1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
9 SMTP connection" log lines.
11 JH/02 Option default value updates:
12 - queue_fast_ramp (main) true (was false)
13 - remote_max_parallel (main) 4 (was 2)
15 JH/03 Cache static regex pattern compilations, for use by ACLs.
17 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
18 Make the rewrite never match and keep the logging. Trust the
19 admin to be using verify=header-syntax (to actually reject the message).
25 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
26 after reception to before a subsequent reception. This should
27 mean slightly faster delivery, and also confirmation of reception
30 JH/02 Move from using the pcre library to pcre2. The former is no longer
31 being developed or supported (by the original developer).
33 JH/03 Constification work in the filters module required a major version
34 bump for the local-scan API. Specifically, the "headers_charset"
35 global which is visible via the API is now const and may therefore
36 not be modified by local-scan code.
38 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
39 sendfile() didi not account for the way the ClamAV driver code called it.
41 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
42 time check was being done for every character; replace that with one
45 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
46 sent was prefixed with a length byte.
48 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
49 RFC 5321. Previously Dovecot (at least) would log errors during
52 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
53 in 4.61 without a reason logged. Binaries will be bigger, which might
54 matter on diskspace-constrained systems, but debug is easier.
56 JH/09 Fix macro-definition during "-be" expansion testing. The move to
57 write-protected store for macros had not accounted for these runtime
58 additions; fix by removing this protection for "-be" mode.
60 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
61 handing out large-numbered file descriptors, violating the usual Unix
62 assumption (and required by Posix) that the lowest possible number will be
63 allocated by the kernel when a new one is needed. In the daemon, and any
64 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
65 are not useable for FD_SET() [and hence select()] and overwrite the stack.
66 Assorted crashes happen.
68 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
69 main-section options or in a connect ACL, the value from the first ever
70 connection was never replaced for subsequent connections. Found by
73 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
74 though only once PCRE2 was introduced: the memory accounting used under
75 debug offset allocations by an int, giving a hard trap in early startup.
76 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
78 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
79 with underbars is given. The write-protection of configuration introduced
80 in 4.95 trapped when normalisation was applied to an option not needing
83 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
85 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
86 failing to close the certificates directory, every hour or any time it
89 JH/16 Debugging initiated by an ACL control now continues through into routing
90 and transport processes. Previously debugging stopped any time Exim
91 re-execs, or for processing a queued message.
93 JH/17 The "expand" debug selector now gives more detail, specifically on the
94 result of expansion operators and items.
96 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
97 bad comparison between the option value and the name of the file to
98 be included was done, and a mismatch was wrongly identified.
99 4.88 to 4.95 are affected.
101 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
103 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
104 supplied as the db file-pair basename. Previously, if a directory
105 path was given, for example via the autoreply "once" option, the DB
106 file.pag and file.dir files would be created in that directory's
109 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
110 "taint" log_selector. These were previously deprecated.
112 JH/22 Fix static address-list lookups to properly return the matched item.
113 Previously only the domain part was returned.
115 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
116 the call into OpenSSL to send a TLS Close was being repeated; this
117 resulted in the library waiting for the peer's Close. If that was never
118 sent we waited forever. Fix by tracking send calls.
120 JH/24 The ${run} expansion item now expands its command string elements after
121 splitting. Previously it was before; the new ordering makes handling
122 zero-length arguments simpler. The old ordering can be obtained by
123 appending a new option "preexpand", after a comma, to the "run".
125 JH/25 Taint-check exec arguments for transport-initiated external processes.
126 Previously, tainted values could be used. This affects "pipe", "lmtp" and
127 "queryprogram" transport, transport-filter, and ETRN commands.
128 The ${run} expansion is also affected: in "preexpand" mode no part of
129 the command line may be tainted, in default mode the executable name
132 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
133 the the facility was not passed across execs, and only the first message
134 passed over a connection could use BDAT; any further ones using DATA.
136 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
137 uses $sending_ip_address and an interface is specified.
138 Previously any use of the local address in the EHLO name disabled
139 PIPECONNECT, the common case being to use the rDNS of it.
141 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
142 resumption. Previously verify failed because no certificate status is
143 passed on the wire for the restarted session. Fix by using the recorded
144 ocsp status of the stored session for the new connection.
146 JH/29 TLS resumption: the key for session lookup in the client now includes
147 more info that a server could potentially use in configuring a TLS
148 session, avoiding oferring mismatching sessions to such a server.
149 Previously only the server IP was used.
151 JH/30 Fix string_copyn() for limit greater than actual string length.
152 Previously the copied amount was the limit, which could result in a
153 overlapping memcpy for newly allocated destination soon after a
154 source string shorter than the limit. Found/investigated by KM.
156 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
157 close; it may be needed for a subsequent connection. This caused a
158 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
160 JH/32 Fix CHUNKING for a second message on a connection when the first was
161 rejected. Previously we did not reset the chunking-offered state, and
162 erroneously rejected the BDAT command. Investigation help from
165 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
166 an empty address. Previously the expansion returned an error.
168 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
169 proxy. Previously these were misparsed, leading to paniclog entries.
175 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
176 related applications. Previously an "H" was used where available info
177 says that "M" should be, so change to match.
179 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
180 as arguments, so an implementation trying to copy these into a local
181 buffer was taking a taint-enforcement trap. Fix by using dynamically
182 created buffers. Similar fix for radius expansion condition.
184 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
185 reasonable, eg. to count headers. Fix by using dynamically created
186 buffers rather than a local. Do similar fixes for ACL actions "dcc",
187 "log_reject_target", "malware" and "spam"; the arguments are expanded
188 so could be handling tainted values.
190 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
191 broken the (no-op) support for this sendmail command. Restore it
192 to doing nothing, silently, and returning good status.
194 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
195 record path was given (or the default used) without a leading directory
196 path, an error occurred on trying to open it. Use the transport's working
199 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
200 transport. Previously it was the name on the DNS A-record; use instead
201 the head of the CNAME chain leading there (if there is one). This seems
202 to align better with RFC 6125.
204 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
205 smtp_accept_max_per_host allocated resources which were not released
206 when the limit was exceeded. This eventually crashed the daemon. Fix
207 by adding a release action in that path.
209 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
210 expanded; previously using tainted values was rejected. Fix by using
211 dynamically-created buffers.
213 JH/09 Relax restrictions on ACL verify condition needing access to message
214 headers. Previously they were only permitted in data and non-smtp ACLs;
215 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
216 not_blind, header_sender and header_names_ascii verification.
218 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
219 Previously a macro used one argument twice; when called with the
220 argument as an expression having side-effects, incorrect operation
221 resulted. Use an inlineable function.
223 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
224 held open for a verify callout. Previously this wan not accounted for
225 and a corrupt onward SMTP conversation resulted.
227 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
228 Previously they were embedded naively in the constructed address; when
229 needed, strip the quoting and quote the entire local_part.
230 Also make the inbound_srs expansion condition handle quoting.
232 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
233 excluded, not matching the documentation.
235 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
236 was given for the sqlite_dbfile a trap resulted.
238 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
239 "name" argument resulted in a trap. There is no reason to disallow such;
240 this was a coding error.
242 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
243 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
244 spent suspended, ignoring the POSIX definition. Previously we assumed
245 it did and a constant offset from real time could be used as a correction.
246 Change to using the same clock source for the start-of-message and the
247 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
248 exists, just to get a clock slightly more aligned to reality.
250 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
251 RFC says it is optional some validators care. The missing char was not
252 intended but triggered by a line-wrap alignment. Discovery and fix by
253 Guillaume Outters, hacked on by JH.
255 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
256 name being quoted was tainted a trap would be taken. Fix by using
257 dynamically created buffers. The routine could have been called by a
258 rewrite with the "h" flag, by using the "-F" command-line option, or
259 by using a "name=" option on a control=submission ACL modifier.
261 JH/19 SPF: change the Authentication-Results expansion component to give
262 smtp.helo when the sender domain is empty. Previously it gave
265 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
266 not in 127.0.0.0/8 to help in spotting list domains taken over by a
267 domain-parking registrar.
269 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
270 Previously when a whitespace character was specified it was not inserted
271 after removing the newline.
273 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
274 the domain part of the recipient address. This overrides any tls_sni
275 option set, which was previously used.
277 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
280 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
281 is_tainted() had an off-by-one error in the overenthusiastic direction.
282 Find and fix by Gavan. Although NetBSD is not a supported platform for
283 4.94 this bug could affect other platforms.
285 PP/01 Fix default prime selection to be consistent.
286 One path used ike23 still, instead of exim.dev.20160529.3; now both
287 execution flows will use the same DH primes (currently
288 exim.dev.20160529.3).
290 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
291 option in smtp transport, to match the documentation. Previously
292 verification was not being done in some cases where it should have been.
294 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
295 than one server was defined and depending on the platform memory layout
296 details, an internal consistency trap could be hit while walking the list
299 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
300 files. The value is used by the authresults expansion item. Previously
301 if this was used in a router or transport, a crash could result.
303 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
304 have a sufficient incidence of EINTR returns from select that an
305 interaction with other operations done by the main daemon loop exposed
306 a bug in the error-handling. This was benign apart from the log
309 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
310 for consistency with delivered "=>" and failed "**" lines. While we're
311 there, handle PRX and TFO.
313 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
314 applied. This resulted, if any header-line rewrite rules were configured,
315 in a panic-log triggerable by sending a message with a long address in
316 a header. Fix by increasing the arbitrary limit to larger than a single
317 (dewrapped) 5322 header line maximum size.
319 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
320 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
321 RFC 6648 which deprecates X- options in protocols as a general practice.
322 Changeover between the implementations is handled by the mechanisms
325 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
326 a remote which uses callout/hold. Previously the local was queued.
328 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
331 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
332 log line. The asterisk is now consistently appended to the remote IP
333 (and port, if given), and will also be provided on defer and fail log
334 lines. Previously it could be placed on the local IP if that was being
335 logged, and was only provided on delivery lines.
337 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
339 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
340 authenticator client_send option. Previously the next char, after a pair
341 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
342 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
343 documentation. There is still no way to get a leading ^ immediately
344 after a NUL (ie. for the password of a PLAIN method authenticator.
346 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
347 files. For bad sizes read, delete the record and whine to paniclog.
349 JH/38 When logging an AUTH failure, as server, do not include sensitive
350 information. Previously, the credentials would be included if given
351 as part of the AUTH command line and an ACL denied authentication.
353 JH/39 Bug 2691: fix $local_part_data. When the matching list element
354 referred to a file, bad data was returned. This likely also affected
357 JH/40 The gsasl authenticator now supports caching of the salted password
358 generated by the client-side implementation. This required the addition
359 of a new variable: $auth4.
361 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
362 left undeleted; the attempt to re-create it then failed - resulting in
363 the usual "SIGHUP tp have daemon reload configuration" to not work.
364 This affected any platform not supporting "abstract" Unix-domain
365 sockets (i.e. not Linux).
367 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
368 recipients" response to RCPT in a later response, with a 250. The
369 previous coding assumed this would not happen, and under PIPELINING
370 would result in both lost and duplicate recipients for a message.
372 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
373 Previously the weighting was incorrectly applied. Similar fix for socks
374 proxies. Found and fixed by Heiko Schlichting.
376 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
377 not handle sub-lists included using the +namedlist syntax. While
378 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
379 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
382 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
383 Previously the permanent pool was used, so the sore could not be freed.
384 This meant a connection with many messages would use continually-growing
387 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
388 per-pool so as not to waste too much space. Previously a constant size
389 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
390 DOS attacks harder. The cost is wasted memory use in the larger blocks.
392 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
393 same space repeatedly, and a smaller process footprint.
395 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
396 for GnuTLS. Previously the name started "%s" which, while not a bug,
397 looked as if if might be one.
399 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
400 "continued connection") make the $proxy_* variables available. Previously
401 the information was not passed across the exec() call for subsequent
402 transport executions. This also mean that the log lines for the
403 messages can show the proxy information.
405 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
406 receive time. With modern systems the difference is significant.
407 The historical behaviour can be restored by disabling (a new) log_selector
408 "queue_time_exclusive".
410 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
411 content) were specifically tested for. Now, also cover expansions
412 resulting in ACL names and inline ACL content.
414 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
415 making it unusable in complex expressions.
417 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
418 fail with a taint-check on the spoolfile name, and leave the message
421 HS/01 Enforce absolute PID file path name.
423 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
425 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
427 PP/02 Bug 2643: Correct TLS DH constants.
428 A missing NUL termination in our code-generation tool had led to some
429 incorrect Diffie-Hellman constants in the Exim source.
430 Reported by kylon94, code-gen tool fix by Simon Arlott.
432 PP/03 Impose security length checks on various command-line options.
433 Fixes CVE-2020-SPRSS reported by Qualys.
435 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
436 better. Reported by Qualys.
438 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
439 providing a particularly obnoxious sender full name.
442 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
444 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
447 PP/08 Change default for recipients_max from unlimited to 50,000.
449 PP/09 Fix security issue with too many recipients on a message (to remove a
450 known security problem if someone does set recipients_max to unlimited,
451 or if local additions add to the recipient list).
452 Fixes CVE-2020-RCPTL reported by Qualys.
454 PP/10 Fix security issue in SMTP verb option parsing
455 Fixes CVE-2020-EXOPT reported by Qualys.
457 PP/11 Fix security issue in BDAT state confusion.
458 Ensure we reset known-good where we know we need to not be reading BDAT
459 data, as a general case fix, and move the places where we switch to BDAT
460 mode until after various protocol state checks.
461 Fixes CVE-2020-BDATA reported by Qualys.
463 HS/03 Die on "/../" in msglog file names
465 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
466 the Exim runtime user are allowed to create files.
468 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
471 QS/03 When reading the output from interpreted forward files we do not
472 pass the pipe between the parent and the interpreting process to
473 executed child processes (if any).
475 QS/04 Always die if requested from internal logging, even is logging is
478 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
479 the API; compilation noo longer completes with DMARC support included.
480 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
481 been reported on other platforms.
483 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
485 JH/56 Make the majority of info read from config files readonly, for defence-in-
486 depth against exploits. Suggestion by Qualys.
487 Not supported on Solaris 10.
489 JH/57 Fix control=fakreject for a custom message containing tainted data.
490 Previously this resulted in a log complaint, due to a re-expansion present
491 since fakereject was originally introduced.
493 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
494 modifier was given, a loop resulted.
496 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
497 pipelined SMTP command followed the BDAT LAST then it would be
498 incorrectly treated as part of the message body, causing a verification
501 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
502 require looser validation rules than those for 821-level addresses,
503 which only permit IP addresses.
509 JH/01 Avoid costly startup code when not strictly needed. This reduces time
510 for some exim process initialisations. It does mean that the logging
511 of TLS configuration problems is only done for the daemon startup.
513 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
515 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
516 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
519 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
520 DKIM signing is being done. Previously a transport_filter would always
521 disable CHUNKING, falling back to traditional DATA.
523 JH/05 Regard command-line recipients as tainted.
525 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
527 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
528 PAM library frees one of the arguments given to it, despite the
529 documentation. Therefore a plain malloc must be used.
531 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
532 on-stack buffers were used, resulting in a taint trap when DSN information
533 copied from a received message was written into the buffer.
535 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
536 the ordering of its ARC headers. This caused a crash.
538 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
539 a new record was being constructed with information from the peer, a trap
542 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
543 installation would get error messages from DMARC verify, when it hit the
544 nonexistent file indicated by the default. Distros wanting DMARC enabled
545 should both provide the file and set the option.
546 Also enforce no DMARC verification for command-line sourced messages.
548 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
549 could, depending on the platform, hang at the STARTTLS response.
551 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
552 message on a connection. Previously if one message had ARC headers and
553 the following one did not, a crash could result when adding an
554 Authentication-Results: header.
556 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
557 Exim main code and Exim-related utities. The introduction of taint
558 tracking also did many adjustments to string handling. Since then, eximon
559 frequently terminated with an assert failure.
561 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
562 check for 452 responses. This slightly helps the inefficieny of doing
563 a large alias-expansion into a recipient-limited target. The max_rcpt
564 transport option still applies (and at the current default, will override
565 the new feature). The check is done for either cause of synch, and forces
566 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
567 connection. The new facility is not tunable at this time.
569 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
570 library live data was being used, so the results became garbage. Make
571 copies while it is still usable.
573 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
574 on delivery deferred (==) and failed (**) lines (if a delivery was
575 attemtped). Previously it was only on completion (=>) lines.
577 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
578 for the expansion of the server_scram_iter and server_scram_salt options.
580 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
581 are now specifically given a NO_DATA response without hitting the system
582 resolver. The library goes on to do the now-standard TXT lookup.
583 Use of dnsdb lookups is not affected.
585 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
586 only retrieve the errormessage once. Previously two calls to dlerror()
587 were used, and the second one (for mainlog/paniclog) retrieved null
590 JH/20 Taint checking: disallow use of tainted data for
591 - the appendfile transport file and directory options
592 - the pipe transport command
593 - the autoreply transport file, log and once options
594 - file names used by the redirect router (including filter files)
596 - paths used by single-key lookups
597 Previously this was permitted.
599 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
600 adjusted the size of a major service buffer; this failed because the
601 buffer was in use at the time. Change to a compile-time increase in the
602 buffer size, when this authenticator is compiled into exim.
604 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
605 previous fast-mode was untenable in the face of glibs using mmap to
606 support larger malloc requests.
608 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
609 New values supported, if defined on system where compiled:
610 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
611 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
613 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
614 running a limited number of proceses in parallel, a benefit is gained. The
615 amount varies with the platform hardware and load. The use of the option
616 queue_run_in_order means we cannot do this, as ordering becomes
619 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
620 had introduced a string-copy (for ensuring NUL-termination) which was not
621 appropriate for that case, which can include embedded NUL bytes in the
622 block of data. Investigation showed the copy to actually be needless, the
623 data being length-specified.
625 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
626 done during a receiving connection, and both used TLS, global info was
627 used rather than per-connection info for tracking the state of data
628 queued for transmission. This could result in a connection hang.
630 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
631 Previously, when delivering serveral messages down a single connection
632 only the first would provide a SIZE. This was due to the size information
633 not being properly tracked.
635 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
636 TAI (at 37 seconds currently), pretend to be in UTC for time-related
637 expansion and logging. Previously, spurious values such as a future
638 minute could be seen.
640 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
641 it could crash from a null-deref. This could also affect the
642 ${addresses: } operator and ${readsock } item.
644 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
645 message following a mime one, the variable was not reset.
647 JH/30 When an pipelined-connect fails at the first response, assume incorrect
648 cached capability (perhaps the peer reneged?) and immediately retry in
651 JH/31 Fix spurious detection of timeout while writing to transport filter.
653 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
654 an attempt to copy the string was made before checking it.
656 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
657 taint of the lookup key was maintained; we now regard the presence in the
658 filesystem as sufficient validation.
660 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
661 argument is supplied.
663 JH/35 The dsearch lookup now requires that the directory is an absolute path.
664 Previously this was not checked, and nonempty relative paths made an
665 access under Exim's current working directory.
667 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
668 Previously no event was raised.
670 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
671 parameter supplied by the sender MAIL FROM command. Previously it was
672 ignored, and only the check_spool_space option value for the required
675 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
676 the size of the signing public-key. Previously it was instead giving
677 the size of the signature hash.
679 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
680 the default. See the (new) dkim_verify_min_keysizes option.
682 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
683 and an ACL use a lookup for checking either the local_part or domain,
684 stale data could be accessed. Ensure that variable references are
685 dropped between messages.
687 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
688 by the client was not checked as pointing within response data before
689 being used. A malicious client could thus cause an out-of-bounds read and
690 possibly gain authentication. Fix by adding the check.
692 JH/42 Internationalisation: change the default for downconversion in the smtp
693 transport to be "if needed". Previously it was "as previously set" for
694 the message, which usually meant "if needed" for message-submission but
695 "no" for everything else. However, MTAs have been seen using SMTPUTF8
696 even when the envelope addresses did not need it, resulting in forwarding
697 failures to non-supporting MTAs. A downconvert in such cases will be
698 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
699 transport. The change does mean that addresses needing conversion will
700 be converted when previously a delivery failure would occur.
702 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
703 response was received it would be used unchecked in a fail-DSN, violating
704 standards on line-length limits. Truncate if needed.
706 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
707 doesn't work. (Additionally add a new main config option to configure the
714 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
715 side, to decode a TLS 1.3 packet capture.
717 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
718 Previously the default library behaviour applied, sending two, each in
721 JH/03 Debug output for ACL now gives the config file name and line number for
724 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
726 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
728 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
729 buffer overrun for (non-chunking) other transports.
731 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
732 TLS1.3, means that a server rejecting a client certificate is not visible
733 to the client until the first read of encrypted data (typically the
734 response to EHLO). Add detection for that case and treat it as a failed
735 TLS connection attempt, so that the normal retry-in-clear can work (if
736 suitably configured).
738 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
739 and/or domain. Found and fixed by Jason Betts.
741 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
742 configuration). If a CNAME target was not a wellformed name pattern, a
745 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
746 the OS reports them interleaved with other addresses.
748 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
749 used both for input and for a verify callout, both encrypted, SMTP
750 responses being sent by the server could be lost. This resulted in
751 dropped connections and sometimes bounces generated by a peer sending
754 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
755 string. Previously it was possible to cause undefined behaviour in a
756 library routine (usually a crash). Found by "zerons".
758 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
761 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
762 API was removed, so update to use the newer ones.
764 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
765 any timeout set, is taking a long time. Previously we would hang on to a
766 rotated logfile "forever" if the input was arriving with long gaps
767 (a previous attempt to fix addressed lack, for a long time, of initial
770 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
771 shared (NFS) environment. The length of the tempfile name is now
772 4 + 16 ("hdr.$message_exim_id") which might break on file
773 systems which restrict the file name length to lower values.
776 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
777 shared (NFS) environment.
779 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
780 did for all versions <4.90). Notably -M, -m, --invert, -I may be
783 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
784 on some platforms for bit 31.
786 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
787 to changes apparently associated with TLS1.3 handling some of the APIs
788 previously used were either nonfunctional or inappropriate. Strings
789 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
790 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
791 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
792 This affects log line X= elements, the $tls_{in,out}_cipher variables,
793 and the use of specific cipher names in the encrypted= ACL condition.
795 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
797 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
798 verification result was not updated unless hosts_require_ocsp applied.
800 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
801 queue_list_requires_admin set to false, non-admin users were denied the
804 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
805 directory-of-certs mode. Previously they were advertised despite the
808 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
809 A single TCP connection by a client will now hold a TLS connection open
810 for multiple message deliveries, by default. Previously the default was to
813 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
814 default. If built with the facility, DANE will be used. The facility
815 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
817 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
818 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
819 must be defined and you must still, unless you define DISABLE_TLS, manage
820 the the include-dir and library-file requirements that go with that
821 choice. Non-TLS builds are still supported.
823 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
826 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
827 default. If the platform supports and has the facility enabled, it will
828 be requested on all coneections.
830 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
831 controlled by the build-time option SUPPORT_PIPE_CONNECT.
833 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
835 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
836 success-DSN messages. Previously the From: header was always the default
837 one for these; the option was ignored.
839 JH/28 Fix the timeout on smtp response to apply to the whole response.
840 Previously it was reset for every read, so a teergrubing peer sending
841 single bytes within the time limit could extend the connection for a
842 long time. Credit to Qualsys Security Advisory Team for the discovery.
844 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
845 delivery address, which leaked information of the results of local
846 forwarding. Change to the original envelope recipient address, per
849 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
850 requested. Previously not bounce was generated and a log entry of
851 error ignored was made.
853 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
855 JH/32 Introduce a general tainting mechanism for values read from the input
856 channel, and values derived from them. Refuse to expand any tainted
857 values, to catch one form of exploit.
859 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
860 was unused and the unexpanded text used for the test. Found and
861 fixed by Ruben Jenster.
863 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
864 an attempt to use a TLS library read routine dereffed a nul pointer,
867 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
868 them in our smtp response.
870 JH/36 Have the general router option retry_use_local_part default to true when
871 any of the restrictive preconditions are set (to anything). Previously it
872 was only for check_local user. The change removes one item of manual
873 configuration which is required for proper retries when a remote router
874 handles a subset of addresses for a domain.
876 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
877 link count into consideration.
879 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
880 caused the extension of big_buffer, the following lines were ignored.
882 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
883 accordance with RFC 2308. Previously there was no expiry, so a longlived
884 receive process (eg. due to ACL delays) versus a short SOA value could
887 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
889 JH/39 Promote DMARC support to mainline.
891 JH/40 Bug 2452: Add a References: header to DSNs.
893 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
894 parameters. The relevant library call is documented as "Deprecated: This
895 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
896 3.6.0, DH parameters are negotiated following RFC7919."
898 HS/06 Change the default of dnssec_request_domains to "*"
900 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
901 carried on and emitted a BDAT command, even when PIPELINING was not
904 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
905 buffer was used for the filename, resulting in a trap when tainted
906 arguments (eg. $domain) were used.
908 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
909 recommended to avoid a possible server-load attack. The feature can be
910 re-enabled via the openssl_options main cofiguration option.
912 JH/45 local_scan API: documented the current smtp_printf() call. This changed
913 for version 4.90 - adding a "more data" boolean to the arguments.
914 Bumped the ABI version number also, this having been missed previously;
915 release versions 4.90 to 4.92.3 inclusive were effectively broken in
916 respect of usage of smtp_printf() by either local_scan code or libraries
917 accessed via the ${dlfunc } expansion item. Both will need coding
918 adjustment for any calls to smtp_printf() to match the new function
919 signature; a FALSE value for the new argument is always safe.
921 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
922 the file-offset (which the Linux syscall does, and exim expects); this
923 resulted in an indefinite loop.
925 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
926 to do ARC verification. The Authentication-Results: header line added
927 by the configuration then had no ARC item.
929 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
930 an incorrect mode was used for reading the result, resulting in it being
937 JH/01 Remove code calling the customisable local_scan function, unless a new
938 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
940 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
941 non-signal-safe functions being used.
943 JH/03 Bug 2269: When presented with a received message having a stupidly large
944 number of DKIM-Signature headers, disable DKIM verification to avoid
945 a resource-consumption attack. The limit is set at twenty.
947 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
948 report of oldest_pass in ${authres } in consequence, and separate out
949 some descriptions of reasons for verification fail.
951 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
952 files in the spool were present and unlocked. A queue-runner could spot
953 them, resulting in a duplicate delivery. Fix that by doing the unlock
954 after the unlink. Investigation by Tim Stewart. Take the opportunity to
955 add more error-checking on spoolfile handling while that code is being
958 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
959 No known attacks, no CVE, this is defensive hardening.
961 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
962 a queue-runner could start a delivery while other operations were ongoing.
963 Cutthrough delivery was a common victim, resulting in duplicate delivery.
964 Found and investigated by Tim Stewart. Fix by using the open message data
965 file handle rather than opening another, and not locally closing it (which
966 releases a lock) for that case, while creating the temporary .eml format
967 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
969 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
970 $sender_verify_failure/$recipient_verify_failure to "random".
972 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
975 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
976 Previously this would segfault.
978 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
981 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
982 like zero, since the resolver should be doing this for us, But we need one
983 as a CNAME but no MX presence gets the CNAME returned; we need to check
984 that doesn't point to an MX to declare it "no MX returned" rather than
985 "error, loop". A new main option is added so the older capability of
986 following some limited number of chain links is maintained.
988 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
990 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
991 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
992 now-deprecated earlier definitions used only specified the range up to TLS
993 1.2 (in the older-version library docs).
995 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
997 JH/15 Rework TLS client-side context management. Stop using a global, and
998 explicitly pass a context around. This enables future use of TLS for
999 connections to service-daemons (eg. malware scanning) while a client smtp
1000 connection is using TLS; with cutthrough connections this is quite likely.
1002 JH/16 Fix ARC verification to do AS checks in reverse order.
1004 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1006 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1007 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1008 Previously the "utf8" would be re-prepended for every additional message.
1010 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1011 Previously thery were accepted, resulting in issues when attempting to
1012 forward messages to a non-supporting MTA.
1014 PP/02 Let -n work with printing macros too, not just options.
1016 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1017 one parent address was copied, and bogus data was used at delivery-logging
1018 time. Either a crash (after delivery) or bogus log data could result.
1019 Discovery and analysis by Tim Stewart.
1021 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1022 Previously if the string ended mid-character, we did not insert the
1023 promised '?' replacement.
1025 PP/04 Documentation: current string operators work on bytes, not codepoints.
1027 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1028 should pack well giving a smaller memory footprint so better caching and
1029 therefore performance. Group the declarations where this can't be done so
1030 that the byte-sized flag variables are not interspersed among pointer
1031 variables, giving a better chance of good packing by the compiler.
1033 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1034 non-null, to avoid issues with sites running BATV. Previously reports were
1035 sent with an empty envelope sender so looked like bounces.
1037 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1038 The ignore_error flag wasn't being returned from the filter subprocess so
1039 was not set for later routers. Investigation and fix by Matthias Kurz.
1041 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1042 and a msg:complete for the whole, when a message is manually removed using
1043 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1045 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1046 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1047 hope that is portable enough.
1049 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1050 requiring a known-CA anchor certificate; make it now rely entirely on the
1051 TLSA as an anchor. Checking the name on the leaf cert against the name
1052 on the A-record for the host is still done for TA (but not for EE mode).
1054 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1055 included in delivery lines for non-proxied connections, when compiled with
1056 SUPPORT_SOCKS and running with proxy logging enabled.
1058 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1059 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1060 move the existing event to fire before the normal logging of message
1061 failure so that custom logging is bracketed by normal logging.
1063 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1064 msg:fail:internal event. Developement by Matthias Kurz.
1066 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1067 far too small for todays use of crypto signatures stored there. Go all
1068 the way to the max DNS message size of 64kB, even though this might be
1069 overmuch for IOT constrained device use.
1071 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1072 copy a string over itself. The library routine is documented as not
1073 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1075 JH/32 For main options check_spool_space and check_inode_space, where the
1076 platform supports 64b integers, support more than the previous 2^31 kB
1077 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1078 the previous G, M, k.
1080 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1081 $authenticated_fail_id variable on authentication failure. Previously
1084 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1085 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1086 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1087 GNUTLS_SEC_PARAM_MEDIUM.
1089 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1090 side. Previously we would continue as if no SNI had been received.
1092 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1093 "<" character, which should be a list-separator specification, we walked
1094 off past the nul-terimation.
1096 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1097 causes) even when the retry time is not yet met. Previously they were
1098 not, meaning that when (say) an account was over-quota and temp-rejecting,
1099 and multiple senders' messages were queued, only one sender would get
1100 notified on each configured delay_warning cycle.
1102 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1104 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1105 cipher-suites, an error can be left on the stack even for a succeeding
1106 accept; this results in impossible error messages when a later operation
1109 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1110 return error codes indicating retry. Under TLS1.3 this becomes required.
1112 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1113 it only wrote the new authenticators, resulting in a lack of tracking of
1114 peer changes of ESMTP extensions until the next cache flush.
1116 JH/41 Fix the loop reading a message header line to check for integer overflow,
1117 and more-often against header_maxsize. Previously a crafted message could
1118 induce a crash of the recive process; now the message is cleanly rejected.
1120 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1121 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1127 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1128 When redis_servers is set to a list of > 1 element, and the Redis servers
1129 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1130 case of MOVED into a DEFER case instead, thus moving the query onto the
1131 next server in the list. For a cluster of N elements, all N servers must
1132 be defined in redis_servers.
1134 GF/02 Catch and remove uninitialized value warning in exiqsumm
1135 Check for existence of @ARGV before looking at $ARGV[0]
1137 JH/01 Replace the store_release() internal interface with store_newblock(),
1138 which internalises the check required to safely use the old one, plus
1139 the allocate and data copy operations duplicated in both (!) of the
1140 extant use locations.
1142 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1143 modifier. This matches the restriction on the commandline.
1145 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1146 Previously only the last row was returned.
1148 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1149 we assumed that tags in the header were well-formed, and parsed the
1150 element content after inspecting only the first char of the tag.
1151 Assumptions at that stage could crash the receive process on malformed
1154 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1155 While running the DKIM ACL we operate on the Permanent memory pool so that
1156 variables created with "set" persist to the DATA ACL. Also (at any time)
1157 DNS lookups that fail create cache records using the Permanent pool. But
1158 expansions release any allocations made on the current pool - so a dnsdb
1159 lookup expansion done in the DKIM ACL releases the memory used for the
1160 DNS negative-cache, and bad things result. Solution is to switch to the
1161 Main pool for expansions.
1162 While we're in that code, add checks on the DNS cache during store_reset,
1163 active in the testsuite.
1164 Problem spotted, and debugging aided, by Wolfgang Breyha.
1166 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1167 When none of the hosts presented to a transport match an already-open
1168 connection, close it and proceed with the list. Previously we would
1169 queue the message. Spotted by Lena with Yahoo, probably involving
1172 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1173 Previously a spurious "250 OK id=" response was appended to the proper
1176 JH/08 The "support for" informational output now, which built with Content
1177 Scanning support, has a line for the malware scanner interfaces compiled
1178 in. Interface can be individually included or not at build time.
1180 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1181 by the template makefile "src/EDITME". The "STREAM" support for an older
1182 ClamAV interface method is removed.
1184 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1185 rows affected is given instead).
1187 JH/11 The runtime Berkeley DB library version is now additionally output by
1188 "exim -d -bV". Previously only the compile-time version was shown.
1190 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1191 SMTP connection. Previously, when one had more recipients than the
1192 first, an abortive onward connection was made. Move to full support for
1193 multiple onward connections in sequence, handling cutthrough connection
1194 for all multi-message initiating connections.
1196 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1197 routers. Previously, a multi-recipient message would fail to match the
1198 onward-connection opened for the first recipient, and cause its closure.
1200 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1201 a timeout on read on a GnuTLS initiating connection, resulting in the
1202 initiating connection being dropped. This mattered most when the callout
1203 was marked defer_ok. Fix to keep the two timeout-detection methods
1206 JH/15 Relax results from ACL control request to enable cutthrough, in
1207 unsupported situations, from error to silently (except under debug)
1208 ignoring. This covers use with PRDR, frozen messages, queue-only and
1211 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1213 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1214 metadata, resulting in a crash in free().
1216 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1217 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1218 Broken also in d185889f4, with init system revamp.
1220 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1221 Previously we abruptly closed the connection after reading a malware-
1222 found indication; now we go on to read the "scan ok" response line,
1225 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1226 ACL. Previously, a crash would result.
1228 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1229 macro text after a replacement (previously it was only once per line) and
1230 by skipping builtin macros when searching for an uppercase lead character.
1232 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1233 for the build is renamed.
1235 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1236 was allocated for every new TLS startup, meaning one per message. Fix
1237 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1239 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1240 reported the original. Fix to report (as far as possible) the ACL
1241 result replacing the original.
1243 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1244 OpenSSL. Certificate information is loaded for every new TLS startup,
1245 and the resources needed to be freed.
1247 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1249 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1252 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1253 DATA response info to the (existing) per-recipient response info for
1254 the "C=" log element. It can have useful tracking info from the
1255 destination system. Patch from Simon Arlott.
1257 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1258 length value. Previously this would segfault.
1260 HS/02 Support Avast multiline protoocol, this allows passing flags to
1261 newer versions of the scanner.
1263 JH/28 Ensure that variables possibly set during message acceptance are marked
1264 dead before release of memory in the daemon loop. This stops complaints
1265 about them when the debug_store option is enabled. Discovered specifically
1266 for sender_rate_period, but applies to a whole set of variables.
1267 Do the same for the queue-runner and queue-list loops, for variables set
1268 from spool message files. Do the same for the SMTP per-message loop, for
1269 certain variables indirectly set in ACL operations.
1271 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1272 as a multi-recipient message from a mailinglist manager). The coding had
1273 an arbitrary cutoff number of characters while checking for more input;
1274 enforced by writing a NUL into the buffer. This corrupted long / fast
1275 input. The problem was exposed more widely when more pipelineing of SMTP
1276 responses was introduced, and one Exim system was feeding another.
1277 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1278 receiving system, and refused recipients seen by the sending system
1279 (propating to people being dropped from mailing lists).
1280 Discovered and pinpointed by David Carter.
1282 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1283 replaced by the ${authresults } expansion.
1285 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1287 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1288 allows proper process termination in container environments.
1290 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1291 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1293 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1294 and "err_perm", deprecated since 4.83 when the RFC-defined words
1295 "temperror" and "permerror" were introduced.
1297 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1298 transport-filters or DKIM-signing. The restriction was lost in the
1299 consolidation of verify-callout and delivery SMTP handling.
1300 Extend the restriction to also cover ARC-signing.
1302 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1303 in defer=pass mode supply a 450 to the initiator. Previously the message
1306 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1307 tls_require_ciphers is used as before.
1309 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1310 "pass_unscanned". Only tmpfails from the scanner are written to
1311 the paniclog, as they may require admin intervention (permission
1312 denied, license issues). Other scanner errors (like decompression
1313 bombs) do not cause a paniclog entry.
1315 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1316 Previously it was possible to log spurious information in receive log
1319 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1320 triggered odd behaviour from Outlook Express clients.
1322 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1325 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1326 since the IETF WG has not yet settled on that versus the original
1327 "bare" representation.
1329 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1330 Previously the millisecond value corrupted the output.
1331 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1332 corrupted the output.
1338 JH/01 Rework error string handling in TLS interface so that the caller in
1339 more cases is responsible for logging. This permits library-sourced
1340 string to be attached to addresses during delivery, and collapses
1341 pairs of long lines into single ones.
1343 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1344 during configuration. Wildcards are allowed and expanded.
1346 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1347 This permits better logging.
1349 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1350 do not close it down and have the child transport start it up again on
1351 the passed-on TCP connection. Instead, proxy the child (and any
1352 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1353 affected: the continued delivery log lines do not have any DNSSEC, TLS
1354 Certificate or OCSP information. TLS cipher information is still logged.
1356 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1357 identical IP addresses on different listening ports. Will also affect
1360 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1361 add noisy ifdef guards to special-case this sillyness.
1362 Patch from Bernd Kuhls.
1364 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1365 than 255 are no longer allowed.
1367 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1368 Disable the session-cache too, which might reduce our load. Since we
1369 currrectly use a new context for every connection, both as server and
1370 client, there is no benefit for these.
1371 GnuTLS appears to not support tickets server-side by default (we don't
1372 call gnutls_session_ticket_enable_server()) but client side is enabled
1373 by default on recent versions (3.1.3 +) unless the PFS priority string
1376 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1377 <https://reproducible-builds.org/specs/source-date-epoch/>.
1379 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1380 the check for any unsuccessful recipients did not notice the limit, and
1381 erroneously found still-pending ones.
1383 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1384 MSG_MORE. Only in-clear (not on TLS connections).
1386 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1387 transport-filter is involved we can buffer the headers in memory for
1388 creating the signature, and read the spool data file once for the
1389 signature and again for transmission.
1391 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1392 4.77 as the kernel support then wasn't solid, having issues in 64bit
1393 mode. Now, it's been long enough. Add support for FreeBSD also.
1395 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1396 case where the routing stage had gathered several addresses to send to
1397 a host before calling the transport for the first, we previously failed
1398 to close down TLS in the old transport process before passing the TCP
1399 connection to the new process. The new one sent a STARTTLS command
1400 which naturally failed, giving a failed delivery and bloating the retry
1401 database. Investigation and fix prototype from Wolfgang Breyha.
1403 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1404 false-negatives in the check that the sender had not preempted a response
1405 or prompt from Exim (running as a server), due to that code's lack of
1406 awareness of the SMTP input buffering.
1408 PP/04 Add commandline_checks_require_admin option.
1409 Exim drops privileges sanely, various checks such as -be aren't a
1410 security problem, as long as you trust local users with access to their
1411 own account. When invoked by services which pass untrusted data to
1412 Exim, this might be an issue. Set this option in main configuration
1413 AND make fixes to the calling application, such as using `--` to stop
1416 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1417 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1418 responses to those, into a single TLS record each way (this usually means
1419 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1422 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1423 affects you only if you're dancing at the edge of the param size limits.
1424 If you are, and this message makes sense to you, then: raise the
1425 configured limit or use OpenSSL 1.1. Nothing we can do for older
1428 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1429 cmdline element to get the documented default one. Previously it was
1432 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1433 are unsuable for non-message-specific reasons - eg. connection timeout,
1434 banner-time rejection.
1436 JH/16 Fix logging of delivery remote port, when specified by router, under
1439 PP/06 Repair manualroute's ability to take options in any order, even if one
1440 is the name of a transport.
1443 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1445 JH/17 Change the list-building routines interface to use the expanding-string
1446 triplet model, for better allocation and copying behaviour.
1448 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1449 Previously it was constructed the first time a possibly-matching string
1450 was met in the configuration file input during startup; now it is done
1453 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1454 compatible one, to avoid the (poorly documented) possibility of a config
1455 file in the working directory redirecting the DB files, possibly correpting
1456 some existing file. CVE-2017-10140 assigned for BDB.
1458 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1459 cache-hot. Previously, although the result was properly cached, the
1460 initial verify call returned a defer.
1462 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1463 the main verify for receipient in uncached-mode.
1465 JH/22 Retire historical build files to an "unsupported" subdir. These are
1466 defined as "ones for which we have no current evidence of testing".
1468 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1469 if present. Previously it was ignored.
1471 JH/24 Start using specified-initialisers in C structure init coding. This is
1472 a C99 feature (it's 2017, so now considered safe).
1474 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1475 if was a fixed-sized field and bitmask ops via macros; it is now more
1478 PP/07 GitHub PR 56: Apply MariaDB build fix.
1479 Patch provided by Jaroslav Škarvada.
1481 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1482 during Coverity cleanups [4.87 JH/47]
1483 Diagnosis and fix provided by Michael Fischer v. Mollard.
1485 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1486 the right size to place the terminating semicolon on its own folded
1487 line, the header hash was calculated to an incorrect value thanks to
1488 the (relaxed) space the fold became.
1490 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1491 and confused the parent.
1493 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1494 which could crash as a result. This could lead to undeliverable messages.
1496 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1499 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1500 matters on fast-turnover and PID-randomising systems, which were getting
1501 out-of-order delivery.
1503 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1504 a possibly-overlapping copy. The symptom was that "Remote host closed
1505 connection in response to HELO" was logged instead of the actual 4xx
1508 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1509 Previously only that bufferd was discarded, resulting in SYMTP command
1512 JH/32 DKIM: when a message has multiple signatures matching an identity given
1513 in dkim_verify_signers, run the dkim acl once for each. Previously only
1514 one run was done. Bug 2189.
1516 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1517 "panic the current process" to "deliberately defer". The panic log is
1518 still written with the problem list name; the mail and reject logs now
1519 get a temp-reject line for the message that was being handled, saying
1520 something like "domains check lookup or other defer". The SMTP 451
1521 message is still "Temporary local problem".
1523 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1524 A crafted sequence of BDAT commands could result in in-use memory beeing
1525 freed. CVE-2017-16943.
1527 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1528 from SMTP input. Previously it was always done; now only done for DATA
1529 and not BDAT commands. CVE-2017-16944.
1531 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1532 to the message (such as an overlong header line). Previously this was
1533 not done and we did not exit BDAT mode. Followon from the previous item
1534 though a different problem.
1540 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1541 than -2003 did; needs libidn2 in addition to libidn.
1543 JH/02 The path option on a pipe transport is now expanded before use.
1545 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1546 Patch provided by "Björn", documentation fix added too.
1548 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1549 missing a wire-to-host endian conversion.
1551 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1552 close after a BDAT command line could be taken as a following command,
1553 giving a synch failure. Fix by only checking for synch immediately
1554 before acknowledging the chunk.
1556 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1557 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1558 macro. Patches provided by Josh Soref.
1560 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1561 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1562 by IANA as a service extension. However, John Klensin suggests that we
1565 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1566 the dkim code may be unix-mode line endings rather than smtp wire-format
1567 CRLF, so prepend a CR to any bare LF.
1569 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1570 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1572 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1573 the first were themselves being wrongly included in the feed into dkim
1574 processing; with most chunk sizes in use this resulted in an incorrect
1575 body hash calculated value.
1577 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1578 DKIM signature block, for verification. Although advised against by
1579 standards it is specifically not ruled illegal.
1581 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1583 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1584 missing a body hash (the bh= tag).
1586 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1587 It seems that HAProxy sends the Proxy Protocol information in clear and
1588 only then does a TLS startup, so do the same.
1590 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1591 TCP connections (such as for Spamd) unless the daemon successfully set
1592 Fast Open mode on its listening sockets. This fixes breakage seen on
1593 too-old kernels or those not configured for Fast Open, at the cost of
1594 requiring both directions being enabled for TFO, and TFO never being used
1595 by non-daemon-related Exim processes.
1597 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1598 endings, at least on the first header line. Try to canonify any that get
1599 past that check, despite the cost.
1601 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1602 now limited to an arbitrary five deep, while parsing addresses with the
1603 strip_excess_angle_brackets option enabled.
1605 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1606 instead leave the unprompted TLS handshake in socket buffer for the
1607 TLS library to consume.
1609 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1611 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1613 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1614 region in one operation, for speed, and this leaves assigned pointers
1615 dangling. Add checks run only under the testsuite which checks all
1616 variables at a store-reset and panics on a dangling pointer; add code
1617 explicitly nulling out all the variables discovered. Fixes one known
1618 bug: a transport crash, where a dangling pointer for $sending_ip_address
1619 originally assigned in a verify callout, is re-used.
1621 PP/06 Drop '.' from @INC in various Perl scripts.
1623 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1625 PP/08 Reduce a number of compilation warnings under clang; building with
1626 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1627 should be warning-free.
1629 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1631 HS/01 Fix portability problems introduced by PP/08 for platforms where
1632 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1634 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1635 chunk. This allows us to accept broken chunked messages. We need a more
1636 general solution here.
1638 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1639 already-broken messages in the queue.
1641 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1643 JH/19 Fix reference counting bug in routing-generated-address tracking.
1649 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1650 supports it and a size is available (ie. the sending peer gave us one).
1652 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1653 years of being deprecated). The replacements are the ACLs
1654 acl_smtp_mime and acl_not_smtp_mime.
1656 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1657 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1658 or even an in-clear connection were permitted. Now, if the host lookup
1659 was dnssec and dane was requested then the host is only used if the
1660 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1661 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1662 if one fails this test.
1663 This means that a poorly-configured remote DNS will make it incommunicado;
1664 but it protects against a DNS-interception attack on it.
1666 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1667 non-noisy when a race steals the message being considered.
1669 JH/05 If main configuration option tls_certificate is unset, generate a
1670 self-signed certificate for inbound TLS connections.
1672 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1673 in rewrites and routers.
1675 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1676 and logged a warning sing 4.83; now they are a configuration file error.
1678 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1679 (lacking @domain). Apply the same qualification processing as RCPT.
1681 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1683 JH/10 Support ${sha256:} applied to a string (as well as the previous
1686 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1687 a cutthrough deliver is pending, as we always want to make a connection.
1688 This also avoids re-routing the message when later placing the cutthrough
1689 connection after a verify cache hit.
1690 Do not update it with the verify result either.
1692 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1693 when routing results in more than one destination address.
1695 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1696 signing (which inhibits the cutthrough capability). Previously only
1697 the presence of an option was tested; now an expansion evaluating as
1698 empty is permissible (obviously it should depend only on data available
1699 when the cutthrough connection is made).
1701 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1702 the relevant preceding SMTP command did not note the pipelining mode.
1704 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1705 Previously they were not counted.
1707 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1708 as one having no matching records. Previously we deferred the message
1709 that needed the lookup.
1711 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1712 distinguished as "(=".
1714 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1715 for missing MX records. Previously it only worked for missing A records.
1717 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1719 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1720 after the data-go-ahead and data-ack. Patch from Jason Betts.
1722 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1723 even for a "none" policy. Patch from Tony Meyer.
1725 JH/22 Fix continued use of a connection for further deliveries. If a port was
1726 specified by a router, it must also match for the delivery to be
1729 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1730 When one of the recipients of a message was unsuitable for the connection
1731 (has no matching addresses), we lost track of needing to mark it
1732 deferred. As a result mail would be lost.
1734 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1736 JH/25 Decoding ACL controls is now done using a binary search; the source code
1737 takes up less space and should be simpler to maintain. Merge the ACL
1738 condition decode tables also, with similar effect.
1740 JH/26 Fix problem with one_time used on a redirect router which returned the
1741 parent address unchanged. A retry would see the parent address marked as
1742 delivered, so not attempt the (identical) child. As a result mail would
1745 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1746 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1747 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1750 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1751 Main config options check_{log,spool}_{inodes,space} are now
1752 100 inodes, 10MB unless set otherwise in the configuration.
1754 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1755 Previously it only applied to the main-section connection policy
1758 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1760 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1761 by me. Added RFC7919 DH primes as an alternative.
1763 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1764 are not in the system include path.
1766 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1767 GnuTLS, when a session startup failed (eg because the client disconnected)
1768 Exim did stdio operations after fclose. This was exposed by a recent
1769 change which nulled out the file handle after the fclose.
1771 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1772 signed directly by the cert-signing cert, rather than an intermediate
1773 OCSP-signing cert. This is the model used by LetsEncrypt.
1775 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1777 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1778 an incoming connection.
1780 HS/02 Bug 1802: Do not half-close the connection after sending a request
1783 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1784 fallback to "prime256v1".
1786 JH/34 SECURITY: Use proper copy of DATA command in error message.
1787 Could leak key material. Remotely exploitable. CVE-2016-9963.
1793 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1794 and 3.4.4 - once the server is enabled to respond to an OCSP request
1795 it does even when not requested, resulting in a stapling non-aware
1796 client dropping the TLS connection.
1798 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1799 support variable-length bit vectors. No functional change.
1801 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1802 The I= interface field on outgoing lines is now after the H= remote
1803 host field, same as incoming lines. There is a separate
1804 outgoing_interface log selector which allows you to disable the
1807 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1808 If not running log_selector +smtp_connection the mainlog would be held
1809 open indefinitely after a "too many connections" event, including to a
1810 deleted file after a log rotate. Leave the per net connection logging
1811 leaving it open for efficiency as that will be quickly detected by the
1812 check on the next write.
1814 HS/01 Bug 1671: Fix post transport crash.
1815 Processing the wait-<transport> messages could crash the delivery
1816 process if the message IDs didn't exist for some reason. When
1817 using 'split_spool_directory=yes' the construction of the spool
1818 file name failed already, exposing the same netto behaviour.
1820 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1821 mime_regex ACL conditions.
1823 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1824 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1825 to HELO, local diagnostic string.
1827 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1828 log line to debug. Even when configured with a tls authenticator many
1829 client connections are expected to not authenticate in this way, so
1830 an authenticate fail is not an error.
1832 HS/02 Add the Exim version string to the process info. This way exiwhat
1833 gives some more detail about the running daemon.
1835 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1836 matter for fast-change records such as DNSBLs.
1838 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1839 retry record, even if constant. There may be multiple transports with
1840 different interface settings and the retry behaviour needs to be kept
1843 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1845 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1847 JH/10 Bug 840: fix log_defer_output option of pipe transport
1849 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1850 hosts_randomize. This matters a lot when combined with mua_wrapper.
1852 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1853 ${quote_pgsql:<string>} operator.
1855 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1857 JH/14 Change header file name for freeradius-client. Relevant if compiling
1858 with Radius support; from the Gentoo tree and checked under Fedora.
1860 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1862 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1863 option transport_filter, meaning no filtering.
1865 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1867 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1868 defaults to "*" (all hosts). The variable is now available when not built
1869 with TLS, default unset, mainly to enable keeping the testsuite sane.
1870 If a server certificate is not supplied (via tls_certificate) an error is
1871 logged, and clients will find TLS connections fail on startup. Presumably
1872 they will retry in-clear.
1873 Packagers of Exim are strongly encouraged to create a server certificate
1874 at installation time.
1876 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1877 with the $config_file variable.
1879 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1880 in transport context, after the attempt, and per-recipient. The latter type
1881 is per host attempted. The event data is the error message, and the errno
1882 information encodes the lookup type (A vs. MX) used for the (first) host,
1883 and the trailing two digits of the smtp 4xx response.
1885 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1886 to write to mainlog (or rejectlog, paniclog) in the window between file
1887 creation and permissions/ownership being changed. Particularly affects
1888 installations where exicyclog is run as root, rather than exim user;
1889 result is that the running daemon panics and dies.
1891 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1893 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1894 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1895 "pri" and "weight". Note that the previous implicit priority given by the
1896 list order is no longer honoured.
1898 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1899 for DKIM processing.
1901 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1902 by defining SUPPORT_SOCKS.
1904 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1905 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1906 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1907 variables are renamed to proxy_{local,external}_{address,port}.
1909 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1910 for a build by defining SUPPORT_I18N
1912 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1913 of the query string, and make ${quote_redis:} do that quoting.
1915 JH/29 Move Events support from Experimental to mainline, enabled by default
1916 and removable for a build by defining DISABLE_EVENT.
1918 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1920 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1921 cached by the daemon.
1923 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1924 by defining LOOKUP_REDIS. The libhiredis library is required.
1926 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1927 keys are given for lookup.
1929 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1930 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1931 only supported when built with TLS support. The PolarSSL SHA routines
1932 are still used when the TLS library is too old for convenient support.
1934 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1935 openssl_options), for security. OpenSSL forces this from version 1.1.0
1936 server-side so match that on older versions.
1938 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1939 allocation for $value could be released as the expansion processing
1940 concluded, but leaving the global pointer active for it.
1942 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1943 and to use the domains and local_parts ACL conditions.
1945 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1946 incorrectly not doubled on cutthrough transmission, hence seen as a
1947 body-termination at the receiving system - resulting in truncated mails.
1948 Commonly the sender saw a TCP-level error, and retransmitted the message
1949 via the normal store-and-forward channel. This could result in duplicates
1950 received - but deduplicating mailstores were liable to retain only the
1951 initial truncated version.
1953 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1955 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1957 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1958 we're in there, support oversigning also; bug 1309.
1960 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1962 HS/04 Add support for keep_environment and add_environment options.
1964 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1965 either intentional arithmetic overflow during PRNG, or testing config-
1968 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1969 delivery resulted in actual delivery. Cancel cutthrough before DATA
1972 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1973 rejects a recipient: pass the reject to the originator.
1975 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1976 Many were false-positives and ignorable, but it's worth fixing the
1979 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1980 for the new environment-manipulation done at startup. Move the routines
1981 from being local to tls.c to being global via the os.c file.
1983 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1984 an extract embedded as result-arg for a map, the first arg for extract
1985 is unavailable so we cannot tell if this is a numbered or keyed
1986 extraction. Accept either.
1992 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1995 JH/02 The smtp transport option "multi_domain" is now expanded.
1997 JH/03 The smtp transport now requests PRDR by default, if the server offers
2000 JH/04 Certificate name checking on server certificates, when exim is a client,
2001 is now done by default. The transport option tls_verify_cert_hostnames
2002 can be used to disable this per-host. The build option
2003 EXPERIMENTAL_CERTNAMES is withdrawn.
2005 JH/05 The value of the tls_verify_certificates smtp transport and main options
2006 default to the word "system" to access the system default CA bundle.
2007 For GnuTLS, only version 3.0.20 or later.
2009 JH/06 Verification of the server certificate for a TLS connection is now tried
2010 (but not required) by default. The verification status is now logged by
2011 default, for both outbound TLS and client-certificate supplying inbound
2014 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2017 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2018 Status Notification (bounce) messages are now MIME format per RFC 3464.
2019 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2020 under the control of the dsn_advertise_hosts option, and routers may
2021 have a dsn_lasthop option.
2023 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2024 default, modifiable by a malware= option. The list separator for
2025 the options can now be changed in the usual way. Bug 68.
2027 JH/10 The smtp_receive_timeout main option is now expanded before use.
2029 JH/11 The incoming_interface log option now also enables logging of the
2030 local interface on delivery outgoing connections.
2032 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2033 if the interface and destination host and port all match.
2035 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2038 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2039 Patch from Andrew Lewis.
2041 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2042 now supports optional time-restrictions, weighting, and priority
2043 modifiers per server. Patch originally by <rommer@active.by>.
2045 JH/16 The spamd_address main option now supports a mixed list of local
2046 and remote servers. Remote servers can be IPv6 addresses, and
2047 specify a port-range.
2049 JH/17 Bug 68: The spamd_address main option now supports an optional
2050 timeout value per server.
2052 JH/18 Bug 1581: Router and transport options headers_add/remove can
2053 now have the list separator specified.
2055 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2058 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2061 JH/21 Support for the A6 type of dns record is withdrawn.
2063 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2064 rather than the verbs used.
2066 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2067 from 255 to 1024 chars.
2069 JH/24 Verification callouts now attempt to use TLS by default.
2071 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2072 are generic router options now. The defaults didn't change.
2074 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2075 Original patch from Alexander Shikoff, worked over by JH.
2077 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2078 files if ZCAT_COMMAND is not executable.
2080 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2082 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2084 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2085 Normally benign, it bites when the pair was led to by a CNAME;
2086 modern usage is to not canonicalize the domain to a CNAME target
2087 (and we were inconsistent anyway for A-only vs AAAA+A).
2089 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2091 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2092 when evaluating $sender_host_dnssec.
2094 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2095 $sender_helo_dnssec variable.
2097 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2099 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2101 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2103 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2104 documented as working, but never had. Support all but $spam_report.
2106 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2107 added for tls authenticator.
2109 HS/03 Add perl_taintmode main config option
2115 TL/01 When running the test suite, the README says that variables such as
2116 no_msglog_check are global and can be placed anywhere in a specific
2117 test's script, however it was observed that placement needed to be near
2118 the beginning for it to behave that way. Changed the runtest perl
2119 script to read through the entire script once to detect and set these
2120 variables, reset to the beginning of the script, and then run through
2121 the script parsing/test process like normal.
2123 TL/02 The BSD's have an arc4random API. One of the functions to induce
2124 adding randomness was arc4random_stir(), but it has been removed in
2125 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2126 function when detected.
2128 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2129 cause callback expansion.
2131 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2132 syntax errors in an expansion can be treated as a string instead of
2133 logging or causing an error, due to the internal use of bool_lax
2134 instead of bool when processing it.
2136 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2137 server certificates when making smtp deliveries.
2139 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2141 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2143 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2145 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2146 Merged patch from Sebastian Wiedenroth.
2148 JH/05 Fix results-pipe from transport process. Several recipients, combined
2149 with certificate use, exposed issues where response data items split
2150 over buffer boundaries were not parsed properly. This eventually
2151 resulted in duplicates being sent. This issue only became common enough
2152 to notice due to the introduction of connection certificate information,
2153 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2155 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2156 size buffer was used, resulting in syntax errors when an expansion
2159 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2160 version 3.3.6 or later.
2162 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2163 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2164 both become "event_action", the variables become $event_name, $event_data
2165 and $event_defer_errno. There is a new variable $verify_mode, usable in
2166 routers, transports and related events. The tls:cert event is now also
2167 raised for inbound connections, if the main configuration event_action
2170 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2171 early OCSP support, but no stapling (appears to be less than 1.0.0).
2173 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2174 server certificate names available under the smtp transport option
2175 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2178 JH/10 Time-related extraction expansions from certificates now use the main
2179 option "timezone" setting for output formatting, and are consistent
2180 between OpenSSL and GnuTLS compilations. Bug 1541.
2182 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2183 encoded parameter in the incoming message. Bug 1558.
2185 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2186 include certificate info, eximon was claiming there were spoolfile
2189 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2191 JH/14 Log delivery-related information more consistently, using the sequence
2192 "H=<name> [<ip>]" wherever possible.
2194 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2195 are problematic for Debian distribution, omit them from the release
2198 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2200 JH/16 Fix string representation of time values on 64bit time_t architectures.
2203 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2204 output list separator was used.
2209 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2210 checkers that were complaining about end of non-void function with no
2213 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2214 This was a regression introduced in 4.83 by another bugfix.
2216 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2218 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2219 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2225 TF/01 Correctly close the server side of TLS when forking for delivery.
2227 When a message was received over SMTP with TLS, Exim failed to clear up
2228 the incoming connection properly after forking off the child process to
2229 deliver the message. In some situations the subsequent outgoing
2230 delivery connection happened to have the same fd number as the incoming
2231 connection previously had. Exim would try to use TLS and fail, logging
2232 a "Bad file descriptor" error.
2234 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2235 utilities have not been installed.
2237 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2238 temporary space as the ACL may create new global variables.
2240 TL/01 LDAP support uses per connection or global context settings, depending
2241 upon the detected version of the libraries at build time.
2243 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2244 to extract and use the src ip:port in logging and expansions as if it
2245 were a direct connection from the outside internet. PPv2 support was
2246 updated based on HAProxy spec change in May 2014.
2248 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2250 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2251 Properly escape header and check for NULL return.
2253 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2256 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2258 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2259 characters in header names, implemented as a verify condition.
2260 Contributed by Michael Fischer v. Mollard.
2262 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2263 results permerror and temperror. Previous values are deprecated but
2264 still accepted. In a future release, err_perm and err_temp will be
2265 completely removed, which will be a backward incompatibility if the
2266 ACL tests for either of these two old results. Patch contributed by
2267 user bes-internal on the mailing list.
2269 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2271 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2272 selectors, in both main and reject logs.
2274 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2277 JH/07 Add malware type "sock" for talking to simple daemon.
2279 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2281 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2282 routers/transports under cutthrough routing.
2284 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2285 numbers. Touch up "bool" conditional to keep the same definition.
2287 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2289 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2291 JH/12 Expand items in router/transport headers_add or headers_remove lists
2292 individually rather than the list as a whole. Bug 1452.
2294 Required for reasonable handling of multiple headers_ options when
2295 they may be empty; requires that headers_remove items with embedded
2296 colons must have them doubled (or the list-separator changed).
2298 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2299 view the policy declared in the DMARC record. Currently, $dmarc_status
2300 is a combined value of both the record presence and the result of the
2303 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2305 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2306 dnslookup router and the smtp transport (applying to the forward
2309 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2310 of ldap servers used for a specific lookup. Patch provided by Heiko
2313 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2314 New variable $lookup_dnssec_authenticated for observability.
2316 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2317 Patch submitted by Lars Timman.
2319 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2321 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2322 Requires trusted mode and valid format message id, aborts otherwise.
2323 Patch contributed by Heiko Schlichting.
2325 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2326 certextract with support for various fields. Bug 1358.
2328 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2329 is requested by default, modifiable by smtp transport option
2332 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2333 operate on certificate variables to give certificate fingerprints
2334 Also new ${sha256:cert_variable}.
2336 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2338 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2339 Christian Aistleitner.
2341 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2343 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2344 file. Patch from Wolfgang Breyha.
2346 JH/25 Expand the coverage of the delivery $host and $host_address to
2347 client authenticators run in verify callout. Bug 1476.
2349 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2350 align with daemon_smtp_ports. Bug 72.
2352 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2353 support and error reporting did not work properly.
2355 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2356 and is readable. Patch from Andrew Colin Kissa.
2358 TL/14 Enhance documentation of ${run expansion and how it parses the
2359 commandline after expansion, particularly in the case when an
2360 unquoted variable expansion results in an empty value.
2362 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2364 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2365 by renaming away from it. Addresses GH issue 15, reported by
2368 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2370 TL/15 SECURITY: prevent double expansion in math comparison functions
2371 (can expand unsanitized data). Not remotely exploitable.
2378 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2380 PP/02 Make -n do something, by making it not do something.
2381 When combined with -bP, the name of an option is not output.
2383 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2386 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2387 $sender_host_name and config options to manage this, and basic check
2390 PP/05 DSCP support for outbound connections and control modifier for inbound.
2392 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2393 (Only plugin which currently uses this is kerberos4, which nobody should
2394 be using, but we should make it available and other future plugins might
2395 conceivably use it, even though it would break NAT; stuff *should* be
2396 using channel bindings instead).
2398 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2399 name; added for Sendmail compatibility; requires admin caller.
2400 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2401 just ignore it); requires trusted caller.
2402 Also parse but ignore: -Ac -Am -X<logfile>
2405 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2407 TL/02 Add +smtp_confirmation as a default logging option.
2409 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2410 Patch by Magnus Holmgren from 2007-02-20.
2412 TL/04 Bugzilla 1281 - Spec typo.
2413 Bugzilla 1283 - Spec typo.
2414 Bugzilla 1290 - Spec grammar fixes.
2416 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2418 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2420 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2421 mailing list by by Dmitry Isaikin.
2423 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2425 JH/02 Support "G" suffix to numbers in ${if comparisons.
2427 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2429 NM/01 Bugzilla 1197 - Spec typo
2430 Bugzilla 1196 - Spec examples corrections
2432 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2434 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2435 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2438 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2439 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2441 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2442 "acl {{name}{arg}...}", and optional args on acl condition
2445 JH/05 Permit multiple router/transport headers_add/remove lines.
2447 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2449 JH/07 Avoid using a waiting database for a single-message-only transport.
2450 Performance patch from Paul Fisher. Bugzilla 1262.
2452 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2455 JH/09 Add $headers_added variable, with content from use of ACL modifier
2456 add_header (but not yet added to the message). Bugzilla 199.
2458 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2459 Pulled from Bugzilla 817 by Wolfgang Breyha.
2461 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2463 (nb: this is the same fix as in Exim 4.80.1)
2465 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2468 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2471 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2473 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2474 advertise SMTP AUTH mechanism to us, instead of a generic
2475 protocol violation error. Also, make Exim more robust to bad
2476 data from the Dovecot auth socket.
2478 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2480 When a queue runner is handling a message, Exim first routes the
2481 recipient addresses, during which it prunes them based on the retry
2482 hints database. After that it attempts to deliver the message to
2483 any remaining recipients. It then updates the hints database using
2486 So if a recipient address works intermittently, it can get repeatedly
2487 deferred at routing time. The retry hints record remains fresh so the
2488 address never reaches the final cutoff time.
2490 This is a fairly common occurrence when a user is bumping up against
2491 their storage quota. Exim had some logic in its local delivery code
2492 to deal with this. However it did not apply to per-recipient defers
2493 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2495 This change adds a proper retry rule check during routing so that the
2496 final cutoff time is checked against the message's age. We only do
2497 this check if there is an address retry record and there is not a
2498 domain retry record; this implies that previous attempts to handle
2499 the address had the retry_use_local_parts option turned on. We use
2500 this as an approximation for the destination being like a local
2501 delivery, as in LMTP.
2503 I suspect this new check makes the old local delivery cutoff check
2504 redundant, but I have not verified this so I left the code in place.
2506 TF/02 Correct gecos expansion when From: is a prefix of the username.
2508 Test 0254 submits a message to Exim with the header
2512 When I ran the test suite under the user fanf2, Exim expanded
2513 the header to contain my full name, whereas it should have added
2514 a Resent-Sender: header. It erroneously treats any prefix of the
2515 username as equal to the username.
2517 This change corrects that bug.
2519 GF/01 DCC debug and logging tidyup
2520 Error conditions log to paniclog rather than rejectlog.
2521 Debug lines prefixed by "DCC: " to remove any ambiguity.
2523 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2525 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2526 Bug spotted by Jeremy Harris; was flawed since initial commit.
2527 Would have resulted in OCSP responses post-SNI triggering an Exim
2528 NULL dereference and crash.
2530 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2532 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2533 Bug detection, analysis and fix by Samuel Thibault.
2534 Bugzilla 1331, Debian bug #698092.
2536 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2538 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2539 Server implementation by Todd Lyons, client by JH.
2540 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2541 config variable "prdr_enable" controls whether the server
2542 advertises the facility. If the client requests PRDR a new
2543 acl_data_smtp_prdr ACL is called once for each recipient, after
2544 the body content is received and before the acl_smtp_data ACL.
2545 The client is controlled by both of: a hosts_try_prdr option
2546 on the smtp transport, and the server advertisement.
2547 Default client logging of deliveries and rejections involving
2548 PRDR are flagged with the string "PRDR".
2550 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2551 fclose(). Diagnosis by Todd Lyons.
2553 PP/17 Update configure.default to handle IPv6 localhost better.
2554 Patch by Alain Williams (plus minor tweaks).
2557 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2558 This is now consistent with GnuTLS, and is now documented: the
2559 previous undocumented portable approach to treating the option as
2560 unset was to force an expansion failure. That still works, and
2561 an empty string is now equivalent.
2563 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2564 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2565 not performing validation itself.
2567 PP/20 Added force_command boolean option to pipe transport.
2568 Patch from Nick Koston, of cPanel Inc.
2570 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2573 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2575 PP/21 Fix eximon continuous updating with timestamped log-files.
2576 Broken in a format-string cleanup in 4.80, missed when I repaired the
2577 other false fix of the same issue.
2578 Report and fix from Heiko Schlichting.
2581 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2582 Report from Prashanth Katuri.
2584 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2585 It's SecureTransport, so affects any MacOS clients which use the
2586 system-integrated TLS libraries, including email clients.
2588 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2589 using a MIME ACL for non-SMTP local injection.
2590 Report and assistance in diagnosis by Warren Baker.
2592 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2594 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2596 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2597 last id that failed so it may be referenced in subsequent ACL's.
2599 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2602 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2603 ldap library initialization, allowing self-signed CA's to be
2604 used. Also properly sets require_cert option later in code by
2605 using NULL (global ldap config) instead of ldap handle (per
2606 session). Bug diagnosis and testing by alxgomz.
2608 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2609 the src/util/ subdirectory.
2611 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2612 renamed to Transport Post Delivery Action by Jeremy Harris, as
2615 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2616 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2617 redis_servers = needs to be configured which will be used by the redis
2618 lookup. Patch from Warren Baker, of The Packet Hub.
2620 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2622 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2623 hostname or reverse DNS when processing a host list. Used suggestions
2624 from multiple comments on this bug.
2626 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2628 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2629 Missed a few lines, added it to make the runtest require no keyboard
2632 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2633 contains upper case chars. Make router use caseful_local_part.
2635 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2636 support when GnuTLS has been built with p11-kit.
2642 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2644 This, or similar/improved, will also be change PP/11 of 4.82.
2650 PP/01 Handle short writes when writing local log-files.
2651 In practice, only affects FreeBSD (8 onwards).
2652 Bugzilla 1053, with thanks to Dmitry Isaikin.
2654 NM/01 Bugzilla 949 - Documentation tweak
2656 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2659 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2661 PP/02 Implemented gsasl authenticator.
2663 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2665 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2666 `pkg-config foo` for cflags/libs.
2668 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2669 with rest of GSASL and with heimdal_gssapi.
2671 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2672 `pkg-config foo` for cflags/libs for the TLS implementation.
2674 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2675 properties get this fed in as external SSF. A number of robustness
2676 and debugging improvements to the cyrus_sasl authenticator.
2678 PP/08 cyrus_sasl server now expands the server_realm option.
2680 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2681 Patch by Jeremy Harris.
2683 PP/10 Added dbmjz lookup type.
2685 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2687 PP/12 MAIL args handles TAB as well as SP, for better interop with
2688 non-compliant senders.
2689 Analysis and variant patch by Todd Lyons.
2691 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2692 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2693 Patch from Dirk Mueller <dmueller@suse.com>
2695 PP/13 tls_peerdn now print-escaped for spool files.
2696 Observed some $tls_peerdn in wild which contained \n, which resulted
2697 in spool file corruption.
2699 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2700 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2701 or write after TLS renegotiation, which otherwise led to messages
2704 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2705 as a tracking header (ie: a signed header comes before the signature).
2706 Patch from Wolfgang Breyha.
2708 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2709 comma-sep list; embedded commas doubled.
2711 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2713 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2715 Report and patch from Dmitry Banschikov.
2717 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2718 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2719 failures. We appear to now support TLS1.1+ with Exim.
2721 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2722 lets Exim select keys and certificates based upon TLS SNI from client.
2723 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2724 before an outbound SMTP session. New log_selector, +tls_sni.
2726 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2727 NULL dereference. Report and patch from Alun Jones.
2729 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2730 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2731 Not seeing resolver debug output on NetBSD, but suspect this is a
2732 resolver implementation change.
2734 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2735 Left warnings. Added "eximon gdb" invocation mode.
2737 PP/21 Defaulting "accept_8bitmime" to true, not false.
2739 PP/22 Added -bw for inetd wait mode support.
2741 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2742 locate the relevant includes and libraries. Made this the default.
2744 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2745 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2747 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2748 This may cause build issues on older platforms.
2750 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2751 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2752 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2753 Added SNI support via GnuTLS too.
2754 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2756 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2758 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2759 Applied second patch from Janne, implementing suggestion to default
2760 multiple-strings-in-record handling to match SPF spec.
2762 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2764 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2765 read-only, out of scope).
2766 Patch from Wolfgang Breyha, report from Stuart Northfield.
2768 PP/29 Fix three issues highlighted by clang analyser static analysis.
2769 Only crash-plausible issue would require the Cambridge-specific
2770 iplookup router and a misconfiguration.
2771 Report from Marcin Mirosław.
2773 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2775 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2776 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2777 As part of this, removing so much warning spew let me fix some minor
2778 real issues in debug logging.
2780 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2781 assignment on my part. Fixed.
2783 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2784 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2785 Janne Snabb (who went above and beyond: thank you).
2787 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2788 string otherwise requires a connection and a bunch more work and it's
2789 relatively easy to get wrong. Should also expose TLS library linkage
2792 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2793 64-bit ${eval} (JH/03).
2795 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2796 GNU libc to support some of the 64-bit stuff, should not lead to
2797 conflicts. Defined before os.h is pulled in, so if a given platform
2798 needs to override this, it can.
2800 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2801 protection layer was required, which is not implemented.
2802 Bugzilla 1254, patch from Wolfgang Breyha.
2804 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2805 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2806 tls_dhparam take prime identifiers. Also unbreak combination of
2807 OpenSSL+DH_params+TLSSNI.
2809 PP/39 Disable SSLv2 by default in OpenSSL support.
2815 PP/01 Solaris build fix for Oracle's LDAP libraries.
2816 Bugzilla 1109, patch from Stephen Usher.
2818 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2820 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2823 TF/02 Fix a couple more cases where we did not log the error message
2824 when unlink() failed. See also change 4.74-TF/03.
2826 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2827 lock up or crash if it happened to be inside a call to libc when it
2828 got a SIGUSR1 from exiwhat.
2830 The SIGUSR1 handler appends the current process status to the process
2831 log which is later printed by exiwhat. It used to use the general
2832 purpose logging code to do this, but several functions it calls are
2833 not safe for signals.
2835 The new output code in the SIGUSR1 handler is specific to the process
2836 log, and simple enough that it's easy to inspect for signal safety.
2837 Removing some special cases also simplifies the general logging code.
2838 Removing the spurious timestamps from the process log simplifies
2841 TF/04 Improved ratelimit ACL condition.
2843 The /noupdate option has been deprecated in favour of /readonly which
2844 has clearer semantics. The /leaky, /strict, and /readonly update modes
2845 are mutually exclusive. The update mode is no longer included in the
2846 database key; it just determines when the database is updated. (This
2847 means that when you upgrade Exim will forget old rate measurements.)
2849 Exim now checks that the per_* options are used with an update mode that
2850 makes sense for the current ACL. For example, when Exim is processing a
2851 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2852 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2853 must specify per_mail/readonly. If you omit the update mode it defaults to
2854 /leaky where that makes sense (as before) or /readonly where required.
2856 The /noupdate option is now undocumented but still supported for
2857 backwards compatibility. It is equivalent to /readonly except that in
2858 ACLs where /readonly is required you may specify /leaky/noupdate or
2859 /strict/noupdate which are treated the same as /readonly.
2861 A useful new feature is the /count= option. This is a generalization
2862 of the per_byte option, so that you can measure the throughput of other
2863 aggregate values. For example, the per_byte option is now equivalent
2864 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2866 The per_rcpt option has been generalized using the /count= mechanism
2867 (though it's more complicated than the per_byte equivalence). When it is
2868 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2869 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2870 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2871 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2872 Note that using per_rcpt with a non-readonly update mode in more than
2873 one ACL will cause the recipients to be double-counted. (The per_mail
2874 and per_byte options don't have this problem.)
2876 The handling of very low rates has changed slightly. If the computed rate
2877 is less than the event's count (usually one) then this event is the first
2878 after a long gap. In this case the rate is set to the same as this event's
2879 count, so that the first message of a spam run is counted properly.
2881 The major new feature is a mechanism for counting the rate of unique
2882 events. The new per_addr option counts the number of different
2883 recipients that someone has sent messages to in the last time period. It
2884 behaves like per_rcpt if all the recipient addresses are different, but
2885 duplicate recipient addresses do not increase the measured rate. Like
2886 the /count= option this is a general mechanism, so the per_addr option
2887 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2888 example, measure the rate that a client uses different sender addresses
2889 with the options per_mail/unique=$sender_address. There are further
2890 details in the main documentation.
2892 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2894 TF/06 Removed a few PCRE remnants.
2896 TF/07 Automatically extract Exim's version number from tags in the git
2897 repository when doing development or release builds.
2899 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2900 Bugzilla 879. Patch from Paul Fisher.
2902 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2903 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2906 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2908 PP/05 Variable $av_failed, true if the AV scanner deferred.
2909 Bugzilla 1078. Patch from John Horne.
2911 PP/06 Stop make process more reliably on build failure.
2912 Bugzilla 1087. Patch from Heiko Schlittermann.
2914 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2915 Bugzilla 1089. Patch from Heiko Schlittermann.
2917 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2918 Bugzilla 1131. Patch from Holger Weiß.
2920 PP/09 Handle IPv6 addresses with SPF.
2921 Bugzilla 860. Patch from Wolfgang Breyha.
2923 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2925 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2928 PP/11 match_* no longer expand right-hand-side by default.
2929 New compile-time build option, EXPAND_LISTMATCH_RHS.
2930 New expansion conditions, "inlist", "inlisti".
2932 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2934 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2936 PP/14 fix log_write() format string regression from TF/03.
2937 Bugzilla 1152. Patch from Dmitry Isaikin.
2943 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2945 PP/02 Harmonised TLS library version reporting; only show if debugging.
2946 Layout now matches that introduced for other libraries in 4.74 PP/03.
2948 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2950 PP/04 New "dns_use_edns0" global option.
2952 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2955 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2956 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2958 TK/01 Updated PolarSSL code to 0.14.2.
2959 Bugzilla 1097. Patch from Andreas Metzler.
2961 PP/07 Catch divide-by-zero in ${eval:...}.
2962 Fixes bugzilla 1102.
2964 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2967 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2968 format-string attack -- SECURITY: remote arbitrary code execution.
2970 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2971 time unintentionally subject to list matching rules, letting the header
2972 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2973 arbitrary string expansion). This allowed for information disclosure.
2975 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2976 INT_MIN/-1 -- value coerced to INT_MAX.
2982 NM/01 Workaround for PCRE version dependency in version reporting
2985 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2986 This fixes portability to compilers other than gcc, notably
2987 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2989 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2990 makefiles for portability to HP-UX and POSIX correctness.
2992 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2993 Also via indented variable definition in the Makefile.
2994 (Debugging by Oliver Heesakkers).
2996 PP/02 Restore caching of spamd results with expanded spamd_address.
2997 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2999 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3000 Improves build reliability. Fix from: Frank Elsner
3002 NM/02 Fix wide character breakage in the rfc2047 coding
3003 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3005 NM/03 Allow underscore in dnslist lookups
3006 Fixes bug 1026. Patch from Graeme Fowler
3008 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3009 Code patches from Adam Ciarcinski of NetBSD.
3011 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3014 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3015 is logged, to avoid truncation. Patch from John Horne.
3017 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3018 Patch from Jakob Hirsch.
3020 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3021 SQL string expansion failure details.
3022 Patch from Andrey Oktyabrski.
3024 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3025 Patch from Simon Arlott.
3027 PP/09 New lookups functionality failed to compile on old gcc which rejects
3028 extern declarations in function scope.
3029 Patch from Oliver Fleischmann
3031 PP/10 Use sig_atomic_t for flags set from signal handlers.
3032 Check getgroups() return and improve debugging.
3033 Fixed developed for diagnosis in bug 927 (which turned out to be
3036 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3037 Patch from Mark Zealey.
3039 PP/12 Bugzilla 1056: Improved spamd server selection.
3040 Patch from Mark Zealey.
3042 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3043 Based on patch from Heiko Schlittermann.
3045 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3046 Patch from Uwe Doering, sign-off by Michael Haardt.
3048 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3049 variable declaration deep within a block. Bug and patch from
3052 PP/15 lookups-Makefile IRIX compatibility coercion.
3054 PP/16 Make DISABLE_DKIM build knob functional.
3056 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3057 Patch by Simon Arlott
3059 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3060 variable argument macros. Our copy now differs from upstream.
3066 TF/01 Failure to get a lock on a hints database can have serious
3067 consequences so log it to the panic log.
3069 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3070 controlled using the smtp_confirmation log selector.
3072 TF/03 Include the error message when we fail to unlink a spool file.
3074 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3075 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3076 for maintaining out-of-tree patches for some time.
3078 PP/01 Bugzilla 139: Documentation and portability issues.
3079 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3080 Handle per-OS dynamic-module compilation flags.
3082 PP/02 Let /dev/null have normal permissions.
3083 The 4.73 fixes were a little too stringent and complained about the
3084 permissions on /dev/null. Exempt it from some checks.
3085 Reported by Andreas M. Kirchwitz.
3087 PP/03 Report version information for many libraries, including
3088 Exim version information for dynamically loaded libraries. Created
3089 version.h, now support a version extension string for distributors
3090 who patch heavily. Dynamic module ABI change.
3092 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3093 privilege escalation vulnerability whereby the Exim run-time user
3094 can cause root to append content of the attacker's choosing to
3097 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3100 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3101 If dropping privileges for untrusted macros, we disabled normal logging
3102 on the basis that it would fail; for the Exim run-time user, this is not
3103 the case, and it resulted in successful deliveries going unlogged.
3104 Fixed. Reported by Andreas Metzler.
3110 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3111 only prepend for the Resent-* case. Fixes regression introduced in
3112 Exim 4.70 by NM/22 for Bugzilla 607.
3114 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3115 increasing numbers of administrators be bitten by this.
3117 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3119 PP/03 Bugzilla 994: added openssl_options main configuration option.
3121 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3123 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3125 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3127 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3128 without a peer certificate, leading to a segfault because of an
3129 assumption that peers always have certificates. Be a little more
3130 paranoid. Problem reported by Martin Tscholak.
3132 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3133 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3134 NB: ClamAV planning to remove STREAM in "middle of 2010".
3135 CL also introduces -bmalware, various -d+acl logging additions and
3136 more caution in buffer sizes.
3138 PP/09 Implemented reverse_ip expansion operator.
3140 PP/10 Bugzilla 937: provide a "debug" ACL control.
3142 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3144 PP/12 Bugzilla 973: Implement --version.
3146 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3148 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3150 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3152 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3153 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3154 ignore trailing whitespace.
3156 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3158 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3161 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3162 Notification from Dr Andrew Aitchison.
3164 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3165 ExtendedDetectionInfo response format.
3166 Notification from John Horne.
3168 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3171 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3172 XSL and documented dependency on system catalogs, with examples of how
3175 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3178 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3179 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3180 configuration file which is writeable by the Exim user or group.
3182 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3183 of configuration files to cover files specified with the -C option if
3184 they are going to be used with root privileges, not just the default
3187 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3188 option (effectively making it always true).
3190 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3191 files to be used while preserving root privileges.
3193 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3194 that rogue child processes cannot use them.
3196 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3197 run-time user, instead of root.
3199 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3200 Exim run-time user without dropping privileges.
3202 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3203 result string, instead of calling string_vformat() twice with the same
3206 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3207 for other users. Others should always drop root privileges if they use
3208 -C on the command line, even for a whitelisted configure file.
3210 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3212 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3218 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3219 $data_path, and $header_path variables; fixed documentation bugs and
3222 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3223 exipick to access non-standard spools, including the "frozen" queue
3226 NM/01 Bugzilla 965: Support mysql stored procedures.
3227 Patch from Alain Williams
3229 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3231 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3232 Patch from Andreas Metzler
3234 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3235 Patch from Kirill Miazine
3237 NM/05 Bugzilla 671: Added umask to procmail example.
3239 JJ/03 installed exipick 20100323.0, fixing doc bug
3241 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3242 directory. Notification and patch from Dan Rosenberg.
3244 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3246 TK/02 Improve log output when DKIM signing operation fails.
3248 MH/01 Treat the transport option dkim_domain as a colon separated
3249 list, not as a single string, and sign the message with each element,
3250 omitting multiple occurences of the same signer.
3252 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3253 Bugzilla 985, 986. Patch by Simon Arlott
3255 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3256 Patch by Simon Arlott
3258 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3259 MBX locking. Notification from Dan Rosenberg.
3265 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3267 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3269 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3271 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3273 NM/04 Bugzilla 915: Flag broken perl installation during build.
3279 TK/01 Added patch by Johannes Berg that expands the main option
3280 "spamd_address" if it starts with a dollar sign.
3282 TK/02 Write list of recipients to X-Envelope-Sender header when building
3283 the mbox-format spool file for content scanning (suggested by Jakob
3286 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3287 (http://www.dcc-servers.net/) support via dccifd. Activated by
3288 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3290 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3291 by Mark Daniel Reidel <mr@df.eu>.
3293 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3294 When building exim an external PCRE library is now needed -
3295 PCRE is a system library on the majority of modern systems.
3296 See entry on PCRE_LIBS in EDITME file.
3298 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3299 conversation. Added nologin parameter to request.
3300 Patch contributed by Kirill Miazine.
3302 TF/01 Do not log submission mode rewrites if they do not change the address.
3304 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3306 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3307 log files in place. Contributed by Roberto Lima.
3309 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3311 TF/03 Bugzilla 615: When checking the local_parts router precondition
3312 after a local_part_suffix or local_part_prefix option, Exim now
3313 does not use the address's named list lookup cache, since this
3314 contains cached lookups for the whole local part.
3316 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3317 Robert Millan. Documentation is in experimental-spec.txt.
3319 TF/04 Bugzilla 668: Fix parallel build (make -j).
3321 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3323 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3324 Patch provided by Jan Srzednicki.
3326 TF/05 Leading white space used to be stripped from $spam_report which
3327 wrecked the formatting. Now it is preserved.
3329 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3330 that they are available at delivery time.
3332 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3334 TF/08 TLS error reporting now respects the incoming_interface and
3335 incoming_port log selectors.
3337 TF/09 Produce a more useful error message if an SMTP transport's hosts
3338 setting expands to an empty string.
3340 NM/06 Bugzilla 744: EXPN did not work under TLS.
3341 Patch provided by Phil Pennock.
3343 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3344 Patch provided by Richard Godbee.
3346 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3347 acl_smtp_notquit, added index entry.
3349 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3350 Patch provided by Eugene Bujak.
3352 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3353 accept(). Patch provided by Maxim Dounin.
3355 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3356 Patch provided by Phil Pennock.
3358 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3360 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3361 Patch provided by Brad "anomie" Jorsch.
3363 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3364 Patch provided by Dean Brooks.
3366 TK/05 Add native DKIM support (does not depend on external libraries).
3368 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3369 Patch provided by Graeme Fowler.
3371 NM/16 Bugzilla 851: Documentation example syntax fix.
3373 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3375 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3378 NM/19 Bugzilla 745: TLS version reporting.
3379 Patch provided by Phil Pennock.
3381 NM/20 Bugzilla 167: bool: condition support.
3382 Patch provided by Phil Pennock.
3384 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3385 clients. Patch provided by Phil Pennock.
3387 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3388 Patch provided by Brad "anomie" Jorsch.
3390 NM/23 Bugzilla 687: Fix misparses in eximstats.
3391 Patch provided by Heiko Schlittermann.
3393 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3394 Patch provided by Heiko Schlittermann.
3396 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3397 plus update to original patch.
3399 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3401 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3402 Patch provided by David Brownlee.
3404 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3406 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3408 NM/30 Bugzilla 888: TLS documentation bugfixes.
3410 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3412 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3413 Unlike the original bugzilla I have changed all shell scripts in src tree.
3415 NM/33 Bugzilla 898: Transport filter timeout fix.
3416 Patch by Todd Rinaldo.
3418 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3419 Patch by Serge Demonchaux.
3421 NM/35 Bugzilla 39: Base64 decode bug fixes.
3422 Patch by Jakob Hirsch.
3424 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3426 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3428 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3430 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3436 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3437 ALT-N's libdkim that I have put here:
3438 http://duncanthrax.net/exim-experimental/
3440 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3441 were called 'true' and it seems that C99 defines that as a reserved
3442 keyword to be used with 'bool' variable types. That means you could
3443 not include C99-style headers which use bools without triggering
3444 build errors in sieve.c.
3446 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3447 as mailq or other aliases. Changed the --help handling significantly
3448 to do whats expected. exim_usage() emits usage/help information.
3450 SC/01 Added the -bylocaldomain option to eximstats.
3452 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3454 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3456 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3462 PH/01 Another patch from the Sieve maintainer.
3464 PH/02 When an IPv6 address is converted to a string for single-key lookup
3465 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3466 dots are used instead of colons so that keys in lsearch files need not
3467 contain colons. This was done some time before quoting was made available
3468 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3469 (notated using the quote facility) so as to distinguish them from IPv4
3470 keys. This meant that lookups for IP addresses in host lists did not work
3471 for iplsearch lookups.
3473 This has been fixed by arranging for IPv6 addresses to be expressed with
3474 colons if the lookup type is iplsearch. This is not incompatible, because
3475 previously such lookups could never work.
3477 The situation is now rather anomalous, since one *can* have colons in
3478 ordinary lsearch keys. However, making the change in all cases is
3479 incompatible and would probably break a number of configurations.
3481 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3484 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3485 conversion specification without a maximum field width, thereby enabling
3486 a rogue spamd server to cause a buffer overflow. While nobody in their
3487 right mind would setup Exim to query an untrusted spamd server, an
3488 attacker that gains access to a server running spamd could potentially
3489 exploit this vulnerability to run arbitrary code as the Exim user.
3491 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3492 $primary_hostname instead of what libspf2 thinks the hosts name is.
3494 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3495 a directory entry by the name of the lookup key. Previously, if a
3496 symlink pointed to a non-existing file or a file in a directory that
3497 Exim lacked permissions to read, a lookup for a key matching that
3498 symlink would fail. Now it is enough that a matching directory entry
3499 exists, symlink or not. (Bugzilla 503.)
3501 PH/03 The body_linecount and body_zerocount variables are now exported in the
3504 PH/04 Added the $dnslist_matched variable.
3506 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3507 This means they are set thereafter only if the connection becomes
3510 PH/06 Added the client_condition to authenticators so that some can be skipped
3511 by clients under certain conditions.
3513 PH/07 The error message for a badly-placed control=no_multiline_responses left
3514 "_responses" off the end of the name.
3516 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3518 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3519 (without spaces) instead of just copying the configuration text.
3521 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3523 PH/11 Added $max_received_linelength.
3525 PH/12 Added +ignore_defer and +include_defer to host lists.
3527 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3528 way in which PCRE > 7.0 is built.
3530 PH/14 Implemented queue_only_load_latch.
3532 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3533 MAIL command. The effect was to mangle the value on 64-bit systems.
3535 PH/16 Another patch from the Sieve maintainer.
3537 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3539 PH/18 If a system quota error occurred while trying to create the file for
3540 a maildir delivery, the message "Mailbox is full" was not appended to the
3541 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3542 only to a quota excession during the actual writing of the file.
3544 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3545 characters?) which causes problems in log lines. The DN values are now
3546 passed through string_printing() before being added to log lines.
3548 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3549 and InterBase are left for another time.)
3551 PH/21 Added message_body_newlines option.
3553 PH/22 Guard against possible overflow in moan_check_errorcopy().
3555 PH/23 POSIX allows open() to be a macro; guard against that.
3557 PH/24 If the recipient of an error message contained an @ in the local part
3558 (suitably quoted, of course), incorrect values were put in $domain and
3559 $local_part during the evaluation of errors_copy.
3565 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3566 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3569 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3570 issue a MAIL command.
3572 PH/02 In an ACL statement such as
3574 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3576 if a client was not listed at all, or was listed with a value other than
3577 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3578 the condition was not true (as it should be), so access was not denied.
3579 The bug was that the ! inversion was incorrectly passed on to the second
3580 item. This has been fixed.
3582 PH/03 Added additional dnslists conditions == and =& which are different from
3583 = and & when the dns lookup returns more than one IP address.
3585 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3586 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3588 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3589 FSYNC, which compiles an option called disable_fsync that allows for
3590 bypassing fsync(). The documentation is heavily laced with warnings.
3592 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3594 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3595 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3596 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3597 including adding "make clean"; (3) Added -fPIC when compiling the test
3598 dynamically loaded module, to get rid of a warning.
3600 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3601 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3602 = 0s. The bug is otherwise harmless.
3604 PH/07 There was a bug in the dovecot authenticator such that the value of
3605 $auth1 could be overwritten, and so not correctly preserved, after a
3606 successful authentication. This usually meant that the value preserved by
3607 the server_setid option was incorrect.
3609 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3611 PH/09 Installed PCRE release 7.0.
3613 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3614 run for batched SMTP input. It is now run at the start of every message
3615 in the batch. While fixing this I discovered that the process information
3616 (output by running exiwhat) was not always getting set for -bs and -bS
3617 input. This is fixed, and it now also says "batched" for BSMTP.
3619 PH/11 Added control=no_pipelining.
3621 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3622 patch, slightly modified), and move the expansion of helo_data till after
3623 the connection is made in the smtp transport (so it can use these
3626 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3628 PH/14 Added log_selector = +pid.
3630 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3632 PH/16 Add ${if forany and ${if forall.
3634 PH/17 Added dsn_from option to vary the From: line in DSNs.
3636 PH/18 Flush SMTP output before performing a callout, unless control =
3637 no_callout_flush is set.
3639 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3640 was true (the default) a successful delivery failed to delete the retry
3641 item, thus causing premature timeout of the address. The bug is now
3644 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3646 PH/21 Long custom messages for fakedefer and fakereject are now split up
3647 into multiline responses in the same way that messages for "deny" and
3648 other ACL rejections are.
3650 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3651 with slight modification.
3653 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3654 draft, changing the syntax and factoring some duplicate code".
3656 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3657 for deliveries of the second and subsequent messages over the same SMTP
3660 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3661 ${reduce, with only minor "tidies".
3663 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3665 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3666 expansion side effects.
3668 PH/27 When a message times out after an over-quota error from an Exim-imposed
3669 quota, the bounce message says "mailbox is full". This message was not
3670 being given when it was a system quota that was exceeded. It now should
3673 MH/03 Made $recipients available in local_scan(). local_scan() already has
3674 better access to the recipient list through recipients_list[], but
3675 $recipients can be useful in postmaster-provided expansion strings.
3677 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3678 in the case of a MAIL command with additional options following the
3679 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3680 were accidentally chopped off.
3682 PH/29 SMTP synchronization checks are implemented when a command is read -
3683 there is a check that no more input is waiting when there shouldn't be
3684 any. However, for some commands, a delay in an ACL can mean that it is
3685 some time before the response is written. In this time, more input might
3686 arrive, invalidly. So now there are extra checks after an ACL has run for
3687 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3688 pipelining has not been advertised.
3690 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3692 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3693 returned in the SMTP response when smtp_return_error_details was false.
3694 This has been fixed.
3696 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3697 instead of the C I/O that was originally supplied, because problems were
3698 reported on Solaris.
3700 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3701 Exim which did not show up earlier: it was assuming that a call to
3702 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3703 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3704 was a macro that became an assignment, so it seemed to work. This has
3705 changed to a proper function call with a void return, hence the compile
3706 error. Exim's code has been fixed.
3708 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3711 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3713 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3715 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3716 because it turns out to be too big or there is a timeout) there is no
3717 "Completed" line in the log. When some messages of this type were
3718 selected by exigrep, they were listed as "not completed". Others were
3719 picked up by some special patterns. I have improved the selection
3720 criteria to be more general.
3722 PH/38 The host_find_failed option in the manualroute router can now be set
3723 to "ignore", to completely ignore a host whose IP address cannot be
3724 found. If all hosts are ignored, the behaviour is controlled by the new
3725 host_all_ignored option.
3727 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3728 homing or because of multiple MX records with /mx) generated more than
3729 one IP address, and the following item turned out to be the local host,
3730 all the secondary addresses of the first item were incorrectly removed
3731 from the list, along with the local host and any following hosts (which
3732 is what is supposed to happen).
3734 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3735 whoever called Exim into the -H file. In the case of the daemon it was
3736 behaving confusingly. When first started, it used values for whoever
3737 started the daemon, but after a SIGHUP it used the Exim user (because it
3738 calls itself on a restart). I have changed the code so that it now always
3741 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3742 message are rejected with the same error (e.g. no authentication or bad
3743 sender address), and a DATA command is nevertheless sent (as can happen
3744 with PIPELINING or a stupid MUA), the error message that was given to the
3745 RCPT commands is included in the rejection of the DATA command. This is
3746 intended to be helpful for MUAs that show only the final error to their
3749 PH/42 Another patch from the Sieve maintainer.
3751 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3752 Eximstats - Fixed some broken HTML links and added missing column headers
3754 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3755 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3757 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3759 PH/43 Yet another patch from the Sieve maintainer.
3761 PH/44 I found a way to check for a TCP/IP connection going away before sending
3762 the response to the final '.' that terminates a message, but only in the
3763 case where the client has not sent further data following the '.'
3764 (unfortunately, this is allowed). However, in many cases there won't be
3765 any further data because there won't be any more messages to send. A call
3766 to select() can be used: if it shows that the input is "ready", there is
3767 either input waiting, or the socket has been closed. An attempt to read
3768 the next input character can distinguish the two cases. Previously, Exim
3769 would have sent an OK response which the client would never have see.
3770 This could lead to message repetition. This fix should cure that, at
3771 least in a lot of common cases.
3773 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3774 advertised in response to EHLO.
3780 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3781 fixed by 4.65/MH/01 (is this a record?) are fixed:
3783 (i) An empty string was always treated as zero by the numeric comparison
3784 operators. This behaviour has been restored.
3786 (ii) It is documented that the numeric comparison operators always treat
3787 their arguments as decimal numbers. This was broken in that numbers
3788 starting with 0 were being interpreted as octal.
3790 While fixing these problems I realized that there was another issue that
3791 hadn't been noticed. Values of message_size_limit (both the global option
3792 and the transport option) were treated as octal if they started with 0.
3793 The documentation was vague. These values are now always treated as
3794 decimal, and I will make that clear in the documentation.
3800 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3801 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3804 MH/01 Don't check that the operands of numeric comparison operators are
3805 integers when their expansion is in "skipping" mode (fixes bug
3806 introduced by 4.64-PH/07).
3808 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3809 child addresses, Exim now panics and dies. Previously, because the count
3810 is held in a short int, deliveries were likely to be lost. As such a
3811 large number of recipients for a single message is ridiculous
3812 (performance will be very, very poor), I have chosen to impose a limit
3813 rather than extend the field.
3819 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3820 leftover -K file (the existence of which was triggered by #402).
3821 While we were at it, introduced process PID as part of the -K
3822 filename. This should rule out race conditions when creating
3825 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3826 processing considerably. Previous code took too long for large mails,
3827 triggering a timeout which in turn triggers #401.
3829 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3830 in the DK code in transports.c. sendfile() is not really portable,
3831 hence the _LINUX specificness.
3833 TF/01 In the add_headers option to the mail command in an Exim filter,
3834 there was a bug that Exim would claim a syntax error in any
3835 header after the first one which had an odd number of characters
3838 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3839 callout verification, Exim cached a "reject" for the entire domain. This
3840 is correct for most verifications, but it is not correct for a recipient
3841 verification with use_sender or use_postmaster set, because in that case
3842 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3843 case of MAIL FROM:<> rejection from other early rejections (e.g.
3844 rejection of HELO). When verifying a recipient using a non-null MAIL
3845 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3846 Whatever the result of the callout, the value of the domain cache is
3847 left unchanged (for any other kind of callout, getting as far as trying
3848 RCPT means that the domain itself is ok).
3850 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3853 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3854 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3855 session, the daemon ignores ECONNECT errors and logs others; it now
3856 ignores EPIPE as well.
3858 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3859 (quoted-printable decoding).
3861 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3862 later the small subsequent patch to fix an introduced bug.
3864 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3866 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3868 PH/08 An error is now given if message_size_limit is specified negative.
3870 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3871 to be given (somewhat) arbitrary names.
3873 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3876 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3877 miscellaneous code fixes
3879 PH/10 Added the log_reject_target ACL modifier to specify where to log
3882 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3883 hostname. This is wrong, because it relates to the incoming message (and
3884 probably the interface on which it is arriving) and not to the outgoing
3885 callout (which could be using a different interface). This has been
3886 changed to use the value of the helo_data option from the smtp transport
3887 instead - this is what is used when a message is actually being sent. If
3888 there is no remote transport (possible with a router that sets up host
3889 addresses), $smtp_active_hostname is used.
3891 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3892 tweaks were necessary in order to get it to work (see also 21 below):
3893 (a) The code assumed that strncpy() returns a negative number on buffer
3894 overflow, which isn't the case. Replaced with Exim's string_format()
3896 (b) There were several signed/unsigned issues. I just did the minimum
3897 hacking in of casts. There is scope for a larger refactoring.
3898 (c) The code used strcasecmp() which is not a standard C function.
3899 Replaced with Exim's strcmpic() function.
3900 (d) The code set only $1; it now sets $auth1 as well.
3901 (e) A simple test gave the error "authentication client didn't specify
3902 service in request". It would seem that Dovecot has changed its
3903 interface. Fortunately there's a specification; I followed it and
3904 changed what the client sends and it appears to be working now.
3906 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3909 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3910 address A is aliased to B and C, where B exists and C does not. Without
3911 -v the output is "A verified" because verification stops after a
3912 successful redirection if more than one address is generated. However,
3913 with -v the child addresses are also verified. Exim was outputting "A
3914 failed to verify" and then showing the successful verification for C,
3915 with its parentage. It now outputs "B failed to verify", showing B's
3916 parentage before showing the successful verification of C.
3918 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3919 look up a TXT record in a specific list after matching in a combined
3922 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3923 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3924 they consult the DNS. I had assumed they would set it the way they
3925 wanted; and indeed my experiments on Linux seem to show that in some
3926 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3927 To be on the safe side, however, I have now made the interface to
3928 host_find_byname() similar to host_find_bydns(), with an argument
3929 containing the DNS resolver options. The host_find_byname() function now
3930 sets these options at its start, just as host_find_bydns() does. The smtp
3931 transport options dns_qualify_single and dns_search_parents are passed to
3932 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3933 of host_find_byname() use the default settings of RES_DEFNAMES
3934 (qualify_single) but not RES_DNSRCH (search_parents).
3936 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3937 spool_read_header() do less string testing, by means of a preliminary
3938 switch on the second character of optional "-foo" lines. (This is
3939 overdue, caused by the large number of possibilities that now exist.
3940 Originally there were few.) While I was there, I also converted the
3941 str(n)cmp tests so they don't re-test the leading "-" and the first
3942 character, in the hope this might squeeze out yet more improvement.
3944 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3945 flag allowing group syntax was set by the header_syntax check but not
3946 turned off, possible causing trouble later; (2) The flag was not being
3947 set at all for the header_verify test, causing "group"-style headers to
3948 be rejected. I have now set it in this case, and also caused header_
3949 verify to ignore an empty address taken from a group. While doing this, I
3950 came across some other cases where the code for allowing group syntax
3951 while scanning a header line wasn't quite right (mostly, not resetting
3952 the flag correctly in the right place). These bugs could have caused
3953 trouble for malformed header lines. I hope it is now all correct.
3955 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3956 with the "reply" argument non-NULL. The code, however (which originally
3957 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3958 but it didn't always do it. This confused somebody who was copying the
3959 code for some other use. I have removed all the tests.
3961 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3962 feature that was used to support insecure browsers during the U.S. crypto
3963 embargo. It requires special client support, and Exim is probably the
3964 only MTA that supported it -- and would never use it because real RSA is
3965 always available. This code has been removed, because it had the bad
3966 effect of slowing Exim down by computing (never used) parameters for the
3967 RSA_EXPORT functionality.
3969 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3970 authenticator to fail if there's a tab character in the incoming data
3971 (there should never be unless someone is messing about, as it's supposed
3972 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3973 if the connection is using TLS or if the remote IP is the same as the
3974 local IP, and the "valid-client-cert option" if a client certificate has
3977 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3978 authenticators. This can be used for authorization after authentication
3979 succeeds. (In the case of plaintext, it servers for both authentication
3982 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3983 if any retry times were supplied.
3985 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3986 connection, where there is no client IP address to check. In this
3987 situation, the verify now always succeeds.
3989 PH/25 Applied John Jetmore's -Mset patch.
3991 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3993 PH/27 In a string expansion for a processed (not raw) header when multiple
3994 headers of the same name were present, leading whitespace was being
3995 removed from all of them, but trailing whitespace was being removed only
3996 from the last one. Now trailing whitespace is removed from each header
3997 before concatenation. Completely empty headers in a concatenation (as
3998 before) are ignored.
4000 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4001 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4003 PH/29 [Removed. This was a change that I later backed out, and forgot to
4004 correct the ChangeLog entry (that I had efficiently created) before
4005 committing the later change.]
4007 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4008 address errors (4xx response to RCPT) over the same connection as other
4009 messages routed to the same hosts. Such deliveries are always "forced",
4010 so retry times are not inspected. This resulted in far too many retries
4011 for the affected addresses. The effect occurred only when there were more
4012 hosts than the hosts_max_try setting in the smtp transport when it had
4013 the 4xx errors. Those hosts that it had tried were not added to the list
4014 of hosts for which the message was waiting, so if all were tried, there
4015 was no problem. Two fixes have been applied:
4017 (i) If there are any address or message errors in an SMTP delivery, none
4018 of the hosts (tried or untried) are now added to the list of hosts
4019 for which the message is waiting, so the message should not be a
4020 candidate for sending over the same connection that was used for a
4021 successful delivery of some other message. This seems entirely
4022 reasonable: after all the message is NOT "waiting for some host".
4023 This is so "obvious" that I'm not sure why it wasn't done
4024 previously. Hope I haven't missed anything, but it can't do any
4025 harm, as the worst effect is to miss an optimization.
4027 (ii) If, despite (i), such a delivery is accidentally attempted, the
4028 routing retry time is respected, so at least it doesn't keep
4029 hammering the server.
4031 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4032 in ${readsocket because some servers need this prod.
4034 PH/32 Added some extra debug output when updating a wait-xxx database.
4036 PH/33 The hint "could be header name not terminated by colon", which has been
4037 given for certain expansion errors for a long time, was not being given
4038 for the ${if def:h_colon_omitted{... case.
4040 PH/34 The spec says: "With one important exception, whenever a domain list is
4041 being scanned, $domain contains the subject domain." There was at least
4042 one case where this was not true.
4044 PH/35 The error "getsockname() failed: connection reset by peer" was being
4045 written to the panic log as well as the main log, but it isn't really
4046 panic-worthy as it just means the connection died rather early on. I have
4047 removed the panic log writing for the ECONNRESET error when getsockname()
4050 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4051 runs only) independently of the message's sender address. This meant
4052 that, if the 4xx error was in fact related to the sender, a different
4053 message to the same recipient with a different sender could confuse
4054 things. In particular, this can happen when sending to a greylisting
4055 server, but other circumstances could also provoke similar problems.
4056 I have changed the default so that the retry time for these errors is now
4057 based a combination of the sender and recipient addresses. This change
4058 can be overridden by setting address_retry_include_sender=false in the
4061 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4062 remote server are returned as part of bounce messages. This was not
4063 happening for LMTP over a pipe (the lmtp transport), but now it is the
4064 same for both kinds of LMTP.
4066 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4067 in header lines that were in fact CNAMEs. This is no longer the case.
4069 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4070 and queue runs started by the daemon processed all messages. This has
4071 been fixed so that -R and -S can now usefully be given with -q<time>.
4073 PH/40 Import PCRE release 6.7 (fixes some bugs).
4075 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4077 PH/42 Give an error if -q is specified more than once.
4079 PH/43 Renamed the variables $interface_address and $interface_port as
4080 $received_ip_address and $received_port, to make it clear that these
4081 values apply to message reception, and not to the outgoing interface when
4082 a message is delivered. (The old names remain recognized, of course.)
4084 PH/44 There was no timeout on the connect() call when using a Unix domain
4085 socket in the ${readsocket expansion. There now is.
4087 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4088 be meaningful with "accept".
4090 SC/01 Eximstats V1.43
4091 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4093 SC/02 Eximstats V1.44
4094 Use a glob alias rather than an array ref in the generated
4095 parser. This improves both readability and performance.
4097 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4098 Collect SpamAssassin and rejection statistics.
4099 Don't display local sender or destination tables unless
4100 there is data to show.
4101 Added average volumes into the top table text output.
4103 SC/04 Eximstats V1.46
4104 Collect data on the number of addresses (recipients)
4105 as well as the number of messages.
4107 SC/05 Eximstats V1.47
4108 Added 'Message too big' to the list of mail rejection
4109 reasons (thanks to Marco Gaiarin).
4111 SC/06 Eximstats V1.48
4112 Mainlog lines which have GMT offsets and are too short to
4113 have a flag are now skipped.
4115 SC/07 Eximstats V1.49 (Alain Williams)
4116 Added the -emptyok flag.
4118 SC/08 Eximstats V1.50
4119 Fixes for obtaining the IP address from reject messages.
4121 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4122 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4123 whitespace changes from 4.64-PH/27
4125 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4128 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4129 are found, allow negative numbers in numeric criteria)
4131 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4133 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4134 to match changes made in 4.64-PH/43
4136 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4138 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4139 "domain" is actually an IP address, and force a failure. This locks out
4140 those revolvers/nameservers that support "A-for-A" lookups, in
4141 contravention of the specifications.
4143 PH/48 When a host name was looked up from an IP address, and the subsequent
4144 forward lookup of the name timed out, the host name was left in
4145 $sender_host_name, contrary to the specification.
4147 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4148 restricted to single-key lookups, Exim was not diagnosing an error if
4149 * or *@ was used with a query-style lookup.
4151 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4153 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4154 long ago, but noone interested enough thought of it. Let's just say that
4155 the "1.1" means that there are some new functions that weren't there at
4156 some point in the past.
4158 PH/51 Error processing for expansion failure of helo_data from an smtp
4159 transport during callout processing was broken.
4161 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4162 tested/used via the -bh/-bhc/-bs options.
4164 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4165 bug, fixed in subsequent PCRE releases).
4167 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4168 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4170 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4176 SC/01 Use a glob alias rather than an array ref in eximstats generated
4177 parser. This improves both readability and performance.
4179 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4180 Don't display local sender or destination tables in eximstats unless
4181 there is data to show.
4182 Added average volumes into the eximstats top table text output.
4184 SC/03 Collect data on the number of addresses (recipients) as well
4185 as the number of messages in eximstats.
4187 TF/01 Correct an error in the documentation for the redirect router. Exim
4188 does (usually) call initgroups() when daemonizing.
4190 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4191 with consistent privilege compared to when running as a daemon.
4193 TF/03 Note in the spec that $authenticated_id is not set for local
4194 submissions from trusted users.
4196 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4197 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4199 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4200 by adding some example configuration directives to the default
4201 configuration file. A little bit of work is required to uncomment the
4202 directives and define how usernames and passwords are checked, but
4203 there is now a framework to start from.
4205 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4206 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4207 without this. I don't know how relevant this is to other LDAP libraries.
4209 PH/02 Add the verb name to the "unknown ACL verb" error.
4211 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4213 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4215 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4216 directory not expanded when it should be if an expanded home directory
4217 was set for the address (which is overridden by the transport).
4219 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4222 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4223 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4224 because it is too late at that time, and has no effect.
4226 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4227 security issue with \' (bugzilla #107). I could not use the
4228 PQescapeStringConn() function, because it needs a PGconn value as one of
4231 PH/08 When testing addresses using -bt, indicate those final addresses that
4232 are duplicates that would not cause an additional delivery. At least one
4233 person was confused, thinking that -bt output corresponded to deliveries.
4234 (Suppressing duplicates isn't a good idea as you lose the information
4235 about possibly different redirections that led to the duplicates.)
4237 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4238 systems where poll() doesn't work, in particular OS X.
4240 PH/10 Added more information to debugging output for retry time not reached.
4242 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4243 operations in malware.c.
4245 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4248 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4249 syslog_duplication set false, log lines that would normally be written
4250 both the the main log and to the reject log were not written to syslog at
4253 PH/14 In the default configuration, change the use of "message" in ACL warn
4254 statements to "add_header".
4256 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4257 not followed by a command (e.g. "seen endif").
4259 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4260 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4263 PH/17 Added extra conditions to the default value of delay_warning_condition
4267 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4268 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4269 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4272 The Auto-Submitted: and various List- headers are standardised, whereas I
4273 don't think Precedence: ever was.
4275 PH/18 Refactored debugging code in route_finduser() to show more information,
4276 in particular, the error code if getpwnam() issues one.
4278 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4279 This is apparently needed in addition to the PH/07 change above to avoid
4280 any possible encoding problems.
4282 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4283 but not after initializing Perl.
4285 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4286 output them only if debugging. By default they are written stderr,
4287 apparently, which is not desirable.
4289 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4292 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4295 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4297 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4298 authenticated or an ident call has been made. Suppress the default
4299 values for $authenticated_id and $authenticated_sender (but permit -oMai
4300 and -oMas) when testing with -bh.
4302 PH/24 Re-jigged the order of the tests in the default configuration so that the
4303 tests for valid domains and recipients precede the DNS black list and CSA
4304 tests, on the grounds that those ones are more expensive.
4306 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4307 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4308 This bug exists in every version of Exim that I still have, right back to
4311 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4312 However, an attempt to turn on case-sensitivity in a regex key by
4313 including (?-i) didn't work because the subject string was already
4314 lowercased, and the effects were non-intuitive. It turns out that a
4315 one-line patch can be used to allow (?-i) to work as expected.
4321 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4322 other effects) broke the use of negated acl sub-conditions.
4324 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4327 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4328 "Deny" causes Exim to reject the incoming connection with a 554 error.
4329 Unfortunately, if there is a major crisis, such as a disk failure,
4330 tcp-wrappers gives "deny", whereas what one would like would be some
4331 kind of temporary error. A kludge has been added to help with this.
4332 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4333 554 error is used if errno is still zero or contains ENOENT (which occurs
4334 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4337 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4339 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4340 errors. Otherwise a message that provokes a temporary error (when other
4341 messages do not) can cause a whole host to time out.
4343 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4344 addresses were routed directly to files or pipes from a redirect router.
4345 File deliveries just didn't batch; pipe deliveries might have suffered
4348 PH/06 A failure to get a lock for a hints database would erroneously always say
4349 "Failed to get write lock", even when it was really a read lock.
4351 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4352 of 0600. This has been changed to use the value of the lockfile_mode
4353 option (which defaults to 0600).
4355 PH/08 Applied small patch from the Sieve maintainer.
4357 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4358 folder from quota calculations, a direct delivery into this folder messed
4359 up the contents of the maildirsize file. This was because the regex was
4360 used only to exclude .Trash (or whatever) when the size of the mailbox
4361 was calculated. There was no check that a delivery was happening into an
4362 excluded directory. This bug has been fixed by ignoring all quota
4363 processing for deliveries into excluded directories.
4365 PH/10 Added the maildirfolder_create_regex option to appendfile.
4371 PH/01 The code for finding all the local interface addresses on a FreeBSD
4372 system running IPv6 was broken. This may well have applied to all BSD
4373 systems, as well as to others that have similar system calls. The broken
4374 code found IPv4 interfaces correctly, but gave incorrect values for the
4375 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4376 that it would not match correctly against @[] and not recognize the IPv6
4379 PH/02 The ipliteral router was not recognizing addresses of the form user@
4380 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4382 PH/03 Added disable_ipv6.
4384 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4385 decoded form, because it is most often used to construct To: headers
4386 lines in autoreplies, and the decoded form may well be syntactically
4387 invalid. However, $reply_address has leading white space removed, and all
4388 newlines turned into spaces so that the autoreply transport does not
4391 PH/05 If group was specified without a user on a router, and no group or user
4392 was specified on a transport, the group from the router was ignored.
4394 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4395 for visible compile-time settings that can be used to change these
4396 numbers, for those that want even more. Backwards compatibility with old
4397 spool files has been maintained. However, going back to a previous Exim
4398 release will lost any variables that are in spool files.
4400 PH/07 Two small changes when running in the test harness: increase delay when
4401 passing a TCP/IP connection to a new process, in case the original
4402 process has to generate a bounce, and remove special handling of
4403 127.0.0.2 (sic), which is no longer necessary.
4405 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4406 be the same on different OS.
4408 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4411 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4412 whether --show-vars was specified or not
4414 JJ/02 exipick: Added support for new ACL variable spool format introduced
4417 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4418 syntactically invalid From: or Reply-to: line, and a filter used this to
4419 generate an autoreply, and therefore failed to obtain an address for the
4420 autoreply, Exim could try to deliver to a non-existent relative file
4421 name, causing unrelated and misleading errors. What now happens is that
4422 it logs this as a hard delivery error, but does not attempt to create a
4425 PH/11 The exinext utility has a -C option for testing purposes, but although
4426 the given file was scanned by exinext itself; it wasn't being passed on
4427 when Exim was called.
4429 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4430 an end-of-file indication when reading a command response.
4432 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4433 compiled. In many other places in Exim, IPv6 addresses are always
4434 recognized, so I have changed this. It also means that IPv4 domain
4435 literals of the form [IPV4:n.n.n.n] are now always recognized.
4437 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4438 used if the router is not running as root, for example, when verifying at
4439 ACL time, or when using -bh. The debugging output from this situation was
4440 non-existent - all you got was a failure to exec. I have made two
4443 (a) Failures to set uid/gid, the current directory, or a process leader
4444 in a subprocess such as that created by queryprogram now generate
4445 suitable debugging output when -d is set.
4447 (b) The queryprogram router detects when it is not running as root,
4448 outputs suitable debugging information if -d is set, and then runs
4449 the subprocess without attempting to change uid/gid.
4451 PH/15 Minor change to Makefile for building test_host (undocumented testing
4454 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4455 additional section of a DNS packet that returns MX or SRV records.
4456 Instead, it always explicitly searches for A/AAAA records. This avoids
4457 major problems that occur when a DNS server includes only records of one
4458 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4459 fixed another bug: if SRV records were looked up and the corresponding
4460 address records were *not* found in the additional section, the port
4461 values from the SRV records were lost.
4463 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4464 using the correct key (the original address) when searching the retry
4465 rules in order to find which one to use for generating the retry hint.
4467 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4468 adding the default one. Similarly, if it contains a Reply-To: header, the
4469 errors_reply_to option, if set, is not used.
4471 PH/19 When calculating a retry time, Exim used to measure the "time since
4472 failure" by looking at the "first failed" field in the retry record. Now
4473 it does not use this if it is later than than the arrival time of the
4474 message. Instead it uses the arrival time. This makes for better
4475 behaviour in cases where some deliveries succeed, thus re-setting the
4476 "first failed" field. An example is a quota failure for a huge message
4477 when small messages continue to be delivered. Without this change, the
4478 "time since failure" will always be short, possible causing more frequent
4479 delivery attempts for the huge message than are intended.
4480 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4482 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4483 $1, $2, $3) because the numerical variables can be reset during some
4484 expansion items (e.g. "match"), thereby losing the authentication data.
4486 PH/21 Make -bV show the size of off_t variables so that the test suite can
4487 decide whether to run tests for quotas > 2G.
4489 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4490 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4491 filecount value is greater than 2G or if a quota value is greater than 2G
4492 on a system where the size of off_t is not greater than 4, a panic error
4495 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4496 never match. The debug and -bh output now contains an explicit error
4497 message indicating a malformed IPv4 address or mask.
4499 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4500 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4501 PH/23 above applies.
4503 PH/25 Do not write to syslog when running in the test harness. The only
4504 occasion when this arises is a failure to open the main or panic logs
4505 (for which there is an explicit test).
4507 PH/26 Added the /no_tell option to "control=freeze".
4509 PH/27 If a host name lookup failed very early in a connection, for example, if
4510 the IP address matched host_lookup and the reverse lookup yielded a name
4511 that did not have a forward lookup, an error message of the form "no IP
4512 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4513 could be logged. Now it outputs the IP address instead of "NULL".
4515 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4516 allows the sender and the authenticated sender to be set when
4517 submitting a message from within Exim. Since child_open_exim() is
4518 documented for local_scan(), the new function should be too.
4520 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4521 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4522 results in an empty string is now treated as unset.
4524 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4526 PH/31 Added sender_verify_fail logging option.
4528 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4529 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4530 tidied the source and removed it altogether.
4532 PH/33 When a queue run was abandoned because the load average was too high, a
4533 log line was always written; now it is written only if the queue_run log
4534 selector is set. In addition, the log line for abandonment now contains
4535 information about the queue run such as the pid. This is always present
4536 in "start" and "stop" lines but was omitted from the "abandon" line.
4538 PH/34 Omit spaces between a header name and the colon in the error message that
4539 is given when verify = headers_syntax fails (if there are lots of them,
4540 the message gets confusing).
4542 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4543 names, as there are now some PTR records that contain slashes. This check
4544 is only to protect against broken name servers that fall over on strange
4545 characters, so the fact that it applies to all lookups doesn't matter.
4547 PH/36 Now that the new test suite is complete, we can remove some of the
4548 special code in Exim that was needed for the old test suite. For example,
4549 sorting DNS records because real resolvers return them in an arbitrary
4550 order. The new test suite's fake resolver always returns records in the
4553 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4554 bounces) except when queue_only is set, to avoid logging races between
4555 the different processes.
4557 PH/38 Panic-die if .include specifies a non-absolute path.
4559 PH/39 A tweak to the "H" retry rule from its user.
4561 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4562 a label. They prevented compilation on older perls.
4564 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4565 a warning to be raised on newish perls.
4567 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4568 on queue. Changes to match documented behaviour of showing count of
4569 messages matching specified criteria.
4571 PH/40 Changed the default ident timeout from 30s to 5s.
4573 PH/41 Added support for the use of login_cap features, on those BSD systems
4574 that have them, for controlling the resources used by pipe deliveries.
4576 PH/42 The content-scanning code uses fopen() to create files in which to put
4577 message data. Previously it was not paying any attention to the mode of
4578 the files. Exim runs with umask(0) because the rest of the code creates
4579 files with open(), and sets the required mode explicitly. Thus, these
4580 files were ending up world-writeable. This was not a big issue, because,
4581 being within the spool directory, they were not world-accessible. I have
4582 created a function called modefopen, which takes an additional mode
4583 argument. It sets umask(777), creates the file, chmods it to the required
4584 mode, then resets the umask. All the relevant calls to fopen() in the
4585 content scanning code have been changed to use this function.
4587 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4588 to 24 hours. This avoids potential overflow problems when processing G
4589 and H retry rules. I suspect nobody ever tinkers with this value.
4591 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4593 PH/45 When the plaintext authenticator is running as a client, the server's
4594 challenges are checked to ensure they are valid base64 strings. By
4595 default, the authentication attempt is cancelled if an invalid string is
4596 received. Setting client_ignore_invalid_base64 true ignores these errors.
4597 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4598 they are received. Thus, the responses can be made to depend on the
4599 challenges. If an invalid string is ignored, an empty string is placed in
4602 PH/46 Messages that are created by the autoreply transport now contains a
4603 References: header, in accordance with RFCs 2822 and 3834.
4605 PH/47 Added authenticated_sender_force to the smtp transport.
4607 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4609 PH/49 Installed latest patch from the Sieve maintainer.
4611 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4612 was also set, the appendfile transport was unnecessarily scanning a
4613 directory of message files (e.g. for maildir delivery) to find the count
4614 of files (along with the size), even though it did not need this
4615 information. It now does the scan only if it needs to find either the
4616 size of the count of files.
4618 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4620 PH/52 Two bugs concerned with error handling when the smtp transport is
4623 (i) Exim was not creating retry information for temporary errors given
4624 for individual recipients after the DATA command when the smtp transport
4625 was used in LMTP mode. This meant that they could be retried too
4626 frequently, and not timed out correctly.
4628 (ii) Exim was setting the flag that allows error details to be returned
4629 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4630 recipients that were returned after the DATA command.
4632 PH/53 This is related to PH/52, but is more general: for any failing address,
4633 when detailed error information was permitted to be returned to the
4634 sender, but the error was temporary, then after the final timeout, only
4635 "retry timeout exceeded" was returned. Now it returns the full error as
4636 well as "retry timeout exceeded".
4638 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4639 do this, and (what is worse) MTAs that accept it.
4641 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4642 will now be deprecated.
4644 PH/56 New os.c-cygwin from the Cygwin maintainer.
4646 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4647 formats (previously only available in exim formats via -bpr, -bpru,
4648 and -bpra. Now also available in native and exiqgrep formats)
4650 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4651 with very large, slow to parse queues
4653 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4655 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4657 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4658 responses to authentication challenges, though it was showing the
4659 challenges; (ii) I've removed the CR characters from the debug output for
4662 PH/58 Allow for the insertion of a newline as well as a space when a string
4663 is turned into more than one encoded-word during RFC 2047 encoding. The
4664 Sieve code now uses this.
4666 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4667 data_4xx, lost_connection, tls_required.
4669 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4670 message was being sent as an SMTP response.
4672 PH/61 Add -l and -k options to exicyclog.
4674 PH/62 When verifying, if an address was redirected to one new address, so that
4675 verification continued, and the new address failed or deferred after
4676 having set something in $address_data, the value of $address_data was not
4677 passed back to the ACL. This was different to the case when no
4678 redirection occurred. The value is now passed back in both cases.
4680 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4681 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4682 use login_cap.h, so on its own it isn't the distinguishing feature. The
4683 new name refers directly to the setclassresources() function.
4685 PH/65 Added configuration files for NetBSD3.
4687 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4689 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4690 is preferred over IPv4.
4692 PH/68 The bounce_return_message and bounce_return_body options were not being
4693 honoured for bounces generated during the reception of non-SMTP messages.
4694 In particular, this applied to messages rejected by the ACL. This bug has
4695 been fixed. However, if bounce_return_message is true and bounce_return_
4696 body is false, the headers that are returned for a non-SMTP message
4697 include only those that have been read before the error was detected.
4698 (In the case of an ACL rejection, they have all been read.)
4700 PH/69 The HTML version of the specification is now built in a directory called
4701 spec_html instead of spec.html, because the latter looks like a path with
4702 a MIME-type, and this confuses some software.
4704 PH/70 Catch two compiler warnings in sieve.c.
4706 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4707 function verify_get_ident() calls ip_connect() to connect a socket, but
4708 if the "connect()" function timed out, ip_connect() used to close the
4709 socket. However, verify_get_ident() also closes the socket later, and in
4710 between Exim writes to the log, which may get opened at this point. When
4711 the socket was closed in ip_connect(), the log could get the same file
4712 descriptor number as the socket. This naturally causes chaos. The fix is
4713 not to close the socket in ip_connect(); the socket should be closed by
4714 the function that creates it. There was only one place in the code where
4715 this was missing, in the iplookup router, which I don't think anybody now
4716 uses, but I've fixed it anyway.
4718 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4719 well as to direct DNS lookups. Otherwise the handling of names in host
4720 lists is inconsistent and therefore confusing.
4726 PH/01 Two changes to the default runtime configuration:
4728 (1) Move the checks for relay_from_hosts and authenticated clients from
4729 after to before the (commented out) DNS black list checks.
4731 (2) Add control=submission to the relay_from_hosts and authenticated
4732 clients checks, on the grounds that messages accepted by these
4733 statements are most likely to be submissions.
4735 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4737 (1) Generate an error if the third argument for the ${prvs expansion is
4740 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4743 (3) Reset the variables that are obtained from the first argument of
4744 ${prvscheck and used in the second argument before leaving the code,
4745 because their memory is reclaimed, so using them afterwards may do
4748 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4749 one (it's much easier than Tom thought :-).
4751 (5) Because of (4), we can now allow for the use of $prvscheck_result
4752 inside the third argument.
4754 PH/03 For some reason, the default setting of PATH when running a command from
4755 a pipe transport was just "/usr/bin". I have changed it to
4758 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4759 anything to be listed in the output from -bV.
4761 PH/05 When a filter generated an autoreply, the entire To: header line was
4762 quoted in the delivery log line, like this:
4764 => >A.N.Other <ano@some.domain> <original@ddress> ...
4766 This has been changed so that it extracts the operative address. There
4767 may be more than one such address. If so, they are comma-separated, like
4770 => >ano@some.domain,ona@other.domain <original@ddress> ...
4772 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4773 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4774 not being looked up in the rDNS to get a host name, Exim was showing the
4775 IP address twice in Received: lines, even though the IP addresses were
4776 identical. For example:
4778 Received: from [1.2.3.4] (helo=[1.2.3.4])
4780 However, if the real host name was known, it was omitting the HELO data
4781 if it matched the actual IP address. This has been tidied up so that it
4782 doesn't show the same IP address twice.
4784 PH/07 When both +timestamp and +memory debugging was on, the value given by
4785 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4786 called by the debug printing, thereby overwriting the timestamp buffer.
4787 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4789 PH/08 When the original message was included in an autoreply transport, it
4790 always said "this is a copy of the message, including all the headers",
4791 even if body_only or headers_only was set. It now gives an appropriate
4794 PH/09 Applied a patch from the Sieve maintainer which:
4796 o fixes some comments
4797 o adds the (disabled) notify extension core
4798 o adds some debug output for the result of if/elsif tests
4799 o points to the current vacation draft in the documentation
4800 and documents the missing references header update
4804 o fixes a bug in processing the envelope test (when testing
4805 multiple envelope elements, the last element determined the
4808 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4809 Electronic Mail") by including:
4811 Auto-submitted: auto-generated
4813 in the messages that it generates (bounce messages and others, such as
4814 warnings). In the case of bounce messages for non-SMTP messages, there was
4815 also a typo: it was using "Auto_submitted" (underscore instead of
4816 hyphen). Since every message generated by Exim is necessarily in response
4817 to another message, thes have all been changed to:
4819 Auto-Submitted: auto-replied
4821 in accordance with these statements in the RFC:
4823 The auto-replied keyword:
4825 - SHOULD be used on messages sent in direct response to another
4826 message by an automatic process,
4828 - MUST NOT be used on manually-generated messages,
4830 - MAY be used on Delivery Status Notifications (DSNs) and Message
4831 Disposition Notifications (MDNs),
4833 - MUST NOT be used on messages generated by automatic or periodic
4834 processes, except for messages which are automatic responses to
4837 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4838 to the default Received: header definition.
4840 PH/12 Added log selector acl_warn_skipped (default on).
4842 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4843 variables because (a) they are in the wrong storage pool and (b) even if
4844 they were copied, it wouldn't work properly because of the caching.
4846 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4847 checking when decoding. Apparently there are clients that generate
4848 overlong encoded strings. Why am I not surprised?
4850 PH/15 If the first argument of "${if match_address" was not empty, but did not
4851 contain an "@" character, Exim crashed. Now it writes a panic log message
4852 and treats the condition as false.
4854 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4856 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4857 extension "envelope-auth". The code is finished and in agreement with
4858 other implementations, but there is no documentation so far and in fact,
4859 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4860 not changing the active code.
4862 Print executed "if" and "elsif" statements when debugging is used. This
4863 helps a great deal to understand what a filter does.
4865 Document more things not specified clearly in RFC3028. I had all this
4866 sorted out, when out of a sudden new issues came to my mind. Oops."
4868 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4871 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4872 canonical form (as documented). However, after a host name lookup from
4873 the IP address, check_host() was doing a simple string comparison with
4874 addresses acquired from the DNS when checking that the found name did
4875 have the original IP as one of its addresses. Since any found IPv6
4876 addresses are likely to be in abbreviated form, the comparison could
4877 fail. Luckily, there already exists a function for doing the comparison
4878 by converting both addresses to binary, so now that is used instead of
4879 the text comparison.
4881 PH/20 There was another similar case to PH/19, when a complete host name was
4882 given in a host list; looking up its IP address could give an abbreviated
4883 form, whereas the current host's name might or might not be abbreviated.
4884 The same fix has been applied.
4890 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4891 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4894 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4895 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4897 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4899 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4900 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4901 Also, for "Auto-Submitted", treat anything other than "no" as
4902 non-personal, in accordance with RFC 3834. (Previously it treated
4903 anything starting "auto-" as non-personal.)
4905 TF/01 The control=submission/name=... option had a problem with syntax
4906 errors if the name included a slash character. The /name= option
4907 now slurps the rest of the string, so it can include any characters
4908 but it must come last in the list of options (after /sender_retain
4911 PH/05 Some modifications to the interface to the fake nameserver for the new
4919 TK/01 Added the "success_on_redirect" address verification option. See
4920 NewStuff for rationale and an example.
4922 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4924 PH/02 Patch to exigrep to allow it to work on syslog lines.
4926 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4927 fread() to skip over the body file's header line, because in Cygwin the
4928 header line is locked and is inaccessible.
4930 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4931 co-exist for some time) to make it clear that it is the Exim ID that is
4932 referenced, not the Message-ID: header line.
4934 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4935 string_format() function, because snprintf() does not exist on all
4938 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4939 ${stat: expansion item.
4941 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4942 protocol synchronization error", to keep the pedants happy.
4944 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4945 well as for IRIX systems, when gcc is being used. See the host.c source
4948 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4950 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4953 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4954 transport and to the smtp transport in LMTP mode.
4956 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4958 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4959 run a filter in a subprocess. This could lead to confusion in subsequent
4960 lookups in the parent process. There should also be a search_tidyup() at
4961 the end of the subprocess.
4963 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4964 only if the host matched helo_try_verify_hosts, which caused the
4965 verification to occur when the EHLO/HELO command was issued. The ACL just
4966 tested the remembered result. Now, if a previous verification attempt has
4967 not happened, "verify = helo" does it there and then.
4969 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4971 TK/03 Fix log output including CR from clamd.
4973 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4974 exist provoked a memory error which could cause a segfault.
4976 PH/15 Installed PCRE 6.2
4978 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4980 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4981 of the problem. Specifically, suggested +O2 rather than +O1 for the
4984 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4986 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4987 record, the port was not correctly logged when the outgoing_port log
4988 selector was set (it logged the transort's default port).
4990 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4991 fallback_hosts, and "hosts" in the smtp transport.
4993 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4994 host errors such as "Connection refused".
4996 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4997 authentication with radiusclient 0.4.9:
4999 - Error returned from rc_read_config was caught wrongly
5000 - Username/password not passed on to radius server due to wrong length.
5002 The presumption is that some radiusclient API changes for 4.51/PH/17
5003 were not taken care of correctly. The code is still untested by me (my
5004 Linux distribution still has 0.3.2 of radiusclient), but it was
5005 contributed by a Radius user.
5007 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5008 expanding the "port" option of the smtp transport.
5010 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5011 while reading a MIME header. Thanks to Tom Hughes for a patch.
5013 PH/24 Include config.h inside local_scan.h so that configuration settings are
5016 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5017 that in an ACL for RCPT (for example), you can examine exactly what was
5020 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5021 commands, but it was not correctly comparing the address with the actual
5022 client host address. Thus, it would show the EHLO address in Received:
5023 header lines when this was not necessary.
5025 PH/27 Added the % operator to ${eval:}.
5027 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5028 it should be ignoring failures (because with -C, for example, it has lost
5029 privilege). It wasn't ignoring creation failures other than "already
5032 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5035 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5036 because an input line was too long, either on its own, or by virtue of
5037 too many continuations, the temporary file was not being removed, and the
5038 return code was incorrect.
5040 PH/31 Missing "BOOL" in function definition in filtertest.c.
5042 PH/32 Applied Sieve patches from the maintainer.
5044 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5046 PH/33 Added "verify = not_blind".
5048 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5049 Local/Makefile (with some defaults set). These are used in built scripts
5050 such as exicyclog, but they have never been used in the exim_install
5051 script (though there are many overriding facilities there). I have
5052 arranged that the exim_install script now takes note of these two
5055 PH/35 Installed configuration files for Dragonfly.
5057 PH/36 When a locally submitted message by a trusted user did not contain a
5058 From: header, and the sender address was obtained from -f or from an SMTP
5059 MAIL command, and the trusted user did not use -F to supply a sender
5060 name, $originator_name was incorrectly used when constructing a From:
5061 header. Furthermore, $originator_name was used for submission mode
5062 messages from external hosts without From: headers in a similar way,
5063 which is clearly wrong.
5065 PH/37 Added control=suppress_local_fixups.
5067 PH/38 When log_selector = +received_sender was set, and the addition of the
5068 sender made the log line's construction buffer exactly full, or one byte
5069 less than full, an overflow happened when the terminating "\n" was
5072 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5073 when the result of a list match is failure because a DNS lookup failed.
5075 PH/40 RM_COMMAND is now used in the building process.
5077 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5078 the "build-* directories that it finds.
5080 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5081 address in a domain literal was a prefix of an interface address.
5083 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5084 when verifying a sender address, unless rewrite_headers is false.
5086 PH/44 Wrote a long comment about why errors_to addresses are verified as
5087 recipients, not senders.
5089 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5090 the ratelimit ACL was added.
5092 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5094 PH/46 Added extra information about PostgreSQL errors to the error string.
5096 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5097 avoiding the need to install special zones in a real server. This is
5098 backwards compatible; if it can't find the fake resolver, it drops back.
5099 Thus, both old and new test suites can be run.
5101 TF/02 Added util/ratelimit.pl
5103 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5104 clock is set back in time.
5106 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5107 Candler <B.Candler@pobox.com>.
5109 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5110 OK for addresses that are the result of redirections.
5112 PH/48 A number of further additions for the benefit of the new test suite,
5113 including a fake gethostbyname() that interfaces to the fake DNS resolver
5116 TF/06 The fix for widen_domains has also been applied to qualify_single and
5117 search_parents which are the other dnslookup options that can cause
5120 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5123 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5125 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5126 service" if it encountered an MX record with an empty target hostname.
5127 The message is now "an MX or SRV record indicated no SMTP service".
5129 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5130 if the DNS of the sending site is misconfigured. This is quite a
5131 common situation. This change restores the behaviour of treating a
5132 helo verification defer as a failure.
5134 PH/51 If self=fail was set on a router, the bounce message did not include the
5135 actual error message.
5141 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5143 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5144 command itself ended in error, the underlying message about the transport
5145 filter timeout was being overwritten with the pipe command error. Now the
5146 underlying error message should be appended to the second error message.
5148 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5150 PH/02 Reduce the amount of output that "make" produces by default. Full output
5151 can still be requested.
5153 PH/03 The warning log line about a condition test deferring for a "warn" verb
5154 was being output only once per connection, rather than after each
5155 occurrence (because it was using the same function as for successful
5156 "warn" verbs). This seems wrong, so I have changed it.
5158 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5159 it should not have, which might have caused a crash in the right
5160 circumstances, but probably never did.
5162 PH/04 Installed a modified version of Tony Finch's patch to make submission
5163 mode fix the return path as well as the Sender: header line, and to
5164 add a /name= option so that you can make the user's friendly name appear
5167 TF/03 Added the control = fakedefer ACL modifier.
5169 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5170 Mark Lowes for thorough testing.
5172 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5174 TK/03 Merged latest SRS patch from Miles Wilton.
5176 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5177 with the definition in sysexits.h (which is #included earlier).
5178 Fortunately, Exim does not actually use EX_OK. The code used to try to
5179 preserve the sysexits.h value, by assuming that macro definitions were
5180 scanned for macro replacements. I have been disabused of this notion,
5181 so now the code just undefines EX_OK before #including unistd.h.
5183 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5184 in the smtp transport. When a block could not be written in a single
5185 write() function, the timeout was being re-applied to each part-write.
5186 This seems wrong - if the receiver was accepting one byte at a time it
5187 would take for ever. The timeout is now adjusted when this happens. It
5188 doesn't have to be particularly precise.
5190 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5191 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5193 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5194 without a domain if the check to <postmaster@domain> fails.
5196 SC/01 Eximstats: added -xls and the ability to specify output files
5197 (patch written by Frank Heydlauf).
5199 SC/02 Eximstats: use FileHandles for outputting results.
5201 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5203 SC/04 Eximstats: fixed display of large numbers with -nvr option
5205 SC/05 Eximstats: fixed merging of reports with empty tables.
5207 SC/06 Eximstats: added the -include_original_destination flag
5209 SC/07 Eximstats: removed tabs and trailing whitespace.
5211 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5213 TK/06 MBOX spool code: Add real "From " MBOX separator line
5214 so the .eml file is really in mbox format (even though
5215 most programs do not really care). Patch from Alex Miller.
5217 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5218 The latter is generated from $received_to and is only set if the
5219 message has one envelope recipient. SA can use these headers,
5220 obviously out-of-the-box. Patch from Alex Miller.
5222 PH/08 The ${def test on a variable was returning false if the variable's
5223 value was "0", contrary to what the specification has always said!
5224 The result should be true unless the variable is empty.
5226 PH/09 The syntax error of a character other than { following "${if
5227 def:variable_name" (after optional whitespace) was not being diagnosed.
5228 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5229 accidental colon was present, for example, could give incorrect results.
5231 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5232 result is used (not including appendfile, where other changes are about
5235 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5236 This involved changing a lot of size variables from int to off_t. It
5237 should work with maildirs and everything.
5239 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5240 spamd dying while we are connected to it.
5242 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5245 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5246 function for BDB 4.3.
5248 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5250 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5251 the sources for PCRE have been re-arranged and more files are now
5254 PH/15 The code I had for printing potentially long long variables in PH/11
5255 above was not the best (it lost precision). The length of off_t variables
5256 is now inspected at build time, and an appropriate printing format (%ld
5257 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5258 to be "long long int" or "long int". This is needed for the internal
5259 formatting function string_vformat().
5261 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5262 the configuration file to be ":syslog", then the script "guesses" where
5263 the logs files are, rather than using the compiled in default. In our
5264 case the guess is not the same as the compiled default, so the script
5265 suddenly stopped working when I started to use syslog. The patch checks
5266 to see if log_file_path is "". If so, it attempts to read it from exim
5267 with no configuration file to get the compiled in version, before it
5268 falls back to the previous guessing code."
5270 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5271 implementing BATV in an Exim configuration. See NewStuff for the gory
5274 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5275 Makefile that are specific to HP-UX.
5277 PH/18 If the "use_postmaster" option was set for a recipient callout together
5278 with the "random" option, the postmaster address was used as the MAIL
5279 FROM address for the random test, but not for the subsequent recipient
5280 test. It is now used for both.
5282 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5283 patch removes a few documentation additions to RFC 3028, because the
5284 latest draft now contains them. It adds the new en;ascii-case comparator
5285 and a new error check for 8bit text in MIME parts. Comparator and
5286 require names are now matched exactly. I enabled the subaddress
5287 extension, but it is not well tested yet (read: it works for me)."
5289 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5290 rework some of the code of TK/09 above to avoid the hardwired use of
5291 "%lld" and "long long". Replaced the call to snprintf() with a call to
5294 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5295 records point to non-existent hosts", "retry timeout exceeded", and
5296 "retry time not reached for any host after a long failure period".
5298 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5299 experimental DomainKeys support:
5301 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5302 (2) On an error such as an illegally used "control", the wrong name for
5303 the control was given.
5305 These problems did NOT occur unless DomainKeys support was compiled.
5307 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5309 PH/24 Added ${if match_ip condition.
5311 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5312 databases so that it will be absolutely obvious if a crash occurs in the
5313 DB library. This is a regular occurrence (often caused by mis-matched
5316 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5317 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5318 picked up on a user's system that detects such things. There doesn't seem
5319 to be a gcc warning option for this - only an attribute that has to be
5320 put on the function's prototype. It seems that in Fedora Core 4 they have
5321 set this on a number of new functions. No doubt there will be more in due
5324 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5325 specify a transport. However, if an address that was verified by such a
5326 router was the subject of a callout, Exim crashed because it tried to
5327 read the rcpt_include_affixes from the non-existent transport. Now it
5328 just assumes that the setting of that option is false. This bug was
5329 introduced by 4.51/PH/31.
5331 PH/28 Changed -d+all to exclude +memory, because that information is very
5332 rarely of interest, but it makes the output a lot bigger. People tend to
5333 do -d+all out of habit.
5335 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5336 code in os-type was giving problems when libc.so lives in lib64, like on
5339 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5340 aren't the modern standard, and it seems that some systems' include files
5341 don't always have them. Exim was already checking for some of the newer
5342 ones like T_AAAA, and defining it itself. I've added checks for all the
5343 record types that Exim uses.
5345 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5346 not automatically generating a new one, as it is supposed to. This
5347 prevented TLS from working. If the file did exist, but contained invalid
5348 data, a new version was generated, as expected. It was only the case of a
5349 non-existent file that was broken.
5351 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5352 with a change in libdomainkeys > 0.64.
5354 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5355 from DNS. If the selector record carries the flag, it now has
5356 precedence over the domain-wide flag.
5358 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5360 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5361 the use of a port name that isn't defined in /etc/services) occurred, the
5362 message was deferred as in a normal delivery, and thus remained on the
5363 spool, instead of being failed because of the mua_wrapper setting. This
5364 is now fixed, and I tidied up some of the mua_wrapper messages at the
5367 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5368 the messages in a hash of arrays rather than using individual hashes.
5369 This is a bit cleaner and results in dramatic memory savings, albeit
5370 at a slight CPU cost.
5372 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5373 as requested by Marc Sherman.
5375 SC/10 Eximstats: added histograms for user specified patterns as requested
5378 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5380 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5381 fopen() in the content-scanning modules that did not already have it.
5387 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5388 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5390 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5392 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5394 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5395 to test Sieve filters that use "vacation".
5397 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5398 that changes the way the GnuTLS parameters are stored in the cache file.
5399 The new format can be generated externally. For backward compatibility,
5400 if the data in the cache doesn't make sense, Exim assumes it has read an
5401 old-format file, and it generates new data and writes a new file. This
5402 means that you can't go back to an older release without removing the
5405 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5406 work if there are any delivery delays because "one_time" forces the
5407 parent to be marked "delivered", so its unseen clone is never tried
5408 again. For this reason, Exim now forbids the simultaneous setting of
5411 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5412 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5413 entry if you want to know the details. The fix, however, neglected to
5414 consider the case when local delivery batching is involved. The test for
5415 "previously delivered" was not happening when checking to see if an
5416 address could be batched with a previous (undelivered) one; under
5417 certain circumstances this could lead to multiple deliveries to the same
5420 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5421 in its include files, and this causes problems building Exim.
5423 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5424 header_syntax) but Exim was just ignoring anything given after a slash.
5425 In particular, this caused confusion with an attempt to use "verify =
5426 reverse_host_lookup/defer_ok". An error is now given when options are
5427 supplied for verify items that do not have them. (Maybe reverse_host_
5428 lookup should have a defer_ok option, but that's a different point.)
5430 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5431 defined by RFC 821) to 2048, because there were problems with some AUTH
5432 commands, and RFC 1869 says the size should be increased for extended
5433 SMTP commands that take arguments.
5435 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5438 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5439 "unknown" error; now it says that the functionality isn't in the binary.
5441 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5442 an address' error message when a string expansion fails (syntax or
5443 whatever). Otherwise the password may appear in the log. Following change
5444 PH/42 below, there is no longer a chance of it appearing in a bounce
5447 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5449 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5450 of its addresses was ever tried. (Bugzilla bug #2.)
5452 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5453 the result incorrectly in the debug output. (It correctly added a newline
5454 to what was transported.)
5456 TF/01 Added $received_time.
5458 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5459 commented out examples of how to interface to a virus scanner and to
5460 SpamAssassin. Also added commented examples of av_scanner and
5461 spamd_address settings.
5463 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5464 and controls are allowed in which ACLs. There were a couple of minor
5465 errors. Some of the entries in the conditions table (which is a table of
5466 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5467 negation of where the condition IS allowed.
5469 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5471 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5472 header file does not have a version number, so I've had to invent a new
5473 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5474 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5475 radiusclient), but it was contributed by a Radius user.
5477 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5478 files or directories, for OpenSSL.
5480 PH/19 When an Exim process that is running as root has to create an Exim log
5481 file, it does so in a subprocess that runs as exim:exim so as to get the
5482 ownership right at creation (otherwise, other Exim processes might see
5483 the file with the wrong ownership). There was no test for failure of this
5484 fork() call, which would lead to the process getting stuck as it waited
5485 for a non-existent subprocess. Forks do occasionally fail when resources
5486 run out. I reviewed all the other calls to fork(); they all seem to check
5489 PH/20 When checking for unexpected SMTP input at connect time (before writing
5490 the banner), Exim was not dealing correctly with a non-positive return
5491 from the read() function. If the client had disconnected by this time,
5492 the result was a log entry for a synchronization error with an empty
5493 string after "input=" when read() returned zero. If read() returned -1
5494 (an event I could not check), uninitialized data bytes were printed.
5495 There were reports of junk text (parts of files, etc) appearing after
5498 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5500 PH/22 Added support for macro redefinition, and (re)definition in between
5501 driver and ACL definitions.
5503 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5504 forgetting to use the resulting value; it was using the unexpanded value.
5506 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5507 hadn't been configured. The fix is from Juergen Kreileder, who
5508 understands it better than I do:
5510 "Here's what I see happening with three configured cyrus_sasl
5511 authenticators configured (plain, login, cram-md5):
5513 On startup auth_cyrus_sasl_init() gets called for each of these.
5514 This means three calls to sasl_listmech() without a specified mech_list.
5515 => SASL tests which mechs of all available mechs actually work
5516 => three warnings about OTP not working
5517 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5519 With the patch, sasl_listmech() also gets called three times. But now
5520 SASL's mech_list option is set to the server_mech specified in the the
5521 authenticator. Or in other words, the answer from sasl_listmech()
5522 gets limited to just the mech you're testing for (which is different
5524 => the return list contains just 'plain' or 'login', 'cram-md5' or
5525 nothing depending on the value of ob->server_mech.
5527 I've just tested the patch: Authentication still works fine,
5528 unavailable mechs specified in the exim configuration are still
5529 caught, and the auth.log warnings about OTP are gone."
5531 PH/25 When debugging is enabled, the contents of the command line are added
5532 to the debugging output, even when log_selector=+arguments is not
5535 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5536 answer is "GNU", and only if the return is "GNU/something" is the answer
5539 PH/27 $acl_verify_message is now set immediately after the failure of a
5540 verification in an ACL, and so is available in subsequent modifiers. In
5541 particular, the message can be preserved by coding like this:
5543 warn !verify = sender
5544 set acl_m0 = $acl_verify_message
5546 Previously, $acl_verify_message was set only while expanding "message"
5547 and "log_message" when a very denied access.
5549 PH/28 Modified OS/os.c-Linux with
5551 -#ifndef OS_LOAD_AVERAGE
5552 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5554 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5555 nomenclature these days.)
5557 PH/29 Installed patch from the Sieve maintainer that adds the options
5558 sieve_useraddress and sieve_subaddress to the redirect router.
5560 PH/30 In these circumstances:
5561 . Two addresses routed to the same list of hosts;
5562 . First host does not offer TLS;
5563 . First host accepts first address;
5564 . First host gives temporary error to second address;
5565 . Second host offers TLS and a TLS session is established;
5566 . Second host accepts second address.
5567 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5568 and peerdn, if requested) that were in fact used only for the second
5571 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5572 attention to any local part prefix or suffix that was matched by the
5573 router that accepted the address. It now behaves in the same way as it
5574 does for delivery: the affixes are removed from the local part unless
5575 rcpt_include_affixes is set on the transport.
5577 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5578 timeout during the DATA phase of an incoming message.
5580 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5581 applied a patch sanctioned by the Sieve maintainer.
5583 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5584 the uid or gid is negative. A case of a negative gid caused this to be
5585 noticed. The fix allows for either to be negative.
5587 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5588 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5591 PH/36 The change PH/12 above was broken. Fixed it.
5593 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5594 the grounds that routing the same address twice would always produce the
5595 same answer. This might have been true once, but it is certainly no
5596 longer true now. Routing a child address may depend on the previous
5597 routing that produced that child. Some complicated redirection strategies
5598 went wrong when messages had multiple recipients, and made Exim's
5599 behaviour dependent on the order in which the addresses were given.
5601 I have moved the duplicate checking until after the routing is complete.
5602 Exim scans the addresses that are assigned to local and remote
5603 transports, and removes any duplicates. This means that more work will be
5604 done, as duplicates will always all be routed, but duplicates are
5605 presumably rare, so I don't expect this is of any significance.
5607 For deliveries to pipes, files, and autoreplies, the duplicate checking
5608 still happens during the routing process, since they are not going to be
5611 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5612 It corrects a timeout issue with spamd. This is Ian's comment: "The
5613 background is that sometimes spamd either never reads data from a
5614 connection it has accepted, or it never writes response data. The exiscan
5615 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5616 blindly assumes that writes won't block so it may never time out."
5618 PH/39 Allow G after quota size as well as K and M.
5620 PH/40 The value set for $authenticated_id in an authenticator may not contain
5621 binary zeroes or newlines because the value is written to log lines and
5622 to spool files. There was no check on this. Now the value is run through
5623 the string_printing() function so that such characters are converted to
5624 printable escape sequences.
5626 PH/41 $message_linecount is a new variable that contains the total number of
5627 lines in the message. Compare $body_linecount, which is the count for the
5630 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5631 bounce and delay warning messages, except in certain special cases, which
5634 (a) An SMTP error message from a remote host;
5635 (b) A message specified in a :fail: redirection;
5636 (c) A message specified in a "fail" command in a system filter;
5637 (d) A message specified in a FAIL return from the queryprogram router;
5638 (e) A message specified by the cannot_route_message router option.
5640 In these cases only, Exim does include the error details in bounce and
5641 warning messages. There are also a few cases where bland messages such
5642 as "unrouteable address" or "local delivery error" are given.
5644 PH/43 $value is now also set for the "else" part of a ${run expansion.
5646 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5647 being worked on, but at least Exim now implements the latest version to
5650 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5651 process to complete was treated as a delivery failure, a timeout while
5652 writing the message to the pipe was logged, but erroneously treated as a
5653 successful delivery. Such timeouts include transport filter timeouts. For
5654 consistency with the overall process timeout, these timeouts are now
5655 treated as errors, giving rise to delivery failures by default. However,
5656 there is now a new Boolean option for the pipe transport called
5657 timeout_defer, which, if set TRUE, converts the failures into defers for
5658 both kinds of timeout. A transport filter timeout is now identified in
5661 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5662 systems where "make" and "gmake" are different, calling "gmake" at top
5663 level broke things. I've arranged for the value of $(MAKE) to be passed
5664 from the Makefile to this script so that it can call the same version of
5668 A note about Exim versions 4.44 and 4.50
5669 ----------------------------------------
5671 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5672 changes of various kinds. As a consequence, a big documentation update was
5673 needed. This delayed the release for rather longer than seemed good, especially
5674 in the light of a couple of (minor) security issues. Therefore, the changes
5675 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5676 release. So 4.44 and 4.50 are in effect two different branches that both start
5679 I have left the 4.50 change log unchanged; it contains all the changes since
5680 4.43. The change log for 4.44 is below; many of its items are identical to
5681 those for 4.50. This seems to be the most sensible way to preserve the
5682 historical information.
5688 1. Minor wording change to the doc/README.SIEVE file.
5690 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5691 computation of the current number of files was incorrect.
5693 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5694 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5697 4. Give more explanation in the error message when the command for a transport
5698 filter fails to execute.
5700 5. There are several places where Exim runs a non-Exim command in a
5701 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5702 was being done only for the command run by the queryprogram router. It is
5703 now done for all such subprocesses. The other cases are: ${run, transport
5704 filters, and the commands run by the lmtp and pipe transports.
5706 6. Added CONFIGURE_GROUP build-time option.
5708 7. Some older OS have a limit of 256 on the maximum number of file
5709 descriptors. Exim was using setrlimit() to set 1000 as a large value
5710 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5711 systems. I've change it so that if it can't get 1000, it tries for 256.
5713 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5714 was an oversight, and furthermore, ever since the addition of extra
5715 controls (e.g. 4.43/32), the checks on when to allow different forms of
5716 "control" were broken. There should now be diagnostics for all cases when a
5717 control that does not make sense is encountered.
5719 9. Added the /retain_sender option to "control=submission".
5721 10. $recipients is now available in the predata ACL (oversight).
5723 11. Tidy the search cache before the fork to do a delivery from a message
5724 received from the command line. Otherwise the child will trigger a lookup
5725 failure and thereby defer the delivery if it tries to use (for example) a
5726 cached ldap connection that the parent has called unbind on.
5728 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5729 of $address_data from the recipient verification was clobbered by the
5730 sender verification.
5732 13. The value of address_data from a sender verification is now available in
5733 $sender_address_data in subsequent conditions in the ACL statement.
5735 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5737 15. Added a new option "connect=<time>" to callout options, to set a different
5740 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5741 was its contents. (It was OK if the option was not defined at all.)
5743 17. A "Completed" log line is now written for messages that are removed from
5744 the spool by the -Mrm option.
5746 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5747 information about exactly what failed.
5749 19. Added -dd to debug only the daemon process.
5751 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5752 handles timeouts, both on the server side and network timeouts. Renamed the
5753 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5755 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5756 It is now set to "smtps".
5758 22. $host_address is now set to the target address during the checking of
5759 ignore_target_hosts.
5761 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5762 being passed; this would have caused $sender_host_name to have been used if
5763 matching the list had actually called for a host name (not very likely,
5764 since this list is usually IP addresses). A host name is now passed as
5767 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5768 code that specifies a non-restarting handler (typically sigaction() in
5769 modern systems) in an attempt to fix a rare and obscure crash bug.
5771 25. Narrowed the window for a race in the daemon that could cause it to ignore
5772 SIGCHLD signals. This is not a major problem, because they are used only to
5773 wake it up if nothing else does.
5775 26. A malformed maildirsize file could cause Exim to calculate negative values
5776 for the mailbox size or file count. Odd effects could occur as a result.
5777 The maildirsize information is now recalculated if the size or filecount
5780 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5781 support for a long time. Removed HAVE_SYS_VFS_H.
5783 28. Installed the latest version of exipick from John Jetmore.
5785 29. In an address list, if the pattern was not a regular expression, an empty
5786 subject address (from a bounce message) matched only if the pattern was an
5787 empty string. Non-empty patterns were not even tested. This was the wrong
5788 because it is perfectly reasonable to use an empty address as part of a
5789 database query. An empty address is now tested by patterns that are
5790 lookups. However, all the other forms of pattern expect the subject to
5791 contain a local part and a domain, and therefore, for them, an empty
5792 address still always fails if the pattern is not itself empty.
5794 30. Exim went into a mad DNS loop when attempting to do a callout where the
5795 host was specified on an smtp transport, and looking it up yielded more
5796 than one IP address.
5798 31. Re-factored the code for checking spool and log partition space into a
5799 function that finds that data and another that does the check. The former
5800 is then used to implement four new variables: $spool_space, $log_space,
5801 $spool_inodes, and $log_inodes.
5803 32. The RFC2047 encoding function was originally intended for short strings
5804 such as real names; it was not keeping to the 75-character limit for
5805 encoded words that the RFC imposes. It now respects the limit, and
5806 generates multiple encoded words if necessary. To be on the safe side, I
5807 have increased the buffer size for the ${rfc2047: expansion operator from
5810 33. It is now permitted to omit both strings after an "if" condition; if the
5811 condition is true, the result is "true". As before, when the second string
5812 is omitted, a false condition yields an empty string. This makes it less
5813 cumbersome to write custom ACL and router conditions.
5815 34. Failure to deliver a bounce message always caused it to be frozen, even if
5816 there was an errors_to setting on the router. The errors_to setting is now
5819 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5820 canonical form (fully expanded) before being placed in
5821 $sender_host_address.
5823 36. The table in the code that translates DNS record types into text (T_A to
5824 "A" for instance) was missing entries for NS and CNAME. It is just possible
5825 that this could have caused confusion if both these types were looked up
5826 for the same domain, because the text type is used as part of Exim's
5827 per-process caching. But the chance of anyone hitting this buglet seems
5830 37. The dnsdb lookup has been extended in a number of ways.
5832 (1) There is a new type, "zns", which walks up the domain tree until it
5833 finds some nameserver records. It should be used with care.
5835 (2) There is a new type, "mxh", which is like "mx" except that it returns
5836 just the host names, not the priorities.
5838 (3) It is now possible to give a list of domains (or IP addresses) to be
5839 looked up. The behaviour when one of the lookups defers can be
5840 controlled by a keyword.
5842 (4) It is now possible to specify the separator character for use when
5843 multiple records are returned.
5845 38. The dnslists ACL condition has been extended: it is now possible to supply
5846 a list of IP addresses and/or domains to be looked up in a particular DNS
5849 39. Added log_selector=+queue_time_overall.
5851 40. When running the queue in the test harness, wait just a tad after forking a
5852 delivery process, to get repeatability of debugging output.
5854 41. Include certificate and key file names in error message when GnuTLS fails
5855 to set them up, because the GnuTLS error message doesn't include the name
5856 of the failing file when there is a problem reading it.
5858 42. Allow both -bf and -bF in the same test run.
5860 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5862 44. The "Exiscan patch" is now merged into the mainline Exim source.
5864 45. Sometimes the final signoff response after QUIT could fail to get
5865 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5866 < 0 before doing a fflush(). This bug looks as though it goes back to the
5867 introduction of TLS in release 3.20, but "sometimes" must have been rare
5868 because the tests only now provoked it.
5870 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5871 (this can affect the format of dates).
5873 47. exim_tidydb, when checking for the continued existence of a message for
5874 which it has found a message-specific retry record, was not finding
5875 messages that were in split spool directories. Consequently, it was
5876 deleting retry records that should have stayed in existence.
5878 48. Steve fixed some bugs in eximstats.
5880 49. The SPA authentication driver was not abandoning authentication and moving
5881 on to the next authenticator when an expansion was forced to fail,
5882 contradicting the general specification for all authenticators. Instead it
5883 was generating a temporary error. It now behaves as specified.
5885 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5886 (the order specifies the preference for clients). The order is now AES256,
5887 AES128, 3DES, ARCFOUR128.
5889 51. Small patch to Sieve code - explicitly set From: when generating an
5892 52. Exim crashed if a remote delivery caused a very long error message to be
5893 recorded - for instance if somebody sent an entire SpamAssassin report back
5894 as a large number of 550 error lines. This bug was coincidentally fixed by
5895 increasing the size of one of Exim's internal buffers (big_buffer) that
5896 happened as part of the Exiscan merge. However, to be on the safe side, I
5897 have made the code more robust (and fixed the comments that describe what
5900 53. Now that there can be additional text after "Completed" in log lines (if
5901 the queue_time_overall log selector is set), a one-byte patch to exigrep
5902 was needed to allow it to recognize "Completed" as not the last thing in
5905 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5906 patch that reportedly fixes this has been added. I am not expert enough to
5907 create a test for it. This is what the patch creator wrote:
5909 "I found a little strange behaviour of ldap code when working with
5910 Windows 2003 AD Domain, where users was placed in more than one
5911 Organization Units. When I tried to give exim partial DN, the exit code
5912 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5913 But simultaneously result of request was absolutely normal ldap result,
5914 so I produce this patch..."
5916 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5917 REFERENCE, so I have modified the code to exclude the patch when that macro
5920 55. Some experimental protocols are using DNS PTR records for new purposes. The
5921 keys for these records are domain names, not reversed IP addresses. The
5922 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5923 leaves it alone. Component reversal etc. now happens only for IP addresses.
5926 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5928 57. Double the size of the debug message buffer (to 2048) so that more of very
5929 long debug lines gets shown.
5931 58. The exicyclog utility now does better if the number of log files to keep
5932 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5934 59. Two changes related to the smtp_active_hostname option:
5936 (1) $smtp_active_hostname is now available as a variable.
5937 (2) The default for smtp_banner uses $smtp_active_hostname instead
5938 of $primary_hostname.
5940 60. The host_aton() function is supposed to be passed a string that is known
5941 to be a valid IP address. However, in the case of IPv6 addresses, it was
5942 not checking this. This is a hostage to fortune. Exim now panics and dies
5943 if the condition is not met. A case was found where this could be provoked
5944 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5945 components; fortuitously, this particular loophole had already been fixed
5946 by change 4.50/55 above.
5948 If there are any other similar loopholes, the new check in host_aton()
5949 itself should stop them being exploited. The report I received stated that
5950 data on the command line could provoke the exploit when Exim was running as
5951 exim, but did not say which command line option was involved. All I could
5952 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5953 running as the user.
5956 61. There was a buffer overflow vulnerability in the SPA authentication code
5957 (which came originally from the Samba project). I have added a test to the
5958 spa_base64_to_bits() function which I hope fixes it.
5961 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5962 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5964 63. The daemon start-up calls getloadavg() while still root for those OS that
5965 need the first call to be done as root, but it missed one case: when
5966 deliver_queue_load_max is set with deliver_drop_privilege. This is
5967 necessary for the benefit of the queue runner, because there is no re-exec
5968 when deliver_drop_privilege is set.
5970 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5971 This has been fixed.
5973 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5974 was in use, was not putting the data itself into the right store pool;
5975 consequently, it could be overwritten for a subsequent message in the same
5976 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5979 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5981 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5982 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5983 the function were treating the return as a boolean value, which happened to
5984 work because 0=false and not-0=true, but is not correct code.
5986 68. The host_aton() function was not handling scoped IPv6 addresses (those
5987 with, for example, "%eth0" on the end) correctly.
5989 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5990 negated items (that is, ~something) in unsigned ints. Some compilers
5991 apparently mutter when there is no cast.
5993 70. If an address verification called from an ACL failed, and did not produce a
5994 user-specific message (i.e. there was only a "system" message), nothing was
5995 put in $acl_verify_message. In this situation, it now puts the system
5998 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5999 session; change 4.31/43 added the unwanted input to the log line - except
6000 that it did not do this in the start of session case. It now does.
6002 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6003 This is wrong and can cause the other end to generate a synchronization
6004 error if it is another Exim or anything else that does the synchronization
6005 check. A QUIT command is no longer sent after a timeout.
6007 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6008 during host lookups.
6010 74. The defer_ok option of callout verification was not working if it was used
6011 when verifying addresses in header lines, that is, for this case:
6013 verify = header_sender/callout=defer_ok
6015 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6016 those file descriptors could be used for SMTP connections. If anything
6017 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6018 could be sent to the SMTP client, causing chaos. The daemon now opens
6019 stdin, stdout, and stderr to /dev/null when it puts itself into the
6022 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6023 log by default. The user can override this with suitable Perl magic.
6025 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6026 the log message when discard triggers, was not working for the DATA ACL or
6027 for the non-SMTP ACL.
6029 78. Error message wording change in sieve.c.
6031 79. If smtp_accept_max_per_host was set, the number of connections could be
6032 restricted to fewer than expected, because the daemon was trying to set up
6033 a new connection before checking whether the processes handling previous
6034 connections had finished. The check for completed processes is now done
6035 earlier. On busy systems, this bug wouldn't be noticed because something
6036 else would have woken the daemon, and it would have reaped the completed
6039 80. If a message was submitted locally by a user whose login name contained one
6040 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6041 It caused a spool format error. I have fixed the spool reading code. A
6042 related problem was that the "from" clause in the Received: line became
6043 illegal because of the space(s). It is now covered by ${quote_local_part.
6045 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6048 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6050 83. Patch from Sieve maintainer to fix unterminated string problem in
6051 "vacation" handling.
6053 84. Some minor changes to the Linux configuration files to help with other
6054 OS variants using glibc.
6056 85. One more patch for Sieve to update vacation handling to latest spec.
6059 ----------------------------------------------------
6060 See the note above about the 4.44 and 4.50 releases.
6061 ----------------------------------------------------
6067 1. Change 4.43/35 introduced a bug that caused file counts to be
6068 incorrectly computed when quota_filecount was set in an appendfile
6071 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6072 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6075 3. Give more explanation in the error message when the command for a transport
6076 filter fails to execute.
6078 4. There are several places where Exim runs a non-Exim command in a
6079 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6080 was being done only for the command run by the queryprogram router. It is
6081 now done for all such subprocesses. The other cases are: ${run, transport
6082 filters, and the commands run by the lmtp and pipe transports.
6084 5. Some older OS have a limit of 256 on the maximum number of file
6085 descriptors. Exim was using setrlimit() to set 1000 as a large value
6086 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6087 systems. I've change it so that if it can't get 1000, it tries for 256.
6089 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6090 was an oversight, and furthermore, ever since the addition of extra
6091 controls (e.g. 4.43/32), the checks on when to allow different forms of
6092 "control" were broken. There should now be diagnostics for all cases when a
6093 control that does not make sense is encountered.
6095 7. $recipients is now available in the predata ACL (oversight).
6097 8. Tidy the search cache before the fork to do a delivery from a message
6098 received from the command line. Otherwise the child will trigger a lookup
6099 failure and thereby defer the delivery if it tries to use (for example) a
6100 cached ldap connection that the parent has called unbind on.
6102 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6103 of $address_data from the recipient verification was clobbered by the
6104 sender verification.
6106 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6107 was its contents. (It was OK if the option was not defined at all.)
6109 11. A "Completed" log line is now written for messages that are removed from
6110 the spool by the -Mrm option.
6112 12. $host_address is now set to the target address during the checking of
6113 ignore_target_hosts.
6115 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6116 being passed; this would have caused $sender_host_name to have been used if
6117 matching the list had actually called for a host name (not very likely,
6118 since this list is usually IP addresses). A host name is now passed as
6121 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6122 code that specifies a non-restarting handler (typically sigaction() in
6123 modern systems) in an attempt to fix a rare and obscure crash bug.
6125 15. Narrowed the window for a race in the daemon that could cause it to ignore
6126 SIGCHLD signals. This is not a major problem, because they are used only to
6127 wake it up if nothing else does.
6129 16. A malformed maildirsize file could cause Exim to calculate negative values
6130 for the mailbox size or file count. Odd effects could occur as a result.
6131 The maildirsize information is now recalculated if the size or filecount
6134 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6135 support for a long time. Removed HAVE_SYS_VFS_H.
6137 18. Updated exipick to current release from John Jetmore.
6139 19. Allow an empty sender to be matched against a lookup in an address list.
6140 Previously the only cases considered were a regular expression, or an
6143 20. Exim went into a mad DNS lookup loop when doing a callout where the
6144 host was specified on the transport, if the DNS lookup yielded more than
6147 21. The RFC2047 encoding function was originally intended for short strings
6148 such as real names; it was not keeping to the 75-character limit for
6149 encoded words that the RFC imposes. It now respects the limit, and
6150 generates multiple encoded words if necessary. To be on the safe side, I
6151 have increased the buffer size for the ${rfc2047: expansion operator from
6154 22. Failure to deliver a bounce message always caused it to be frozen, even if
6155 there was an errors_to setting on the router. The errors_to setting is now
6158 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6159 canonical form (fully expanded) before being placed in
6160 $sender_host_address.
6162 24. Updated eximstats to version 1.33
6164 25. Include certificate and key file names in error message when GnuTLS fails
6165 to set them up, because the GnuTLS error message doesn't include the name
6166 of the failing file when there is a problem reading it.
6168 26. Expand error message when OpenSSL has problems setting up cert/key files.
6171 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6172 (this can affect the format of dates).
6174 28. exim_tidydb, when checking for the continued existence of a message for
6175 which it has found a message-specific retry record, was not finding
6176 messages that were in split spool directories. Consequently, it was
6177 deleting retry records that should have stayed in existence.
6179 29. eximstats updated to version 1.35
6180 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6181 1.35 - bugfix such that pie charts by volume are generated correctly
6183 30. The SPA authentication driver was not abandoning authentication and moving
6184 on to the next authenticator when an expansion was forced to fail,
6185 contradicting the general specification for all authenticators. Instead it
6186 was generating a temporary error. It now behaves as specified.
6188 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6189 (the order specifies the preference for clients). The order is now AES256,
6190 AES128, 3DES, ARCFOUR128.
6192 31. Small patch to Sieve code - explicitly set From: when generating an
6195 32. Exim crashed if a remote delivery caused a very long error message to be
6196 recorded - for instance if somebody sent an entire SpamAssassin report back
6197 as a large number of 550 error lines. This bug was coincidentally fixed by
6198 increasing the size of one of Exim's internal buffers (big_buffer) that
6199 happened as part of the Exiscan merge. However, to be on the safe side, I
6200 have made the code more robust (and fixed the comments that describe what
6203 33. Some experimental protocols are using DNS PTR records for new purposes. The
6204 keys for these records are domain names, not reversed IP addresses. The
6205 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6206 leaves it alone. Component reversal etc. now happens only for IP addresses.
6209 34. The host_aton() function is supposed to be passed a string that is known
6210 to be a valid IP address. However, in the case of IPv6 addresses, it was
6211 not checking this. This is a hostage to fortune. Exim now panics and dies
6212 if the condition is not met. A case was found where this could be provoked
6213 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6214 components; fortuitously, this particular loophole had already been fixed
6215 by change 4.50/55 or 4.44/33 above.
6217 If there are any other similar loopholes, the new check in host_aton()
6218 itself should stop them being exploited. The report I received stated that
6219 data on the command line could provoke the exploit when Exim was running as
6220 exim, but did not say which command line option was involved. All I could
6221 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6222 running as the user.
6225 35. There was a buffer overflow vulnerability in the SPA authentication code
6226 (which came originally from the Samba project). I have added a test to the
6227 spa_base64_to_bits() function which I hope fixes it.
6230 36. The daemon start-up calls getloadavg() while still root for those OS that
6231 need the first call to be done as root, but it missed one case: when
6232 deliver_queue_load_max is set with deliver_drop_privilege. This is
6233 necessary for the benefit of the queue runner, because there is no re-exec
6234 when deliver_drop_privilege is set.
6236 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6237 was in use, was not putting the data itself into the right store pool;
6238 consequently, it could be overwritten for a subsequent message in the same
6239 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6242 38. Sometimes the final signoff response after QUIT could fail to get
6243 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6244 < 0 before doing a fflush(). This bug looks as though it goes back to the
6245 introduction of TLS in release 3.20, but "sometimes" must have been rare
6246 because the tests only now provoked it.
6252 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6253 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6254 Now it returns an integer. A number of calls were still expecting a T/F
6255 return. Fortuitously, in all cases, the tests worked in OK situations,
6256 which is the norm. However, things would have gone wrong on any write
6257 failures on the smtp file descriptor. This function is used when sending
6258 messages over SMTP and also when doing verify callouts.
6260 2. When Exim is called to do synchronous delivery of a locally submitted
6261 message (the -odf or -odi options), it no longer closes stderr before doing
6264 3. Implemented the mua_wrapper option.
6266 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6268 5. Implemented the functions header_remove(), header_testname(),
6269 header_add_at_position(), and receive_remove_recipient(), and exported them
6272 6. If an ACL "warn" statement specified the addition of headers, Exim already
6273 inserted X-ACL-Warn: at the start if there was no header name. However, it
6274 was not making this test for the second and subsequent header lines if
6275 there were newlines in the string. This meant that an invalid header could
6276 be inserted if Exim was badly configured.
6278 7. Allow an ACL "warn" statement to add header lines at the start or after all
6279 the Received: headers, as well as at the end.
6281 8. Added the rcpt_4xx retry error code.
6283 9. Added postmaster_mailfrom=xxx to callout verification option.
6285 10. Added mailfrom=xxxx to the callout verification option, for verify=
6288 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6289 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6291 12. Inserted some casts to stop certain compilers warning when using pointer
6292 differences as field lengths or precisions in printf-type calls (mostly
6293 affecting debugging statements).
6295 13. Added optional readline() support for -be (dynamically loaded).
6297 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6298 same clock tick as a message's arrival, so that its received time was the
6299 same as the "first fail" time on the retry record, and that message
6300 remained on the queue past the ultimate address timeout, every queue runner
6301 would try a delivery (because it was past the ultimate address timeout) but
6302 after another failure, the ultimate address timeout, which should have then
6303 bounced the address, did not kick in. This was a "< instead of <=" error;
6304 in most cases the first failure would have been in the next clock tick
6305 after the received time, and all would be well.
6307 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6308 being recognized when the domain list was tested by the match_domain
6309 condition in an expansion string.
6311 16. Added the ${str2b64: operator.
6313 17. Exim was always calling setrlimit() to set a large limit for the number of
6314 processes, without checking whether the existing limit was already
6315 adequate. (It did check for the limit on file descriptors.) Furthermore,
6316 errors from getrlimit() and setrlimit() were being ignored. Now they are
6317 logged to the main and panic logs, but Exim does carry on, to try to do its
6318 job under whatever limits there are.
6320 18. Imported PCRE 5.0.
6322 19. Trivial typo in log message " temporarily refused connection" (the leading
6325 20. If the log selector return_path_on_delivery was set and an address was
6326 redirected to /dev/null, the delivery process crashed because it assumed
6327 that a return path would always be set for a "successful" delivery. In this
6328 case, the whole delivery is bypassed as an optimization, and therefore no
6331 21. Internal re-arrangement: the function for sending a challenge and reading
6332 a response while authentication was assuming a zero-terminated challenge
6333 string. It's now changed to take a pointer and a length, to allow for
6334 binary data in such strings.
6336 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6338 23. Exim was not respecting finduser_retries when seeking the login of the
6339 uid under which it was called; it was always trying 10 times. (The default
6340 setting of finduser_retries is zero.) Also, it was sleeping after the final
6341 failure, which is pointless.
6343 24. Implemented tls_on_connect_ports.
6345 25. Implemented acl_smtp_predata.
6347 26. If the domain in control=submission is set empty, Exim assumes that the
6348 authenticated id is a complete email address when it generates From: or
6349 Sender: header lines.
6351 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6352 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6353 chown and chgrp in /bin and hostname in /usr/bin.
6355 28. Exim was keeping the "process log" file open after each use, just as it
6356 does for the main log. This opens the possibility of it remaining open for
6357 long periods when the USR1 signal hits a daemon. Occasional processlog
6358 errors were reported, that could have been caused by this. Anyway, it seems
6359 much more sensible not to leave this file open at all, so that is what now
6362 29. The long-running daemon process does not normally write to the log once it
6363 has entered its main loop, and it closes the log before doing so. This is
6364 so that log files can straightforwardly be renamed and moved. However,
6365 there are a couple of unusual error situations where the daemon does write
6366 log entries, and I had neglected to close the log afterwards.
6368 30. The text of an SMTP error response that was received during a remote
6369 delivery was being truncated at 512 bytes. This is too short for some of
6370 the long messages that one sometimes sees. I've increased the limit to
6373 31. It is now possible to make retry rules that apply only when a message has a
6374 specific sender, in particular, an empty sender.
6376 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6377 it possible to be selective about when SMTP synchronization is enforced.
6379 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6381 32. Implemented hosts_connection_nolog.
6383 33. Added an ACL for QUIT.
6385 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6388 35. Added mailbox_size and mailbox_filecount to appendfile.
6390 36. Added control = no_multiline_responses to ACLs.
6392 37. There was a bug in the logic of the code that waits for the clock to tick
6393 in the case where the clock went backwards by a substantial amount such
6394 that the microsecond fraction of "now" was more than the microsecond
6395 fraction of "then" (but the whole seconds number was less).
6397 38. Added support for the libradius Radius client library this is found on
6398 FreeBSD (previously only the radiusclient library was supported).
6404 1. When certain lookups returned multiple values in the form name=value, the
6405 quoting of the values was not always being done properly. Specifically:
6406 (a) If the value started with a double quote, but contained no whitespace,
6408 (b) If the value contained whitespace other than a space character (i.e.
6409 tabs or newlines or carriage returns) it was not quoted.
6410 This fix has been applied to the mysql and pgsql lookups by writing a
6411 separate quoting function and calling it from the lookup code. The fix
6412 should probably also be applied to nisplus, ibase and oracle lookups, but
6413 since I cannot test any of those, I have not disturbed their existing code.
6415 2. A hit in the callout cache for a specific address caused a log line with no
6416 reason for rejecting RCPT. Now it says "Previous (cached) callout
6417 verification failure".
6419 3. There was an off-by-one bug in the queryprogram router. An over-long
6420 return line was truncated at 256 instead of 255 characters, thereby
6421 overflowing its buffer with the terminating zero. As well as fixing this, I
6422 have increased the buffer size to 1024 (and made a note to document this).
6424 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6425 when Exim is waiting for an SMTP response from a remote server, Exim
6426 restarts its select() call on the socket, thereby resetting its timeout.
6427 This is not a problem when such interrupts are rare. Somebody set up a cron
6428 job to run exiwhat every 2 minutes, which is less than the normal select()
6429 timeout (5 or 10 minutes). This meant that the select() timeout never
6430 kicked in because it was always reset. I have fixed this by comparing the
6431 time when an interrupt arrives with the time at the start of the first call
6432 to select(). If more time than the timeout has elapsed, the interrupt is
6433 treated as a timeout.
6435 5. Some internal re-factoring in preparation for the addition of Sieve
6436 extensions (by MH). In particular, the "personal" test is moved to a
6437 separate function, and given an option for scanning Cc: and Bcc: (which is
6438 not set for Exim filters).
6440 6. When Exim created an email address using the login of the caller as the
6441 local part (e.g. when creating a From: or Sender: header line), it was not
6442 quoting the local part when it contained special characters such as @.
6444 7. Installed new OpenBSD configuration files.
6446 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6447 try to make them clearer.
6449 9. Callout options, other than the timeout value, were being ignored when
6450 verifying sender addresses in header lines. For example, when using
6452 verify = header_sender/callout=no_cache
6454 the cache was (incorrectly) being used.
6456 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6457 only the Cygwin environment.
6459 11. When return_path_on_delivery was set as a log selector, if different remote
6460 addresses in the same message used different return paths and parallel
6461 remote delivery occurred, the wrong values would sometimes be logged.
6462 (Whenever a remote delivery process finished, the return path value from
6463 the most recently started remote delivery process was logged.)
6465 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6466 lines when AUTH and/or TLS are in use. This is the "received protocol"
6467 field. Exim used to use "asmtp" for authenticated SMTP, without any
6468 indication (in the protocol name) for TLS use. Now it follows the RFC and
6469 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6470 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6471 names appear in log lines as well as in Received: header lines.
6473 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6474 extensions, and comparison tests, and to fix some bugs.
6476 14. Changes to the "personal" filter test:
6478 (1) The test was buggy in that it was just doing the equivalent of
6479 "contains" tests on header lines. For example, if a user's address was
6480 anne@some.where, the "personal" test would incorrectly be true for
6482 To: susanne@some.where
6484 This test is now done by extracting each address from the header in turn,
6485 and checking the entire address. Other tests that are part of "personal"
6486 are now done using regular expressions (for example, to check local parts
6487 of addresses in From: header lines).
6489 (2) The list of non-personal local parts in From: addresses has been
6490 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6491 taken from the Sieve specification recommendations.
6493 (3) If the message contains any header line starting with "List-" it is
6494 treated as non-personal.
6496 (4) The test for "circular" in the Subject: header line has been removed
6497 because it now seems ill-conceived.
6499 15. Minor typos in src/EDITME comments corrected.
6501 16. Installed latest exipick from John Jetmore.
6503 17. If headers_add on a router specified a text string that was too long for
6504 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6505 of string_sprintf() is now avoided.
6507 18. $message_body_size was not set (it was always zero) when running the DATA
6508 ACL and the local_scan() function.
6510 19. For the "mail" command in an Exim filter, no default was being set for
6511 the once_repeat time, causing a random time value to be used if "once" was
6512 specified. (If the value happened to be <= 0, no repeat happened.) The
6513 default is now 0s, meaning "never repeat". The "vacation" command was OK
6514 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6515 (I found it when inspecting the code).
6517 20. There is now an overall timeout for performing a callout verification. It
6518 defaults to 4 times the callout timeout, which applies to individual SMTP
6519 commands during the callout. The overall timeout applies when there is more
6520 than one host that can be tried. The timeout is checked before trying the
6521 next host. This prevents very long delays if there are a large number of
6522 hosts and all are timing out (e.g. when the network connections are timing
6523 out). The value of the overall timeout can be changed by specifying an
6524 additional sub-option for "callout", called "maxwait". For example:
6526 verify = sender/callout=5s,maxwait=20s
6528 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6529 to the end before writing, but this should make it even safer).
6531 22. Exim was forgetting that it had advertised PIPELINING for the second and
6532 subsequent messages on an SMTP connection. It was also not resetting its
6533 memory on STARTTLS and an internal HELO.
6535 23. When Exim logs an SMTP synchronization error within a session, it now
6536 records whether PIPELINING has been advertised or not.
6538 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6539 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6540 rather than long int.
6542 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6544 26. Added the never_mail option to autoreply.
6550 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6551 crash if the getsockname() call failed; this can happen if a connection is
6552 closed very soon after it is established. The problem was simply in the
6553 order in which certain operations were done, causing Exim to try to write
6554 to the SMTP stream before it had set up the file descriptor. The bug has
6555 been fixed by making things happen in the correct order.
6561 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6562 before the connection was closed, thus losing the rejection response.
6564 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6565 some early Solaris releases, but causes trouble in current releases where
6566 socklen_t is defined.
6568 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6571 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6574 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6575 that releasing the top of it at the end releases what was used for sub-
6576 expansions (unless the block got too big). However, discard this block if
6577 the first thing is a variable or header, so that we can use its block when
6578 it is dynamic (useful for very large $message_headers, for example).
6580 6. Lookups now cache *every* query, not just the most recent. A new, separate
6581 store pool is used for this. It can be recovered when all lookup caches are
6582 flushed. Lookups now release memory at the end of their result strings.
6583 This has involved some general refactoring of the lookup sources.
6585 7. Some code has been added to the store_xxx() functions to reduce the amount
6586 of flapping under certain conditions.
6588 8. log_incoming_interface used to affect only the <= reception log lines. Now
6589 it causes the local interface and port to be added to several more SMTP log
6590 lines, for example "SMTP connection from", and rejection lines.
6592 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6594 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6596 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6597 could be overwritten at the end of the current message (or the start of a
6598 new message if it was set in a HELO ACL). The value is now preserved for
6599 the duration of the SMTP connection.
6601 12. If a transport had a headers_rewrite setting, and a matching header line
6602 contained an unqualified address, that address was qualified, even if it
6603 did not match any rewriting rules. The underlying bug was that the values
6604 of the flags that permit the existence of unqualified sender and recipient
6605 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6606 non-local messages, and by -bnq for local messages) were not being
6607 preserved with the message after it was received.
6609 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6610 "next input=" as part of the text comprising the host identity instead of
6611 the correct text. The code was using the same buffer for two different
6612 strings. However, depending on which order the printing function evaluated
6613 its arguments, the bug did not always show up. Under Linux, for example, my
6614 test suite worked just fine.
6616 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6617 doesn't work with some older versions of Perl. It has been changed to "my",
6618 which in any case is probably the better facility to use.
6620 15. A really picky compiler found some instances of statements for creating
6621 error messages that either had too many or two few arguments for the format
6624 16. The size of the buffer for calls to the DNS resolver has been increased
6625 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6626 for addresses that have a lot of PTR records. This alleviates a problem; it
6627 does not fully solve it.
6629 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6630 buffer now truncates the list and logs the incident, which is the same
6631 action as happens when Exim is looking up a host name and its aliases.
6632 Previously in this situation something unpredictable would happen;
6633 sometimes it was "internal error: store_reset failed".
6635 18. If a server dropped the connection unexpectedly when an Exim client was
6636 using GnuTLS and trying to read a response, the client delivery process
6637 crashed while trying to generate an error log message.
6639 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6640 string, for example:
6642 warn message = H1: something\nH2: something
6644 the text was added as a single header line from Exim's point of view
6645 though it ended up OK in the delivered message. However, searching for the
6646 second and subsequent header lines using $h_h2: did not work. This has been
6647 fixed. Similarly, if a system filter added multiple headers in this way,
6648 the routers could not see them.
6650 20. Expanded the error message when iplsearch is called with an invalid key to
6651 suggest using net-iplsearch in a host list.
6653 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6654 ACLs are no longer actually imposed (and a message to that effect is
6657 22. If a "gecos" field in a passwd entry contained escaped characters, in
6658 particular, if it contained a \" sequence, Exim got it wrong when building
6659 a From: or a Sender: header from that name. A second bug also caused
6660 incorrect handling when an unquoted " was present following a character
6661 that needed quoting.
6663 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6664 was not being matched caselessly.
6666 24. Arranged for all hyphens in the exim.8 source to be escaped with
6669 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6670 properly. Recipient callout cache records were still being keyed to include
6671 the sender, even when use_sender was set false. This led to far more
6672 callouts that were necessary. The sender is no longer included in the key
6673 when use_sender is false.
6675 26. Added "control = submission" modifier to ACLs.
6677 27. Added the ${base62d: operator to decode base 62 numbers.
6679 28. dnsdb lookups can now access SRV records.
6681 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6682 the configuration file.
6684 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6685 (-v) mode. This makes the output for a verbose queue run more intelligible.
6687 31. Added a use_postmaster feature to recipient callouts.
6689 32. Added the $body_zerocount variable, containing the number of binary zero
6690 bytes in the message body.
6692 33. The time of last modification of the "new" subdirectory is now used as the
6693 "mailbox time last read" when there is a quota error for a maildir
6696 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6698 35. Added +ignore_unknown as a special item in host lists.
6700 36. Code for decoding IPv6 addresses in host lists is now included, even if
6701 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6702 address was recognized as an IP address, but was then not correctly decoded
6703 into binary, causing unexpected and incorrect effects when compared with
6710 1. Very minor rewording of debugging text in manualroute to say "list of
6711 hosts" instead of "hostlist".
6713 2. If verify=header_syntax was set, and a header line with an unqualified
6714 address (no domain) and a large number of spaces between the end of the
6715 name and the colon was received, the reception process suffered a buffer
6716 overflow, and (when I tested it) crashed. This was caused by some obsolete
6717 code that should have been removed. The fix is to remove it!
6719 3. When running in the test harness, delay a bit after writing a bounce
6720 message to get a bit more predictability in the log output.
6722 4. Added a call to search_tidyup() just before forking a reception process. In
6723 theory, someone could use a lookup in the expansion of smtp_accept_max_
6724 per_host which, without the tidyup, could leave open a database connection.
6726 5. Added the variables $recipient_data and $sender_data which get set from a
6727 lookup success in an ACL "recipients" or "senders" condition, or a router
6728 "senders" option, similar to $domain_data and $local_part_data.
6730 6. Moved the writing of debug_print from before to after the "senders" test
6733 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6734 problems for message scanning, either using a data ACL, or using
6735 local_scan() because the Received: header was not generated till after they
6736 were called (in order to set the time as the time of reception completion).
6737 I have revised the way this works. The header is now generated after the
6738 body is received, but before the ACL or local_scan() are called. After they
6739 are run, the timestamp in the header is updated.
6745 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6746 before starting a queue runner without re-exec. This happened only when
6747 deliver_drop_privilege was set or when the Exim user was set to root. The
6748 effect of the bug was that timeouts during subsequent deliveries caused
6749 crashes instead of being properly handled. The handler is now left at its
6750 default (and expected) setting.
6752 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6753 message, again when deliver_drop_privilege is set or Exim is run as root.
6754 The bug described in (1) was not present in this case, but the tidying up
6755 of the other signals was missing. I have made the two cases consistent.
6757 3. The ignore_target_hosts setting on a manualroute router was being ignored
6758 for hosts that were looked up using the /MX notation.
6760 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6763 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6764 operated on the sender address. After changing the $sender_address to <>
6765 for the sender address verify, Exim was re-instated it as the original
6766 (before rewriting) address, but remembering that it had rewritten it, so it
6767 wasn't rewriting it again. This bug also had the effect of breaking the
6768 sender address verification caching when the sender address was rewritten.
6770 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6771 This has been changed so that if the ip literal address matches
6772 ignore_target_hosts, the router declines.
6774 7. Added expansion conditions match_domain, match_address, and match_local_
6775 part (NOT match_host).
6777 8. The placeholder for the Received: header didn't have a length field set.
6779 9. Added code to Exim itself and to exim_lock to test for a specific race
6780 condition that could lead to file corruption when using MBX delivery. The
6781 issue is with the lockfile that is created in /tmp. If this file is removed
6782 after a process has opened it but before that process has acquired a lock,
6783 there is the potential for a second process to recreate the file and also
6784 acquire a lock. This could lead to two Exim processes writing to the file
6785 at the same time. The added code performs the same test as UW imapd; it
6786 checks after acquiring the lock that its file descriptor still refers to
6787 the same named file.
6789 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6790 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6793 11. Added the smtp_active_hostname option. If used, this will typically be made
6794 to depend on the incoming interface address. Because $interface_address is
6795 not set up until the daemon has forked a reception process, error responses
6796 that can happen earlier (such as "too many connections") no longer contain
6799 12. If an expansion in a condition on a "warn" statement fails because a lookup
6800 defers, the "warn" statement is abandoned, and the next ACL statement is
6801 processed. Previously this caused the whole ACL to be aborted.
6803 13. Added the iplsearch lookup type.
6805 14. Added ident_timeout as a log selector.
6807 15. Added tls_certificate_verified as a log selector.
6809 16. Added a global option tls_require_ciphers (compare the smtp transport
6810 option of the same name). This controls incoming TLS connections.
6812 17. I finally figured out how to make tls_require_ciphers do a similar thing
6813 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6814 before starting the TLS session.
6816 18. Tabs are now shown as \t in -bP output.
6818 19. If the log selector return_path_on_delivery was set, Exim crashed when
6819 bouncing a message because it had too many Received: header lines.
6821 20. If two routers both had headers_remove settings, and the first one included
6822 a superfluous trailing colon, the final name in the first list and the
6823 first name in the second list were incorrectly joined into one item (with a
6824 colon in the middle).
6830 1. Added -C and -D options to the exinext utility, mainly to make it easier
6831 to include in the automated testing, but these could be helpful when
6832 multiple configurations are in use.
6834 2. The exinext utility was not formatting the output nicely when there was
6835 an alternate port involved in the retry record key, nor when there was a
6836 message id as well (for retries that were specific to a specific message
6837 and a specific host). It was also confused by IPv6 addresses, because of
6838 the additional colons they contain. I have fixed the IPv4 problem, and
6839 patched it up to do a reasonable job for IPv6.
6841 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6842 delivery, the log line now contains "pipelined" if PIPELINING was used.
6844 4. An SMTP transport process used to panic and die if the bind() call to set
6845 an explicit outgoing interface failed. This has been changed; it is now
6846 treated in the same way as a connect() failure.
6848 5. A reference to $sender_host_name in the part of a conditional expansion
6849 that was being skipped was still causing a DNS lookup. This no longer
6852 6. The def: expansion condition was not recognizing references to header lines
6853 that used bh_ and bheader_.
6855 7. Added the _cache feature to named lists.
6857 8. The code for checking quota_filecount in the appendfile transport was
6858 allowing one more file than it should have been.
6860 9. For compatibility with Sendmail, the command line option
6868 and sets the incoming protocol and host name (for trusted callers). The
6869 host name and its colon can be omitted when only the protocol is to be set.
6870 Note the Exim already has two private options, -pd and -ps, that refer to
6871 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6872 "s", but I don't think that's a major issue.
6874 10. A number of refactoring changes to the code, none of which should affect
6877 (a) The number of logging options was getting close to filling up the
6878 32-bit word that was used as a bit map. I have split them into two classes:
6879 those that are passed in the argument to log_write(), and those that are
6880 only ever tested independently outside of that function. These are now in
6881 separate 32-bit words, so there is plenty of room for expansion again.
6882 There is no change in the user interface or the logging behaviour.
6884 (b) When building, for example, log lines, the code previously used a
6885 macro that called string_cat() twice, in order to add two strings. This is
6886 not really sufficiently general. Furthermore, there was one instance where
6887 it was actually wrong because one of the argument was used twice, and in
6888 one call a function was used. (As it happened, calling the function twice
6889 did not affect the overall behaviour.) The macro has been replaced by a
6890 function that can join an arbitrary number of extra strings onto a growing
6893 (c) The code for expansion conditions now uses a table and a binary chop
6894 instead of a serial search (which was left over from when there were very
6895 few conditions). Also, it now recognizes conditions like "pam" even when
6896 the relevant support is not compiled in: a suitably worded error message is
6897 given if an attempt is made to use such a condition.
6899 11. Added ${time_interval:xxxxx}.
6901 12. A bug was causing one of the ddress fields not to be passed back correctly
6902 from remote delivery subprocesses. The field in question was not being
6903 subsequently used, so this caused to problems in practice.
6905 13. Added new log selectors queue_time and deliver_time.
6907 14. Might have fixed a bug in maildirsizefile handling that threw up
6908 "unexpected character" debug warnings, and recalculated the data
6909 unnecessarily. In any case, I expanded the warning message to give more
6912 15. Added the message "Restricted characters in address" to the statements in
6913 the default ACL that block characters like @ and % in local parts.
6915 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6916 Three changes have been made:
6918 (a) There was a serious bug; a negative response to MAIL caused the whole
6919 recipient domain to be cached as invalid, thereby blocking all messages
6920 to all local parts at the same domain, from all senders. This bug has
6921 been fixed. The domain is no longer cached after a negative response to
6922 MAIL if the sender used is not empty.
6924 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6927 (c) A new callout option, "use_sender" has been added for people who want
6928 the modified behaviour.
6934 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6937 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6938 indeed breaks things for older releases.
6940 3. Added additional logging to the case where there is a problem reading data
6941 from a filter that is running in a subprocess using a pipe, in order to
6942 try to track down a specific problem.
6944 4. Testing facility fudge: when running in the test harness and attempting
6945 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6946 getting "No route to host". Convert this to a timeout.
6948 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6951 6. Some OS don't have socklen_t but use size_t instead. This affects the
6952 fifth argument of getsockopt() amongst other things. This is now
6953 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6954 can be set for individual OS. I have set it for SunOS5, OSF1, and
6955 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6956 some earlier ones do not.
6958 7. Change 4.30/15 was not doing the test caselessly.
6960 8. The standard form for an IPv6 address literal was being rejected by address
6961 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6962 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6963 this, as well as the form without the "IPv6" on the front (but only when
6964 address literals are enabled, of course).
6966 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6968 10. Exim crashed if a message with an empty sender address specified by -f
6969 encountered a router with an errors_to setting. This could be provoked only
6970 by a command such as
6974 where an empty string was supplied; "<>" did not hit this bug.
6976 11. Installed PCRE release 4.5.
6978 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6979 remained set. It is now erased.
6981 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6982 times from message ids (which are base 36 rather than the normal 62).
6984 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6985 were being counted as actual protocol errors, and logged if the log
6986 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6987 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6988 and DATA following a set of rejected RCPTs do not count as protocol errors.
6989 In other words, Exim assumes they were pipelined, though this may not
6990 actually be the case. Of course, in all cases the client gets an
6991 appropriate error code.
6993 15. If a lookup fails in an ACL condition, a message about the failure may
6994 be available; it is used if testing the ACL cannot continue, because most
6995 such messages specify what the cause of the deferral is. However, some
6996 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6997 that caused an old message to be retained and used if a later statement
6998 caused a defer, replacing the real cause of the deferral.
7000 16. If an IP address had so many PTR records that the DNS lookup buffer
7001 was not large enough to hold them, Exim could crash while trying to process
7002 the truncated data. It now detects and logs this case.
7004 17. Further to 4.21/58, another change has been made: if (and only if) the
7005 first line of a message (the first header line) ends with CRLF, a bare LF
7006 in a subsequent header line has a space inserted after it, so as not to
7007 terminate the header.
7009 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7010 unnecessarily, used atoi() instead of strtol(), and didn't check the
7011 termination when getting file sizes from file names by regex.
7013 19. Completely re-implemented the support for maildirsize files, in the light
7014 of a number of problems with the previous contributed implementation
7015 (4.30/29). In particular:
7017 . If the quota is zero, the maildirsize file is maintained, but no quota is
7020 . If the maildir directory does not exist, it is created before any attempt
7021 to write a maildirsize file.
7023 . The quota value in the file is just a cache; if the quota is changed in
7024 the transport, the new value overrides.
7026 . A regular expression is available for excluding directories from the
7029 20. The autoreply transport checks the characters in options that define the
7030 message's headers; it allows continued headers, but it was checking with
7031 isspace() after an embedded newline instead of explicitly looking for a
7034 21. If all the "regular" hosts to which an address was routed had passed their
7035 expiry times, and had not reached their retry times, the address was
7036 bounced, even if fallback hosts were defined. Now Exim should go on to try
7039 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7040 equivalent code in the SMTP transport. Some hosts send humungous responses
7041 to HELO/EHLO, more than 1024 it seems.
7043 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7044 gives compiler warnings in some environments. I've now done it "properly",
7047 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7048 (because of problems with the built-in one) was declared to return uschar *
7049 instead of char *, causing compiler failure.
7051 25. Fixed a file descriptor leak when processing alias/forward files.
7053 26. Fixed a minor format string issue in dbfn.c.
7055 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7057 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7058 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7061 29. When a sender address is verified, it is cached, to save repeating the test
7062 when there is more than one recipient in a message. However, when the
7063 verification involves a callout, it is possible for different callout
7064 options to be set for different recipients. It is too complicated to keep
7065 track of this in the cache, so now Exim always runs a verification when a
7066 callout is required, relying on the callout cache for the optimization.
7067 The overhead is duplication of the address routing, but this should not be
7070 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7071 to be verified with callout=postmaster, and the main callout worked but the
7072 postmaster check failed, the verification correctly failed. However, if a
7073 subsequent RCPT command asked for sender verification *without* the
7074 postmaster check, incorrect caching caused this verification also to fail,
7077 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7078 it was not caching the DNS options (qualify_single, search_parents) that
7079 were used when the lookup failed. A subsequent lookup with different
7080 options therefore always gave the same answer, though there were cases
7081 where it should not have. (Example: a "domains = !$mx_any" option on a
7082 dnslookup router: the "domains" option is always processed without any
7083 widening, but the router might have qualify_single set.) Now Exim uses the
7084 cached value only when the same options are set.
7086 32. Added John Jetmore's "exipick" utility to the distribution.
7088 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7089 than a timeout (e.g. a certificate is required, and is not provided), an
7090 Exim server now closes the connection immediately. Previously it waited for
7091 the client to close - but if the client is SSL, it seems that they each
7092 wait for each other, leading to a delay before one of them times out.
7094 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7095 maintained 0.8.x compatibility because I don't think many are using it, and
7096 it is clearly obsolete.
7098 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7101 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7102 client certificate was expired. A simple patch fixes this, though I don't
7103 understand the full logic of why the verify callback is called multiple
7106 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7107 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7108 which causes problems with some clients (such as the Certicom SSL Plus
7109 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7110 disables the coutermeasure allowing Eudora to connect."
7112 38. Exim was not checking that a write() to a log file succeeded. This could
7113 lead to Bad Things if a log got too big, in particular if it hit a file
7114 size limit. Exim now panics and dies if it cannot write to a log file, just
7115 as it does if it cannot open a log file.
7117 39. Modified OS/Makefile-Linux so that it now contains
7119 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7121 The two -D definitions ensure that Exim is compiled with large file
7122 support, which makes it possible to handle log files that are bigger than
7125 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7126 instance) a domain was checked against a named list that involved a lookup,
7127 causing $domain_data to be set, then another domain was checked against the
7128 same list, then the first domain was re-checked, the value of $domain_data
7129 after the final check could be wrong. In particular, if the second check
7130 failed, it could be set empty. This bug probably also applied to
7133 41. The strip_trailing_dot option was not being applied to the address given
7134 with the -f command-line option.
7136 42. The code for reading a message's header from the spool was incrementing
7137 $received_count, but never initializing it. This meant that the value was
7138 incorrect (doubled) while delivering a message in the same process in which
7139 it was received. In the most common configuration of Exim, this never
7140 happens - a fresh exec is done - but it can happen when
7141 deliver_drop_privilege is set.
7143 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7144 it now includes up to 150 characters of the unexpected data in the log
7147 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7148 and building data strings. The size of both of these buffers was 10 000
7149 bytes - far larger than anybody would *ever* want, thought I. Needless to
7150 say, somebody hit the limit. I have increased the maximum line length to
7151 20 000 and the maximum data length of concatenated lines to 100 000. I have
7152 also fixed two bugs, because there was no checking on these buffers. Tsk,
7153 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7154 buffer is too small.
7156 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7157 lsearch lookups. Now it does.
7159 46. When parsing a route_list item in a manualroute router, a fixed-length
7160 buffer was used for the list of hosts. I made this 1024 bytes long,
7161 thinking that nobody would ever have a list of hosts that long. Wrong.
7162 Somebody had a whole pile of complicated expansion conditions, and the
7163 string was silently truncated, leading to an expansion error. It turns out
7164 that it is easier to change to an unlimited length (owing to other changes
7165 that have happened since this code was originally written) than to build
7166 structure for giving a limitation error. The length of the item that
7167 expands into the list of hosts is now unlimited.
7169 47. The lsearch lookup could not handle data where the length of text line was
7170 more than 4095 characters. Such lines were truncated, leading to shortened
7171 data being returned. It should now handle lines of any length.
7173 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7174 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7177 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7178 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7179 no longer happens in comment lines. A list of replacements is now placed
7180 at the head of all of the source files, except those whose only change is
7181 to replace PERL_COMMAND in the very first #! line.
7183 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7184 messages on the queue, with a bottom-up merge sort, using code contributed
7185 by Michael Haardt. This should make operations like -bp somewhat faster on
7186 large queues. It won't affect queue runners, except when queue_run_in_order
7189 51. Installed eximstats 1.31 in the distribution.
7191 52. Added support for SRV lookups to the dnslookup router.
7193 53. If an ACL referred to $message_body or $message_body_end, the value was not
7194 reset for any messages that followed in the same SMTP session.
7196 54. The store-handling optimization for building very long strings was not
7197 differentiating between the different store pools. I don't think this
7198 actually made any difference in practice, but I've tidied it.
7200 55. While running the routers to verify a sender address, $sender_address
7201 was still set to the sender address. This is wrong, because when routing to
7202 send a bounce to the sender, it would be empty. Therefore, I have changed
7203 it so that, while verifying a sender address, $sender_address is set to <>.
7204 (There is no change to what happens when verifying a recipient address.)
7206 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7207 target A or AAAA records (if not already returned) without resetting the
7208 qualify_single or search_parents options of the DNS resolver. These are
7209 inappropriate in this case because the targets of MX and SRV records must
7210 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7211 target that, when qualified, matched something in the local domain. These
7212 two options are now turned off when doing these lookups.
7214 57. It seems that at least some releases of Reiserfs (which does not have the
7215 concept of a fixed number of inodes) returns zero and not -1 for the
7216 number of available inodes. This interacted badly with check_spool_inodes,
7217 which assumed that -1 was the "no such thing" setting. What I have done is
7218 to check that the total number of inodes is greater than zero before doing
7219 the test of how many are available.
7221 58. When a "warn" ACL statement has a log_message modifier, the message is
7222 remembered, and not repeated. This is to avoid a lot of repetition when a
7223 message has many recipients that cause the same warning to be written.
7224 However, Exim was preserving the list of already written lines for an
7225 entire SMTP session, which doesn't seem right. The memory is now reset if a
7226 new message is started.
7228 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7229 debugging output unless log_rewrite was also set.
7231 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7232 of (int)(handle) when we know that handle contains (void *)(-1).
7234 61. The Exim daemon panic-logs an error return when it closes the incoming
7235 connection. However "connection reset by peer" seems to be common, and
7236 isn't really an error worthy of noting specially, so that particular error
7239 62. When Exim is trying to find all the local interfaces, it used to panic and
7240 die if the ioctl to get the interface flags failed. However, it seems that
7241 on at least one OS (Solaris 9) it is possible to have an interface that is
7242 included in the list of interfaces, but for which you get a failure error
7243 for this call. This happens when the interface is not "plumbed" into a
7244 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7245 failure of the "get flags" call assumes that the interface is down.
7247 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7248 makes life easier for people who are doing arithmetic on fields extracted
7249 from dates, where you often get leading zeros that should not be
7250 interpreted as octal.
7252 64. Added qualify_domain to the redirect router, to override the global
7255 65. If a pathologically long header line contained very many addresses (the
7256 report of this problem mentioned 10 000) and each of them was rewritten,
7257 Exim could use up a very large amount of memory. (It kept on making new
7258 copies of the header line as it rewrote, and never released the old ones.)
7259 At the expense of a bit more processing, the header rewriting function has
7260 been changed so that it no longer eats memory in this way.
7262 66. The generation of the Received: header has been moved from the time that a
7263 message starts to be received, to the time that it finishes. The timestamp
7264 in the Received: header should now be very close to that of the <= log
7265 line. There are two side-effects of this change:
7267 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7268 logged header lines no longer include the local Received: line, because
7269 it has not yet been created. The same applies to a copy of the message
7270 that is returned to a non-SMTP sender when a message is rejected.
7272 (b) When a filter file is tested using -bf, no additional Received: header
7273 is added to the test message. After some thought, I decided that this
7276 This change does not affect the value of $received_for. It is still set
7277 after address rewriting, but before local_scan() is called.
7279 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7281 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7282 gave an unhelpful panic error message, and a defer error. I have managed to
7283 change this behaviour so that it now rejects any supplied certificate,
7284 which seems right, as the list of acceptable certificates is empty.
7286 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7287 gave an unhelpful defer error. I have not managed to make this reject any
7288 supplied certificates, but the error message it gives is "no certificate
7289 supplied", which is not helpful.
7291 70. exigrep's output now also includes lines that are not associated with any
7292 message, but which match the given pattern. Implemented by a patch from
7293 Martin Sluka, which also tidied up the Perl a bit.
7295 71. Recipient callout verification, like sender verification, was using <> in
7296 the MAIL FROM command. This isn't really the right thing, since the actual
7297 sender may affect whether the remote host accepts the recipient or not. I
7298 have changed it to use the actual sender in the callout; this means that
7299 the cache record is now keyed on a recipient/sender pair, not just the
7300 recipient address. There doesn't seem to be a real danger of callout loops,
7301 since a callout by the remote host to check the sender would use <>.
7302 [SEE ABOVE: changed after hitting problems.]
7304 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7305 temporary errors. However, in the case of such a code being given after
7306 the end of a data transmission (i.e. after ".") Exim was failing to write
7307 a retry record for the message. (Yes, there was some broken host that was
7308 actually sending 8xx at this point.)
7310 73. An unknown lookup type in a host list could cause Exim to panic-die when
7311 the list was checked. (An example that provoked this was putting <; in the
7312 middle of a list instead of at the start.) If this happened during a DATA
7313 ACL check, a -D file could be left lying around. This kind of configuration
7314 error no longer causes Exim to die; instead it causes a defer error. The
7315 incident is still logged to the main and panic logs.
7317 74. Buglet left over from Exim 3 conversion. The message "too many messages
7318 in one connection" was written to the rejectlog but not the mainlog, except
7319 when address rewriting (yes!) was being logged.
7321 75. Added write_rejectlog option.
7323 76. When a system filter was run not as root (that is, when system_filter_user
7324 was set), the values of the $n variables were not being returned to the
7325 main process; thus, they were not subsequently available in the $sn
7328 77. Added +return_path_on_delivery log selector.
7330 78. A connection timeout was being treated differently from recipients deferred
7331 when testing hosts_max_try with a message that was older than the host's
7332 retry timeout. (The host should not be counted, thus allowing all hosts to
7333 be tried at least once before bouncing.) This may have been the cause of an
7334 occasionally reported bug whereby a message would remain on the queue
7335 longer than the retry timeout, but would be bounced if a delivery was
7336 forced. I say "may" because I never totally pinned down the problem;
7337 setting up timeout/retry tests is difficult. See also the next item.
7339 79. The ultimate address timeout was not being applied to errors that involved
7340 a combination of host plus message (for example, a timeout on a MAIL
7341 command). When an address resolved to a number of possible hosts, and they
7342 were not all tried for each delivery (e.g. because of hosts_max_try), a
7343 message could remain on the queue longer than the retry timeout.
7345 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7348 81. Fixed an obscure SMTP outgoing bug which required at least the following
7349 conditions: (a) there was another message waiting for the same server;
7350 (b) the server returned 5xx to all RCPT commands in the first message so
7351 that the message was not completed; (c) the server dropped the connection
7352 or gave a negative response to the RSET that Exim sends to abort the
7353 transaction. The observed case was a dropped connection after DATA that had
7354 been sent in pipelining mode. That is, the server had advertised PIPELINING
7355 but was not implementing it correctly. The effect of the bug was incorrect
7356 behaviour, such as trying another host, and this could lead to a crash.
7362 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7363 and daemon.c were passed as pointers to ints; they should have been
7364 pointers to socklen_t variables (which are typically unsigned ints).
7366 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7369 3. Fixed a really odd bug that affected only the testing scheme; patching a
7370 certain fixed string in the binary changed the value of another string that
7371 happened to be identical to the end of the original first string.
7373 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7374 name", it returns that address as the IP address. On some operating
7375 systems (e.g. Solaris), it also passes back the IP address string as the
7376 "host name". However, on others (e.g. Linux), it passes back an empty
7377 string. Exim wasn't checking for this, and was changing the host name to an
7378 empty string, assuming it had been canonicalized.
7380 5. Although rare, it is permitted to have more than one PTR record for a given
7381 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7382 all the names associated with an address, because they do in Solaris.
7383 However, it seems that they do not in Linux for data that comes from the
7384 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7385 I found this out when I moved to a new Linux workstation and tried to run
7386 the Exim test suite.
7388 To get round this problem I have changed the code so that it now does its
7389 own call to the DNS to look up PTR records when searching for a host name.
7390 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7391 addresses that are only in /etc/hosts are still found.
7393 This behaviour is, however, controlled by an option called host_lookup_
7394 order, which defaults to "bydns:byaddr". If people want to use the other
7395 order, or indeed, just use one or the other means of lookup, they can
7396 specify it in this variable.
7398 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7399 some operating systems, this comes back from gethostbyaddr() as an empty
7400 string, and this is what Exim used to test for. However, it seems that in
7401 other systems, "." is yielded. Exim now tests for this case too.
7403 7. The values of check_spool_space and check_log_space are now held internally
7404 as a number of kilobytes instead of an absolute number of bytes. If a
7405 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7406 kilobyte. This means that much larger values can be stored.
7408 8. Exim monitor: an attempt to get the action menu when not actually pointing
7409 at a message produces an empty menu entitled "No message selected". This
7410 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7411 no entries in it ("Shell widget menu has zero width and/or height"). So I
7412 have added a single, blank menu entry in this case.
7414 9. Added ${quote_local_part.
7416 10. MIME decoding is now applied to the contents of Subject: header lines when
7419 11. Now that a reference to $sender_host_address automatically causes a reverse
7420 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7421 host lookup before query-style lookups in lists that might use this
7422 variable. This has therefore been abolished, and the "net-" prefix is no
7423 longer necessary for query-style lookups.
7425 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7426 have been a typo for LFLAGS, so it has been changed.
7428 13. The install script calls Exim with "-C /dev/null" in order to find the
7429 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7430 to be output. However, since Exim outputs its version number before the
7431 error, it didn't break the script. It just looked ugly. I fixed this by
7432 always allowing "-C /dev/null" if the caller is root.
7434 14. Ignore overlarge ACL variable number when reading spool file - insurance
7435 against a later release with more variables having written the file.
7437 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7438 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7439 this, as well as the form without the "IPv6" on the front.
7441 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7442 OS/Makefile-Darwin file.
7444 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7445 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7447 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7448 variable. However, some OS use ints for inodes. I've added cast to long int
7449 to get rid of the compiler warning.
7451 19. I had forgotten to lock out "/../" in configuration file names when
7452 ALT_CONFIG_PREFIX was set.
7454 20. Routers used for verification do not need to specify transports. However,
7455 if such a router generated a host list, and callout was configured, Exim
7456 crashed, because it could not find a port number from the (non-existent)
7457 transport. It now assumes port 25 in this circumstance.
7459 21. Added the -t option to exigrep.
7461 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7462 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7463 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7464 them is set, LOOKUP_LSEARCH is forced.
7466 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7468 24. Added sender and host information to the "rejected by local_scan()" log
7469 line; previously there was no indication of these.
7471 25. Added .include_if_exists.
7473 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7474 Linux. It turns out that not all file systems support this. Apparently some
7475 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7476 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7477 not supported on the file descriptor, is now ignored when Exim is trying to
7478 sync a directory. This applies only to Linux.
7480 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7482 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7483 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7486 29. Added support for maildirsize files from supplied patch (modified a bit).
7488 30. The use of :fail: followed by an empty string could lead Exim to respond to
7489 sender verification failures with (e.g.):
7491 550 Verification failed for <xxx>
7492 550 Sender verify failed
7494 where the first response line was missing the '-' that indicates it is not
7495 the final line of the response.
7497 31. The loop for finding the name of the user that called Exim had a hardwired
7498 limit of 10; it now uses the value of finduser_retries, which is used for
7499 all other user lookups.
7501 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7504 33. Exim was neglecting to zero errno before one call of strtol() when
7505 expanding a string and expecting an integer value. On some systems this
7506 resulted in spurious "integer overflow" errors. Also, it was casting the
7507 result into an int without checking.
7509 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7510 did not work. The code looks as if it has *never* worked, though it appears
7511 to have been documented since at least release 1.62. I have made it work.
7513 35. The "timeout_DNS" error in retry rules, also documented since at least
7514 1.62, also never worked. As it isn't clear exactly what this means, and
7515 clearly it isn't a major issue, I have abolished the feature by treating it
7516 as "timeout", and writing a warning to the main and panic logs.
7518 36. The display of retry rules for -brt wasn't always showing the error code
7521 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7522 timeout_connect_A, timeout_connect_MX.
7524 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7525 to the empty sender.
7527 39. The daemon was not analysing the content of -oX till after it had closed
7528 stderr and disconnected from the controlling terminal. This meant that any
7529 syntax errors were only noted on the panic log, and the return code from
7530 the command was 0. By re-arranging the code a little, I've made the
7531 decoding happen first, so such errors now appear on stderr, and the return
7532 code is 1. However, the actual setting up of the sockets still happens in
7533 the disconnected process, so errors there are still only recorded on the
7536 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7537 connections (as happens on some IP stacks) was logged at start up time as
7538 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7539 it from "IPv6 and IPv4", which means that two separate sockets are being
7542 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7543 says whether AF_INET or AF_INET6 was passed as an argument.
7545 42. Exiwhat output was messed up when time zones were included in log
7548 43. Exiwhat now gives more information about the daemon's listening ports,
7549 and whether -tls-on-connect was used.
7551 44. The "port" option of the smtp transport is now expanded.
7553 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7554 silently ignored. Now an error message is written to the main and panic
7557 46. There's a new ACL modifier called "logwrite" which writes to a log file
7558 as soon as it is encountered.
7560 47. Added $local_user_uid and $local_user_gid at routing time.
7562 48. Exim crashed when trying to verify a sender address that was being
7565 49. Exim was recognizing only a space character after ".include". It now also
7566 recognizes a tab character.
7568 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7569 extracting the relevant information from the specification. The man page no
7570 longer contains scrambled data for the -d option, and I've added a section
7571 at the front about calling Exim under different names.
7573 51. Added "extra_headers" argument to the "mail" command in filter files.
7575 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7578 53. Installed eximstats 1.29.
7580 54. Added transport_filter_timeout as a generic transport option.
7582 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7583 Cc: header lines. This was required by RFC 822, but it not required by RFC
7586 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7587 incoming messages that did not have them. Now it does so only if the
7588 message originates locally, that is, if there is no associated remote host
7589 address. When Resent- header lines are present, this applies to the Resent-
7590 lines rather than the non-Resent- lines.
7592 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7593 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7595 58. Messages for configuration errors now include the name of the main
7596 configuration file - useful now that there may be more than one file in a
7597 list (.included file names were always shown).
7599 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7600 for those rare installations that do not start the daemon as root or run it
7601 setuid root. I've cut out the call to initgroups() if the daemon is not
7604 60. The Exim user and group can now be bound into the binary as text strings
7605 that are looked up at the start of Exim's processing.
7607 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7609 62. Added $mailstore_basename variable.
7611 63. Installed patch to sieve.c from Michael Haardt.
7613 64. When Exim failed to open the panic log after failing to open the main log,
7614 the original message it was trying to log was written to stderr and debug
7615 output, but if they were not available (the usual case in production), it
7616 was lost. Now it is written to syslog before the two lines that record the
7617 failures to open the logs.
7619 65. Users' Exim filters run in subprocesses under the user's uid. It is
7620 possible for a "deliver" command or an alias in a "personal" command to
7621 provoke an address rewrite. If logging of address rewriting is configured,
7622 this fails because the process is not running as root or exim. There may be
7623 a better way of dealing with this, but for the moment (because 4.30 needs
7624 to be released), I have disabled address rewrite logging when running a
7625 filter in a non-root, non-exim process.
7631 1. The buildconfig auxiliary program wasn't quoting the value set for
7632 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7633 not defined. This bug was masked in 4.22 by the effect that was fixed in
7636 2. Some messages that were rejected after a message id was allocated were
7637 shown as "incomplete" by exigrep. It no longer does this for messages that
7638 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7640 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7641 have allow_domain_literals set, the ID did not get logged in the <= line.
7642 Domain literals are now always recognized in Message-ID: header lines.
7644 4. The first argument for a ${extract expansion item is the key name or field
7645 number. Leading and trailing spaces in this item were not being ignored,
7646 causing some misleading effects.
7648 5. When deliver_drop_privilege was set, single queue runner processes started
7649 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7650 same command in the process it spins off) were not dropping privilege.
7652 6. When the daemon running as "exim" started a queue runner, it always
7653 re-executed Exim in the spun-off process. This is a waste of effort when
7654 deliver_drop_privilege is set. The new process now just calls the
7655 queue-runner function directly.
7661 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7664 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7665 ignored. Though the use of -oP was forcing the writing of a pid file, it
7666 was always written to the default place.
7668 3. If the message "no IP address found for host xxxx" is generated during
7669 incoming verification, it is now followed by identification of the incoming
7670 connection (so you can more easily find what provoked it).
7672 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7674 5. Added some features to "harden" Exim a bit more against certain attacks:
7676 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7677 be put in Local/Makefile. This is like the never_users runtime option,
7678 but it cannot be overridden. The default setting is "root".
7680 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7681 prefix string with which any file named in a -C command line option
7684 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7685 is retained for -C and -D only if the caller of Exim is root. Without
7686 it, the exim user may also use -C and -D and retain privilege.
7688 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7689 command line option is disabled.
7691 6. Macro names set by the -D option must start with an upper case letter, just
7692 like macro names defined in the configuration file.
7694 7. Added "dereference=" facility to LDAP.
7696 8. Two instances of the typo "uknown" in the source files are fixed.
7698 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7699 the Configure-Makefile script screwed up while processing it.
7701 10. Incorporated PCRE 4.4.
7703 11. The SMTP synchronization check was not operating right at the start of an
7704 SMTP session. For example, it could not catch a HELO sent before the client
7705 waited for the greeting. There is now a check for outstanding input at the
7706 point when the greeting is written. Because of the duplex, asynchronous
7707 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7708 way, but not yet received, when the check is performed.
7710 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7711 on TCP/IP sockets, because this apparently causes some broken clients to
7714 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7715 unchanged) from the Cygwin maintainer.
7717 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7718 was supported instead of testing for mbx. Effectively a typo.
7720 15. The spa authenticator server code was not checking that the input it
7721 received was valid base64.
7723 16. The debug output line for the "set" modifier in ACLs was not showing the
7724 name of the variable that was being set.
7726 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7728 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7729 lookup on its own. Something else was needed to trigger the lookup. For
7730 example, a match in host_lookup or the need for a host name in a host list.
7731 Now, if $sender_host_name is referenced and the host name has not yet been
7732 looked up, a lookup is performed. If the lookup fails, the variable remains
7733 empty, and $host_lookup_failed is set to "1".
7735 19. Added "eqi" as a case-independent comparison operator.
7737 20. The saslauthd authentication condition could segfault if neither service
7738 nor realm was specified.
7740 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7741 error message that was logged was misleading, and incoming SMTP
7742 connections were dropped. The message is now more accurate, and temporary
7743 errors are given to SMTP connections.
7745 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7746 (except RSET) with a 421 error, until QUIT is received. However, it was
7747 failing to send a response to QUIT.
7749 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7750 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7751 ACL is now run after the helo_try_verify_hosts code.
7753 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7754 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7755 case-independent, so other case variants are also recognized). Apparently
7756 some systems use these upper case variants.
7758 25. If more than two messages were waiting for the same host, and a transport
7759 filter was specified for the transport, Exim sent two messages over the
7760 same TCP/IP connection, and then failed with "socket operation on non-
7761 socket" when it tried to send the third.
7763 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7765 27. The extern definition of crypt16() in expand.c was not being excluded when
7766 the OS had its own crypt16() function.
7768 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7769 as a preferred synonym for return_size_limit, both as an option and as an
7772 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7774 30. Changed the default configuration ACL to relax the local part checking rule
7775 for addresses that are not in any local domains. For these addresses,
7776 slashes and pipe symbols are allowed within local parts, but the sequence
7777 /../ is explicitly forbidden.
7779 31. SPA server authentication was not clearing the challenge buffer before
7782 32. log_message in a "warn" ACL statement was writing to the reject log as
7783 well as to the main log, which contradicts the documentation and doesn't
7784 seem right (because no rejection is happening). So I have stopped it.
7786 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7787 However, I am unable to do any testing of this.
7789 34. Fixed an infelicity in the appendfile transport. When checking directories
7790 for a mailbox, to see if any needed to be created, it was accidentally
7791 using path names with one or more superfluous leading slashes; tracing
7792 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7794 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7795 contents are added to the log line that is written for every discarded
7796 recipient. (Previously a log_message setting was ignored.)
7798 36. The ${quote: operator now quotes the string if it is empty.
7800 37. The install script runs exim in order to find its version number. If for
7801 some reason other than non-existence or emptiness, which it checks, it
7802 could not run './exim', it was installing it with an empty version number,
7803 i.e. as "exim-". This error state is now caught, and the installation is
7806 38. An argument was missing from the function that creates an error message
7807 when Exim fails to connect to the socket for saslauthd authentication.
7808 This could cause Exim to crash, or give a corrupted message.
7810 39. Added isip, isip4, and isip6 to ${if conditions.
7812 40. The ACL variables $acl_xx are now saved with the message, and can be
7813 accessed later in routers, transports, and filters.
7815 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7816 strings in the file are not string-expanded.
7818 42. If a MAIL command specified a SIZE value that was too large to fit into an
7819 int variable, the check against message_size_limit failed. Such values are
7820 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7821 day this will have to be increased, but I don't think I want to be around
7822 when emails are that large.
7829 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7830 iconv() is not standard in FreeBSD.
7832 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7833 IPv6 enabled. The observed symptom was a segmentation fault on return from
7834 the function os_common_find_running_interfaces() in src/os.c.
7836 3. In the check_special_case() function in daemon.c I had used "errno" as an
7837 argument name, which causes warnings on some systems. This was basically a
7838 typo, since it was named "eno" in the comments!
7840 4. The code that waits for the clock to tick (at a resolution of some fraction
7841 of a second) so as to ensure message-id uniqueness was always waiting for
7842 at least one whole tick, when it could have waited for less. [This is
7843 almost certainly not relevant at current processor speeds, where it is
7844 unlikely to ever wait at all. But we try to future-proof.]
7846 5. The function that sleeps for a time interval that includes fractions of a
7847 second contained a race. It did not block SIGALRM between setting the
7848 timer, and suspending (a couple of lines later). If the interval was short
7849 and the sigsuspend() was delayed until after it had expired, the suspension
7850 never ended. On busy systems this could lead to processes getting stuck for
7853 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7854 process, before it forks any delivery processes. The open lookup caching
7855 mechanism meant that the open file or database connection was passed into
7856 the delivery process. The problem was that delivery processes always tidy
7857 up cached lookup data. This could cause a problem for the next delivery
7858 process started by the queue runner, because the external queue runner
7859 process does not know about the closure. So the next delivery process
7860 still has data in the lookup cache. In the case of a file lookup, there was
7861 no problem because closing a file descriptor in a subprocess doesn't affect
7862 the parent. However, if the lookup was caching a connection to a database,
7863 the connection was closed, and the second delivery process was likely to
7864 see errors such as "PGSQL: query failed: server closed the connection
7865 unexpectedly". The problem has been fixed by closing all cached lookups
7866 in a queue runner before running a delivery process.
7868 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7869 seem to have the "const" qualifier which it has on other OS. I've
7872 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7873 *of the same type* (client or server) with the same public name that an
7874 error should be diagnosed.
7876 9. When Exim looked up a host name for an IP address, but failed to find the
7877 original IP address when looking up the host name (a safety check), it
7878 output the message "<ip address> does not match any IP for NULL", which was
7879 confusing, to say the least. The bug was that the host name should have
7880 appeared instead of "NULL".
7882 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7883 user that is built into the binary, and the -C or -D options is used, root
7884 privilege is dropped before the configuration file is read. In addition,
7885 logging is switched to stderr instead of the normal log files. If the
7886 configuration then re-defines the Exim user, the unprivileged environment
7887 is probably not what is expected, so Exim logs a panic warning message (but
7890 However, if deliver_drop_privilege is set, the unprivileged state may well
7891 be exactly what is intended, so the warning has been cut out in that case,
7892 and Exim is allowed to try to write to its normal log files.
7898 1. smtp_return_error_details was not giving details for temporary sender
7899 or receiver verification errors.
7901 2. Diagnose a configuration error if two authenticators have the same public
7904 3. Exim used not to create the message log file for a message until the first
7905 delivery attempt. This could be confusing when incoming messages were held
7906 for policy or load reasons. The message log file is now created at the time
7907 the message is received, and an initial "Received" line is written to it.
7909 4. The automatically generated man page for command line options had a minor
7910 bug that caused no ill effects; however, a more serious problem was that
7911 the procedure for building the man page automatically didn't always
7912 operate. Consequently, release 4.20 contains an out-of-date version. This
7913 shouldn't happen again.
7915 5. When building Exim with embedded Perl support, the script that builds the
7916 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7917 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7919 6. The freeze_tell option was not being used for messages that were frozen on
7920 arrival, either by an ACL or by local_scan().
7922 7. Added the smtp_incomplete_transaction log selector.
7924 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7925 was accepting AUTH without a new EHLO.
7927 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7928 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7931 10. Logging of TCP/IP connections (when configured) now happens in the main
7932 daemon process instead of the child process, so that the TCP/IP connection
7933 count is more accurate (but it can never be perfect).
7935 11. The use of "drop" in a nested ACL was not being handled correctly in the
7936 outer ACL. Now, if condition failure induced by the nested "drop" causes
7937 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7938 or "require"), the connection is dropped.
7940 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7941 that yield "discard" can now be used with an "accept" or a "discard" verb,
7942 but an error is generated for any others (because I can't see a useful way
7943 to define what should happen).
7945 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7946 are now processed in the same way as lines in the Exim configuration file.
7947 In particular, continuation lines are supported.
7949 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7951 15. Added -ti meaning -t -i.
7953 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7954 domains, and warn by logging if others are found.
7956 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7957 returned from GIFCONF when Exim is trying to find the list of interfaces on
7958 a host. The code in os.c has been modified to copy each ifreq to an aligned
7959 structure in all cases.
7961 Also, in some cases, the returned ifreq's were being copied to a 'struct
7962 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7963 means the last couple of bytes of an IPv6 address could be chopped if the
7964 ifreq contained only a normal sockaddr (14 bytes storage).
7966 18. Named domain lists were not supported in the hosts_treat_as_local option.
7967 An entry such as +xxxx was not recognized, and was treated as a literal
7970 19. Ensure that header lines added by a DATA ACL are included in the reject log
7971 if the ACL subsequently rejects the message.
7973 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7974 MD5 (which is deprecated).
7976 21. When testing a filter file using -bf, Exim was writing a message when it
7977 took the sender from a "From " line in the message, but it was not doing so
7978 when it took $return_path from a Return-Path: header line. It now does.
7980 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7981 with a valid header line field name (a series of printing characters
7982 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7984 23. Changed "disc" in the source to "disk" to conform to the documentation and
7985 the book and for uniformity.
7987 24. Ignore Sendmail's -Ooption=value command line item.
7989 25. When execve() failed while trying to run a command in a pipe transport,
7990 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7991 could be confused with a return value of 69 from the command itself. This
7992 has been changed to 127, the value the shell returns if it is asked to run
7993 a non-existent command. The wording for the related log line suggests a
7994 non-existent command as the problem.
7996 26. If received_header_text expands to an empty string, do not add a Received:
7997 header line to the message. (Well, it adds a token one on the spool, but
7998 marks it "old" so that it doesn't get used or transmitted.)
8000 27. Installed eximstats 1.28 (addition of -nt option).
8002 28. There was no check for failure on the call to getsockname() in the daemon
8003 code. This can fail if there is a shortage of resources on the system, with
8004 ENOMEM, for example. A temporary error is now given on failure.
8006 29. Contrary to the C standard, it seems that in some environments, the
8007 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8008 program. Exim now does this explicitly; it affects the formatting of
8009 timestamps using strftime().
8011 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8012 complaints. I've now initialized all the variables, to avoid this.
8014 32. Header lines added by a system filter were not being "seen" during
8015 transport-time rewrites.
8017 33. The info_callback() function passed to OpenSSL is set up with type void
8018 (*)(SSL *, int, int), as described somewhere. However, when calling the
8019 function (actually a macro) that sets it up, the type void(*)() is
8020 expected. I've put in a cast to prevent warnings from picky compilers.
8022 34. If a DNS black list lookup found a CNAME record, but there were no A
8023 records associated with the domain it pointed at, Exim crashed.
8025 35. If a DNS black list lookup returned more than one A record, Exim ignored
8026 all but the first. It now scans all returned addresses if a particular IP
8027 value is being sought. In this situation, the contents of the
8028 $dnslist_value variable are a list of all the addresses, separated by a
8031 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8032 to accept a host name and all its aliases if the forward lookup for any of
8033 them yielded the IP address of the incoming connection. Now it accepts only
8034 those names whose forward lookup yields the correct IP address. Any other
8035 names are discarded. This closes a loophole whereby a rogue DNS
8036 administrator could create reverse DNS records to break through a
8037 wildcarded host restriction in an ACL.
8039 37. If a user filter or a system filter that ran in a subprocess used any of
8040 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8041 the wrong values were passed to the pipe command ($thisaddress had the
8042 value of $0, $0 had the value of $1, etc). This bug was introduced by
8043 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8045 38. Improved the line breaking for long SMTP error messages from ACLs.
8046 Previously, if there was no break point between 40 and 75 characters, Exim
8047 left the rest of the message alone. Two changes have been made: (a) I've
8048 reduced the minimum length to 35 characters; (b) if it can't find a break
8049 point between 35 and 75 characters, it looks ahead and uses the first one
8050 that it finds. This may give the occasional overlong line, but at least the
8051 remaining text gets split now.
8053 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8054 file descriptors might be low, and that setting 1000 would always raise it.
8055 It turns out that in some environments, the limit is already over 1000 and
8056 that lowering it causes trouble. So now Exim takes care not to decrease it.
8058 40. When delivering a message, the value of $return_path is set to $sender_
8059 address at the start of routing (routers may change the value). By an
8060 oversight, this default was not being set up when an address was tested by
8061 -bt or -bv, which affected the outcome if any router or filter referred to
8064 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8065 log only when "message" or "log_message" are set. However, if one of the
8066 conditions was an address verification, or a call to a nested ACL, the
8067 messages generated by the underlying test were being passed through. This
8068 no longer happens. The underlying message is available in $acl_verify_
8069 message for both "message" and "log_message" expansions, so it can be
8070 passed through if needed.
8072 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8073 new expansion $bh_ to give the encoded byte string without charset
8074 translation. Translation happens only if iconv() is available; HAVE_ICONV
8075 indicates this at build time. HEADERS_CHARSET gives the charset to
8076 translate to; headers_charset can change it in the configuration, and
8077 "headers charset" can change it in an individual filter file.
8079 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8080 that creates RFC 2047 encoded "words" labels them as that charset instead
8081 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8082 expansion operator; (ii) when Exim creates a From: line for a local
8083 message; (iii) when a header line is rewritten to include a "phrase" part.
8085 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8086 buggy, causing it to skip the first lines of messages whose message ID
8087 ended in 'D'. This would not have bitten before Exim release 4.14, because
8088 message IDs were unlikely to end in 'D' before then. The effect was to have
8089 incorrect size information for certain domains.
8091 45. #include "config.h" was missing at the start of the crypt16.c module. This
8092 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8095 46. If there was a timeout during a "random" callout check, Exim treated it as
8096 a failure of the random address, and carried on sending RSET and the real
8097 address. If the delay was just some slowness somewhere, the response to the
8098 original RCPT would be taken as a response to RSET and so on, causing
8099 mayhem of various kinds.
8101 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8102 when I implemented it. It didn't allow for the fact that some option values
8103 may legitimately be negative (e.g. size_addition), and it didn't even do
8104 the right test for positive values.
8106 48. Domain names in DNS records are case-independent. Exim always looks them up
8107 in lower case. Some resolvers return domain names in exactly the case they
8108 appear in the zone file, that is, they may contain uppercase letters. Not
8109 all resolvers do this - some return always lower case. Exim was treating a
8110 change of case by a resolver as a change of domain, similar to a widening
8111 of a domain abbreviation. This triggered its re-routing code and so it was
8112 trying to route what was effectively the same domain again. This normally
8113 caused routing to fail (because the router wouldn't handle the domain
8114 twice). Now Exim checks for this case specially, and just changes the
8115 casing of the domain that it ultimately uses when it transmits the message
8118 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8121 50. If a filter generated a file delivery with a non-absolute name (possible if
8122 no home directory exists for the router), the forbid_file option was not
8125 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8126 the existing equality matching.
8128 52. Exim was using ints instead of ino_t variables in some places where it was
8129 dealing with inode numbers.
8131 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8132 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8133 called TMPDIR, and if it finds it is different, it changes its value.
8135 54. The smtp_printf() function is now made available to local_scan() so
8136 additional output lines can be written before returning. There is also an
8137 smtp_fflush() function to enable the detection of a dropped connection.
8138 The variables smtp_input and smtp_batched_input are exported to
8141 55. Changed the default runtime configuration: the message "Unknown user"
8142 has been removed from the ACL, and instead placed on the localuser router,
8143 using the cannot_route_message feature. This means that any verification
8144 failures that generate their own messages won't get overridden. Similarly,
8145 the "Unrouteable address" message that was in the ACL for unverifiable
8146 relay addresses has also been removed.
8148 56. Added hosts_avoid_esmtp to the smtp transport.
8150 57. The exicyclog script was not checking for the esoteric option
8151 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8152 will work only if exicyclog is run under the appropriate euid.
8154 58. Following a discussion on the list, the rules by which Exim recognises line
8155 endings on incoming messages have been changed. The -dropcr and drop_cr
8156 options are now no-ops, retained only for backwards compatibility. The
8157 following line terminators are recognized: LF CRLF CR. However, special
8158 processing applies to CR:
8160 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8161 nor a local message in the state where . is a terminator.
8163 (ii) If a bare CR is encountered in a header line, an extra space is added
8164 after the line terminator so as not to end the header. The reasoning
8165 behind this is that bare CRs in header lines are most likely either
8166 to be mistakes, or people trying to play silly games.
8168 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8169 was being incorrectly given as 18 bytes larger than it should have been.
8170 This is a VOB (very old bug).
8172 60. This may never have affected anything current, but just in case it has:
8173 When the local host is found other than at the start of a list of hosts,
8174 the local host, those with the same MX, and any that follow, are discarded.
8175 When the list in question was part of a longer list of hosts, the following
8176 hosts (not currently being processed) were also being discarded. This no
8177 longer happens. I'm not sure if this situation could ever has previously
8180 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8183 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8184 header. This is something that is recommended in a new Internet Draft, and
8185 is something that is documented as being done by Sendmail. There are two
8186 possible values. For messages generated by the autoreply transport, Exim
8189 Auto-Submitted: auto-replied
8191 whereas for all other generated messages (e.g. bounces) it adds
8193 Auto-Submitted: auto-generated
8195 63. The "personal" condition in filters now includes a test for the
8196 Auto-Submitted: header. If it contains the string "auto-" the message it
8197 not considered personal.
8199 64. Added rcpt_include_affixes as a generic transport option.
8201 65. Added queue_only_override (default true).
8203 66. Added the syslog_duplication option.
8205 67. If what should have been the first header line of a message consisted of
8206 a space followed by a colon, Exim was mis-interpreting it as a header line.
8207 It isn't of course - it is syntactically invalid and should therefore be
8208 treated as the start of the message body. The misbehaviour could have
8209 caused a number of strange effects, including loss of data in subsequent
8210 header lines, and spool format errors.
8212 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8213 client host had authenticated. This control can now be exercised by an ACL
8214 for more flexibility.
8216 69. By default, callouts do not happen when testing with -bh. There is now a
8217 variant, -bhc, which does actually run the callout code, including
8218 consulting and updating the callout cache.
8220 70. Added support for saslauthd authentication, courtesy of Alexander
8223 71. If statvfs() failed on the spool or log directories while checking their
8224 size for availability, Exim confusingly gave the error "space shortage".
8225 Furthermore, in debugging mode it crashed with a floating point exception.
8226 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8227 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8228 serious problem, Exim now writes to the main and panic logs when this
8229 happens, with details of the failure. It then refuses to accept the
8230 incoming message, giving the message "spool directory problem" or "log
8231 directory problem" with a 421 code for SMTP messages.
8233 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8234 0, 1, and 2 exist, because some OS complain for execs without them (see
8235 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8236 descriptors. However, the code omitted to check that the open succeeded,
8237 causing mysterious errors if for some reason the permissions on /dev/null
8238 got screwed. Now Exim writes a message to the main and panic logs, and
8239 bombs out if it can't open /dev/null.
8241 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8242 interact so that it is all more flexible. It is supposed to remain
8243 backwards compatible. Also added extra_local_interfaces.
8245 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8246 to bomb out with an assertion failure - to the client this appears as a
8247 connection drop. This problem occurs in the part of the code that was taken
8248 from the Samba project. Fortunately, the assertion is in a very simple
8249 function, so I have fixed this by reproducing the function inline in the
8250 one place where it is called, and arranging for authentication to fail
8251 instead of killing the process with assert().
8253 75. The SPA client code was not working when the server requested OEM rather
8254 than Unicode encoding.
8256 76. Added code to make require_files with a specific uid setting more usable in
8257 the case where statting the file as root fails - usually a non-root-mounted
8258 NFS file system. When this happens and the failure is EACCES, Exim now
8259 forks a subprocess and does the per-uid checking as the relevant uid.
8261 77. Added process_log_path.
8263 78. If log_file_path was not explicitly set, a setting of check_log_space or
8264 check_log_inodes was ignored.
8266 79. If a space check for the spool or log partitions fails, the incident is now
8267 logged. Of course, in the latter case the data may get lost...
8269 80. Added the %p formatting code to string_format() so that it can be used to
8270 print addresses in debug_print(). Adjusted all the address printing in the
8271 debugging in store.c to use %p rather than %d.
8273 81. There was a concern that a line of code in smtp_in.c could overflow a
8274 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8275 initially expressed, the concern was not well-founded, because trailing
8276 spaces are removed early. However, if the trailing spaces were followed by
8277 a NULL, they did not get removed, so the overflow was possible. Two fixes
8280 (a) I re-wrote the offending code in a cleaner fashion.
8281 (b) If an incoming SMTP command contains a NULL character, it is rejected
8284 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8285 runs initgroups(), so that if the Exim user is in any additional groups,
8286 they will be used during message reception.
8292 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob