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.
8 Since Exim version 4.90
11 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
12 Previously only the last row was returned.
14 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
15 we assumed that tags in the header were well-formed, and parsed the
16 element content after inspecting only the first char of the tag.
17 Assumptions at that stage could crash the receive process on malformed
24 JH/01 Rework error string handling in TLS interface so that the caller in
25 more cases is responsible for logging. This permits library-sourced
26 string to be attached to addresses during delivery, and collapses
27 pairs of long lines into single ones.
29 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
30 during configuration. Wildcards are allowed and expanded.
32 JH/02 Rework error string handling in DKIM to pass more info back to callers.
33 This permits better logging.
35 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
36 do not close it down and have the child transport start it up again on
37 the passed-on TCP connection. Instead, proxy the child (and any
38 subsequent ones) for TLS via a unix-domain socket channel. Logging is
39 affected: the continued delivery log lines do not have any DNSSEC, TLS
40 Certificate or OCSP information. TLS cipher information is still logged.
42 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
43 identical IP addresses on different listening ports. Will also affect
46 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
47 add noisy ifdef guards to special-case this sillyness.
48 Patch from Bernd Kuhls.
50 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
51 than 255 are no longer allowed.
53 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
54 Disable the session-cache too, which might reduce our load. Since we
55 currrectly use a new context for every connection, both as server and
56 client, there is no benefit for these.
57 GnuTLS appears to not support tickets server-side by default (we don't
58 call gnutls_session_ticket_enable_server()) but client side is enabled
59 by default on recent versions (3.1.3 +) unless the PFS priority string
62 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
63 <https://reproducible-builds.org/specs/source-date-epoch/>.
65 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
66 the check for any unsuccessful recipients did not notice the limit, and
67 erroneously found still-pending ones.
69 JH/08 Pipeline CHUNKING command and data together, on kernels that support
70 MSG_MORE. Only in-clear (not on TLS connections).
72 JH/09 Avoid using a temporary file during transport using dkim. Unless a
73 transport-filter is involved we can buffer the headers in memory for
74 creating the signature, and read the spool data file once for the
75 signature and again for transmission.
77 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
78 4.77 as the kernel support then wasn't solid, having issues in 64bit
79 mode. Now, it's been long enough. Add support for FreeBSD also.
81 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
82 case where the routing stage had gathered several addresses to send to
83 a host before calling the transport for the first, we previously failed
84 to close down TLS in the old transport process before passing the TCP
85 connection to the new process. The new one sent a STARTTLS command
86 which naturally failed, giving a failed delivery and bloating the retry
87 database. Investigation and fix prototype from Wolfgang Breyha.
89 JH/12 Fix check on SMTP command input synchronisation. Previously there were
90 false-negatives in the check that the sender had not preempted a response
91 or prompt from Exim (running as a server), due to that code's lack of
92 awareness of the SMTP input buffering.
94 PP/04 Add commandline_checks_require_admin option.
95 Exim drops privileges sanely, various checks such as -be aren't a
96 security problem, as long as you trust local users with access to their
97 own account. When invoked by services which pass untrusted data to
98 Exim, this might be an issue. Set this option in main configuration
99 AND make fixes to the calling application, such as using `--` to stop
102 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
103 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
104 responses to those, into a single TLS record each way (this usually means
105 a single packet). As a side issue, smtp_enforce_sync now works on TLS
108 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
109 affects you only if you're dancing at the edge of the param size limits.
110 If you are, and this message makes sense to you, then: raise the
111 configured limit or use OpenSSL 1.1. Nothing we can do for older
114 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
115 cmdline element to get the documented default one. Previously it was
118 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
119 are unsuable for non-message-specific reasons - eg. connection timeout,
120 banner-time rejection.
122 JH/16 Fix logging of delivery remote port, when specified by router, under
125 PP/06 Repair manualroute's ability to take options in any order, even if one
126 is the name of a transport.
129 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
131 JH/17 Change the list-building routines interface to use the expanding-string
132 triplet model, for better allocation and copying behaviour.
134 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
135 Previously it was constructed the first time a possibly-matching string
136 was met in the configuration file input during startup; now it is done
139 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
140 compatible one, to avoid the (poorly documented) possibility of a config
141 file in the working directory redirecting the DB files, possibly correpting
142 some existing file. CVE-2017-10140 assigned for BDB.
144 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
145 cache-hot. Previously, although the result was properly cached, the
146 initial verify call returned a defer.
148 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
149 the main verify for receipient in uncached-mode.
151 JH/22 Retire historical build files to an "unsupported" subdir. These are
152 defined as "ones for which we have no current evidence of testing".
154 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
155 if present. Previously it was ignored.
157 JH/24 Start using specified-initialisers in C structure init coding. This is
158 a C99 feature (it's 2017, so now considered safe).
160 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
161 if was a fixed-sized field and bitmask ops via macros; it is now more
164 PP/07 GitHub PR 56: Apply MariaDB build fix.
165 Patch provided by Jaroslav Škarvada.
167 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
168 during Coverity cleanups [4.87 JH/47]
169 Diagnosis and fix provided by Michael Fischer v. Mollard.
171 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
172 the right size to place the terminating semicolon on its own folded
173 line, the header hash was calculated to an incorrect value thanks to
174 the (relaxed) space the fold became.
176 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
177 and confused the parent.
179 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
180 which could crash as a result. This could lead to undeliverable messages.
182 JH/28 Logging: "next input sent too soon" now shows where input was truncated
185 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
186 matters on fast-turnover and PID-randomising systems, which were getting
187 out-of-order delivery.
189 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
190 a possibly-overlapping copy. The symptom was that "Remote host closed
191 connection in response to HELO" was logged instead of the actual 4xx
194 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
195 Previously only that bufferd was discarded, resulting in SYMTP command
198 JH/32 DKIM: when a message has multiple signatures matching an identity given
199 in dkim_verify_signers, run the dkim acl once for each. Previously only
200 one run was done. Bug 2189.
202 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
203 "panic the current process" to "deliberately defer". The panic log is
204 still written with the problem list name; the mail and reject logs now
205 get a temp-reject line for the message that was being handled, saying
206 something like "domains check lookup or other defer". The SMTP 451
207 message is still "Temporary local problem".
209 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
210 A crafted sequence of BDAT commands could result in in-use memory beeing
211 freed. CVE-2017-16943.
213 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
214 from SMTP input. Previously it was always done; now only done for DATA
215 and not BDAT commands. CVE-2017-16944.
217 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
218 to the message (such as an overlong header line). Previously this was
219 not done and we did not exit BDAT mode. Followon from the previous item
220 though a different problem.
226 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
227 than -2003 did; needs libidn2 in addition to libidn.
229 JH/02 The path option on a pipe transport is now expanded before use.
231 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
232 Patch provided by "Björn", documentation fix added too.
234 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
235 missing a wire-to-host endian conversion.
237 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
238 close after a BDAT command line could be taken as a following command,
239 giving a synch failure. Fix by only checking for synch immediately
240 before acknowledging the chunk.
242 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
243 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
244 macro. Patches provided by Josh Soref.
246 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
247 Previously we did not; the RFC seems ambiguous and VRFY is not listed
248 by IANA as a service extension. However, John Klensin suggests that we
251 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
252 the dkim code may be unix-mode line endings rather than smtp wire-format
253 CRLF, so prepend a CR to any bare LF.
255 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
256 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
258 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
259 the first were themselves being wrongly included in the feed into dkim
260 processing; with most chunk sizes in use this resulted in an incorrect
261 body hash calculated value.
263 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
264 DKIM signature block, for verification. Although advised against by
265 standards it is specifically not ruled illegal.
267 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
269 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
270 missing a body hash (the bh= tag).
272 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
273 It seems that HAProxy sends the Proxy Protocol information in clear and
274 only then does a TLS startup, so do the same.
276 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
277 TCP connections (such as for Spamd) unless the daemon successfully set
278 Fast Open mode on its listening sockets. This fixes breakage seen on
279 too-old kernels or those not configured for Fast Open, at the cost of
280 requiring both directions being enabled for TFO, and TFO never being used
281 by non-daemon-related Exim processes.
283 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
284 endings, at least on the first header line. Try to canonify any that get
285 past that check, despite the cost.
287 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
288 now limited to an arbitrary five deep, while parsing addresses with the
289 strip_excess_angle_brackets option enabled.
291 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
292 instead leave the unprompted TLS handshake in socket buffer for the
293 TLS library to consume.
295 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
297 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
299 JH/16 Drop variables when they go out of scope. Memory management drops a whole
300 region in one operation, for speed, and this leaves assigned pointers
301 dangling. Add checks run only under the testsuite which checks all
302 variables at a store-reset and panics on a dangling pointer; add code
303 explicitly nulling out all the variables discovered. Fixes one known
304 bug: a transport crash, where a dangling pointer for $sending_ip_address
305 originally assigned in a verify callout, is re-used.
307 PP/06 Drop '.' from @INC in various Perl scripts.
309 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
311 PP/08 Reduce a number of compilation warnings under clang; building with
312 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
313 should be warning-free.
315 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
317 HS/01 Fix portability problems introduced by PP/08 for platforms where
318 realloc(NULL) is not equivalent to malloc() [SunOS et al].
320 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
321 chunk. This allows us to accept broken chunked messages. We need a more
322 general solution here.
324 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
325 already-broken messages in the queue.
327 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
329 JH/19 Fix reference counting bug in routing-generated-address tracking.
335 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
336 supports it and a size is available (ie. the sending peer gave us one).
338 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
339 years of being deprecated). The replacements are the ACLs
340 acl_smtp_mime and acl_not_smtp_mime.
342 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
343 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
344 or even an in-clear connection were permitted. Now, if the host lookup
345 was dnssec and dane was requested then the host is only used if the
346 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
347 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
348 if one fails this test.
349 This means that a poorly-configured remote DNS will make it incommunicado;
350 but it protects against a DNS-interception attack on it.
352 JH/04 Bug 1810: make continued-use of an open smtp transport connection
353 non-noisy when a race steals the message being considered.
355 JH/05 If main configuration option tls_certificate is unset, generate a
356 self-signed certificate for inbound TLS connections.
358 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
359 in rewrites and routers.
361 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
362 and logged a warning sing 4.83; now they are a configuration file error.
364 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
365 (lacking @domain). Apply the same qualification processing as RCPT.
367 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
369 JH/10 Support ${sha256:} applied to a string (as well as the previous
372 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
373 a cutthrough deliver is pending, as we always want to make a connection.
374 This also avoids re-routing the message when later placing the cutthrough
375 connection after a verify cache hit.
376 Do not update it with the verify result either.
378 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
379 when routing results in more than one destination address.
381 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
382 signing (which inhibits the cutthrough capability). Previously only
383 the presence of an option was tested; now an expansion evaluating as
384 empty is permissible (obviously it should depend only on data available
385 when the cutthrough connection is made).
387 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
388 the relevant preceding SMTP command did not note the pipelining mode.
390 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
391 Previously they were not counted.
393 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
394 as one having no matching records. Previously we deferred the message
395 that needed the lookup.
397 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
398 distinguished as "(=".
400 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
401 for missing MX records. Previously it only worked for missing A records.
403 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
405 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
406 after the data-go-ahead and data-ack. Patch from Jason Betts.
408 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
409 even for a "none" policy. Patch from Tony Meyer.
411 JH/22 Fix continued use of a connection for further deliveries. If a port was
412 specified by a router, it must also match for the delivery to be
415 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
416 When one of the recipients of a message was unsuitable for the connection
417 (has no matching addresses), we lost track of needing to mark it
418 deferred. As a result mail would be lost.
420 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
422 JH/25 Decoding ACL controls is now done using a binary search; the source code
423 takes up less space and should be simpler to maintain. Merge the ACL
424 condition decode tables also, with similar effect.
426 JH/26 Fix problem with one_time used on a redirect router which returned the
427 parent address unchanged. A retry would see the parent address marked as
428 delivered, so not attempt the (identical) child. As a result mail would
431 JH/27 Fix a possible security hole, wherein a process operating with the Exim
432 UID can gain a root shell. Credit to http://www.halfdog.net/ for
433 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
436 JH/28 Enable {spool,log} filesystem space and inode checks as default.
437 Main config options check_{log,spool}_{inodes,space} are now
438 100 inodes, 10MB unless set otherwise in the configuration.
440 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
441 Previously it only applied to the main-section connection policy
444 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
446 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
447 by me. Added RFC7919 DH primes as an alternative.
449 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
450 are not in the system include path.
452 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
453 GnuTLS, when a session startup failed (eg because the client disconnected)
454 Exim did stdio operations after fclose. This was exposed by a recent
455 change which nulled out the file handle after the fclose.
457 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
458 signed directly by the cert-signing cert, rather than an intermediate
459 OCSP-signing cert. This is the model used by LetsEncrypt.
461 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
463 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
464 an incoming connection.
466 HS/02 Bug 1802: Do not half-close the connection after sending a request
469 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
470 fallback to "prime256v1".
472 JH/34 SECURITY: Use proper copy of DATA command in error message.
473 Could leak key material. Remotely exploitable. CVE-2016-9963.
479 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
480 and 3.4.4 - once the server is enabled to respond to an OCSP request
481 it does even when not requested, resulting in a stapling non-aware
482 client dropping the TLS connection.
484 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
485 support variable-length bit vectors. No functional change.
487 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
488 The I= interface field on outgoing lines is now after the H= remote
489 host field, same as incoming lines. There is a separate
490 outgoing_interface log selector which allows you to disable the
493 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
494 If not running log_selector +smtp_connection the mainlog would be held
495 open indefinitely after a "too many connections" event, including to a
496 deleted file after a log rotate. Leave the per net connection logging
497 leaving it open for efficiency as that will be quickly detected by the
498 check on the next write.
500 HS/01 Bug 1671: Fix post transport crash.
501 Processing the wait-<transport> messages could crash the delivery
502 process if the message IDs didn't exist for some reason. When
503 using 'split_spool_directory=yes' the construction of the spool
504 file name failed already, exposing the same netto behaviour.
506 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
507 mime_regex ACL conditions.
509 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
510 to DSN fail messages (bounces): remote IP, remote greeting, remote response
511 to HELO, local diagnostic string.
513 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
514 log line to debug. Even when configured with a tls authenticator many
515 client connections are expected to not authenticate in this way, so
516 an authenticate fail is not an error.
518 HS/02 Add the Exim version string to the process info. This way exiwhat
519 gives some more detail about the running daemon.
521 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
522 matter for fast-change records such as DNSBLs.
524 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
525 retry record, even if constant. There may be multiple transports with
526 different interface settings and the retry behaviour needs to be kept
529 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
531 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
533 JH/10 Bug 840: fix log_defer_output option of pipe transport
535 JH/11 Bug 830: use same host for all RCPTS of a message, even under
536 hosts_randomize. This matters a lot when combined with mua_wrapper.
538 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
539 ${quote_pgsql:<string>} operator.
541 JH/13 Bug 1708: avoid misaligned access in cached lookup.
543 JH/14 Change header file name for freeradius-client. Relevant if compiling
544 with Radius support; from the Gentoo tree and checked under Fedora.
546 JH/15 Bug 1712: Introduce $prdr_requested flag variable
548 JH/16 Bug 1714: Permit an empty string as expansion result for transport
549 option transport_filter, meaning no filtering.
551 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
553 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
554 defaults to "*" (all hosts). The variable is now available when not built
555 with TLS, default unset, mainly to enable keeping the testsuite sane.
556 If a server certificate is not supplied (via tls_certificate) an error is
557 logged, and clients will find TLS connections fail on startup. Presumably
558 they will retry in-clear.
559 Packagers of Exim are strongly encouraged to create a server certificate
560 at installation time.
562 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
563 with the $config_file variable.
565 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
566 in transport context, after the attempt, and per-recipient. The latter type
567 is per host attempted. The event data is the error message, and the errno
568 information encodes the lookup type (A vs. MX) used for the (first) host,
569 and the trailing two digits of the smtp 4xx response.
571 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
572 to write to mainlog (or rejectlog, paniclog) in the window between file
573 creation and permissions/ownership being changed. Particularly affects
574 installations where exicyclog is run as root, rather than exim user;
575 result is that the running daemon panics and dies.
577 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
579 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
580 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
581 "pri" and "weight". Note that the previous implicit priority given by the
582 list order is no longer honoured.
584 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
587 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
588 by defining SUPPORT_SOCKS.
590 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
591 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
592 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
593 variables are renamed to proxy_{local,external}_{address,port}.
595 JH/27 Move Internationalisation support from Experimental to mainline, enabled
596 for a build by defining SUPPORT_I18N
598 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
599 of the query string, and make ${quote_redis:} do that quoting.
601 JH/29 Move Events support from Experimental to mainline, enabled by default
602 and removable for a build by defining DISABLE_EVENT.
604 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
606 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
607 cached by the daemon.
609 JH/32 Move Redis support from Experimental to mainline, enabled for a build
610 by defining LOOKUP_REDIS. The libhiredis library is required.
612 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
613 keys are given for lookup.
615 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
616 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
617 only supported when built with TLS support. The PolarSSL SHA routines
618 are still used when the TLS library is too old for convenient support.
620 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
621 openssl_options), for security. OpenSSL forces this from version 1.1.0
622 server-side so match that on older versions.
624 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
625 allocation for $value could be released as the expansion processing
626 concluded, but leaving the global pointer active for it.
628 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
629 and to use the domains and local_parts ACL conditions.
631 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
632 incorrectly not doubled on cutthrough transmission, hence seen as a
633 body-termination at the receiving system - resulting in truncated mails.
634 Commonly the sender saw a TCP-level error, and retransmitted the message
635 via the normal store-and-forward channel. This could result in duplicates
636 received - but deduplicating mailstores were liable to retain only the
637 initial truncated version.
639 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
641 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
643 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
644 we're in there, support oversigning also; bug 1309.
646 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
648 HS/04 Add support for keep_environment and add_environment options.
650 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
651 either intentional arithmetic overflow during PRNG, or testing config-
654 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
655 delivery resulted in actual delivery. Cancel cutthrough before DATA
658 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
659 rejects a recipient: pass the reject to the originator.
661 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
662 Many were false-positives and ignorable, but it's worth fixing the
665 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
666 for the new environment-manipulation done at startup. Move the routines
667 from being local to tls.c to being global via the os.c file.
669 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
670 an extract embedded as result-arg for a map, the first arg for extract
671 is unavailable so we cannot tell if this is a numbered or keyed
672 extraction. Accept either.
678 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
681 JH/02 The smtp transport option "multi_domain" is now expanded.
683 JH/03 The smtp transport now requests PRDR by default, if the server offers
686 JH/04 Certificate name checking on server certificates, when exim is a client,
687 is now done by default. The transport option tls_verify_cert_hostnames
688 can be used to disable this per-host. The build option
689 EXPERIMENTAL_CERTNAMES is withdrawn.
691 JH/05 The value of the tls_verify_certificates smtp transport and main options
692 default to the word "system" to access the system default CA bundle.
693 For GnuTLS, only version 3.0.20 or later.
695 JH/06 Verification of the server certificate for a TLS connection is now tried
696 (but not required) by default. The verification status is now logged by
697 default, for both outbound TLS and client-certificate supplying inbound
700 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
703 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
704 Status Notification (bounce) messages are now MIME format per RFC 3464.
705 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
706 under the control of the dsn_advertise_hosts option, and routers may
707 have a dsn_lasthop option.
709 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
710 default, modifiable by a malware= option. The list separator for
711 the options can now be changed in the usual way. Bug 68.
713 JH/10 The smtp_receive_timeout main option is now expanded before use.
715 JH/11 The incoming_interface log option now also enables logging of the
716 local interface on delivery outgoing connections.
718 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
719 if the interface and destination host and port all match.
721 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
724 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
725 Patch from Andrew Lewis.
727 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
728 now supports optional time-restrictions, weighting, and priority
729 modifiers per server. Patch originally by <rommer@active.by>.
731 JH/16 The spamd_address main option now supports a mixed list of local
732 and remote servers. Remote servers can be IPv6 addresses, and
733 specify a port-range.
735 JH/17 Bug 68: The spamd_address main option now supports an optional
736 timeout value per server.
738 JH/18 Bug 1581: Router and transport options headers_add/remove can
739 now have the list separator specified.
741 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
744 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
747 JH/21 Support for the A6 type of dns record is withdrawn.
749 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
750 rather than the verbs used.
752 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
753 from 255 to 1024 chars.
755 JH/24 Verification callouts now attempt to use TLS by default.
757 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
758 are generic router options now. The defaults didn't change.
760 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
761 Original patch from Alexander Shikoff, worked over by JH.
763 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
764 files if ZCAT_COMMAND is not executable.
766 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
768 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
770 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
771 Normally benign, it bites when the pair was led to by a CNAME;
772 modern usage is to not canonicalize the domain to a CNAME target
773 (and we were inconsistent anyway for A-only vs AAAA+A).
775 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
777 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
778 when evaluating $sender_host_dnssec.
780 JH/31 Check the HELO verification lookup for DNSSEC, adding new
781 $sender_helo_dnssec variable.
783 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
785 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
787 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
789 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
790 documented as working, but never had. Support all but $spam_report.
792 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
793 added for tls authenticator.
795 HS/03 Add perl_taintmode main config option
801 TL/01 When running the test suite, the README says that variables such as
802 no_msglog_check are global and can be placed anywhere in a specific
803 test's script, however it was observed that placement needed to be near
804 the beginning for it to behave that way. Changed the runtest perl
805 script to read through the entire script once to detect and set these
806 variables, reset to the beginning of the script, and then run through
807 the script parsing/test process like normal.
809 TL/02 The BSD's have an arc4random API. One of the functions to induce
810 adding randomness was arc4random_stir(), but it has been removed in
811 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
812 function when detected.
814 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
815 cause callback expansion.
817 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
818 syntax errors in an expansion can be treated as a string instead of
819 logging or causing an error, due to the internal use of bool_lax
820 instead of bool when processing it.
822 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
823 server certificates when making smtp deliveries.
825 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
827 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
829 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
831 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
832 Merged patch from Sebastian Wiedenroth.
834 JH/05 Fix results-pipe from transport process. Several recipients, combined
835 with certificate use, exposed issues where response data items split
836 over buffer boundaries were not parsed properly. This eventually
837 resulted in duplicates being sent. This issue only became common enough
838 to notice due to the introduction of connection certificate information,
839 the item size being so much larger. Found and fixed by Wolfgang Breyha.
841 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
842 size buffer was used, resulting in syntax errors when an expansion
845 JH/07 Add support for directories of certificates when compiled with a GnuTLS
846 version 3.3.6 or later.
848 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
849 is EXPERIMENTAL_EVENT, the main-configuration and transport options
850 both become "event_action", the variables become $event_name, $event_data
851 and $event_defer_errno. There is a new variable $verify_mode, usable in
852 routers, transports and related events. The tls:cert event is now also
853 raised for inbound connections, if the main configuration event_action
856 TL/06 In test suite, disable OCSP for old versions of openssl which contained
857 early OCSP support, but no stapling (appears to be less than 1.0.0).
859 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
860 server certificate names available under the smtp transport option
861 "tls_verify_cert_hostname" now do not permit multi-component wildcard
864 JH/10 Time-related extraction expansions from certificates now use the main
865 option "timezone" setting for output formatting, and are consistent
866 between OpenSSL and GnuTLS compilations. Bug 1541.
868 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
869 encoded parameter in the incoming message. Bug 1558.
871 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
872 include certificate info, eximon was claiming there were spoolfile
875 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
877 JH/14 Log delivery-related information more consistently, using the sequence
878 "H=<name> [<ip>]" wherever possible.
880 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
881 are problematic for Debian distribution, omit them from the release
884 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
886 JH/16 Fix string representation of time values on 64bit time_t architectures.
889 JH/17 Fix a null-indirection in certextract expansions when a nondefault
890 output list separator was used.
895 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
896 checkers that were complaining about end of non-void function with no
899 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
900 This was a regression introduced in 4.83 by another bugfix.
902 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
904 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
905 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
911 TF/01 Correctly close the server side of TLS when forking for delivery.
913 When a message was received over SMTP with TLS, Exim failed to clear up
914 the incoming connection properly after forking off the child process to
915 deliver the message. In some situations the subsequent outgoing
916 delivery connection happened to have the same fd number as the incoming
917 connection previously had. Exim would try to use TLS and fail, logging
918 a "Bad file descriptor" error.
920 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
921 utilities have not been installed.
923 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
924 temporary space as the ACL may create new global variables.
926 TL/01 LDAP support uses per connection or global context settings, depending
927 upon the detected version of the libraries at build time.
929 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
930 to extract and use the src ip:port in logging and expansions as if it
931 were a direct connection from the outside internet. PPv2 support was
932 updated based on HAProxy spec change in May 2014.
934 JH/02 Add ${listextract {number}{list}{success}{fail}}.
936 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
937 Properly escape header and check for NULL return.
939 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
942 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
944 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
945 characters in header names, implemented as a verify condition.
946 Contributed by Michael Fischer v. Mollard.
948 TL/05 Rename SPF condition results err_perm and err_temp to standardized
949 results permerror and temperror. Previous values are deprecated but
950 still accepted. In a future release, err_perm and err_temp will be
951 completely removed, which will be a backward incompatibility if the
952 ACL tests for either of these two old results. Patch contributed by
953 user bes-internal on the mailing list.
955 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
957 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
958 selectors, in both main and reject logs.
960 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
963 JH/07 Add malware type "sock" for talking to simple daemon.
965 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
967 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
968 routers/transports under cutthrough routing.
970 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
971 numbers. Touch up "bool" conditional to keep the same definition.
973 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
975 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
977 JH/12 Expand items in router/transport headers_add or headers_remove lists
978 individually rather than the list as a whole. Bug 1452.
980 Required for reasonable handling of multiple headers_ options when
981 they may be empty; requires that headers_remove items with embedded
982 colons must have them doubled (or the list-separator changed).
984 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
985 view the policy declared in the DMARC record. Currently, $dmarc_status
986 is a combined value of both the record presence and the result of the
989 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
991 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
992 dnslookup router and the smtp transport (applying to the forward
995 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
996 of ldap servers used for a specific lookup. Patch provided by Heiko
999 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1000 New variable $lookup_dnssec_authenticated for observability.
1002 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1003 Patch submitted by Lars Timman.
1005 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1007 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1008 Requires trusted mode and valid format message id, aborts otherwise.
1009 Patch contributed by Heiko Schlichting.
1011 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1012 certextract with support for various fields. Bug 1358.
1014 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1015 is requested by default, modifiable by smtp transport option
1018 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1019 operate on certificate variables to give certificate fingerprints
1020 Also new ${sha256:cert_variable}.
1022 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1024 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1025 Christian Aistleitner.
1027 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1029 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1030 file. Patch from Wolfgang Breyha.
1032 JH/25 Expand the coverage of the delivery $host and $host_address to
1033 client authenticators run in verify callout. Bug 1476.
1035 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1036 align with daemon_smtp_ports. Bug 72.
1038 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1039 support and error reporting did not work properly.
1041 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1042 and is readable. Patch from Andrew Colin Kissa.
1044 TL/14 Enhance documentation of ${run expansion and how it parses the
1045 commandline after expansion, particularly in the case when an
1046 unquoted variable expansion results in an empty value.
1048 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1050 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1051 by renaming away from it. Addresses GH issue 15, reported by
1054 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1056 TL/15 SECURITY: prevent double expansion in math comparison functions
1057 (can expand unsanitized data). Not remotely exploitable.
1064 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1066 PP/02 Make -n do something, by making it not do something.
1067 When combined with -bP, the name of an option is not output.
1069 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1072 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1073 $sender_host_name and config options to manage this, and basic check
1076 PP/05 DSCP support for outbound connections and control modifier for inbound.
1078 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1079 (Only plugin which currently uses this is kerberos4, which nobody should
1080 be using, but we should make it available and other future plugins might
1081 conceivably use it, even though it would break NAT; stuff *should* be
1082 using channel bindings instead).
1084 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1085 name; added for Sendmail compatibility; requires admin caller.
1086 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1087 just ignore it); requires trusted caller.
1088 Also parse but ignore: -Ac -Am -X<logfile>
1091 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1093 TL/02 Add +smtp_confirmation as a default logging option.
1095 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1096 Patch by Magnus Holmgren from 2007-02-20.
1098 TL/04 Bugzilla 1281 - Spec typo.
1099 Bugzilla 1283 - Spec typo.
1100 Bugzilla 1290 - Spec grammar fixes.
1102 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1104 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1106 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1107 mailing list by by Dmitry Isaikin.
1109 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1111 JH/02 Support "G" suffix to numbers in ${if comparisons.
1113 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1115 NM/01 Bugzilla 1197 - Spec typo
1116 Bugzilla 1196 - Spec examples corrections
1118 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1120 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1121 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1124 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1125 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1127 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1128 "acl {{name}{arg}...}", and optional args on acl condition
1131 JH/05 Permit multiple router/transport headers_add/remove lines.
1133 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1135 JH/07 Avoid using a waiting database for a single-message-only transport.
1136 Performance patch from Paul Fisher. Bugzilla 1262.
1138 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1141 JH/09 Add $headers_added variable, with content from use of ACL modifier
1142 add_header (but not yet added to the message). Bugzilla 199.
1144 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1145 Pulled from Bugzilla 817 by Wolfgang Breyha.
1147 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1149 (nb: this is the same fix as in Exim 4.80.1)
1151 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1154 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1157 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1159 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1160 advertise SMTP AUTH mechanism to us, instead of a generic
1161 protocol violation error. Also, make Exim more robust to bad
1162 data from the Dovecot auth socket.
1164 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1166 When a queue runner is handling a message, Exim first routes the
1167 recipient addresses, during which it prunes them based on the retry
1168 hints database. After that it attempts to deliver the message to
1169 any remaining recipients. It then updates the hints database using
1172 So if a recipient address works intermittently, it can get repeatedly
1173 deferred at routing time. The retry hints record remains fresh so the
1174 address never reaches the final cutoff time.
1176 This is a fairly common occurrence when a user is bumping up against
1177 their storage quota. Exim had some logic in its local delivery code
1178 to deal with this. However it did not apply to per-recipient defers
1179 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1181 This change adds a proper retry rule check during routing so that the
1182 final cutoff time is checked against the message's age. We only do
1183 this check if there is an address retry record and there is not a
1184 domain retry record; this implies that previous attempts to handle
1185 the address had the retry_use_local_parts option turned on. We use
1186 this as an approximation for the destination being like a local
1187 delivery, as in LMTP.
1189 I suspect this new check makes the old local delivery cutoff check
1190 redundant, but I have not verified this so I left the code in place.
1192 TF/02 Correct gecos expansion when From: is a prefix of the username.
1194 Test 0254 submits a message to Exim with the header
1198 When I ran the test suite under the user fanf2, Exim expanded
1199 the header to contain my full name, whereas it should have added
1200 a Resent-Sender: header. It erroneously treats any prefix of the
1201 username as equal to the username.
1203 This change corrects that bug.
1205 GF/01 DCC debug and logging tidyup
1206 Error conditions log to paniclog rather than rejectlog.
1207 Debug lines prefixed by "DCC: " to remove any ambiguity.
1209 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1211 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1212 Bug spotted by Jeremy Harris; was flawed since initial commit.
1213 Would have resulted in OCSP responses post-SNI triggering an Exim
1214 NULL dereference and crash.
1216 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1218 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1219 Bug detection, analysis and fix by Samuel Thibault.
1220 Bugzilla 1331, Debian bug #698092.
1222 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1224 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1225 Server implementation by Todd Lyons, client by JH.
1226 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1227 config variable "prdr_enable" controls whether the server
1228 advertises the facility. If the client requests PRDR a new
1229 acl_data_smtp_prdr ACL is called once for each recipient, after
1230 the body content is received and before the acl_smtp_data ACL.
1231 The client is controlled by both of: a hosts_try_prdr option
1232 on the smtp transport, and the server advertisement.
1233 Default client logging of deliveries and rejections involving
1234 PRDR are flagged with the string "PRDR".
1236 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1237 fclose(). Diagnosis by Todd Lyons.
1239 PP/17 Update configure.default to handle IPv6 localhost better.
1240 Patch by Alain Williams (plus minor tweaks).
1243 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1244 This is now consistent with GnuTLS, and is now documented: the
1245 previous undocumented portable approach to treating the option as
1246 unset was to force an expansion failure. That still works, and
1247 an empty string is now equivalent.
1249 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1250 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1251 not performing validation itself.
1253 PP/20 Added force_command boolean option to pipe transport.
1254 Patch from Nick Koston, of cPanel Inc.
1256 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1259 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1261 PP/21 Fix eximon continuous updating with timestamped log-files.
1262 Broken in a format-string cleanup in 4.80, missed when I repaired the
1263 other false fix of the same issue.
1264 Report and fix from Heiko Schlichting.
1267 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1268 Report from Prashanth Katuri.
1270 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1271 It's SecureTransport, so affects any MacOS clients which use the
1272 system-integrated TLS libraries, including email clients.
1274 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1275 using a MIME ACL for non-SMTP local injection.
1276 Report and assistance in diagnosis by Warren Baker.
1278 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1280 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1282 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1283 last id that failed so it may be referenced in subsequent ACL's.
1285 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1288 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1289 ldap library initialization, allowing self-signed CA's to be
1290 used. Also properly sets require_cert option later in code by
1291 using NULL (global ldap config) instead of ldap handle (per
1292 session). Bug diagnosis and testing by alxgomz.
1294 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1295 the src/util/ subdirectory.
1297 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1298 renamed to Transport Post Delivery Action by Jeremy Harris, as
1301 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1302 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1303 redis_servers = needs to be configured which will be used by the redis
1304 lookup. Patch from Warren Baker, of The Packet Hub.
1306 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1308 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1309 hostname or reverse DNS when processing a host list. Used suggestions
1310 from multiple comments on this bug.
1312 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1314 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1315 Missed a few lines, added it to make the runtest require no keyboard
1318 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1319 contains upper case chars. Make router use caseful_local_part.
1321 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1322 support when GnuTLS has been built with p11-kit.
1328 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1330 This, or similar/improved, will also be change PP/11 of 4.82.
1336 PP/01 Handle short writes when writing local log-files.
1337 In practice, only affects FreeBSD (8 onwards).
1338 Bugzilla 1053, with thanks to Dmitry Isaikin.
1340 NM/01 Bugzilla 949 - Documentation tweak
1342 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1345 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1347 PP/02 Implemented gsasl authenticator.
1349 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1351 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1352 `pkg-config foo` for cflags/libs.
1354 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1355 with rest of GSASL and with heimdal_gssapi.
1357 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1358 `pkg-config foo` for cflags/libs for the TLS implementation.
1360 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1361 properties get this fed in as external SSF. A number of robustness
1362 and debugging improvements to the cyrus_sasl authenticator.
1364 PP/08 cyrus_sasl server now expands the server_realm option.
1366 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1367 Patch by Jeremy Harris.
1369 PP/10 Added dbmjz lookup type.
1371 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1373 PP/12 MAIL args handles TAB as well as SP, for better interop with
1374 non-compliant senders.
1375 Analysis and variant patch by Todd Lyons.
1377 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1378 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1379 Patch from Dirk Mueller <dmueller@suse.com>
1381 PP/13 tls_peerdn now print-escaped for spool files.
1382 Observed some $tls_peerdn in wild which contained \n, which resulted
1383 in spool file corruption.
1385 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1386 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1387 or write after TLS renegotiation, which otherwise led to messages
1390 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1391 as a tracking header (ie: a signed header comes before the signature).
1392 Patch from Wolfgang Breyha.
1394 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1395 comma-sep list; embedded commas doubled.
1397 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1399 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1401 Report and patch from Dmitry Banschikov.
1403 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1404 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1405 failures. We appear to now support TLS1.1+ with Exim.
1407 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1408 lets Exim select keys and certificates based upon TLS SNI from client.
1409 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1410 before an outbound SMTP session. New log_selector, +tls_sni.
1412 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1413 NULL dereference. Report and patch from Alun Jones.
1415 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1416 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1417 Not seeing resolver debug output on NetBSD, but suspect this is a
1418 resolver implementation change.
1420 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1421 Left warnings. Added "eximon gdb" invocation mode.
1423 PP/21 Defaulting "accept_8bitmime" to true, not false.
1425 PP/22 Added -bw for inetd wait mode support.
1427 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1428 locate the relevant includes and libraries. Made this the default.
1430 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1431 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1433 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1434 This may cause build issues on older platforms.
1436 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1437 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1438 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1439 Added SNI support via GnuTLS too.
1440 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1442 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1444 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1445 Applied second patch from Janne, implementing suggestion to default
1446 multiple-strings-in-record handling to match SPF spec.
1448 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1450 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1451 read-only, out of scope).
1452 Patch from Wolfgang Breyha, report from Stuart Northfield.
1454 PP/29 Fix three issues highlighted by clang analyser static analysis.
1455 Only crash-plausible issue would require the Cambridge-specific
1456 iplookup router and a misconfiguration.
1457 Report from Marcin Mirosław.
1459 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1461 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1462 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1463 As part of this, removing so much warning spew let me fix some minor
1464 real issues in debug logging.
1466 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1467 assignment on my part. Fixed.
1469 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1470 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1471 Janne Snabb (who went above and beyond: thank you).
1473 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1474 string otherwise requires a connection and a bunch more work and it's
1475 relatively easy to get wrong. Should also expose TLS library linkage
1478 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1479 64-bit ${eval} (JH/03).
1481 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1482 GNU libc to support some of the 64-bit stuff, should not lead to
1483 conflicts. Defined before os.h is pulled in, so if a given platform
1484 needs to override this, it can.
1486 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1487 protection layer was required, which is not implemented.
1488 Bugzilla 1254, patch from Wolfgang Breyha.
1490 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1491 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1492 tls_dhparam take prime identifiers. Also unbreak combination of
1493 OpenSSL+DH_params+TLSSNI.
1495 PP/39 Disable SSLv2 by default in OpenSSL support.
1501 PP/01 Solaris build fix for Oracle's LDAP libraries.
1502 Bugzilla 1109, patch from Stephen Usher.
1504 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1506 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1509 TF/02 Fix a couple more cases where we did not log the error message
1510 when unlink() failed. See also change 4.74-TF/03.
1512 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1513 lock up or crash if it happened to be inside a call to libc when it
1514 got a SIGUSR1 from exiwhat.
1516 The SIGUSR1 handler appends the current process status to the process
1517 log which is later printed by exiwhat. It used to use the general
1518 purpose logging code to do this, but several functions it calls are
1519 not safe for signals.
1521 The new output code in the SIGUSR1 handler is specific to the process
1522 log, and simple enough that it's easy to inspect for signal safety.
1523 Removing some special cases also simplifies the general logging code.
1524 Removing the spurious timestamps from the process log simplifies
1527 TF/04 Improved ratelimit ACL condition.
1529 The /noupdate option has been deprecated in favour of /readonly which
1530 has clearer semantics. The /leaky, /strict, and /readonly update modes
1531 are mutually exclusive. The update mode is no longer included in the
1532 database key; it just determines when the database is updated. (This
1533 means that when you upgrade Exim will forget old rate measurements.)
1535 Exim now checks that the per_* options are used with an update mode that
1536 makes sense for the current ACL. For example, when Exim is processing a
1537 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1538 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1539 must specify per_mail/readonly. If you omit the update mode it defaults to
1540 /leaky where that makes sense (as before) or /readonly where required.
1542 The /noupdate option is now undocumented but still supported for
1543 backwards compatibility. It is equivalent to /readonly except that in
1544 ACLs where /readonly is required you may specify /leaky/noupdate or
1545 /strict/noupdate which are treated the same as /readonly.
1547 A useful new feature is the /count= option. This is a generalization
1548 of the per_byte option, so that you can measure the throughput of other
1549 aggregate values. For example, the per_byte option is now equivalent
1550 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1552 The per_rcpt option has been generalized using the /count= mechanism
1553 (though it's more complicated than the per_byte equivalence). When it is
1554 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1555 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1556 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1557 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1558 Note that using per_rcpt with a non-readonly update mode in more than
1559 one ACL will cause the recipients to be double-counted. (The per_mail
1560 and per_byte options don't have this problem.)
1562 The handling of very low rates has changed slightly. If the computed rate
1563 is less than the event's count (usually one) then this event is the first
1564 after a long gap. In this case the rate is set to the same as this event's
1565 count, so that the first message of a spam run is counted properly.
1567 The major new feature is a mechanism for counting the rate of unique
1568 events. The new per_addr option counts the number of different
1569 recipients that someone has sent messages to in the last time period. It
1570 behaves like per_rcpt if all the recipient addresses are different, but
1571 duplicate recipient addresses do not increase the measured rate. Like
1572 the /count= option this is a general mechanism, so the per_addr option
1573 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1574 example, measure the rate that a client uses different sender addresses
1575 with the options per_mail/unique=$sender_address. There are further
1576 details in the main documentation.
1578 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1580 TF/06 Removed a few PCRE remnants.
1582 TF/07 Automatically extract Exim's version number from tags in the git
1583 repository when doing development or release builds.
1585 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1586 Bugzilla 879. Patch from Paul Fisher.
1588 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1589 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1592 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1594 PP/05 Variable $av_failed, true if the AV scanner deferred.
1595 Bugzilla 1078. Patch from John Horne.
1597 PP/06 Stop make process more reliably on build failure.
1598 Bugzilla 1087. Patch from Heiko Schlittermann.
1600 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1601 Bugzilla 1089. Patch from Heiko Schlittermann.
1603 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1604 Bugzilla 1131. Patch from Holger Weiß.
1606 PP/09 Handle IPv6 addresses with SPF.
1607 Bugzilla 860. Patch from Wolfgang Breyha.
1609 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1611 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1614 PP/11 match_* no longer expand right-hand-side by default.
1615 New compile-time build option, EXPAND_LISTMATCH_RHS.
1616 New expansion conditions, "inlist", "inlisti".
1618 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1620 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1622 PP/14 fix log_write() format string regression from TF/03.
1623 Bugzilla 1152. Patch from Dmitry Isaikin.
1629 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1631 PP/02 Harmonised TLS library version reporting; only show if debugging.
1632 Layout now matches that introduced for other libraries in 4.74 PP/03.
1634 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1636 PP/04 New "dns_use_edns0" global option.
1638 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1641 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1642 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1644 TK/01 Updated PolarSSL code to 0.14.2.
1645 Bugzilla 1097. Patch from Andreas Metzler.
1647 PP/07 Catch divide-by-zero in ${eval:...}.
1648 Fixes bugzilla 1102.
1650 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1653 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1654 format-string attack -- SECURITY: remote arbitrary code execution.
1656 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1657 time unintentionally subject to list matching rules, letting the header
1658 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1659 arbitrary string expansion). This allowed for information disclosure.
1661 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1662 INT_MIN/-1 -- value coerced to INT_MAX.
1668 NM/01 Workaround for PCRE version dependency in version reporting
1671 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1672 This fixes portability to compilers other than gcc, notably
1673 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1675 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1676 makefiles for portability to HP-UX and POSIX correctness.
1678 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1679 Also via indented variable definition in the Makefile.
1680 (Debugging by Oliver Heesakkers).
1682 PP/02 Restore caching of spamd results with expanded spamd_address.
1683 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1685 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1686 Improves build reliability. Fix from: Frank Elsner
1688 NM/02 Fix wide character breakage in the rfc2047 coding
1689 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1691 NM/03 Allow underscore in dnslist lookups
1692 Fixes bug 1026. Patch from Graeme Fowler
1694 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1695 Code patches from Adam Ciarcinski of NetBSD.
1697 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1700 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1701 is logged, to avoid truncation. Patch from John Horne.
1703 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1704 Patch from Jakob Hirsch.
1706 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1707 SQL string expansion failure details.
1708 Patch from Andrey Oktyabrski.
1710 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1711 Patch from Simon Arlott.
1713 PP/09 New lookups functionality failed to compile on old gcc which rejects
1714 extern declarations in function scope.
1715 Patch from Oliver Fleischmann
1717 PP/10 Use sig_atomic_t for flags set from signal handlers.
1718 Check getgroups() return and improve debugging.
1719 Fixed developed for diagnosis in bug 927 (which turned out to be
1722 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1723 Patch from Mark Zealey.
1725 PP/12 Bugzilla 1056: Improved spamd server selection.
1726 Patch from Mark Zealey.
1728 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1729 Based on patch from Heiko Schlittermann.
1731 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1732 Patch from Uwe Doering, sign-off by Michael Haardt.
1734 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1735 variable declaration deep within a block. Bug and patch from
1738 PP/15 lookups-Makefile IRIX compatibility coercion.
1740 PP/16 Make DISABLE_DKIM build knob functional.
1742 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1743 Patch by Simon Arlott
1745 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1746 variable argument macros. Our copy now differs from upstream.
1752 TF/01 Failure to get a lock on a hints database can have serious
1753 consequences so log it to the panic log.
1755 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1756 controlled using the smtp_confirmation log selector.
1758 TF/03 Include the error message when we fail to unlink a spool file.
1760 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1761 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1762 for maintaining out-of-tree patches for some time.
1764 PP/01 Bugzilla 139: Documentation and portability issues.
1765 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1766 Handle per-OS dynamic-module compilation flags.
1768 PP/02 Let /dev/null have normal permissions.
1769 The 4.73 fixes were a little too stringent and complained about the
1770 permissions on /dev/null. Exempt it from some checks.
1771 Reported by Andreas M. Kirchwitz.
1773 PP/03 Report version information for many libraries, including
1774 Exim version information for dynamically loaded libraries. Created
1775 version.h, now support a version extension string for distributors
1776 who patch heavily. Dynamic module ABI change.
1778 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1779 privilege escalation vulnerability whereby the Exim run-time user
1780 can cause root to append content of the attacker's choosing to
1783 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1786 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1787 If dropping privileges for untrusted macros, we disabled normal logging
1788 on the basis that it would fail; for the Exim run-time user, this is not
1789 the case, and it resulted in successful deliveries going unlogged.
1790 Fixed. Reported by Andreas Metzler.
1796 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1797 only prepend for the Resent-* case. Fixes regression introduced in
1798 Exim 4.70 by NM/22 for Bugzilla 607.
1800 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1801 increasing numbers of administrators be bitten by this.
1803 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1805 PP/03 Bugzilla 994: added openssl_options main configuration option.
1807 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1809 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1811 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1813 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1814 without a peer certificate, leading to a segfault because of an
1815 assumption that peers always have certificates. Be a little more
1816 paranoid. Problem reported by Martin Tscholak.
1818 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1819 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1820 NB: ClamAV planning to remove STREAM in "middle of 2010".
1821 CL also introduces -bmalware, various -d+acl logging additions and
1822 more caution in buffer sizes.
1824 PP/09 Implemented reverse_ip expansion operator.
1826 PP/10 Bugzilla 937: provide a "debug" ACL control.
1828 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1830 PP/12 Bugzilla 973: Implement --version.
1832 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1834 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1836 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1838 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1839 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1840 ignore trailing whitespace.
1842 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1844 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1847 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1848 Notification from Dr Andrew Aitchison.
1850 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1851 ExtendedDetectionInfo response format.
1852 Notification from John Horne.
1854 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1857 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1858 XSL and documented dependency on system catalogs, with examples of how
1861 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1864 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1865 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1866 configuration file which is writeable by the Exim user or group.
1868 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1869 of configuration files to cover files specified with the -C option if
1870 they are going to be used with root privileges, not just the default
1873 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1874 option (effectively making it always true).
1876 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1877 files to be used while preserving root privileges.
1879 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1880 that rogue child processes cannot use them.
1882 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1883 run-time user, instead of root.
1885 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1886 Exim run-time user without dropping privileges.
1888 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1889 result string, instead of calling string_vformat() twice with the same
1892 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1893 for other users. Others should always drop root privileges if they use
1894 -C on the command line, even for a whitelisted configure file.
1896 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1898 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1904 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1905 $data_path, and $header_path variables; fixed documentation bugs and
1908 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1909 exipick to access non-standard spools, including the "frozen" queue
1912 NM/01 Bugzilla 965: Support mysql stored procedures.
1913 Patch from Alain Williams
1915 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1917 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1918 Patch from Andreas Metzler
1920 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1921 Patch from Kirill Miazine
1923 NM/05 Bugzilla 671: Added umask to procmail example.
1925 JJ/03 installed exipick 20100323.0, fixing doc bug
1927 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1928 directory. Notification and patch from Dan Rosenberg.
1930 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1932 TK/02 Improve log output when DKIM signing operation fails.
1934 MH/01 Treat the transport option dkim_domain as a colon separated
1935 list, not as a single string, and sign the message with each element,
1936 omitting multiple occurences of the same signer.
1938 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1939 Bugzilla 985, 986. Patch by Simon Arlott
1941 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1942 Patch by Simon Arlott
1944 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1945 MBX locking. Notification from Dan Rosenberg.
1951 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1953 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1955 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1957 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1959 NM/04 Bugzilla 915: Flag broken perl installation during build.
1965 TK/01 Added patch by Johannes Berg that expands the main option
1966 "spamd_address" if it starts with a dollar sign.
1968 TK/02 Write list of recipients to X-Envelope-Sender header when building
1969 the mbox-format spool file for content scanning (suggested by Jakob
1972 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1973 (http://www.dcc-servers.net/) support via dccifd. Activated by
1974 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1976 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1977 by Mark Daniel Reidel <mr@df.eu>.
1979 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1980 When building exim an external PCRE library is now needed -
1981 PCRE is a system library on the majority of modern systems.
1982 See entry on PCRE_LIBS in EDITME file.
1984 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1985 conversation. Added nologin parameter to request.
1986 Patch contributed by Kirill Miazine.
1988 TF/01 Do not log submission mode rewrites if they do not change the address.
1990 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1992 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1993 log files in place. Contributed by Roberto Lima.
1995 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1997 TF/03 Bugzilla 615: When checking the local_parts router precondition
1998 after a local_part_suffix or local_part_prefix option, Exim now
1999 does not use the address's named list lookup cache, since this
2000 contains cached lookups for the whole local part.
2002 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2003 Robert Millan. Documentation is in experimental-spec.txt.
2005 TF/04 Bugzilla 668: Fix parallel build (make -j).
2007 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2009 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2010 Patch provided by Jan Srzednicki.
2012 TF/05 Leading white space used to be stripped from $spam_report which
2013 wrecked the formatting. Now it is preserved.
2015 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2016 that they are available at delivery time.
2018 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2020 TF/08 TLS error reporting now respects the incoming_interface and
2021 incoming_port log selectors.
2023 TF/09 Produce a more useful error message if an SMTP transport's hosts
2024 setting expands to an empty string.
2026 NM/06 Bugzilla 744: EXPN did not work under TLS.
2027 Patch provided by Phil Pennock.
2029 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2030 Patch provided by Richard Godbee.
2032 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2033 acl_smtp_notquit, added index entry.
2035 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2036 Patch provided by Eugene Bujak.
2038 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2039 accept(). Patch provided by Maxim Dounin.
2041 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2042 Patch provided by Phil Pennock.
2044 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2046 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2047 Patch provided by Brad "anomie" Jorsch.
2049 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2050 Patch provided by Dean Brooks.
2052 TK/05 Add native DKIM support (does not depend on external libraries).
2054 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2055 Patch provided by Graeme Fowler.
2057 NM/16 Bugzilla 851: Documentation example syntax fix.
2059 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2061 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2064 NM/19 Bugzilla 745: TLS version reporting.
2065 Patch provided by Phil Pennock.
2067 NM/20 Bugzilla 167: bool: condition support.
2068 Patch provided by Phil Pennock.
2070 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2071 clients. Patch provided by Phil Pennock.
2073 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2074 Patch provided by Brad "anomie" Jorsch.
2076 NM/23 Bugzilla 687: Fix misparses in eximstats.
2077 Patch provided by Heiko Schlittermann.
2079 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2080 Patch provided by Heiko Schlittermann.
2082 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2083 plus update to original patch.
2085 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2087 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2088 Patch provided by David Brownlee.
2090 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2092 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2094 NM/30 Bugzilla 888: TLS documentation bugfixes.
2096 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2098 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2099 Unlike the original bugzilla I have changed all shell scripts in src tree.
2101 NM/33 Bugzilla 898: Transport filter timeout fix.
2102 Patch by Todd Rinaldo.
2104 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2105 Patch by Serge Demonchaux.
2107 NM/35 Bugzilla 39: Base64 decode bug fixes.
2108 Patch by Jakob Hirsch.
2110 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2112 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2114 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2116 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2122 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2123 ALT-N's libdkim that I have put here:
2124 http://duncanthrax.net/exim-experimental/
2126 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2127 were called 'true' and it seems that C99 defines that as a reserved
2128 keyword to be used with 'bool' variable types. That means you could
2129 not include C99-style headers which use bools without triggering
2130 build errors in sieve.c.
2132 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2133 as mailq or other aliases. Changed the --help handling significantly
2134 to do whats expected. exim_usage() emits usage/help information.
2136 SC/01 Added the -bylocaldomain option to eximstats.
2138 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2140 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2142 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2148 PH/01 Another patch from the Sieve maintainer.
2150 PH/02 When an IPv6 address is converted to a string for single-key lookup
2151 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2152 dots are used instead of colons so that keys in lsearch files need not
2153 contain colons. This was done some time before quoting was made available
2154 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2155 (notated using the quote facility) so as to distinguish them from IPv4
2156 keys. This meant that lookups for IP addresses in host lists did not work
2157 for iplsearch lookups.
2159 This has been fixed by arranging for IPv6 addresses to be expressed with
2160 colons if the lookup type is iplsearch. This is not incompatible, because
2161 previously such lookups could never work.
2163 The situation is now rather anomalous, since one *can* have colons in
2164 ordinary lsearch keys. However, making the change in all cases is
2165 incompatible and would probably break a number of configurations.
2167 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2170 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2171 conversion specification without a maximum field width, thereby enabling
2172 a rogue spamd server to cause a buffer overflow. While nobody in their
2173 right mind would setup Exim to query an untrusted spamd server, an
2174 attacker that gains access to a server running spamd could potentially
2175 exploit this vulnerability to run arbitrary code as the Exim user.
2177 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2178 $primary_hostname instead of what libspf2 thinks the hosts name is.
2180 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2181 a directory entry by the name of the lookup key. Previously, if a
2182 symlink pointed to a non-existing file or a file in a directory that
2183 Exim lacked permissions to read, a lookup for a key matching that
2184 symlink would fail. Now it is enough that a matching directory entry
2185 exists, symlink or not. (Bugzilla 503.)
2187 PH/03 The body_linecount and body_zerocount variables are now exported in the
2190 PH/04 Added the $dnslist_matched variable.
2192 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2193 This means they are set thereafter only if the connection becomes
2196 PH/06 Added the client_condition to authenticators so that some can be skipped
2197 by clients under certain conditions.
2199 PH/07 The error message for a badly-placed control=no_multiline_responses left
2200 "_responses" off the end of the name.
2202 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2204 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2205 (without spaces) instead of just copying the configuration text.
2207 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2209 PH/11 Added $max_received_linelength.
2211 PH/12 Added +ignore_defer and +include_defer to host lists.
2213 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2214 way in which PCRE > 7.0 is built.
2216 PH/14 Implemented queue_only_load_latch.
2218 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2219 MAIL command. The effect was to mangle the value on 64-bit systems.
2221 PH/16 Another patch from the Sieve maintainer.
2223 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2225 PH/18 If a system quota error occurred while trying to create the file for
2226 a maildir delivery, the message "Mailbox is full" was not appended to the
2227 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2228 only to a quota excession during the actual writing of the file.
2230 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2231 characters?) which causes problems in log lines. The DN values are now
2232 passed through string_printing() before being added to log lines.
2234 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2235 and InterBase are left for another time.)
2237 PH/21 Added message_body_newlines option.
2239 PH/22 Guard against possible overflow in moan_check_errorcopy().
2241 PH/23 POSIX allows open() to be a macro; guard against that.
2243 PH/24 If the recipient of an error message contained an @ in the local part
2244 (suitably quoted, of course), incorrect values were put in $domain and
2245 $local_part during the evaluation of errors_copy.
2251 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2252 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2255 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2256 issue a MAIL command.
2258 PH/02 In an ACL statement such as
2260 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2262 if a client was not listed at all, or was listed with a value other than
2263 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2264 the condition was not true (as it should be), so access was not denied.
2265 The bug was that the ! inversion was incorrectly passed on to the second
2266 item. This has been fixed.
2268 PH/03 Added additional dnslists conditions == and =& which are different from
2269 = and & when the dns lookup returns more than one IP address.
2271 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2272 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2274 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2275 FSYNC, which compiles an option called disable_fsync that allows for
2276 bypassing fsync(). The documentation is heavily laced with warnings.
2278 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2280 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2281 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2282 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2283 including adding "make clean"; (3) Added -fPIC when compiling the test
2284 dynamically loaded module, to get rid of a warning.
2286 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2287 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2288 = 0s. The bug is otherwise harmless.
2290 PH/07 There was a bug in the dovecot authenticator such that the value of
2291 $auth1 could be overwritten, and so not correctly preserved, after a
2292 successful authentication. This usually meant that the value preserved by
2293 the server_setid option was incorrect.
2295 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2297 PH/09 Installed PCRE release 7.0.
2299 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2300 run for batched SMTP input. It is now run at the start of every message
2301 in the batch. While fixing this I discovered that the process information
2302 (output by running exiwhat) was not always getting set for -bs and -bS
2303 input. This is fixed, and it now also says "batched" for BSMTP.
2305 PH/11 Added control=no_pipelining.
2307 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2308 patch, slightly modified), and move the expansion of helo_data till after
2309 the connection is made in the smtp transport (so it can use these
2312 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2314 PH/14 Added log_selector = +pid.
2316 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2318 PH/16 Add ${if forany and ${if forall.
2320 PH/17 Added dsn_from option to vary the From: line in DSNs.
2322 PH/18 Flush SMTP output before performing a callout, unless control =
2323 no_callout_flush is set.
2325 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2326 was true (the default) a successful delivery failed to delete the retry
2327 item, thus causing premature timeout of the address. The bug is now
2330 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2332 PH/21 Long custom messages for fakedefer and fakereject are now split up
2333 into multiline responses in the same way that messages for "deny" and
2334 other ACL rejections are.
2336 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2337 with slight modification.
2339 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2340 draft, changing the syntax and factoring some duplicate code".
2342 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2343 for deliveries of the second and subsequent messages over the same SMTP
2346 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2347 ${reduce, with only minor "tidies".
2349 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2351 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2352 expansion side effects.
2354 PH/27 When a message times out after an over-quota error from an Exim-imposed
2355 quota, the bounce message says "mailbox is full". This message was not
2356 being given when it was a system quota that was exceeded. It now should
2359 MH/03 Made $recipients available in local_scan(). local_scan() already has
2360 better access to the recipient list through recipients_list[], but
2361 $recipients can be useful in postmaster-provided expansion strings.
2363 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2364 in the case of a MAIL command with additional options following the
2365 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2366 were accidentally chopped off.
2368 PH/29 SMTP synchronization checks are implemented when a command is read -
2369 there is a check that no more input is waiting when there shouldn't be
2370 any. However, for some commands, a delay in an ACL can mean that it is
2371 some time before the response is written. In this time, more input might
2372 arrive, invalidly. So now there are extra checks after an ACL has run for
2373 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2374 pipelining has not been advertised.
2376 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2378 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2379 returned in the SMTP response when smtp_return_error_details was false.
2380 This has been fixed.
2382 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2383 instead of the C I/O that was originally supplied, because problems were
2384 reported on Solaris.
2386 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2387 Exim which did not show up earlier: it was assuming that a call to
2388 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2389 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2390 was a macro that became an assignment, so it seemed to work. This has
2391 changed to a proper function call with a void return, hence the compile
2392 error. Exim's code has been fixed.
2394 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2397 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2399 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2401 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2402 because it turns out to be too big or there is a timeout) there is no
2403 "Completed" line in the log. When some messages of this type were
2404 selected by exigrep, they were listed as "not completed". Others were
2405 picked up by some special patterns. I have improved the selection
2406 criteria to be more general.
2408 PH/38 The host_find_failed option in the manualroute router can now be set
2409 to "ignore", to completely ignore a host whose IP address cannot be
2410 found. If all hosts are ignored, the behaviour is controlled by the new
2411 host_all_ignored option.
2413 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2414 homing or because of multiple MX records with /mx) generated more than
2415 one IP address, and the following item turned out to be the local host,
2416 all the secondary addresses of the first item were incorrectly removed
2417 from the list, along with the local host and any following hosts (which
2418 is what is supposed to happen).
2420 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2421 whoever called Exim into the -H file. In the case of the daemon it was
2422 behaving confusingly. When first started, it used values for whoever
2423 started the daemon, but after a SIGHUP it used the Exim user (because it
2424 calls itself on a restart). I have changed the code so that it now always
2427 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2428 message are rejected with the same error (e.g. no authentication or bad
2429 sender address), and a DATA command is nevertheless sent (as can happen
2430 with PIPELINING or a stupid MUA), the error message that was given to the
2431 RCPT commands is included in the rejection of the DATA command. This is
2432 intended to be helpful for MUAs that show only the final error to their
2435 PH/42 Another patch from the Sieve maintainer.
2437 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2438 Eximstats - Fixed some broken HTML links and added missing column headers
2440 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2441 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2443 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2445 PH/43 Yet another patch from the Sieve maintainer.
2447 PH/44 I found a way to check for a TCP/IP connection going away before sending
2448 the response to the final '.' that terminates a message, but only in the
2449 case where the client has not sent further data following the '.'
2450 (unfortunately, this is allowed). However, in many cases there won't be
2451 any further data because there won't be any more messages to send. A call
2452 to select() can be used: if it shows that the input is "ready", there is
2453 either input waiting, or the socket has been closed. An attempt to read
2454 the next input character can distinguish the two cases. Previously, Exim
2455 would have sent an OK response which the client would never have see.
2456 This could lead to message repetition. This fix should cure that, at
2457 least in a lot of common cases.
2459 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2460 advertised in response to EHLO.
2466 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2467 fixed by 4.65/MH/01 (is this a record?) are fixed:
2469 (i) An empty string was always treated as zero by the numeric comparison
2470 operators. This behaviour has been restored.
2472 (ii) It is documented that the numeric comparison operators always treat
2473 their arguments as decimal numbers. This was broken in that numbers
2474 starting with 0 were being interpreted as octal.
2476 While fixing these problems I realized that there was another issue that
2477 hadn't been noticed. Values of message_size_limit (both the global option
2478 and the transport option) were treated as octal if they started with 0.
2479 The documentation was vague. These values are now always treated as
2480 decimal, and I will make that clear in the documentation.
2486 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2487 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2490 MH/01 Don't check that the operands of numeric comparison operators are
2491 integers when their expansion is in "skipping" mode (fixes bug
2492 introduced by 4.64-PH/07).
2494 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2495 child addresses, Exim now panics and dies. Previously, because the count
2496 is held in a short int, deliveries were likely to be lost. As such a
2497 large number of recipients for a single message is ridiculous
2498 (performance will be very, very poor), I have chosen to impose a limit
2499 rather than extend the field.
2505 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2506 leftover -K file (the existence of which was triggered by #402).
2507 While we were at it, introduced process PID as part of the -K
2508 filename. This should rule out race conditions when creating
2511 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2512 processing considerably. Previous code took too long for large mails,
2513 triggering a timeout which in turn triggers #401.
2515 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2516 in the DK code in transports.c. sendfile() is not really portable,
2517 hence the _LINUX specificness.
2519 TF/01 In the add_headers option to the mail command in an Exim filter,
2520 there was a bug that Exim would claim a syntax error in any
2521 header after the first one which had an odd number of characters
2524 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2525 callout verification, Exim cached a "reject" for the entire domain. This
2526 is correct for most verifications, but it is not correct for a recipient
2527 verification with use_sender or use_postmaster set, because in that case
2528 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2529 case of MAIL FROM:<> rejection from other early rejections (e.g.
2530 rejection of HELO). When verifying a recipient using a non-null MAIL
2531 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2532 Whatever the result of the callout, the value of the domain cache is
2533 left unchanged (for any other kind of callout, getting as far as trying
2534 RCPT means that the domain itself is ok).
2536 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2539 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2540 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2541 session, the daemon ignores ECONNECT errors and logs others; it now
2542 ignores EPIPE as well.
2544 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2545 (quoted-printable decoding).
2547 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2548 later the small subsequent patch to fix an introduced bug.
2550 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2552 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2554 PH/08 An error is now given if message_size_limit is specified negative.
2556 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2557 to be given (somewhat) arbitrary names.
2559 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2562 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2563 miscellaneous code fixes
2565 PH/10 Added the log_reject_target ACL modifier to specify where to log
2568 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2569 hostname. This is wrong, because it relates to the incoming message (and
2570 probably the interface on which it is arriving) and not to the outgoing
2571 callout (which could be using a different interface). This has been
2572 changed to use the value of the helo_data option from the smtp transport
2573 instead - this is what is used when a message is actually being sent. If
2574 there is no remote transport (possible with a router that sets up host
2575 addresses), $smtp_active_hostname is used.
2577 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2578 tweaks were necessary in order to get it to work (see also 21 below):
2579 (a) The code assumed that strncpy() returns a negative number on buffer
2580 overflow, which isn't the case. Replaced with Exim's string_format()
2582 (b) There were several signed/unsigned issues. I just did the minimum
2583 hacking in of casts. There is scope for a larger refactoring.
2584 (c) The code used strcasecmp() which is not a standard C function.
2585 Replaced with Exim's strcmpic() function.
2586 (d) The code set only $1; it now sets $auth1 as well.
2587 (e) A simple test gave the error "authentication client didn't specify
2588 service in request". It would seem that Dovecot has changed its
2589 interface. Fortunately there's a specification; I followed it and
2590 changed what the client sends and it appears to be working now.
2592 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2595 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2596 address A is aliased to B and C, where B exists and C does not. Without
2597 -v the output is "A verified" because verification stops after a
2598 successful redirection if more than one address is generated. However,
2599 with -v the child addresses are also verified. Exim was outputting "A
2600 failed to verify" and then showing the successful verification for C,
2601 with its parentage. It now outputs "B failed to verify", showing B's
2602 parentage before showing the successful verification of C.
2604 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2605 look up a TXT record in a specific list after matching in a combined
2608 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2609 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2610 they consult the DNS. I had assumed they would set it the way they
2611 wanted; and indeed my experiments on Linux seem to show that in some
2612 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2613 To be on the safe side, however, I have now made the interface to
2614 host_find_byname() similar to host_find_bydns(), with an argument
2615 containing the DNS resolver options. The host_find_byname() function now
2616 sets these options at its start, just as host_find_bydns() does. The smtp
2617 transport options dns_qualify_single and dns_search_parents are passed to
2618 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2619 of host_find_byname() use the default settings of RES_DEFNAMES
2620 (qualify_single) but not RES_DNSRCH (search_parents).
2622 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2623 spool_read_header() do less string testing, by means of a preliminary
2624 switch on the second character of optional "-foo" lines. (This is
2625 overdue, caused by the large number of possibilities that now exist.
2626 Originally there were few.) While I was there, I also converted the
2627 str(n)cmp tests so they don't re-test the leading "-" and the first
2628 character, in the hope this might squeeze out yet more improvement.
2630 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2631 flag allowing group syntax was set by the header_syntax check but not
2632 turned off, possible causing trouble later; (2) The flag was not being
2633 set at all for the header_verify test, causing "group"-style headers to
2634 be rejected. I have now set it in this case, and also caused header_
2635 verify to ignore an empty address taken from a group. While doing this, I
2636 came across some other cases where the code for allowing group syntax
2637 while scanning a header line wasn't quite right (mostly, not resetting
2638 the flag correctly in the right place). These bugs could have caused
2639 trouble for malformed header lines. I hope it is now all correct.
2641 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2642 with the "reply" argument non-NULL. The code, however (which originally
2643 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2644 but it didn't always do it. This confused somebody who was copying the
2645 code for some other use. I have removed all the tests.
2647 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2648 feature that was used to support insecure browsers during the U.S. crypto
2649 embargo. It requires special client support, and Exim is probably the
2650 only MTA that supported it -- and would never use it because real RSA is
2651 always available. This code has been removed, because it had the bad
2652 effect of slowing Exim down by computing (never used) parameters for the
2653 RSA_EXPORT functionality.
2655 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2656 authenticator to fail if there's a tab character in the incoming data
2657 (there should never be unless someone is messing about, as it's supposed
2658 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2659 if the connection is using TLS or if the remote IP is the same as the
2660 local IP, and the "valid-client-cert option" if a client certificate has
2663 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2664 authenticators. This can be used for authorization after authentication
2665 succeeds. (In the case of plaintext, it servers for both authentication
2668 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2669 if any retry times were supplied.
2671 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2672 connection, where there is no client IP address to check. In this
2673 situation, the verify now always succeeds.
2675 PH/25 Applied John Jetmore's -Mset patch.
2677 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2679 PH/27 In a string expansion for a processed (not raw) header when multiple
2680 headers of the same name were present, leading whitespace was being
2681 removed from all of them, but trailing whitespace was being removed only
2682 from the last one. Now trailing whitespace is removed from each header
2683 before concatenation. Completely empty headers in a concatenation (as
2684 before) are ignored.
2686 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2687 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2689 PH/29 [Removed. This was a change that I later backed out, and forgot to
2690 correct the ChangeLog entry (that I had efficiently created) before
2691 committing the later change.]
2693 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2694 address errors (4xx response to RCPT) over the same connection as other
2695 messages routed to the same hosts. Such deliveries are always "forced",
2696 so retry times are not inspected. This resulted in far too many retries
2697 for the affected addresses. The effect occurred only when there were more
2698 hosts than the hosts_max_try setting in the smtp transport when it had
2699 the 4xx errors. Those hosts that it had tried were not added to the list
2700 of hosts for which the message was waiting, so if all were tried, there
2701 was no problem. Two fixes have been applied:
2703 (i) If there are any address or message errors in an SMTP delivery, none
2704 of the hosts (tried or untried) are now added to the list of hosts
2705 for which the message is waiting, so the message should not be a
2706 candidate for sending over the same connection that was used for a
2707 successful delivery of some other message. This seems entirely
2708 reasonable: after all the message is NOT "waiting for some host".
2709 This is so "obvious" that I'm not sure why it wasn't done
2710 previously. Hope I haven't missed anything, but it can't do any
2711 harm, as the worst effect is to miss an optimization.
2713 (ii) If, despite (i), such a delivery is accidentally attempted, the
2714 routing retry time is respected, so at least it doesn't keep
2715 hammering the server.
2717 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2718 in ${readsocket because some servers need this prod.
2720 PH/32 Added some extra debug output when updating a wait-xxx database.
2722 PH/33 The hint "could be header name not terminated by colon", which has been
2723 given for certain expansion errors for a long time, was not being given
2724 for the ${if def:h_colon_omitted{... case.
2726 PH/34 The spec says: "With one important exception, whenever a domain list is
2727 being scanned, $domain contains the subject domain." There was at least
2728 one case where this was not true.
2730 PH/35 The error "getsockname() failed: connection reset by peer" was being
2731 written to the panic log as well as the main log, but it isn't really
2732 panic-worthy as it just means the connection died rather early on. I have
2733 removed the panic log writing for the ECONNRESET error when getsockname()
2736 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2737 runs only) independently of the message's sender address. This meant
2738 that, if the 4xx error was in fact related to the sender, a different
2739 message to the same recipient with a different sender could confuse
2740 things. In particular, this can happen when sending to a greylisting
2741 server, but other circumstances could also provoke similar problems.
2742 I have changed the default so that the retry time for these errors is now
2743 based a combination of the sender and recipient addresses. This change
2744 can be overridden by setting address_retry_include_sender=false in the
2747 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2748 remote server are returned as part of bounce messages. This was not
2749 happening for LMTP over a pipe (the lmtp transport), but now it is the
2750 same for both kinds of LMTP.
2752 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2753 in header lines that were in fact CNAMEs. This is no longer the case.
2755 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2756 and queue runs started by the daemon processed all messages. This has
2757 been fixed so that -R and -S can now usefully be given with -q<time>.
2759 PH/40 Import PCRE release 6.7 (fixes some bugs).
2761 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2763 PH/42 Give an error if -q is specified more than once.
2765 PH/43 Renamed the variables $interface_address and $interface_port as
2766 $received_ip_address and $received_port, to make it clear that these
2767 values apply to message reception, and not to the outgoing interface when
2768 a message is delivered. (The old names remain recognized, of course.)
2770 PH/44 There was no timeout on the connect() call when using a Unix domain
2771 socket in the ${readsocket expansion. There now is.
2773 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2774 be meaningful with "accept".
2776 SC/01 Eximstats V1.43
2777 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2779 SC/02 Eximstats V1.44
2780 Use a glob alias rather than an array ref in the generated
2781 parser. This improves both readability and performance.
2783 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2784 Collect SpamAssassin and rejection statistics.
2785 Don't display local sender or destination tables unless
2786 there is data to show.
2787 Added average volumes into the top table text output.
2789 SC/04 Eximstats V1.46
2790 Collect data on the number of addresses (recipients)
2791 as well as the number of messages.
2793 SC/05 Eximstats V1.47
2794 Added 'Message too big' to the list of mail rejection
2795 reasons (thanks to Marco Gaiarin).
2797 SC/06 Eximstats V1.48
2798 Mainlog lines which have GMT offsets and are too short to
2799 have a flag are now skipped.
2801 SC/07 Eximstats V1.49 (Alain Williams)
2802 Added the -emptyok flag.
2804 SC/08 Eximstats V1.50
2805 Fixes for obtaining the IP address from reject messages.
2807 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2808 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2809 whitespace changes from 4.64-PH/27
2811 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2814 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2815 are found, allow negative numbers in numeric criteria)
2817 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2819 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2820 to match changes made in 4.64-PH/43
2822 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2824 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2825 "domain" is actually an IP address, and force a failure. This locks out
2826 those revolvers/nameservers that support "A-for-A" lookups, in
2827 contravention of the specifications.
2829 PH/48 When a host name was looked up from an IP address, and the subsequent
2830 forward lookup of the name timed out, the host name was left in
2831 $sender_host_name, contrary to the specification.
2833 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2834 restricted to single-key lookups, Exim was not diagnosing an error if
2835 * or *@ was used with a query-style lookup.
2837 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2839 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2840 long ago, but noone interested enough thought of it. Let's just say that
2841 the "1.1" means that there are some new functions that weren't there at
2842 some point in the past.
2844 PH/51 Error processing for expansion failure of helo_data from an smtp
2845 transport during callout processing was broken.
2847 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2848 tested/used via the -bh/-bhc/-bs options.
2850 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2851 bug, fixed in subsequent PCRE releases).
2853 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2854 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2856 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2862 SC/01 Use a glob alias rather than an array ref in eximstats generated
2863 parser. This improves both readability and performance.
2865 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2866 Don't display local sender or destination tables in eximstats unless
2867 there is data to show.
2868 Added average volumes into the eximstats top table text output.
2870 SC/03 Collect data on the number of addresses (recipients) as well
2871 as the number of messages in eximstats.
2873 TF/01 Correct an error in the documentation for the redirect router. Exim
2874 does (usually) call initgroups() when daemonizing.
2876 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2877 with consistent privilege compared to when running as a daemon.
2879 TF/03 Note in the spec that $authenticated_id is not set for local
2880 submissions from trusted users.
2882 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2883 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2885 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2886 by adding some example configuration directives to the default
2887 configuration file. A little bit of work is required to uncomment the
2888 directives and define how usernames and passwords are checked, but
2889 there is now a framework to start from.
2891 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2892 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2893 without this. I don't know how relevant this is to other LDAP libraries.
2895 PH/02 Add the verb name to the "unknown ACL verb" error.
2897 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2899 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2901 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2902 directory not expanded when it should be if an expanded home directory
2903 was set for the address (which is overridden by the transport).
2905 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2908 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2909 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2910 because it is too late at that time, and has no effect.
2912 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2913 security issue with \' (bugzilla #107). I could not use the
2914 PQescapeStringConn() function, because it needs a PGconn value as one of
2917 PH/08 When testing addresses using -bt, indicate those final addresses that
2918 are duplicates that would not cause an additional delivery. At least one
2919 person was confused, thinking that -bt output corresponded to deliveries.
2920 (Suppressing duplicates isn't a good idea as you lose the information
2921 about possibly different redirections that led to the duplicates.)
2923 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2924 systems where poll() doesn't work, in particular OS X.
2926 PH/10 Added more information to debugging output for retry time not reached.
2928 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2929 operations in malware.c.
2931 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2934 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2935 syslog_duplication set false, log lines that would normally be written
2936 both the the main log and to the reject log were not written to syslog at
2939 PH/14 In the default configuration, change the use of "message" in ACL warn
2940 statements to "add_header".
2942 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2943 not followed by a command (e.g. "seen endif").
2945 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2946 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2949 PH/17 Added extra conditions to the default value of delay_warning_condition
2953 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2954 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2955 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2958 The Auto-Submitted: and various List- headers are standardised, whereas I
2959 don't think Precedence: ever was.
2961 PH/18 Refactored debugging code in route_finduser() to show more information,
2962 in particular, the error code if getpwnam() issues one.
2964 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2965 This is apparently needed in addition to the PH/07 change above to avoid
2966 any possible encoding problems.
2968 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2969 but not after initializing Perl.
2971 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2972 output them only if debugging. By default they are written stderr,
2973 apparently, which is not desirable.
2975 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2978 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2981 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2983 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2984 authenticated or an ident call has been made. Suppress the default
2985 values for $authenticated_id and $authenticated_sender (but permit -oMai
2986 and -oMas) when testing with -bh.
2988 PH/24 Re-jigged the order of the tests in the default configuration so that the
2989 tests for valid domains and recipients precede the DNS black list and CSA
2990 tests, on the grounds that those ones are more expensive.
2992 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2993 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2994 This bug exists in every version of Exim that I still have, right back to
2997 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2998 However, an attempt to turn on case-sensitivity in a regex key by
2999 including (?-i) didn't work because the subject string was already
3000 lowercased, and the effects were non-intuitive. It turns out that a
3001 one-line patch can be used to allow (?-i) to work as expected.
3007 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3008 other effects) broke the use of negated acl sub-conditions.
3010 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3013 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3014 "Deny" causes Exim to reject the incoming connection with a 554 error.
3015 Unfortunately, if there is a major crisis, such as a disk failure,
3016 tcp-wrappers gives "deny", whereas what one would like would be some
3017 kind of temporary error. A kludge has been added to help with this.
3018 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3019 554 error is used if errno is still zero or contains ENOENT (which occurs
3020 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3023 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3025 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3026 errors. Otherwise a message that provokes a temporary error (when other
3027 messages do not) can cause a whole host to time out.
3029 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3030 addresses were routed directly to files or pipes from a redirect router.
3031 File deliveries just didn't batch; pipe deliveries might have suffered
3034 PH/06 A failure to get a lock for a hints database would erroneously always say
3035 "Failed to get write lock", even when it was really a read lock.
3037 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3038 of 0600. This has been changed to use the value of the lockfile_mode
3039 option (which defaults to 0600).
3041 PH/08 Applied small patch from the Sieve maintainer.
3043 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3044 folder from quota calculations, a direct delivery into this folder messed
3045 up the contents of the maildirsize file. This was because the regex was
3046 used only to exclude .Trash (or whatever) when the size of the mailbox
3047 was calculated. There was no check that a delivery was happening into an
3048 excluded directory. This bug has been fixed by ignoring all quota
3049 processing for deliveries into excluded directories.
3051 PH/10 Added the maildirfolder_create_regex option to appendfile.
3057 PH/01 The code for finding all the local interface addresses on a FreeBSD
3058 system running IPv6 was broken. This may well have applied to all BSD
3059 systems, as well as to others that have similar system calls. The broken
3060 code found IPv4 interfaces correctly, but gave incorrect values for the
3061 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3062 that it would not match correctly against @[] and not recognize the IPv6
3065 PH/02 The ipliteral router was not recognizing addresses of the form user@
3066 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3068 PH/03 Added disable_ipv6.
3070 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3071 decoded form, because it is most often used to construct To: headers
3072 lines in autoreplies, and the decoded form may well be syntactically
3073 invalid. However, $reply_address has leading white space removed, and all
3074 newlines turned into spaces so that the autoreply transport does not
3077 PH/05 If group was specified without a user on a router, and no group or user
3078 was specified on a transport, the group from the router was ignored.
3080 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3081 for visible compile-time settings that can be used to change these
3082 numbers, for those that want even more. Backwards compatibility with old
3083 spool files has been maintained. However, going back to a previous Exim
3084 release will lost any variables that are in spool files.
3086 PH/07 Two small changes when running in the test harness: increase delay when
3087 passing a TCP/IP connection to a new process, in case the original
3088 process has to generate a bounce, and remove special handling of
3089 127.0.0.2 (sic), which is no longer necessary.
3091 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3092 be the same on different OS.
3094 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3097 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3098 whether --show-vars was specified or not
3100 JJ/02 exipick: Added support for new ACL variable spool format introduced
3103 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3104 syntactically invalid From: or Reply-to: line, and a filter used this to
3105 generate an autoreply, and therefore failed to obtain an address for the
3106 autoreply, Exim could try to deliver to a non-existent relative file
3107 name, causing unrelated and misleading errors. What now happens is that
3108 it logs this as a hard delivery error, but does not attempt to create a
3111 PH/11 The exinext utility has a -C option for testing purposes, but although
3112 the given file was scanned by exinext itself; it wasn't being passed on
3113 when Exim was called.
3115 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3116 an end-of-file indication when reading a command response.
3118 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3119 compiled. In many other places in Exim, IPv6 addresses are always
3120 recognized, so I have changed this. It also means that IPv4 domain
3121 literals of the form [IPV4:n.n.n.n] are now always recognized.
3123 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3124 used if the router is not running as root, for example, when verifying at
3125 ACL time, or when using -bh. The debugging output from this situation was
3126 non-existent - all you got was a failure to exec. I have made two
3129 (a) Failures to set uid/gid, the current directory, or a process leader
3130 in a subprocess such as that created by queryprogram now generate
3131 suitable debugging output when -d is set.
3133 (b) The queryprogram router detects when it is not running as root,
3134 outputs suitable debugging information if -d is set, and then runs
3135 the subprocess without attempting to change uid/gid.
3137 PH/15 Minor change to Makefile for building test_host (undocumented testing
3140 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3141 additional section of a DNS packet that returns MX or SRV records.
3142 Instead, it always explicitly searches for A/AAAA records. This avoids
3143 major problems that occur when a DNS server includes only records of one
3144 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3145 fixed another bug: if SRV records were looked up and the corresponding
3146 address records were *not* found in the additional section, the port
3147 values from the SRV records were lost.
3149 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3150 using the correct key (the original address) when searching the retry
3151 rules in order to find which one to use for generating the retry hint.
3153 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3154 adding the default one. Similarly, if it contains a Reply-To: header, the
3155 errors_reply_to option, if set, is not used.
3157 PH/19 When calculating a retry time, Exim used to measure the "time since
3158 failure" by looking at the "first failed" field in the retry record. Now
3159 it does not use this if it is later than than the arrival time of the
3160 message. Instead it uses the arrival time. This makes for better
3161 behaviour in cases where some deliveries succeed, thus re-setting the
3162 "first failed" field. An example is a quota failure for a huge message
3163 when small messages continue to be delivered. Without this change, the
3164 "time since failure" will always be short, possible causing more frequent
3165 delivery attempts for the huge message than are intended.
3166 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3168 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3169 $1, $2, $3) because the numerical variables can be reset during some
3170 expansion items (e.g. "match"), thereby losing the authentication data.
3172 PH/21 Make -bV show the size of off_t variables so that the test suite can
3173 decide whether to run tests for quotas > 2G.
3175 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3176 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3177 filecount value is greater than 2G or if a quota value is greater than 2G
3178 on a system where the size of off_t is not greater than 4, a panic error
3181 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3182 never match. The debug and -bh output now contains an explicit error
3183 message indicating a malformed IPv4 address or mask.
3185 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3186 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3187 PH/23 above applies.
3189 PH/25 Do not write to syslog when running in the test harness. The only
3190 occasion when this arises is a failure to open the main or panic logs
3191 (for which there is an explicit test).
3193 PH/26 Added the /no_tell option to "control=freeze".
3195 PH/27 If a host name lookup failed very early in a connection, for example, if
3196 the IP address matched host_lookup and the reverse lookup yielded a name
3197 that did not have a forward lookup, an error message of the form "no IP
3198 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3199 could be logged. Now it outputs the IP address instead of "NULL".
3201 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3202 allows the sender and the authenticated sender to be set when
3203 submitting a message from within Exim. Since child_open_exim() is
3204 documented for local_scan(), the new function should be too.
3206 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3207 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3208 results in an empty string is now treated as unset.
3210 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3212 PH/31 Added sender_verify_fail logging option.
3214 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3215 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3216 tidied the source and removed it altogether.
3218 PH/33 When a queue run was abandoned because the load average was too high, a
3219 log line was always written; now it is written only if the queue_run log
3220 selector is set. In addition, the log line for abandonment now contains
3221 information about the queue run such as the pid. This is always present
3222 in "start" and "stop" lines but was omitted from the "abandon" line.
3224 PH/34 Omit spaces between a header name and the colon in the error message that
3225 is given when verify = headers_syntax fails (if there are lots of them,
3226 the message gets confusing).
3228 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3229 names, as there are now some PTR records that contain slashes. This check
3230 is only to protect against broken name servers that fall over on strange
3231 characters, so the fact that it applies to all lookups doesn't matter.
3233 PH/36 Now that the new test suite is complete, we can remove some of the
3234 special code in Exim that was needed for the old test suite. For example,
3235 sorting DNS records because real resolvers return them in an arbitrary
3236 order. The new test suite's fake resolver always returns records in the
3239 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3240 bounces) except when queue_only is set, to avoid logging races between
3241 the different processes.
3243 PH/38 Panic-die if .include specifies a non-absolute path.
3245 PH/39 A tweak to the "H" retry rule from its user.
3247 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3248 a label. They prevented compilation on older perls.
3250 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3251 a warning to be raised on newish perls.
3253 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3254 on queue. Changes to match documented behaviour of showing count of
3255 messages matching specified criteria.
3257 PH/40 Changed the default ident timeout from 30s to 5s.
3259 PH/41 Added support for the use of login_cap features, on those BSD systems
3260 that have them, for controlling the resources used by pipe deliveries.
3262 PH/42 The content-scanning code uses fopen() to create files in which to put
3263 message data. Previously it was not paying any attention to the mode of
3264 the files. Exim runs with umask(0) because the rest of the code creates
3265 files with open(), and sets the required mode explicitly. Thus, these
3266 files were ending up world-writeable. This was not a big issue, because,
3267 being within the spool directory, they were not world-accessible. I have
3268 created a function called modefopen, which takes an additional mode
3269 argument. It sets umask(777), creates the file, chmods it to the required
3270 mode, then resets the umask. All the relevant calls to fopen() in the
3271 content scanning code have been changed to use this function.
3273 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3274 to 24 hours. This avoids potential overflow problems when processing G
3275 and H retry rules. I suspect nobody ever tinkers with this value.
3277 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3279 PH/45 When the plaintext authenticator is running as a client, the server's
3280 challenges are checked to ensure they are valid base64 strings. By
3281 default, the authentication attempt is cancelled if an invalid string is
3282 received. Setting client_ignore_invalid_base64 true ignores these errors.
3283 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3284 they are received. Thus, the responses can be made to depend on the
3285 challenges. If an invalid string is ignored, an empty string is placed in
3288 PH/46 Messages that are created by the autoreply transport now contains a
3289 References: header, in accordance with RFCs 2822 and 3834.
3291 PH/47 Added authenticated_sender_force to the smtp transport.
3293 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3295 PH/49 Installed latest patch from the Sieve maintainer.
3297 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3298 was also set, the appendfile transport was unnecessarily scanning a
3299 directory of message files (e.g. for maildir delivery) to find the count
3300 of files (along with the size), even though it did not need this
3301 information. It now does the scan only if it needs to find either the
3302 size of the count of files.
3304 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3306 PH/52 Two bugs concerned with error handling when the smtp transport is
3309 (i) Exim was not creating retry information for temporary errors given
3310 for individual recipients after the DATA command when the smtp transport
3311 was used in LMTP mode. This meant that they could be retried too
3312 frequently, and not timed out correctly.
3314 (ii) Exim was setting the flag that allows error details to be returned
3315 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3316 recipients that were returned after the DATA command.
3318 PH/53 This is related to PH/52, but is more general: for any failing address,
3319 when detailed error information was permitted to be returned to the
3320 sender, but the error was temporary, then after the final timeout, only
3321 "retry timeout exceeded" was returned. Now it returns the full error as
3322 well as "retry timeout exceeded".
3324 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3325 do this, and (what is worse) MTAs that accept it.
3327 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3328 will now be deprecated.
3330 PH/56 New os.c-cygwin from the Cygwin maintainer.
3332 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3333 formats (previously only available in exim formats via -bpr, -bpru,
3334 and -bpra. Now also available in native and exiqgrep formats)
3336 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3337 with very large, slow to parse queues
3339 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3341 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3343 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3344 responses to authentication challenges, though it was showing the
3345 challenges; (ii) I've removed the CR characters from the debug output for
3348 PH/58 Allow for the insertion of a newline as well as a space when a string
3349 is turned into more than one encoded-word during RFC 2047 encoding. The
3350 Sieve code now uses this.
3352 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3353 data_4xx, lost_connection, tls_required.
3355 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3356 message was being sent as an SMTP response.
3358 PH/61 Add -l and -k options to exicyclog.
3360 PH/62 When verifying, if an address was redirected to one new address, so that
3361 verification continued, and the new address failed or deferred after
3362 having set something in $address_data, the value of $address_data was not
3363 passed back to the ACL. This was different to the case when no
3364 redirection occurred. The value is now passed back in both cases.
3366 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3367 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3368 use login_cap.h, so on its own it isn't the distinguishing feature. The
3369 new name refers directly to the setclassresources() function.
3371 PH/65 Added configuration files for NetBSD3.
3373 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3375 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3376 is preferred over IPv4.
3378 PH/68 The bounce_return_message and bounce_return_body options were not being
3379 honoured for bounces generated during the reception of non-SMTP messages.
3380 In particular, this applied to messages rejected by the ACL. This bug has
3381 been fixed. However, if bounce_return_message is true and bounce_return_
3382 body is false, the headers that are returned for a non-SMTP message
3383 include only those that have been read before the error was detected.
3384 (In the case of an ACL rejection, they have all been read.)
3386 PH/69 The HTML version of the specification is now built in a directory called
3387 spec_html instead of spec.html, because the latter looks like a path with
3388 a MIME-type, and this confuses some software.
3390 PH/70 Catch two compiler warnings in sieve.c.
3392 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3393 function verify_get_ident() calls ip_connect() to connect a socket, but
3394 if the "connect()" function timed out, ip_connect() used to close the
3395 socket. However, verify_get_ident() also closes the socket later, and in
3396 between Exim writes to the log, which may get opened at this point. When
3397 the socket was closed in ip_connect(), the log could get the same file
3398 descriptor number as the socket. This naturally causes chaos. The fix is
3399 not to close the socket in ip_connect(); the socket should be closed by
3400 the function that creates it. There was only one place in the code where
3401 this was missing, in the iplookup router, which I don't think anybody now
3402 uses, but I've fixed it anyway.
3404 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3405 well as to direct DNS lookups. Otherwise the handling of names in host
3406 lists is inconsistent and therefore confusing.
3412 PH/01 Two changes to the default runtime configuration:
3414 (1) Move the checks for relay_from_hosts and authenticated clients from
3415 after to before the (commented out) DNS black list checks.
3417 (2) Add control=submission to the relay_from_hosts and authenticated
3418 clients checks, on the grounds that messages accepted by these
3419 statements are most likely to be submissions.
3421 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3423 (1) Generate an error if the third argument for the ${prvs expansion is
3426 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3429 (3) Reset the variables that are obtained from the first argument of
3430 ${prvscheck and used in the second argument before leaving the code,
3431 because their memory is reclaimed, so using them afterwards may do
3434 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3435 one (it's much easier than Tom thought :-).
3437 (5) Because of (4), we can now allow for the use of $prvscheck_result
3438 inside the third argument.
3440 PH/03 For some reason, the default setting of PATH when running a command from
3441 a pipe transport was just "/usr/bin". I have changed it to
3444 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3445 anything to be listed in the output from -bV.
3447 PH/05 When a filter generated an autoreply, the entire To: header line was
3448 quoted in the delivery log line, like this:
3450 => >A.N.Other <ano@some.domain> <original@ddress> ...
3452 This has been changed so that it extracts the operative address. There
3453 may be more than one such address. If so, they are comma-separated, like
3456 => >ano@some.domain,ona@other.domain <original@ddress> ...
3458 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3459 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3460 not being looked up in the rDNS to get a host name, Exim was showing the
3461 IP address twice in Received: lines, even though the IP addresses were
3462 identical. For example:
3464 Received: from [1.2.3.4] (helo=[1.2.3.4])
3466 However, if the real host name was known, it was omitting the HELO data
3467 if it matched the actual IP address. This has been tidied up so that it
3468 doesn't show the same IP address twice.
3470 PH/07 When both +timestamp and +memory debugging was on, the value given by
3471 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3472 called by the debug printing, thereby overwriting the timestamp buffer.
3473 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3475 PH/08 When the original message was included in an autoreply transport, it
3476 always said "this is a copy of the message, including all the headers",
3477 even if body_only or headers_only was set. It now gives an appropriate
3480 PH/09 Applied a patch from the Sieve maintainer which:
3482 o fixes some comments
3483 o adds the (disabled) notify extension core
3484 o adds some debug output for the result of if/elsif tests
3485 o points to the current vacation draft in the documentation
3486 and documents the missing references header update
3490 o fixes a bug in processing the envelope test (when testing
3491 multiple envelope elements, the last element determined the
3494 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3495 Electronic Mail") by including:
3497 Auto-submitted: auto-generated
3499 in the messages that it generates (bounce messages and others, such as
3500 warnings). In the case of bounce messages for non-SMTP messages, there was
3501 also a typo: it was using "Auto_submitted" (underscore instead of
3502 hyphen). Since every message generated by Exim is necessarily in response
3503 to another message, thes have all been changed to:
3505 Auto-Submitted: auto-replied
3507 in accordance with these statements in the RFC:
3509 The auto-replied keyword:
3511 - SHOULD be used on messages sent in direct response to another
3512 message by an automatic process,
3514 - MUST NOT be used on manually-generated messages,
3516 - MAY be used on Delivery Status Notifications (DSNs) and Message
3517 Disposition Notifications (MDNs),
3519 - MUST NOT be used on messages generated by automatic or periodic
3520 processes, except for messages which are automatic responses to
3523 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3524 to the default Received: header definition.
3526 PH/12 Added log selector acl_warn_skipped (default on).
3528 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3529 variables because (a) they are in the wrong storage pool and (b) even if
3530 they were copied, it wouldn't work properly because of the caching.
3532 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3533 checking when decoding. Apparently there are clients that generate
3534 overlong encoded strings. Why am I not surprised?
3536 PH/15 If the first argument of "${if match_address" was not empty, but did not
3537 contain an "@" character, Exim crashed. Now it writes a panic log message
3538 and treats the condition as false.
3540 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3542 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3543 extension "envelope-auth". The code is finished and in agreement with
3544 other implementations, but there is no documentation so far and in fact,
3545 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3546 not changing the active code.
3548 Print executed "if" and "elsif" statements when debugging is used. This
3549 helps a great deal to understand what a filter does.
3551 Document more things not specified clearly in RFC3028. I had all this
3552 sorted out, when out of a sudden new issues came to my mind. Oops."
3554 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3557 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3558 canonical form (as documented). However, after a host name lookup from
3559 the IP address, check_host() was doing a simple string comparison with
3560 addresses acquired from the DNS when checking that the found name did
3561 have the original IP as one of its addresses. Since any found IPv6
3562 addresses are likely to be in abbreviated form, the comparison could
3563 fail. Luckily, there already exists a function for doing the comparison
3564 by converting both addresses to binary, so now that is used instead of
3565 the text comparison.
3567 PH/20 There was another similar case to PH/19, when a complete host name was
3568 given in a host list; looking up its IP address could give an abbreviated
3569 form, whereas the current host's name might or might not be abbreviated.
3570 The same fix has been applied.
3576 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3577 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3580 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3581 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3583 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3585 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3586 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3587 Also, for "Auto-Submitted", treat anything other than "no" as
3588 non-personal, in accordance with RFC 3834. (Previously it treated
3589 anything starting "auto-" as non-personal.)
3591 TF/01 The control=submission/name=... option had a problem with syntax
3592 errors if the name included a slash character. The /name= option
3593 now slurps the rest of the string, so it can include any characters
3594 but it must come last in the list of options (after /sender_retain
3597 PH/05 Some modifications to the interface to the fake nameserver for the new
3605 TK/01 Added the "success_on_redirect" address verification option. See
3606 NewStuff for rationale and an example.
3608 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3610 PH/02 Patch to exigrep to allow it to work on syslog lines.
3612 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3613 fread() to skip over the body file's header line, because in Cygwin the
3614 header line is locked and is inaccessible.
3616 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3617 co-exist for some time) to make it clear that it is the Exim ID that is
3618 referenced, not the Message-ID: header line.
3620 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3621 string_format() function, because snprintf() does not exist on all
3624 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3625 ${stat: expansion item.
3627 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3628 protocol synchronization error", to keep the pedants happy.
3630 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3631 well as for IRIX systems, when gcc is being used. See the host.c source
3634 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3636 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3639 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3640 transport and to the smtp transport in LMTP mode.
3642 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3644 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3645 run a filter in a subprocess. This could lead to confusion in subsequent
3646 lookups in the parent process. There should also be a search_tidyup() at
3647 the end of the subprocess.
3649 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3650 only if the host matched helo_try_verify_hosts, which caused the
3651 verification to occur when the EHLO/HELO command was issued. The ACL just
3652 tested the remembered result. Now, if a previous verification attempt has
3653 not happened, "verify = helo" does it there and then.
3655 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3657 TK/03 Fix log output including CR from clamd.
3659 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3660 exist provoked a memory error which could cause a segfault.
3662 PH/15 Installed PCRE 6.2
3664 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3666 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3667 of the problem. Specifically, suggested +O2 rather than +O1 for the
3670 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3672 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3673 record, the port was not correctly logged when the outgoing_port log
3674 selector was set (it logged the transort's default port).
3676 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3677 fallback_hosts, and "hosts" in the smtp transport.
3679 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3680 host errors such as "Connection refused".
3682 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3683 authentication with radiusclient 0.4.9:
3685 - Error returned from rc_read_config was caught wrongly
3686 - Username/password not passed on to radius server due to wrong length.
3688 The presumption is that some radiusclient API changes for 4.51/PH/17
3689 were not taken care of correctly. The code is still untested by me (my
3690 Linux distribution still has 0.3.2 of radiusclient), but it was
3691 contributed by a Radius user.
3693 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3694 expanding the "port" option of the smtp transport.
3696 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3697 while reading a MIME header. Thanks to Tom Hughes for a patch.
3699 PH/24 Include config.h inside local_scan.h so that configuration settings are
3702 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3703 that in an ACL for RCPT (for example), you can examine exactly what was
3706 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3707 commands, but it was not correctly comparing the address with the actual
3708 client host address. Thus, it would show the EHLO address in Received:
3709 header lines when this was not necessary.
3711 PH/27 Added the % operator to ${eval:}.
3713 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3714 it should be ignoring failures (because with -C, for example, it has lost
3715 privilege). It wasn't ignoring creation failures other than "already
3718 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3721 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3722 because an input line was too long, either on its own, or by virtue of
3723 too many continuations, the temporary file was not being removed, and the
3724 return code was incorrect.
3726 PH/31 Missing "BOOL" in function definition in filtertest.c.
3728 PH/32 Applied Sieve patches from the maintainer.
3730 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3732 PH/33 Added "verify = not_blind".
3734 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3735 Local/Makefile (with some defaults set). These are used in built scripts
3736 such as exicyclog, but they have never been used in the exim_install
3737 script (though there are many overriding facilities there). I have
3738 arranged that the exim_install script now takes note of these two
3741 PH/35 Installed configuration files for Dragonfly.
3743 PH/36 When a locally submitted message by a trusted user did not contain a
3744 From: header, and the sender address was obtained from -f or from an SMTP
3745 MAIL command, and the trusted user did not use -F to supply a sender
3746 name, $originator_name was incorrectly used when constructing a From:
3747 header. Furthermore, $originator_name was used for submission mode
3748 messages from external hosts without From: headers in a similar way,
3749 which is clearly wrong.
3751 PH/37 Added control=suppress_local_fixups.
3753 PH/38 When log_selector = +received_sender was set, and the addition of the
3754 sender made the log line's construction buffer exactly full, or one byte
3755 less than full, an overflow happened when the terminating "\n" was
3758 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3759 when the result of a list match is failure because a DNS lookup failed.
3761 PH/40 RM_COMMAND is now used in the building process.
3763 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3764 the "build-* directories that it finds.
3766 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3767 address in a domain literal was a prefix of an interface address.
3769 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3770 when verifying a sender address, unless rewrite_headers is false.
3772 PH/44 Wrote a long comment about why errors_to addresses are verified as
3773 recipients, not senders.
3775 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3776 the ratelimit ACL was added.
3778 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3780 PH/46 Added extra information about PostgreSQL errors to the error string.
3782 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3783 avoiding the need to install special zones in a real server. This is
3784 backwards compatible; if it can't find the fake resolver, it drops back.
3785 Thus, both old and new test suites can be run.
3787 TF/02 Added util/ratelimit.pl
3789 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3790 clock is set back in time.
3792 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3793 Candler <B.Candler@pobox.com>.
3795 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3796 OK for addresses that are the result of redirections.
3798 PH/48 A number of further additions for the benefit of the new test suite,
3799 including a fake gethostbyname() that interfaces to the fake DNS resolver
3802 TF/06 The fix for widen_domains has also been applied to qualify_single and
3803 search_parents which are the other dnslookup options that can cause
3806 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3809 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3811 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3812 service" if it encountered an MX record with an empty target hostname.
3813 The message is now "an MX or SRV record indicated no SMTP service".
3815 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3816 if the DNS of the sending site is misconfigured. This is quite a
3817 common situation. This change restores the behaviour of treating a
3818 helo verification defer as a failure.
3820 PH/51 If self=fail was set on a router, the bounce message did not include the
3821 actual error message.
3827 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3829 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3830 command itself ended in error, the underlying message about the transport
3831 filter timeout was being overwritten with the pipe command error. Now the
3832 underlying error message should be appended to the second error message.
3834 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3836 PH/02 Reduce the amount of output that "make" produces by default. Full output
3837 can still be requested.
3839 PH/03 The warning log line about a condition test deferring for a "warn" verb
3840 was being output only once per connection, rather than after each
3841 occurrence (because it was using the same function as for successful
3842 "warn" verbs). This seems wrong, so I have changed it.
3844 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3845 it should not have, which might have caused a crash in the right
3846 circumstances, but probably never did.
3848 PH/04 Installed a modified version of Tony Finch's patch to make submission
3849 mode fix the return path as well as the Sender: header line, and to
3850 add a /name= option so that you can make the user's friendly name appear
3853 TF/03 Added the control = fakedefer ACL modifier.
3855 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3856 Mark Lowes for thorough testing.
3858 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3860 TK/03 Merged latest SRS patch from Miles Wilton.
3862 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3863 with the definition in sysexits.h (which is #included earlier).
3864 Fortunately, Exim does not actually use EX_OK. The code used to try to
3865 preserve the sysexits.h value, by assuming that macro definitions were
3866 scanned for macro replacements. I have been disabused of this notion,
3867 so now the code just undefines EX_OK before #including unistd.h.
3869 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3870 in the smtp transport. When a block could not be written in a single
3871 write() function, the timeout was being re-applied to each part-write.
3872 This seems wrong - if the receiver was accepting one byte at a time it
3873 would take for ever. The timeout is now adjusted when this happens. It
3874 doesn't have to be particularly precise.
3876 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3877 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3879 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3880 without a domain if the check to <postmaster@domain> fails.
3882 SC/01 Eximstats: added -xls and the ability to specify output files
3883 (patch written by Frank Heydlauf).
3885 SC/02 Eximstats: use FileHandles for outputting results.
3887 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3889 SC/04 Eximstats: fixed display of large numbers with -nvr option
3891 SC/05 Eximstats: fixed merging of reports with empty tables.
3893 SC/06 Eximstats: added the -include_original_destination flag
3895 SC/07 Eximstats: removed tabs and trailing whitespace.
3897 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3899 TK/06 MBOX spool code: Add real "From " MBOX separator line
3900 so the .eml file is really in mbox format (even though
3901 most programs do not really care). Patch from Alex Miller.
3903 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3904 The latter is generated from $received_to and is only set if the
3905 message has one envelope recipient. SA can use these headers,
3906 obviously out-of-the-box. Patch from Alex Miller.
3908 PH/08 The ${def test on a variable was returning false if the variable's
3909 value was "0", contrary to what the specification has always said!
3910 The result should be true unless the variable is empty.
3912 PH/09 The syntax error of a character other than { following "${if
3913 def:variable_name" (after optional whitespace) was not being diagnosed.
3914 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3915 accidental colon was present, for example, could give incorrect results.
3917 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3918 result is used (not including appendfile, where other changes are about
3921 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3922 This involved changing a lot of size variables from int to off_t. It
3923 should work with maildirs and everything.
3925 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3926 spamd dying while we are connected to it.
3928 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3931 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3932 function for BDB 4.3.
3934 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3936 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3937 the sources for PCRE have been re-arranged and more files are now
3940 PH/15 The code I had for printing potentially long long variables in PH/11
3941 above was not the best (it lost precision). The length of off_t variables
3942 is now inspected at build time, and an appropriate printing format (%ld
3943 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3944 to be "long long int" or "long int". This is needed for the internal
3945 formatting function string_vformat().
3947 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3948 the configuration file to be ":syslog", then the script "guesses" where
3949 the logs files are, rather than using the compiled in default. In our
3950 case the guess is not the same as the compiled default, so the script
3951 suddenly stopped working when I started to use syslog. The patch checks
3952 to see if log_file_path is "". If so, it attempts to read it from exim
3953 with no configuration file to get the compiled in version, before it
3954 falls back to the previous guessing code."
3956 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3957 implementing BATV in an Exim configuration. See NewStuff for the gory
3960 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3961 Makefile that are specific to HP-UX.
3963 PH/18 If the "use_postmaster" option was set for a recipient callout together
3964 with the "random" option, the postmaster address was used as the MAIL
3965 FROM address for the random test, but not for the subsequent recipient
3966 test. It is now used for both.
3968 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3969 patch removes a few documentation additions to RFC 3028, because the
3970 latest draft now contains them. It adds the new en;ascii-case comparator
3971 and a new error check for 8bit text in MIME parts. Comparator and
3972 require names are now matched exactly. I enabled the subaddress
3973 extension, but it is not well tested yet (read: it works for me)."
3975 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3976 rework some of the code of TK/09 above to avoid the hardwired use of
3977 "%lld" and "long long". Replaced the call to snprintf() with a call to
3980 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3981 records point to non-existent hosts", "retry timeout exceeded", and
3982 "retry time not reached for any host after a long failure period".
3984 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3985 experimental DomainKeys support:
3987 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3988 (2) On an error such as an illegally used "control", the wrong name for
3989 the control was given.
3991 These problems did NOT occur unless DomainKeys support was compiled.
3993 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3995 PH/24 Added ${if match_ip condition.
3997 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3998 databases so that it will be absolutely obvious if a crash occurs in the
3999 DB library. This is a regular occurrence (often caused by mis-matched
4002 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4003 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4004 picked up on a user's system that detects such things. There doesn't seem
4005 to be a gcc warning option for this - only an attribute that has to be
4006 put on the function's prototype. It seems that in Fedora Core 4 they have
4007 set this on a number of new functions. No doubt there will be more in due
4010 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4011 specify a transport. However, if an address that was verified by such a
4012 router was the subject of a callout, Exim crashed because it tried to
4013 read the rcpt_include_affixes from the non-existent transport. Now it
4014 just assumes that the setting of that option is false. This bug was
4015 introduced by 4.51/PH/31.
4017 PH/28 Changed -d+all to exclude +memory, because that information is very
4018 rarely of interest, but it makes the output a lot bigger. People tend to
4019 do -d+all out of habit.
4021 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4022 code in os-type was giving problems when libc.so lives in lib64, like on
4025 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4026 aren't the modern standard, and it seems that some systems' include files
4027 don't always have them. Exim was already checking for some of the newer
4028 ones like T_AAAA, and defining it itself. I've added checks for all the
4029 record types that Exim uses.
4031 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4032 not automatically generating a new one, as it is supposed to. This
4033 prevented TLS from working. If the file did exist, but contained invalid
4034 data, a new version was generated, as expected. It was only the case of a
4035 non-existent file that was broken.
4037 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4038 with a change in libdomainkeys > 0.64.
4040 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4041 from DNS. If the selector record carries the flag, it now has
4042 precedence over the domain-wide flag.
4044 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4046 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4047 the use of a port name that isn't defined in /etc/services) occurred, the
4048 message was deferred as in a normal delivery, and thus remained on the
4049 spool, instead of being failed because of the mua_wrapper setting. This
4050 is now fixed, and I tidied up some of the mua_wrapper messages at the
4053 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4054 the messages in a hash of arrays rather than using individual hashes.
4055 This is a bit cleaner and results in dramatic memory savings, albeit
4056 at a slight CPU cost.
4058 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4059 as requested by Marc Sherman.
4061 SC/10 Eximstats: added histograms for user specified patterns as requested
4064 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4066 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4067 fopen() in the content-scanning modules that did not already have it.
4073 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4074 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4076 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4078 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4080 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4081 to test Sieve filters that use "vacation".
4083 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4084 that changes the way the GnuTLS parameters are stored in the cache file.
4085 The new format can be generated externally. For backward compatibility,
4086 if the data in the cache doesn't make sense, Exim assumes it has read an
4087 old-format file, and it generates new data and writes a new file. This
4088 means that you can't go back to an older release without removing the
4091 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4092 work if there are any delivery delays because "one_time" forces the
4093 parent to be marked "delivered", so its unseen clone is never tried
4094 again. For this reason, Exim now forbids the simultaneous setting of
4097 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4098 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4099 entry if you want to know the details. The fix, however, neglected to
4100 consider the case when local delivery batching is involved. The test for
4101 "previously delivered" was not happening when checking to see if an
4102 address could be batched with a previous (undelivered) one; under
4103 certain circumstances this could lead to multiple deliveries to the same
4106 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4107 in its include files, and this causes problems building Exim.
4109 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4110 header_syntax) but Exim was just ignoring anything given after a slash.
4111 In particular, this caused confusion with an attempt to use "verify =
4112 reverse_host_lookup/defer_ok". An error is now given when options are
4113 supplied for verify items that do not have them. (Maybe reverse_host_
4114 lookup should have a defer_ok option, but that's a different point.)
4116 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4117 defined by RFC 821) to 2048, because there were problems with some AUTH
4118 commands, and RFC 1869 says the size should be increased for extended
4119 SMTP commands that take arguments.
4121 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4124 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4125 "unknown" error; now it says that the functionality isn't in the binary.
4127 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4128 an address' error message when a string expansion fails (syntax or
4129 whatever). Otherwise the password may appear in the log. Following change
4130 PH/42 below, there is no longer a chance of it appearing in a bounce
4133 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4135 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4136 of its addresses was ever tried. (Bugzilla bug #2.)
4138 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4139 the result incorrectly in the debug output. (It correctly added a newline
4140 to what was transported.)
4142 TF/01 Added $received_time.
4144 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4145 commented out examples of how to interface to a virus scanner and to
4146 SpamAssassin. Also added commented examples of av_scanner and
4147 spamd_address settings.
4149 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4150 and controls are allowed in which ACLs. There were a couple of minor
4151 errors. Some of the entries in the conditions table (which is a table of
4152 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4153 negation of where the condition IS allowed.
4155 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4157 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4158 header file does not have a version number, so I've had to invent a new
4159 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4160 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4161 radiusclient), but it was contributed by a Radius user.
4163 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4164 files or directories, for OpenSSL.
4166 PH/19 When an Exim process that is running as root has to create an Exim log
4167 file, it does so in a subprocess that runs as exim:exim so as to get the
4168 ownership right at creation (otherwise, other Exim processes might see
4169 the file with the wrong ownership). There was no test for failure of this
4170 fork() call, which would lead to the process getting stuck as it waited
4171 for a non-existent subprocess. Forks do occasionally fail when resources
4172 run out. I reviewed all the other calls to fork(); they all seem to check
4175 PH/20 When checking for unexpected SMTP input at connect time (before writing
4176 the banner), Exim was not dealing correctly with a non-positive return
4177 from the read() function. If the client had disconnected by this time,
4178 the result was a log entry for a synchronization error with an empty
4179 string after "input=" when read() returned zero. If read() returned -1
4180 (an event I could not check), uninitialized data bytes were printed.
4181 There were reports of junk text (parts of files, etc) appearing after
4184 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4186 PH/22 Added support for macro redefinition, and (re)definition in between
4187 driver and ACL definitions.
4189 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4190 forgetting to use the resulting value; it was using the unexpanded value.
4192 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4193 hadn't been configured. The fix is from Juergen Kreileder, who
4194 understands it better than I do:
4196 "Here's what I see happening with three configured cyrus_sasl
4197 authenticators configured (plain, login, cram-md5):
4199 On startup auth_cyrus_sasl_init() gets called for each of these.
4200 This means three calls to sasl_listmech() without a specified mech_list.
4201 => SASL tests which mechs of all available mechs actually work
4202 => three warnings about OTP not working
4203 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4205 With the patch, sasl_listmech() also gets called three times. But now
4206 SASL's mech_list option is set to the server_mech specified in the the
4207 authenticator. Or in other words, the answer from sasl_listmech()
4208 gets limited to just the mech you're testing for (which is different
4210 => the return list contains just 'plain' or 'login', 'cram-md5' or
4211 nothing depending on the value of ob->server_mech.
4213 I've just tested the patch: Authentication still works fine,
4214 unavailable mechs specified in the exim configuration are still
4215 caught, and the auth.log warnings about OTP are gone."
4217 PH/25 When debugging is enabled, the contents of the command line are added
4218 to the debugging output, even when log_selector=+arguments is not
4221 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4222 answer is "GNU", and only if the return is "GNU/something" is the answer
4225 PH/27 $acl_verify_message is now set immediately after the failure of a
4226 verification in an ACL, and so is available in subsequent modifiers. In
4227 particular, the message can be preserved by coding like this:
4229 warn !verify = sender
4230 set acl_m0 = $acl_verify_message
4232 Previously, $acl_verify_message was set only while expanding "message"
4233 and "log_message" when a very denied access.
4235 PH/28 Modified OS/os.c-Linux with
4237 -#ifndef OS_LOAD_AVERAGE
4238 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4240 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4241 nomenclature these days.)
4243 PH/29 Installed patch from the Sieve maintainer that adds the options
4244 sieve_useraddress and sieve_subaddress to the redirect router.
4246 PH/30 In these circumstances:
4247 . Two addresses routed to the same list of hosts;
4248 . First host does not offer TLS;
4249 . First host accepts first address;
4250 . First host gives temporary error to second address;
4251 . Second host offers TLS and a TLS session is established;
4252 . Second host accepts second address.
4253 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4254 and peerdn, if requested) that were in fact used only for the second
4257 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4258 attention to any local part prefix or suffix that was matched by the
4259 router that accepted the address. It now behaves in the same way as it
4260 does for delivery: the affixes are removed from the local part unless
4261 rcpt_include_affixes is set on the transport.
4263 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4264 timeout during the DATA phase of an incoming message.
4266 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4267 applied a patch sanctioned by the Sieve maintainer.
4269 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4270 the uid or gid is negative. A case of a negative gid caused this to be
4271 noticed. The fix allows for either to be negative.
4273 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4274 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4277 PH/36 The change PH/12 above was broken. Fixed it.
4279 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4280 the grounds that routing the same address twice would always produce the
4281 same answer. This might have been true once, but it is certainly no
4282 longer true now. Routing a child address may depend on the previous
4283 routing that produced that child. Some complicated redirection strategies
4284 went wrong when messages had multiple recipients, and made Exim's
4285 behaviour dependent on the order in which the addresses were given.
4287 I have moved the duplicate checking until after the routing is complete.
4288 Exim scans the addresses that are assigned to local and remote
4289 transports, and removes any duplicates. This means that more work will be
4290 done, as duplicates will always all be routed, but duplicates are
4291 presumably rare, so I don't expect this is of any significance.
4293 For deliveries to pipes, files, and autoreplies, the duplicate checking
4294 still happens during the routing process, since they are not going to be
4297 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4298 It corrects a timeout issue with spamd. This is Ian's comment: "The
4299 background is that sometimes spamd either never reads data from a
4300 connection it has accepted, or it never writes response data. The exiscan
4301 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4302 blindly assumes that writes won't block so it may never time out."
4304 PH/39 Allow G after quota size as well as K and M.
4306 PH/40 The value set for $authenticated_id in an authenticator may not contain
4307 binary zeroes or newlines because the value is written to log lines and
4308 to spool files. There was no check on this. Now the value is run through
4309 the string_printing() function so that such characters are converted to
4310 printable escape sequences.
4312 PH/41 $message_linecount is a new variable that contains the total number of
4313 lines in the message. Compare $body_linecount, which is the count for the
4316 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4317 bounce and delay warning messages, except in certain special cases, which
4320 (a) An SMTP error message from a remote host;
4321 (b) A message specified in a :fail: redirection;
4322 (c) A message specified in a "fail" command in a system filter;
4323 (d) A message specified in a FAIL return from the queryprogram router;
4324 (e) A message specified by the cannot_route_message router option.
4326 In these cases only, Exim does include the error details in bounce and
4327 warning messages. There are also a few cases where bland messages such
4328 as "unrouteable address" or "local delivery error" are given.
4330 PH/43 $value is now also set for the "else" part of a ${run expansion.
4332 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4333 being worked on, but at least Exim now implements the latest version to
4336 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4337 process to complete was treated as a delivery failure, a timeout while
4338 writing the message to the pipe was logged, but erroneously treated as a
4339 successful delivery. Such timeouts include transport filter timeouts. For
4340 consistency with the overall process timeout, these timeouts are now
4341 treated as errors, giving rise to delivery failures by default. However,
4342 there is now a new Boolean option for the pipe transport called
4343 timeout_defer, which, if set TRUE, converts the failures into defers for
4344 both kinds of timeout. A transport filter timeout is now identified in
4347 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4348 systems where "make" and "gmake" are different, calling "gmake" at top
4349 level broke things. I've arranged for the value of $(MAKE) to be passed
4350 from the Makefile to this script so that it can call the same version of
4354 A note about Exim versions 4.44 and 4.50
4355 ----------------------------------------
4357 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4358 changes of various kinds. As a consequence, a big documentation update was
4359 needed. This delayed the release for rather longer than seemed good, especially
4360 in the light of a couple of (minor) security issues. Therefore, the changes
4361 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4362 release. So 4.44 and 4.50 are in effect two different branches that both start
4365 I have left the 4.50 change log unchanged; it contains all the changes since
4366 4.43. The change log for 4.44 is below; many of its items are identical to
4367 those for 4.50. This seems to be the most sensible way to preserve the
4368 historical information.
4374 1. Minor wording change to the doc/README.SIEVE file.
4376 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4377 computation of the current number of files was incorrect.
4379 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4380 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4383 4. Give more explanation in the error message when the command for a transport
4384 filter fails to execute.
4386 5. There are several places where Exim runs a non-Exim command in a
4387 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4388 was being done only for the command run by the queryprogram router. It is
4389 now done for all such subprocesses. The other cases are: ${run, transport
4390 filters, and the commands run by the lmtp and pipe transports.
4392 6. Added CONFIGURE_GROUP build-time option.
4394 7. Some older OS have a limit of 256 on the maximum number of file
4395 descriptors. Exim was using setrlimit() to set 1000 as a large value
4396 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4397 systems. I've change it so that if it can't get 1000, it tries for 256.
4399 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4400 was an oversight, and furthermore, ever since the addition of extra
4401 controls (e.g. 4.43/32), the checks on when to allow different forms of
4402 "control" were broken. There should now be diagnostics for all cases when a
4403 control that does not make sense is encountered.
4405 9. Added the /retain_sender option to "control=submission".
4407 10. $recipients is now available in the predata ACL (oversight).
4409 11. Tidy the search cache before the fork to do a delivery from a message
4410 received from the command line. Otherwise the child will trigger a lookup
4411 failure and thereby defer the delivery if it tries to use (for example) a
4412 cached ldap connection that the parent has called unbind on.
4414 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4415 of $address_data from the recipient verification was clobbered by the
4416 sender verification.
4418 13. The value of address_data from a sender verification is now available in
4419 $sender_address_data in subsequent conditions in the ACL statement.
4421 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4423 15. Added a new option "connect=<time>" to callout options, to set a different
4426 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4427 was its contents. (It was OK if the option was not defined at all.)
4429 17. A "Completed" log line is now written for messages that are removed from
4430 the spool by the -Mrm option.
4432 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4433 information about exactly what failed.
4435 19. Added -dd to debug only the daemon process.
4437 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4438 handles timeouts, both on the server side and network timeouts. Renamed the
4439 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4441 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4442 It is now set to "smtps".
4444 22. $host_address is now set to the target address during the checking of
4445 ignore_target_hosts.
4447 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4448 being passed; this would have caused $sender_host_name to have been used if
4449 matching the list had actually called for a host name (not very likely,
4450 since this list is usually IP addresses). A host name is now passed as
4453 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4454 code that specifies a non-restarting handler (typically sigaction() in
4455 modern systems) in an attempt to fix a rare and obscure crash bug.
4457 25. Narrowed the window for a race in the daemon that could cause it to ignore
4458 SIGCHLD signals. This is not a major problem, because they are used only to
4459 wake it up if nothing else does.
4461 26. A malformed maildirsize file could cause Exim to calculate negative values
4462 for the mailbox size or file count. Odd effects could occur as a result.
4463 The maildirsize information is now recalculated if the size or filecount
4466 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4467 support for a long time. Removed HAVE_SYS_VFS_H.
4469 28. Installed the latest version of exipick from John Jetmore.
4471 29. In an address list, if the pattern was not a regular expression, an empty
4472 subject address (from a bounce message) matched only if the pattern was an
4473 empty string. Non-empty patterns were not even tested. This was the wrong
4474 because it is perfectly reasonable to use an empty address as part of a
4475 database query. An empty address is now tested by patterns that are
4476 lookups. However, all the other forms of pattern expect the subject to
4477 contain a local part and a domain, and therefore, for them, an empty
4478 address still always fails if the pattern is not itself empty.
4480 30. Exim went into a mad DNS loop when attempting to do a callout where the
4481 host was specified on an smtp transport, and looking it up yielded more
4482 than one IP address.
4484 31. Re-factored the code for checking spool and log partition space into a
4485 function that finds that data and another that does the check. The former
4486 is then used to implement four new variables: $spool_space, $log_space,
4487 $spool_inodes, and $log_inodes.
4489 32. The RFC2047 encoding function was originally intended for short strings
4490 such as real names; it was not keeping to the 75-character limit for
4491 encoded words that the RFC imposes. It now respects the limit, and
4492 generates multiple encoded words if necessary. To be on the safe side, I
4493 have increased the buffer size for the ${rfc2047: expansion operator from
4496 33. It is now permitted to omit both strings after an "if" condition; if the
4497 condition is true, the result is "true". As before, when the second string
4498 is omitted, a false condition yields an empty string. This makes it less
4499 cumbersome to write custom ACL and router conditions.
4501 34. Failure to deliver a bounce message always caused it to be frozen, even if
4502 there was an errors_to setting on the router. The errors_to setting is now
4505 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4506 canonical form (fully expanded) before being placed in
4507 $sender_host_address.
4509 36. The table in the code that translates DNS record types into text (T_A to
4510 "A" for instance) was missing entries for NS and CNAME. It is just possible
4511 that this could have caused confusion if both these types were looked up
4512 for the same domain, because the text type is used as part of Exim's
4513 per-process caching. But the chance of anyone hitting this buglet seems
4516 37. The dnsdb lookup has been extended in a number of ways.
4518 (1) There is a new type, "zns", which walks up the domain tree until it
4519 finds some nameserver records. It should be used with care.
4521 (2) There is a new type, "mxh", which is like "mx" except that it returns
4522 just the host names, not the priorities.
4524 (3) It is now possible to give a list of domains (or IP addresses) to be
4525 looked up. The behaviour when one of the lookups defers can be
4526 controlled by a keyword.
4528 (4) It is now possible to specify the separator character for use when
4529 multiple records are returned.
4531 38. The dnslists ACL condition has been extended: it is now possible to supply
4532 a list of IP addresses and/or domains to be looked up in a particular DNS
4535 39. Added log_selector=+queue_time_overall.
4537 40. When running the queue in the test harness, wait just a tad after forking a
4538 delivery process, to get repeatability of debugging output.
4540 41. Include certificate and key file names in error message when GnuTLS fails
4541 to set them up, because the GnuTLS error message doesn't include the name
4542 of the failing file when there is a problem reading it.
4544 42. Allow both -bf and -bF in the same test run.
4546 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4548 44. The "Exiscan patch" is now merged into the mainline Exim source.
4550 45. Sometimes the final signoff response after QUIT could fail to get
4551 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4552 < 0 before doing a fflush(). This bug looks as though it goes back to the
4553 introduction of TLS in release 3.20, but "sometimes" must have been rare
4554 because the tests only now provoked it.
4556 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4557 (this can affect the format of dates).
4559 47. exim_tidydb, when checking for the continued existence of a message for
4560 which it has found a message-specific retry record, was not finding
4561 messages that were in split spool directories. Consequently, it was
4562 deleting retry records that should have stayed in existence.
4564 48. Steve fixed some bugs in eximstats.
4566 49. The SPA authentication driver was not abandoning authentication and moving
4567 on to the next authenticator when an expansion was forced to fail,
4568 contradicting the general specification for all authenticators. Instead it
4569 was generating a temporary error. It now behaves as specified.
4571 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4572 (the order specifies the preference for clients). The order is now AES256,
4573 AES128, 3DES, ARCFOUR128.
4575 51. Small patch to Sieve code - explicitly set From: when generating an
4578 52. Exim crashed if a remote delivery caused a very long error message to be
4579 recorded - for instance if somebody sent an entire SpamAssassin report back
4580 as a large number of 550 error lines. This bug was coincidentally fixed by
4581 increasing the size of one of Exim's internal buffers (big_buffer) that
4582 happened as part of the Exiscan merge. However, to be on the safe side, I
4583 have made the code more robust (and fixed the comments that describe what
4586 53. Now that there can be additional text after "Completed" in log lines (if
4587 the queue_time_overall log selector is set), a one-byte patch to exigrep
4588 was needed to allow it to recognize "Completed" as not the last thing in
4591 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4592 patch that reportedly fixes this has been added. I am not expert enough to
4593 create a test for it. This is what the patch creator wrote:
4595 "I found a little strange behaviour of ldap code when working with
4596 Windows 2003 AD Domain, where users was placed in more than one
4597 Organization Units. When I tried to give exim partial DN, the exit code
4598 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4599 But simultaneously result of request was absolutely normal ldap result,
4600 so I produce this patch..."
4602 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4603 REFERENCE, so I have modified the code to exclude the patch when that macro
4606 55. Some experimental protocols are using DNS PTR records for new purposes. The
4607 keys for these records are domain names, not reversed IP addresses. The
4608 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4609 leaves it alone. Component reversal etc. now happens only for IP addresses.
4612 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4614 57. Double the size of the debug message buffer (to 2048) so that more of very
4615 long debug lines gets shown.
4617 58. The exicyclog utility now does better if the number of log files to keep
4618 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4620 59. Two changes related to the smtp_active_hostname option:
4622 (1) $smtp_active_hostname is now available as a variable.
4623 (2) The default for smtp_banner uses $smtp_active_hostname instead
4624 of $primary_hostname.
4626 60. The host_aton() function is supposed to be passed a string that is known
4627 to be a valid IP address. However, in the case of IPv6 addresses, it was
4628 not checking this. This is a hostage to fortune. Exim now panics and dies
4629 if the condition is not met. A case was found where this could be provoked
4630 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4631 components; fortuitously, this particular loophole had already been fixed
4632 by change 4.50/55 above.
4634 If there are any other similar loopholes, the new check in host_aton()
4635 itself should stop them being exploited. The report I received stated that
4636 data on the command line could provoke the exploit when Exim was running as
4637 exim, but did not say which command line option was involved. All I could
4638 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4639 running as the user.
4642 61. There was a buffer overflow vulnerability in the SPA authentication code
4643 (which came originally from the Samba project). I have added a test to the
4644 spa_base64_to_bits() function which I hope fixes it.
4647 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4648 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4650 63. The daemon start-up calls getloadavg() while still root for those OS that
4651 need the first call to be done as root, but it missed one case: when
4652 deliver_queue_load_max is set with deliver_drop_privilege. This is
4653 necessary for the benefit of the queue runner, because there is no re-exec
4654 when deliver_drop_privilege is set.
4656 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4657 This has been fixed.
4659 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4660 was in use, was not putting the data itself into the right store pool;
4661 consequently, it could be overwritten for a subsequent message in the same
4662 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4665 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4667 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4668 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4669 the function were treating the return as a boolean value, which happened to
4670 work because 0=false and not-0=true, but is not correct code.
4672 68. The host_aton() function was not handling scoped IPv6 addresses (those
4673 with, for example, "%eth0" on the end) correctly.
4675 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4676 negated items (that is, ~something) in unsigned ints. Some compilers
4677 apparently mutter when there is no cast.
4679 70. If an address verification called from an ACL failed, and did not produce a
4680 user-specific message (i.e. there was only a "system" message), nothing was
4681 put in $acl_verify_message. In this situation, it now puts the system
4684 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4685 session; change 4.31/43 added the unwanted input to the log line - except
4686 that it did not do this in the start of session case. It now does.
4688 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4689 This is wrong and can cause the other end to generate a synchronization
4690 error if it is another Exim or anything else that does the synchronization
4691 check. A QUIT command is no longer sent after a timeout.
4693 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4694 during host lookups.
4696 74. The defer_ok option of callout verification was not working if it was used
4697 when verifying addresses in header lines, that is, for this case:
4699 verify = header_sender/callout=defer_ok
4701 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4702 those file descriptors could be used for SMTP connections. If anything
4703 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4704 could be sent to the SMTP client, causing chaos. The daemon now opens
4705 stdin, stdout, and stderr to /dev/null when it puts itself into the
4708 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4709 log by default. The user can override this with suitable Perl magic.
4711 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4712 the log message when discard triggers, was not working for the DATA ACL or
4713 for the non-SMTP ACL.
4715 78. Error message wording change in sieve.c.
4717 79. If smtp_accept_max_per_host was set, the number of connections could be
4718 restricted to fewer than expected, because the daemon was trying to set up
4719 a new connection before checking whether the processes handling previous
4720 connections had finished. The check for completed processes is now done
4721 earlier. On busy systems, this bug wouldn't be noticed because something
4722 else would have woken the daemon, and it would have reaped the completed
4725 80. If a message was submitted locally by a user whose login name contained one
4726 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4727 It caused a spool format error. I have fixed the spool reading code. A
4728 related problem was that the "from" clause in the Received: line became
4729 illegal because of the space(s). It is now covered by ${quote_local_part.
4731 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4734 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4736 83. Patch from Sieve maintainer to fix unterminated string problem in
4737 "vacation" handling.
4739 84. Some minor changes to the Linux configuration files to help with other
4740 OS variants using glibc.
4742 85. One more patch for Sieve to update vacation handling to latest spec.
4745 ----------------------------------------------------
4746 See the note above about the 4.44 and 4.50 releases.
4747 ----------------------------------------------------
4753 1. Change 4.43/35 introduced a bug that caused file counts to be
4754 incorrectly computed when quota_filecount was set in an appendfile
4757 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4758 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4761 3. Give more explanation in the error message when the command for a transport
4762 filter fails to execute.
4764 4. There are several places where Exim runs a non-Exim command in a
4765 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4766 was being done only for the command run by the queryprogram router. It is
4767 now done for all such subprocesses. The other cases are: ${run, transport
4768 filters, and the commands run by the lmtp and pipe transports.
4770 5. Some older OS have a limit of 256 on the maximum number of file
4771 descriptors. Exim was using setrlimit() to set 1000 as a large value
4772 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4773 systems. I've change it so that if it can't get 1000, it tries for 256.
4775 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4776 was an oversight, and furthermore, ever since the addition of extra
4777 controls (e.g. 4.43/32), the checks on when to allow different forms of
4778 "control" were broken. There should now be diagnostics for all cases when a
4779 control that does not make sense is encountered.
4781 7. $recipients is now available in the predata ACL (oversight).
4783 8. Tidy the search cache before the fork to do a delivery from a message
4784 received from the command line. Otherwise the child will trigger a lookup
4785 failure and thereby defer the delivery if it tries to use (for example) a
4786 cached ldap connection that the parent has called unbind on.
4788 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4789 of $address_data from the recipient verification was clobbered by the
4790 sender verification.
4792 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4793 was its contents. (It was OK if the option was not defined at all.)
4795 11. A "Completed" log line is now written for messages that are removed from
4796 the spool by the -Mrm option.
4798 12. $host_address is now set to the target address during the checking of
4799 ignore_target_hosts.
4801 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4802 being passed; this would have caused $sender_host_name to have been used if
4803 matching the list had actually called for a host name (not very likely,
4804 since this list is usually IP addresses). A host name is now passed as
4807 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4808 code that specifies a non-restarting handler (typically sigaction() in
4809 modern systems) in an attempt to fix a rare and obscure crash bug.
4811 15. Narrowed the window for a race in the daemon that could cause it to ignore
4812 SIGCHLD signals. This is not a major problem, because they are used only to
4813 wake it up if nothing else does.
4815 16. A malformed maildirsize file could cause Exim to calculate negative values
4816 for the mailbox size or file count. Odd effects could occur as a result.
4817 The maildirsize information is now recalculated if the size or filecount
4820 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4821 support for a long time. Removed HAVE_SYS_VFS_H.
4823 18. Updated exipick to current release from John Jetmore.
4825 19. Allow an empty sender to be matched against a lookup in an address list.
4826 Previously the only cases considered were a regular expression, or an
4829 20. Exim went into a mad DNS lookup loop when doing a callout where the
4830 host was specified on the transport, if the DNS lookup yielded more than
4833 21. The RFC2047 encoding function was originally intended for short strings
4834 such as real names; it was not keeping to the 75-character limit for
4835 encoded words that the RFC imposes. It now respects the limit, and
4836 generates multiple encoded words if necessary. To be on the safe side, I
4837 have increased the buffer size for the ${rfc2047: expansion operator from
4840 22. Failure to deliver a bounce message always caused it to be frozen, even if
4841 there was an errors_to setting on the router. The errors_to setting is now
4844 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4845 canonical form (fully expanded) before being placed in
4846 $sender_host_address.
4848 24. Updated eximstats to version 1.33
4850 25. Include certificate and key file names in error message when GnuTLS fails
4851 to set them up, because the GnuTLS error message doesn't include the name
4852 of the failing file when there is a problem reading it.
4854 26. Expand error message when OpenSSL has problems setting up cert/key files.
4857 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4858 (this can affect the format of dates).
4860 28. exim_tidydb, when checking for the continued existence of a message for
4861 which it has found a message-specific retry record, was not finding
4862 messages that were in split spool directories. Consequently, it was
4863 deleting retry records that should have stayed in existence.
4865 29. eximstats updated to version 1.35
4866 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4867 1.35 - bugfix such that pie charts by volume are generated correctly
4869 30. The SPA authentication driver was not abandoning authentication and moving
4870 on to the next authenticator when an expansion was forced to fail,
4871 contradicting the general specification for all authenticators. Instead it
4872 was generating a temporary error. It now behaves as specified.
4874 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4875 (the order specifies the preference for clients). The order is now AES256,
4876 AES128, 3DES, ARCFOUR128.
4878 31. Small patch to Sieve code - explicitly set From: when generating an
4881 32. Exim crashed if a remote delivery caused a very long error message to be
4882 recorded - for instance if somebody sent an entire SpamAssassin report back
4883 as a large number of 550 error lines. This bug was coincidentally fixed by
4884 increasing the size of one of Exim's internal buffers (big_buffer) that
4885 happened as part of the Exiscan merge. However, to be on the safe side, I
4886 have made the code more robust (and fixed the comments that describe what
4889 33. Some experimental protocols are using DNS PTR records for new purposes. The
4890 keys for these records are domain names, not reversed IP addresses. The
4891 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4892 leaves it alone. Component reversal etc. now happens only for IP addresses.
4895 34. The host_aton() function is supposed to be passed a string that is known
4896 to be a valid IP address. However, in the case of IPv6 addresses, it was
4897 not checking this. This is a hostage to fortune. Exim now panics and dies
4898 if the condition is not met. A case was found where this could be provoked
4899 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4900 components; fortuitously, this particular loophole had already been fixed
4901 by change 4.50/55 or 4.44/33 above.
4903 If there are any other similar loopholes, the new check in host_aton()
4904 itself should stop them being exploited. The report I received stated that
4905 data on the command line could provoke the exploit when Exim was running as
4906 exim, but did not say which command line option was involved. All I could
4907 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4908 running as the user.
4911 35. There was a buffer overflow vulnerability in the SPA authentication code
4912 (which came originally from the Samba project). I have added a test to the
4913 spa_base64_to_bits() function which I hope fixes it.
4916 36. The daemon start-up calls getloadavg() while still root for those OS that
4917 need the first call to be done as root, but it missed one case: when
4918 deliver_queue_load_max is set with deliver_drop_privilege. This is
4919 necessary for the benefit of the queue runner, because there is no re-exec
4920 when deliver_drop_privilege is set.
4922 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4923 was in use, was not putting the data itself into the right store pool;
4924 consequently, it could be overwritten for a subsequent message in the same
4925 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4928 38. Sometimes the final signoff response after QUIT could fail to get
4929 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4930 < 0 before doing a fflush(). This bug looks as though it goes back to the
4931 introduction of TLS in release 3.20, but "sometimes" must have been rare
4932 because the tests only now provoked it.
4938 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4939 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4940 Now it returns an integer. A number of calls were still expecting a T/F
4941 return. Fortuitously, in all cases, the tests worked in OK situations,
4942 which is the norm. However, things would have gone wrong on any write
4943 failures on the smtp file descriptor. This function is used when sending
4944 messages over SMTP and also when doing verify callouts.
4946 2. When Exim is called to do synchronous delivery of a locally submitted
4947 message (the -odf or -odi options), it no longer closes stderr before doing
4950 3. Implemented the mua_wrapper option.
4952 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4954 5. Implemented the functions header_remove(), header_testname(),
4955 header_add_at_position(), and receive_remove_recipient(), and exported them
4958 6. If an ACL "warn" statement specified the addition of headers, Exim already
4959 inserted X-ACL-Warn: at the start if there was no header name. However, it
4960 was not making this test for the second and subsequent header lines if
4961 there were newlines in the string. This meant that an invalid header could
4962 be inserted if Exim was badly configured.
4964 7. Allow an ACL "warn" statement to add header lines at the start or after all
4965 the Received: headers, as well as at the end.
4967 8. Added the rcpt_4xx retry error code.
4969 9. Added postmaster_mailfrom=xxx to callout verification option.
4971 10. Added mailfrom=xxxx to the callout verification option, for verify=
4974 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4975 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4977 12. Inserted some casts to stop certain compilers warning when using pointer
4978 differences as field lengths or precisions in printf-type calls (mostly
4979 affecting debugging statements).
4981 13. Added optional readline() support for -be (dynamically loaded).
4983 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4984 same clock tick as a message's arrival, so that its received time was the
4985 same as the "first fail" time on the retry record, and that message
4986 remained on the queue past the ultimate address timeout, every queue runner
4987 would try a delivery (because it was past the ultimate address timeout) but
4988 after another failure, the ultimate address timeout, which should have then
4989 bounced the address, did not kick in. This was a "< instead of <=" error;
4990 in most cases the first failure would have been in the next clock tick
4991 after the received time, and all would be well.
4993 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4994 being recognized when the domain list was tested by the match_domain
4995 condition in an expansion string.
4997 16. Added the ${str2b64: operator.
4999 17. Exim was always calling setrlimit() to set a large limit for the number of
5000 processes, without checking whether the existing limit was already
5001 adequate. (It did check for the limit on file descriptors.) Furthermore,
5002 errors from getrlimit() and setrlimit() were being ignored. Now they are
5003 logged to the main and panic logs, but Exim does carry on, to try to do its
5004 job under whatever limits there are.
5006 18. Imported PCRE 5.0.
5008 19. Trivial typo in log message " temporarily refused connection" (the leading
5011 20. If the log selector return_path_on_delivery was set and an address was
5012 redirected to /dev/null, the delivery process crashed because it assumed
5013 that a return path would always be set for a "successful" delivery. In this
5014 case, the whole delivery is bypassed as an optimization, and therefore no
5017 21. Internal re-arrangement: the function for sending a challenge and reading
5018 a response while authentication was assuming a zero-terminated challenge
5019 string. It's now changed to take a pointer and a length, to allow for
5020 binary data in such strings.
5022 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5024 23. Exim was not respecting finduser_retries when seeking the login of the
5025 uid under which it was called; it was always trying 10 times. (The default
5026 setting of finduser_retries is zero.) Also, it was sleeping after the final
5027 failure, which is pointless.
5029 24. Implemented tls_on_connect_ports.
5031 25. Implemented acl_smtp_predata.
5033 26. If the domain in control=submission is set empty, Exim assumes that the
5034 authenticated id is a complete email address when it generates From: or
5035 Sender: header lines.
5037 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5038 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5039 chown and chgrp in /bin and hostname in /usr/bin.
5041 28. Exim was keeping the "process log" file open after each use, just as it
5042 does for the main log. This opens the possibility of it remaining open for
5043 long periods when the USR1 signal hits a daemon. Occasional processlog
5044 errors were reported, that could have been caused by this. Anyway, it seems
5045 much more sensible not to leave this file open at all, so that is what now
5048 29. The long-running daemon process does not normally write to the log once it
5049 has entered its main loop, and it closes the log before doing so. This is
5050 so that log files can straightforwardly be renamed and moved. However,
5051 there are a couple of unusual error situations where the daemon does write
5052 log entries, and I had neglected to close the log afterwards.
5054 30. The text of an SMTP error response that was received during a remote
5055 delivery was being truncated at 512 bytes. This is too short for some of
5056 the long messages that one sometimes sees. I've increased the limit to
5059 31. It is now possible to make retry rules that apply only when a message has a
5060 specific sender, in particular, an empty sender.
5062 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5063 it possible to be selective about when SMTP synchronization is enforced.
5065 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5067 32. Implemented hosts_connection_nolog.
5069 33. Added an ACL for QUIT.
5071 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5074 35. Added mailbox_size and mailbox_filecount to appendfile.
5076 36. Added control = no_multiline_responses to ACLs.
5078 37. There was a bug in the logic of the code that waits for the clock to tick
5079 in the case where the clock went backwards by a substantial amount such
5080 that the microsecond fraction of "now" was more than the microsecond
5081 fraction of "then" (but the whole seconds number was less).
5083 38. Added support for the libradius Radius client library this is found on
5084 FreeBSD (previously only the radiusclient library was supported).
5090 1. When certain lookups returned multiple values in the form name=value, the
5091 quoting of the values was not always being done properly. Specifically:
5092 (a) If the value started with a double quote, but contained no whitespace,
5094 (b) If the value contained whitespace other than a space character (i.e.
5095 tabs or newlines or carriage returns) it was not quoted.
5096 This fix has been applied to the mysql and pgsql lookups by writing a
5097 separate quoting function and calling it from the lookup code. The fix
5098 should probably also be applied to nisplus, ibase and oracle lookups, but
5099 since I cannot test any of those, I have not disturbed their existing code.
5101 2. A hit in the callout cache for a specific address caused a log line with no
5102 reason for rejecting RCPT. Now it says "Previous (cached) callout
5103 verification failure".
5105 3. There was an off-by-one bug in the queryprogram router. An over-long
5106 return line was truncated at 256 instead of 255 characters, thereby
5107 overflowing its buffer with the terminating zero. As well as fixing this, I
5108 have increased the buffer size to 1024 (and made a note to document this).
5110 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5111 when Exim is waiting for an SMTP response from a remote server, Exim
5112 restarts its select() call on the socket, thereby resetting its timeout.
5113 This is not a problem when such interrupts are rare. Somebody set up a cron
5114 job to run exiwhat every 2 minutes, which is less than the normal select()
5115 timeout (5 or 10 minutes). This meant that the select() timeout never
5116 kicked in because it was always reset. I have fixed this by comparing the
5117 time when an interrupt arrives with the time at the start of the first call
5118 to select(). If more time than the timeout has elapsed, the interrupt is
5119 treated as a timeout.
5121 5. Some internal re-factoring in preparation for the addition of Sieve
5122 extensions (by MH). In particular, the "personal" test is moved to a
5123 separate function, and given an option for scanning Cc: and Bcc: (which is
5124 not set for Exim filters).
5126 6. When Exim created an email address using the login of the caller as the
5127 local part (e.g. when creating a From: or Sender: header line), it was not
5128 quoting the local part when it contained special characters such as @.
5130 7. Installed new OpenBSD configuration files.
5132 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5133 try to make them clearer.
5135 9. Callout options, other than the timeout value, were being ignored when
5136 verifying sender addresses in header lines. For example, when using
5138 verify = header_sender/callout=no_cache
5140 the cache was (incorrectly) being used.
5142 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5143 only the Cygwin environment.
5145 11. When return_path_on_delivery was set as a log selector, if different remote
5146 addresses in the same message used different return paths and parallel
5147 remote delivery occurred, the wrong values would sometimes be logged.
5148 (Whenever a remote delivery process finished, the return path value from
5149 the most recently started remote delivery process was logged.)
5151 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5152 lines when AUTH and/or TLS are in use. This is the "received protocol"
5153 field. Exim used to use "asmtp" for authenticated SMTP, without any
5154 indication (in the protocol name) for TLS use. Now it follows the RFC and
5155 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5156 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5157 names appear in log lines as well as in Received: header lines.
5159 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5160 extensions, and comparison tests, and to fix some bugs.
5162 14. Changes to the "personal" filter test:
5164 (1) The test was buggy in that it was just doing the equivalent of
5165 "contains" tests on header lines. For example, if a user's address was
5166 anne@some.where, the "personal" test would incorrectly be true for
5168 To: susanne@some.where
5170 This test is now done by extracting each address from the header in turn,
5171 and checking the entire address. Other tests that are part of "personal"
5172 are now done using regular expressions (for example, to check local parts
5173 of addresses in From: header lines).
5175 (2) The list of non-personal local parts in From: addresses has been
5176 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5177 taken from the Sieve specification recommendations.
5179 (3) If the message contains any header line starting with "List-" it is
5180 treated as non-personal.
5182 (4) The test for "circular" in the Subject: header line has been removed
5183 because it now seems ill-conceived.
5185 15. Minor typos in src/EDITME comments corrected.
5187 16. Installed latest exipick from John Jetmore.
5189 17. If headers_add on a router specified a text string that was too long for
5190 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5191 of string_sprintf() is now avoided.
5193 18. $message_body_size was not set (it was always zero) when running the DATA
5194 ACL and the local_scan() function.
5196 19. For the "mail" command in an Exim filter, no default was being set for
5197 the once_repeat time, causing a random time value to be used if "once" was
5198 specified. (If the value happened to be <= 0, no repeat happened.) The
5199 default is now 0s, meaning "never repeat". The "vacation" command was OK
5200 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5201 (I found it when inspecting the code).
5203 20. There is now an overall timeout for performing a callout verification. It
5204 defaults to 4 times the callout timeout, which applies to individual SMTP
5205 commands during the callout. The overall timeout applies when there is more
5206 than one host that can be tried. The timeout is checked before trying the
5207 next host. This prevents very long delays if there are a large number of
5208 hosts and all are timing out (e.g. when the network connections are timing
5209 out). The value of the overall timeout can be changed by specifying an
5210 additional sub-option for "callout", called "maxwait". For example:
5212 verify = sender/callout=5s,maxwait=20s
5214 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5215 to the end before writing, but this should make it even safer).
5217 22. Exim was forgetting that it had advertised PIPELINING for the second and
5218 subsequent messages on an SMTP connection. It was also not resetting its
5219 memory on STARTTLS and an internal HELO.
5221 23. When Exim logs an SMTP synchronization error within a session, it now
5222 records whether PIPELINING has been advertised or not.
5224 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5225 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5226 rather than long int.
5228 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5230 26. Added the never_mail option to autoreply.
5236 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5237 crash if the getsockname() call failed; this can happen if a connection is
5238 closed very soon after it is established. The problem was simply in the
5239 order in which certain operations were done, causing Exim to try to write
5240 to the SMTP stream before it had set up the file descriptor. The bug has
5241 been fixed by making things happen in the correct order.
5247 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5248 before the connection was closed, thus losing the rejection response.
5250 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5251 some early Solaris releases, but causes trouble in current releases where
5252 socklen_t is defined.
5254 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5257 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5260 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5261 that releasing the top of it at the end releases what was used for sub-
5262 expansions (unless the block got too big). However, discard this block if
5263 the first thing is a variable or header, so that we can use its block when
5264 it is dynamic (useful for very large $message_headers, for example).
5266 6. Lookups now cache *every* query, not just the most recent. A new, separate
5267 store pool is used for this. It can be recovered when all lookup caches are
5268 flushed. Lookups now release memory at the end of their result strings.
5269 This has involved some general refactoring of the lookup sources.
5271 7. Some code has been added to the store_xxx() functions to reduce the amount
5272 of flapping under certain conditions.
5274 8. log_incoming_interface used to affect only the <= reception log lines. Now
5275 it causes the local interface and port to be added to several more SMTP log
5276 lines, for example "SMTP connection from", and rejection lines.
5278 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5280 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5282 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5283 could be overwritten at the end of the current message (or the start of a
5284 new message if it was set in a HELO ACL). The value is now preserved for
5285 the duration of the SMTP connection.
5287 12. If a transport had a headers_rewrite setting, and a matching header line
5288 contained an unqualified address, that address was qualified, even if it
5289 did not match any rewriting rules. The underlying bug was that the values
5290 of the flags that permit the existence of unqualified sender and recipient
5291 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5292 non-local messages, and by -bnq for local messages) were not being
5293 preserved with the message after it was received.
5295 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5296 "next input=" as part of the text comprising the host identity instead of
5297 the correct text. The code was using the same buffer for two different
5298 strings. However, depending on which order the printing function evaluated
5299 its arguments, the bug did not always show up. Under Linux, for example, my
5300 test suite worked just fine.
5302 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5303 doesn't work with some older versions of Perl. It has been changed to "my",
5304 which in any case is probably the better facility to use.
5306 15. A really picky compiler found some instances of statements for creating
5307 error messages that either had too many or two few arguments for the format
5310 16. The size of the buffer for calls to the DNS resolver has been increased
5311 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5312 for addresses that have a lot of PTR records. This alleviates a problem; it
5313 does not fully solve it.
5315 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5316 buffer now truncates the list and logs the incident, which is the same
5317 action as happens when Exim is looking up a host name and its aliases.
5318 Previously in this situation something unpredictable would happen;
5319 sometimes it was "internal error: store_reset failed".
5321 18. If a server dropped the connection unexpectedly when an Exim client was
5322 using GnuTLS and trying to read a response, the client delivery process
5323 crashed while trying to generate an error log message.
5325 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5326 string, for example:
5328 warn message = H1: something\nH2: something
5330 the text was added as a single header line from Exim's point of view
5331 though it ended up OK in the delivered message. However, searching for the
5332 second and subsequent header lines using $h_h2: did not work. This has been
5333 fixed. Similarly, if a system filter added multiple headers in this way,
5334 the routers could not see them.
5336 20. Expanded the error message when iplsearch is called with an invalid key to
5337 suggest using net-iplsearch in a host list.
5339 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5340 ACLs are no longer actually imposed (and a message to that effect is
5343 22. If a "gecos" field in a passwd entry contained escaped characters, in
5344 particular, if it contained a \" sequence, Exim got it wrong when building
5345 a From: or a Sender: header from that name. A second bug also caused
5346 incorrect handling when an unquoted " was present following a character
5347 that needed quoting.
5349 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5350 was not being matched caselessly.
5352 24. Arranged for all hyphens in the exim.8 source to be escaped with
5355 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5356 properly. Recipient callout cache records were still being keyed to include
5357 the sender, even when use_sender was set false. This led to far more
5358 callouts that were necessary. The sender is no longer included in the key
5359 when use_sender is false.
5361 26. Added "control = submission" modifier to ACLs.
5363 27. Added the ${base62d: operator to decode base 62 numbers.
5365 28. dnsdb lookups can now access SRV records.
5367 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5368 the configuration file.
5370 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5371 (-v) mode. This makes the output for a verbose queue run more intelligible.
5373 31. Added a use_postmaster feature to recipient callouts.
5375 32. Added the $body_zerocount variable, containing the number of binary zero
5376 bytes in the message body.
5378 33. The time of last modification of the "new" subdirectory is now used as the
5379 "mailbox time last read" when there is a quota error for a maildir
5382 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5384 35. Added +ignore_unknown as a special item in host lists.
5386 36. Code for decoding IPv6 addresses in host lists is now included, even if
5387 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5388 address was recognized as an IP address, but was then not correctly decoded
5389 into binary, causing unexpected and incorrect effects when compared with
5396 1. Very minor rewording of debugging text in manualroute to say "list of
5397 hosts" instead of "hostlist".
5399 2. If verify=header_syntax was set, and a header line with an unqualified
5400 address (no domain) and a large number of spaces between the end of the
5401 name and the colon was received, the reception process suffered a buffer
5402 overflow, and (when I tested it) crashed. This was caused by some obsolete
5403 code that should have been removed. The fix is to remove it!
5405 3. When running in the test harness, delay a bit after writing a bounce
5406 message to get a bit more predictability in the log output.
5408 4. Added a call to search_tidyup() just before forking a reception process. In
5409 theory, someone could use a lookup in the expansion of smtp_accept_max_
5410 per_host which, without the tidyup, could leave open a database connection.
5412 5. Added the variables $recipient_data and $sender_data which get set from a
5413 lookup success in an ACL "recipients" or "senders" condition, or a router
5414 "senders" option, similar to $domain_data and $local_part_data.
5416 6. Moved the writing of debug_print from before to after the "senders" test
5419 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5420 problems for message scanning, either using a data ACL, or using
5421 local_scan() because the Received: header was not generated till after they
5422 were called (in order to set the time as the time of reception completion).
5423 I have revised the way this works. The header is now generated after the
5424 body is received, but before the ACL or local_scan() are called. After they
5425 are run, the timestamp in the header is updated.
5431 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5432 before starting a queue runner without re-exec. This happened only when
5433 deliver_drop_privilege was set or when the Exim user was set to root. The
5434 effect of the bug was that timeouts during subsequent deliveries caused
5435 crashes instead of being properly handled. The handler is now left at its
5436 default (and expected) setting.
5438 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5439 message, again when deliver_drop_privilege is set or Exim is run as root.
5440 The bug described in (1) was not present in this case, but the tidying up
5441 of the other signals was missing. I have made the two cases consistent.
5443 3. The ignore_target_hosts setting on a manualroute router was being ignored
5444 for hosts that were looked up using the /MX notation.
5446 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5449 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5450 operated on the sender address. After changing the $sender_address to <>
5451 for the sender address verify, Exim was re-instated it as the original
5452 (before rewriting) address, but remembering that it had rewritten it, so it
5453 wasn't rewriting it again. This bug also had the effect of breaking the
5454 sender address verification caching when the sender address was rewritten.
5456 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5457 This has been changed so that if the ip literal address matches
5458 ignore_target_hosts, the router declines.
5460 7. Added expansion conditions match_domain, match_address, and match_local_
5461 part (NOT match_host).
5463 8. The placeholder for the Received: header didn't have a length field set.
5465 9. Added code to Exim itself and to exim_lock to test for a specific race
5466 condition that could lead to file corruption when using MBX delivery. The
5467 issue is with the lockfile that is created in /tmp. If this file is removed
5468 after a process has opened it but before that process has acquired a lock,
5469 there is the potential for a second process to recreate the file and also
5470 acquire a lock. This could lead to two Exim processes writing to the file
5471 at the same time. The added code performs the same test as UW imapd; it
5472 checks after acquiring the lock that its file descriptor still refers to
5473 the same named file.
5475 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5476 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5479 11. Added the smtp_active_hostname option. If used, this will typically be made
5480 to depend on the incoming interface address. Because $interface_address is
5481 not set up until the daemon has forked a reception process, error responses
5482 that can happen earlier (such as "too many connections") no longer contain
5485 12. If an expansion in a condition on a "warn" statement fails because a lookup
5486 defers, the "warn" statement is abandoned, and the next ACL statement is
5487 processed. Previously this caused the whole ACL to be aborted.
5489 13. Added the iplsearch lookup type.
5491 14. Added ident_timeout as a log selector.
5493 15. Added tls_certificate_verified as a log selector.
5495 16. Added a global option tls_require_ciphers (compare the smtp transport
5496 option of the same name). This controls incoming TLS connections.
5498 17. I finally figured out how to make tls_require_ciphers do a similar thing
5499 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5500 before starting the TLS session.
5502 18. Tabs are now shown as \t in -bP output.
5504 19. If the log selector return_path_on_delivery was set, Exim crashed when
5505 bouncing a message because it had too many Received: header lines.
5507 20. If two routers both had headers_remove settings, and the first one included
5508 a superfluous trailing colon, the final name in the first list and the
5509 first name in the second list were incorrectly joined into one item (with a
5510 colon in the middle).
5516 1. Added -C and -D options to the exinext utility, mainly to make it easier
5517 to include in the automated testing, but these could be helpful when
5518 multiple configurations are in use.
5520 2. The exinext utility was not formatting the output nicely when there was
5521 an alternate port involved in the retry record key, nor when there was a
5522 message id as well (for retries that were specific to a specific message
5523 and a specific host). It was also confused by IPv6 addresses, because of
5524 the additional colons they contain. I have fixed the IPv4 problem, and
5525 patched it up to do a reasonable job for IPv6.
5527 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5528 delivery, the log line now contains "pipelined" if PIPELINING was used.
5530 4. An SMTP transport process used to panic and die if the bind() call to set
5531 an explicit outgoing interface failed. This has been changed; it is now
5532 treated in the same way as a connect() failure.
5534 5. A reference to $sender_host_name in the part of a conditional expansion
5535 that was being skipped was still causing a DNS lookup. This no longer
5538 6. The def: expansion condition was not recognizing references to header lines
5539 that used bh_ and bheader_.
5541 7. Added the _cache feature to named lists.
5543 8. The code for checking quota_filecount in the appendfile transport was
5544 allowing one more file than it should have been.
5546 9. For compatibility with Sendmail, the command line option
5554 and sets the incoming protocol and host name (for trusted callers). The
5555 host name and its colon can be omitted when only the protocol is to be set.
5556 Note the Exim already has two private options, -pd and -ps, that refer to
5557 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5558 "s", but I don't think that's a major issue.
5560 10. A number of refactoring changes to the code, none of which should affect
5563 (a) The number of logging options was getting close to filling up the
5564 32-bit word that was used as a bit map. I have split them into two classes:
5565 those that are passed in the argument to log_write(), and those that are
5566 only ever tested independently outside of that function. These are now in
5567 separate 32-bit words, so there is plenty of room for expansion again.
5568 There is no change in the user interface or the logging behaviour.
5570 (b) When building, for example, log lines, the code previously used a
5571 macro that called string_cat() twice, in order to add two strings. This is
5572 not really sufficiently general. Furthermore, there was one instance where
5573 it was actually wrong because one of the argument was used twice, and in
5574 one call a function was used. (As it happened, calling the function twice
5575 did not affect the overall behaviour.) The macro has been replaced by a
5576 function that can join an arbitrary number of extra strings onto a growing
5579 (c) The code for expansion conditions now uses a table and a binary chop
5580 instead of a serial search (which was left over from when there were very
5581 few conditions). Also, it now recognizes conditions like "pam" even when
5582 the relevant support is not compiled in: a suitably worded error message is
5583 given if an attempt is made to use such a condition.
5585 11. Added ${time_interval:xxxxx}.
5587 12. A bug was causing one of the ddress fields not to be passed back correctly
5588 from remote delivery subprocesses. The field in question was not being
5589 subsequently used, so this caused to problems in practice.
5591 13. Added new log selectors queue_time and deliver_time.
5593 14. Might have fixed a bug in maildirsizefile handling that threw up
5594 "unexpected character" debug warnings, and recalculated the data
5595 unnecessarily. In any case, I expanded the warning message to give more
5598 15. Added the message "Restricted characters in address" to the statements in
5599 the default ACL that block characters like @ and % in local parts.
5601 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5602 Three changes have been made:
5604 (a) There was a serious bug; a negative response to MAIL caused the whole
5605 recipient domain to be cached as invalid, thereby blocking all messages
5606 to all local parts at the same domain, from all senders. This bug has
5607 been fixed. The domain is no longer cached after a negative response to
5608 MAIL if the sender used is not empty.
5610 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5613 (c) A new callout option, "use_sender" has been added for people who want
5614 the modified behaviour.
5620 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5623 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5624 indeed breaks things for older releases.
5626 3. Added additional logging to the case where there is a problem reading data
5627 from a filter that is running in a subprocess using a pipe, in order to
5628 try to track down a specific problem.
5630 4. Testing facility fudge: when running in the test harness and attempting
5631 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5632 getting "No route to host". Convert this to a timeout.
5634 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5637 6. Some OS don't have socklen_t but use size_t instead. This affects the
5638 fifth argument of getsockopt() amongst other things. This is now
5639 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5640 can be set for individual OS. I have set it for SunOS5, OSF1, and
5641 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5642 some earlier ones do not.
5644 7. Change 4.30/15 was not doing the test caselessly.
5646 8. The standard form for an IPv6 address literal was being rejected by address
5647 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5648 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5649 this, as well as the form without the "IPv6" on the front (but only when
5650 address literals are enabled, of course).
5652 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5654 10. Exim crashed if a message with an empty sender address specified by -f
5655 encountered a router with an errors_to setting. This could be provoked only
5656 by a command such as
5660 where an empty string was supplied; "<>" did not hit this bug.
5662 11. Installed PCRE release 4.5.
5664 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5665 remained set. It is now erased.
5667 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5668 times from message ids (which are base 36 rather than the normal 62).
5670 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5671 were being counted as actual protocol errors, and logged if the log
5672 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5673 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5674 and DATA following a set of rejected RCPTs do not count as protocol errors.
5675 In other words, Exim assumes they were pipelined, though this may not
5676 actually be the case. Of course, in all cases the client gets an
5677 appropriate error code.
5679 15. If a lookup fails in an ACL condition, a message about the failure may
5680 be available; it is used if testing the ACL cannot continue, because most
5681 such messages specify what the cause of the deferral is. However, some
5682 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5683 that caused an old message to be retained and used if a later statement
5684 caused a defer, replacing the real cause of the deferral.
5686 16. If an IP address had so many PTR records that the DNS lookup buffer
5687 was not large enough to hold them, Exim could crash while trying to process
5688 the truncated data. It now detects and logs this case.
5690 17. Further to 4.21/58, another change has been made: if (and only if) the
5691 first line of a message (the first header line) ends with CRLF, a bare LF
5692 in a subsequent header line has a space inserted after it, so as not to
5693 terminate the header.
5695 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5696 unnecessarily, used atoi() instead of strtol(), and didn't check the
5697 termination when getting file sizes from file names by regex.
5699 19. Completely re-implemented the support for maildirsize files, in the light
5700 of a number of problems with the previous contributed implementation
5701 (4.30/29). In particular:
5703 . If the quota is zero, the maildirsize file is maintained, but no quota is
5706 . If the maildir directory does not exist, it is created before any attempt
5707 to write a maildirsize file.
5709 . The quota value in the file is just a cache; if the quota is changed in
5710 the transport, the new value overrides.
5712 . A regular expression is available for excluding directories from the
5715 20. The autoreply transport checks the characters in options that define the
5716 message's headers; it allows continued headers, but it was checking with
5717 isspace() after an embedded newline instead of explicitly looking for a
5720 21. If all the "regular" hosts to which an address was routed had passed their
5721 expiry times, and had not reached their retry times, the address was
5722 bounced, even if fallback hosts were defined. Now Exim should go on to try
5725 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5726 equivalent code in the SMTP transport. Some hosts send humungous responses
5727 to HELO/EHLO, more than 1024 it seems.
5729 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5730 gives compiler warnings in some environments. I've now done it "properly",
5733 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5734 (because of problems with the built-in one) was declared to return uschar *
5735 instead of char *, causing compiler failure.
5737 25. Fixed a file descriptor leak when processing alias/forward files.
5739 26. Fixed a minor format string issue in dbfn.c.
5741 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5743 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5744 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5747 29. When a sender address is verified, it is cached, to save repeating the test
5748 when there is more than one recipient in a message. However, when the
5749 verification involves a callout, it is possible for different callout
5750 options to be set for different recipients. It is too complicated to keep
5751 track of this in the cache, so now Exim always runs a verification when a
5752 callout is required, relying on the callout cache for the optimization.
5753 The overhead is duplication of the address routing, but this should not be
5756 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5757 to be verified with callout=postmaster, and the main callout worked but the
5758 postmaster check failed, the verification correctly failed. However, if a
5759 subsequent RCPT command asked for sender verification *without* the
5760 postmaster check, incorrect caching caused this verification also to fail,
5763 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5764 it was not caching the DNS options (qualify_single, search_parents) that
5765 were used when the lookup failed. A subsequent lookup with different
5766 options therefore always gave the same answer, though there were cases
5767 where it should not have. (Example: a "domains = !$mx_any" option on a
5768 dnslookup router: the "domains" option is always processed without any
5769 widening, but the router might have qualify_single set.) Now Exim uses the
5770 cached value only when the same options are set.
5772 32. Added John Jetmore's "exipick" utility to the distribution.
5774 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5775 than a timeout (e.g. a certificate is required, and is not provided), an
5776 Exim server now closes the connection immediately. Previously it waited for
5777 the client to close - but if the client is SSL, it seems that they each
5778 wait for each other, leading to a delay before one of them times out.
5780 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5781 maintained 0.8.x compatibility because I don't think many are using it, and
5782 it is clearly obsolete.
5784 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5787 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5788 client certificate was expired. A simple patch fixes this, though I don't
5789 understand the full logic of why the verify callback is called multiple
5792 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5793 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5794 which causes problems with some clients (such as the Certicom SSL Plus
5795 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5796 disables the coutermeasure allowing Eudora to connect."
5798 38. Exim was not checking that a write() to a log file succeeded. This could
5799 lead to Bad Things if a log got too big, in particular if it hit a file
5800 size limit. Exim now panics and dies if it cannot write to a log file, just
5801 as it does if it cannot open a log file.
5803 39. Modified OS/Makefile-Linux so that it now contains
5805 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5807 The two -D definitions ensure that Exim is compiled with large file
5808 support, which makes it possible to handle log files that are bigger than
5811 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5812 instance) a domain was checked against a named list that involved a lookup,
5813 causing $domain_data to be set, then another domain was checked against the
5814 same list, then the first domain was re-checked, the value of $domain_data
5815 after the final check could be wrong. In particular, if the second check
5816 failed, it could be set empty. This bug probably also applied to
5819 41. The strip_trailing_dot option was not being applied to the address given
5820 with the -f command-line option.
5822 42. The code for reading a message's header from the spool was incrementing
5823 $received_count, but never initializing it. This meant that the value was
5824 incorrect (doubled) while delivering a message in the same process in which
5825 it was received. In the most common configuration of Exim, this never
5826 happens - a fresh exec is done - but it can happen when
5827 deliver_drop_privilege is set.
5829 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5830 it now includes up to 150 characters of the unexpected data in the log
5833 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5834 and building data strings. The size of both of these buffers was 10 000
5835 bytes - far larger than anybody would *ever* want, thought I. Needless to
5836 say, somebody hit the limit. I have increased the maximum line length to
5837 20 000 and the maximum data length of concatenated lines to 100 000. I have
5838 also fixed two bugs, because there was no checking on these buffers. Tsk,
5839 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5840 buffer is too small.
5842 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5843 lsearch lookups. Now it does.
5845 46. When parsing a route_list item in a manualroute router, a fixed-length
5846 buffer was used for the list of hosts. I made this 1024 bytes long,
5847 thinking that nobody would ever have a list of hosts that long. Wrong.
5848 Somebody had a whole pile of complicated expansion conditions, and the
5849 string was silently truncated, leading to an expansion error. It turns out
5850 that it is easier to change to an unlimited length (owing to other changes
5851 that have happened since this code was originally written) than to build
5852 structure for giving a limitation error. The length of the item that
5853 expands into the list of hosts is now unlimited.
5855 47. The lsearch lookup could not handle data where the length of text line was
5856 more than 4095 characters. Such lines were truncated, leading to shortened
5857 data being returned. It should now handle lines of any length.
5859 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5860 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5863 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5864 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5865 no longer happens in comment lines. A list of replacements is now placed
5866 at the head of all of the source files, except those whose only change is
5867 to replace PERL_COMMAND in the very first #! line.
5869 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5870 messages on the queue, with a bottom-up merge sort, using code contributed
5871 by Michael Haardt. This should make operations like -bp somewhat faster on
5872 large queues. It won't affect queue runners, except when queue_run_in_order
5875 51. Installed eximstats 1.31 in the distribution.
5877 52. Added support for SRV lookups to the dnslookup router.
5879 53. If an ACL referred to $message_body or $message_body_end, the value was not
5880 reset for any messages that followed in the same SMTP session.
5882 54. The store-handling optimization for building very long strings was not
5883 differentiating between the different store pools. I don't think this
5884 actually made any difference in practice, but I've tidied it.
5886 55. While running the routers to verify a sender address, $sender_address
5887 was still set to the sender address. This is wrong, because when routing to
5888 send a bounce to the sender, it would be empty. Therefore, I have changed
5889 it so that, while verifying a sender address, $sender_address is set to <>.
5890 (There is no change to what happens when verifying a recipient address.)
5892 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5893 target A or AAAA records (if not already returned) without resetting the
5894 qualify_single or search_parents options of the DNS resolver. These are
5895 inappropriate in this case because the targets of MX and SRV records must
5896 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5897 target that, when qualified, matched something in the local domain. These
5898 two options are now turned off when doing these lookups.
5900 57. It seems that at least some releases of Reiserfs (which does not have the
5901 concept of a fixed number of inodes) returns zero and not -1 for the
5902 number of available inodes. This interacted badly with check_spool_inodes,
5903 which assumed that -1 was the "no such thing" setting. What I have done is
5904 to check that the total number of inodes is greater than zero before doing
5905 the test of how many are available.
5907 58. When a "warn" ACL statement has a log_message modifier, the message is
5908 remembered, and not repeated. This is to avoid a lot of repetition when a
5909 message has many recipients that cause the same warning to be written.
5910 However, Exim was preserving the list of already written lines for an
5911 entire SMTP session, which doesn't seem right. The memory is now reset if a
5912 new message is started.
5914 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5915 debugging output unless log_rewrite was also set.
5917 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5918 of (int)(handle) when we know that handle contains (void *)(-1).
5920 61. The Exim daemon panic-logs an error return when it closes the incoming
5921 connection. However "connection reset by peer" seems to be common, and
5922 isn't really an error worthy of noting specially, so that particular error
5925 62. When Exim is trying to find all the local interfaces, it used to panic and
5926 die if the ioctl to get the interface flags failed. However, it seems that
5927 on at least one OS (Solaris 9) it is possible to have an interface that is
5928 included in the list of interfaces, but for which you get a failure error
5929 for this call. This happens when the interface is not "plumbed" into a
5930 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5931 failure of the "get flags" call assumes that the interface is down.
5933 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5934 makes life easier for people who are doing arithmetic on fields extracted
5935 from dates, where you often get leading zeros that should not be
5936 interpreted as octal.
5938 64. Added qualify_domain to the redirect router, to override the global
5941 65. If a pathologically long header line contained very many addresses (the
5942 report of this problem mentioned 10 000) and each of them was rewritten,
5943 Exim could use up a very large amount of memory. (It kept on making new
5944 copies of the header line as it rewrote, and never released the old ones.)
5945 At the expense of a bit more processing, the header rewriting function has
5946 been changed so that it no longer eats memory in this way.
5948 66. The generation of the Received: header has been moved from the time that a
5949 message starts to be received, to the time that it finishes. The timestamp
5950 in the Received: header should now be very close to that of the <= log
5951 line. There are two side-effects of this change:
5953 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5954 logged header lines no longer include the local Received: line, because
5955 it has not yet been created. The same applies to a copy of the message
5956 that is returned to a non-SMTP sender when a message is rejected.
5958 (b) When a filter file is tested using -bf, no additional Received: header
5959 is added to the test message. After some thought, I decided that this
5962 This change does not affect the value of $received_for. It is still set
5963 after address rewriting, but before local_scan() is called.
5965 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5967 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5968 gave an unhelpful panic error message, and a defer error. I have managed to
5969 change this behaviour so that it now rejects any supplied certificate,
5970 which seems right, as the list of acceptable certificates is empty.
5972 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5973 gave an unhelpful defer error. I have not managed to make this reject any
5974 supplied certificates, but the error message it gives is "no certificate
5975 supplied", which is not helpful.
5977 70. exigrep's output now also includes lines that are not associated with any
5978 message, but which match the given pattern. Implemented by a patch from
5979 Martin Sluka, which also tidied up the Perl a bit.
5981 71. Recipient callout verification, like sender verification, was using <> in
5982 the MAIL FROM command. This isn't really the right thing, since the actual
5983 sender may affect whether the remote host accepts the recipient or not. I
5984 have changed it to use the actual sender in the callout; this means that
5985 the cache record is now keyed on a recipient/sender pair, not just the
5986 recipient address. There doesn't seem to be a real danger of callout loops,
5987 since a callout by the remote host to check the sender would use <>.
5988 [SEE ABOVE: changed after hitting problems.]
5990 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5991 temporary errors. However, in the case of such a code being given after
5992 the end of a data transmission (i.e. after ".") Exim was failing to write
5993 a retry record for the message. (Yes, there was some broken host that was
5994 actually sending 8xx at this point.)
5996 73. An unknown lookup type in a host list could cause Exim to panic-die when
5997 the list was checked. (An example that provoked this was putting <; in the
5998 middle of a list instead of at the start.) If this happened during a DATA
5999 ACL check, a -D file could be left lying around. This kind of configuration
6000 error no longer causes Exim to die; instead it causes a defer error. The
6001 incident is still logged to the main and panic logs.
6003 74. Buglet left over from Exim 3 conversion. The message "too many messages
6004 in one connection" was written to the rejectlog but not the mainlog, except
6005 when address rewriting (yes!) was being logged.
6007 75. Added write_rejectlog option.
6009 76. When a system filter was run not as root (that is, when system_filter_user
6010 was set), the values of the $n variables were not being returned to the
6011 main process; thus, they were not subsequently available in the $sn
6014 77. Added +return_path_on_delivery log selector.
6016 78. A connection timeout was being treated differently from recipients deferred
6017 when testing hosts_max_try with a message that was older than the host's
6018 retry timeout. (The host should not be counted, thus allowing all hosts to
6019 be tried at least once before bouncing.) This may have been the cause of an
6020 occasionally reported bug whereby a message would remain on the queue
6021 longer than the retry timeout, but would be bounced if a delivery was
6022 forced. I say "may" because I never totally pinned down the problem;
6023 setting up timeout/retry tests is difficult. See also the next item.
6025 79. The ultimate address timeout was not being applied to errors that involved
6026 a combination of host plus message (for example, a timeout on a MAIL
6027 command). When an address resolved to a number of possible hosts, and they
6028 were not all tried for each delivery (e.g. because of hosts_max_try), a
6029 message could remain on the queue longer than the retry timeout.
6031 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6034 81. Fixed an obscure SMTP outgoing bug which required at least the following
6035 conditions: (a) there was another message waiting for the same server;
6036 (b) the server returned 5xx to all RCPT commands in the first message so
6037 that the message was not completed; (c) the server dropped the connection
6038 or gave a negative response to the RSET that Exim sends to abort the
6039 transaction. The observed case was a dropped connection after DATA that had
6040 been sent in pipelining mode. That is, the server had advertised PIPELINING
6041 but was not implementing it correctly. The effect of the bug was incorrect
6042 behaviour, such as trying another host, and this could lead to a crash.
6048 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6049 and daemon.c were passed as pointers to ints; they should have been
6050 pointers to socklen_t variables (which are typically unsigned ints).
6052 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6055 3. Fixed a really odd bug that affected only the testing scheme; patching a
6056 certain fixed string in the binary changed the value of another string that
6057 happened to be identical to the end of the original first string.
6059 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6060 name", it returns that address as the IP address. On some operating
6061 systems (e.g. Solaris), it also passes back the IP address string as the
6062 "host name". However, on others (e.g. Linux), it passes back an empty
6063 string. Exim wasn't checking for this, and was changing the host name to an
6064 empty string, assuming it had been canonicalized.
6066 5. Although rare, it is permitted to have more than one PTR record for a given
6067 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6068 all the names associated with an address, because they do in Solaris.
6069 However, it seems that they do not in Linux for data that comes from the
6070 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6071 I found this out when I moved to a new Linux workstation and tried to run
6072 the Exim test suite.
6074 To get round this problem I have changed the code so that it now does its
6075 own call to the DNS to look up PTR records when searching for a host name.
6076 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6077 addresses that are only in /etc/hosts are still found.
6079 This behaviour is, however, controlled by an option called host_lookup_
6080 order, which defaults to "bydns:byaddr". If people want to use the other
6081 order, or indeed, just use one or the other means of lookup, they can
6082 specify it in this variable.
6084 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6085 some operating systems, this comes back from gethostbyaddr() as an empty
6086 string, and this is what Exim used to test for. However, it seems that in
6087 other systems, "." is yielded. Exim now tests for this case too.
6089 7. The values of check_spool_space and check_log_space are now held internally
6090 as a number of kilobytes instead of an absolute number of bytes. If a
6091 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6092 kilobyte. This means that much larger values can be stored.
6094 8. Exim monitor: an attempt to get the action menu when not actually pointing
6095 at a message produces an empty menu entitled "No message selected". This
6096 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6097 no entries in it ("Shell widget menu has zero width and/or height"). So I
6098 have added a single, blank menu entry in this case.
6100 9. Added ${quote_local_part.
6102 10. MIME decoding is now applied to the contents of Subject: header lines when
6105 11. Now that a reference to $sender_host_address automatically causes a reverse
6106 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6107 host lookup before query-style lookups in lists that might use this
6108 variable. This has therefore been abolished, and the "net-" prefix is no
6109 longer necessary for query-style lookups.
6111 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6112 have been a typo for LFLAGS, so it has been changed.
6114 13. The install script calls Exim with "-C /dev/null" in order to find the
6115 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6116 to be output. However, since Exim outputs its version number before the
6117 error, it didn't break the script. It just looked ugly. I fixed this by
6118 always allowing "-C /dev/null" if the caller is root.
6120 14. Ignore overlarge ACL variable number when reading spool file - insurance
6121 against a later release with more variables having written the file.
6123 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6124 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6125 this, as well as the form without the "IPv6" on the front.
6127 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6128 OS/Makefile-Darwin file.
6130 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6131 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6133 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6134 variable. However, some OS use ints for inodes. I've added cast to long int
6135 to get rid of the compiler warning.
6137 19. I had forgotten to lock out "/../" in configuration file names when
6138 ALT_CONFIG_PREFIX was set.
6140 20. Routers used for verification do not need to specify transports. However,
6141 if such a router generated a host list, and callout was configured, Exim
6142 crashed, because it could not find a port number from the (non-existent)
6143 transport. It now assumes port 25 in this circumstance.
6145 21. Added the -t option to exigrep.
6147 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6148 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6149 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6150 them is set, LOOKUP_LSEARCH is forced.
6152 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6154 24. Added sender and host information to the "rejected by local_scan()" log
6155 line; previously there was no indication of these.
6157 25. Added .include_if_exists.
6159 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6160 Linux. It turns out that not all file systems support this. Apparently some
6161 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6162 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6163 not supported on the file descriptor, is now ignored when Exim is trying to
6164 sync a directory. This applies only to Linux.
6166 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6168 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6169 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6172 29. Added support for maildirsize files from supplied patch (modified a bit).
6174 30. The use of :fail: followed by an empty string could lead Exim to respond to
6175 sender verification failures with (e.g.):
6177 550 Verification failed for <xxx>
6178 550 Sender verify failed
6180 where the first response line was missing the '-' that indicates it is not
6181 the final line of the response.
6183 31. The loop for finding the name of the user that called Exim had a hardwired
6184 limit of 10; it now uses the value of finduser_retries, which is used for
6185 all other user lookups.
6187 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6190 33. Exim was neglecting to zero errno before one call of strtol() when
6191 expanding a string and expecting an integer value. On some systems this
6192 resulted in spurious "integer overflow" errors. Also, it was casting the
6193 result into an int without checking.
6195 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6196 did not work. The code looks as if it has *never* worked, though it appears
6197 to have been documented since at least release 1.62. I have made it work.
6199 35. The "timeout_DNS" error in retry rules, also documented since at least
6200 1.62, also never worked. As it isn't clear exactly what this means, and
6201 clearly it isn't a major issue, I have abolished the feature by treating it
6202 as "timeout", and writing a warning to the main and panic logs.
6204 36. The display of retry rules for -brt wasn't always showing the error code
6207 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6208 timeout_connect_A, timeout_connect_MX.
6210 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6211 to the empty sender.
6213 39. The daemon was not analysing the content of -oX till after it had closed
6214 stderr and disconnected from the controlling terminal. This meant that any
6215 syntax errors were only noted on the panic log, and the return code from
6216 the command was 0. By re-arranging the code a little, I've made the
6217 decoding happen first, so such errors now appear on stderr, and the return
6218 code is 1. However, the actual setting up of the sockets still happens in
6219 the disconnected process, so errors there are still only recorded on the
6222 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6223 connections (as happens on some IP stacks) was logged at start up time as
6224 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6225 it from "IPv6 and IPv4", which means that two separate sockets are being
6228 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6229 says whether AF_INET or AF_INET6 was passed as an argument.
6231 42. Exiwhat output was messed up when time zones were included in log
6234 43. Exiwhat now gives more information about the daemon's listening ports,
6235 and whether -tls-on-connect was used.
6237 44. The "port" option of the smtp transport is now expanded.
6239 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6240 silently ignored. Now an error message is written to the main and panic
6243 46. There's a new ACL modifier called "logwrite" which writes to a log file
6244 as soon as it is encountered.
6246 47. Added $local_user_uid and $local_user_gid at routing time.
6248 48. Exim crashed when trying to verify a sender address that was being
6251 49. Exim was recognizing only a space character after ".include". It now also
6252 recognizes a tab character.
6254 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6255 extracting the relevant information from the specification. The man page no
6256 longer contains scrambled data for the -d option, and I've added a section
6257 at the front about calling Exim under different names.
6259 51. Added "extra_headers" argument to the "mail" command in filter files.
6261 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6264 53. Installed eximstats 1.29.
6266 54. Added transport_filter_timeout as a generic transport option.
6268 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6269 Cc: header lines. This was required by RFC 822, but it not required by RFC
6272 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6273 incoming messages that did not have them. Now it does so only if the
6274 message originates locally, that is, if there is no associated remote host
6275 address. When Resent- header lines are present, this applies to the Resent-
6276 lines rather than the non-Resent- lines.
6278 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6279 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6281 58. Messages for configuration errors now include the name of the main
6282 configuration file - useful now that there may be more than one file in a
6283 list (.included file names were always shown).
6285 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6286 for those rare installations that do not start the daemon as root or run it
6287 setuid root. I've cut out the call to initgroups() if the daemon is not
6290 60. The Exim user and group can now be bound into the binary as text strings
6291 that are looked up at the start of Exim's processing.
6293 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6295 62. Added $mailstore_basename variable.
6297 63. Installed patch to sieve.c from Michael Haardt.
6299 64. When Exim failed to open the panic log after failing to open the main log,
6300 the original message it was trying to log was written to stderr and debug
6301 output, but if they were not available (the usual case in production), it
6302 was lost. Now it is written to syslog before the two lines that record the
6303 failures to open the logs.
6305 65. Users' Exim filters run in subprocesses under the user's uid. It is
6306 possible for a "deliver" command or an alias in a "personal" command to
6307 provoke an address rewrite. If logging of address rewriting is configured,
6308 this fails because the process is not running as root or exim. There may be
6309 a better way of dealing with this, but for the moment (because 4.30 needs
6310 to be released), I have disabled address rewrite logging when running a
6311 filter in a non-root, non-exim process.
6317 1. The buildconfig auxiliary program wasn't quoting the value set for
6318 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6319 not defined. This bug was masked in 4.22 by the effect that was fixed in
6322 2. Some messages that were rejected after a message id was allocated were
6323 shown as "incomplete" by exigrep. It no longer does this for messages that
6324 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6326 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6327 have allow_domain_literals set, the ID did not get logged in the <= line.
6328 Domain literals are now always recognized in Message-ID: header lines.
6330 4. The first argument for a ${extract expansion item is the key name or field
6331 number. Leading and trailing spaces in this item were not being ignored,
6332 causing some misleading effects.
6334 5. When deliver_drop_privilege was set, single queue runner processes started
6335 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6336 same command in the process it spins off) were not dropping privilege.
6338 6. When the daemon running as "exim" started a queue runner, it always
6339 re-executed Exim in the spun-off process. This is a waste of effort when
6340 deliver_drop_privilege is set. The new process now just calls the
6341 queue-runner function directly.
6347 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6350 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6351 ignored. Though the use of -oP was forcing the writing of a pid file, it
6352 was always written to the default place.
6354 3. If the message "no IP address found for host xxxx" is generated during
6355 incoming verification, it is now followed by identification of the incoming
6356 connection (so you can more easily find what provoked it).
6358 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6360 5. Added some features to "harden" Exim a bit more against certain attacks:
6362 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6363 be put in Local/Makefile. This is like the never_users runtime option,
6364 but it cannot be overridden. The default setting is "root".
6366 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6367 prefix string with which any file named in a -C command line option
6370 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6371 is retained for -C and -D only if the caller of Exim is root. Without
6372 it, the exim user may also use -C and -D and retain privilege.
6374 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6375 command line option is disabled.
6377 6. Macro names set by the -D option must start with an upper case letter, just
6378 like macro names defined in the configuration file.
6380 7. Added "dereference=" facility to LDAP.
6382 8. Two instances of the typo "uknown" in the source files are fixed.
6384 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6385 the Configure-Makefile script screwed up while processing it.
6387 10. Incorporated PCRE 4.4.
6389 11. The SMTP synchronization check was not operating right at the start of an
6390 SMTP session. For example, it could not catch a HELO sent before the client
6391 waited for the greeting. There is now a check for outstanding input at the
6392 point when the greeting is written. Because of the duplex, asynchronous
6393 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6394 way, but not yet received, when the check is performed.
6396 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6397 on TCP/IP sockets, because this apparently causes some broken clients to
6400 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6401 unchanged) from the Cygwin maintainer.
6403 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6404 was supported instead of testing for mbx. Effectively a typo.
6406 15. The spa authenticator server code was not checking that the input it
6407 received was valid base64.
6409 16. The debug output line for the "set" modifier in ACLs was not showing the
6410 name of the variable that was being set.
6412 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6414 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6415 lookup on its own. Something else was needed to trigger the lookup. For
6416 example, a match in host_lookup or the need for a host name in a host list.
6417 Now, if $sender_host_name is referenced and the host name has not yet been
6418 looked up, a lookup is performed. If the lookup fails, the variable remains
6419 empty, and $host_lookup_failed is set to "1".
6421 19. Added "eqi" as a case-independent comparison operator.
6423 20. The saslauthd authentication condition could segfault if neither service
6424 nor realm was specified.
6426 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6427 error message that was logged was misleading, and incoming SMTP
6428 connections were dropped. The message is now more accurate, and temporary
6429 errors are given to SMTP connections.
6431 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6432 (except RSET) with a 421 error, until QUIT is received. However, it was
6433 failing to send a response to QUIT.
6435 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6436 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6437 ACL is now run after the helo_try_verify_hosts code.
6439 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6440 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6441 case-independent, so other case variants are also recognized). Apparently
6442 some systems use these upper case variants.
6444 25. If more than two messages were waiting for the same host, and a transport
6445 filter was specified for the transport, Exim sent two messages over the
6446 same TCP/IP connection, and then failed with "socket operation on non-
6447 socket" when it tried to send the third.
6449 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6451 27. The extern definition of crypt16() in expand.c was not being excluded when
6452 the OS had its own crypt16() function.
6454 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6455 as a preferred synonym for return_size_limit, both as an option and as an
6458 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6460 30. Changed the default configuration ACL to relax the local part checking rule
6461 for addresses that are not in any local domains. For these addresses,
6462 slashes and pipe symbols are allowed within local parts, but the sequence
6463 /../ is explicitly forbidden.
6465 31. SPA server authentication was not clearing the challenge buffer before
6468 32. log_message in a "warn" ACL statement was writing to the reject log as
6469 well as to the main log, which contradicts the documentation and doesn't
6470 seem right (because no rejection is happening). So I have stopped it.
6472 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6473 However, I am unable to do any testing of this.
6475 34. Fixed an infelicity in the appendfile transport. When checking directories
6476 for a mailbox, to see if any needed to be created, it was accidentally
6477 using path names with one or more superfluous leading slashes; tracing
6478 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6480 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6481 contents are added to the log line that is written for every discarded
6482 recipient. (Previously a log_message setting was ignored.)
6484 36. The ${quote: operator now quotes the string if it is empty.
6486 37. The install script runs exim in order to find its version number. If for
6487 some reason other than non-existence or emptiness, which it checks, it
6488 could not run './exim', it was installing it with an empty version number,
6489 i.e. as "exim-". This error state is now caught, and the installation is
6492 38. An argument was missing from the function that creates an error message
6493 when Exim fails to connect to the socket for saslauthd authentication.
6494 This could cause Exim to crash, or give a corrupted message.
6496 39. Added isip, isip4, and isip6 to ${if conditions.
6498 40. The ACL variables $acl_xx are now saved with the message, and can be
6499 accessed later in routers, transports, and filters.
6501 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6502 strings in the file are not string-expanded.
6504 42. If a MAIL command specified a SIZE value that was too large to fit into an
6505 int variable, the check against message_size_limit failed. Such values are
6506 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6507 day this will have to be increased, but I don't think I want to be around
6508 when emails are that large.
6515 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6516 iconv() is not standard in FreeBSD.
6518 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6519 IPv6 enabled. The observed symptom was a segmentation fault on return from
6520 the function os_common_find_running_interfaces() in src/os.c.
6522 3. In the check_special_case() function in daemon.c I had used "errno" as an
6523 argument name, which causes warnings on some systems. This was basically a
6524 typo, since it was named "eno" in the comments!
6526 4. The code that waits for the clock to tick (at a resolution of some fraction
6527 of a second) so as to ensure message-id uniqueness was always waiting for
6528 at least one whole tick, when it could have waited for less. [This is
6529 almost certainly not relevant at current processor speeds, where it is
6530 unlikely to ever wait at all. But we try to future-proof.]
6532 5. The function that sleeps for a time interval that includes fractions of a
6533 second contained a race. It did not block SIGALRM between setting the
6534 timer, and suspending (a couple of lines later). If the interval was short
6535 and the sigsuspend() was delayed until after it had expired, the suspension
6536 never ended. On busy systems this could lead to processes getting stuck for
6539 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6540 process, before it forks any delivery processes. The open lookup caching
6541 mechanism meant that the open file or database connection was passed into
6542 the delivery process. The problem was that delivery processes always tidy
6543 up cached lookup data. This could cause a problem for the next delivery
6544 process started by the queue runner, because the external queue runner
6545 process does not know about the closure. So the next delivery process
6546 still has data in the lookup cache. In the case of a file lookup, there was
6547 no problem because closing a file descriptor in a subprocess doesn't affect
6548 the parent. However, if the lookup was caching a connection to a database,
6549 the connection was closed, and the second delivery process was likely to
6550 see errors such as "PGSQL: query failed: server closed the connection
6551 unexpectedly". The problem has been fixed by closing all cached lookups
6552 in a queue runner before running a delivery process.
6554 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6555 seem to have the "const" qualifier which it has on other OS. I've
6558 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6559 *of the same type* (client or server) with the same public name that an
6560 error should be diagnosed.
6562 9. When Exim looked up a host name for an IP address, but failed to find the
6563 original IP address when looking up the host name (a safety check), it
6564 output the message "<ip address> does not match any IP for NULL", which was
6565 confusing, to say the least. The bug was that the host name should have
6566 appeared instead of "NULL".
6568 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6569 user that is built into the binary, and the -C or -D options is used, root
6570 privilege is dropped before the configuration file is read. In addition,
6571 logging is switched to stderr instead of the normal log files. If the
6572 configuration then re-defines the Exim user, the unprivileged environment
6573 is probably not what is expected, so Exim logs a panic warning message (but
6576 However, if deliver_drop_privilege is set, the unprivileged state may well
6577 be exactly what is intended, so the warning has been cut out in that case,
6578 and Exim is allowed to try to write to its normal log files.
6584 1. smtp_return_error_details was not giving details for temporary sender
6585 or receiver verification errors.
6587 2. Diagnose a configuration error if two authenticators have the same public
6590 3. Exim used not to create the message log file for a message until the first
6591 delivery attempt. This could be confusing when incoming messages were held
6592 for policy or load reasons. The message log file is now created at the time
6593 the message is received, and an initial "Received" line is written to it.
6595 4. The automatically generated man page for command line options had a minor
6596 bug that caused no ill effects; however, a more serious problem was that
6597 the procedure for building the man page automatically didn't always
6598 operate. Consequently, release 4.20 contains an out-of-date version. This
6599 shouldn't happen again.
6601 5. When building Exim with embedded Perl support, the script that builds the
6602 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6603 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6605 6. The freeze_tell option was not being used for messages that were frozen on
6606 arrival, either by an ACL or by local_scan().
6608 7. Added the smtp_incomplete_transaction log selector.
6610 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6611 was accepting AUTH without a new EHLO.
6613 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6614 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6617 10. Logging of TCP/IP connections (when configured) now happens in the main
6618 daemon process instead of the child process, so that the TCP/IP connection
6619 count is more accurate (but it can never be perfect).
6621 11. The use of "drop" in a nested ACL was not being handled correctly in the
6622 outer ACL. Now, if condition failure induced by the nested "drop" causes
6623 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6624 or "require"), the connection is dropped.
6626 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6627 that yield "discard" can now be used with an "accept" or a "discard" verb,
6628 but an error is generated for any others (because I can't see a useful way
6629 to define what should happen).
6631 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6632 are now processed in the same way as lines in the Exim configuration file.
6633 In particular, continuation lines are supported.
6635 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6637 15. Added -ti meaning -t -i.
6639 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6640 domains, and warn by logging if others are found.
6642 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6643 returned from GIFCONF when Exim is trying to find the list of interfaces on
6644 a host. The code in os.c has been modified to copy each ifreq to an aligned
6645 structure in all cases.
6647 Also, in some cases, the returned ifreq's were being copied to a 'struct
6648 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6649 means the last couple of bytes of an IPv6 address could be chopped if the
6650 ifreq contained only a normal sockaddr (14 bytes storage).
6652 18. Named domain lists were not supported in the hosts_treat_as_local option.
6653 An entry such as +xxxx was not recognized, and was treated as a literal
6656 19. Ensure that header lines added by a DATA ACL are included in the reject log
6657 if the ACL subsequently rejects the message.
6659 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6660 MD5 (which is deprecated).
6662 21. When testing a filter file using -bf, Exim was writing a message when it
6663 took the sender from a "From " line in the message, but it was not doing so
6664 when it took $return_path from a Return-Path: header line. It now does.
6666 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6667 with a valid header line field name (a series of printing characters
6668 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6670 23. Changed "disc" in the source to "disk" to conform to the documentation and
6671 the book and for uniformity.
6673 24. Ignore Sendmail's -Ooption=value command line item.
6675 25. When execve() failed while trying to run a command in a pipe transport,
6676 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6677 could be confused with a return value of 69 from the command itself. This
6678 has been changed to 127, the value the shell returns if it is asked to run
6679 a non-existent command. The wording for the related log line suggests a
6680 non-existent command as the problem.
6682 26. If received_header_text expands to an empty string, do not add a Received:
6683 header line to the message. (Well, it adds a token one on the spool, but
6684 marks it "old" so that it doesn't get used or transmitted.)
6686 27. Installed eximstats 1.28 (addition of -nt option).
6688 28. There was no check for failure on the call to getsockname() in the daemon
6689 code. This can fail if there is a shortage of resources on the system, with
6690 ENOMEM, for example. A temporary error is now given on failure.
6692 29. Contrary to the C standard, it seems that in some environments, the
6693 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6694 program. Exim now does this explicitly; it affects the formatting of
6695 timestamps using strftime().
6697 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6698 complaints. I've now initialized all the variables, to avoid this.
6700 32. Header lines added by a system filter were not being "seen" during
6701 transport-time rewrites.
6703 33. The info_callback() function passed to OpenSSL is set up with type void
6704 (*)(SSL *, int, int), as described somewhere. However, when calling the
6705 function (actually a macro) that sets it up, the type void(*)() is
6706 expected. I've put in a cast to prevent warnings from picky compilers.
6708 34. If a DNS black list lookup found a CNAME record, but there were no A
6709 records associated with the domain it pointed at, Exim crashed.
6711 35. If a DNS black list lookup returned more than one A record, Exim ignored
6712 all but the first. It now scans all returned addresses if a particular IP
6713 value is being sought. In this situation, the contents of the
6714 $dnslist_value variable are a list of all the addresses, separated by a
6717 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6718 to accept a host name and all its aliases if the forward lookup for any of
6719 them yielded the IP address of the incoming connection. Now it accepts only
6720 those names whose forward lookup yields the correct IP address. Any other
6721 names are discarded. This closes a loophole whereby a rogue DNS
6722 administrator could create reverse DNS records to break through a
6723 wildcarded host restriction in an ACL.
6725 37. If a user filter or a system filter that ran in a subprocess used any of
6726 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6727 the wrong values were passed to the pipe command ($thisaddress had the
6728 value of $0, $0 had the value of $1, etc). This bug was introduced by
6729 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6731 38. Improved the line breaking for long SMTP error messages from ACLs.
6732 Previously, if there was no break point between 40 and 75 characters, Exim
6733 left the rest of the message alone. Two changes have been made: (a) I've
6734 reduced the minimum length to 35 characters; (b) if it can't find a break
6735 point between 35 and 75 characters, it looks ahead and uses the first one
6736 that it finds. This may give the occasional overlong line, but at least the
6737 remaining text gets split now.
6739 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6740 file descriptors might be low, and that setting 1000 would always raise it.
6741 It turns out that in some environments, the limit is already over 1000 and
6742 that lowering it causes trouble. So now Exim takes care not to decrease it.
6744 40. When delivering a message, the value of $return_path is set to $sender_
6745 address at the start of routing (routers may change the value). By an
6746 oversight, this default was not being set up when an address was tested by
6747 -bt or -bv, which affected the outcome if any router or filter referred to
6750 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6751 log only when "message" or "log_message" are set. However, if one of the
6752 conditions was an address verification, or a call to a nested ACL, the
6753 messages generated by the underlying test were being passed through. This
6754 no longer happens. The underlying message is available in $acl_verify_
6755 message for both "message" and "log_message" expansions, so it can be
6756 passed through if needed.
6758 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6759 new expansion $bh_ to give the encoded byte string without charset
6760 translation. Translation happens only if iconv() is available; HAVE_ICONV
6761 indicates this at build time. HEADERS_CHARSET gives the charset to
6762 translate to; headers_charset can change it in the configuration, and
6763 "headers charset" can change it in an individual filter file.
6765 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6766 that creates RFC 2047 encoded "words" labels them as that charset instead
6767 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6768 expansion operator; (ii) when Exim creates a From: line for a local
6769 message; (iii) when a header line is rewritten to include a "phrase" part.
6771 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6772 buggy, causing it to skip the first lines of messages whose message ID
6773 ended in 'D'. This would not have bitten before Exim release 4.14, because
6774 message IDs were unlikely to end in 'D' before then. The effect was to have
6775 incorrect size information for certain domains.
6777 45. #include "config.h" was missing at the start of the crypt16.c module. This
6778 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6781 46. If there was a timeout during a "random" callout check, Exim treated it as
6782 a failure of the random address, and carried on sending RSET and the real
6783 address. If the delay was just some slowness somewhere, the response to the
6784 original RCPT would be taken as a response to RSET and so on, causing
6785 mayhem of various kinds.
6787 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6788 when I implemented it. It didn't allow for the fact that some option values
6789 may legitimately be negative (e.g. size_addition), and it didn't even do
6790 the right test for positive values.
6792 48. Domain names in DNS records are case-independent. Exim always looks them up
6793 in lower case. Some resolvers return domain names in exactly the case they
6794 appear in the zone file, that is, they may contain uppercase letters. Not
6795 all resolvers do this - some return always lower case. Exim was treating a
6796 change of case by a resolver as a change of domain, similar to a widening
6797 of a domain abbreviation. This triggered its re-routing code and so it was
6798 trying to route what was effectively the same domain again. This normally
6799 caused routing to fail (because the router wouldn't handle the domain
6800 twice). Now Exim checks for this case specially, and just changes the
6801 casing of the domain that it ultimately uses when it transmits the message
6804 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6807 50. If a filter generated a file delivery with a non-absolute name (possible if
6808 no home directory exists for the router), the forbid_file option was not
6811 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6812 the existing equality matching.
6814 52. Exim was using ints instead of ino_t variables in some places where it was
6815 dealing with inode numbers.
6817 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6818 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6819 called TMPDIR, and if it finds it is different, it changes its value.
6821 54. The smtp_printf() function is now made available to local_scan() so
6822 additional output lines can be written before returning. There is also an
6823 smtp_fflush() function to enable the detection of a dropped connection.
6824 The variables smtp_input and smtp_batched_input are exported to
6827 55. Changed the default runtime configuration: the message "Unknown user"
6828 has been removed from the ACL, and instead placed on the localuser router,
6829 using the cannot_route_message feature. This means that any verification
6830 failures that generate their own messages won't get overridden. Similarly,
6831 the "Unrouteable address" message that was in the ACL for unverifiable
6832 relay addresses has also been removed.
6834 56. Added hosts_avoid_esmtp to the smtp transport.
6836 57. The exicyclog script was not checking for the esoteric option
6837 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6838 will work only if exicyclog is run under the appropriate euid.
6840 58. Following a discussion on the list, the rules by which Exim recognises line
6841 endings on incoming messages have been changed. The -dropcr and drop_cr
6842 options are now no-ops, retained only for backwards compatibility. The
6843 following line terminators are recognized: LF CRLF CR. However, special
6844 processing applies to CR:
6846 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6847 nor a local message in the state where . is a terminator.
6849 (ii) If a bare CR is encountered in a header line, an extra space is added
6850 after the line terminator so as not to end the header. The reasoning
6851 behind this is that bare CRs in header lines are most likely either
6852 to be mistakes, or people trying to play silly games.
6854 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6855 was being incorrectly given as 18 bytes larger than it should have been.
6856 This is a VOB (very old bug).
6858 60. This may never have affected anything current, but just in case it has:
6859 When the local host is found other than at the start of a list of hosts,
6860 the local host, those with the same MX, and any that follow, are discarded.
6861 When the list in question was part of a longer list of hosts, the following
6862 hosts (not currently being processed) were also being discarded. This no
6863 longer happens. I'm not sure if this situation could ever has previously
6866 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6869 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6870 header. This is something that is recommended in a new Internet Draft, and
6871 is something that is documented as being done by Sendmail. There are two
6872 possible values. For messages generated by the autoreply transport, Exim
6875 Auto-Submitted: auto-replied
6877 whereas for all other generated messages (e.g. bounces) it adds
6879 Auto-Submitted: auto-generated
6881 63. The "personal" condition in filters now includes a test for the
6882 Auto-Submitted: header. If it contains the string "auto-" the message it
6883 not considered personal.
6885 64. Added rcpt_include_affixes as a generic transport option.
6887 65. Added queue_only_override (default true).
6889 66. Added the syslog_duplication option.
6891 67. If what should have been the first header line of a message consisted of
6892 a space followed by a colon, Exim was mis-interpreting it as a header line.
6893 It isn't of course - it is syntactically invalid and should therefore be
6894 treated as the start of the message body. The misbehaviour could have
6895 caused a number of strange effects, including loss of data in subsequent
6896 header lines, and spool format errors.
6898 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6899 client host had authenticated. This control can now be exercised by an ACL
6900 for more flexibility.
6902 69. By default, callouts do not happen when testing with -bh. There is now a
6903 variant, -bhc, which does actually run the callout code, including
6904 consulting and updating the callout cache.
6906 70. Added support for saslauthd authentication, courtesy of Alexander
6909 71. If statvfs() failed on the spool or log directories while checking their
6910 size for availability, Exim confusingly gave the error "space shortage".
6911 Furthermore, in debugging mode it crashed with a floating point exception.
6912 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6913 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6914 serious problem, Exim now writes to the main and panic logs when this
6915 happens, with details of the failure. It then refuses to accept the
6916 incoming message, giving the message "spool directory problem" or "log
6917 directory problem" with a 421 code for SMTP messages.
6919 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6920 0, 1, and 2 exist, because some OS complain for execs without them (see
6921 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6922 descriptors. However, the code omitted to check that the open succeeded,
6923 causing mysterious errors if for some reason the permissions on /dev/null
6924 got screwed. Now Exim writes a message to the main and panic logs, and
6925 bombs out if it can't open /dev/null.
6927 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6928 interact so that it is all more flexible. It is supposed to remain
6929 backwards compatible. Also added extra_local_interfaces.
6931 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6932 to bomb out with an assertion failure - to the client this appears as a
6933 connection drop. This problem occurs in the part of the code that was taken
6934 from the Samba project. Fortunately, the assertion is in a very simple
6935 function, so I have fixed this by reproducing the function inline in the
6936 one place where it is called, and arranging for authentication to fail
6937 instead of killing the process with assert().
6939 75. The SPA client code was not working when the server requested OEM rather
6940 than Unicode encoding.
6942 76. Added code to make require_files with a specific uid setting more usable in
6943 the case where statting the file as root fails - usually a non-root-mounted
6944 NFS file system. When this happens and the failure is EACCES, Exim now
6945 forks a subprocess and does the per-uid checking as the relevant uid.
6947 77. Added process_log_path.
6949 78. If log_file_path was not explicitly set, a setting of check_log_space or
6950 check_log_inodes was ignored.
6952 79. If a space check for the spool or log partitions fails, the incident is now
6953 logged. Of course, in the latter case the data may get lost...
6955 80. Added the %p formatting code to string_format() so that it can be used to
6956 print addresses in debug_print(). Adjusted all the address printing in the
6957 debugging in store.c to use %p rather than %d.
6959 81. There was a concern that a line of code in smtp_in.c could overflow a
6960 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6961 initially expressed, the concern was not well-founded, because trailing
6962 spaces are removed early. However, if the trailing spaces were followed by
6963 a NULL, they did not get removed, so the overflow was possible. Two fixes
6966 (a) I re-wrote the offending code in a cleaner fashion.
6967 (b) If an incoming SMTP command contains a NULL character, it is rejected
6970 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6971 runs initgroups(), so that if the Exim user is in any additional groups,
6972 they will be used during message reception.
6978 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob