski of NetBSD.
1705 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1708 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1709 is logged, to avoid truncation. Patch from John Horne.
1711 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1712 Patch from Jakob Hirsch.
1714 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1715 SQL string expansion failure details.
1716 Patch from Andrey Oktyabrski.
1718 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1719 Patch from Simon Arlott.
1721 PP/09 New lookups functionality failed to compile on old gcc which rejects
1722 extern declarations in function scope.
1723 Patch from Oliver Fleischmann
1725 PP/10 Use sig_atomic_t for flags set from signal handlers.
1726 Check getgroups() return and improve debugging.
1727 Fixed developed for diagnosis in bug 927 (which turned out to be
1730 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1731 Patch from Mark Zealey.
1733 PP/12 Bugzilla 1056: Improved spamd server selection.
1734 Patch from Mark Zealey.
1736 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1737 Based on patch from Heiko Schlittermann.
1739 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1740 Patch from Uwe Doering, sign-off by Michael Haardt.
1742 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1743 variable declaration deep within a block. Bug and patch from
1746 PP/15 lookups-Makefile IRIX compatibility coercion.
1748 PP/16 Make DISABLE_DKIM build knob functional.
1750 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1751 Patch by Simon Arlott
1753 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1754 variable argument macros. Our copy now differs from upstream.
1760 TF/01 Failure to get a lock on a hints database can have serious
1761 consequences so log it to the panic log.
1763 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1764 controlled using the smtp_confirmation log selector.
1766 TF/03 Include the error message when we fail to unlink a spool file.
1768 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1769 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1770 for maintaining out-of-tree patches for some time.
1772 PP/01 Bugzilla 139: Documentation and portability issues.
1773 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1774 Handle per-OS dynamic-module compilation flags.
1776 PP/02 Let /dev/null have normal permissions.
1777 The 4.73 fixes were a little too stringent and complained about the
1778 permissions on /dev/null. Exempt it from some checks.
1779 Reported by Andreas M. Kirchwitz.
1781 PP/03 Report version information for many libraries, including
1782 Exim version information for dynamically loaded libraries. Created
1783 version.h, now support a version extension string for distributors
1784 who patch heavily. Dynamic module ABI change.
1786 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1787 privilege escalation vulnerability whereby the Exim run-time user
1788 can cause root to append content of the attacker's choosing to
1791 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1794 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1795 If dropping privileges for untrusted macros, we disabled normal logging
1796 on the basis that it would fail; for the Exim run-time user, this is not
1797 the case, and it resulted in successful deliveries going unlogged.
1798 Fixed. Reported by Andreas Metzler.
1804 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1805 only prepend for the Resent-* case. Fixes regression introduced in
1806 Exim 4.70 by NM/22 for Bugzilla 607.
1808 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1809 increasing numbers of administrators be bitten by this.
1811 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1813 PP/03 Bugzilla 994: added openssl_options main configuration option.
1815 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1817 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1819 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1821 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1822 without a peer certificate, leading to a segfault because of an
1823 assumption that peers always have certificates. Be a little more
1824 paranoid. Problem reported by Martin Tscholak.
1826 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1827 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1828 NB: ClamAV planning to remove STREAM in "middle of 2010".
1829 CL also introduces -bmalware, various -d+acl logging additions and
1830 more caution in buffer sizes.
1832 PP/09 Implemented reverse_ip expansion operator.
1834 PP/10 Bugzilla 937: provide a "debug" ACL control.
1836 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1838 PP/12 Bugzilla 973: Implement --version.
1840 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1842 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1844 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1846 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1847 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1848 ignore trailing whitespace.
1850 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1852 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1855 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1856 Notification from Dr Andrew Aitchison.
1858 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1859 ExtendedDetectionInfo response format.
1860 Notification from John Horne.
1862 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1865 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1866 XSL and documented dependency on system catalogs, with examples of how
1869 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1872 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1873 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1874 configuration file which is writeable by the Exim user or group.
1876 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1877 of configuration files to cover files specified with the -C option if
1878 they are going to be used with root privileges, not just the default
1881 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1882 option (effectively making it always true).
1884 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1885 files to be used while preserving root privileges.
1887 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1888 that rogue child processes cannot use them.
1890 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1891 run-time user, instead of root.
1893 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1894 Exim run-time user without dropping privileges.
1896 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1897 result string, instead of calling string_vformat() twice with the same
1900 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1901 for other users. Others should always drop root privileges if they use
1902 -C on the command line, even for a whitelisted configure file.
1904 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1906 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1912 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1913 $data_path, and $header_path variables; fixed documentation bugs and
1916 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1917 exipick to access non-standard spools, including the "frozen" queue
1920 NM/01 Bugzilla 965: Support mysql stored procedures.
1921 Patch from Alain Williams
1923 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1925 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1926 Patch from Andreas Metzler
1928 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1929 Patch from Kirill Miazine
1931 NM/05 Bugzilla 671: Added umask to procmail example.
1933 JJ/03 installed exipick 20100323.0, fixing doc bug
1935 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1936 directory. Notification and patch from Dan Rosenberg.
1938 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1940 TK/02 Improve log output when DKIM signing operation fails.
1942 MH/01 Treat the transport option dkim_domain as a colon separated
1943 list, not as a single string, and sign the message with each element,
1944 omitting multiple occurences of the same signer.
1946 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1947 Bugzilla 985, 986. Patch by Simon Arlott
1949 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1950 Patch by Simon Arlott
1952 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1953 MBX locking. Notification from Dan Rosenberg.
1959 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1961 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1963 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1965 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1967 NM/04 Bugzilla 915: Flag broken perl installation during build.
1973 TK/01 Added patch by Johannes Berg that expands the main option
1974 "spamd_address" if it starts with a dollar sign.
1976 TK/02 Write list of recipients to X-Envelope-Sender header when building
1977 the mbox-format spool file for content scanning (suggested by Jakob
1980 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1981 (http://www.dcc-servers.net/) support via dccifd. Activated by
1982 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1984 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1985 by Mark Daniel Reidel <mr@df.eu>.
1987 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1988 When building exim an external PCRE library is now needed -
1989 PCRE is a system library on the majority of modern systems.
1990 See entry on PCRE_LIBS in EDITME file.
1992 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1993 conversation. Added nologin parameter to request.
1994 Patch contributed by Kirill Miazine.
1996 TF/01 Do not log submission mode rewrites if they do not change the address.
1998 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2000 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2001 log files in place. Contributed by Roberto Lima.
2003 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2005 TF/03 Bugzilla 615: When checking the local_parts router precondition
2006 after a local_part_suffix or local_part_prefix option, Exim now
2007 does not use the address's named list lookup cache, since this
2008 contains cached lookups for the whole local part.
2010 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2011 Robert Millan. Documentation is in experimental-spec.txt.
2013 TF/04 Bugzilla 668: Fix parallel build (make -j).
2015 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2017 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2018 Patch provided by Jan Srzednicki.
2020 TF/05 Leading white space used to be stripped from $spam_report which
2021 wrecked the formatting. Now it is preserved.
2023 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2024 that they are available at delivery time.
2026 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2028 TF/08 TLS error reporting now respects the incoming_interface and
2029 incoming_port log selectors.
2031 TF/09 Produce a more useful error message if an SMTP transport's hosts
2032 setting expands to an empty string.
2034 NM/06 Bugzilla 744: EXPN did not work under TLS.
2035 Patch provided by Phil Pennock.
2037 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2038 Patch provided by Richard Godbee.
2040 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2041 acl_smtp_notquit, added index entry.
2043 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2044 Patch provided by Eugene Bujak.
2046 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2047 accept(). Patch provided by Maxim Dounin.
2049 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2050 Patch provided by Phil Pennock.
2052 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2054 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2055 Patch provided by Brad "anomie" Jorsch.
2057 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2058 Patch provided by Dean Brooks.
2060 TK/05 Add native DKIM support (does not depend on external libraries).
2062 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2063 Patch provided by Graeme Fowler.
2065 NM/16 Bugzilla 851: Documentation example syntax fix.
2067 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2069 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2072 NM/19 Bugzilla 745: TLS version reporting.
2073 Patch provided by Phil Pennock.
2075 NM/20 Bugzilla 167: bool: condition support.
2076 Patch provided by Phil Pennock.
2078 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2079 clients. Patch provided by Phil Pennock.
2081 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2082 Patch provided by Brad "anomie" Jorsch.
2084 NM/23 Bugzilla 687: Fix misparses in eximstats.
2085 Patch provided by Heiko Schlittermann.
2087 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2088 Patch provided by Heiko Schlittermann.
2090 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2091 plus update to original patch.
2093 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2095 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2096 Patch provided by David Brownlee.
2098 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2100 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2102 NM/30 Bugzilla 888: TLS documentation bugfixes.
2104 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2106 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2107 Unlike the original bugzilla I have changed all shell scripts in src tree.
2109 NM/33 Bugzilla 898: Transport filter timeout fix.
2110 Patch by Todd Rinaldo.
2112 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2113 Patch by Serge Demonchaux.
2115 NM/35 Bugzilla 39: Base64 decode bug fixes.
2116 Patch by Jakob Hirsch.
2118 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2120 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2122 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2124 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2130 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2131 ALT-N's libdkim that I have put here:
2132 http://duncanthrax.net/exim-experimental/
2134 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2135 were called 'true' and it seems that C99 defines that as a reserved
2136 keyword to be used with 'bool' variable types. That means you could
2137 not include C99-style headers which use bools without triggering
2138 build errors in sieve.c.
2140 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2141 as mailq or other aliases. Changed the --help handling significantly
2142 to do whats expected. exim_usage() emits usage/help information.
2144 SC/01 Added the -bylocaldomain option to eximstats.
2146 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2148 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2150 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2156 PH/01 Another patch from the Sieve maintainer.
2158 PH/02 When an IPv6 address is converted to a string for single-key lookup
2159 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2160 dots are used instead of colons so that keys in lsearch files need not
2161 contain colons. This was done some time before quoting was made available
2162 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2163 (notated using the quote facility) so as to distinguish them from IPv4
2164 keys. This meant that lookups for IP addresses in host lists did not work
2165 for iplsearch lookups.
2167 This has been fixed by arranging for IPv6 addresses to be expressed with
2168 colons if the lookup type is iplsearch. This is not incompatible, because
2169 previously such lookups could never work.
2171 The situation is now rather anomalous, since one *can* have colons in
2172 ordinary lsearch keys. However, making the change in all cases is
2173 incompatible and would probably break a number of configurations.
2175 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2178 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2179 conversion specification without a maximum field width, thereby enabling
2180 a rogue spamd server to cause a buffer overflow. While nobody in their
2181 right mind would setup Exim to query an untrusted spamd server, an
2182 attacker that gains access to a server running spamd could potentially
2183 exploit this vulnerability to run arbitrary code as the Exim user.
2185 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2186 $primary_hostname instead of what libspf2 thinks the hosts name is.
2188 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2189 a directory entry by the name of the lookup key. Previously, if a
2190 symlink pointed to a non-existing file or a file in a directory that
2191 Exim lacked permissions to read, a lookup for a key matching that
2192 symlink would fail. Now it is enough that a matching directory entry
2193 exists, symlink or not. (Bugzilla 503.)
2195 PH/03 The body_linecount and body_zerocount variables are now exported in the
2198 PH/04 Added the $dnslist_matched variable.
2200 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2201 This means they are set thereafter only if the connection becomes
2204 PH/06 Added the client_condition to authenticators so that some can be skipped
2205 by clients under certain conditions.
2207 PH/07 The error message for a badly-placed control=no_multiline_responses left
2208 "_responses" off the end of the name.
2210 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2212 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2213 (without spaces) instead of just copying the configuration text.
2215 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2217 PH/11 Added $max_received_linelength.
2219 PH/12 Added +ignore_defer and +include_defer to host lists.
2221 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2222 way in which PCRE > 7.0 is built.
2224 PH/14 Implemented queue_only_load_latch.
2226 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2227 MAIL command. The effect was to mangle the value on 64-bit systems.
2229 PH/16 Another patch from the Sieve maintainer.
2231 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2233 PH/18 If a system quota error occurred while trying to create the file for
2234 a maildir delivery, the message "Mailbox is full" was not appended to the
2235 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2236 only to a quota excession during the actual writing of the file.
2238 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2239 characters?) which causes problems in log lines. The DN values are now
2240 passed through string_printing() before being added to log lines.
2242 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2243 and InterBase are left for another time.)
2245 PH/21 Added message_body_newlines option.
2247 PH/22 Guard against possible overflow in moan_check_errorcopy().
2249 PH/23 POSIX allows open() to be a macro; guard against that.
2251 PH/24 If the recipient of an error message contained an @ in the local part
2252 (suitably quoted, of course), incorrect values were put in $domain and
2253 $local_part during the evaluation of errors_copy.
2259 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2260 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2263 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2264 issue a MAIL command.
2266 PH/02 In an ACL statement such as
2268 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2270 if a client was not listed at all, or was listed with a value other than
2271 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2272 the condition was not true (as it should be), so access was not denied.
2273 The bug was that the ! inversion was incorrectly passed on to the second
2274 item. This has been fixed.
2276 PH/03 Added additional dnslists conditions == and =& which are different from
2277 = and & when the dns lookup returns more than one IP address.
2279 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2280 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2282 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2283 FSYNC, which compiles an option called disable_fsync that allows for
2284 bypassing fsync(). The documentation is heavily laced with warnings.
2286 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2288 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2289 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2290 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2291 including adding "make clean"; (3) Added -fPIC when compiling the test
2292 dynamically loaded module, to get rid of a warning.
2294 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2295 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2296 = 0s. The bug is otherwise harmless.
2298 PH/07 There was a bug in the dovecot authenticator such that the value of
2299 $auth1 could be overwritten, and so not correctly preserved, after a
2300 successful authentication. This usually meant that the value preserved by
2301 the server_setid option was incorrect.
2303 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2305 PH/09 Installed PCRE release 7.0.
2307 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2308 run for batched SMTP input. It is now run at the start of every message
2309 in the batch. While fixing this I discovered that the process information
2310 (output by running exiwhat) was not always getting set for -bs and -bS
2311 input. This is fixed, and it now also says "batched" for BSMTP.
2313 PH/11 Added control=no_pipelining.
2315 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2316 patch, slightly modified), and move the expansion of helo_data till after
2317 the connection is made in the smtp transport (so it can use these
2320 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2322 PH/14 Added log_selector = +pid.
2324 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2326 PH/16 Add ${if forany and ${if forall.
2328 PH/17 Added dsn_from option to vary the From: line in DSNs.
2330 PH/18 Flush SMTP output before performing a callout, unless control =
2331 no_callout_flush is set.
2333 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2334 was true (the default) a successful delivery failed to delete the retry
2335 item, thus causing premature timeout of the address. The bug is now
2338 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2340 PH/21 Long custom messages for fakedefer and fakereject are now split up
2341 into multiline responses in the same way that messages for "deny" and
2342 other ACL rejections are.
2344 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2345 with slight modification.
2347 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2348 draft, changing the syntax and factoring some duplicate code".
2350 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2351 for deliveries of the second and subsequent messages over the same SMTP
2354 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2355 ${reduce, with only minor "tidies".
2357 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2359 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2360 expansion side effects.
2362 PH/27 When a message times out after an over-quota error from an Exim-imposed
2363 quota, the bounce message says "mailbox is full". This message was not
2364 being given when it was a system quota that was exceeded. It now should
2367 MH/03 Made $recipients available in local_scan(). local_scan() already has
2368 better access to the recipient list through recipients_list[], but
2369 $recipients can be useful in postmaster-provided expansion strings.
2371 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2372 in the case of a MAIL command with additional options following the
2373 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2374 were accidentally chopped off.
2376 PH/29 SMTP synchronization checks are implemented when a command is read -
2377 there is a check that no more input is waiting when there shouldn't be
2378 any. However, for some commands, a delay in an ACL can mean that it is
2379 some time before the response is written. In this time, more input might
2380 arrive, invalidly. So now there are extra checks after an ACL has run for
2381 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2382 pipelining has not been advertised.
2384 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2386 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2387 returned in the SMTP response when smtp_return_error_details was false.
2388 This has been fixed.
2390 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2391 instead of the C I/O that was originally supplied, because problems were
2392 reported on Solaris.
2394 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2395 Exim which did not show up earlier: it was assuming that a call to
2396 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2397 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2398 was a macro that became an assignment, so it seemed to work. This has
2399 changed to a proper function call with a void return, hence the compile
2400 error. Exim's code has been fixed.
2402 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2405 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2407 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2409 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2410 because it turns out to be too big or there is a timeout) there is no
2411 "Completed" line in the log. When some messages of this type were
2412 selected by exigrep, they were listed as "not completed". Others were
2413 picked up by some special patterns. I have improved the selection
2414 criteria to be more general.
2416 PH/38 The host_find_failed option in the manualroute router can now be set
2417 to "ignore", to completely ignore a host whose IP address cannot be
2418 found. If all hosts are ignored, the behaviour is controlled by the new
2419 host_all_ignored option.
2421 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2422 homing or because of multiple MX records with /mx) generated more than
2423 one IP address, and the following item turned out to be the local host,
2424 all the secondary addresses of the first item were incorrectly removed
2425 from the list, along with the local host and any following hosts (which
2426 is what is supposed to happen).
2428 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2429 whoever called Exim into the -H file. In the case of the daemon it was
2430 behaving confusingly. When first started, it used values for whoever
2431 started the daemon, but after a SIGHUP it used the Exim user (because it
2432 calls itself on a restart). I have changed the code so that it now always
2435 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2436 message are rejected with the same error (e.g. no authentication or bad
2437 sender address), and a DATA command is nevertheless sent (as can happen
2438 with PIPELINING or a stupid MUA), the error message that was given to the
2439 RCPT commands is included in the rejection of the DATA command. This is
2440 intended to be helpful for MUAs that show only the final error to their
2443 PH/42 Another patch from the Sieve maintainer.
2445 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2446 Eximstats - Fixed some broken HTML links and added missing column headers
2448 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2449 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2451 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2453 PH/43 Yet another patch from the Sieve maintainer.
2455 PH/44 I found a way to check for a TCP/IP connection going away before sending
2456 the response to the final '.' that terminates a message, but only in the
2457 case where the client has not sent further data following the '.'
2458 (unfortunately, this is allowed). However, in many cases there won't be
2459 any further data because there won't be any more messages to send. A call
2460 to select() can be used: if it shows that the input is "ready", there is
2461 either input waiting, or the socket has been closed. An attempt to read
2462 the next input character can distinguish the two cases. Previously, Exim
2463 would have sent an OK response which the client would never have see.
2464 This could lead to message repetition. This fix should cure that, at
2465 least in a lot of common cases.
2467 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2468 advertised in response to EHLO.
2474 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2475 fixed by 4.65/MH/01 (is this a record?) are fixed:
2477 (i) An empty string was always treated as zero by the numeric comparison
2478 operators. This behaviour has been restored.
2480 (ii) It is documented that the numeric comparison operators always treat
2481 their arguments as decimal numbers. This was broken in that numbers
2482 starting with 0 were being interpreted as octal.
2484 While fixing these problems I realized that there was another issue that
2485 hadn't been noticed. Values of message_size_limit (both the global option
2486 and the transport option) were treated as octal if they started with 0.
2487 The documentation was vague. These values are now always treated as
2488 decimal, and I will make that clear in the documentation.
2494 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2495 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2498 MH/01 Don't check that the operands of numeric comparison operators are
2499 integers when their expansion is in "skipping" mode (fixes bug
2500 introduced by 4.64-PH/07).
2502 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2503 child addresses, Exim now panics and dies. Previously, because the count
2504 is held in a short int, deliveries were likely to be lost. As such a
2505 large number of recipients for a single message is ridiculous
2506 (performance will be very, very poor), I have chosen to impose a limit
2507 rather than extend the field.
2513 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2514 leftover -K file (the existence of which was triggered by #402).
2515 While we were at it, introduced process PID as part of the -K
2516 filename. This should rule out race conditions when creating
2519 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2520 processing considerably. Previous code took too long for large mails,
2521 triggering a timeout which in turn triggers #401.
2523 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2524 in the DK code in transports.c. sendfile() is not really portable,
2525 hence the _LINUX specificness.
2527 TF/01 In the add_headers option to the mail command in an Exim filter,
2528 there was a bug that Exim would claim a syntax error in any
2529 header after the first one which had an odd number of characters
2532 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2533 callout verification, Exim cached a "reject" for the entire domain. This
2534 is correct for most verifications, but it is not correct for a recipient
2535 verification with use_sender or use_postmaster set, because in that case
2536 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2537 case of MAIL FROM:<> rejection from other early rejections (e.g.
2538 rejection of HELO). When verifying a recipient using a non-null MAIL
2539 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2540 Whatever the result of the callout, the value of the domain cache is
2541 left unchanged (for any other kind of callout, getting as far as trying
2542 RCPT means that the domain itself is ok).
2544 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2547 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2548 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2549 session, the daemon ignores ECONNECT errors and logs others; it now
2550 ignores EPIPE as well.
2552 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2553 (quoted-printable decoding).
2555 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2556 later the small subsequent patch to fix an introduced bug.
2558 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2560 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2562 PH/08 An error is now given if message_size_limit is specified negative.
2564 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2565 to be given (somewhat) arbitrary names.
2567 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2570 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2571 miscellaneous code fixes
2573 PH/10 Added the log_reject_target ACL modifier to specify where to log
2576 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2577 hostname. This is wrong, because it relates to the incoming message (and
2578 probably the interface on which it is arriving) and not to the outgoing
2579 callout (which could be using a different interface). This has been
2580 changed to use the value of the helo_data option from the smtp transport
2581 instead - this is what is used when a message is actually being sent. If
2582 there is no remote transport (possible with a router that sets up host
2583 addresses), $smtp_active_hostname is used.
2585 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2586 tweaks were necessary in order to get it to work (see also 21 below):
2587 (a) The code assumed that strncpy() returns a negative number on buffer
2588 overflow, which isn't the case. Replaced with Exim's string_format()
2590 (b) There were several signed/unsigned issues. I just did the minimum
2591 hacking in of casts. There is scope for a larger refactoring.
2592 (c) The code used strcasecmp() which is not a standard C function.
2593 Replaced with Exim's strcmpic() function.
2594 (d) The code set only $1; it now sets $auth1 as well.
2595 (e) A simple test gave the error "authentication client didn't specify
2596 service in request". It would seem that Dovecot has changed its
2597 interface. Fortunately there's a specification; I followed it and
2598 changed what the client sends and it appears to be working now.
2600 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2603 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2604 address A is aliased to B and C, where B exists and C does not. Without
2605 -v the output is "A verified" because verification stops after a
2606 successful redirection if more than one address is generated. However,
2607 with -v the child addresses are also verified. Exim was outputting "A
2608 failed to verify" and then showing the successful verification for C,
2609 with its parentage. It now outputs "B failed to verify", showing B's
2610 parentage before showing the successful verification of C.
2612 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2613 look up a TXT record in a specific list after matching in a combined
2616 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2617 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2618 they consult the DNS. I had assumed they would set it the way they
2619 wanted; and indeed my experiments on Linux seem to show that in some
2620 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2621 To be on the safe side, however, I have now made the interface to
2622 host_find_byname() similar to host_find_bydns(), with an argument
2623 containing the DNS resolver options. The host_find_byname() function now
2624 sets these options at its start, just as host_find_bydns() does. The smtp
2625 transport options dns_qualify_single and dns_search_parents are passed to
2626 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2627 of host_find_byname() use the default settings of RES_DEFNAMES
2628 (qualify_single) but not RES_DNSRCH (search_parents).
2630 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2631 spool_read_header() do less string testing, by means of a preliminary
2632 switch on the second character of optional "-foo" lines. (This is
2633 overdue, caused by the large number of possibilities that now exist.
2634 Originally there were few.) While I was there, I also converted the
2635 str(n)cmp tests so they don't re-test the leading "-" and the first
2636 character, in the hope this might squeeze out yet more improvement.
2638 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2639 flag allowing group syntax was set by the header_syntax check but not
2640 turned off, possible causing trouble later; (2) The flag was not being
2641 set at all for the header_verify test, causing "group"-style headers to
2642 be rejected. I have now set it in this case, and also caused header_
2643 verify to ignore an empty address taken from a group. While doing this, I
2644 came across some other cases where the code for allowing group syntax
2645 while scanning a header line wasn't quite right (mostly, not resetting
2646 the flag correctly in the right place). These bugs could have caused
2647 trouble for malformed header lines. I hope it is now all correct.
2649 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2650 with the "reply" argument non-NULL. The code, however (which originally
2651 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2652 but it didn't always do it. This confused somebody who was copying the
2653 code for some other use. I have removed all the tests.
2655 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2656 feature that was used to support insecure browsers during the U.S. crypto
2657 embargo. It requires special client support, and Exim is probably the
2658 only MTA that supported it -- and would never use it because real RSA is
2659 always available. This code has been removed, because it had the bad
2660 effect of slowing Exim down by computing (never used) parameters for the
2661 RSA_EXPORT functionality.
2663 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2664 authenticator to fail if there's a tab character in the incoming data
2665 (there should never be unless someone is messing about, as it's supposed
2666 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2667 if the connection is using TLS or if the remote IP is the same as the
2668 local IP, and the "valid-client-cert option" if a client certificate has
2671 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2672 authenticators. This can be used for authorization after authentication
2673 succeeds. (In the case of plaintext, it servers for both authentication
2676 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2677 if any retry times were supplied.
2679 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2680 connection, where there is no client IP address to check. In this
2681 situation, the verify now always succeeds.
2683 PH/25 Applied John Jetmore's -Mset patch.
2685 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2687 PH/27 In a string expansion for a processed (not raw) header when multiple
2688 headers of the same name were present, leading whitespace was being
2689 removed from all of them, but trailing whitespace was being removed only
2690 from the last one. Now trailing whitespace is removed from each header
2691 before concatenation. Completely empty headers in a concatenation (as
2692 before) are ignored.
2694 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2695 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2697 PH/29 [Removed. This was a change that I later backed out, and forgot to
2698 correct the ChangeLog entry (that I had efficiently created) before
2699 committing the later change.]
2701 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2702 address errors (4xx response to RCPT) over the same connection as other
2703 messages routed to the same hosts. Such deliveries are always "forced",
2704 so retry times are not inspected. This resulted in far too many retries
2705 for the affected addresses. The effect occurred only when there were more
2706 hosts than the hosts_max_try setting in the smtp transport when it had
2707 the 4xx errors. Those hosts that it had tried were not added to the list
2708 of hosts for which the message was waiting, so if all were tried, there
2709 was no problem. Two fixes have been applied:
2711 (i) If there are any address or message errors in an SMTP delivery, none
2712 of the hosts (tried or untried) are now added to the list of hosts
2713 for which the message is waiting, so the message should not be a
2714 candidate for sending over the same connection that was used for a
2715 successful delivery of some other message. This seems entirely
2716 reasonable: after all the message is NOT "waiting for some host".
2717 This is so "obvious" that I'm not sure why it wasn't done
2718 previously. Hope I haven't missed anything, but it can't do any
2719 harm, as the worst effect is to miss an optimization.
2721 (ii) If, despite (i), such a delivery is accidentally attempted, the
2722 routing retry time is respected, so at least it doesn't keep
2723 hammering the server.
2725 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2726 in ${readsocket because some servers need this prod.
2728 PH/32 Added some extra debug output when updating a wait-xxx database.
2730 PH/33 The hint "could be header name not terminated by colon", which has been
2731 given for certain expansion errors for a long time, was not being given
2732 for the ${if def:h_colon_omitted{... case.
2734 PH/34 The spec says: "With one important exception, whenever a domain list is
2735 being scanned, $domain contains the subject domain." There was at least
2736 one case where this was not true.
2738 PH/35 The error "getsockname() failed: connection reset by peer" was being
2739 written to the panic log as well as the main log, but it isn't really
2740 panic-worthy as it just means the connection died rather early on. I have
2741 removed the panic log writing for the ECONNRESET error when getsockname()
2744 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2745 runs only) independently of the message's sender address. This meant
2746 that, if the 4xx error was in fact related to the sender, a different
2747 message to the same recipient with a different sender could confuse
2748 things. In particular, this can happen when sending to a greylisting
2749 server, but other circumstances could also provoke similar problems.
2750 I have changed the default so that the retry time for these errors is now
2751 based a combination of the sender and recipient addresses. This change
2752 can be overridden by setting address_retry_include_sender=false in the
2755 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2756 remote server are returned as part of bounce messages. This was not
2757 happening for LMTP over a pipe (the lmtp transport), but now it is the
2758 same for both kinds of LMTP.
2760 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2761 in header lines that were in fact CNAMEs. This is no longer the case.
2763 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2764 and queue runs started by the daemon processed all messages. This has
2765 been fixed so that -R and -S can now usefully be given with -q<time>.
2767 PH/40 Import PCRE release 6.7 (fixes some bugs).
2769 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2771 PH/42 Give an error if -q is specified more than once.
2773 PH/43 Renamed the variables $interface_address and $interface_port as
2774 $received_ip_address and $received_port, to make it clear that these
2775 values apply to message reception, and not to the outgoing interface when
2776 a message is delivered. (The old names remain recognized, of course.)
2778 PH/44 There was no timeout on the connect() call when using a Unix domain
2779 socket in the ${readsocket expansion. There now is.
2781 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2782 be meaningful with "accept".
2784 SC/01 Eximstats V1.43
2785 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2787 SC/02 Eximstats V1.44
2788 Use a glob alias rather than an array ref in the generated
2789 parser. This improves both readability and performance.
2791 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2792 Collect SpamAssassin and rejection statistics.
2793 Don't display local sender or destination tables unless
2794 there is data to show.
2795 Added average volumes into the top table text output.
2797 SC/04 Eximstats V1.46
2798 Collect data on the number of addresses (recipients)
2799 as well as the number of messages.
2801 SC/05 Eximstats V1.47
2802 Added 'Message too big' to the list of mail rejection
2803 reasons (thanks to Marco Gaiarin).
2805 SC/06 Eximstats V1.48
2806 Mainlog lines which have GMT offsets and are too short to
2807 have a flag are now skipped.
2809 SC/07 Eximstats V1.49 (Alain Williams)
2810 Added the -emptyok flag.
2812 SC/08 Eximstats V1.50
2813 Fixes for obtaining the IP address from reject messages.
2815 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2816 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2817 whitespace changes from 4.64-PH/27
2819 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2822 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2823 are found, allow negative numbers in numeric criteria)
2825 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2827 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2828 to match changes made in 4.64-PH/43
2830 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2832 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2833 "domain" is actually an IP address, and force a failure. This locks out
2834 those revolvers/nameservers that support "A-for-A" lookups, in
2835 contravention of the specifications.
2837 PH/48 When a host name was looked up from an IP address, and the subsequent
2838 forward lookup of the name timed out, the host name was left in
2839 $sender_host_name, contrary to the specification.
2841 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2842 restricted to single-key lookups, Exim was not diagnosing an error if
2843 * or *@ was used with a query-style lookup.
2845 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2847 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2848 long ago, but noone interested enough thought of it. Let's just say that
2849 the "1.1" means that there are some new functions that weren't there at
2850 some point in the past.
2852 PH/51 Error processing for expansion failure of helo_data from an smtp
2853 transport during callout processing was broken.
2855 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2856 tested/used via the -bh/-bhc/-bs options.
2858 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2859 bug, fixed in subsequent PCRE releases).
2861 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2862 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2864 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2870 SC/01 Use a glob alias rather than an array ref in eximstats generated
2871 parser. This improves both readability and performance.
2873 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2874 Don't display local sender or destination tables in eximstats unless
2875 there is data to show.
2876 Added average volumes into the eximstats top table text output.
2878 SC/03 Collect data on the number of addresses (recipients) as well
2879 as the number of messages in eximstats.
2881 TF/01 Correct an error in the documentation for the redirect router. Exim
2882 does (usually) call initgroups() when daemonizing.
2884 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2885 with consistent privilege compared to when running as a daemon.
2887 TF/03 Note in the spec that $authenticated_id is not set for local
2888 submissions from trusted users.
2890 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2891 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2893 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2894 by adding some example configuration directives to the default
2895 configuration file. A little bit of work is required to uncomment the
2896 directives and define how usernames and passwords are checked, but
2897 there is now a framework to start from.
2899 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2900 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2901 without this. I don't know how relevant this is to other LDAP libraries.
2903 PH/02 Add the verb name to the "unknown ACL verb" error.
2905 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2907 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2909 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2910 directory not expanded when it should be if an expanded home directory
2911 was set for the address (which is overridden by the transport).
2913 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2916 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2917 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2918 because it is too late at that time, and has no effect.
2920 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2921 security issue with \' (bugzilla #107). I could not use the
2922 PQescapeStringConn() function, because it needs a PGconn value as one of
2925 PH/08 When testing addresses using -bt, indicate those final addresses that
2926 are duplicates that would not cause an additional delivery. At least one
2927 person was confused, thinking that -bt output corresponded to deliveries.
2928 (Suppressing duplicates isn't a good idea as you lose the information
2929 about possibly different redirections that led to the duplicates.)
2931 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2932 systems where poll() doesn't work, in particular OS X.
2934 PH/10 Added more information to debugging output for retry time not reached.
2936 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2937 operations in malware.c.
2939 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2942 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2943 syslog_duplication set false, log lines that would normally be written
2944 both the the main log and to the reject log were not written to syslog at
2947 PH/14 In the default configuration, change the use of "message" in ACL warn
2948 statements to "add_header".
2950 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2951 not followed by a command (e.g. "seen endif").
2953 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2954 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2957 PH/17 Added extra conditions to the default value of delay_warning_condition
2961 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2962 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2963 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2966 The Auto-Submitted: and various List- headers are standardised, whereas I
2967 don't think Precedence: ever was.
2969 PH/18 Refactored debugging code in route_finduser() to show more information,
2970 in particular, the error code if getpwnam() issues one.
2972 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2973 This is apparently needed in addition to the PH/07 change above to avoid
2974 any possible encoding problems.
2976 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2977 but not after initializing Perl.
2979 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2980 output them only if debugging. By default they are written stderr,
2981 apparently, which is not desirable.
2983 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2986 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2989 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2991 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2992 authenticated or an ident call has been made. Suppress the default
2993 values for $authenticated_id and $authenticated_sender (but permit -oMai
2994 and -oMas) when testing with -bh.
2996 PH/24 Re-jigged the order of the tests in the default configuration so that the
2997 tests for valid domains and recipients precede the DNS black list and CSA
2998 tests, on the grounds that those ones are more expensive.
3000 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3001 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3002 This bug exists in every version of Exim that I still have, right back to
3005 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3006 However, an attempt to turn on case-sensitivity in a regex key by
3007 including (?-i) didn't work because the subject string was already
3008 lowercased, and the effects were non-intuitive. It turns out that a
3009 one-line patch can be used to allow (?-i) to work as expected.
3015 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3016 other effects) broke the use of negated acl sub-conditions.
3018 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3021 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3022 "Deny" causes Exim to reject the incoming connection with a 554 error.
3023 Unfortunately, if there is a major crisis, such as a disk failure,
3024 tcp-wrappers gives "deny", whereas what one would like would be some
3025 kind of temporary error. A kludge has been added to help with this.
3026 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3027 554 error is used if errno is still zero or contains ENOENT (which occurs
3028 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3031 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3033 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3034 errors. Otherwise a message that provokes a temporary error (when other
3035 messages do not) can cause a whole host to time out.
3037 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3038 addresses were routed directly to files or pipes from a redirect router.
3039 File deliveries just didn't batch; pipe deliveries might have suffered
3042 PH/06 A failure to get a lock for a hints database would erroneously always say
3043 "Failed to get write lock", even when it was really a read lock.
3045 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3046 of 0600. This has been changed to use the value of the lockfile_mode
3047 option (which defaults to 0600).
3049 PH/08 Applied small patch from the Sieve maintainer.
3051 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3052 folder from quota calculations, a direct delivery into this folder messed
3053 up the contents of the maildirsize file. This was because the regex was
3054 used only to exclude .Trash (or whatever) when the size of the mailbox
3055 was calculated. There was no check that a delivery was happening into an
3056 excluded directory. This bug has been fixed by ignoring all quota
3057 processing for deliveries into excluded directories.
3059 PH/10 Added the maildirfolder_create_regex option to appendfile.
3065 PH/01 The code for finding all the local interface addresses on a FreeBSD
3066 system running IPv6 was broken. This may well have applied to all BSD
3067 systems, as well as to others that have similar system calls. The broken
3068 code found IPv4 interfaces correctly, but gave incorrect values for the
3069 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3070 that it would not match correctly against @[] and not recognize the IPv6
3073 PH/02 The ipliteral router was not recognizing addresses of the form user@
3074 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3076 PH/03 Added disable_ipv6.
3078 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3079 decoded form, because it is most often used to construct To: headers
3080 lines in autoreplies, and the decoded form may well be syntactically
3081 invalid. However, $reply_address has leading white space removed, and all
3082 newlines turned into spaces so that the autoreply transport does not
3085 PH/05 If group was specified without a user on a router, and no group or user
3086 was specified on a transport, the group from the router was ignored.
3088 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3089 for visible compile-time settings that can be used to change these
3090 numbers, for those that want even more. Backwards compatibility with old
3091 spool files has been maintained. However, going back to a previous Exim
3092 release will lost any variables that are in spool files.
3094 PH/07 Two small changes when running in the test harness: increase delay when
3095 passing a TCP/IP connection to a new process, in case the original
3096 process has to generate a bounce, and remove special handling of
3097 127.0.0.2 (sic), which is no longer necessary.
3099 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3100 be the same on different OS.
3102 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3105 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3106 whether --show-vars was specified or not
3108 JJ/02 exipick: Added support for new ACL variable spool format introduced
3111 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3112 syntactically invalid From: or Reply-to: line, and a filter used this to
3113 generate an autoreply, and therefore failed to obtain an address for the
3114 autoreply, Exim could try to deliver to a non-existent relative file
3115 name, causing unrelated and misleading errors. What now happens is that
3116 it logs this as a hard delivery error, but does not attempt to create a
3119 PH/11 The exinext utility has a -C option for testing purposes, but although
3120 the given file was scanned by exinext itself; it wasn't being passed on
3121 when Exim was called.
3123 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3124 an end-of-file indication when reading a command response.
3126 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3127 compiled. In many other places in Exim, IPv6 addresses are always
3128 recognized, so I have changed this. It also means that IPv4 domain
3129 literals of the form [IPV4:n.n.n.n] are now always recognized.
3131 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3132 used if the router is not running as root, for example, when verifying at
3133 ACL time, or when using -bh. The debugging output from this situation was
3134 non-existent - all you got was a failure to exec. I have made two
3137 (a) Failures to set uid/gid, the current directory, or a process leader
3138 in a subprocess such as that created by queryprogram now generate
3139 suitable debugging output when -d is set.
3141 (b) The queryprogram router detects when it is not running as root,
3142 outputs suitable debugging information if -d is set, and then runs
3143 the subprocess without attempting to change uid/gid.
3145 PH/15 Minor change to Makefile for building test_host (undocumented testing
3148 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3149 additional section of a DNS packet that returns MX or SRV records.
3150 Instead, it always explicitly searches for A/AAAA records. This avoids
3151 major problems that occur when a DNS server includes only records of one
3152 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3153 fixed another bug: if SRV records were looked up and the corresponding
3154 address records were *not* found in the additional section, the port
3155 values from the SRV records were lost.
3157 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3158 using the correct key (the original address) when searching the retry
3159 rules in order to find which one to use for generating the retry hint.
3161 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3162 adding the default one. Similarly, if it contains a Reply-To: header, the
3163 errors_reply_to option, if set, is not used.
3165 PH/19 When calculating a retry time, Exim used to measure the "time since
3166 failure" by looking at the "first failed" field in the retry record. Now
3167 it does not use this if it is later than than the arrival time of the
3168 message. Instead it uses the arrival time. This makes for better
3169 behaviour in cases where some deliveries succeed, thus re-setting the
3170 "first failed" field. An example is a quota failure for a huge message
3171 when small messages continue to be delivered. Without this change, the
3172 "time since failure" will always be short, possible causing more frequent
3173 delivery attempts for the huge message than are intended.
3174 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3176 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3177 $1, $2, $3) because the numerical variables can be reset during some
3178 expansion items (e.g. "match"), thereby losing the authentication data.
3180 PH/21 Make -bV show the size of off_t variables so that the test suite can
3181 decide whether to run tests for quotas > 2G.
3183 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3184 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3185 filecount value is greater than 2G or if a quota value is greater than 2G
3186 on a system where the size of off_t is not greater than 4, a panic error
3189 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3190 never match. The debug and -bh output now contains an explicit error
3191 message indicating a malformed IPv4 address or mask.
3193 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3194 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3195 PH/23 above applies.
3197 PH/25 Do not write to syslog when running in the test harness. The only
3198 occasion when this arises is a failure to open the main or panic logs
3199 (for which there is an explicit test).
3201 PH/26 Added the /no_tell option to "control=freeze".
3203 PH/27 If a host name lookup failed very early in a connection, for example, if
3204 the IP address matched host_lookup and the reverse lookup yielded a name
3205 that did not have a forward lookup, an error message of the form "no IP
3206 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3207 could be logged. Now it outputs the IP address instead of "NULL".
3209 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3210 allows the sender and the authenticated sender to be set when
3211 submitting a message from within Exim. Since child_open_exim() is
3212 documented for local_scan(), the new function should be too.
3214 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3215 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3216 results in an empty string is now treated as unset.
3218 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3220 PH/31 Added sender_verify_fail logging option.
3222 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3223 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3224 tidied the source and removed it altogether.
3226 PH/33 When a queue run was abandoned because the load average was too high, a
3227 log line was always written; now it is written only if the queue_run log
3228 selector is set. In addition, the log line for abandonment now contains
3229 information about the queue run such as the pid. This is always present
3230 in "start" and "stop" lines but was omitted from the "abandon" line.
3232 PH/34 Omit spaces between a header name and the colon in the error message that
3233 is given when verify = headers_syntax fails (if there are lots of them,
3234 the message gets confusing).
3236 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3237 names, as there are now some PTR records that contain slashes. This check
3238 is only to protect against broken name servers that fall over on strange
3239 characters, so the fact that it applies to all lookups doesn't matter.
3241 PH/36 Now that the new test suite is complete, we can remove some of the
3242 special code in Exim that was needed for the old test suite. For example,
3243 sorting DNS records because real resolvers return them in an arbitrary
3244 order. The new test suite's fake resolver always returns records in the
3247 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3248 bounces) except when queue_only is set, to avoid logging races between
3249 the different processes.
3251 PH/38 Panic-die if .include specifies a non-absolute path.
3253 PH/39 A tweak to the "H" retry rule from its user.
3255 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3256 a label. They prevented compilation on older perls.
3258 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3259 a warning to be raised on newish perls.
3261 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3262 on queue. Changes to match documented behaviour of showing count of
3263 messages matching specified criteria.
3265 PH/40 Changed the default ident timeout from 30s to 5s.
3267 PH/41 Added support for the use of login_cap features, on those BSD systems
3268 that have them, for controlling the resources used by pipe deliveries.
3270 PH/42 The content-scanning code uses fopen() to create files in which to put
3271 message data. Previously it was not paying any attention to the mode of
3272 the files. Exim runs with umask(0) because the rest of the code creates
3273 files with open(), and sets the required mode explicitly. Thus, these
3274 files were ending up world-writeable. This was not a big issue, because,
3275 being within the spool directory, they were not world-accessible. I have
3276 created a function called modefopen, which takes an additional mode
3277 argument. It sets umask(777), creates the file, chmods it to the required
3278 mode, then resets the umask. All the relevant calls to fopen() in the
3279 content scanning code have been changed to use this function.
3281 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3282 to 24 hours. This avoids potential overflow problems when processing G
3283 and H retry rules. I suspect nobody ever tinkers with this value.
3285 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3287 PH/45 When the plaintext authenticator is running as a client, the server's
3288 challenges are checked to ensure they are valid base64 strings. By
3289 default, the authentication attempt is cancelled if an invalid string is
3290 received. Setting client_ignore_invalid_base64 true ignores these errors.
3291 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3292 they are received. Thus, the responses can be made to depend on the
3293 challenges. If an invalid string is ignored, an empty string is placed in
3296 PH/46 Messages that are created by the autoreply transport now contains a
3297 References: header, in accordance with RFCs 2822 and 3834.
3299 PH/47 Added authenticated_sender_force to the smtp transport.
3301 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3303 PH/49 Installed latest patch from the Sieve maintainer.
3305 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3306 was also set, the appendfile transport was unnecessarily scanning a
3307 directory of message files (e.g. for maildir delivery) to find the count
3308 of files (along with the size), even though it did not need this
3309 information. It now does the scan only if it needs to find either the
3310 size of the count of files.
3312 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3314 PH/52 Two bugs concerned with error handling when the smtp transport is
3317 (i) Exim was not creating retry information for temporary errors given
3318 for individual recipients after the DATA command when the smtp transport
3319 was used in LMTP mode. This meant that they could be retried too
3320 frequently, and not timed out correctly.
3322 (ii) Exim was setting the flag that allows error details to be returned
3323 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3324 recipients that were returned after the DATA command.
3326 PH/53 This is related to PH/52, but is more general: for any failing address,
3327 when detailed error information was permitted to be returned to the
3328 sender, but the error was temporary, then after the final timeout, only
3329 "retry timeout exceeded" was returned. Now it returns the full error as
3330 well as "retry timeout exceeded".
3332 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3333 do this, and (what is worse) MTAs that accept it.
3335 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3336 will now be deprecated.
3338 PH/56 New os.c-cygwin from the Cygwin maintainer.
3340 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3341 formats (previously only available in exim formats via -bpr, -bpru,
3342 and -bpra. Now also available in native and exiqgrep formats)
3344 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3345 with very large, slow to parse queues
3347 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3349 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3351 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3352 responses to authentication challenges, though it was showing the
3353 challenges; (ii) I've removed the CR characters from the debug output for
3356 PH/58 Allow for the insertion of a newline as well as a space when a string
3357 is turned into more than one encoded-word during RFC 2047 encoding. The
3358 Sieve code now uses this.
3360 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3361 data_4xx, lost_connection, tls_required.
3363 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3364 message was being sent as an SMTP response.
3366 PH/61 Add -l and -k options to exicyclog.
3368 PH/62 When verifying, if an address was redirected to one new address, so that
3369 verification continued, and the new address failed or deferred after
3370 having set something in $address_data, the value of $address_data was not
3371 passed back to the ACL. This was different to the case when no
3372 redirection occurred. The value is now passed back in both cases.
3374 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3375 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3376 use login_cap.h, so on its own it isn't the distinguishing feature. The
3377 new name refers directly to the setclassresources() function.
3379 PH/65 Added configuration files for NetBSD3.
3381 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3383 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3384 is preferred over IPv4.
3386 PH/68 The bounce_return_message and bounce_return_body options were not being
3387 honoured for bounces generated during the reception of non-SMTP messages.
3388 In particular, this applied to messages rejected by the ACL. This bug has
3389 been fixed. However, if bounce_return_message is true and bounce_return_
3390 body is false, the headers that are returned for a non-SMTP message
3391 include only those that have been read before the error was detected.
3392 (In the case of an ACL rejection, they have all been read.)
3394 PH/69 The HTML version of the specification is now built in a directory called
3395 spec_html instead of spec.html, because the latter looks like a path with
3396 a MIME-type, and this confuses some software.
3398 PH/70 Catch two compiler warnings in sieve.c.
3400 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3401 function verify_get_ident() calls ip_connect() to connect a socket, but
3402 if the "connect()" function timed out, ip_connect() used to close the
3403 socket. However, verify_get_ident() also closes the socket later, and in
3404 between Exim writes to the log, which may get opened at this point. When
3405 the socket was closed in ip_connect(), the log could get the same file
3406 descriptor number as the socket. This naturally causes chaos. The fix is
3407 not to close the socket in ip_connect(); the socket should be closed by
3408 the function that creates it. There was only one place in the code where
3409 this was missing, in the iplookup router, which I don't think anybody now
3410 uses, but I've fixed it anyway.
3412 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3413 well as to direct DNS lookups. Otherwise the handling of names in host
3414 lists is inconsistent and therefore confusing.
3420 PH/01 Two changes to the default runtime configuration:
3422 (1) Move the checks for relay_from_hosts and authenticated clients from
3423 after to before the (commented out) DNS black list checks.
3425 (2) Add control=submission to the relay_from_hosts and authenticated
3426 clients checks, on the grounds that messages accepted by these
3427 statements are most likely to be submissions.
3429 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3431 (1) Generate an error if the third argument for the ${prvs expansion is
3434 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3437 (3) Reset the variables that are obtained from the first argument of
3438 ${prvscheck and used in the second argument before leaving the code,
3439 because their memory is reclaimed, so using them afterwards may do
3442 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3443 one (it's much easier than Tom thought :-).
3445 (5) Because of (4), we can now allow for the use of $prvscheck_result
3446 inside the third argument.
3448 PH/03 For some reason, the default setting of PATH when running a command from
3449 a pipe transport was just "/usr/bin". I have changed it to
3452 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3453 anything to be listed in the output from -bV.
3455 PH/05 When a filter generated an autoreply, the entire To: header line was
3456 quoted in the delivery log line, like this:
3458 => >A.N.Other <ano@some.domain> <original@ddress> ...
3460 This has been changed so that it extracts the operative address. There
3461 may be more than one such address. If so, they are comma-separated, like
3464 => >ano@some.domain,ona@other.domain <original@ddress> ...
3466 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3467 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3468 not being looked up in the rDNS to get a host name, Exim was showing the
3469 IP address twice in Received: lines, even though the IP addresses were
3470 identical. For example:
3472 Received: from [1.2.3.4] (helo=[1.2.3.4])
3474 However, if the real host name was known, it was omitting the HELO data
3475 if it matched the actual IP address. This has been tidied up so that it
3476 doesn't show the same IP address twice.
3478 PH/07 When both +timestamp and +memory debugging was on, the value given by
3479 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3480 called by the debug printing, thereby overwriting the timestamp buffer.
3481 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3483 PH/08 When the original message was included in an autoreply transport, it
3484 always said "this is a copy of the message, including all the headers",
3485 even if body_only or headers_only was set. It now gives an appropriate
3488 PH/09 Applied a patch from the Sieve maintainer which:
3490 o fixes some comments
3491 o adds the (disabled) notify extension core
3492 o adds some debug output for the result of if/elsif tests
3493 o points to the current vacation draft in the documentation
3494 and documents the missing references header update
3498 o fixes a bug in processing the envelope test (when testing
3499 multiple envelope elements, the last element determined the
3502 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3503 Electronic Mail") by including:
3505 Auto-submitted: auto-generated
3507 in the messages that it generates (bounce messages and others, such as
3508 warnings). In the case of bounce messages for non-SMTP messages, there was
3509 also a typo: it was using "Auto_submitted" (underscore instead of
3510 hyphen). Since every message generated by Exim is necessarily in response
3511 to another message, thes have all been changed to:
3513 Auto-Submitted: auto-replied
3515 in accordance with these statements in the RFC:
3517 The auto-replied keyword:
3519 - SHOULD be used on messages sent in direct response to another
3520 message by an automatic process,
3522 - MUST NOT be used on manually-generated messages,
3524 - MAY be used on Delivery Status Notifications (DSNs) and Message
3525 Disposition Notifications (MDNs),
3527 - MUST NOT be used on messages generated by automatic or periodic
3528 processes, except for messages which are automatic responses to
3531 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3532 to the default Received: header definition.
3534 PH/12 Added log selector acl_warn_skipped (default on).
3536 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3537 variables because (a) they are in the wrong storage pool and (b) even if
3538 they were copied, it wouldn't work properly because of the caching.
3540 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3541 checking when decoding. Apparently there are clients that generate
3542 overlong encoded strings. Why am I not surprised?
3544 PH/15 If the first argument of "${if match_address" was not empty, but did not
3545 contain an "@" character, Exim crashed. Now it writes a panic log message
3546 and treats the condition as false.
3548 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3550 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3551 extension "envelope-auth". The code is finished and in agreement with
3552 other implementations, but there is no documentation so far and in fact,
3553 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3554 not changing the active code.
3556 Print executed "if" and "elsif" statements when debugging is used. This
3557 helps a great deal to understand what a filter does.
3559 Document more things not specified clearly in RFC3028. I had all this
3560 sorted out, when out of a sudden new issues came to my mind. Oops."
3562 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3565 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3566 canonical form (as documented). However, after a host name lookup from
3567 the IP address, check_host() was doing a simple string comparison with
3568 addresses acquired from the DNS when checking that the found name did
3569 have the original IP as one of its addresses. Since any found IPv6
3570 addresses are likely to be in abbreviated form, the comparison could
3571 fail. Luckily, there already exists a function for doing the comparison
3572 by converting both addresses to binary, so now that is used instead of
3573 the text comparison.
3575 PH/20 There was another similar case to PH/19, when a complete host name was
3576 given in a host list; looking up its IP address could give an abbreviated
3577 form, whereas the current host's name might or might not be abbreviated.
3578 The same fix has been applied.
3584 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3585 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3588 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3589 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3591 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3593 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3594 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3595 Also, for "Auto-Submitted", treat anything other than "no" as
3596 non-personal, in accordance with RFC 3834. (Previously it treated
3597 anything starting "auto-" as non-personal.)
3599 TF/01 The control=submission/name=... option had a problem with syntax
3600 errors if the name included a slash character. The /name= option
3601 now slurps the rest of the string, so it can include any characters
3602 but it must come last in the list of options (after /sender_retain
3605 PH/05 Some modifications to the interface to the fake nameserver for the new
3613 TK/01 Added the "success_on_redirect" address verification option. See
3614 NewStuff for rationale and an example.
3616 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3618 PH/02 Patch to exigrep to allow it to work on syslog lines.
3620 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3621 fread() to skip over the body file's header line, because in Cygwin the
3622 header line is locked and is inaccessible.
3624 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3625 co-exist for some time) to make it clear that it is the Exim ID that is
3626 referenced, not the Message-ID: header line.
3628 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3629 string_format() function, because snprintf() does not exist on all
3632 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3633 ${stat: expansion item.
3635 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3636 protocol synchronization error", to keep the pedants happy.
3638 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3639 well as for IRIX systems, when gcc is being used. See the host.c source
3642 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3644 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3647 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3648 transport and to the smtp transport in LMTP mode.
3650 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3652 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3653 run a filter in a subprocess. This could lead to confusion in subsequent
3654 lookups in the parent process. There should also be a search_tidyup() at
3655 the end of the subprocess.
3657 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3658 only if the host matched helo_try_verify_hosts, which caused the
3659 verification to occur when the EHLO/HELO command was issued. The ACL just
3660 tested the remembered result. Now, if a previous verification attempt has
3661 not happened, "verify = helo" does it there and then.
3663 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3665 TK/03 Fix log output including CR from clamd.
3667 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3668 exist provoked a memory error which could cause a segfault.
3670 PH/15 Installed PCRE 6.2
3672 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3674 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3675 of the problem. Specifically, suggested +O2 rather than +O1 for the
3678 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3680 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3681 record, the port was not correctly logged when the outgoing_port log
3682 selector was set (it logged the transort's default port).
3684 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3685 fallback_hosts, and "hosts" in the smtp transport.
3687 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3688 host errors such as "Connection refused".
3690 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3691 authentication with radiusclient 0.4.9:
3693 - Error returned from rc_read_config was caught wrongly
3694 - Username/password not passed on to radius server due to wrong length.
3696 The presumption is that some radiusclient API changes for 4.51/PH/17
3697 were not taken care of correctly. The code is still untested by me (my
3698 Linux distribution still has 0.3.2 of radiusclient), but it was
3699 contributed by a Radius user.
3701 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3702 expanding the "port" option of the smtp transport.
3704 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3705 while reading a MIME header. Thanks to Tom Hughes for a patch.
3707 PH/24 Include config.h inside local_scan.h so that configuration settings are
3710 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3711 that in an ACL for RCPT (for example), you can examine exactly what was
3714 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3715 commands, but it was not correctly comparing the address with the actual
3716 client host address. Thus, it would show the EHLO address in Received:
3717 header lines when this was not necessary.
3719 PH/27 Added the % operator to ${eval:}.
3721 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3722 it should be ignoring failures (because with -C, for example, it has lost
3723 privilege). It wasn't ignoring creation failures other than "already
3726 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3729 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3730 because an input line was too long, either on its own, or by virtue of
3731 too many continuations, the temporary file was not being removed, and the
3732 return code was incorrect.
3734 PH/31 Missing "BOOL" in function definition in filtertest.c.
3736 PH/32 Applied Sieve patches from the maintainer.
3738 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3740 PH/33 Added "verify = not_blind".
3742 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3743 Local/Makefile (with some defaults set). These are used in built scripts
3744 such as exicyclog, but they have never been used in the exim_install
3745 script (though there are many overriding facilities there). I have
3746 arranged that the exim_install script now takes note of these two
3749 PH/35 Installed configuration files for Dragonfly.
3751 PH/36 When a locally submitted message by a trusted user did not contain a
3752 From: header, and the sender address was obtained from -f or from an SMTP
3753 MAIL command, and the trusted user did not use -F to supply a sender
3754 name, $originator_name was incorrectly used when constructing a From:
3755 header. Furthermore, $originator_name was used for submission mode
3756 messages from external hosts without From: headers in a similar way,
3757 which is clearly wrong.
3759 PH/37 Added control=suppress_local_fixups.
3761 PH/38 When log_selector = +received_sender was set, and the addition of the
3762 sender made the log line's construction buffer exactly full, or one byte
3763 less than full, an overflow happened when the terminating "\n" was
3766 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3767 when the result of a list match is failure because a DNS lookup failed.
3769 PH/40 RM_COMMAND is now used in the building process.
3771 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3772 the "build-* directories that it finds.
3774 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3775 address in a domain literal was a prefix of an interface address.
3777 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3778 when verifying a sender address, unless rewrite_headers is false.
3780 PH/44 Wrote a long comment about why errors_to addresses are verified as
3781 recipients, not senders.
3783 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3784 the ratelimit ACL was added.
3786 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3788 PH/46 Added extra information about PostgreSQL errors to the error string.
3790 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3791 avoiding the need to install special zones in a real server. This is
3792 backwards compatible; if it can't find the fake resolver, it drops back.
3793 Thus, both old and new test suites can be run.
3795 TF/02 Added util/ratelimit.pl
3797 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3798 clock is set back in time.
3800 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3801 Candler <B.Candler@pobox.com>.
3803 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3804 OK for addresses that are the result of redirections.
3806 PH/48 A number of further additions for the benefit of the new test suite,
3807 including a fake gethostbyname() that interfaces to the fake DNS resolver
3810 TF/06 The fix for widen_domains has also been applied to qualify_single and
3811 search_parents which are the other dnslookup options that can cause
3814 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3817 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3819 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3820 service" if it encountered an MX record with an empty target hostname.
3821 The message is now "an MX or SRV record indicated no SMTP service".
3823 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3824 if the DNS of the sending site is misconfigured. This is quite a
3825 common situation. This change restores the behaviour of treating a
3826 helo verification defer as a failure.
3828 PH/51 If self=fail was set on a router, the bounce message did not include the
3829 actual error message.
3835 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3837 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3838 command itself ended in error, the underlying message about the transport
3839 filter timeout was being overwritten with the pipe command error. Now the
3840 underlying error message should be appended to the second error message.
3842 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3844 PH/02 Reduce the amount of output that "make" produces by default. Full output
3845 can still be requested.
3847 PH/03 The warning log line about a condition test deferring for a "warn" verb
3848 was being output only once per connection, rather than after each
3849 occurrence (because it was using the same function as for successful
3850 "warn" verbs). This seems wrong, so I have changed it.
3852 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3853 it should not have, which might have caused a crash in the right
3854 circumstances, but probably never did.
3856 PH/04 Installed a modified version of Tony Finch's patch to make submission
3857 mode fix the return path as well as the Sender: header line, and to
3858 add a /name= option so that you can make the user's friendly name appear
3861 TF/03 Added the control = fakedefer ACL modifier.
3863 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3864 Mark Lowes for thorough testing.
3866 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3868 TK/03 Merged latest SRS patch from Miles Wilton.
3870 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3871 with the definition in sysexits.h (which is #included earlier).
3872 Fortunately, Exim does not actually use EX_OK. The code used to try to
3873 preserve the sysexits.h value, by assuming that macro definitions were
3874 scanned for macro replacements. I have been disabused of this notion,
3875 so now the code just undefines EX_OK before #including unistd.h.
3877 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3878 in the smtp transport. When a block could not be written in a single
3879 write() function, the timeout was being re-applied to each part-write.
3880 This seems wrong - if the receiver was accepting one byte at a time it
3881 would take for ever. The timeout is now adjusted when this happens. It
3882 doesn't have to be particularly precise.
3884 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3885 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3887 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3888 without a domain if the check to <postmaster@domain> fails.
3890 SC/01 Eximstats: added -xls and the ability to specify output files
3891 (patch written by Frank Heydlauf).
3893 SC/02 Eximstats: use FileHandles for outputting results.
3895 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3897 SC/04 Eximstats: fixed display of large numbers with -nvr option
3899 SC/05 Eximstats: fixed merging of reports with empty tables.
3901 SC/06 Eximstats: added the -include_original_destination flag
3903 SC/07 Eximstats: removed tabs and trailing whitespace.
3905 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3907 TK/06 MBOX spool code: Add real "From " MBOX separator line
3908 so the .eml file is really in mbox format (even though
3909 most programs do not really care). Patch from Alex Miller.
3911 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3912 The latter is generated from $received_to and is only set if the
3913 message has one envelope recipient. SA can use these headers,
3914 obviously out-of-the-box. Patch from Alex Miller.
3916 PH/08 The ${def test on a variable was returning false if the variable's
3917 value was "0", contrary to what the specification has always said!
3918 The result should be true unless the variable is empty.
3920 PH/09 The syntax error of a character other than { following "${if
3921 def:variable_name" (after optional whitespace) was not being diagnosed.
3922 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3923 accidental colon was present, for example, could give incorrect results.
3925 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3926 result is used (not including appendfile, where other changes are about
3929 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3930 This involved changing a lot of size variables from int to off_t. It
3931 should work with maildirs and everything.
3933 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3934 spamd dying while we are connected to it.
3936 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3939 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3940 function for BDB 4.3.
3942 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3944 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3945 the sources for PCRE have been re-arranged and more files are now
3948 PH/15 The code I had for printing potentially long long variables in PH/11
3949 above was not the best (it lost precision). The length of off_t variables
3950 is now inspected at build time, and an appropriate printing format (%ld
3951 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3952 to be "long long int" or "long int". This is needed for the internal
3953 formatting function string_vformat().
3955 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3956 the configuration file to be ":syslog", then the script "guesses" where
3957 the logs files are, rather than using the compiled in default. In our
3958 case the guess is not the same as the compiled default, so the script
3959 suddenly stopped working when I started to use syslog. The patch checks
3960 to see if log_file_path is "". If so, it attempts to read it from exim
3961 with no configuration file to get the compiled in version, before it
3962 falls back to the previous guessing code."
3964 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3965 implementing BATV in an Exim configuration. See NewStuff for the gory
3968 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3969 Makefile that are specific to HP-UX.
3971 PH/18 If the "use_postmaster" option was set for a recipient callout together
3972 with the "random" option, the postmaster address was used as the MAIL
3973 FROM address for the random test, but not for the subsequent recipient
3974 test. It is now used for both.
3976 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3977 patch removes a few documentation additions to RFC 3028, because the
3978 latest draft now contains them. It adds the new en;ascii-case comparator
3979 and a new error check for 8bit text in MIME parts. Comparator and
3980 require names are now matched exactly. I enabled the subaddress
3981 extension, but it is not well tested yet (read: it works for me)."
3983 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3984 rework some of the code of TK/09 above to avoid the hardwired use of
3985 "%lld" and "long long". Replaced the call to snprintf() with a call to
3988 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3989 records point to non-existent hosts", "retry timeout exceeded", and
3990 "retry time not reached for any host after a long failure period".
3992 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3993 experimental DomainKeys support:
3995 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3996 (2) On an error such as an illegally used "control", the wrong name for
3997 the control was given.
3999 These problems did NOT occur unless DomainKeys support was compiled.
4001 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4003 PH/24 Added ${if match_ip condition.
4005 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4006 databases so that it will be absolutely obvious if a crash occurs in the
4007 DB library. This is a regular occurrence (often caused by mis-matched
4010 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4011 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4012 picked up on a user's system that detects such things. There doesn't seem
4013 to be a gcc warning option for this - only an attribute that has to be
4014 put on the function's prototype. It seems that in Fedora Core 4 they have
4015 set this on a number of new functions. No doubt there will be more in due
4018 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4019 specify a transport. However, if an address that was verified by such a
4020 router was the subject of a callout, Exim crashed because it tried to
4021 read the rcpt_include_affixes from the non-existent transport. Now it
4022 just assumes that the setting of that option is false. This bug was
4023 introduced by 4.51/PH/31.
4025 PH/28 Changed -d+all to exclude +memory, because that information is very
4026 rarely of interest, but it makes the output a lot bigger. People tend to
4027 do -d+all out of habit.
4029 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4030 code in os-type was giving problems when libc.so lives in lib64, like on
4033 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4034 aren't the modern standard, and it seems that some systems' include files
4035 don't always have them. Exim was already checking for some of the newer
4036 ones like T_AAAA, and defining it itself. I've added checks for all the
4037 record types that Exim uses.
4039 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4040 not automatically generating a new one, as it is supposed to. This
4041 prevented TLS from working. If the file did exist, but contained invalid
4042 data, a new version was generated, as expected. It was only the case of a
4043 non-existent file that was broken.
4045 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4046 with a change in libdomainkeys > 0.64.
4048 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4049 from DNS. If the selector record carries the flag, it now has
4050 precedence over the domain-wide flag.
4052 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4054 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4055 the use of a port name that isn't defined in /etc/services) occurred, the
4056 message was deferred as in a normal delivery, and thus remained on the
4057 spool, instead of being failed because of the mua_wrapper setting. This
4058 is now fixed, and I tidied up some of the mua_wrapper messages at the
4061 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4062 the messages in a hash of arrays rather than using individual hashes.
4063 This is a bit cleaner and results in dramatic memory savings, albeit
4064 at a slight CPU cost.
4066 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4067 as requested by Marc Sherman.
4069 SC/10 Eximstats: added histograms for user specified patterns as requested
4072 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4074 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4075 fopen() in the content-scanning modules that did not already have it.
4081 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4082 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4084 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4086 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4088 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4089 to test Sieve filters that use "vacation".
4091 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4092 that changes the way the GnuTLS parameters are stored in the cache file.
4093 The new format can be generated externally. For backward compatibility,
4094 if the data in the cache doesn't make sense, Exim assumes it has read an
4095 old-format file, and it generates new data and writes a new file. This
4096 means that you can't go back to an older release without removing the
4099 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4100 work if there are any delivery delays because "one_time" forces the
4101 parent to be marked "delivered", so its unseen clone is never tried
4102 again. For this reason, Exim now forbids the simultaneous setting of
4105 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4106 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4107 entry if you want to know the details. The fix, however, neglected to
4108 consider the case when local delivery batching is involved. The test for
4109 "previously delivered" was not happening when checking to see if an
4110 address could be batched with a previous (undelivered) one; under
4111 certain circumstances this could lead to multiple deliveries to the same
4114 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4115 in its include files, and this causes problems building Exim.
4117 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4118 header_syntax) but Exim was just ignoring anything given after a slash.
4119 In particular, this caused confusion with an attempt to use "verify =
4120 reverse_host_lookup/defer_ok". An error is now given when options are
4121 supplied for verify items that do not have them. (Maybe reverse_host_
4122 lookup should have a defer_ok option, but that's a different point.)
4124 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4125 defined by RFC 821) to 2048, because there were problems with some AUTH
4126 commands, and RFC 1869 says the size should be increased for extended
4127 SMTP commands that take arguments.
4129 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4132 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4133 "unknown" error; now it says that the functionality isn't in the binary.
4135 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4136 an address' error message when a string expansion fails (syntax or
4137 whatever). Otherwise the password may appear in the log. Following change
4138 PH/42 below, there is no longer a chance of it appearing in a bounce
4141 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4143 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4144 of its addresses was ever tried. (Bugzilla bug #2.)
4146 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4147 the result incorrectly in the debug output. (It correctly added a newline
4148 to what was transported.)
4150 TF/01 Added $received_time.
4152 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4153 commented out examples of how to interface to a virus scanner and to
4154 SpamAssassin. Also added commented examples of av_scanner and
4155 spamd_address settings.
4157 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4158 and controls are allowed in which ACLs. There were a couple of minor
4159 errors. Some of the entries in the conditions table (which is a table of
4160 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4161 negation of where the condition IS allowed.
4163 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4165 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4166 header file does not have a version number, so I've had to invent a new
4167 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4168 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4169 radiusclient), but it was contributed by a Radius user.
4171 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4172 files or directories, for OpenSSL.
4174 PH/19 When an Exim process that is running as root has to create an Exim log
4175 file, it does so in a subprocess that runs as exim:exim so as to get the
4176 ownership right at creation (otherwise, other Exim processes might see
4177 the file with the wrong ownership). There was no test for failure of this
4178 fork() call, which would lead to the process getting stuck as it waited
4179 for a non-existent subprocess. Forks do occasionally fail when resources
4180 run out. I reviewed all the other calls to fork(); they all seem to check
4183 PH/20 When checking for unexpected SMTP input at connect time (before writing
4184 the banner), Exim was not dealing correctly with a non-positive return
4185 from the read() function. If the client had disconnected by this time,
4186 the result was a log entry for a synchronization error with an empty
4187 string after "input=" when read() returned zero. If read() returned -1
4188 (an event I could not check), uninitialized data bytes were printed.
4189 There were reports of junk text (parts of files, etc) appearing after
4192 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4194 PH/22 Added support for macro redefinition, and (re)definition in between
4195 driver and ACL definitions.
4197 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4198 forgetting to use the resulting value; it was using the unexpanded value.
4200 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4201 hadn't been configured. The fix is from Juergen Kreileder, who
4202 understands it better than I do:
4204 "Here's what I see happening with three configured cyrus_sasl
4205 authenticators configured (plain, login, cram-md5):
4207 On startup auth_cyrus_sasl_init() gets called for each of these.
4208 This means three calls to sasl_listmech() without a specified mech_list.
4209 => SASL tests which mechs of all available mechs actually work
4210 => three warnings about OTP not working
4211 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4213 With the patch, sasl_listmech() also gets called three times. But now
4214 SASL's mech_list option is set to the server_mech specified in the the
4215 authenticator. Or in other words, the answer from sasl_listmech()
4216 gets limited to just the mech you're testing for (which is different
4218 => the return list contains just 'plain' or 'login', 'cram-md5' or
4219 nothing depending on the value of ob->server_mech.
4221 I've just tested the patch: Authentication still works fine,
4222 unavailable mechs specified in the exim configuration are still
4223 caught, and the auth.log warnings about OTP are gone."
4225 PH/25 When debugging is enabled, the contents of the command line are added
4226 to the debugging output, even when log_selector=+arguments is not
4229 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4230 answer is "GNU", and only if the return is "GNU/something" is the answer
4233 PH/27 $acl_verify_message is now set immediately after the failure of a
4234 verification in an ACL, and so is available in subsequent modifiers. In
4235 particular, the message can be preserved by coding like this:
4237 warn !verify = sender
4238 set acl_m0 = $acl_verify_message
4240 Previously, $acl_verify_message was set only while expanding "message"
4241 and "log_message" when a very denied access.
4243 PH/28 Modified OS/os.c-Linux with
4245 -#ifndef OS_LOAD_AVERAGE
4246 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4248 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4249 nomenclature these days.)
4251 PH/29 Installed patch from the Sieve maintainer that adds the options
4252 sieve_useraddress and sieve_subaddress to the redirect router.
4254 PH/30 In these circumstances:
4255 . Two addresses routed to the same list of hosts;
4256 . First host does not offer TLS;
4257 . First host accepts first address;
4258 . First host gives temporary error to second address;
4259 . Second host offers TLS and a TLS session is established;
4260 . Second host accepts second address.
4261 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4262 and peerdn, if requested) that were in fact used only for the second
4265 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4266 attention to any local part prefix or suffix that was matched by the
4267 router that accepted the address. It now behaves in the same way as it
4268 does for delivery: the affixes are removed from the local part unless
4269 rcpt_include_affixes is set on the transport.
4271 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4272 timeout during the DATA phase of an incoming message.
4274 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4275 applied a patch sanctioned by the Sieve maintainer.
4277 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4278 the uid or gid is negative. A case of a negative gid caused this to be
4279 noticed. The fix allows for either to be negative.
4281 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4282 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4285 PH/36 The change PH/12 above was broken. Fixed it.
4287 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4288 the grounds that routing the same address twice would always produce the
4289 same answer. This might have been true once, but it is certainly no
4290 longer true now. Routing a child address may depend on the previous
4291 routing that produced that child. Some complicated redirection strategies
4292 went wrong when messages had multiple recipients, and made Exim's
4293 behaviour dependent on the order in which the addresses were given.
4295 I have moved the duplicate checking until after the routing is complete.
4296 Exim scans the addresses that are assigned to local and remote
4297 transports, and removes any duplicates. This means that more work will be
4298 done, as duplicates will always all be routed, but duplicates are
4299 presumably rare, so I don't expect this is of any significance.
4301 For deliveries to pipes, files, and autoreplies, the duplicate checking
4302 still happens during the routing process, since they are not going to be
4305 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4306 It corrects a timeout issue with spamd. This is Ian's comment: "The
4307 background is that sometimes spamd either never reads data from a
4308 connection it has accepted, or it never writes response data. The exiscan
4309 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4310 blindly assumes that writes won't block so it may never time out."
4312 PH/39 Allow G after quota size as well as K and M.
4314 PH/40 The value set for $authenticated_id in an authenticator may not contain
4315 binary zeroes or newlines because the value is written to log lines and
4316 to spool files. There was no check on this. Now the value is run through
4317 the string_printing() function so that such characters are converted to
4318 printable escape sequences.
4320 PH/41 $message_linecount is a new variable that contains the total number of
4321 lines in the message. Compare $body_linecount, which is the count for the
4324 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4325 bounce and delay warning messages, except in certain special cases, which
4328 (a) An SMTP error message from a remote host;
4329 (b) A message specified in a :fail: redirection;
4330 (c) A message specified in a "fail" command in a system filter;
4331 (d) A message specified in a FAIL return from the queryprogram router;
4332 (e) A message specified by the cannot_route_message router option.
4334 In these cases only, Exim does include the error details in bounce and
4335 warning messages. There are also a few cases where bland messages such
4336 as "unrouteable address" or "local delivery error" are given.
4338 PH/43 $value is now also set for the "else" part of a ${run expansion.
4340 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4341 being worked on, but at least Exim now implements the latest version to
4344 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4345 process to complete was treated as a delivery failure, a timeout while
4346 writing the message to the pipe was logged, but erroneously treated as a
4347 successful delivery. Such timeouts include transport filter timeouts. For
4348 consistency with the overall process timeout, these timeouts are now
4349 treated as errors, giving rise to delivery failures by default. However,
4350 there is now a new Boolean option for the pipe transport called
4351 timeout_defer, which, if set TRUE, converts the failures into defers for
4352 both kinds of timeout. A transport filter timeout is now identified in
4355 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4356 systems where "make" and "gmake" are different, calling "gmake" at top
4357 level broke things. I've arranged for the value of $(MAKE) to be passed
4358 from the Makefile to this script so that it can call the same version of
4362 A note about Exim versions 4.44 and 4.50
4363 ----------------------------------------
4365 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4366 changes of various kinds. As a consequence, a big documentation update was
4367 needed. This delayed the release for rather longer than seemed good, especially
4368 in the light of a couple of (minor) security issues. Therefore, the changes
4369 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4370 release. So 4.44 and 4.50 are in effect two different branches that both start
4373 I have left the 4.50 change log unchanged; it contains all the changes since
4374 4.43. The change log for 4.44 is below; many of its items are identical to
4375 those for 4.50. This seems to be the most sensible way to preserve the
4376 historical information.
4382 1. Minor wording change to the doc/README.SIEVE file.
4384 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4385 computation of the current number of files was incorrect.
4387 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4388 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4391 4. Give more explanation in the error message when the command for a transport
4392 filter fails to execute.
4394 5. There are several places where Exim runs a non-Exim command in a
4395 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4396 was being done only for the command run by the queryprogram router. It is
4397 now done for all such subprocesses. The other cases are: ${run, transport
4398 filters, and the commands run by the lmtp and pipe transports.
4400 6. Added CONFIGURE_GROUP build-time option.
4402 7. Some older OS have a limit of 256 on the maximum number of file
4403 descriptors. Exim was using setrlimit() to set 1000 as a large value
4404 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4405 systems. I've change it so that if it can't get 1000, it tries for 256.
4407 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4408 was an oversight, and furthermore, ever since the addition of extra
4409 controls (e.g. 4.43/32), the checks on when to allow different forms of
4410 "control" were broken. There should now be diagnostics for all cases when a
4411 control that does not make sense is encountered.
4413 9. Added the /retain_sender option to "control=submission".
4415 10. $recipients is now available in the predata ACL (oversight).
4417 11. Tidy the search cache before the fork to do a delivery from a message
4418 received from the command line. Otherwise the child will trigger a lookup
4419 failure and thereby defer the delivery if it tries to use (for example) a
4420 cached ldap connection that the parent has called unbind on.
4422 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4423 of $address_data from the recipient verification was clobbered by the
4424 sender verification.
4426 13. The value of address_data from a sender verification is now available in
4427 $sender_address_data in subsequent conditions in the ACL statement.
4429 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4431 15. Added a new option "connect=<time>" to callout options, to set a different
4434 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4435 was its contents. (It was OK if the option was not defined at all.)
4437 17. A "Completed" log line is now written for messages that are removed from
4438 the spool by the -Mrm option.
4440 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4441 information about exactly what failed.
4443 19. Added -dd to debug only the daemon process.
4445 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4446 handles timeouts, both on the server side and network timeouts. Renamed the
4447 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4449 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4450 It is now set to "smtps".
4452 22. $host_address is now set to the target address during the checking of
4453 ignore_target_hosts.
4455 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4456 being passed; this would have caused $sender_host_name to have been used if
4457 matching the list had actually called for a host name (not very likely,
4458 since this list is usually IP addresses). A host name is now passed as
4461 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4462 code that specifies a non-restarting handler (typically sigaction() in
4463 modern systems) in an attempt to fix a rare and obscure crash bug.
4465 25. Narrowed the window for a race in the daemon that could cause it to ignore
4466 SIGCHLD signals. This is not a major problem, because they are used only to
4467 wake it up if nothing else does.
4469 26. A malformed maildirsize file could cause Exim to calculate negative values
4470 for the mailbox size or file count. Odd effects could occur as a result.
4471 The maildirsize information is now recalculated if the size or filecount
4474 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4475 support for a long time. Removed HAVE_SYS_VFS_H.
4477 28. Installed the latest version of exipick from John Jetmore.
4479 29. In an address list, if the pattern was not a regular expression, an empty
4480 subject address (from a bounce message) matched only if the pattern was an
4481 empty string. Non-empty patterns were not even tested. This was the wrong
4482 because it is perfectly reasonable to use an empty address as part of a
4483 database query. An empty address is now tested by patterns that are
4484 lookups. However, all the other forms of pattern expect the subject to
4485 contain a local part and a domain, and therefore, for them, an empty
4486 address still always fails if the pattern is not itself empty.
4488 30. Exim went into a mad DNS loop when attempting to do a callout where the
4489 host was specified on an smtp transport, and looking it up yielded more
4490 than one IP address.
4492 31. Re-factored the code for checking spool and log partition space into a
4493 function that finds that data and another that does the check. The former
4494 is then used to implement four new variables: $spool_space, $log_space,
4495 $spool_inodes, and $log_inodes.
4497 32. The RFC2047 encoding function was originally intended for short strings
4498 such as real names; it was not keeping to the 75-character limit for
4499 encoded words that the RFC imposes. It now respects the limit, and
4500 generates multiple encoded words if necessary. To be on the safe side, I
4501 have increased the buffer size for the ${rfc2047: expansion operator from
4504 33. It is now permitted to omit both strings after an "if" condition; if the
4505 condition is true, the result is "true". As before, when the second string
4506 is omitted, a false condition yields an empty string. This makes it less
4507 cumbersome to write custom ACL and router conditions.
4509 34. Failure to deliver a bounce message always caused it to be frozen, even if
4510 there was an errors_to setting on the router. The errors_to setting is now
4513 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4514 canonical form (fully expanded) before being placed in
4515 $sender_host_address.
4517 36. The table in the code that translates DNS record types into text (T_A to
4518 "A" for instance) was missing entries for NS and CNAME. It is just possible
4519 that this could have caused confusion if both these types were looked up
4520 for the same domain, because the text type is used as part of Exim's
4521 per-process caching. But the chance of anyone hitting this buglet seems
4524 37. The dnsdb lookup has been extended in a number of ways.
4526 (1) There is a new type, "zns", which walks up the domain tree until it
4527 finds some nameserver records. It should be used with care.
4529 (2) There is a new type, "mxh", which is like "mx" except that it returns
4530 just the host names, not the priorities.
4532 (3) It is now possible to give a list of domains (or IP addresses) to be
4533 looked up. The behaviour when one of the lookups defers can be
4534 controlled by a keyword.
4536 (4) It is now possible to specify the separator character for use when
4537 multiple records are returned.
4539 38. The dnslists ACL condition has been extended: it is now possible to supply
4540 a list of IP addresses and/or domains to be looked up in a particular DNS
4543 39. Added log_selector=+queue_time_overall.
4545 40. When running the queue in the test harness, wait just a tad after forking a
4546 delivery process, to get repeatability of debugging output.
4548 41. Include certificate and key file names in error message when GnuTLS fails
4549 to set them up, because the GnuTLS error message doesn't include the name
4550 of the failing file when there is a problem reading it.
4552 42. Allow both -bf and -bF in the same test run.
4554 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4556 44. The "Exiscan patch" is now merged into the mainline Exim source.
4558 45. Sometimes the final signoff response after QUIT could fail to get
4559 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4560 < 0 before doing a fflush(). This bug looks as though it goes back to the
4561 introduction of TLS in release 3.20, but "sometimes" must have been rare
4562 because the tests only now provoked it.
4564 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4565 (this can affect the format of dates).
4567 47. exim_tidydb, when checking for the continued existence of a message for
4568 which it has found a message-specific retry record, was not finding
4569 messages that were in split spool directories. Consequently, it was
4570 deleting retry records that should have stayed in existence.
4572 48. Steve fixed some bugs in eximstats.
4574 49. The SPA authentication driver was not abandoning authentication and moving
4575 on to the next authenticator when an expansion was forced to fail,
4576 contradicting the general specification for all authenticators. Instead it
4577 was generating a temporary error. It now behaves as specified.
4579 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4580 (the order specifies the preference for clients). The order is now AES256,
4581 AES128, 3DES, ARCFOUR128.
4583 51. Small patch to Sieve code - explicitly set From: when generating an
4586 52. Exim crashed if a remote delivery caused a very long error message to be
4587 recorded - for instance if somebody sent an entire SpamAssassin report back
4588 as a large number of 550 error lines. This bug was coincidentally fixed by
4589 increasing the size of one of Exim's internal buffers (big_buffer) that
4590 happened as part of the Exiscan merge. However, to be on the safe side, I
4591 have made the code more robust (and fixed the comments that describe what
4594 53. Now that there can be additional text after "Completed" in log lines (if
4595 the queue_time_overall log selector is set), a one-byte patch to exigrep
4596 was needed to allow it to recognize "Completed" as not the last thing in
4599 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4600 patch that reportedly fixes this has been added. I am not expert enough to
4601 create a test for it. This is what the patch creator wrote:
4603 "I found a little strange behaviour of ldap code when working with
4604 Windows 2003 AD Domain, where users was placed in more than one
4605 Organization Units. When I tried to give exim partial DN, the exit code
4606 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4607 But simultaneously result of request was absolutely normal ldap result,
4608 so I produce this patch..."
4610 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4611 REFERENCE, so I have modified the code to exclude the patch when that macro
4614 55. Some experimental protocols are using DNS PTR records for new purposes. The
4615 keys for these records are domain names, not reversed IP addresses. The
4616 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4617 leaves it alone. Component reversal etc. now happens only for IP addresses.
4620 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4622 57. Double the size of the debug message buffer (to 2048) so that more of very
4623 long debug lines gets shown.
4625 58. The exicyclog utility now does better if the number of log files to keep
4626 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4628 59. Two changes related to the smtp_active_hostname option:
4630 (1) $smtp_active_hostname is now available as a variable.
4631 (2) The default for smtp_banner uses $smtp_active_hostname instead
4632 of $primary_hostname.
4634 60. The host_aton() function is supposed to be passed a string that is known
4635 to be a valid IP address. However, in the case of IPv6 addresses, it was
4636 not checking this. This is a hostage to fortune. Exim now panics and dies
4637 if the condition is not met. A case was found where this could be provoked
4638 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4639 components; fortuitously, this particular loophole had already been fixed
4640 by change 4.50/55 above.
4642 If there are any other similar loopholes, the new check in host_aton()
4643 itself should stop them being exploited. The report I received stated that
4644 data on the command line could provoke the exploit when Exim was running as
4645 exim, but did not say which command line option was involved. All I could
4646 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4647 running as the user.
4650 61. There was a buffer overflow vulnerability in the SPA authentication code
4651 (which came originally from the Samba project). I have added a test to the
4652 spa_base64_to_bits() function which I hope fixes it.
4655 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4656 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4658 63. The daemon start-up calls getloadavg() while still root for those OS that
4659 need the first call to be done as root, but it missed one case: when
4660 deliver_queue_load_max is set with deliver_drop_privilege. This is
4661 necessary for the benefit of the queue runner, because there is no re-exec
4662 when deliver_drop_privilege is set.
4664 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4665 This has been fixed.
4667 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4668 was in use, was not putting the data itself into the right store pool;
4669 consequently, it could be overwritten for a subsequent message in the same
4670 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4673 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4675 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4676 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4677 the function were treating the return as a boolean value, which happened to
4678 work because 0=false and not-0=true, but is not correct code.
4680 68. The host_aton() function was not handling scoped IPv6 addresses (those
4681 with, for example, "%eth0" on the end) correctly.
4683 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4684 negated items (that is, ~something) in unsigned ints. Some compilers
4685 apparently mutter when there is no cast.
4687 70. If an address verification called from an ACL failed, and did not produce a
4688 user-specific message (i.e. there was only a "system" message), nothing was
4689 put in $acl_verify_message. In this situation, it now puts the system
4692 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4693 session; change 4.31/43 added the unwanted input to the log line - except
4694 that it did not do this in the start of session case. It now does.
4696 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4697 This is wrong and can cause the other end to generate a synchronization
4698 error if it is another Exim or anything else that does the synchronization
4699 check. A QUIT command is no longer sent after a timeout.
4701 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4702 during host lookups.
4704 74. The defer_ok option of callout verification was not working if it was used
4705 when verifying addresses in header lines, that is, for this case:
4707 verify = header_sender/callout=defer_ok
4709 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4710 those file descriptors could be used for SMTP connections. If anything
4711 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4712 could be sent to the SMTP client, causing chaos. The daemon now opens
4713 stdin, stdout, and stderr to /dev/null when it puts itself into the
4716 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4717 log by default. The user can override this with suitable Perl magic.
4719 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4720 the log message when discard triggers, was not working for the DATA ACL or
4721 for the non-SMTP ACL.
4723 78. Error message wording change in sieve.c.
4725 79. If smtp_accept_max_per_host was set, the number of connections could be
4726 restricted to fewer than expected, because the daemon was trying to set up
4727 a new connection before checking whether the processes handling previous
4728 connections had finished. The check for completed processes is now done
4729 earlier. On busy systems, this bug wouldn't be noticed because something
4730 else would have woken the daemon, and it would have reaped the completed
4733 80. If a message was submitted locally by a user whose login name contained one
4734 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4735 It caused a spool format error. I have fixed the spool reading code. A
4736 related problem was that the "from" clause in the Received: line became
4737 illegal because of the space(s). It is now covered by ${quote_local_part.
4739 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4742 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4744 83. Patch from Sieve maintainer to fix unterminated string problem in
4745 "vacation" handling.
4747 84. Some minor changes to the Linux configuration files to help with other
4748 OS variants using glibc.
4750 85. One more patch for Sieve to update vacation handling to latest spec.
4753 ----------------------------------------------------
4754 See the note above about the 4.44 and 4.50 releases.
4755 ----------------------------------------------------
4761 1. Change 4.43/35 introduced a bug that caused file counts to be
4762 incorrectly computed when quota_filecount was set in an appendfile
4765 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4766 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4769 3. Give more explanation in the error message when the command for a transport
4770 filter fails to execute.
4772 4. There are several places where Exim runs a non-Exim command in a
4773 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4774 was being done only for the command run by the queryprogram router. It is
4775 now done for all such subprocesses. The other cases are: ${run, transport
4776 filters, and the commands run by the lmtp and pipe transports.
4778 5. Some older OS have a limit of 256 on the maximum number of file
4779 descriptors. Exim was using setrlimit() to set 1000 as a large value
4780 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4781 systems. I've change it so that if it can't get 1000, it tries for 256.
4783 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4784 was an oversight, and furthermore, ever since the addition of extra
4785 controls (e.g. 4.43/32), the checks on when to allow different forms of
4786 "control" were broken. There should now be diagnostics for all cases when a
4787 control that does not make sense is encountered.
4789 7. $recipients is now available in the predata ACL (oversight).
4791 8. Tidy the search cache before the fork to do a delivery from a message
4792 received from the command line. Otherwise the child will trigger a lookup
4793 failure and thereby defer the delivery if it tries to use (for example) a
4794 cached ldap connection that the parent has called unbind on.
4796 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4797 of $address_data from the recipient verification was clobbered by the
4798 sender verification.
4800 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4801 was its contents. (It was OK if the option was not defined at all.)
4803 11. A "Completed" log line is now written for messages that are removed from
4804 the spool by the -Mrm option.
4806 12. $host_address is now set to the target address during the checking of
4807 ignore_target_hosts.
4809 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4810 being passed; this would have caused $sender_host_name to have been used if
4811 matching the list had actually called for a host name (not very likely,
4812 since this list is usually IP addresses). A host name is now passed as
4815 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4816 code that specifies a non-restarting handler (typically sigaction() in
4817 modern systems) in an attempt to fix a rare and obscure crash bug.
4819 15. Narrowed the window for a race in the daemon that could cause it to ignore
4820 SIGCHLD signals. This is not a major problem, because they are used only to
4821 wake it up if nothing else does.
4823 16. A malformed maildirsize file could cause Exim to calculate negative values
4824 for the mailbox size or file count. Odd effects could occur as a result.
4825 The maildirsize information is now recalculated if the size or filecount
4828 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4829 support for a long time. Removed HAVE_SYS_VFS_H.
4831 18. Updated exipick to current release from John Jetmore.
4833 19. Allow an empty sender to be matched against a lookup in an address list.
4834 Previously the only cases considered were a regular expression, or an
4837 20. Exim went into a mad DNS lookup loop when doing a callout where the
4838 host was specified on the transport, if the DNS lookup yielded more than
4841 21. The RFC2047 encoding function was originally intended for short strings
4842 such as real names; it was not keeping to the 75-character limit for
4843 encoded words that the RFC imposes. It now respects the limit, and
4844 generates multiple encoded words if necessary. To be on the safe side, I
4845 have increased the buffer size for the ${rfc2047: expansion operator from
4848 22. Failure to deliver a bounce message always caused it to be frozen, even if
4849 there was an errors_to setting on the router. The errors_to setting is now
4852 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4853 canonical form (fully expanded) before being placed in
4854 $sender_host_address.
4856 24. Updated eximstats to version 1.33
4858 25. Include certificate and key file names in error message when GnuTLS fails
4859 to set them up, because the GnuTLS error message doesn't include the name
4860 of the failing file when there is a problem reading it.
4862 26. Expand error message when OpenSSL has problems setting up cert/key files.
4865 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4866 (this can affect the format of dates).
4868 28. exim_tidydb, when checking for the continued existence of a message for
4869 which it has found a message-specific retry record, was not finding
4870 messages that were in split spool directories. Consequently, it was
4871 deleting retry records that should have stayed in existence.
4873 29. eximstats updated to version 1.35
4874 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4875 1.35 - bugfix such that pie charts by volume are generated correctly
4877 30. The SPA authentication driver was not abandoning authentication and moving
4878 on to the next authenticator when an expansion was forced to fail,
4879 contradicting the general specification for all authenticators. Instead it
4880 was generating a temporary error. It now behaves as specified.
4882 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4883 (the order specifies the preference for clients). The order is now AES256,
4884 AES128, 3DES, ARCFOUR128.
4886 31. Small patch to Sieve code - explicitly set From: when generating an
4889 32. Exim crashed if a remote delivery caused a very long error message to be
4890 recorded - for instance if somebody sent an entire SpamAssassin report back
4891 as a large number of 550 error lines. This bug was coincidentally fixed by
4892 increasing the size of one of Exim's internal buffers (big_buffer) that
4893 happened as part of the Exiscan merge. However, to be on the safe side, I
4894 have made the code more robust (and fixed the comments that describe what
4897 33. Some experimental protocols are using DNS PTR records for new purposes. The
4898 keys for these records are domain names, not reversed IP addresses. The
4899 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4900 leaves it alone. Component reversal etc. now happens only for IP addresses.
4903 34. The host_aton() function is supposed to be passed a string that is known
4904 to be a valid IP address. However, in the case of IPv6 addresses, it was
4905 not checking this. This is a hostage to fortune. Exim now panics and dies
4906 if the condition is not met. A case was found where this could be provoked
4907 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4908 components; fortuitously, this particular loophole had already been fixed
4909 by change 4.50/55 or 4.44/33 above.
4911 If there are any other similar loopholes, the new check in host_aton()
4912 itself should stop them being exploited. The report I received stated that
4913 data on the command line could provoke the exploit when Exim was running as
4914 exim, but did not say which command line option was involved. All I could
4915 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4916 running as the user.
4919 35. There was a buffer overflow vulnerability in the SPA authentication code
4920 (which came originally from the Samba project). I have added a test to the
4921 spa_base64_to_bits() function which I hope fixes it.
4924 36. The daemon start-up calls getloadavg() while still root for those OS that
4925 need the first call to be done as root, but it missed one case: when
4926 deliver_queue_load_max is set with deliver_drop_privilege. This is
4927 necessary for the benefit of the queue runner, because there is no re-exec
4928 when deliver_drop_privilege is set.
4930 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4931 was in use, was not putting the data itself into the right store pool;
4932 consequently, it could be overwritten for a subsequent message in the same
4933 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4936 38. Sometimes the final signoff response after QUIT could fail to get
4937 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4938 < 0 before doing a fflush(). This bug looks as though it goes back to the
4939 introduction of TLS in release 3.20, but "sometimes" must have been rare
4940 because the tests only now provoked it.
4946 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4947 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4948 Now it returns an integer. A number of calls were still expecting a T/F
4949 return. Fortuitously, in all cases, the tests worked in OK situations,
4950 which is the norm. However, things would have gone wrong on any write
4951 failures on the smtp file descriptor. This function is used when sending
4952 messages over SMTP and also when doing verify callouts.
4954 2. When Exim is called to do synchronous delivery of a locally submitted
4955 message (the -odf or -odi options), it no longer closes stderr before doing
4958 3. Implemented the mua_wrapper option.
4960 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4962 5. Implemented the functions header_remove(), header_testname(),
4963 header_add_at_position(), and receive_remove_recipient(), and exported them
4966 6. If an ACL "warn" statement specified the addition of headers, Exim already
4967 inserted X-ACL-Warn: at the start if there was no header name. However, it
4968 was not making this test for the second and subsequent header lines if
4969 there were newlines in the string. This meant that an invalid header could
4970 be inserted if Exim was badly configured.
4972 7. Allow an ACL "warn" statement to add header lines at the start or after all
4973 the Received: headers, as well as at the end.
4975 8. Added the rcpt_4xx retry error code.
4977 9. Added postmaster_mailfrom=xxx to callout verification option.
4979 10. Added mailfrom=xxxx to the callout verification option, for verify=
4982 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4983 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4985 12. Inserted some casts to stop certain compilers warning when using pointer
4986 differences as field lengths or precisions in printf-type calls (mostly
4987 affecting debugging statements).
4989 13. Added optional readline() support for -be (dynamically loaded).
4991 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4992 same clock tick as a message's arrival, so that its received time was the
4993 same as the "first fail" time on the retry record, and that message
4994 remained on the queue past the ultimate address timeout, every queue runner
4995 would try a delivery (because it was past the ultimate address timeout) but
4996 after another failure, the ultimate address timeout, which should have then
4997 bounced the address, did not kick in. This was a "< instead of <=" error;
4998 in most cases the first failure would have been in the next clock tick
4999 after the received time, and all would be well.
5001 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5002 being recognized when the domain list was tested by the match_domain
5003 condition in an expansion string.
5005 16. Added the ${str2b64: operator.
5007 17. Exim was always calling setrlimit() to set a large limit for the number of
5008 processes, without checking whether the existing limit was already
5009 adequate. (It did check for the limit on file descriptors.) Furthermore,
5010 errors from getrlimit() and setrlimit() were being ignored. Now they are
5011 logged to the main and panic logs, but Exim does carry on, to try to do its
5012 job under whatever limits there are.
5014 18. Imported PCRE 5.0.
5016 19. Trivial typo in log message " temporarily refused connection" (the leading
5019 20. If the log selector return_path_on_delivery was set and an address was
5020 redirected to /dev/null, the delivery process crashed because it assumed
5021 that a return path would always be set for a "successful" delivery. In this
5022 case, the whole delivery is bypassed as an optimization, and therefore no
5025 21. Internal re-arrangement: the function for sending a challenge and reading
5026 a response while authentication was assuming a zero-terminated challenge
5027 string. It's now changed to take a pointer and a length, to allow for
5028 binary data in such strings.
5030 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5032 23. Exim was not respecting finduser_retries when seeking the login of the
5033 uid under which it was called; it was always trying 10 times. (The default
5034 setting of finduser_retries is zero.) Also, it was sleeping after the final
5035 failure, which is pointless.
5037 24. Implemented tls_on_connect_ports.
5039 25. Implemented acl_smtp_predata.
5041 26. If the domain in control=submission is set empty, Exim assumes that the
5042 authenticated id is a complete email address when it generates From: or
5043 Sender: header lines.
5045 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5046 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5047 chown and chgrp in /bin and hostname in /usr/bin.
5049 28. Exim was keeping the "process log" file open after each use, just as it
5050 does for the main log. This opens the possibility of it remaining open for
5051 long periods when the USR1 signal hits a daemon. Occasional processlog
5052 errors were reported, that could have been caused by this. Anyway, it seems
5053 much more sensible not to leave this file open at all, so that is what now
5056 29. The long-running daemon process does not normally write to the log once it
5057 has entered its main loop, and it closes the log before doing so. This is
5058 so that log files can straightforwardly be renamed and moved. However,
5059 there are a couple of unusual error situations where the daemon does write
5060 log entries, and I had neglected to close the log afterwards.
5062 30. The text of an SMTP error response that was received during a remote
5063 delivery was being truncated at 512 bytes. This is too short for some of
5064 the long messages that one sometimes sees. I've increased the limit to
5067 31. It is now possible to make retry rules that apply only when a message has a
5068 specific sender, in particular, an empty sender.
5070 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5071 it possible to be selective about when SMTP synchronization is enforced.
5073 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5075 32. Implemented hosts_connection_nolog.
5077 33. Added an ACL for QUIT.
5079 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5082 35. Added mailbox_size and mailbox_filecount to appendfile.
5084 36. Added control = no_multiline_responses to ACLs.
5086 37. There was a bug in the logic of the code that waits for the clock to tick
5087 in the case where the clock went backwards by a substantial amount such
5088 that the microsecond fraction of "now" was more than the microsecond
5089 fraction of "then" (but the whole seconds number was less).
5091 38. Added support for the libradius Radius client library this is found on
5092 FreeBSD (previously only the radiusclient library was supported).
5098 1. When certain lookups returned multiple values in the form name=value, the
5099 quoting of the values was not always being done properly. Specifically:
5100 (a) If the value started with a double quote, but contained no whitespace,
5102 (b) If the value contained whitespace other than a space character (i.e.
5103 tabs or newlines or carriage returns) it was not quoted.
5104 This fix has been applied to the mysql and pgsql lookups by writing a
5105 separate quoting function and calling it from the lookup code. The fix
5106 should probably also be applied to nisplus, ibase and oracle lookups, but
5107 since I cannot test any of those, I have not disturbed their existing code.
5109 2. A hit in the callout cache for a specific address caused a log line with no
5110 reason for rejecting RCPT. Now it says "Previous (cached) callout
5111 verification failure".
5113 3. There was an off-by-one bug in the queryprogram router. An over-long
5114 return line was truncated at 256 instead of 255 characters, thereby
5115 overflowing its buffer with the terminating zero. As well as fixing this, I
5116 have increased the buffer size to 1024 (and made a note to document this).
5118 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5119 when Exim is waiting for an SMTP response from a remote server, Exim
5120 restarts its select() call on the socket, thereby resetting its timeout.
5121 This is not a problem when such interrupts are rare. Somebody set up a cron
5122 job to run exiwhat every 2 minutes, which is less than the normal select()
5123 timeout (5 or 10 minutes). This meant that the select() timeout never
5124 kicked in because it was always reset. I have fixed this by comparing the
5125 time when an interrupt arrives with the time at the start of the first call
5126 to select(). If more time than the timeout has elapsed, the interrupt is
5127 treated as a timeout.
5129 5. Some internal re-factoring in preparation for the addition of Sieve
5130 extensions (by MH). In particular, the "personal" test is moved to a
5131 separate function, and given an option for scanning Cc: and Bcc: (which is
5132 not set for Exim filters).
5134 6. When Exim created an email address using the login of the caller as the
5135 local part (e.g. when creating a From: or Sender: header line), it was not
5136 quoting the local part when it contained special characters such as @.
5138 7. Installed new OpenBSD configuration files.
5140 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5141 try to make them clearer.
5143 9. Callout options, other than the timeout value, were being ignored when
5144 verifying sender addresses in header lines. For example, when using
5146 verify = header_sender/callout=no_cache
5148 the cache was (incorrectly) being used.
5150 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5151 only the Cygwin environment.
5153 11. When return_path_on_delivery was set as a log selector, if different remote
5154 addresses in the same message used different return paths and parallel
5155 remote delivery occurred, the wrong values would sometimes be logged.
5156 (Whenever a remote delivery process finished, the return path value from
5157 the most recently started remote delivery process was logged.)
5159 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5160 lines when AUTH and/or TLS are in use. This is the "received protocol"
5161 field. Exim used to use "asmtp" for authenticated SMTP, without any
5162 indication (in the protocol name) for TLS use. Now it follows the RFC and
5163 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5164 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5165 names appear in log lines as well as in Received: header lines.
5167 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5168 extensions, and comparison tests, and to fix some bugs.
5170 14. Changes to the "personal" filter test:
5172 (1) The test was buggy in that it was just doing the equivalent of
5173 "contains" tests on header lines. For example, if a user's address was
5174 anne@some.where, the "personal" test would incorrectly be true for
5176 To: susanne@some.where
5178 This test is now done by extracting each address from the header in turn,
5179 and checking the entire address. Other tests that are part of "personal"
5180 are now done using regular expressions (for example, to check local parts
5181 of addresses in From: header lines).
5183 (2) The list of non-personal local parts in From: addresses has been
5184 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5185 taken from the Sieve specification recommendations.
5187 (3) If the message contains any header line starting with "List-" it is
5188 treated as non-personal.
5190 (4) The test for "circular" in the Subject: header line has been removed
5191 because it now seems ill-conceived.
5193 15. Minor typos in src/EDITME comments corrected.
5195 16. Installed latest exipick from John Jetmore.
5197 17. If headers_add on a router specified a text string that was too long for
5198 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5199 of string_sprintf() is now avoided.
5201 18. $message_body_size was not set (it was always zero) when running the DATA
5202 ACL and the local_scan() function.
5204 19. For the "mail" command in an Exim filter, no default was being set for
5205 the once_repeat time, causing a random time value to be used if "once" was
5206 specified. (If the value happened to be <= 0, no repeat happened.) The
5207 default is now 0s, meaning "never repeat". The "vacation" command was OK
5208 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5209 (I found it when inspecting the code).
5211 20. There is now an overall timeout for performing a callout verification. It
5212 defaults to 4 times the callout timeout, which applies to individual SMTP
5213 commands during the callout. The overall timeout applies when there is more
5214 than one host that can be tried. The timeout is checked before trying the
5215 next host. This prevents very long delays if there are a large number of
5216 hosts and all are timing out (e.g. when the network connections are timing
5217 out). The value of the overall timeout can be changed by specifying an
5218 additional sub-option for "callout", called "maxwait". For example:
5220 verify = sender/callout=5s,maxwait=20s
5222 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5223 to the end before writing, but this should make it even safer).
5225 22. Exim was forgetting that it had advertised PIPELINING for the second and
5226 subsequent messages on an SMTP connection. It was also not resetting its
5227 memory on STARTTLS and an internal HELO.
5229 23. When Exim logs an SMTP synchronization error within a session, it now
5230 records whether PIPELINING has been advertised or not.
5232 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5233 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5234 rather than long int.
5236 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5238 26. Added the never_mail option to autoreply.
5244 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5245 crash if the getsockname() call failed; this can happen if a connection is
5246 closed very soon after it is established. The problem was simply in the
5247 order in which certain operations were done, causing Exim to try to write
5248 to the SMTP stream before it had set up the file descriptor. The bug has
5249 been fixed by making things happen in the correct order.
5255 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5256 before the connection was closed, thus losing the rejection response.
5258 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5259 some early Solaris releases, but causes trouble in current releases where
5260 socklen_t is defined.
5262 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5265 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5268 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5269 that releasing the top of it at the end releases what was used for sub-
5270 expansions (unless the block got too big). However, discard this block if
5271 the first thing is a variable or header, so that we can use its block when
5272 it is dynamic (useful for very large $message_headers, for example).
5274 6. Lookups now cache *every* query, not just the most recent. A new, separate
5275 store pool is used for this. It can be recovered when all lookup caches are
5276 flushed. Lookups now release memory at the end of their result strings.
5277 This has involved some general refactoring of the lookup sources.
5279 7. Some code has been added to the store_xxx() functions to reduce the amount
5280 of flapping under certain conditions.
5282 8. log_incoming_interface used to affect only the <= reception log lines. Now
5283 it causes the local interface and port to be added to several more SMTP log
5284 lines, for example "SMTP connection from", and rejection lines.
5286 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5288 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5290 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5291 could be overwritten at the end of the current message (or the start of a
5292 new message if it was set in a HELO ACL). The value is now preserved for
5293 the duration of the SMTP connection.
5295 12. If a transport had a headers_rewrite setting, and a matching header line
5296 contained an unqualified address, that address was qualified, even if it
5297 did not match any rewriting rules. The underlying bug was that the values
5298 of the flags that permit the existence of unqualified sender and recipient
5299 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5300 non-local messages, and by -bnq for local messages) were not being
5301 preserved with the message after it was received.
5303 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5304 "next input=" as part of the text comprising the host identity instead of
5305 the correct text. The code was using the same buffer for two different
5306 strings. However, depending on which order the printing function evaluated
5307 its arguments, the bug did not always show up. Under Linux, for example, my
5308 test suite worked just fine.
5310 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5311 doesn't work with some older versions of Perl. It has been changed to "my",
5312 which in any case is probably the better facility to use.
5314 15. A really picky compiler found some instances of statements for creating
5315 error messages that either had too many or two few arguments for the format
5318 16. The size of the buffer for calls to the DNS resolver has been increased
5319 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5320 for addresses that have a lot of PTR records. This alleviates a problem; it
5321 does not fully solve it.
5323 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5324 buffer now truncates the list and logs the incident, which is the same
5325 action as happens when Exim is looking up a host name and its aliases.
5326 Previously in this situation something unpredictable would happen;
5327 sometimes it was "internal error: store_reset failed".
5329 18. If a server dropped the connection unexpectedly when an Exim client was
5330 using GnuTLS and trying to read a response, the client delivery process
5331 crashed while trying to generate an error log message.
5333 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5334 string, for example:
5336 warn message = H1: something\nH2: something
5338 the text was added as a single header line from Exim's point of view
5339 though it ended up OK in the delivered message. However, searching for the
5340 second and subsequent header lines using $h_h2: did not work. This has been
5341 fixed. Similarly, if a system filter added multiple headers in this way,
5342 the routers could not see them.
5344 20. Expanded the error message when iplsearch is called with an invalid key to
5345 suggest using net-iplsearch in a host list.
5347 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5348 ACLs are no longer actually imposed (and a message to that effect is
5351 22. If a "gecos" field in a passwd entry contained escaped characters, in
5352 particular, if it contained a \" sequence, Exim got it wrong when building
5353 a From: or a Sender: header from that name. A second bug also caused
5354 incorrect handling when an unquoted " was present following a character
5355 that needed quoting.
5357 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5358 was not being matched caselessly.
5360 24. Arranged for all hyphens in the exim.8 source to be escaped with
5363 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5364 properly. Recipient callout cache records were still being keyed to include
5365 the sender, even when use_sender was set false. This led to far more
5366 callouts that were necessary. The sender is no longer included in the key
5367 when use_sender is false.
5369 26. Added "control = submission" modifier to ACLs.
5371 27. Added the ${base62d: operator to decode base 62 numbers.
5373 28. dnsdb lookups can now access SRV records.
5375 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5376 the configuration file.
5378 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5379 (-v) mode. This makes the output for a verbose queue run more intelligible.
5381 31. Added a use_postmaster feature to recipient callouts.
5383 32. Added the $body_zerocount variable, containing the number of binary zero
5384 bytes in the message body.
5386 33. The time of last modification of the "new" subdirectory is now used as the
5387 "mailbox time last read" when there is a quota error for a maildir
5390 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5392 35. Added +ignore_unknown as a special item in host lists.
5394 36. Code for decoding IPv6 addresses in host lists is now included, even if
5395 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5396 address was recognized as an IP address, but was then not correctly decoded
5397 into binary, causing unexpected and incorrect effects when compared with
5404 1. Very minor rewording of debugging text in manualroute to say "list of
5405 hosts" instead of "hostlist".
5407 2. If verify=header_syntax was set, and a header line with an unqualified
5408 address (no domain) and a large number of spaces between the end of the
5409 name and the colon was received, the reception process suffered a buffer
5410 overflow, and (when I tested it) crashed. This was caused by some obsolete
5411 code that should have been removed. The fix is to remove it!
5413 3. When running in the test harness, delay a bit after writing a bounce
5414 message to get a bit more predictability in the log output.
5416 4. Added a call to search_tidyup() just before forking a reception process. In
5417 theory, someone could use a lookup in the expansion of smtp_accept_max_
5418 per_host which, without the tidyup, could leave open a database connection.
5420 5. Added the variables $recipient_data and $sender_data which get set from a
5421 lookup success in an ACL "recipients" or "senders" condition, or a router
5422 "senders" option, similar to $domain_data and $local_part_data.
5424 6. Moved the writing of debug_print from before to after the "senders" test
5427 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5428 problems for message scanning, either using a data ACL, or using
5429 local_scan() because the Received: header was not generated till after they
5430 were called (in order to set the time as the time of reception completion).
5431 I have revised the way this works. The header is now generated after the
5432 body is received, but before the ACL or local_scan() are called. After they
5433 are run, the timestamp in the header is updated.
5439 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5440 before starting a queue runner without re-exec. This happened only when
5441 deliver_drop_privilege was set or when the Exim user was set to root. The
5442 effect of the bug was that timeouts during subsequent deliveries caused
5443 crashes instead of being properly handled. The handler is now left at its
5444 default (and expected) setting.
5446 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5447 message, again when deliver_drop_privilege is set or Exim is run as root.
5448 The bug described in (1) was not present in this case, but the tidying up
5449 of the other signals was missing. I have made the two cases consistent.
5451 3. The ignore_target_hosts setting on a manualroute router was being ignored
5452 for hosts that were looked up using the /MX notation.
5454 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5457 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5458 operated on the sender address. After changing the $sender_address to <>
5459 for the sender address verify, Exim was re-instated it as the original
5460 (before rewriting) address, but remembering that it had rewritten it, so it
5461 wasn't rewriting it again. This bug also had the effect of breaking the
5462 sender address verification caching when the sender address was rewritten.
5464 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5465 This has been changed so that if the ip literal address matches
5466 ignore_target_hosts, the router declines.
5468 7. Added expansion conditions match_domain, match_address, and match_local_
5469 part (NOT match_host).
5471 8. The placeholder for the Received: header didn't have a length field set.
5473 9. Added code to Exim itself and to exim_lock to test for a specific race
5474 condition that could lead to file corruption when using MBX delivery. The
5475 issue is with the lockfile that is created in /tmp. If this file is removed
5476 after a process has opened it but before that process has acquired a lock,
5477 there is the potential for a second process to recreate the file and also
5478 acquire a lock. This could lead to two Exim processes writing to the file
5479 at the same time. The added code performs the same test as UW imapd; it
5480 checks after acquiring the lock that its file descriptor still refers to
5481 the same named file.
5483 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5484 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5487 11. Added the smtp_active_hostname option. If used, this will typically be made
5488 to depend on the incoming interface address. Because $interface_address is
5489 not set up until the daemon has forked a reception process, error responses
5490 that can happen earlier (such as "too many connections") no longer contain
5493 12. If an expansion in a condition on a "warn" statement fails because a lookup
5494 defers, the "warn" statement is abandoned, and the next ACL statement is
5495 processed. Previously this caused the whole ACL to be aborted.
5497 13. Added the iplsearch lookup type.
5499 14. Added ident_timeout as a log selector.
5501 15. Added tls_certificate_verified as a log selector.
5503 16. Added a global option tls_require_ciphers (compare the smtp transport
5504 option of the same name). This controls incoming TLS connections.
5506 17. I finally figured out how to make tls_require_ciphers do a similar thing
5507 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5508 before starting the TLS session.
5510 18. Tabs are now shown as \t in -bP output.
5512 19. If the log selector return_path_on_delivery was set, Exim crashed when
5513 bouncing a message because it had too many Received: header lines.
5515 20. If two routers both had headers_remove settings, and the first one included
5516 a superfluous trailing colon, the final name in the first list and the
5517 first name in the second list were incorrectly joined into one item (with a
5518 colon in the middle).
5524 1. Added -C and -D options to the exinext utility, mainly to make it easier
5525 to include in the automated testing, but these could be helpful when
5526 multiple configurations are in use.
5528 2. The exinext utility was not formatting the output nicely when there was
5529 an alternate port involved in the retry record key, nor when there was a
5530 message id as well (for retries that were specific to a specific message
5531 and a specific host). It was also confused by IPv6 addresses, because of
5532 the additional colons they contain. I have fixed the IPv4 problem, and
5533 patched it up to do a reasonable job for IPv6.
5535 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5536 delivery, the log line now contains "pipelined" if PIPELINING was used.
5538 4. An SMTP transport process used to panic and die if the bind() call to set
5539 an explicit outgoing interface failed. This has been changed; it is now
5540 treated in the same way as a connect() failure.
5542 5. A reference to $sender_host_name in the part of a conditional expansion
5543 that was being skipped was still causing a DNS lookup. This no longer
5546 6. The def: expansion condition was not recognizing references to header lines
5547 that used bh_ and bheader_.
5549 7. Added the _cache feature to named lists.
5551 8. The code for checking quota_filecount in the appendfile transport was
5552 allowing one more file than it should have been.
5554 9. For compatibility with Sendmail, the command line option
5562 and sets the incoming protocol and host name (for trusted callers). The
5563 host name and its colon can be omitted when only the protocol is to be set.
5564 Note the Exim already has two private options, -pd and -ps, that refer to
5565 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5566 "s", but I don't think that's a major issue.
5568 10. A number of refactoring changes to the code, none of which should affect
5571 (a) The number of logging options was getting close to filling up the
5572 32-bit word that was used as a bit map. I have split them into two classes:
5573 those that are passed in the argument to log_write(), and those that are
5574 only ever tested independently outside of that function. These are now in
5575 separate 32-bit words, so there is plenty of room for expansion again.
5576 There is no change in the user interface or the logging behaviour.
5578 (b) When building, for example, log lines, the code previously used a
5579 macro that called string_cat() twice, in order to add two strings. This is
5580 not really sufficiently general. Furthermore, there was one instance where
5581 it was actually wrong because one of the argument was used twice, and in
5582 one call a function was used. (As it happened, calling the function twice
5583 did not affect the overall behaviour.) The macro has been replaced by a
5584 function that can join an arbitrary number of extra strings onto a growing
5587 (c) The code for expansion conditions now uses a table and a binary chop
5588 instead of a serial search (which was left over from when there were very
5589 few conditions). Also, it now recognizes conditions like "pam" even when
5590 the relevant support is not compiled in: a suitably worded error message is
5591 given if an attempt is made to use such a condition.
5593 11. Added ${time_interval:xxxxx}.
5595 12. A bug was causing one of the ddress fields not to be passed back correctly
5596 from remote delivery subprocesses. The field in question was not being
5597 subsequently used, so this caused to problems in practice.
5599 13. Added new log selectors queue_time and deliver_time.
5601 14. Might have fixed a bug in maildirsizefile handling that threw up
5602 "unexpected character" debug warnings, and recalculated the data
5603 unnecessarily. In any case, I expanded the warning message to give more
5606 15. Added the message "Restricted characters in address" to the statements in
5607 the default ACL that block characters like @ and % in local parts.
5609 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5610 Three changes have been made:
5612 (a) There was a serious bug; a negative response to MAIL caused the whole
5613 recipient domain to be cached as invalid, thereby blocking all messages
5614 to all local parts at the same domain, from all senders. This bug has
5615 been fixed. The domain is no longer cached after a negative response to
5616 MAIL if the sender used is not empty.
5618 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5621 (c) A new callout option, "use_sender" has been added for people who want
5622 the modified behaviour.
5628 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5631 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5632 indeed breaks things for older releases.
5634 3. Added additional logging to the case where there is a problem reading data
5635 from a filter that is running in a subprocess using a pipe, in order to
5636 try to track down a specific problem.
5638 4. Testing facility fudge: when running in the test harness and attempting
5639 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5640 getting "No route to host". Convert this to a timeout.
5642 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5645 6. Some OS don't have socklen_t but use size_t instead. This affects the
5646 fifth argument of getsockopt() amongst other things. This is now
5647 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5648 can be set for individual OS. I have set it for SunOS5, OSF1, and
5649 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5650 some earlier ones do not.
5652 7. Change 4.30/15 was not doing the test caselessly.
5654 8. The standard form for an IPv6 address literal was being rejected by address
5655 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5656 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5657 this, as well as the form without the "IPv6" on the front (but only when
5658 address literals are enabled, of course).
5660 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5662 10. Exim crashed if a message with an empty sender address specified by -f
5663 encountered a router with an errors_to setting. This could be provoked only
5664 by a command such as
5668 where an empty string was supplied; "<>" did not hit this bug.
5670 11. Installed PCRE release 4.5.
5672 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5673 remained set. It is now erased.
5675 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5676 times from message ids (which are base 36 rather than the normal 62).
5678 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5679 were being counted as actual protocol errors, and logged if the log
5680 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5681 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5682 and DATA following a set of rejected RCPTs do not count as protocol errors.
5683 In other words, Exim assumes they were pipelined, though this may not
5684 actually be the case. Of course, in all cases the client gets an
5685 appropriate error code.
5687 15. If a lookup fails in an ACL condition, a message about the failure may
5688 be available; it is used if testing the ACL cannot continue, because most
5689 such messages specify what the cause of the deferral is. However, some
5690 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5691 that caused an old message to be retained and used if a later statement
5692 caused a defer, replacing the real cause of the deferral.
5694 16. If an IP address had so many PTR records that the DNS lookup buffer
5695 was not large enough to hold them, Exim could crash while trying to process
5696 the truncated data. It now detects and logs this case.
5698 17. Further to 4.21/58, another change has been made: if (and only if) the
5699 first line of a message (the first header line) ends with CRLF, a bare LF
5700 in a subsequent header line has a space inserted after it, so as not to
5701 terminate the header.
5703 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5704 unnecessarily, used atoi() instead of strtol(), and didn't check the
5705 termination when getting file sizes from file names by regex.
5707 19. Completely re-implemented the support for maildirsize files, in the light
5708 of a number of problems with the previous contributed implementation
5709 (4.30/29). In particular:
5711 . If the quota is zero, the maildirsize file is maintained, but no quota is
5714 . If the maildir directory does not exist, it is created before any attempt
5715 to write a maildirsize file.
5717 . The quota value in the file is just a cache; if the quota is changed in
5718 the transport, the new value overrides.
5720 . A regular expression is available for excluding directories from the
5723 20. The autoreply transport checks the characters in options that define the
5724 message's headers; it allows continued headers, but it was checking with
5725 isspace() after an embedded newline instead of explicitly looking for a
5728 21. If all the "regular" hosts to which an address was routed had passed their
5729 expiry times, and had not reached their retry times, the address was
5730 bounced, even if fallback hosts were defined. Now Exim should go on to try
5733 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5734 equivalent code in the SMTP transport. Some hosts send humungous responses
5735 to HELO/EHLO, more than 1024 it seems.
5737 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5738 gives compiler warnings in some environments. I've now done it "properly",
5741 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5742 (because of problems with the built-in one) was declared to return uschar *
5743 instead of char *, causing compiler failure.
5745 25. Fixed a file descriptor leak when processing alias/forward files.
5747 26. Fixed a minor format string issue in dbfn.c.
5749 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5751 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5752 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5755 29. When a sender address is verified, it is cached, to save repeating the test
5756 when there is more than one recipient in a message. However, when the
5757 verification involves a callout, it is possible for different callout
5758 options to be set for different recipients. It is too complicated to keep
5759 track of this in the cache, so now Exim always runs a verification when a
5760 callout is required, relying on the callout cache for the optimization.
5761 The overhead is duplication of the address routing, but this should not be
5764 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5765 to be verified with callout=postmaster, and the main callout worked but the
5766 postmaster check failed, the verification correctly failed. However, if a
5767 subsequent RCPT command asked for sender verification *without* the
5768 postmaster check, incorrect caching caused this verification also to fail,
5771 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5772 it was not caching the DNS options (qualify_single, search_parents) that
5773 were used when the lookup failed. A subsequent lookup with different
5774 options therefore always gave the same answer, though there were cases
5775 where it should not have. (Example: a "domains = !$mx_any" option on a
5776 dnslookup router: the "domains" option is always processed without any
5777 widening, but the router might have qualify_single set.) Now Exim uses the
5778 cached value only when the same options are set.
5780 32. Added John Jetmore's "exipick" utility to the distribution.
5782 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5783 than a timeout (e.g. a certificate is required, and is not provided), an
5784 Exim server now closes the connection immediately. Previously it waited for
5785 the client to close - but if the client is SSL, it seems that they each
5786 wait for each other, leading to a delay before one of them times out.
5788 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5789 maintained 0.8.x compatibility because I don't think many are using it, and
5790 it is clearly obsolete.
5792 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5795 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5796 client certificate was expired. A simple patch fixes this, though I don't
5797 understand the full logic of why the verify callback is called multiple
5800 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5801 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5802 which causes problems with some clients (such as the Certicom SSL Plus
5803 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5804 disables the coutermeasure allowing Eudora to connect."
5806 38. Exim was not checking that a write() to a log file succeeded. This could
5807 lead to Bad Things if a log got too big, in particular if it hit a file
5808 size limit. Exim now panics and dies if it cannot write to a log file, just
5809 as it does if it cannot open a log file.
5811 39. Modified OS/Makefile-Linux so that it now contains
5813 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5815 The two -D definitions ensure that Exim is compiled with large file
5816 support, which makes it possible to handle log files that are bigger than
5819 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5820 instance) a domain was checked against a named list that involved a lookup,
5821 causing $domain_data to be set, then another domain was checked against the
5822 same list, then the first domain was re-checked, the value of $domain_data
5823 after the final check could be wrong. In particular, if the second check
5824 failed, it could be set empty. This bug probably also applied to
5827 41. The strip_trailing_dot option was not being applied to the address given
5828 with the -f command-line option.
5830 42. The code for reading a message's header from the spool was incrementing
5831 $received_count, but never initializing it. This meant that the value was
5832 incorrect (doubled) while delivering a message in the same process in which
5833 it was received. In the most common configuration of Exim, this never
5834 happens - a fresh exec is done - but it can happen when
5835 deliver_drop_privilege is set.
5837 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5838 it now includes up to 150 characters of the unexpected data in the log
5841 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5842 and building data strings. The size of both of these buffers was 10 000
5843 bytes - far larger than anybody would *ever* want, thought I. Needless to
5844 say, somebody hit the limit. I have increased the maximum line length to
5845 20 000 and the maximum data length of concatenated lines to 100 000. I have
5846 also fixed two bugs, because there was no checking on these buffers. Tsk,
5847 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5848 buffer is too small.
5850 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5851 lsearch lookups. Now it does.
5853 46. When parsing a route_list item in a manualroute router, a fixed-length
5854 buffer was used for the list of hosts. I made this 1024 bytes long,
5855 thinking that nobody would ever have a list of hosts that long. Wrong.
5856 Somebody had a whole pile of complicated expansion conditions, and the
5857 string was silently truncated, leading to an expansion error. It turns out
5858 that it is easier to change to an unlimited length (owing to other changes
5859 that have happened since this code was originally written) than to build
5860 structure for giving a limitation error. The length of the item that
5861 expands into the list of hosts is now unlimited.
5863 47. The lsearch lookup could not handle data where the length of text line was
5864 more than 4095 characters. Such lines were truncated, leading to shortened
5865 data being returned. It should now handle lines of any length.
5867 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5868 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5871 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5872 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5873 no longer happens in comment lines. A list of replacements is now placed
5874 at the head of all of the source files, except those whose only change is
5875 to replace PERL_COMMAND in the very first #! line.
5877 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5878 messages on the queue, with a bottom-up merge sort, using code contributed
5879 by Michael Haardt. This should make operations like -bp somewhat faster on
5880 large queues. It won't affect queue runners, except when queue_run_in_order
5883 51. Installed eximstats 1.31 in the distribution.
5885 52. Added support for SRV lookups to the dnslookup router.
5887 53. If an ACL referred to $message_body or $message_body_end, the value was not
5888 reset for any messages that followed in the same SMTP session.
5890 54. The store-handling optimization for building very long strings was not
5891 differentiating between the different store pools. I don't think this
5892 actually made any difference in practice, but I've tidied it.
5894 55. While running the routers to verify a sender address, $sender_address
5895 was still set to the sender address. This is wrong, because when routing to
5896 send a bounce to the sender, it would be empty. Therefore, I have changed
5897 it so that, while verifying a sender address, $sender_address is set to <>.
5898 (There is no change to what happens when verifying a recipient address.)
5900 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5901 target A or AAAA records (if not already returned) without resetting the
5902 qualify_single or search_parents options of the DNS resolver. These are
5903 inappropriate in this case because the targets of MX and SRV records must
5904 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5905 target that, when qualified, matched something in the local domain. These
5906 two options are now turned off when doing these lookups.
5908 57. It seems that at least some releases of Reiserfs (which does not have the
5909 concept of a fixed number of inodes) returns zero and not -1 for the
5910 number of available inodes. This interacted badly with check_spool_inodes,
5911 which assumed that -1 was the "no such thing" setting. What I have done is
5912 to check that the total number of inodes is greater than zero before doing
5913 the test of how many are available.
5915 58. When a "warn" ACL statement has a log_message modifier, the message is
5916 remembered, and not repeated. This is to avoid a lot of repetition when a
5917 message has many recipients that cause the same warning to be written.
5918 However, Exim was preserving the list of already written lines for an
5919 entire SMTP session, which doesn't seem right. The memory is now reset if a
5920 new message is started.
5922 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5923 debugging output unless log_rewrite was also set.
5925 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5926 of (int)(handle) when we know that handle contains (void *)(-1).
5928 61. The Exim daemon panic-logs an error return when it closes the incoming
5929 connection. However "connection reset by peer" seems to be common, and
5930 isn't really an error worthy of noting specially, so that particular error
5933 62. When Exim is trying to find all the local interfaces, it used to panic and
5934 die if the ioctl to get the interface flags failed. However, it seems that
5935 on at least one OS (Solaris 9) it is possible to have an interface that is
5936 included in the list of interfaces, but for which you get a failure error
5937 for this call. This happens when the interface is not "plumbed" into a
5938 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5939 failure of the "get flags" call assumes that the interface is down.
5941 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5942 makes life easier for people who are doing arithmetic on fields extracted
5943 from dates, where you often get leading zeros that should not be
5944 interpreted as octal.
5946 64. Added qualify_domain to the redirect router, to override the global
5949 65. If a pathologically long header line contained very many addresses (the
5950 report of this problem mentioned 10 000) and each of them was rewritten,
5951 Exim could use up a very large amount of memory. (It kept on making new
5952 copies of the header line as it rewrote, and never released the old ones.)
5953 At the expense of a bit more processing, the header rewriting function has
5954 been changed so that it no longer eats memory in this way.
5956 66. The generation of the Received: header has been moved from the time that a
5957 message starts to be received, to the time that it finishes. The timestamp
5958 in the Received: header should now be very close to that of the <= log
5959 line. There are two side-effects of this change:
5961 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5962 logged header lines no longer include the local Received: line, because
5963 it has not yet been created. The same applies to a copy of the message
5964 that is returned to a non-SMTP sender when a message is rejected.
5966 (b) When a filter file is tested using -bf, no additional Received: header
5967 is added to the test message. After some thought, I decided that this
5970 This change does not affect the value of $received_for. It is still set
5971 after address rewriting, but before local_scan() is called.
5973 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5975 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5976 gave an unhelpful panic error message, and a defer error. I have managed to
5977 change this behaviour so that it now rejects any supplied certificate,
5978 which seems right, as the list of acceptable certificates is empty.
5980 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5981 gave an unhelpful defer error. I have not managed to make this reject any
5982 supplied certificates, but the error message it gives is "no certificate
5983 supplied", which is not helpful.
5985 70. exigrep's output now also includes lines that are not associated with any
5986 message, but which match the given pattern. Implemented by a patch from
5987 Martin Sluka, which also tidied up the Perl a bit.
5989 71. Recipient callout verification, like sender verification, was using <> in
5990 the MAIL FROM command. This isn't really the right thing, since the actual
5991 sender may affect whether the remote host accepts the recipient or not. I
5992 have changed it to use the actual sender in the callout; this means that
5993 the cache record is now keyed on a recipient/sender pair, not just the
5994 recipient address. There doesn't seem to be a real danger of callout loops,
5995 since a callout by the remote host to check the sender would use <>.
5996 [SEE ABOVE: changed after hitting problems.]
5998 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5999 temporary errors. However, in the case of such a code being given after
6000 the end of a data transmission (i.e. after ".") Exim was failing to write
6001 a retry record for the message. (Yes, there was some broken host that was
6002 actually sending 8xx at this point.)
6004 73. An unknown lookup type in a host list could cause Exim to panic-die when
6005 the list was checked. (An example that provoked this was putting <; in the
6006 middle of a list instead of at the start.) If this happened during a DATA
6007 ACL check, a -D file could be left lying around. This kind of configuration
6008 error no longer causes Exim to die; instead it causes a defer error. The
6009 incident is still logged to the main and panic logs.
6011 74. Buglet left over from Exim 3 conversion. The message "too many messages
6012 in one connection" was written to the rejectlog but not the mainlog, except
6013 when address rewriting (yes!) was being logged.
6015 75. Added write_rejectlog option.
6017 76. When a system filter was run not as root (that is, when system_filter_user
6018 was set), the values of the $n variables were not being returned to the
6019 main process; thus, they were not subsequently available in the $sn
6022 77. Added +return_path_on_delivery log selector.
6024 78. A connection timeout was being treated differently from recipients deferred
6025 when testing hosts_max_try with a message that was older than the host's
6026 retry timeout. (The host should not be counted, thus allowing all hosts to
6027 be tried at least once before bouncing.) This may have been the cause of an
6028 occasionally reported bug whereby a message would remain on the queue
6029 longer than the retry timeout, but would be bounced if a delivery was
6030 forced. I say "may" because I never totally pinned down the problem;
6031 setting up timeout/retry tests is difficult. See also the next item.
6033 79. The ultimate address timeout was not being applied to errors that involved
6034 a combination of host plus message (for example, a timeout on a MAIL
6035 command). When an address resolved to a number of possible hosts, and they
6036 were not all tried for each delivery (e.g. because of hosts_max_try), a
6037 message could remain on the queue longer than the retry timeout.
6039 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6042 81. Fixed an obscure SMTP outgoing bug which required at least the following
6043 conditions: (a) there was another message waiting for the same server;
6044 (b) the server returned 5xx to all RCPT commands in the first message so
6045 that the message was not completed; (c) the server dropped the connection
6046 or gave a negative response to the RSET that Exim sends to abort the
6047 transaction. The observed case was a dropped connection after DATA that had
6048 been sent in pipelining mode. That is, the server had advertised PIPELINING
6049 but was not implementing it correctly. The effect of the bug was incorrect
6050 behaviour, such as trying another host, and this could lead to a crash.
6056 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6057 and daemon.c were passed as pointers to ints; they should have been
6058 pointers to socklen_t variables (which are typically unsigned ints).
6060 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6063 3. Fixed a really odd bug that affected only the testing scheme; patching a
6064 certain fixed string in the binary changed the value of another string that
6065 happened to be identical to the end of the original first string.
6067 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6068 name", it returns that address as the IP address. On some operating
6069 systems (e.g. Solaris), it also passes back the IP address string as the
6070 "host name". However, on others (e.g. Linux), it passes back an empty
6071 string. Exim wasn't checking for this, and was changing the host name to an
6072 empty string, assuming it had been canonicalized.
6074 5. Although rare, it is permitted to have more than one PTR record for a given
6075 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6076 all the names associated with an address, because they do in Solaris.
6077 However, it seems that they do not in Linux for data that comes from the
6078 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6079 I found this out when I moved to a new Linux workstation and tried to run
6080 the Exim test suite.
6082 To get round this problem I have changed the code so that it now does its
6083 own call to the DNS to look up PTR records when searching for a host name.
6084 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6085 addresses that are only in /etc/hosts are still found.
6087 This behaviour is, however, controlled by an option called host_lookup_
6088 order, which defaults to "bydns:byaddr". If people want to use the other
6089 order, or indeed, just use one or the other means of lookup, they can
6090 specify it in this variable.
6092 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6093 some operating systems, this comes back from gethostbyaddr() as an empty
6094 string, and this is what Exim used to test for. However, it seems that in
6095 other systems, "." is yielded. Exim now tests for this case too.
6097 7. The values of check_spool_space and check_log_space are now held internally
6098 as a number of kilobytes instead of an absolute number of bytes. If a
6099 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6100 kilobyte. This means that much larger values can be stored.
6102 8. Exim monitor: an attempt to get the action menu when not actually pointing
6103 at a message produces an empty menu entitled "No message selected". This
6104 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6105 no entries in it ("Shell widget menu has zero width and/or height"). So I
6106 have added a single, blank menu entry in this case.
6108 9. Added ${quote_local_part.
6110 10. MIME decoding is now applied to the contents of Subject: header lines when
6113 11. Now that a reference to $sender_host_address automatically causes a reverse
6114 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6115 host lookup before query-style lookups in lists that might use this
6116 variable. This has therefore been abolished, and the "net-" prefix is no
6117 longer necessary for query-style lookups.
6119 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6120 have been a typo for LFLAGS, so it has been changed.
6122 13. The install script calls Exim with "-C /dev/null" in order to find the
6123 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6124 to be output. However, since Exim outputs its version number before the
6125 error, it didn't break the script. It just looked ugly. I fixed this by
6126 always allowing "-C /dev/null" if the caller is root.
6128 14. Ignore overlarge ACL variable number when reading spool file - insurance
6129 against a later release with more variables having written the file.
6131 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6132 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6133 this, as well as the form without the "IPv6" on the front.
6135 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6136 OS/Makefile-Darwin file.
6138 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6139 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6141 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6142 variable. However, some OS use ints for inodes. I've added cast to long int
6143 to get rid of the compiler warning.
6145 19. I had forgotten to lock out "/../" in configuration file names when
6146 ALT_CONFIG_PREFIX was set.
6148 20. Routers used for verification do not need to specify transports. However,
6149 if such a router generated a host list, and callout was configured, Exim
6150 crashed, because it could not find a port number from the (non-existent)
6151 transport. It now assumes port 25 in this circumstance.
6153 21. Added the -t option to exigrep.
6155 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6156 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6157 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6158 them is set, LOOKUP_LSEARCH is forced.
6160 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6162 24. Added sender and host information to the "rejected by local_scan()" log
6163 line; previously there was no indication of these.
6165 25. Added .include_if_exists.
6167 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6168 Linux. It turns out that not all file systems support this. Apparently some
6169 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6170 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6171 not supported on the file descriptor, is now ignored when Exim is trying to
6172 sync a directory. This applies only to Linux.
6174 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6176 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6177 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6180 29. Added support for maildirsize files from supplied patch (modified a bit).
6182 30. The use of :fail: followed by an empty string could lead Exim to respond to
6183 sender verification failures with (e.g.):
6185 550 Verification failed for <xxx>
6186 550 Sender verify failed
6188 where the first response line was missing the '-' that indicates it is not
6189 the final line of the response.
6191 31. The loop for finding the name of the user that called Exim had a hardwired
6192 limit of 10; it now uses the value of finduser_retries, which is used for
6193 all other user lookups.
6195 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6198 33. Exim was neglecting to zero errno before one call of strtol() when
6199 expanding a string and expecting an integer value. On some systems this
6200 resulted in spurious "integer overflow" errors. Also, it was casting the
6201 result into an int without checking.
6203 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6204 did not work. The code looks as if it has *never* worked, though it appears
6205 to have been documented since at least release 1.62. I have made it work.
6207 35. The "timeout_DNS" error in retry rules, also documented since at least
6208 1.62, also never worked. As it isn't clear exactly what this means, and
6209 clearly it isn't a major issue, I have abolished the feature by treating it
6210 as "timeout", and writing a warning to the main and panic logs.
6212 36. The display of retry rules for -brt wasn't always showing the error code
6215 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6216 timeout_connect_A, timeout_connect_MX.
6218 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6219 to the empty sender.
6221 39. The daemon was not analysing the content of -oX till after it had closed
6222 stderr and disconnected from the controlling terminal. This meant that any
6223 syntax errors were only noted on the panic log, and the return code from
6224 the command was 0. By re-arranging the code a little, I've made the
6225 decoding happen first, so such errors now appear on stderr, and the return
6226 code is 1. However, the actual setting up of the sockets still happens in
6227 the disconnected process, so errors there are still only recorded on the
6230 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6231 connections (as happens on some IP stacks) was logged at start up time as
6232 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6233 it from "IPv6 and IPv4", which means that two separate sockets are being
6236 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6237 says whether AF_INET or AF_INET6 was passed as an argument.
6239 42. Exiwhat output was messed up when time zones were included in log
6242 43. Exiwhat now gives more information about the daemon's listening ports,
6243 and whether -tls-on-connect was used.
6245 44. The "port" option of the smtp transport is now expanded.
6247 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6248 silently ignored. Now an error message is written to the main and panic
6251 46. There's a new ACL modifier called "logwrite" which writes to a log file
6252 as soon as it is encountered.
6254 47. Added $local_user_uid and $local_user_gid at routing time.
6256 48. Exim crashed when trying to verify a sender address that was being
6259 49. Exim was recognizing only a space character after ".include". It now also
6260 recognizes a tab character.
6262 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6263 extracting the relevant information from the specification. The man page no
6264 longer contains scrambled data for the -d option, and I've added a section
6265 at the front about calling Exim under different names.
6267 51. Added "extra_headers" argument to the "mail" command in filter files.
6269 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6272 53. Installed eximstats 1.29.
6274 54. Added transport_filter_timeout as a generic transport option.
6276 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6277 Cc: header lines. This was required by RFC 822, but it not required by RFC
6280 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6281 incoming messages that did not have them. Now it does so only if the
6282 message originates locally, that is, if there is no associated remote host
6283 address. When Resent- header lines are present, this applies to the Resent-
6284 lines rather than the non-Resent- lines.
6286 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6287 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6289 58. Messages for configuration errors now include the name of the main
6290 configuration file - useful now that there may be more than one file in a
6291 list (.included file names were always shown).
6293 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6294 for those rare installations that do not start the daemon as root or run it
6295 setuid root. I've cut out the call to initgroups() if the daemon is not
6298 60. The Exim user and group can now be bound into the binary as text strings
6299 that are looked up at the start of Exim's processing.
6301 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6303 62. Added $mailstore_basename variable.
6305 63. Installed patch to sieve.c from Michael Haardt.
6307 64. When Exim failed to open the panic log after failing to open the main log,
6308 the original message it was trying to log was written to stderr and debug
6309 output, but if they were not available (the usual case in production), it
6310 was lost. Now it is written to syslog before the two lines that record the
6311 failures to open the logs.
6313 65. Users' Exim filters run in subprocesses under the user's uid. It is
6314 possible for a "deliver" command or an alias in a "personal" command to
6315 provoke an address rewrite. If logging of address rewriting is configured,
6316 this fails because the process is not running as root or exim. There may be
6317 a better way of dealing with this, but for the moment (because 4.30 needs
6318 to be released), I have disabled address rewrite logging when running a
6319 filter in a non-root, non-exim process.
6325 1. The buildconfig auxiliary program wasn't quoting the value set for
6326 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6327 not defined. This bug was masked in 4.22 by the effect that was fixed in
6330 2. Some messages that were rejected after a message id was allocated were
6331 shown as "incomplete" by exigrep. It no longer does this for messages that
6332 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6334 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6335 have allow_domain_literals set, the ID did not get logged in the <= line.
6336 Domain literals are now always recognized in Message-ID: header lines.
6338 4. The first argument for a ${extract expansion item is the key name or field
6339 number. Leading and trailing spaces in this item were not being ignored,
6340 causing some misleading effects.
6342 5. When deliver_drop_privilege was set, single queue runner processes started
6343 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6344 same command in the process it spins off) were not dropping privilege.
6346 6. When the daemon running as "exim" started a queue runner, it always
6347 re-executed Exim in the spun-off process. This is a waste of effort when
6348 deliver_drop_privilege is set. The new process now just calls the
6349 queue-runner function directly.
6355 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6358 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6359 ignored. Though the use of -oP was forcing the writing of a pid file, it
6360 was always written to the default place.
6362 3. If the message "no IP address found for host xxxx" is generated during
6363 incoming verification, it is now followed by identification of the incoming
6364 connection (so you can more easily find what provoked it).
6366 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6368 5. Added some features to "harden" Exim a bit more against certain attacks:
6370 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6371 be put in Local/Makefile. This is like the never_users runtime option,
6372 but it cannot be overridden. The default setting is "root".
6374 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6375 prefix string with which any file named in a -C command line option
6378 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6379 is retained for -C and -D only if the caller of Exim is root. Without
6380 it, the exim user may also use -C and -D and retain privilege.
6382 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6383 command line option is disabled.
6385 6. Macro names set by the -D option must start with an upper case letter, just
6386 like macro names defined in the configuration file.
6388 7. Added "dereference=" facility to LDAP.
6390 8. Two instances of the typo "uknown" in the source files are fixed.
6392 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6393 the Configure-Makefile script screwed up while processing it.
6395 10. Incorporated PCRE 4.4.
6397 11. The SMTP synchronization check was not operating right at the start of an
6398 SMTP session. For example, it could not catch a HELO sent before the client
6399 waited for the greeting. There is now a check for outstanding input at the
6400 point when the greeting is written. Because of the duplex, asynchronous
6401 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6402 way, but not yet received, when the check is performed.
6404 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6405 on TCP/IP sockets, because this apparently causes some broken clients to
6408 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6409 unchanged) from the Cygwin maintainer.
6411 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6412 was supported instead of testing for mbx. Effectively a typo.
6414 15. The spa authenticator server code was not checking that the input it
6415 received was valid base64.
6417 16. The debug output line for the "set" modifier in ACLs was not showing the
6418 name of the variable that was being set.
6420 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6422 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6423 lookup on its own. Something else was needed to trigger the lookup. For
6424 example, a match in host_lookup or the need for a host name in a host list.
6425 Now, if $sender_host_name is referenced and the host name has not yet been
6426 looked up, a lookup is performed. If the lookup fails, the variable remains
6427 empty, and $host_lookup_failed is set to "1".
6429 19. Added "eqi" as a case-independent comparison operator.
6431 20. The saslauthd authentication condition could segfault if neither service
6432 nor realm was specified.
6434 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6435 error message that was logged was misleading, and incoming SMTP
6436 connections were dropped. The message is now more accurate, and temporary
6437 errors are given to SMTP connections.
6439 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6440 (except RSET) with a 421 error, until QUIT is received. However, it was
6441 failing to send a response to QUIT.
6443 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6444 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6445 ACL is now run after the helo_try_verify_hosts code.
6447 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6448 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6449 case-independent, so other case variants are also recognized). Apparently
6450 some systems use these upper case variants.
6452 25. If more than two messages were waiting for the same host, and a transport
6453 filter was specified for the transport, Exim sent two messages over the
6454 same TCP/IP connection, and then failed with "socket operation on non-
6455 socket" when it tried to send the third.
6457 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6459 27. The extern definition of crypt16() in expand.c was not being excluded when
6460 the OS had its own crypt16() function.
6462 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6463 as a preferred synonym for return_size_limit, both as an option and as an
6466 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6468 30. Changed the default configuration ACL to relax the local part checking rule
6469 for addresses that are not in any local domains. For these addresses,
6470 slashes and pipe symbols are allowed within local parts, but the sequence
6471 /../ is explicitly forbidden.
6473 31. SPA server authentication was not clearing the challenge buffer before
6476 32. log_message in a "warn" ACL statement was writing to the reject log as
6477 well as to the main log, which contradicts the documentation and doesn't
6478 seem right (because no rejection is happening). So I have stopped it.
6480 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6481 However, I am unable to do any testing of this.
6483 34. Fixed an infelicity in the appendfile transport. When checking directories
6484 for a mailbox, to see if any needed to be created, it was accidentally
6485 using path names with one or more superfluous leading slashes; tracing
6486 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6488 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6489 contents are added to the log line that is written for every discarded
6490 recipient. (Previously a log_message setting was ignored.)
6492 36. The ${quote: operator now quotes the string if it is empty.
6494 37. The install script runs exim in order to find its version number. If for
6495 some reason other than non-existence or emptiness, which it checks, it
6496 could not run './exim', it was installing it with an empty version number,
6497 i.e. as "exim-". This error state is now caught, and the installation is
6500 38. An argument was missing from the function that creates an error message
6501 when Exim fails to connect to the socket for saslauthd authentication.
6502 This could cause Exim to crash, or give a corrupted message.
6504 39. Added isip, isip4, and isip6 to ${if conditions.
6506 40. The ACL variables $acl_xx are now saved with the message, and can be
6507 accessed later in routers, transports, and filters.
6509 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6510 strings in the file are not string-expanded.
6512 42. If a MAIL command specified a SIZE value that was too large to fit into an
6513 int variable, the check against message_size_limit failed. Such values are
6514 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6515 day this will have to be increased, but I don't think I want to be around
6516 when emails are that large.
6523 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6524 iconv() is not standard in FreeBSD.
6526 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6527 IPv6 enabled. The observed symptom was a segmentation fault on return from
6528 the function os_common_find_running_interfaces() in src/os.c.
6530 3. In the check_special_case() function in daemon.c I had used "errno" as an
6531 argument name, which causes warnings on some systems. This was basically a
6532 typo, since it was named "eno" in the comments!
6534 4. The code that waits for the clock to tick (at a resolution of some fraction
6535 of a second) so as to ensure message-id uniqueness was always waiting for
6536 at least one whole tick, when it could have waited for less. [This is
6537 almost certainly not relevant at current processor speeds, where it is
6538 unlikely to ever wait at all. But we try to future-proof.]
6540 5. The function that sleeps for a time interval that includes fractions of a
6541 second contained a race. It did not block SIGALRM between setting the
6542 timer, and suspending (a couple of lines later). If the interval was short
6543 and the sigsuspend() was delayed until after it had expired, the suspension
6544 never ended. On busy systems this could lead to processes getting stuck for
6547 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6548 process, before it forks any delivery processes. The open lookup caching
6549 mechanism meant that the open file or database connection was passed into
6550 the delivery process. The problem was that delivery processes always tidy
6551 up cached lookup data. This could cause a problem for the next delivery
6552 process started by the queue runner, because the external queue runner
6553 process does not know about the closure. So the next delivery process
6554 still has data in the lookup cache. In the case of a file lookup, there was
6555 no problem because closing a file descriptor in a subprocess doesn't affect
6556 the parent. However, if the lookup was caching a connection to a database,
6557 the connection was closed, and the second delivery process was likely to
6558 see errors such as "PGSQL: query failed: server closed the connection
6559 unexpectedly". The problem has been fixed by closing all cached lookups
6560 in a queue runner before running a delivery process.
6562 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6563 seem to have the "const" qualifier which it has on other OS. I've
6566 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6567 *of the same type* (client or server) with the same public name that an
6568 error should be diagnosed.
6570 9. When Exim looked up a host name for an IP address, but failed to find the
6571 original IP address when looking up the host name (a safety check), it
6572 output the message "<ip address> does not match any IP for NULL", which was
6573 confusing, to say the least. The bug was that the host name should have
6574 appeared instead of "NULL".
6576 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6577 user that is built into the binary, and the -C or -D options is used, root
6578 privilege is dropped before the configuration file is read. In addition,
6579 logging is switched to stderr instead of the normal log files. If the
6580 configuration then re-defines the Exim user, the unprivileged environment
6581 is probably not what is expected, so Exim logs a panic warning message (but
6584 However, if deliver_drop_privilege is set, the unprivileged state may well
6585 be exactly what is intended, so the warning has been cut out in that case,
6586 and Exim is allowed to try to write to its normal log files.
6592 1. smtp_return_error_details was not giving details for temporary sender
6593 or receiver verification errors.
6595 2. Diagnose a configuration error if two authenticators have the same public
6598 3. Exim used not to create the message log file for a message until the first
6599 delivery attempt. This could be confusing when incoming messages were held
6600 for policy or load reasons. The message log file is now created at the time
6601 the message is received, and an initial "Received" line is written to it.
6603 4. The automatically generated man page for command line options had a minor
6604 bug that caused no ill effects; however, a more serious problem was that
6605 the procedure for building the man page automatically didn't always
6606 operate. Consequently, release 4.20 contains an out-of-date version. This
6607 shouldn't happen again.
6609 5. When building Exim with embedded Perl support, the script that builds the
6610 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6611 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6613 6. The freeze_tell option was not being used for messages that were frozen on
6614 arrival, either by an ACL or by local_scan().
6616 7. Added the smtp_incomplete_transaction log selector.
6618 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6619 was accepting AUTH without a new EHLO.
6621 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6622 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6625 10. Logging of TCP/IP connections (when configured) now happens in the main
6626 daemon process instead of the child process, so that the TCP/IP connection
6627 count is more accurate (but it can never be perfect).
6629 11. The use of "drop" in a nested ACL was not being handled correctly in the
6630 outer ACL. Now, if condition failure induced by the nested "drop" causes
6631 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6632 or "require"), the connection is dropped.
6634 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6635 that yield "discard" can now be used with an "accept" or a "discard" verb,
6636 but an error is generated for any others (because I can't see a useful way
6637 to define what should happen).
6639 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6640 are now processed in the same way as lines in the Exim configuration file.
6641 In particular, continuation lines are supported.
6643 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6645 15. Added -ti meaning -t -i.
6647 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6648 domains, and warn by logging if others are found.
6650 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6651 returned from GIFCONF when Exim is trying to find the list of interfaces on
6652 a host. The code in os.c has been modified to copy each ifreq to an aligned
6653 structure in all cases.
6655 Also, in some cases, the returned ifreq's were being copied to a 'struct
6656 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6657 means the last couple of bytes of an IPv6 address could be chopped if the
6658 ifreq contained only a normal sockaddr (14 bytes storage).
6660 18. Named domain lists were not supported in the hosts_treat_as_local option.
6661 An entry such as +xxxx was not recognized, and was treated as a literal
6664 19. Ensure that header lines added by a DATA ACL are included in the reject log
6665 if the ACL subsequently rejects the message.
6667 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6668 MD5 (which is deprecated).
6670 21. When testing a filter file using -bf, Exim was writing a message when it
6671 took the sender from a "From " line in the message, but it was not doing so
6672 when it took $return_path from a Return-Path: header line. It now does.
6674 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6675 with a valid header line field name (a series of printing characters
6676 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6678 23. Changed "disc" in the source to "disk" to conform to the documentation and
6679 the book and for uniformity.
6681 24. Ignore Sendmail's -Ooption=value command line item.
6683 25. When execve() failed while trying to run a command in a pipe transport,
6684 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6685 could be confused with a return value of 69 from the command itself. This
6686 has been changed to 127, the value the shell returns if it is asked to run
6687 a non-existent command. The wording for the related log line suggests a
6688 non-existent command as the problem.
6690 26. If received_header_text expands to an empty string, do not add a Received:
6691 header line to the message. (Well, it adds a token one on the spool, but
6692 marks it "old" so that it doesn't get used or transmitted.)
6694 27. Installed eximstats 1.28 (addition of -nt option).
6696 28. There was no check for failure on the call to getsockname() in the daemon
6697 code. This can fail if there is a shortage of resources on the system, with
6698 ENOMEM, for example. A temporary error is now given on failure.
6700 29. Contrary to the C standard, it seems that in some environments, the
6701 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6702 program. Exim now does this explicitly; it affects the formatting of
6703 timestamps using strftime().
6705 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6706 complaints. I've now initialized all the variables, to avoid this.
6708 32. Header lines added by a system filter were not being "seen" during
6709 transport-time rewrites.
6711 33. The info_callback() function passed to OpenSSL is set up with type void
6712 (*)(SSL *, int, int), as described somewhere. However, when calling the
6713 function (actually a macro) that sets it up, the type void(*)() is
6714 expected. I've put in a cast to prevent warnings from picky compilers.
6716 34. If a DNS black list lookup found a CNAME record, but there were no A
6717 records associated with the domain it pointed at, Exim crashed.
6719 35. If a DNS black list lookup returned more than one A record, Exim ignored
6720 all but the first. It now scans all returned addresses if a particular IP
6721 value is being sought. In this situation, the contents of the
6722 $dnslist_value variable are a list of all the addresses, separated by a
6725 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6726 to accept a host name and all its aliases if the forward lookup for any of
6727 them yielded the IP address of the incoming connection. Now it accepts only
6728 those names whose forward lookup yields the correct IP address. Any other
6729 names are discarded. This closes a loophole whereby a rogue DNS
6730 administrator could create reverse DNS records to break through a
6731 wildcarded host restriction in an ACL.
6733 37. If a user filter or a system filter that ran in a subprocess used any of
6734 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6735 the wrong values were passed to the pipe command ($thisaddress had the
6736 value of $0, $0 had the value of $1, etc). This bug was introduced by
6737 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6739 38. Improved the line breaking for long SMTP error messages from ACLs.
6740 Previously, if there was no break point between 40 and 75 characters, Exim
6741 left the rest of the message alone. Two changes have been made: (a) I've
6742 reduced the minimum length to 35 characters; (b) if it can't find a break
6743 point between 35 and 75 characters, it looks ahead and uses the first one
6744 that it finds. This may give the occasional overlong line, but at least the
6745 remaining text gets split now.
6747 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6748 file descriptors might be low, and that setting 1000 would always raise it.
6749 It turns out that in some environments, the limit is already over 1000 and
6750 that lowering it causes trouble. So now Exim takes care not to decrease it.
6752 40. When delivering a message, the value of $return_path is set to $sender_
6753 address at the start of routing (routers may change the value). By an
6754 oversight, this default was not being set up when an address was tested by
6755 -bt or -bv, which affected the outcome if any router or filter referred to
6758 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6759 log only when "message" or "log_message" are set. However, if one of the
6760 conditions was an address verification, or a call to a nested ACL, the
6761 messages generated by the underlying test were being passed through. This
6762 no longer happens. The underlying message is available in $acl_verify_
6763 message for both "message" and "log_message" expansions, so it can be
6764 passed through if needed.
6766 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6767 new expansion $bh_ to give the encoded byte string without charset
6768 translation. Translation happens only if iconv() is available; HAVE_ICONV
6769 indicates this at build time. HEADERS_CHARSET gives the charset to
6770 translate to; headers_charset can change it in the configuration, and
6771 "headers charset" can change it in an individual filter file.
6773 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6774 that creates RFC 2047 encoded "words" labels them as that charset instead
6775 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6776 expansion operator; (ii) when Exim creates a From: line for a local
6777 message; (iii) when a header line is rewritten to include a "phrase" part.
6779 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6780 buggy, causing it to skip the first lines of messages whose message ID
6781 ended in 'D'. This would not have bitten before Exim release 4.14, because
6782 message IDs were unlikely to end in 'D' before then. The effect was to have
6783 incorrect size information for certain domains.
6785 45. #include "config.h" was missing at the start of the crypt16.c module. This
6786 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6789 46. If there was a timeout during a "random" callout check, Exim treated it as
6790 a failure of the random address, and carried on sending RSET and the real
6791 address. If the delay was just some slowness somewhere, the response to the
6792 original RCPT would be taken as a response to RSET and so on, causing
6793 mayhem of various kinds.
6795 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6796 when I implemented it. It didn't allow for the fact that some option values
6797 may legitimately be negative (e.g. size_addition), and it didn't even do
6798 the right test for positive values.
6800 48. Domain names in DNS records are case-independent. Exim always looks them up
6801 in lower case. Some resolvers return domain names in exactly the case they
6802 appear in the zone file, that is, they may contain uppercase letters. Not
6803 all resolvers do this - some return always lower case. Exim was treating a
6804 change of case by a resolver as a change of domain, similar to a widening
6805 of a domain abbreviation. This triggered its re-routing code and so it was
6806 trying to route what was effectively the same domain again. This normally
6807 caused routing to fail (because the router wouldn't handle the domain
6808 twice). Now Exim checks for this case specially, and just changes the
6809 casing of the domain that it ultimately uses when it transmits the message
6812 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6815 50. If a filter generated a file delivery with a non-absolute name (possible if
6816 no home directory exists for the router), the forbid_file option was not
6819 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6820 the existing equality matching.
6822 52. Exim was using ints instead of ino_t variables in some places where it was
6823 dealing with inode numbers.
6825 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6826 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6827 called TMPDIR, and if it finds it is different, it changes its value.
6829 54. The smtp_printf() function is now made available to local_scan() so
6830 additional output lines can be written before returning. There is also an
6831 smtp_fflush() function to enable the detection of a dropped connection.
6832 The variables smtp_input and smtp_batched_input are exported to
6835 55. Changed the default runtime configuration: the message "Unknown user"
6836 has been removed from the ACL, and instead placed on the localuser router,
6837 using the cannot_route_message feature. This means that any verification
6838 failures that generate their own messages won't get overridden. Similarly,
6839 the "Unrouteable address" message that was in the ACL for unverifiable
6840 relay addresses has also been removed.
6842 56. Added hosts_avoid_esmtp to the smtp transport.
6844 57. The exicyclog script was not checking for the esoteric option
6845 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6846 will work only if exicyclog is run under the appropriate euid.
6848 58. Following a discussion on the list, the rules by which Exim recognises line
6849 endings on incoming messages have been changed. The -dropcr and drop_cr
6850 options are now no-ops, retained only for backwards compatibility. The
6851 following line terminators are recognized: LF CRLF CR. However, special
6852 processing applies to CR:
6854 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6855 nor a local message in the state where . is a terminator.
6857 (ii) If a bare CR is encountered in a header line, an extra space is added
6858 after the line terminator so as not to end the header. The reasoning
6859 behind this is that bare CRs in header lines are most likely either
6860 to be mistakes, or people trying to play silly games.
6862 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6863 was being incorrectly given as 18 bytes larger than it should have been.
6864 This is a VOB (very old bug).
6866 60. This may never have affected anything current, but just in case it has:
6867 When the local host is found other than at the start of a list of hosts,
6868 the local host, those with the same MX, and any that follow, are discarded.
6869 When the list in question was part of a longer list of hosts, the following
6870 hosts (not currently being processed) were also being discarded. This no
6871 longer happens. I'm not sure if this situation could ever has previously
6874 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6877 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6878 header. This is something that is recommended in a new Internet Draft, and
6879 is something that is documented as being done by Sendmail. There are two
6880 possible values. For messages generated by the autoreply transport, Exim
6883 Auto-Submitted: auto-replied
6885 whereas for all other generated messages (e.g. bounces) it adds
6887 Auto-Submitted: auto-generated
6889 63. The "personal" condition in filters now includes a test for the
6890 Auto-Submitted: header. If it contains the string "auto-" the message it
6891 not considered personal.
6893 64. Added rcpt_include_affixes as a generic transport option.
6895 65. Added queue_only_override (default true).
6897 66. Added the syslog_duplication option.
6899 67. If what should have been the first header line of a message consisted of
6900 a space followed by a colon, Exim was mis-interpreting it as a header line.
6901 It isn't of course - it is syntactically invalid and should therefore be
6902 treated as the start of the message body. The misbehaviour could have
6903 caused a number of strange effects, including loss of data in subsequent
6904 header lines, and spool format errors.
6906 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6907 client host had authenticated. This control can now be exercised by an ACL
6908 for more flexibility.
6910 69. By default, callouts do not happen when testing with -bh. There is now a
6911 variant, -bhc, which does actually run the callout code, including
6912 consulting and updating the callout cache.
6914 70. Added support for saslauthd authentication, courtesy of Alexander
6917 71. If statvfs() failed on the spool or log directories while checking their
6918 size for availability, Exim confusingly gave the error "space shortage".
6919 Furthermore, in debugging mode it crashed with a floating point exception.
6920 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6921 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6922 serious problem, Exim now writes to the main and panic logs when this
6923 happens, with details of the failure. It then refuses to accept the
6924 incoming message, giving the message "spool directory problem" or "log
6925 directory problem" with a 421 code for SMTP messages.
6927 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6928 0, 1, and 2 exist, because some OS complain for execs without them (see
6929 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6930 descriptors. However, the code omitted to check that the open succeeded,
6931 causing mysterious errors if for some reason the permissions on /dev/null
6932 got screwed. Now Exim writes a message to the main and panic logs, and
6933 bombs out if it can't open /dev/null.
6935 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6936 interact so that it is all more flexible. It is supposed to remain
6937 backwards compatible. Also added extra_local_interfaces.
6939 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6940 to bomb out with an assertion failure - to the client this appears as a
6941 connection drop. This problem occurs in the part of the code that was taken
6942 from the Samba project. Fortunately, the assertion is in a very simple
6943 function, so I have fixed this by reproducing the function inline in the
6944 one place where it is called, and arranging for authentication to fail
6945 instead of killing the process with assert().
6947 75. The SPA client code was not working when the server requested OEM rather
6948 than Unicode encoding.
6950 76. Added code to make require_files with a specific uid setting more usable in
6951 the case where statting the file as root fails - usually a non-root-mounted
6952 NFS file system. When this happens and the failure is EACCES, Exim now
6953 forks a subprocess and does the per-uid checking as the relevant uid.
6955 77. Added process_log_path.
6957 78. If log_file_path was not explicitly set, a setting of check_log_space or
6958 check_log_inodes was ignored.
6960 79. If a space check for the spool or log partitions fails, the incident is now
6961 logged. Of course, in the latter case the data may get lost...
6963 80. Added the %p formatting code to string_format() so that it can be used to
6964 print addresses in debug_print(). Adjusted all the address printing in the
6965 debugging in store.c to use %p rather than %d.
6967 81. There was a concern that a line of code in smtp_in.c could overflow a
6968 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6969 initially expressed, the concern was not well-founded, because trailing
6970 spaces are removed early. However, if the trailing spaces were followed by
6971 a NULL, they did not get removed, so the overflow was possible. Two fixes
6974 (a) I re-wrote the offending code in a cleaner fashion.
6975 (b) If an incoming SMTP command contains a NULL character, it is rejected
6978 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6979 runs initgroups(), so that if the Exim user is in any additional groups,
6980 they will be used during message reception.
6986 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob