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 Support list of dkim results in the dkim_status ACL condition, making
9 it more usable in the data ACL.
11 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
12 Previously This was only logged, on the assumption that errors would be
13 seen for a previous fflush(). However, a fuse filesystem has been
14 reported as showing this an error for the fclose(). The spool is now in
15 an uncertain state, and we have logged and responded acceptance. Change
16 this to respond with a temp-reject, wipe spoolfiles, and log the error
19 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
20 during a recipient verify callout, a QUIT command was attempted on the
21 now-closed callout channel, causing a paniclog entry.
23 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
24 a connection_reject log_selector, under tls_on_connect. Previously
25 with this combination, when the connect ACL rejected, a spurious
26 paniclog entry was made.
28 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
29 of loadbalancer-detection for resumption, in 4.96 - which tries to
30 use the EHLO response. SMTPS does not have one at the time it is starting
31 TLS. Change the default for the smtp transport host_name_extract option
32 to be a static string, for TLS-on-connect cases; meaning that resumption
33 will always be attempted (unless deliberately overriden).
35 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
36 chunk-separator specification. This was broken by hardening introduced
39 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
40 included the -H with the id; this also messed up exiqgrep.
42 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
43 was not properly skipped and empty elements would cause mis-parsing.
44 Tighten parsing of DKIM header records. Previously, all but lowercase
45 alpha chars would be ignored in potential tag names.
47 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
48 being searched. Previously we only had one for IPv4 addresses. Per the
49 documentation, the error results by default in a no-match result for the
50 list. It is logged if the unknown_in_list log_selector is used.
52 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
53 the routing operation. Previously it would silently stop routing the
56 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
57 info-request event close in time to a scheduled run timer could result in
58 the latter being missed, and no further queue scheduled runs being
59 initiated. This ouwld be more likely on high-load systems.
61 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
62 LF-only mode (as detected from the first header line). Previously we did
63 accept that in (normal) CRLF mode; this has been raised as a possible
64 attack scenario (under the name "smtp smuggling").
66 JH/13 Add an fdatasync call for the received message data file in spool, before
67 loggging reception and sending the SMTP ack. Previously we only flushed
68 the stdio buffer so there was still the possibility of a disk error.
70 JH/14 Bug 3061: Avoid a split log line when trtying to rewrite a malformed
71 address. Previously, for the last address in a header line (commonly
72 there is only one) the terminating newline was part of the logged
75 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
76 header, when parsing for address-qualification. Previously one was only
77 written if there were rewrite rules.
79 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
82 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
83 the multiple-queue-runners facility for 4.97 broking this, giving only
84 a one-time run of the queue.
86 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
93 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
94 SMTP connection" log lines.
96 JH/02 Option default value updates:
97 - queue_fast_ramp (main) true (was false)
98 - remote_max_parallel (main) 4 (was 2)
100 JH/03 Cache static regex pattern compilations, for use by ACLs.
102 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
103 Make the rewrite never match and keep the logging. Trust the
104 admin to be using verify=header-syntax (to actually reject the message).
106 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
107 (under Linux) we watch the dir containing the final file; previously
108 it would be the dir with the first symlink. We still do not monitor
111 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
112 dn_expand() is happy to pass them through.
114 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
115 the file with the proof had an unchanged name, the new proof(s) were
116 loaded on top of the old ones (and nover used; the old ones were stapled).
118 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
119 more than one message arrived in a single connection a reference from
120 the earlier message could be re-used. Often a sigsegv resulted.
121 These variables were introduced in Exim 4.87.
122 Debug help from Graeme Fowler.
124 JH/09 Fix ${filter } for conditions that modify $value. Previously the
125 modified version would be used in construction the result, and a memory
128 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
129 Find and fix by Jasen Betts.
131 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
132 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
133 the systemwide configuration to override the Exim config.
135 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
136 API changes in libopendmarc.
138 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
139 pid 1, in the normal "background daemon" mode, having to drop process-
140 group leadership also lost track of needing to create listener sockets.
142 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
143 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
145 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
146 a capture group which obtained no text (eg. "(abc)*" matching zero
147 occurrences) could cause a segfault if the corresponding $<n> was
150 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
151 included a close-brace character (eg. it itself used an expansion) an
154 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
155 starting TLS. Previously it was after, meaning that attackers on such
156 ports had to be screened using the host_reject_connection main config
157 option. The new sequence aligns better with the STARTTLS behaviour, and
158 permits defences against crypto-processing load attacks, even though it
159 is strictly an incompatible change.
160 Also, avoid sending any SMTP fail response for either the connect ACL
161 or host_reject_connection, for TLS-on-connect ports.
163 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
164 Previously this was not permitted, but it makes reasonable sense.
165 While there, restore a restriction on using it from a connect ACL; given
166 the change JH/16 it could only return false (and before 4.91 was not
169 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
170 was exactly sized compared to the log buffer, a crash occurred with the
171 misleading message "bad memory reference; pool not found".
172 Found and traced by Jasen Betts.
174 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
175 dns_again_means_nonexist included an element causing a DNS lookup which
176 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
177 included (though probably not limited to) a process crash from stack
178 memory limit, or from excessive open files. Replace this with a paniclog
179 whine (as this is likely a configuration error), and returning
182 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
183 this always failed, probably leading to the usual downgrade to in-clear
186 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
187 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
188 to not checking that list for these lookups.
190 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
193 JH/23 Fix crash in string expansions. Previously, if an empty variable was
194 immediately followed by an expansion operator, a null-indirection read
195 was done, killing the process.
197 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
198 include an SMTP response string which is longer than that supported
199 by the delivering transport. Alleviate by wrapping such lines before
202 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
203 chars (RFC limit). Previously a limit of 12 items was made, which with
204 a not-impossible References: in the message being bounced could still
205 be over-large and get stopped in the transport.
207 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
208 close. Previously a bare socket close was done.
210 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
213 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
214 message being considered for delivery was already being handled by
215 another process, and having an SMTP connection already open, the function
216 to close it tried to use an uninitialized variable. This would afftect
217 high-volume sites more, especially when running mailing-list-style loads.
218 Pollution of logs was the major effect, as the other process delivered
219 the message. Found and partly investigated by Graeme Fowler.
221 JH/29 Change format of the internal ID used for message identification. The old
222 version only supported 31 bits for a PID element; the new 64 (on systems
223 which can use Base-62 encoding, which is all currently supported ones
224 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
225 and must use Base-36). The new ID is 23 characters rather than 16, and is
226 visible in various places - notably logs, message headers, and spool file
227 names. Various of the ancillary utilities also have to know the format.
228 As well as the expanded PID portion, the sub-second part of the time
229 recorded in the ID is expanded to support finer precision. Theoretically
230 this permits a receive rate from a single comms channel of better than the
232 The major timestamp part of the ID is not changed; at 6 characters it is
233 usable until about year 3700.
234 Updating from previously releases is fully supported: old-format spool
235 files are still usable, and the utilities support both formats. New
236 message will use the new format. The one hints-DB file type which uses
237 message-IDs (the transport wait- DB) will be discarded if an old-format ID
238 is seen; new ones will be built with only new-format IDs.
239 Optionally, a utility can be used to convert spool files from old to new,
240 but this is only an efficiency measure not a requirement for operation
241 Downgrading from new to old requires running a provided utility, having
242 first stopped all operations. This will convert any spool files from new
243 back to old (losing time-precision and PID information) and remove any
244 wait- hints databases.
246 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
247 we treated them as item separators when parsing for a list item, but they
248 need to be protected by the doublequotes. While there, add handling for
251 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
252 Found and fixed by Jasen Betts. No testcase for this as my usual text
253 editor insists on emitting only valid UTF-8.
255 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
256 a null-indirection SIGSEGV for the receive process.
258 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
259 -bh use. Internal checking would spot this and take a panic.
261 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
262 In 4.96 this would expand to empty.
264 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
265 certificate. Find and fix by Andreas Metzler.
267 JH/36 Add ARC info to DMARC hostory records.
269 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
270 or fakedefer. Previously the sender could discover that the message
271 had in fact been accepted.
273 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
274 sequences. Previously the input was not noted as being tainted; notably
275 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
276 bad coding of authenticators.
278 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
279 and ${tr...}. Found and diagnosed by Heiko Schlichting.
281 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
282 could be triggered by externally-supplied input. Found by Trend Micro.
285 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
286 be triggered by externally-controlled input. Found by Trend Micro.
289 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
290 be triggered by externally-controlled input. Found by Trend Micro.
293 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
294 Make the rewrite never match and keep the logging. Trust the
295 admin to be using verify=header-syntax (to actually reject the message).
297 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
300 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
306 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
307 after reception to before a subsequent reception. This should
308 mean slightly faster delivery, and also confirmation of reception
311 JH/02 Move from using the pcre library to pcre2. The former is no longer
312 being developed or supported (by the original developer).
314 JH/03 Constification work in the filters module required a major version
315 bump for the local-scan API. Specifically, the "headers_charset"
316 global which is visible via the API is now const and may therefore
317 not be modified by local-scan code.
319 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
320 sendfile() didi not account for the way the ClamAV driver code called it.
322 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
323 time check was being done for every character; replace that with one
326 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
327 sent was prefixed with a length byte.
329 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
330 RFC 5321. Previously Dovecot (at least) would log errors during
333 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
334 in 4.61 without a reason logged. Binaries will be bigger, which might
335 matter on diskspace-constrained systems, but debug is easier.
337 JH/09 Fix macro-definition during "-be" expansion testing. The move to
338 write-protected store for macros had not accounted for these runtime
339 additions; fix by removing this protection for "-be" mode.
341 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
342 handing out large-numbered file descriptors, violating the usual Unix
343 assumption (and required by Posix) that the lowest possible number will be
344 allocated by the kernel when a new one is needed. In the daemon, and any
345 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
346 are not useable for FD_SET() [and hence select()] and overwrite the stack.
347 Assorted crashes happen.
349 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
350 main-section options or in a connect ACL, the value from the first ever
351 connection was never replaced for subsequent connections. Found by
354 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
355 though only once PCRE2 was introduced: the memory accounting used under
356 debug offset allocations by an int, giving a hard trap in early startup.
357 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
359 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
360 with underbars is given. The write-protection of configuration introduced
361 in 4.95 trapped when normalisation was applied to an option not needing
364 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
366 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
367 failing to close the certificates directory, every hour or any time it
370 JH/16 Debugging initiated by an ACL control now continues through into routing
371 and transport processes. Previously debugging stopped any time Exim
372 re-execs, or for processing a queued message.
374 JH/17 The "expand" debug selector now gives more detail, specifically on the
375 result of expansion operators and items.
377 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
378 bad comparison between the option value and the name of the file to
379 be included was done, and a mismatch was wrongly identified.
380 4.88 to 4.95 are affected.
382 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
384 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
385 supplied as the db file-pair basename. Previously, if a directory
386 path was given, for example via the autoreply "once" option, the DB
387 file.pag and file.dir files would be created in that directory's
390 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
391 "taint" log_selector. These were previously deprecated.
393 JH/22 Fix static address-list lookups to properly return the matched item.
394 Previously only the domain part was returned.
396 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
397 the call into OpenSSL to send a TLS Close was being repeated; this
398 resulted in the library waiting for the peer's Close. If that was never
399 sent we waited forever. Fix by tracking send calls.
401 JH/24 The ${run} expansion item now expands its command string elements after
402 splitting. Previously it was before; the new ordering makes handling
403 zero-length arguments simpler. The old ordering can be obtained by
404 appending a new option "preexpand", after a comma, to the "run".
406 JH/25 Taint-check exec arguments for transport-initiated external processes.
407 Previously, tainted values could be used. This affects "pipe", "lmtp" and
408 "queryprogram" transport, transport-filter, and ETRN commands.
409 The ${run} expansion is also affected: in "preexpand" mode no part of
410 the command line may be tainted, in default mode the executable name
413 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
414 the the facility was not passed across execs, and only the first message
415 passed over a connection could use BDAT; any further ones using DATA.
417 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
418 uses $sending_ip_address and an interface is specified.
419 Previously any use of the local address in the EHLO name disabled
420 PIPECONNECT, the common case being to use the rDNS of it.
422 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
423 resumption. Previously verify failed because no certificate status is
424 passed on the wire for the restarted session. Fix by using the recorded
425 ocsp status of the stored session for the new connection.
427 JH/29 TLS resumption: the key for session lookup in the client now includes
428 more info that a server could potentially use in configuring a TLS
429 session, avoiding oferring mismatching sessions to such a server.
430 Previously only the server IP was used.
432 JH/30 Fix string_copyn() for limit greater than actual string length.
433 Previously the copied amount was the limit, which could result in a
434 overlapping memcpy for newly allocated destination soon after a
435 source string shorter than the limit. Found/investigated by KM.
437 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
438 close; it may be needed for a subsequent connection. This caused a
439 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
441 JH/32 Fix CHUNKING for a second message on a connection when the first was
442 rejected. Previously we did not reset the chunking-offered state, and
443 erroneously rejected the BDAT command. Investigation help from
446 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
447 an empty address. Previously the expansion returned an error.
449 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
450 proxy. Previously these were misparsed, leading to paniclog entries.
456 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
457 related applications. Previously an "H" was used where available info
458 says that "M" should be, so change to match.
460 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
461 as arguments, so an implementation trying to copy these into a local
462 buffer was taking a taint-enforcement trap. Fix by using dynamically
463 created buffers. Similar fix for radius expansion condition.
465 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
466 reasonable, eg. to count headers. Fix by using dynamically created
467 buffers rather than a local. Do similar fixes for ACL actions "dcc",
468 "log_reject_target", "malware" and "spam"; the arguments are expanded
469 so could be handling tainted values.
471 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
472 broken the (no-op) support for this sendmail command. Restore it
473 to doing nothing, silently, and returning good status.
475 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
476 record path was given (or the default used) without a leading directory
477 path, an error occurred on trying to open it. Use the transport's working
480 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
481 transport. Previously it was the name on the DNS A-record; use instead
482 the head of the CNAME chain leading there (if there is one). This seems
483 to align better with RFC 6125.
485 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
486 smtp_accept_max_per_host allocated resources which were not released
487 when the limit was exceeded. This eventually crashed the daemon. Fix
488 by adding a release action in that path.
490 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
491 expanded; previously using tainted values was rejected. Fix by using
492 dynamically-created buffers.
494 JH/09 Relax restrictions on ACL verify condition needing access to message
495 headers. Previously they were only permitted in data and non-smtp ACLs;
496 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
497 not_blind, header_sender and header_names_ascii verification.
499 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
500 Previously a macro used one argument twice; when called with the
501 argument as an expression having side-effects, incorrect operation
502 resulted. Use an inlineable function.
504 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
505 held open for a verify callout. Previously this wan not accounted for
506 and a corrupt onward SMTP conversation resulted.
508 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
509 Previously they were embedded naively in the constructed address; when
510 needed, strip the quoting and quote the entire local_part.
511 Also make the inbound_srs expansion condition handle quoting.
513 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
514 excluded, not matching the documentation.
516 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
517 was given for the sqlite_dbfile a trap resulted.
519 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
520 "name" argument resulted in a trap. There is no reason to disallow such;
521 this was a coding error.
523 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
524 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
525 spent suspended, ignoring the POSIX definition. Previously we assumed
526 it did and a constant offset from real time could be used as a correction.
527 Change to using the same clock source for the start-of-message and the
528 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
529 exists, just to get a clock slightly more aligned to reality.
531 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
532 RFC says it is optional some validators care. The missing char was not
533 intended but triggered by a line-wrap alignment. Discovery and fix by
534 Guillaume Outters, hacked on by JH.
536 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
537 name being quoted was tainted a trap would be taken. Fix by using
538 dynamically created buffers. The routine could have been called by a
539 rewrite with the "h" flag, by using the "-F" command-line option, or
540 by using a "name=" option on a control=submission ACL modifier.
542 JH/19 SPF: change the Authentication-Results expansion component to give
543 smtp.helo when the sender domain is empty. Previously it gave
546 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
547 not in 127.0.0.0/8 to help in spotting list domains taken over by a
548 domain-parking registrar.
550 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
551 Previously when a whitespace character was specified it was not inserted
552 after removing the newline.
554 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
555 the domain part of the recipient address. This overrides any tls_sni
556 option set, which was previously used.
558 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
561 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
562 is_tainted() had an off-by-one error in the overenthusiastic direction.
563 Find and fix by Gavan. Although NetBSD is not a supported platform for
564 4.94 this bug could affect other platforms.
566 PP/01 Fix default prime selection to be consistent.
567 One path used ike23 still, instead of exim.dev.20160529.3; now both
568 execution flows will use the same DH primes (currently
569 exim.dev.20160529.3).
571 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
572 option in smtp transport, to match the documentation. Previously
573 verification was not being done in some cases where it should have been.
575 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
576 than one server was defined and depending on the platform memory layout
577 details, an internal consistency trap could be hit while walking the list
580 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
581 files. The value is used by the authresults expansion item. Previously
582 if this was used in a router or transport, a crash could result.
584 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
585 have a sufficient incidence of EINTR returns from select that an
586 interaction with other operations done by the main daemon loop exposed
587 a bug in the error-handling. This was benign apart from the log
590 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
591 for consistency with delivered "=>" and failed "**" lines. While we're
592 there, handle PRX and TFO.
594 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
595 applied. This resulted, if any header-line rewrite rules were configured,
596 in a panic-log triggerable by sending a message with a long address in
597 a header. Fix by increasing the arbitrary limit to larger than a single
598 (dewrapped) 5322 header line maximum size.
600 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
601 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
602 RFC 6648 which deprecates X- options in protocols as a general practice.
603 Changeover between the implementations is handled by the mechanisms
606 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
607 a remote which uses callout/hold. Previously the local was queued.
609 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
612 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
613 log line. The asterisk is now consistently appended to the remote IP
614 (and port, if given), and will also be provided on defer and fail log
615 lines. Previously it could be placed on the local IP if that was being
616 logged, and was only provided on delivery lines.
618 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
620 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
621 authenticator client_send option. Previously the next char, after a pair
622 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
623 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
624 documentation. There is still no way to get a leading ^ immediately
625 after a NUL (ie. for the password of a PLAIN method authenticator.
627 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
628 files. For bad sizes read, delete the record and whine to paniclog.
630 JH/38 When logging an AUTH failure, as server, do not include sensitive
631 information. Previously, the credentials would be included if given
632 as part of the AUTH command line and an ACL denied authentication.
634 JH/39 Bug 2691: fix $local_part_data. When the matching list element
635 referred to a file, bad data was returned. This likely also affected
638 JH/40 The gsasl authenticator now supports caching of the salted password
639 generated by the client-side implementation. This required the addition
640 of a new variable: $auth4.
642 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
643 left undeleted; the attempt to re-create it then failed - resulting in
644 the usual "SIGHUP tp have daemon reload configuration" to not work.
645 This affected any platform not supporting "abstract" Unix-domain
646 sockets (i.e. not Linux).
648 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
649 recipients" response to RCPT in a later response, with a 250. The
650 previous coding assumed this would not happen, and under PIPELINING
651 would result in both lost and duplicate recipients for a message.
653 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
654 Previously the weighting was incorrectly applied. Similar fix for socks
655 proxies. Found and fixed by Heiko Schlichting.
657 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
658 not handle sub-lists included using the +namedlist syntax. While
659 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
660 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
663 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
664 Previously the permanent pool was used, so the sore could not be freed.
665 This meant a connection with many messages would use continually-growing
668 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
669 per-pool so as not to waste too much space. Previously a constant size
670 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
671 DOS attacks harder. The cost is wasted memory use in the larger blocks.
673 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
674 same space repeatedly, and a smaller process footprint.
676 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
677 for GnuTLS. Previously the name started "%s" which, while not a bug,
678 looked as if if might be one.
680 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
681 "continued connection") make the $proxy_* variables available. Previously
682 the information was not passed across the exec() call for subsequent
683 transport executions. This also mean that the log lines for the
684 messages can show the proxy information.
686 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
687 receive time. With modern systems the difference is significant.
688 The historical behaviour can be restored by disabling (a new) log_selector
689 "queue_time_exclusive".
691 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
692 content) were specifically tested for. Now, also cover expansions
693 resulting in ACL names and inline ACL content.
695 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
696 making it unusable in complex expressions.
698 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
699 fail with a taint-check on the spoolfile name, and leave the message
702 HS/01 Enforce absolute PID file path name.
704 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
706 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
708 PP/02 Bug 2643: Correct TLS DH constants.
709 A missing NUL termination in our code-generation tool had led to some
710 incorrect Diffie-Hellman constants in the Exim source.
711 Reported by kylon94, code-gen tool fix by Simon Arlott.
713 PP/03 Impose security length checks on various command-line options.
714 Fixes CVE-2020-SPRSS reported by Qualys.
716 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
717 better. Reported by Qualys.
719 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
720 providing a particularly obnoxious sender full name.
723 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
725 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
728 PP/08 Change default for recipients_max from unlimited to 50,000.
730 PP/09 Fix security issue with too many recipients on a message (to remove a
731 known security problem if someone does set recipients_max to unlimited,
732 or if local additions add to the recipient list).
733 Fixes CVE-2020-RCPTL reported by Qualys.
735 PP/10 Fix security issue in SMTP verb option parsing
736 Fixes CVE-2020-EXOPT reported by Qualys.
738 PP/11 Fix security issue in BDAT state confusion.
739 Ensure we reset known-good where we know we need to not be reading BDAT
740 data, as a general case fix, and move the places where we switch to BDAT
741 mode until after various protocol state checks.
742 Fixes CVE-2020-BDATA reported by Qualys.
744 HS/03 Die on "/../" in msglog file names
746 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
747 the Exim runtime user are allowed to create files.
749 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
752 QS/03 When reading the output from interpreted forward files we do not
753 pass the pipe between the parent and the interpreting process to
754 executed child processes (if any).
756 QS/04 Always die if requested from internal logging, even is logging is
759 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
760 the API; compilation noo longer completes with DMARC support included.
761 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
762 been reported on other platforms.
764 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
766 JH/56 Make the majority of info read from config files readonly, for defence-in-
767 depth against exploits. Suggestion by Qualys.
768 Not supported on Solaris 10.
770 JH/57 Fix control=fakreject for a custom message containing tainted data.
771 Previously this resulted in a log complaint, due to a re-expansion present
772 since fakereject was originally introduced.
774 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
775 modifier was given, a loop resulted.
777 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
778 pipelined SMTP command followed the BDAT LAST then it would be
779 incorrectly treated as part of the message body, causing a verification
782 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
783 require looser validation rules than those for 821-level addresses,
784 which only permit IP addresses.
790 JH/01 Avoid costly startup code when not strictly needed. This reduces time
791 for some exim process initialisations. It does mean that the logging
792 of TLS configuration problems is only done for the daemon startup.
794 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
796 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
797 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
800 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
801 DKIM signing is being done. Previously a transport_filter would always
802 disable CHUNKING, falling back to traditional DATA.
804 JH/05 Regard command-line recipients as tainted.
806 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
808 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
809 PAM library frees one of the arguments given to it, despite the
810 documentation. Therefore a plain malloc must be used.
812 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
813 on-stack buffers were used, resulting in a taint trap when DSN information
814 copied from a received message was written into the buffer.
816 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
817 the ordering of its ARC headers. This caused a crash.
819 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
820 a new record was being constructed with information from the peer, a trap
823 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
824 installation would get error messages from DMARC verify, when it hit the
825 nonexistent file indicated by the default. Distros wanting DMARC enabled
826 should both provide the file and set the option.
827 Also enforce no DMARC verification for command-line sourced messages.
829 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
830 could, depending on the platform, hang at the STARTTLS response.
832 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
833 message on a connection. Previously if one message had ARC headers and
834 the following one did not, a crash could result when adding an
835 Authentication-Results: header.
837 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
838 Exim main code and Exim-related utities. The introduction of taint
839 tracking also did many adjustments to string handling. Since then, eximon
840 frequently terminated with an assert failure.
842 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
843 check for 452 responses. This slightly helps the inefficieny of doing
844 a large alias-expansion into a recipient-limited target. The max_rcpt
845 transport option still applies (and at the current default, will override
846 the new feature). The check is done for either cause of synch, and forces
847 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
848 connection. The new facility is not tunable at this time.
850 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
851 library live data was being used, so the results became garbage. Make
852 copies while it is still usable.
854 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
855 on delivery deferred (==) and failed (**) lines (if a delivery was
856 attemtped). Previously it was only on completion (=>) lines.
858 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
859 for the expansion of the server_scram_iter and server_scram_salt options.
861 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
862 are now specifically given a NO_DATA response without hitting the system
863 resolver. The library goes on to do the now-standard TXT lookup.
864 Use of dnsdb lookups is not affected.
866 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
867 only retrieve the errormessage once. Previously two calls to dlerror()
868 were used, and the second one (for mainlog/paniclog) retrieved null
871 JH/20 Taint checking: disallow use of tainted data for
872 - the appendfile transport file and directory options
873 - the pipe transport command
874 - the autoreply transport file, log and once options
875 - file names used by the redirect router (including filter files)
877 - paths used by single-key lookups
878 Previously this was permitted.
880 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
881 adjusted the size of a major service buffer; this failed because the
882 buffer was in use at the time. Change to a compile-time increase in the
883 buffer size, when this authenticator is compiled into exim.
885 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
886 previous fast-mode was untenable in the face of glibs using mmap to
887 support larger malloc requests.
889 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
890 New values supported, if defined on system where compiled:
891 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
892 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
894 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
895 running a limited number of proceses in parallel, a benefit is gained. The
896 amount varies with the platform hardware and load. The use of the option
897 queue_run_in_order means we cannot do this, as ordering becomes
900 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
901 had introduced a string-copy (for ensuring NUL-termination) which was not
902 appropriate for that case, which can include embedded NUL bytes in the
903 block of data. Investigation showed the copy to actually be needless, the
904 data being length-specified.
906 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
907 done during a receiving connection, and both used TLS, global info was
908 used rather than per-connection info for tracking the state of data
909 queued for transmission. This could result in a connection hang.
911 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
912 Previously, when delivering serveral messages down a single connection
913 only the first would provide a SIZE. This was due to the size information
914 not being properly tracked.
916 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
917 TAI (at 37 seconds currently), pretend to be in UTC for time-related
918 expansion and logging. Previously, spurious values such as a future
919 minute could be seen.
921 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
922 it could crash from a null-deref. This could also affect the
923 ${addresses: } operator and ${readsock } item.
925 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
926 message following a mime one, the variable was not reset.
928 JH/30 When an pipelined-connect fails at the first response, assume incorrect
929 cached capability (perhaps the peer reneged?) and immediately retry in
932 JH/31 Fix spurious detection of timeout while writing to transport filter.
934 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
935 an attempt to copy the string was made before checking it.
937 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
938 taint of the lookup key was maintained; we now regard the presence in the
939 filesystem as sufficient validation.
941 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
942 argument is supplied.
944 JH/35 The dsearch lookup now requires that the directory is an absolute path.
945 Previously this was not checked, and nonempty relative paths made an
946 access under Exim's current working directory.
948 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
949 Previously no event was raised.
951 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
952 parameter supplied by the sender MAIL FROM command. Previously it was
953 ignored, and only the check_spool_space option value for the required
956 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
957 the size of the signing public-key. Previously it was instead giving
958 the size of the signature hash.
960 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
961 the default. See the (new) dkim_verify_min_keysizes option.
963 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
964 and an ACL use a lookup for checking either the local_part or domain,
965 stale data could be accessed. Ensure that variable references are
966 dropped between messages.
968 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
969 by the client was not checked as pointing within response data before
970 being used. A malicious client could thus cause an out-of-bounds read and
971 possibly gain authentication. Fix by adding the check.
973 JH/42 Internationalisation: change the default for downconversion in the smtp
974 transport to be "if needed". Previously it was "as previously set" for
975 the message, which usually meant "if needed" for message-submission but
976 "no" for everything else. However, MTAs have been seen using SMTPUTF8
977 even when the envelope addresses did not need it, resulting in forwarding
978 failures to non-supporting MTAs. A downconvert in such cases will be
979 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
980 transport. The change does mean that addresses needing conversion will
981 be converted when previously a delivery failure would occur.
983 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
984 response was received it would be used unchecked in a fail-DSN, violating
985 standards on line-length limits. Truncate if needed.
987 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
988 doesn't work. (Additionally add a new main config option to configure the
995 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
996 side, to decode a TLS 1.3 packet capture.
998 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
999 Previously the default library behaviour applied, sending two, each in
1000 its own TCP segment.
1002 JH/03 Debug output for ACL now gives the config file name and line number for
1005 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1007 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1009 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1010 buffer overrun for (non-chunking) other transports.
1012 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1013 TLS1.3, means that a server rejecting a client certificate is not visible
1014 to the client until the first read of encrypted data (typically the
1015 response to EHLO). Add detection for that case and treat it as a failed
1016 TLS connection attempt, so that the normal retry-in-clear can work (if
1017 suitably configured).
1019 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1020 and/or domain. Found and fixed by Jason Betts.
1022 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1023 configuration). If a CNAME target was not a wellformed name pattern, a
1026 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1027 the OS reports them interleaved with other addresses.
1029 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1030 used both for input and for a verify callout, both encrypted, SMTP
1031 responses being sent by the server could be lost. This resulted in
1032 dropped connections and sometimes bounces generated by a peer sending
1035 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1036 string. Previously it was possible to cause undefined behaviour in a
1037 library routine (usually a crash). Found by "zerons".
1039 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1042 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1043 API was removed, so update to use the newer ones.
1045 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1046 any timeout set, is taking a long time. Previously we would hang on to a
1047 rotated logfile "forever" if the input was arriving with long gaps
1048 (a previous attempt to fix addressed lack, for a long time, of initial
1051 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1052 shared (NFS) environment. The length of the tempfile name is now
1053 4 + 16 ("hdr.$message_exim_id") which might break on file
1054 systems which restrict the file name length to lower values.
1055 (It was "hdr.$pid".)
1057 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1058 shared (NFS) environment.
1060 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1061 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1064 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1065 on some platforms for bit 31.
1067 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1068 to changes apparently associated with TLS1.3 handling some of the APIs
1069 previously used were either nonfunctional or inappropriate. Strings
1070 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1071 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1072 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1073 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1074 and the use of specific cipher names in the encrypted= ACL condition.
1076 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1078 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1079 verification result was not updated unless hosts_require_ocsp applied.
1081 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1082 queue_list_requires_admin set to false, non-admin users were denied the
1085 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1086 directory-of-certs mode. Previously they were advertised despite the
1089 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1090 A single TCP connection by a client will now hold a TLS connection open
1091 for multiple message deliveries, by default. Previously the default was to
1094 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1095 default. If built with the facility, DANE will be used. The facility
1096 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1098 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1099 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1100 must be defined and you must still, unless you define DISABLE_TLS, manage
1101 the the include-dir and library-file requirements that go with that
1102 choice. Non-TLS builds are still supported.
1104 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1107 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1108 default. If the platform supports and has the facility enabled, it will
1109 be requested on all coneections.
1111 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1112 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1114 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1116 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1117 success-DSN messages. Previously the From: header was always the default
1118 one for these; the option was ignored.
1120 JH/28 Fix the timeout on smtp response to apply to the whole response.
1121 Previously it was reset for every read, so a teergrubing peer sending
1122 single bytes within the time limit could extend the connection for a
1123 long time. Credit to Qualsys Security Advisory Team for the discovery.
1125 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1126 delivery address, which leaked information of the results of local
1127 forwarding. Change to the original envelope recipient address, per
1130 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1131 requested. Previously not bounce was generated and a log entry of
1132 error ignored was made.
1134 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1136 JH/32 Introduce a general tainting mechanism for values read from the input
1137 channel, and values derived from them. Refuse to expand any tainted
1138 values, to catch one form of exploit.
1140 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1141 was unused and the unexpanded text used for the test. Found and
1142 fixed by Ruben Jenster.
1144 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1145 an attempt to use a TLS library read routine dereffed a nul pointer,
1148 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1149 them in our smtp response.
1151 JH/36 Have the general router option retry_use_local_part default to true when
1152 any of the restrictive preconditions are set (to anything). Previously it
1153 was only for check_local user. The change removes one item of manual
1154 configuration which is required for proper retries when a remote router
1155 handles a subset of addresses for a domain.
1157 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1158 link count into consideration.
1160 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1161 caused the extension of big_buffer, the following lines were ignored.
1163 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1164 accordance with RFC 2308. Previously there was no expiry, so a longlived
1165 receive process (eg. due to ACL delays) versus a short SOA value could
1168 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1170 JH/39 Promote DMARC support to mainline.
1172 JH/40 Bug 2452: Add a References: header to DSNs.
1174 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1175 parameters. The relevant library call is documented as "Deprecated: This
1176 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1177 3.6.0, DH parameters are negotiated following RFC7919."
1179 HS/06 Change the default of dnssec_request_domains to "*"
1181 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1182 carried on and emitted a BDAT command, even when PIPELINING was not
1185 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1186 buffer was used for the filename, resulting in a trap when tainted
1187 arguments (eg. $domain) were used.
1189 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1190 recommended to avoid a possible server-load attack. The feature can be
1191 re-enabled via the openssl_options main cofiguration option.
1193 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1194 for version 4.90 - adding a "more data" boolean to the arguments.
1195 Bumped the ABI version number also, this having been missed previously;
1196 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1197 respect of usage of smtp_printf() by either local_scan code or libraries
1198 accessed via the ${dlfunc } expansion item. Both will need coding
1199 adjustment for any calls to smtp_printf() to match the new function
1200 signature; a FALSE value for the new argument is always safe.
1202 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1203 the file-offset (which the Linux syscall does, and exim expects); this
1204 resulted in an indefinite loop.
1206 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1207 to do ARC verification. The Authentication-Results: header line added
1208 by the configuration then had no ARC item.
1210 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1211 an incorrect mode was used for reading the result, resulting in it being
1218 JH/01 Remove code calling the customisable local_scan function, unless a new
1219 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1221 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1222 non-signal-safe functions being used.
1224 JH/03 Bug 2269: When presented with a received message having a stupidly large
1225 number of DKIM-Signature headers, disable DKIM verification to avoid
1226 a resource-consumption attack. The limit is set at twenty.
1228 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1229 report of oldest_pass in ${authres } in consequence, and separate out
1230 some descriptions of reasons for verification fail.
1232 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1233 files in the spool were present and unlocked. A queue-runner could spot
1234 them, resulting in a duplicate delivery. Fix that by doing the unlock
1235 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1236 add more error-checking on spoolfile handling while that code is being
1239 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1240 No known attacks, no CVE, this is defensive hardening.
1242 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1243 a queue-runner could start a delivery while other operations were ongoing.
1244 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1245 Found and investigated by Tim Stewart. Fix by using the open message data
1246 file handle rather than opening another, and not locally closing it (which
1247 releases a lock) for that case, while creating the temporary .eml format
1248 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1250 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1251 $sender_verify_failure/$recipient_verify_failure to "random".
1253 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1256 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1257 Previously this would segfault.
1259 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1262 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1263 like zero, since the resolver should be doing this for us, But we need one
1264 as a CNAME but no MX presence gets the CNAME returned; we need to check
1265 that doesn't point to an MX to declare it "no MX returned" rather than
1266 "error, loop". A new main option is added so the older capability of
1267 following some limited number of chain links is maintained.
1269 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1271 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1272 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1273 now-deprecated earlier definitions used only specified the range up to TLS
1274 1.2 (in the older-version library docs).
1276 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1278 JH/15 Rework TLS client-side context management. Stop using a global, and
1279 explicitly pass a context around. This enables future use of TLS for
1280 connections to service-daemons (eg. malware scanning) while a client smtp
1281 connection is using TLS; with cutthrough connections this is quite likely.
1283 JH/16 Fix ARC verification to do AS checks in reverse order.
1285 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1287 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1288 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1289 Previously the "utf8" would be re-prepended for every additional message.
1291 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1292 Previously thery were accepted, resulting in issues when attempting to
1293 forward messages to a non-supporting MTA.
1295 PP/02 Let -n work with printing macros too, not just options.
1297 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1298 one parent address was copied, and bogus data was used at delivery-logging
1299 time. Either a crash (after delivery) or bogus log data could result.
1300 Discovery and analysis by Tim Stewart.
1302 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1303 Previously if the string ended mid-character, we did not insert the
1304 promised '?' replacement.
1306 PP/04 Documentation: current string operators work on bytes, not codepoints.
1308 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1309 should pack well giving a smaller memory footprint so better caching and
1310 therefore performance. Group the declarations where this can't be done so
1311 that the byte-sized flag variables are not interspersed among pointer
1312 variables, giving a better chance of good packing by the compiler.
1314 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1315 non-null, to avoid issues with sites running BATV. Previously reports were
1316 sent with an empty envelope sender so looked like bounces.
1318 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1319 The ignore_error flag wasn't being returned from the filter subprocess so
1320 was not set for later routers. Investigation and fix by Matthias Kurz.
1322 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1323 and a msg:complete for the whole, when a message is manually removed using
1324 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1326 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1327 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1328 hope that is portable enough.
1330 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1331 requiring a known-CA anchor certificate; make it now rely entirely on the
1332 TLSA as an anchor. Checking the name on the leaf cert against the name
1333 on the A-record for the host is still done for TA (but not for EE mode).
1335 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1336 included in delivery lines for non-proxied connections, when compiled with
1337 SUPPORT_SOCKS and running with proxy logging enabled.
1339 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1340 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1341 move the existing event to fire before the normal logging of message
1342 failure so that custom logging is bracketed by normal logging.
1344 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1345 msg:fail:internal event. Developement by Matthias Kurz.
1347 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1348 far too small for todays use of crypto signatures stored there. Go all
1349 the way to the max DNS message size of 64kB, even though this might be
1350 overmuch for IOT constrained device use.
1352 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1353 copy a string over itself. The library routine is documented as not
1354 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1356 JH/32 For main options check_spool_space and check_inode_space, where the
1357 platform supports 64b integers, support more than the previous 2^31 kB
1358 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1359 the previous G, M, k.
1361 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1362 $authenticated_fail_id variable on authentication failure. Previously
1365 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1366 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1367 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1368 GNUTLS_SEC_PARAM_MEDIUM.
1370 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1371 side. Previously we would continue as if no SNI had been received.
1373 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1374 "<" character, which should be a list-separator specification, we walked
1375 off past the nul-terimation.
1377 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1378 causes) even when the retry time is not yet met. Previously they were
1379 not, meaning that when (say) an account was over-quota and temp-rejecting,
1380 and multiple senders' messages were queued, only one sender would get
1381 notified on each configured delay_warning cycle.
1383 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1385 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1386 cipher-suites, an error can be left on the stack even for a succeeding
1387 accept; this results in impossible error messages when a later operation
1390 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1391 return error codes indicating retry. Under TLS1.3 this becomes required.
1393 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1394 it only wrote the new authenticators, resulting in a lack of tracking of
1395 peer changes of ESMTP extensions until the next cache flush.
1397 JH/41 Fix the loop reading a message header line to check for integer overflow,
1398 and more-often against header_maxsize. Previously a crafted message could
1399 induce a crash of the recive process; now the message is cleanly rejected.
1401 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1402 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1408 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1409 When redis_servers is set to a list of > 1 element, and the Redis servers
1410 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1411 case of MOVED into a DEFER case instead, thus moving the query onto the
1412 next server in the list. For a cluster of N elements, all N servers must
1413 be defined in redis_servers.
1415 GF/02 Catch and remove uninitialized value warning in exiqsumm
1416 Check for existence of @ARGV before looking at $ARGV[0]
1418 JH/01 Replace the store_release() internal interface with store_newblock(),
1419 which internalises the check required to safely use the old one, plus
1420 the allocate and data copy operations duplicated in both (!) of the
1421 extant use locations.
1423 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1424 modifier. This matches the restriction on the commandline.
1426 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1427 Previously only the last row was returned.
1429 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1430 we assumed that tags in the header were well-formed, and parsed the
1431 element content after inspecting only the first char of the tag.
1432 Assumptions at that stage could crash the receive process on malformed
1435 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1436 While running the DKIM ACL we operate on the Permanent memory pool so that
1437 variables created with "set" persist to the DATA ACL. Also (at any time)
1438 DNS lookups that fail create cache records using the Permanent pool. But
1439 expansions release any allocations made on the current pool - so a dnsdb
1440 lookup expansion done in the DKIM ACL releases the memory used for the
1441 DNS negative-cache, and bad things result. Solution is to switch to the
1442 Main pool for expansions.
1443 While we're in that code, add checks on the DNS cache during store_reset,
1444 active in the testsuite.
1445 Problem spotted, and debugging aided, by Wolfgang Breyha.
1447 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1448 When none of the hosts presented to a transport match an already-open
1449 connection, close it and proceed with the list. Previously we would
1450 queue the message. Spotted by Lena with Yahoo, probably involving
1453 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1454 Previously a spurious "250 OK id=" response was appended to the proper
1457 JH/08 The "support for" informational output now, which built with Content
1458 Scanning support, has a line for the malware scanner interfaces compiled
1459 in. Interface can be individually included or not at build time.
1461 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1462 by the template makefile "src/EDITME". The "STREAM" support for an older
1463 ClamAV interface method is removed.
1465 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1466 rows affected is given instead).
1468 JH/11 The runtime Berkeley DB library version is now additionally output by
1469 "exim -d -bV". Previously only the compile-time version was shown.
1471 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1472 SMTP connection. Previously, when one had more recipients than the
1473 first, an abortive onward connection was made. Move to full support for
1474 multiple onward connections in sequence, handling cutthrough connection
1475 for all multi-message initiating connections.
1477 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1478 routers. Previously, a multi-recipient message would fail to match the
1479 onward-connection opened for the first recipient, and cause its closure.
1481 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1482 a timeout on read on a GnuTLS initiating connection, resulting in the
1483 initiating connection being dropped. This mattered most when the callout
1484 was marked defer_ok. Fix to keep the two timeout-detection methods
1487 JH/15 Relax results from ACL control request to enable cutthrough, in
1488 unsupported situations, from error to silently (except under debug)
1489 ignoring. This covers use with PRDR, frozen messages, queue-only and
1492 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1494 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1495 metadata, resulting in a crash in free().
1497 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1498 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1499 Broken also in d185889f4, with init system revamp.
1501 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1502 Previously we abruptly closed the connection after reading a malware-
1503 found indication; now we go on to read the "scan ok" response line,
1506 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1507 ACL. Previously, a crash would result.
1509 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1510 macro text after a replacement (previously it was only once per line) and
1511 by skipping builtin macros when searching for an uppercase lead character.
1513 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1514 for the build is renamed.
1516 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1517 was allocated for every new TLS startup, meaning one per message. Fix
1518 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1520 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1521 reported the original. Fix to report (as far as possible) the ACL
1522 result replacing the original.
1524 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1525 OpenSSL. Certificate information is loaded for every new TLS startup,
1526 and the resources needed to be freed.
1528 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1530 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1533 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1534 DATA response info to the (existing) per-recipient response info for
1535 the "C=" log element. It can have useful tracking info from the
1536 destination system. Patch from Simon Arlott.
1538 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1539 length value. Previously this would segfault.
1541 HS/02 Support Avast multiline protoocol, this allows passing flags to
1542 newer versions of the scanner.
1544 JH/28 Ensure that variables possibly set during message acceptance are marked
1545 dead before release of memory in the daemon loop. This stops complaints
1546 about them when the debug_store option is enabled. Discovered specifically
1547 for sender_rate_period, but applies to a whole set of variables.
1548 Do the same for the queue-runner and queue-list loops, for variables set
1549 from spool message files. Do the same for the SMTP per-message loop, for
1550 certain variables indirectly set in ACL operations.
1552 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1553 as a multi-recipient message from a mailinglist manager). The coding had
1554 an arbitrary cutoff number of characters while checking for more input;
1555 enforced by writing a NUL into the buffer. This corrupted long / fast
1556 input. The problem was exposed more widely when more pipelineing of SMTP
1557 responses was introduced, and one Exim system was feeding another.
1558 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1559 receiving system, and refused recipients seen by the sending system
1560 (propating to people being dropped from mailing lists).
1561 Discovered and pinpointed by David Carter.
1563 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1564 replaced by the ${authresults } expansion.
1566 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1568 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1569 allows proper process termination in container environments.
1571 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1572 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1574 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1575 and "err_perm", deprecated since 4.83 when the RFC-defined words
1576 "temperror" and "permerror" were introduced.
1578 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1579 transport-filters or DKIM-signing. The restriction was lost in the
1580 consolidation of verify-callout and delivery SMTP handling.
1581 Extend the restriction to also cover ARC-signing.
1583 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1584 in defer=pass mode supply a 450 to the initiator. Previously the message
1587 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1588 tls_require_ciphers is used as before.
1590 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1591 "pass_unscanned". Only tmpfails from the scanner are written to
1592 the paniclog, as they may require admin intervention (permission
1593 denied, license issues). Other scanner errors (like decompression
1594 bombs) do not cause a paniclog entry.
1596 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1597 Previously it was possible to log spurious information in receive log
1600 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1601 triggered odd behaviour from Outlook Express clients.
1603 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1606 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1607 since the IETF WG has not yet settled on that versus the original
1608 "bare" representation.
1610 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1611 Previously the millisecond value corrupted the output.
1612 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1613 corrupted the output.
1619 JH/01 Rework error string handling in TLS interface so that the caller in
1620 more cases is responsible for logging. This permits library-sourced
1621 string to be attached to addresses during delivery, and collapses
1622 pairs of long lines into single ones.
1624 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1625 during configuration. Wildcards are allowed and expanded.
1627 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1628 This permits better logging.
1630 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1631 do not close it down and have the child transport start it up again on
1632 the passed-on TCP connection. Instead, proxy the child (and any
1633 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1634 affected: the continued delivery log lines do not have any DNSSEC, TLS
1635 Certificate or OCSP information. TLS cipher information is still logged.
1637 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1638 identical IP addresses on different listening ports. Will also affect
1641 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1642 add noisy ifdef guards to special-case this sillyness.
1643 Patch from Bernd Kuhls.
1645 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1646 than 255 are no longer allowed.
1648 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1649 Disable the session-cache too, which might reduce our load. Since we
1650 currrectly use a new context for every connection, both as server and
1651 client, there is no benefit for these.
1652 GnuTLS appears to not support tickets server-side by default (we don't
1653 call gnutls_session_ticket_enable_server()) but client side is enabled
1654 by default on recent versions (3.1.3 +) unless the PFS priority string
1657 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1658 <https://reproducible-builds.org/specs/source-date-epoch/>.
1660 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1661 the check for any unsuccessful recipients did not notice the limit, and
1662 erroneously found still-pending ones.
1664 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1665 MSG_MORE. Only in-clear (not on TLS connections).
1667 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1668 transport-filter is involved we can buffer the headers in memory for
1669 creating the signature, and read the spool data file once for the
1670 signature and again for transmission.
1672 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1673 4.77 as the kernel support then wasn't solid, having issues in 64bit
1674 mode. Now, it's been long enough. Add support for FreeBSD also.
1676 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1677 case where the routing stage had gathered several addresses to send to
1678 a host before calling the transport for the first, we previously failed
1679 to close down TLS in the old transport process before passing the TCP
1680 connection to the new process. The new one sent a STARTTLS command
1681 which naturally failed, giving a failed delivery and bloating the retry
1682 database. Investigation and fix prototype from Wolfgang Breyha.
1684 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1685 false-negatives in the check that the sender had not preempted a response
1686 or prompt from Exim (running as a server), due to that code's lack of
1687 awareness of the SMTP input buffering.
1689 PP/04 Add commandline_checks_require_admin option.
1690 Exim drops privileges sanely, various checks such as -be aren't a
1691 security problem, as long as you trust local users with access to their
1692 own account. When invoked by services which pass untrusted data to
1693 Exim, this might be an issue. Set this option in main configuration
1694 AND make fixes to the calling application, such as using `--` to stop
1697 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1698 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1699 responses to those, into a single TLS record each way (this usually means
1700 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1703 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1704 affects you only if you're dancing at the edge of the param size limits.
1705 If you are, and this message makes sense to you, then: raise the
1706 configured limit or use OpenSSL 1.1. Nothing we can do for older
1709 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1710 cmdline element to get the documented default one. Previously it was
1713 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1714 are unsuable for non-message-specific reasons - eg. connection timeout,
1715 banner-time rejection.
1717 JH/16 Fix logging of delivery remote port, when specified by router, under
1720 PP/06 Repair manualroute's ability to take options in any order, even if one
1721 is the name of a transport.
1724 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1726 JH/17 Change the list-building routines interface to use the expanding-string
1727 triplet model, for better allocation and copying behaviour.
1729 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1730 Previously it was constructed the first time a possibly-matching string
1731 was met in the configuration file input during startup; now it is done
1734 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1735 compatible one, to avoid the (poorly documented) possibility of a config
1736 file in the working directory redirecting the DB files, possibly correpting
1737 some existing file. CVE-2017-10140 assigned for BDB.
1739 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1740 cache-hot. Previously, although the result was properly cached, the
1741 initial verify call returned a defer.
1743 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1744 the main verify for receipient in uncached-mode.
1746 JH/22 Retire historical build files to an "unsupported" subdir. These are
1747 defined as "ones for which we have no current evidence of testing".
1749 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1750 if present. Previously it was ignored.
1752 JH/24 Start using specified-initialisers in C structure init coding. This is
1753 a C99 feature (it's 2017, so now considered safe).
1755 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1756 if was a fixed-sized field and bitmask ops via macros; it is now more
1759 PP/07 GitHub PR 56: Apply MariaDB build fix.
1760 Patch provided by Jaroslav Škarvada.
1762 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1763 during Coverity cleanups [4.87 JH/47]
1764 Diagnosis and fix provided by Michael Fischer v. Mollard.
1766 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1767 the right size to place the terminating semicolon on its own folded
1768 line, the header hash was calculated to an incorrect value thanks to
1769 the (relaxed) space the fold became.
1771 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1772 and confused the parent.
1774 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1775 which could crash as a result. This could lead to undeliverable messages.
1777 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1780 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1781 matters on fast-turnover and PID-randomising systems, which were getting
1782 out-of-order delivery.
1784 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1785 a possibly-overlapping copy. The symptom was that "Remote host closed
1786 connection in response to HELO" was logged instead of the actual 4xx
1789 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1790 Previously only that bufferd was discarded, resulting in SYMTP command
1793 JH/32 DKIM: when a message has multiple signatures matching an identity given
1794 in dkim_verify_signers, run the dkim acl once for each. Previously only
1795 one run was done. Bug 2189.
1797 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1798 "panic the current process" to "deliberately defer". The panic log is
1799 still written with the problem list name; the mail and reject logs now
1800 get a temp-reject line for the message that was being handled, saying
1801 something like "domains check lookup or other defer". The SMTP 451
1802 message is still "Temporary local problem".
1804 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1805 A crafted sequence of BDAT commands could result in in-use memory beeing
1806 freed. CVE-2017-16943.
1808 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1809 from SMTP input. Previously it was always done; now only done for DATA
1810 and not BDAT commands. CVE-2017-16944.
1812 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1813 to the message (such as an overlong header line). Previously this was
1814 not done and we did not exit BDAT mode. Followon from the previous item
1815 though a different problem.
1821 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1822 than -2003 did; needs libidn2 in addition to libidn.
1824 JH/02 The path option on a pipe transport is now expanded before use.
1826 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1827 Patch provided by "Björn", documentation fix added too.
1829 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1830 missing a wire-to-host endian conversion.
1832 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1833 close after a BDAT command line could be taken as a following command,
1834 giving a synch failure. Fix by only checking for synch immediately
1835 before acknowledging the chunk.
1837 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1838 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1839 macro. Patches provided by Josh Soref.
1841 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1842 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1843 by IANA as a service extension. However, John Klensin suggests that we
1846 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1847 the dkim code may be unix-mode line endings rather than smtp wire-format
1848 CRLF, so prepend a CR to any bare LF.
1850 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1851 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1853 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1854 the first were themselves being wrongly included in the feed into dkim
1855 processing; with most chunk sizes in use this resulted in an incorrect
1856 body hash calculated value.
1858 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1859 DKIM signature block, for verification. Although advised against by
1860 standards it is specifically not ruled illegal.
1862 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1864 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1865 missing a body hash (the bh= tag).
1867 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1868 It seems that HAProxy sends the Proxy Protocol information in clear and
1869 only then does a TLS startup, so do the same.
1871 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1872 TCP connections (such as for Spamd) unless the daemon successfully set
1873 Fast Open mode on its listening sockets. This fixes breakage seen on
1874 too-old kernels or those not configured for Fast Open, at the cost of
1875 requiring both directions being enabled for TFO, and TFO never being used
1876 by non-daemon-related Exim processes.
1878 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1879 endings, at least on the first header line. Try to canonify any that get
1880 past that check, despite the cost.
1882 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1883 now limited to an arbitrary five deep, while parsing addresses with the
1884 strip_excess_angle_brackets option enabled.
1886 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1887 instead leave the unprompted TLS handshake in socket buffer for the
1888 TLS library to consume.
1890 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1892 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1894 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1895 region in one operation, for speed, and this leaves assigned pointers
1896 dangling. Add checks run only under the testsuite which checks all
1897 variables at a store-reset and panics on a dangling pointer; add code
1898 explicitly nulling out all the variables discovered. Fixes one known
1899 bug: a transport crash, where a dangling pointer for $sending_ip_address
1900 originally assigned in a verify callout, is re-used.
1902 PP/06 Drop '.' from @INC in various Perl scripts.
1904 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1906 PP/08 Reduce a number of compilation warnings under clang; building with
1907 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1908 should be warning-free.
1910 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1912 HS/01 Fix portability problems introduced by PP/08 for platforms where
1913 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1915 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1916 chunk. This allows us to accept broken chunked messages. We need a more
1917 general solution here.
1919 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1920 already-broken messages in the queue.
1922 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1924 JH/19 Fix reference counting bug in routing-generated-address tracking.
1930 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1931 supports it and a size is available (ie. the sending peer gave us one).
1933 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1934 years of being deprecated). The replacements are the ACLs
1935 acl_smtp_mime and acl_not_smtp_mime.
1937 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1938 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1939 or even an in-clear connection were permitted. Now, if the host lookup
1940 was dnssec and dane was requested then the host is only used if the
1941 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1942 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1943 if one fails this test.
1944 This means that a poorly-configured remote DNS will make it incommunicado;
1945 but it protects against a DNS-interception attack on it.
1947 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1948 non-noisy when a race steals the message being considered.
1950 JH/05 If main configuration option tls_certificate is unset, generate a
1951 self-signed certificate for inbound TLS connections.
1953 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1954 in rewrites and routers.
1956 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1957 and logged a warning sing 4.83; now they are a configuration file error.
1959 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1960 (lacking @domain). Apply the same qualification processing as RCPT.
1962 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1964 JH/10 Support ${sha256:} applied to a string (as well as the previous
1967 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1968 a cutthrough deliver is pending, as we always want to make a connection.
1969 This also avoids re-routing the message when later placing the cutthrough
1970 connection after a verify cache hit.
1971 Do not update it with the verify result either.
1973 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1974 when routing results in more than one destination address.
1976 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1977 signing (which inhibits the cutthrough capability). Previously only
1978 the presence of an option was tested; now an expansion evaluating as
1979 empty is permissible (obviously it should depend only on data available
1980 when the cutthrough connection is made).
1982 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1983 the relevant preceding SMTP command did not note the pipelining mode.
1985 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1986 Previously they were not counted.
1988 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1989 as one having no matching records. Previously we deferred the message
1990 that needed the lookup.
1992 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1993 distinguished as "(=".
1995 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1996 for missing MX records. Previously it only worked for missing A records.
1998 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2000 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2001 after the data-go-ahead and data-ack. Patch from Jason Betts.
2003 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2004 even for a "none" policy. Patch from Tony Meyer.
2006 JH/22 Fix continued use of a connection for further deliveries. If a port was
2007 specified by a router, it must also match for the delivery to be
2010 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2011 When one of the recipients of a message was unsuitable for the connection
2012 (has no matching addresses), we lost track of needing to mark it
2013 deferred. As a result mail would be lost.
2015 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2017 JH/25 Decoding ACL controls is now done using a binary search; the source code
2018 takes up less space and should be simpler to maintain. Merge the ACL
2019 condition decode tables also, with similar effect.
2021 JH/26 Fix problem with one_time used on a redirect router which returned the
2022 parent address unchanged. A retry would see the parent address marked as
2023 delivered, so not attempt the (identical) child. As a result mail would
2026 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2027 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2028 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2031 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2032 Main config options check_{log,spool}_{inodes,space} are now
2033 100 inodes, 10MB unless set otherwise in the configuration.
2035 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2036 Previously it only applied to the main-section connection policy
2039 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2041 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2042 by me. Added RFC7919 DH primes as an alternative.
2044 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2045 are not in the system include path.
2047 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2048 GnuTLS, when a session startup failed (eg because the client disconnected)
2049 Exim did stdio operations after fclose. This was exposed by a recent
2050 change which nulled out the file handle after the fclose.
2052 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2053 signed directly by the cert-signing cert, rather than an intermediate
2054 OCSP-signing cert. This is the model used by LetsEncrypt.
2056 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2058 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2059 an incoming connection.
2061 HS/02 Bug 1802: Do not half-close the connection after sending a request
2064 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2065 fallback to "prime256v1".
2067 JH/34 SECURITY: Use proper copy of DATA command in error message.
2068 Could leak key material. Remotely exploitable. CVE-2016-9963.
2074 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2075 and 3.4.4 - once the server is enabled to respond to an OCSP request
2076 it does even when not requested, resulting in a stapling non-aware
2077 client dropping the TLS connection.
2079 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2080 support variable-length bit vectors. No functional change.
2082 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2083 The I= interface field on outgoing lines is now after the H= remote
2084 host field, same as incoming lines. There is a separate
2085 outgoing_interface log selector which allows you to disable the
2088 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2089 If not running log_selector +smtp_connection the mainlog would be held
2090 open indefinitely after a "too many connections" event, including to a
2091 deleted file after a log rotate. Leave the per net connection logging
2092 leaving it open for efficiency as that will be quickly detected by the
2093 check on the next write.
2095 HS/01 Bug 1671: Fix post transport crash.
2096 Processing the wait-<transport> messages could crash the delivery
2097 process if the message IDs didn't exist for some reason. When
2098 using 'split_spool_directory=yes' the construction of the spool
2099 file name failed already, exposing the same netto behaviour.
2101 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2102 mime_regex ACL conditions.
2104 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2105 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2106 to HELO, local diagnostic string.
2108 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2109 log line to debug. Even when configured with a tls authenticator many
2110 client connections are expected to not authenticate in this way, so
2111 an authenticate fail is not an error.
2113 HS/02 Add the Exim version string to the process info. This way exiwhat
2114 gives some more detail about the running daemon.
2116 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2117 matter for fast-change records such as DNSBLs.
2119 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2120 retry record, even if constant. There may be multiple transports with
2121 different interface settings and the retry behaviour needs to be kept
2124 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2126 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2128 JH/10 Bug 840: fix log_defer_output option of pipe transport
2130 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2131 hosts_randomize. This matters a lot when combined with mua_wrapper.
2133 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2134 ${quote_pgsql:<string>} operator.
2136 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2138 JH/14 Change header file name for freeradius-client. Relevant if compiling
2139 with Radius support; from the Gentoo tree and checked under Fedora.
2141 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2143 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2144 option transport_filter, meaning no filtering.
2146 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2148 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2149 defaults to "*" (all hosts). The variable is now available when not built
2150 with TLS, default unset, mainly to enable keeping the testsuite sane.
2151 If a server certificate is not supplied (via tls_certificate) an error is
2152 logged, and clients will find TLS connections fail on startup. Presumably
2153 they will retry in-clear.
2154 Packagers of Exim are strongly encouraged to create a server certificate
2155 at installation time.
2157 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2158 with the $config_file variable.
2160 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2161 in transport context, after the attempt, and per-recipient. The latter type
2162 is per host attempted. The event data is the error message, and the errno
2163 information encodes the lookup type (A vs. MX) used for the (first) host,
2164 and the trailing two digits of the smtp 4xx response.
2166 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2167 to write to mainlog (or rejectlog, paniclog) in the window between file
2168 creation and permissions/ownership being changed. Particularly affects
2169 installations where exicyclog is run as root, rather than exim user;
2170 result is that the running daemon panics and dies.
2172 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2174 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2175 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2176 "pri" and "weight". Note that the previous implicit priority given by the
2177 list order is no longer honoured.
2179 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2180 for DKIM processing.
2182 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2183 by defining SUPPORT_SOCKS.
2185 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2186 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2187 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2188 variables are renamed to proxy_{local,external}_{address,port}.
2190 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2191 for a build by defining SUPPORT_I18N
2193 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2194 of the query string, and make ${quote_redis:} do that quoting.
2196 JH/29 Move Events support from Experimental to mainline, enabled by default
2197 and removable for a build by defining DISABLE_EVENT.
2199 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2201 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2202 cached by the daemon.
2204 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2205 by defining LOOKUP_REDIS. The libhiredis library is required.
2207 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2208 keys are given for lookup.
2210 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2211 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2212 only supported when built with TLS support. The PolarSSL SHA routines
2213 are still used when the TLS library is too old for convenient support.
2215 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2216 openssl_options), for security. OpenSSL forces this from version 1.1.0
2217 server-side so match that on older versions.
2219 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2220 allocation for $value could be released as the expansion processing
2221 concluded, but leaving the global pointer active for it.
2223 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2224 and to use the domains and local_parts ACL conditions.
2226 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2227 incorrectly not doubled on cutthrough transmission, hence seen as a
2228 body-termination at the receiving system - resulting in truncated mails.
2229 Commonly the sender saw a TCP-level error, and retransmitted the message
2230 via the normal store-and-forward channel. This could result in duplicates
2231 received - but deduplicating mailstores were liable to retain only the
2232 initial truncated version.
2234 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2236 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2238 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2239 we're in there, support oversigning also; bug 1309.
2241 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2243 HS/04 Add support for keep_environment and add_environment options.
2245 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2246 either intentional arithmetic overflow during PRNG, or testing config-
2249 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2250 delivery resulted in actual delivery. Cancel cutthrough before DATA
2253 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2254 rejects a recipient: pass the reject to the originator.
2256 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2257 Many were false-positives and ignorable, but it's worth fixing the
2260 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2261 for the new environment-manipulation done at startup. Move the routines
2262 from being local to tls.c to being global via the os.c file.
2264 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2265 an extract embedded as result-arg for a map, the first arg for extract
2266 is unavailable so we cannot tell if this is a numbered or keyed
2267 extraction. Accept either.
2273 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2276 JH/02 The smtp transport option "multi_domain" is now expanded.
2278 JH/03 The smtp transport now requests PRDR by default, if the server offers
2281 JH/04 Certificate name checking on server certificates, when exim is a client,
2282 is now done by default. The transport option tls_verify_cert_hostnames
2283 can be used to disable this per-host. The build option
2284 EXPERIMENTAL_CERTNAMES is withdrawn.
2286 JH/05 The value of the tls_verify_certificates smtp transport and main options
2287 default to the word "system" to access the system default CA bundle.
2288 For GnuTLS, only version 3.0.20 or later.
2290 JH/06 Verification of the server certificate for a TLS connection is now tried
2291 (but not required) by default. The verification status is now logged by
2292 default, for both outbound TLS and client-certificate supplying inbound
2295 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2298 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2299 Status Notification (bounce) messages are now MIME format per RFC 3464.
2300 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2301 under the control of the dsn_advertise_hosts option, and routers may
2302 have a dsn_lasthop option.
2304 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2305 default, modifiable by a malware= option. The list separator for
2306 the options can now be changed in the usual way. Bug 68.
2308 JH/10 The smtp_receive_timeout main option is now expanded before use.
2310 JH/11 The incoming_interface log option now also enables logging of the
2311 local interface on delivery outgoing connections.
2313 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2314 if the interface and destination host and port all match.
2316 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2319 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2320 Patch from Andrew Lewis.
2322 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2323 now supports optional time-restrictions, weighting, and priority
2324 modifiers per server. Patch originally by <rommer@active.by>.
2326 JH/16 The spamd_address main option now supports a mixed list of local
2327 and remote servers. Remote servers can be IPv6 addresses, and
2328 specify a port-range.
2330 JH/17 Bug 68: The spamd_address main option now supports an optional
2331 timeout value per server.
2333 JH/18 Bug 1581: Router and transport options headers_add/remove can
2334 now have the list separator specified.
2336 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2339 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2342 JH/21 Support for the A6 type of dns record is withdrawn.
2344 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2345 rather than the verbs used.
2347 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2348 from 255 to 1024 chars.
2350 JH/24 Verification callouts now attempt to use TLS by default.
2352 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2353 are generic router options now. The defaults didn't change.
2355 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2356 Original patch from Alexander Shikoff, worked over by JH.
2358 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2359 files if ZCAT_COMMAND is not executable.
2361 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2363 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2365 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2366 Normally benign, it bites when the pair was led to by a CNAME;
2367 modern usage is to not canonicalize the domain to a CNAME target
2368 (and we were inconsistent anyway for A-only vs AAAA+A).
2370 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2372 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2373 when evaluating $sender_host_dnssec.
2375 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2376 $sender_helo_dnssec variable.
2378 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2380 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2382 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2384 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2385 documented as working, but never had. Support all but $spam_report.
2387 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2388 added for tls authenticator.
2390 HS/03 Add perl_taintmode main config option
2396 TL/01 When running the test suite, the README says that variables such as
2397 no_msglog_check are global and can be placed anywhere in a specific
2398 test's script, however it was observed that placement needed to be near
2399 the beginning for it to behave that way. Changed the runtest perl
2400 script to read through the entire script once to detect and set these
2401 variables, reset to the beginning of the script, and then run through
2402 the script parsing/test process like normal.
2404 TL/02 The BSD's have an arc4random API. One of the functions to induce
2405 adding randomness was arc4random_stir(), but it has been removed in
2406 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2407 function when detected.
2409 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2410 cause callback expansion.
2412 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2413 syntax errors in an expansion can be treated as a string instead of
2414 logging or causing an error, due to the internal use of bool_lax
2415 instead of bool when processing it.
2417 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2418 server certificates when making smtp deliveries.
2420 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2422 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2424 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2426 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2427 Merged patch from Sebastian Wiedenroth.
2429 JH/05 Fix results-pipe from transport process. Several recipients, combined
2430 with certificate use, exposed issues where response data items split
2431 over buffer boundaries were not parsed properly. This eventually
2432 resulted in duplicates being sent. This issue only became common enough
2433 to notice due to the introduction of connection certificate information,
2434 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2436 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2437 size buffer was used, resulting in syntax errors when an expansion
2440 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2441 version 3.3.6 or later.
2443 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2444 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2445 both become "event_action", the variables become $event_name, $event_data
2446 and $event_defer_errno. There is a new variable $verify_mode, usable in
2447 routers, transports and related events. The tls:cert event is now also
2448 raised for inbound connections, if the main configuration event_action
2451 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2452 early OCSP support, but no stapling (appears to be less than 1.0.0).
2454 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2455 server certificate names available under the smtp transport option
2456 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2459 JH/10 Time-related extraction expansions from certificates now use the main
2460 option "timezone" setting for output formatting, and are consistent
2461 between OpenSSL and GnuTLS compilations. Bug 1541.
2463 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2464 encoded parameter in the incoming message. Bug 1558.
2466 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2467 include certificate info, eximon was claiming there were spoolfile
2470 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2472 JH/14 Log delivery-related information more consistently, using the sequence
2473 "H=<name> [<ip>]" wherever possible.
2475 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2476 are problematic for Debian distribution, omit them from the release
2479 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2481 JH/16 Fix string representation of time values on 64bit time_t architectures.
2484 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2485 output list separator was used.
2490 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2491 checkers that were complaining about end of non-void function with no
2494 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2495 This was a regression introduced in 4.83 by another bugfix.
2497 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2499 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2500 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2506 TF/01 Correctly close the server side of TLS when forking for delivery.
2508 When a message was received over SMTP with TLS, Exim failed to clear up
2509 the incoming connection properly after forking off the child process to
2510 deliver the message. In some situations the subsequent outgoing
2511 delivery connection happened to have the same fd number as the incoming
2512 connection previously had. Exim would try to use TLS and fail, logging
2513 a "Bad file descriptor" error.
2515 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2516 utilities have not been installed.
2518 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2519 temporary space as the ACL may create new global variables.
2521 TL/01 LDAP support uses per connection or global context settings, depending
2522 upon the detected version of the libraries at build time.
2524 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2525 to extract and use the src ip:port in logging and expansions as if it
2526 were a direct connection from the outside internet. PPv2 support was
2527 updated based on HAProxy spec change in May 2014.
2529 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2531 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2532 Properly escape header and check for NULL return.
2534 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2537 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2539 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2540 characters in header names, implemented as a verify condition.
2541 Contributed by Michael Fischer v. Mollard.
2543 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2544 results permerror and temperror. Previous values are deprecated but
2545 still accepted. In a future release, err_perm and err_temp will be
2546 completely removed, which will be a backward incompatibility if the
2547 ACL tests for either of these two old results. Patch contributed by
2548 user bes-internal on the mailing list.
2550 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2552 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2553 selectors, in both main and reject logs.
2555 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2558 JH/07 Add malware type "sock" for talking to simple daemon.
2560 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2562 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2563 routers/transports under cutthrough routing.
2565 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2566 numbers. Touch up "bool" conditional to keep the same definition.
2568 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2570 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2572 JH/12 Expand items in router/transport headers_add or headers_remove lists
2573 individually rather than the list as a whole. Bug 1452.
2575 Required for reasonable handling of multiple headers_ options when
2576 they may be empty; requires that headers_remove items with embedded
2577 colons must have them doubled (or the list-separator changed).
2579 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2580 view the policy declared in the DMARC record. Currently, $dmarc_status
2581 is a combined value of both the record presence and the result of the
2584 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2586 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2587 dnslookup router and the smtp transport (applying to the forward
2590 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2591 of ldap servers used for a specific lookup. Patch provided by Heiko
2594 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2595 New variable $lookup_dnssec_authenticated for observability.
2597 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2598 Patch submitted by Lars Timman.
2600 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2602 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2603 Requires trusted mode and valid format message id, aborts otherwise.
2604 Patch contributed by Heiko Schlichting.
2606 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2607 certextract with support for various fields. Bug 1358.
2609 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2610 is requested by default, modifiable by smtp transport option
2613 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2614 operate on certificate variables to give certificate fingerprints
2615 Also new ${sha256:cert_variable}.
2617 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2619 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2620 Christian Aistleitner.
2622 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2624 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2625 file. Patch from Wolfgang Breyha.
2627 JH/25 Expand the coverage of the delivery $host and $host_address to
2628 client authenticators run in verify callout. Bug 1476.
2630 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2631 align with daemon_smtp_ports. Bug 72.
2633 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2634 support and error reporting did not work properly.
2636 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2637 and is readable. Patch from Andrew Colin Kissa.
2639 TL/14 Enhance documentation of ${run expansion and how it parses the
2640 commandline after expansion, particularly in the case when an
2641 unquoted variable expansion results in an empty value.
2643 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2645 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2646 by renaming away from it. Addresses GH issue 15, reported by
2649 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2651 TL/15 SECURITY: prevent double expansion in math comparison functions
2652 (can expand unsanitized data). Not remotely exploitable.
2659 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2661 PP/02 Make -n do something, by making it not do something.
2662 When combined with -bP, the name of an option is not output.
2664 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2667 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2668 $sender_host_name and config options to manage this, and basic check
2671 PP/05 DSCP support for outbound connections and control modifier for inbound.
2673 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2674 (Only plugin which currently uses this is kerberos4, which nobody should
2675 be using, but we should make it available and other future plugins might
2676 conceivably use it, even though it would break NAT; stuff *should* be
2677 using channel bindings instead).
2679 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2680 name; added for Sendmail compatibility; requires admin caller.
2681 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2682 just ignore it); requires trusted caller.
2683 Also parse but ignore: -Ac -Am -X<logfile>
2686 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2688 TL/02 Add +smtp_confirmation as a default logging option.
2690 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2691 Patch by Magnus Holmgren from 2007-02-20.
2693 TL/04 Bugzilla 1281 - Spec typo.
2694 Bugzilla 1283 - Spec typo.
2695 Bugzilla 1290 - Spec grammar fixes.
2697 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2699 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2701 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2702 mailing list by by Dmitry Isaikin.
2704 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2706 JH/02 Support "G" suffix to numbers in ${if comparisons.
2708 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2710 NM/01 Bugzilla 1197 - Spec typo
2711 Bugzilla 1196 - Spec examples corrections
2713 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2715 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2716 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2719 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2720 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2722 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2723 "acl {{name}{arg}...}", and optional args on acl condition
2726 JH/05 Permit multiple router/transport headers_add/remove lines.
2728 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2730 JH/07 Avoid using a waiting database for a single-message-only transport.
2731 Performance patch from Paul Fisher. Bugzilla 1262.
2733 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2736 JH/09 Add $headers_added variable, with content from use of ACL modifier
2737 add_header (but not yet added to the message). Bugzilla 199.
2739 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2740 Pulled from Bugzilla 817 by Wolfgang Breyha.
2742 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2744 (nb: this is the same fix as in Exim 4.80.1)
2746 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2749 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2752 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2754 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2755 advertise SMTP AUTH mechanism to us, instead of a generic
2756 protocol violation error. Also, make Exim more robust to bad
2757 data from the Dovecot auth socket.
2759 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2761 When a queue runner is handling a message, Exim first routes the
2762 recipient addresses, during which it prunes them based on the retry
2763 hints database. After that it attempts to deliver the message to
2764 any remaining recipients. It then updates the hints database using
2767 So if a recipient address works intermittently, it can get repeatedly
2768 deferred at routing time. The retry hints record remains fresh so the
2769 address never reaches the final cutoff time.
2771 This is a fairly common occurrence when a user is bumping up against
2772 their storage quota. Exim had some logic in its local delivery code
2773 to deal with this. However it did not apply to per-recipient defers
2774 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2776 This change adds a proper retry rule check during routing so that the
2777 final cutoff time is checked against the message's age. We only do
2778 this check if there is an address retry record and there is not a
2779 domain retry record; this implies that previous attempts to handle
2780 the address had the retry_use_local_parts option turned on. We use
2781 this as an approximation for the destination being like a local
2782 delivery, as in LMTP.
2784 I suspect this new check makes the old local delivery cutoff check
2785 redundant, but I have not verified this so I left the code in place.
2787 TF/02 Correct gecos expansion when From: is a prefix of the username.
2789 Test 0254 submits a message to Exim with the header
2793 When I ran the test suite under the user fanf2, Exim expanded
2794 the header to contain my full name, whereas it should have added
2795 a Resent-Sender: header. It erroneously treats any prefix of the
2796 username as equal to the username.
2798 This change corrects that bug.
2800 GF/01 DCC debug and logging tidyup
2801 Error conditions log to paniclog rather than rejectlog.
2802 Debug lines prefixed by "DCC: " to remove any ambiguity.
2804 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2806 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2807 Bug spotted by Jeremy Harris; was flawed since initial commit.
2808 Would have resulted in OCSP responses post-SNI triggering an Exim
2809 NULL dereference and crash.
2811 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2813 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2814 Bug detection, analysis and fix by Samuel Thibault.
2815 Bugzilla 1331, Debian bug #698092.
2817 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2819 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2820 Server implementation by Todd Lyons, client by JH.
2821 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2822 config variable "prdr_enable" controls whether the server
2823 advertises the facility. If the client requests PRDR a new
2824 acl_data_smtp_prdr ACL is called once for each recipient, after
2825 the body content is received and before the acl_smtp_data ACL.
2826 The client is controlled by both of: a hosts_try_prdr option
2827 on the smtp transport, and the server advertisement.
2828 Default client logging of deliveries and rejections involving
2829 PRDR are flagged with the string "PRDR".
2831 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2832 fclose(). Diagnosis by Todd Lyons.
2834 PP/17 Update configure.default to handle IPv6 localhost better.
2835 Patch by Alain Williams (plus minor tweaks).
2838 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2839 This is now consistent with GnuTLS, and is now documented: the
2840 previous undocumented portable approach to treating the option as
2841 unset was to force an expansion failure. That still works, and
2842 an empty string is now equivalent.
2844 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2845 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2846 not performing validation itself.
2848 PP/20 Added force_command boolean option to pipe transport.
2849 Patch from Nick Koston, of cPanel Inc.
2851 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2854 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2856 PP/21 Fix eximon continuous updating with timestamped log-files.
2857 Broken in a format-string cleanup in 4.80, missed when I repaired the
2858 other false fix of the same issue.
2859 Report and fix from Heiko Schlichting.
2862 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2863 Report from Prashanth Katuri.
2865 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2866 It's SecureTransport, so affects any MacOS clients which use the
2867 system-integrated TLS libraries, including email clients.
2869 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2870 using a MIME ACL for non-SMTP local injection.
2871 Report and assistance in diagnosis by Warren Baker.
2873 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2875 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2877 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2878 last id that failed so it may be referenced in subsequent ACL's.
2880 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2883 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2884 ldap library initialization, allowing self-signed CA's to be
2885 used. Also properly sets require_cert option later in code by
2886 using NULL (global ldap config) instead of ldap handle (per
2887 session). Bug diagnosis and testing by alxgomz.
2889 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2890 the src/util/ subdirectory.
2892 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2893 renamed to Transport Post Delivery Action by Jeremy Harris, as
2896 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2897 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2898 redis_servers = needs to be configured which will be used by the redis
2899 lookup. Patch from Warren Baker, of The Packet Hub.
2901 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2903 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2904 hostname or reverse DNS when processing a host list. Used suggestions
2905 from multiple comments on this bug.
2907 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2909 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2910 Missed a few lines, added it to make the runtest require no keyboard
2913 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2914 contains upper case chars. Make router use caseful_local_part.
2916 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2917 support when GnuTLS has been built with p11-kit.
2923 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2925 This, or similar/improved, will also be change PP/11 of 4.82.
2931 PP/01 Handle short writes when writing local log-files.
2932 In practice, only affects FreeBSD (8 onwards).
2933 Bugzilla 1053, with thanks to Dmitry Isaikin.
2935 NM/01 Bugzilla 949 - Documentation tweak
2937 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2940 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2942 PP/02 Implemented gsasl authenticator.
2944 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2946 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2947 `pkg-config foo` for cflags/libs.
2949 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2950 with rest of GSASL and with heimdal_gssapi.
2952 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2953 `pkg-config foo` for cflags/libs for the TLS implementation.
2955 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2956 properties get this fed in as external SSF. A number of robustness
2957 and debugging improvements to the cyrus_sasl authenticator.
2959 PP/08 cyrus_sasl server now expands the server_realm option.
2961 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2962 Patch by Jeremy Harris.
2964 PP/10 Added dbmjz lookup type.
2966 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2968 PP/12 MAIL args handles TAB as well as SP, for better interop with
2969 non-compliant senders.
2970 Analysis and variant patch by Todd Lyons.
2972 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2973 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2974 Patch from Dirk Mueller <dmueller@suse.com>
2976 PP/13 tls_peerdn now print-escaped for spool files.
2977 Observed some $tls_peerdn in wild which contained \n, which resulted
2978 in spool file corruption.
2980 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2981 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2982 or write after TLS renegotiation, which otherwise led to messages
2985 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2986 as a tracking header (ie: a signed header comes before the signature).
2987 Patch from Wolfgang Breyha.
2989 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2990 comma-sep list; embedded commas doubled.
2992 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2994 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2996 Report and patch from Dmitry Banschikov.
2998 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2999 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3000 failures. We appear to now support TLS1.1+ with Exim.
3002 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3003 lets Exim select keys and certificates based upon TLS SNI from client.
3004 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3005 before an outbound SMTP session. New log_selector, +tls_sni.
3007 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3008 NULL dereference. Report and patch from Alun Jones.
3010 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3011 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3012 Not seeing resolver debug output on NetBSD, but suspect this is a
3013 resolver implementation change.
3015 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3016 Left warnings. Added "eximon gdb" invocation mode.
3018 PP/21 Defaulting "accept_8bitmime" to true, not false.
3020 PP/22 Added -bw for inetd wait mode support.
3022 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3023 locate the relevant includes and libraries. Made this the default.
3025 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3026 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3028 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3029 This may cause build issues on older platforms.
3031 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3032 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3033 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3034 Added SNI support via GnuTLS too.
3035 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3037 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3039 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3040 Applied second patch from Janne, implementing suggestion to default
3041 multiple-strings-in-record handling to match SPF spec.
3043 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3045 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3046 read-only, out of scope).
3047 Patch from Wolfgang Breyha, report from Stuart Northfield.
3049 PP/29 Fix three issues highlighted by clang analyser static analysis.
3050 Only crash-plausible issue would require the Cambridge-specific
3051 iplookup router and a misconfiguration.
3052 Report from Marcin Mirosław.
3054 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3056 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3057 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3058 As part of this, removing so much warning spew let me fix some minor
3059 real issues in debug logging.
3061 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3062 assignment on my part. Fixed.
3064 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3065 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3066 Janne Snabb (who went above and beyond: thank you).
3068 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3069 string otherwise requires a connection and a bunch more work and it's
3070 relatively easy to get wrong. Should also expose TLS library linkage
3073 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3074 64-bit ${eval} (JH/03).
3076 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3077 GNU libc to support some of the 64-bit stuff, should not lead to
3078 conflicts. Defined before os.h is pulled in, so if a given platform
3079 needs to override this, it can.
3081 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3082 protection layer was required, which is not implemented.
3083 Bugzilla 1254, patch from Wolfgang Breyha.
3085 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3086 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3087 tls_dhparam take prime identifiers. Also unbreak combination of
3088 OpenSSL+DH_params+TLSSNI.
3090 PP/39 Disable SSLv2 by default in OpenSSL support.
3096 PP/01 Solaris build fix for Oracle's LDAP libraries.
3097 Bugzilla 1109, patch from Stephen Usher.
3099 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3101 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3104 TF/02 Fix a couple more cases where we did not log the error message
3105 when unlink() failed. See also change 4.74-TF/03.
3107 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3108 lock up or crash if it happened to be inside a call to libc when it
3109 got a SIGUSR1 from exiwhat.
3111 The SIGUSR1 handler appends the current process status to the process
3112 log which is later printed by exiwhat. It used to use the general
3113 purpose logging code to do this, but several functions it calls are
3114 not safe for signals.
3116 The new output code in the SIGUSR1 handler is specific to the process
3117 log, and simple enough that it's easy to inspect for signal safety.
3118 Removing some special cases also simplifies the general logging code.
3119 Removing the spurious timestamps from the process log simplifies
3122 TF/04 Improved ratelimit ACL condition.
3124 The /noupdate option has been deprecated in favour of /readonly which
3125 has clearer semantics. The /leaky, /strict, and /readonly update modes
3126 are mutually exclusive. The update mode is no longer included in the
3127 database key; it just determines when the database is updated. (This
3128 means that when you upgrade Exim will forget old rate measurements.)
3130 Exim now checks that the per_* options are used with an update mode that
3131 makes sense for the current ACL. For example, when Exim is processing a
3132 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3133 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3134 must specify per_mail/readonly. If you omit the update mode it defaults to
3135 /leaky where that makes sense (as before) or /readonly where required.
3137 The /noupdate option is now undocumented but still supported for
3138 backwards compatibility. It is equivalent to /readonly except that in
3139 ACLs where /readonly is required you may specify /leaky/noupdate or
3140 /strict/noupdate which are treated the same as /readonly.
3142 A useful new feature is the /count= option. This is a generalization
3143 of the per_byte option, so that you can measure the throughput of other
3144 aggregate values. For example, the per_byte option is now equivalent
3145 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3147 The per_rcpt option has been generalized using the /count= mechanism
3148 (though it's more complicated than the per_byte equivalence). When it is
3149 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3150 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3151 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3152 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3153 Note that using per_rcpt with a non-readonly update mode in more than
3154 one ACL will cause the recipients to be double-counted. (The per_mail
3155 and per_byte options don't have this problem.)
3157 The handling of very low rates has changed slightly. If the computed rate
3158 is less than the event's count (usually one) then this event is the first
3159 after a long gap. In this case the rate is set to the same as this event's
3160 count, so that the first message of a spam run is counted properly.
3162 The major new feature is a mechanism for counting the rate of unique
3163 events. The new per_addr option counts the number of different
3164 recipients that someone has sent messages to in the last time period. It
3165 behaves like per_rcpt if all the recipient addresses are different, but
3166 duplicate recipient addresses do not increase the measured rate. Like
3167 the /count= option this is a general mechanism, so the per_addr option
3168 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3169 example, measure the rate that a client uses different sender addresses
3170 with the options per_mail/unique=$sender_address. There are further
3171 details in the main documentation.
3173 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3175 TF/06 Removed a few PCRE remnants.
3177 TF/07 Automatically extract Exim's version number from tags in the git
3178 repository when doing development or release builds.
3180 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3181 Bugzilla 879. Patch from Paul Fisher.
3183 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3184 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3187 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3189 PP/05 Variable $av_failed, true if the AV scanner deferred.
3190 Bugzilla 1078. Patch from John Horne.
3192 PP/06 Stop make process more reliably on build failure.
3193 Bugzilla 1087. Patch from Heiko Schlittermann.
3195 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3196 Bugzilla 1089. Patch from Heiko Schlittermann.
3198 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3199 Bugzilla 1131. Patch from Holger Weiß.
3201 PP/09 Handle IPv6 addresses with SPF.
3202 Bugzilla 860. Patch from Wolfgang Breyha.
3204 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3206 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3209 PP/11 match_* no longer expand right-hand-side by default.
3210 New compile-time build option, EXPAND_LISTMATCH_RHS.
3211 New expansion conditions, "inlist", "inlisti".
3213 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3215 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3217 PP/14 fix log_write() format string regression from TF/03.
3218 Bugzilla 1152. Patch from Dmitry Isaikin.
3224 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3226 PP/02 Harmonised TLS library version reporting; only show if debugging.
3227 Layout now matches that introduced for other libraries in 4.74 PP/03.
3229 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3231 PP/04 New "dns_use_edns0" global option.
3233 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3236 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3237 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3239 TK/01 Updated PolarSSL code to 0.14.2.
3240 Bugzilla 1097. Patch from Andreas Metzler.
3242 PP/07 Catch divide-by-zero in ${eval:...}.
3243 Fixes bugzilla 1102.
3245 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3248 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3249 format-string attack -- SECURITY: remote arbitrary code execution.
3251 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3252 time unintentionally subject to list matching rules, letting the header
3253 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3254 arbitrary string expansion). This allowed for information disclosure.
3256 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3257 INT_MIN/-1 -- value coerced to INT_MAX.
3263 NM/01 Workaround for PCRE version dependency in version reporting
3266 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3267 This fixes portability to compilers other than gcc, notably
3268 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3270 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3271 makefiles for portability to HP-UX and POSIX correctness.
3273 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3274 Also via indented variable definition in the Makefile.
3275 (Debugging by Oliver Heesakkers).
3277 PP/02 Restore caching of spamd results with expanded spamd_address.
3278 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3280 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3281 Improves build reliability. Fix from: Frank Elsner
3283 NM/02 Fix wide character breakage in the rfc2047 coding
3284 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3286 NM/03 Allow underscore in dnslist lookups
3287 Fixes bug 1026. Patch from Graeme Fowler
3289 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3290 Code patches from Adam Ciarcinski of NetBSD.
3292 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3295 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3296 is logged, to avoid truncation. Patch from John Horne.
3298 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3299 Patch from Jakob Hirsch.
3301 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3302 SQL string expansion failure details.
3303 Patch from Andrey Oktyabrski.
3305 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3306 Patch from Simon Arlott.
3308 PP/09 New lookups functionality failed to compile on old gcc which rejects
3309 extern declarations in function scope.
3310 Patch from Oliver Fleischmann
3312 PP/10 Use sig_atomic_t for flags set from signal handlers.
3313 Check getgroups() return and improve debugging.
3314 Fixed developed for diagnosis in bug 927 (which turned out to be
3317 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3318 Patch from Mark Zealey.
3320 PP/12 Bugzilla 1056: Improved spamd server selection.
3321 Patch from Mark Zealey.
3323 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3324 Based on patch from Heiko Schlittermann.
3326 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3327 Patch from Uwe Doering, sign-off by Michael Haardt.
3329 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3330 variable declaration deep within a block. Bug and patch from
3333 PP/15 lookups-Makefile IRIX compatibility coercion.
3335 PP/16 Make DISABLE_DKIM build knob functional.
3337 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3338 Patch by Simon Arlott
3340 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3341 variable argument macros. Our copy now differs from upstream.
3347 TF/01 Failure to get a lock on a hints database can have serious
3348 consequences so log it to the panic log.
3350 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3351 controlled using the smtp_confirmation log selector.
3353 TF/03 Include the error message when we fail to unlink a spool file.
3355 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3356 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3357 for maintaining out-of-tree patches for some time.
3359 PP/01 Bugzilla 139: Documentation and portability issues.
3360 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3361 Handle per-OS dynamic-module compilation flags.
3363 PP/02 Let /dev/null have normal permissions.
3364 The 4.73 fixes were a little too stringent and complained about the
3365 permissions on /dev/null. Exempt it from some checks.
3366 Reported by Andreas M. Kirchwitz.
3368 PP/03 Report version information for many libraries, including
3369 Exim version information for dynamically loaded libraries. Created
3370 version.h, now support a version extension string for distributors
3371 who patch heavily. Dynamic module ABI change.
3373 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3374 privilege escalation vulnerability whereby the Exim run-time user
3375 can cause root to append content of the attacker's choosing to
3378 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3381 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3382 If dropping privileges for untrusted macros, we disabled normal logging
3383 on the basis that it would fail; for the Exim run-time user, this is not
3384 the case, and it resulted in successful deliveries going unlogged.
3385 Fixed. Reported by Andreas Metzler.
3391 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3392 only prepend for the Resent-* case. Fixes regression introduced in
3393 Exim 4.70 by NM/22 for Bugzilla 607.
3395 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3396 increasing numbers of administrators be bitten by this.
3398 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3400 PP/03 Bugzilla 994: added openssl_options main configuration option.
3402 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3404 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3406 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3408 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3409 without a peer certificate, leading to a segfault because of an
3410 assumption that peers always have certificates. Be a little more
3411 paranoid. Problem reported by Martin Tscholak.
3413 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3414 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3415 NB: ClamAV planning to remove STREAM in "middle of 2010".
3416 CL also introduces -bmalware, various -d+acl logging additions and
3417 more caution in buffer sizes.
3419 PP/09 Implemented reverse_ip expansion operator.
3421 PP/10 Bugzilla 937: provide a "debug" ACL control.
3423 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3425 PP/12 Bugzilla 973: Implement --version.
3427 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3429 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3431 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3433 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3434 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3435 ignore trailing whitespace.
3437 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3439 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3442 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3443 Notification from Dr Andrew Aitchison.
3445 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3446 ExtendedDetectionInfo response format.
3447 Notification from John Horne.
3449 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3452 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3453 XSL and documented dependency on system catalogs, with examples of how
3456 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3459 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3460 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3461 configuration file which is writeable by the Exim user or group.
3463 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3464 of configuration files to cover files specified with the -C option if
3465 they are going to be used with root privileges, not just the default
3468 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3469 option (effectively making it always true).
3471 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3472 files to be used while preserving root privileges.
3474 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3475 that rogue child processes cannot use them.
3477 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3478 run-time user, instead of root.
3480 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3481 Exim run-time user without dropping privileges.
3483 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3484 result string, instead of calling string_vformat() twice with the same
3487 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3488 for other users. Others should always drop root privileges if they use
3489 -C on the command line, even for a whitelisted configure file.
3491 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3493 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3499 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3500 $data_path, and $header_path variables; fixed documentation bugs and
3503 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3504 exipick to access non-standard spools, including the "frozen" queue
3507 NM/01 Bugzilla 965: Support mysql stored procedures.
3508 Patch from Alain Williams
3510 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3512 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3513 Patch from Andreas Metzler
3515 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3516 Patch from Kirill Miazine
3518 NM/05 Bugzilla 671: Added umask to procmail example.
3520 JJ/03 installed exipick 20100323.0, fixing doc bug
3522 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3523 directory. Notification and patch from Dan Rosenberg.
3525 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3527 TK/02 Improve log output when DKIM signing operation fails.
3529 MH/01 Treat the transport option dkim_domain as a colon separated
3530 list, not as a single string, and sign the message with each element,
3531 omitting multiple occurences of the same signer.
3533 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3534 Bugzilla 985, 986. Patch by Simon Arlott
3536 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3537 Patch by Simon Arlott
3539 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3540 MBX locking. Notification from Dan Rosenberg.
3546 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3548 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3550 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3552 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3554 NM/04 Bugzilla 915: Flag broken perl installation during build.
3560 TK/01 Added patch by Johannes Berg that expands the main option
3561 "spamd_address" if it starts with a dollar sign.
3563 TK/02 Write list of recipients to X-Envelope-Sender header when building
3564 the mbox-format spool file for content scanning (suggested by Jakob
3567 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3568 (http://www.dcc-servers.net/) support via dccifd. Activated by
3569 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3571 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3572 by Mark Daniel Reidel <mr@df.eu>.
3574 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3575 When building exim an external PCRE library is now needed -
3576 PCRE is a system library on the majority of modern systems.
3577 See entry on PCRE_LIBS in EDITME file.
3579 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3580 conversation. Added nologin parameter to request.
3581 Patch contributed by Kirill Miazine.
3583 TF/01 Do not log submission mode rewrites if they do not change the address.
3585 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3587 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3588 log files in place. Contributed by Roberto Lima.
3590 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3592 TF/03 Bugzilla 615: When checking the local_parts router precondition
3593 after a local_part_suffix or local_part_prefix option, Exim now
3594 does not use the address's named list lookup cache, since this
3595 contains cached lookups for the whole local part.
3597 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3598 Robert Millan. Documentation is in experimental-spec.txt.
3600 TF/04 Bugzilla 668: Fix parallel build (make -j).
3602 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3604 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3605 Patch provided by Jan Srzednicki.
3607 TF/05 Leading white space used to be stripped from $spam_report which
3608 wrecked the formatting. Now it is preserved.
3610 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3611 that they are available at delivery time.
3613 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3615 TF/08 TLS error reporting now respects the incoming_interface and
3616 incoming_port log selectors.
3618 TF/09 Produce a more useful error message if an SMTP transport's hosts
3619 setting expands to an empty string.
3621 NM/06 Bugzilla 744: EXPN did not work under TLS.
3622 Patch provided by Phil Pennock.
3624 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3625 Patch provided by Richard Godbee.
3627 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3628 acl_smtp_notquit, added index entry.
3630 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3631 Patch provided by Eugene Bujak.
3633 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3634 accept(). Patch provided by Maxim Dounin.
3636 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3637 Patch provided by Phil Pennock.
3639 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3641 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3642 Patch provided by Brad "anomie" Jorsch.
3644 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3645 Patch provided by Dean Brooks.
3647 TK/05 Add native DKIM support (does not depend on external libraries).
3649 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3650 Patch provided by Graeme Fowler.
3652 NM/16 Bugzilla 851: Documentation example syntax fix.
3654 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3656 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3659 NM/19 Bugzilla 745: TLS version reporting.
3660 Patch provided by Phil Pennock.
3662 NM/20 Bugzilla 167: bool: condition support.
3663 Patch provided by Phil Pennock.
3665 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3666 clients. Patch provided by Phil Pennock.
3668 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3669 Patch provided by Brad "anomie" Jorsch.
3671 NM/23 Bugzilla 687: Fix misparses in eximstats.
3672 Patch provided by Heiko Schlittermann.
3674 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3675 Patch provided by Heiko Schlittermann.
3677 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3678 plus update to original patch.
3680 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3682 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3683 Patch provided by David Brownlee.
3685 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3687 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3689 NM/30 Bugzilla 888: TLS documentation bugfixes.
3691 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3693 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3694 Unlike the original bugzilla I have changed all shell scripts in src tree.
3696 NM/33 Bugzilla 898: Transport filter timeout fix.
3697 Patch by Todd Rinaldo.
3699 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3700 Patch by Serge Demonchaux.
3702 NM/35 Bugzilla 39: Base64 decode bug fixes.
3703 Patch by Jakob Hirsch.
3705 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3707 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3709 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3711 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3717 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3718 ALT-N's libdkim that I have put here:
3719 http://duncanthrax.net/exim-experimental/
3721 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3722 were called 'true' and it seems that C99 defines that as a reserved
3723 keyword to be used with 'bool' variable types. That means you could
3724 not include C99-style headers which use bools without triggering
3725 build errors in sieve.c.
3727 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3728 as mailq or other aliases. Changed the --help handling significantly
3729 to do whats expected. exim_usage() emits usage/help information.
3731 SC/01 Added the -bylocaldomain option to eximstats.
3733 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3735 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3737 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3743 PH/01 Another patch from the Sieve maintainer.
3745 PH/02 When an IPv6 address is converted to a string for single-key lookup
3746 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3747 dots are used instead of colons so that keys in lsearch files need not
3748 contain colons. This was done some time before quoting was made available
3749 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3750 (notated using the quote facility) so as to distinguish them from IPv4
3751 keys. This meant that lookups for IP addresses in host lists did not work
3752 for iplsearch lookups.
3754 This has been fixed by arranging for IPv6 addresses to be expressed with
3755 colons if the lookup type is iplsearch. This is not incompatible, because
3756 previously such lookups could never work.
3758 The situation is now rather anomalous, since one *can* have colons in
3759 ordinary lsearch keys. However, making the change in all cases is
3760 incompatible and would probably break a number of configurations.
3762 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3765 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3766 conversion specification without a maximum field width, thereby enabling
3767 a rogue spamd server to cause a buffer overflow. While nobody in their
3768 right mind would setup Exim to query an untrusted spamd server, an
3769 attacker that gains access to a server running spamd could potentially
3770 exploit this vulnerability to run arbitrary code as the Exim user.
3772 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3773 $primary_hostname instead of what libspf2 thinks the hosts name is.
3775 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3776 a directory entry by the name of the lookup key. Previously, if a
3777 symlink pointed to a non-existing file or a file in a directory that
3778 Exim lacked permissions to read, a lookup for a key matching that
3779 symlink would fail. Now it is enough that a matching directory entry
3780 exists, symlink or not. (Bugzilla 503.)
3782 PH/03 The body_linecount and body_zerocount variables are now exported in the
3785 PH/04 Added the $dnslist_matched variable.
3787 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3788 This means they are set thereafter only if the connection becomes
3791 PH/06 Added the client_condition to authenticators so that some can be skipped
3792 by clients under certain conditions.
3794 PH/07 The error message for a badly-placed control=no_multiline_responses left
3795 "_responses" off the end of the name.
3797 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3799 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3800 (without spaces) instead of just copying the configuration text.
3802 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3804 PH/11 Added $max_received_linelength.
3806 PH/12 Added +ignore_defer and +include_defer to host lists.
3808 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3809 way in which PCRE > 7.0 is built.
3811 PH/14 Implemented queue_only_load_latch.
3813 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3814 MAIL command. The effect was to mangle the value on 64-bit systems.
3816 PH/16 Another patch from the Sieve maintainer.
3818 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3820 PH/18 If a system quota error occurred while trying to create the file for
3821 a maildir delivery, the message "Mailbox is full" was not appended to the
3822 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3823 only to a quota excession during the actual writing of the file.
3825 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3826 characters?) which causes problems in log lines. The DN values are now
3827 passed through string_printing() before being added to log lines.
3829 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3830 and InterBase are left for another time.)
3832 PH/21 Added message_body_newlines option.
3834 PH/22 Guard against possible overflow in moan_check_errorcopy().
3836 PH/23 POSIX allows open() to be a macro; guard against that.
3838 PH/24 If the recipient of an error message contained an @ in the local part
3839 (suitably quoted, of course), incorrect values were put in $domain and
3840 $local_part during the evaluation of errors_copy.
3846 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3847 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3850 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3851 issue a MAIL command.
3853 PH/02 In an ACL statement such as
3855 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3857 if a client was not listed at all, or was listed with a value other than
3858 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3859 the condition was not true (as it should be), so access was not denied.
3860 The bug was that the ! inversion was incorrectly passed on to the second
3861 item. This has been fixed.
3863 PH/03 Added additional dnslists conditions == and =& which are different from
3864 = and & when the dns lookup returns more than one IP address.
3866 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3867 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3869 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3870 FSYNC, which compiles an option called disable_fsync that allows for
3871 bypassing fsync(). The documentation is heavily laced with warnings.
3873 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3875 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3876 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3877 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3878 including adding "make clean"; (3) Added -fPIC when compiling the test
3879 dynamically loaded module, to get rid of a warning.
3881 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3882 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3883 = 0s. The bug is otherwise harmless.
3885 PH/07 There was a bug in the dovecot authenticator such that the value of
3886 $auth1 could be overwritten, and so not correctly preserved, after a
3887 successful authentication. This usually meant that the value preserved by
3888 the server_setid option was incorrect.
3890 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3892 PH/09 Installed PCRE release 7.0.
3894 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3895 run for batched SMTP input. It is now run at the start of every message
3896 in the batch. While fixing this I discovered that the process information
3897 (output by running exiwhat) was not always getting set for -bs and -bS
3898 input. This is fixed, and it now also says "batched" for BSMTP.
3900 PH/11 Added control=no_pipelining.
3902 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3903 patch, slightly modified), and move the expansion of helo_data till after
3904 the connection is made in the smtp transport (so it can use these
3907 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3909 PH/14 Added log_selector = +pid.
3911 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3913 PH/16 Add ${if forany and ${if forall.
3915 PH/17 Added dsn_from option to vary the From: line in DSNs.
3917 PH/18 Flush SMTP output before performing a callout, unless control =
3918 no_callout_flush is set.
3920 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3921 was true (the default) a successful delivery failed to delete the retry
3922 item, thus causing premature timeout of the address. The bug is now
3925 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3927 PH/21 Long custom messages for fakedefer and fakereject are now split up
3928 into multiline responses in the same way that messages for "deny" and
3929 other ACL rejections are.
3931 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3932 with slight modification.
3934 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3935 draft, changing the syntax and factoring some duplicate code".
3937 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3938 for deliveries of the second and subsequent messages over the same SMTP
3941 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3942 ${reduce, with only minor "tidies".
3944 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3946 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3947 expansion side effects.
3949 PH/27 When a message times out after an over-quota error from an Exim-imposed
3950 quota, the bounce message says "mailbox is full". This message was not
3951 being given when it was a system quota that was exceeded. It now should
3954 MH/03 Made $recipients available in local_scan(). local_scan() already has
3955 better access to the recipient list through recipients_list[], but
3956 $recipients can be useful in postmaster-provided expansion strings.
3958 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3959 in the case of a MAIL command with additional options following the
3960 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3961 were accidentally chopped off.
3963 PH/29 SMTP synchronization checks are implemented when a command is read -
3964 there is a check that no more input is waiting when there shouldn't be
3965 any. However, for some commands, a delay in an ACL can mean that it is
3966 some time before the response is written. In this time, more input might
3967 arrive, invalidly. So now there are extra checks after an ACL has run for
3968 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3969 pipelining has not been advertised.
3971 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3973 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3974 returned in the SMTP response when smtp_return_error_details was false.
3975 This has been fixed.
3977 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3978 instead of the C I/O that was originally supplied, because problems were
3979 reported on Solaris.
3981 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3982 Exim which did not show up earlier: it was assuming that a call to
3983 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3984 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3985 was a macro that became an assignment, so it seemed to work. This has
3986 changed to a proper function call with a void return, hence the compile
3987 error. Exim's code has been fixed.
3989 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3992 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3994 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3996 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3997 because it turns out to be too big or there is a timeout) there is no
3998 "Completed" line in the log. When some messages of this type were
3999 selected by exigrep, they were listed as "not completed". Others were
4000 picked up by some special patterns. I have improved the selection
4001 criteria to be more general.
4003 PH/38 The host_find_failed option in the manualroute router can now be set
4004 to "ignore", to completely ignore a host whose IP address cannot be
4005 found. If all hosts are ignored, the behaviour is controlled by the new
4006 host_all_ignored option.
4008 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4009 homing or because of multiple MX records with /mx) generated more than
4010 one IP address, and the following item turned out to be the local host,
4011 all the secondary addresses of the first item were incorrectly removed
4012 from the list, along with the local host and any following hosts (which
4013 is what is supposed to happen).
4015 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4016 whoever called Exim into the -H file. In the case of the daemon it was
4017 behaving confusingly. When first started, it used values for whoever
4018 started the daemon, but after a SIGHUP it used the Exim user (because it
4019 calls itself on a restart). I have changed the code so that it now always
4022 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4023 message are rejected with the same error (e.g. no authentication or bad
4024 sender address), and a DATA command is nevertheless sent (as can happen
4025 with PIPELINING or a stupid MUA), the error message that was given to the
4026 RCPT commands is included in the rejection of the DATA command. This is
4027 intended to be helpful for MUAs that show only the final error to their
4030 PH/42 Another patch from the Sieve maintainer.
4032 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4033 Eximstats - Fixed some broken HTML links and added missing column headers
4035 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4036 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4038 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4040 PH/43 Yet another patch from the Sieve maintainer.
4042 PH/44 I found a way to check for a TCP/IP connection going away before sending
4043 the response to the final '.' that terminates a message, but only in the
4044 case where the client has not sent further data following the '.'
4045 (unfortunately, this is allowed). However, in many cases there won't be
4046 any further data because there won't be any more messages to send. A call
4047 to select() can be used: if it shows that the input is "ready", there is
4048 either input waiting, or the socket has been closed. An attempt to read
4049 the next input character can distinguish the two cases. Previously, Exim
4050 would have sent an OK response which the client would never have see.
4051 This could lead to message repetition. This fix should cure that, at
4052 least in a lot of common cases.
4054 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4055 advertised in response to EHLO.
4061 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4062 fixed by 4.65/MH/01 (is this a record?) are fixed:
4064 (i) An empty string was always treated as zero by the numeric comparison
4065 operators. This behaviour has been restored.
4067 (ii) It is documented that the numeric comparison operators always treat
4068 their arguments as decimal numbers. This was broken in that numbers
4069 starting with 0 were being interpreted as octal.
4071 While fixing these problems I realized that there was another issue that
4072 hadn't been noticed. Values of message_size_limit (both the global option
4073 and the transport option) were treated as octal if they started with 0.
4074 The documentation was vague. These values are now always treated as
4075 decimal, and I will make that clear in the documentation.
4081 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4082 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4085 MH/01 Don't check that the operands of numeric comparison operators are
4086 integers when their expansion is in "skipping" mode (fixes bug
4087 introduced by 4.64-PH/07).
4089 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4090 child addresses, Exim now panics and dies. Previously, because the count
4091 is held in a short int, deliveries were likely to be lost. As such a
4092 large number of recipients for a single message is ridiculous
4093 (performance will be very, very poor), I have chosen to impose a limit
4094 rather than extend the field.
4100 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4101 leftover -K file (the existence of which was triggered by #402).
4102 While we were at it, introduced process PID as part of the -K
4103 filename. This should rule out race conditions when creating
4106 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4107 processing considerably. Previous code took too long for large mails,
4108 triggering a timeout which in turn triggers #401.
4110 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4111 in the DK code in transports.c. sendfile() is not really portable,
4112 hence the _LINUX specificness.
4114 TF/01 In the add_headers option to the mail command in an Exim filter,
4115 there was a bug that Exim would claim a syntax error in any
4116 header after the first one which had an odd number of characters
4119 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4120 callout verification, Exim cached a "reject" for the entire domain. This
4121 is correct for most verifications, but it is not correct for a recipient
4122 verification with use_sender or use_postmaster set, because in that case
4123 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4124 case of MAIL FROM:<> rejection from other early rejections (e.g.
4125 rejection of HELO). When verifying a recipient using a non-null MAIL
4126 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4127 Whatever the result of the callout, the value of the domain cache is
4128 left unchanged (for any other kind of callout, getting as far as trying
4129 RCPT means that the domain itself is ok).
4131 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4134 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4135 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4136 session, the daemon ignores ECONNECT errors and logs others; it now
4137 ignores EPIPE as well.
4139 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4140 (quoted-printable decoding).
4142 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4143 later the small subsequent patch to fix an introduced bug.
4145 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4147 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4149 PH/08 An error is now given if message_size_limit is specified negative.
4151 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4152 to be given (somewhat) arbitrary names.
4154 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4157 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4158 miscellaneous code fixes
4160 PH/10 Added the log_reject_target ACL modifier to specify where to log
4163 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4164 hostname. This is wrong, because it relates to the incoming message (and
4165 probably the interface on which it is arriving) and not to the outgoing
4166 callout (which could be using a different interface). This has been
4167 changed to use the value of the helo_data option from the smtp transport
4168 instead - this is what is used when a message is actually being sent. If
4169 there is no remote transport (possible with a router that sets up host
4170 addresses), $smtp_active_hostname is used.
4172 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4173 tweaks were necessary in order to get it to work (see also 21 below):
4174 (a) The code assumed that strncpy() returns a negative number on buffer
4175 overflow, which isn't the case. Replaced with Exim's string_format()
4177 (b) There were several signed/unsigned issues. I just did the minimum
4178 hacking in of casts. There is scope for a larger refactoring.
4179 (c) The code used strcasecmp() which is not a standard C function.
4180 Replaced with Exim's strcmpic() function.
4181 (d) The code set only $1; it now sets $auth1 as well.
4182 (e) A simple test gave the error "authentication client didn't specify
4183 service in request". It would seem that Dovecot has changed its
4184 interface. Fortunately there's a specification; I followed it and
4185 changed what the client sends and it appears to be working now.
4187 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4190 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4191 address A is aliased to B and C, where B exists and C does not. Without
4192 -v the output is "A verified" because verification stops after a
4193 successful redirection if more than one address is generated. However,
4194 with -v the child addresses are also verified. Exim was outputting "A
4195 failed to verify" and then showing the successful verification for C,
4196 with its parentage. It now outputs "B failed to verify", showing B's
4197 parentage before showing the successful verification of C.
4199 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4200 look up a TXT record in a specific list after matching in a combined
4203 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4204 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4205 they consult the DNS. I had assumed they would set it the way they
4206 wanted; and indeed my experiments on Linux seem to show that in some
4207 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4208 To be on the safe side, however, I have now made the interface to
4209 host_find_byname() similar to host_find_bydns(), with an argument
4210 containing the DNS resolver options. The host_find_byname() function now
4211 sets these options at its start, just as host_find_bydns() does. The smtp
4212 transport options dns_qualify_single and dns_search_parents are passed to
4213 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4214 of host_find_byname() use the default settings of RES_DEFNAMES
4215 (qualify_single) but not RES_DNSRCH (search_parents).
4217 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4218 spool_read_header() do less string testing, by means of a preliminary
4219 switch on the second character of optional "-foo" lines. (This is
4220 overdue, caused by the large number of possibilities that now exist.
4221 Originally there were few.) While I was there, I also converted the
4222 str(n)cmp tests so they don't re-test the leading "-" and the first
4223 character, in the hope this might squeeze out yet more improvement.
4225 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4226 flag allowing group syntax was set by the header_syntax check but not
4227 turned off, possible causing trouble later; (2) The flag was not being
4228 set at all for the header_verify test, causing "group"-style headers to
4229 be rejected. I have now set it in this case, and also caused header_
4230 verify to ignore an empty address taken from a group. While doing this, I
4231 came across some other cases where the code for allowing group syntax
4232 while scanning a header line wasn't quite right (mostly, not resetting
4233 the flag correctly in the right place). These bugs could have caused
4234 trouble for malformed header lines. I hope it is now all correct.
4236 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4237 with the "reply" argument non-NULL. The code, however (which originally
4238 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4239 but it didn't always do it. This confused somebody who was copying the
4240 code for some other use. I have removed all the tests.
4242 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4243 feature that was used to support insecure browsers during the U.S. crypto
4244 embargo. It requires special client support, and Exim is probably the
4245 only MTA that supported it -- and would never use it because real RSA is
4246 always available. This code has been removed, because it had the bad
4247 effect of slowing Exim down by computing (never used) parameters for the
4248 RSA_EXPORT functionality.
4250 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4251 authenticator to fail if there's a tab character in the incoming data
4252 (there should never be unless someone is messing about, as it's supposed
4253 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4254 if the connection is using TLS or if the remote IP is the same as the
4255 local IP, and the "valid-client-cert option" if a client certificate has
4258 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4259 authenticators. This can be used for authorization after authentication
4260 succeeds. (In the case of plaintext, it servers for both authentication
4263 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4264 if any retry times were supplied.
4266 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4267 connection, where there is no client IP address to check. In this
4268 situation, the verify now always succeeds.
4270 PH/25 Applied John Jetmore's -Mset patch.
4272 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4274 PH/27 In a string expansion for a processed (not raw) header when multiple
4275 headers of the same name were present, leading whitespace was being
4276 removed from all of them, but trailing whitespace was being removed only
4277 from the last one. Now trailing whitespace is removed from each header
4278 before concatenation. Completely empty headers in a concatenation (as
4279 before) are ignored.
4281 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4282 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4284 PH/29 [Removed. This was a change that I later backed out, and forgot to
4285 correct the ChangeLog entry (that I had efficiently created) before
4286 committing the later change.]
4288 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4289 address errors (4xx response to RCPT) over the same connection as other
4290 messages routed to the same hosts. Such deliveries are always "forced",
4291 so retry times are not inspected. This resulted in far too many retries
4292 for the affected addresses. The effect occurred only when there were more
4293 hosts than the hosts_max_try setting in the smtp transport when it had
4294 the 4xx errors. Those hosts that it had tried were not added to the list
4295 of hosts for which the message was waiting, so if all were tried, there
4296 was no problem. Two fixes have been applied:
4298 (i) If there are any address or message errors in an SMTP delivery, none
4299 of the hosts (tried or untried) are now added to the list of hosts
4300 for which the message is waiting, so the message should not be a
4301 candidate for sending over the same connection that was used for a
4302 successful delivery of some other message. This seems entirely
4303 reasonable: after all the message is NOT "waiting for some host".
4304 This is so "obvious" that I'm not sure why it wasn't done
4305 previously. Hope I haven't missed anything, but it can't do any
4306 harm, as the worst effect is to miss an optimization.
4308 (ii) If, despite (i), such a delivery is accidentally attempted, the
4309 routing retry time is respected, so at least it doesn't keep
4310 hammering the server.
4312 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4313 in ${readsocket because some servers need this prod.
4315 PH/32 Added some extra debug output when updating a wait-xxx database.
4317 PH/33 The hint "could be header name not terminated by colon", which has been
4318 given for certain expansion errors for a long time, was not being given
4319 for the ${if def:h_colon_omitted{... case.
4321 PH/34 The spec says: "With one important exception, whenever a domain list is
4322 being scanned, $domain contains the subject domain." There was at least
4323 one case where this was not true.
4325 PH/35 The error "getsockname() failed: connection reset by peer" was being
4326 written to the panic log as well as the main log, but it isn't really
4327 panic-worthy as it just means the connection died rather early on. I have
4328 removed the panic log writing for the ECONNRESET error when getsockname()
4331 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4332 runs only) independently of the message's sender address. This meant
4333 that, if the 4xx error was in fact related to the sender, a different
4334 message to the same recipient with a different sender could confuse
4335 things. In particular, this can happen when sending to a greylisting
4336 server, but other circumstances could also provoke similar problems.
4337 I have changed the default so that the retry time for these errors is now
4338 based a combination of the sender and recipient addresses. This change
4339 can be overridden by setting address_retry_include_sender=false in the
4342 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4343 remote server are returned as part of bounce messages. This was not
4344 happening for LMTP over a pipe (the lmtp transport), but now it is the
4345 same for both kinds of LMTP.
4347 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4348 in header lines that were in fact CNAMEs. This is no longer the case.
4350 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4351 and queue runs started by the daemon processed all messages. This has
4352 been fixed so that -R and -S can now usefully be given with -q<time>.
4354 PH/40 Import PCRE release 6.7 (fixes some bugs).
4356 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4358 PH/42 Give an error if -q is specified more than once.
4360 PH/43 Renamed the variables $interface_address and $interface_port as
4361 $received_ip_address and $received_port, to make it clear that these
4362 values apply to message reception, and not to the outgoing interface when
4363 a message is delivered. (The old names remain recognized, of course.)
4365 PH/44 There was no timeout on the connect() call when using a Unix domain
4366 socket in the ${readsocket expansion. There now is.
4368 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4369 be meaningful with "accept".
4371 SC/01 Eximstats V1.43
4372 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4374 SC/02 Eximstats V1.44
4375 Use a glob alias rather than an array ref in the generated
4376 parser. This improves both readability and performance.
4378 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4379 Collect SpamAssassin and rejection statistics.
4380 Don't display local sender or destination tables unless
4381 there is data to show.
4382 Added average volumes into the top table text output.
4384 SC/04 Eximstats V1.46
4385 Collect data on the number of addresses (recipients)
4386 as well as the number of messages.
4388 SC/05 Eximstats V1.47
4389 Added 'Message too big' to the list of mail rejection
4390 reasons (thanks to Marco Gaiarin).
4392 SC/06 Eximstats V1.48
4393 Mainlog lines which have GMT offsets and are too short to
4394 have a flag are now skipped.
4396 SC/07 Eximstats V1.49 (Alain Williams)
4397 Added the -emptyok flag.
4399 SC/08 Eximstats V1.50
4400 Fixes for obtaining the IP address from reject messages.
4402 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4403 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4404 whitespace changes from 4.64-PH/27
4406 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4409 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4410 are found, allow negative numbers in numeric criteria)
4412 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4414 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4415 to match changes made in 4.64-PH/43
4417 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4419 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4420 "domain" is actually an IP address, and force a failure. This locks out
4421 those revolvers/nameservers that support "A-for-A" lookups, in
4422 contravention of the specifications.
4424 PH/48 When a host name was looked up from an IP address, and the subsequent
4425 forward lookup of the name timed out, the host name was left in
4426 $sender_host_name, contrary to the specification.
4428 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4429 restricted to single-key lookups, Exim was not diagnosing an error if
4430 * or *@ was used with a query-style lookup.
4432 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4434 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4435 long ago, but noone interested enough thought of it. Let's just say that
4436 the "1.1" means that there are some new functions that weren't there at
4437 some point in the past.
4439 PH/51 Error processing for expansion failure of helo_data from an smtp
4440 transport during callout processing was broken.
4442 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4443 tested/used via the -bh/-bhc/-bs options.
4445 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4446 bug, fixed in subsequent PCRE releases).
4448 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4449 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4451 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4457 SC/01 Use a glob alias rather than an array ref in eximstats generated
4458 parser. This improves both readability and performance.
4460 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4461 Don't display local sender or destination tables in eximstats unless
4462 there is data to show.
4463 Added average volumes into the eximstats top table text output.
4465 SC/03 Collect data on the number of addresses (recipients) as well
4466 as the number of messages in eximstats.
4468 TF/01 Correct an error in the documentation for the redirect router. Exim
4469 does (usually) call initgroups() when daemonizing.
4471 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4472 with consistent privilege compared to when running as a daemon.
4474 TF/03 Note in the spec that $authenticated_id is not set for local
4475 submissions from trusted users.
4477 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4478 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4480 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4481 by adding some example configuration directives to the default
4482 configuration file. A little bit of work is required to uncomment the
4483 directives and define how usernames and passwords are checked, but
4484 there is now a framework to start from.
4486 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4487 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4488 without this. I don't know how relevant this is to other LDAP libraries.
4490 PH/02 Add the verb name to the "unknown ACL verb" error.
4492 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4494 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4496 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4497 directory not expanded when it should be if an expanded home directory
4498 was set for the address (which is overridden by the transport).
4500 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4503 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4504 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4505 because it is too late at that time, and has no effect.
4507 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4508 security issue with \' (bugzilla #107). I could not use the
4509 PQescapeStringConn() function, because it needs a PGconn value as one of
4512 PH/08 When testing addresses using -bt, indicate those final addresses that
4513 are duplicates that would not cause an additional delivery. At least one
4514 person was confused, thinking that -bt output corresponded to deliveries.
4515 (Suppressing duplicates isn't a good idea as you lose the information
4516 about possibly different redirections that led to the duplicates.)
4518 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4519 systems where poll() doesn't work, in particular OS X.
4521 PH/10 Added more information to debugging output for retry time not reached.
4523 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4524 operations in malware.c.
4526 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4529 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4530 syslog_duplication set false, log lines that would normally be written
4531 both the the main log and to the reject log were not written to syslog at
4534 PH/14 In the default configuration, change the use of "message" in ACL warn
4535 statements to "add_header".
4537 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4538 not followed by a command (e.g. "seen endif").
4540 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4541 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4544 PH/17 Added extra conditions to the default value of delay_warning_condition
4548 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4549 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4550 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4553 The Auto-Submitted: and various List- headers are standardised, whereas I
4554 don't think Precedence: ever was.
4556 PH/18 Refactored debugging code in route_finduser() to show more information,
4557 in particular, the error code if getpwnam() issues one.
4559 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4560 This is apparently needed in addition to the PH/07 change above to avoid
4561 any possible encoding problems.
4563 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4564 but not after initializing Perl.
4566 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4567 output them only if debugging. By default they are written stderr,
4568 apparently, which is not desirable.
4570 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4573 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4576 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4578 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4579 authenticated or an ident call has been made. Suppress the default
4580 values for $authenticated_id and $authenticated_sender (but permit -oMai
4581 and -oMas) when testing with -bh.
4583 PH/24 Re-jigged the order of the tests in the default configuration so that the
4584 tests for valid domains and recipients precede the DNS black list and CSA
4585 tests, on the grounds that those ones are more expensive.
4587 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4588 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4589 This bug exists in every version of Exim that I still have, right back to
4592 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4593 However, an attempt to turn on case-sensitivity in a regex key by
4594 including (?-i) didn't work because the subject string was already
4595 lowercased, and the effects were non-intuitive. It turns out that a
4596 one-line patch can be used to allow (?-i) to work as expected.
4602 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4603 other effects) broke the use of negated acl sub-conditions.
4605 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4608 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4609 "Deny" causes Exim to reject the incoming connection with a 554 error.
4610 Unfortunately, if there is a major crisis, such as a disk failure,
4611 tcp-wrappers gives "deny", whereas what one would like would be some
4612 kind of temporary error. A kludge has been added to help with this.
4613 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4614 554 error is used if errno is still zero or contains ENOENT (which occurs
4615 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4618 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4620 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4621 errors. Otherwise a message that provokes a temporary error (when other
4622 messages do not) can cause a whole host to time out.
4624 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4625 addresses were routed directly to files or pipes from a redirect router.
4626 File deliveries just didn't batch; pipe deliveries might have suffered
4629 PH/06 A failure to get a lock for a hints database would erroneously always say
4630 "Failed to get write lock", even when it was really a read lock.
4632 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4633 of 0600. This has been changed to use the value of the lockfile_mode
4634 option (which defaults to 0600).
4636 PH/08 Applied small patch from the Sieve maintainer.
4638 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4639 folder from quota calculations, a direct delivery into this folder messed
4640 up the contents of the maildirsize file. This was because the regex was
4641 used only to exclude .Trash (or whatever) when the size of the mailbox
4642 was calculated. There was no check that a delivery was happening into an
4643 excluded directory. This bug has been fixed by ignoring all quota
4644 processing for deliveries into excluded directories.
4646 PH/10 Added the maildirfolder_create_regex option to appendfile.
4652 PH/01 The code for finding all the local interface addresses on a FreeBSD
4653 system running IPv6 was broken. This may well have applied to all BSD
4654 systems, as well as to others that have similar system calls. The broken
4655 code found IPv4 interfaces correctly, but gave incorrect values for the
4656 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4657 that it would not match correctly against @[] and not recognize the IPv6
4660 PH/02 The ipliteral router was not recognizing addresses of the form user@
4661 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4663 PH/03 Added disable_ipv6.
4665 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4666 decoded form, because it is most often used to construct To: headers
4667 lines in autoreplies, and the decoded form may well be syntactically
4668 invalid. However, $reply_address has leading white space removed, and all
4669 newlines turned into spaces so that the autoreply transport does not
4672 PH/05 If group was specified without a user on a router, and no group or user
4673 was specified on a transport, the group from the router was ignored.
4675 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4676 for visible compile-time settings that can be used to change these
4677 numbers, for those that want even more. Backwards compatibility with old
4678 spool files has been maintained. However, going back to a previous Exim
4679 release will lost any variables that are in spool files.
4681 PH/07 Two small changes when running in the test harness: increase delay when
4682 passing a TCP/IP connection to a new process, in case the original
4683 process has to generate a bounce, and remove special handling of
4684 127.0.0.2 (sic), which is no longer necessary.
4686 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4687 be the same on different OS.
4689 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4692 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4693 whether --show-vars was specified or not
4695 JJ/02 exipick: Added support for new ACL variable spool format introduced
4698 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4699 syntactically invalid From: or Reply-to: line, and a filter used this to
4700 generate an autoreply, and therefore failed to obtain an address for the
4701 autoreply, Exim could try to deliver to a non-existent relative file
4702 name, causing unrelated and misleading errors. What now happens is that
4703 it logs this as a hard delivery error, but does not attempt to create a
4706 PH/11 The exinext utility has a -C option for testing purposes, but although
4707 the given file was scanned by exinext itself; it wasn't being passed on
4708 when Exim was called.
4710 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4711 an end-of-file indication when reading a command response.
4713 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4714 compiled. In many other places in Exim, IPv6 addresses are always
4715 recognized, so I have changed this. It also means that IPv4 domain
4716 literals of the form [IPV4:n.n.n.n] are now always recognized.
4718 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4719 used if the router is not running as root, for example, when verifying at
4720 ACL time, or when using -bh. The debugging output from this situation was
4721 non-existent - all you got was a failure to exec. I have made two
4724 (a) Failures to set uid/gid, the current directory, or a process leader
4725 in a subprocess such as that created by queryprogram now generate
4726 suitable debugging output when -d is set.
4728 (b) The queryprogram router detects when it is not running as root,
4729 outputs suitable debugging information if -d is set, and then runs
4730 the subprocess without attempting to change uid/gid.
4732 PH/15 Minor change to Makefile for building test_host (undocumented testing
4735 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4736 additional section of a DNS packet that returns MX or SRV records.
4737 Instead, it always explicitly searches for A/AAAA records. This avoids
4738 major problems that occur when a DNS server includes only records of one
4739 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4740 fixed another bug: if SRV records were looked up and the corresponding
4741 address records were *not* found in the additional section, the port
4742 values from the SRV records were lost.
4744 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4745 using the correct key (the original address) when searching the retry
4746 rules in order to find which one to use for generating the retry hint.
4748 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4749 adding the default one. Similarly, if it contains a Reply-To: header, the
4750 errors_reply_to option, if set, is not used.
4752 PH/19 When calculating a retry time, Exim used to measure the "time since
4753 failure" by looking at the "first failed" field in the retry record. Now
4754 it does not use this if it is later than than the arrival time of the
4755 message. Instead it uses the arrival time. This makes for better
4756 behaviour in cases where some deliveries succeed, thus re-setting the
4757 "first failed" field. An example is a quota failure for a huge message
4758 when small messages continue to be delivered. Without this change, the
4759 "time since failure" will always be short, possible causing more frequent
4760 delivery attempts for the huge message than are intended.
4761 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4763 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4764 $1, $2, $3) because the numerical variables can be reset during some
4765 expansion items (e.g. "match"), thereby losing the authentication data.
4767 PH/21 Make -bV show the size of off_t variables so that the test suite can
4768 decide whether to run tests for quotas > 2G.
4770 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4771 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4772 filecount value is greater than 2G or if a quota value is greater than 2G
4773 on a system where the size of off_t is not greater than 4, a panic error
4776 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4777 never match. The debug and -bh output now contains an explicit error
4778 message indicating a malformed IPv4 address or mask.
4780 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4781 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4782 PH/23 above applies.
4784 PH/25 Do not write to syslog when running in the test harness. The only
4785 occasion when this arises is a failure to open the main or panic logs
4786 (for which there is an explicit test).
4788 PH/26 Added the /no_tell option to "control=freeze".
4790 PH/27 If a host name lookup failed very early in a connection, for example, if
4791 the IP address matched host_lookup and the reverse lookup yielded a name
4792 that did not have a forward lookup, an error message of the form "no IP
4793 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4794 could be logged. Now it outputs the IP address instead of "NULL".
4796 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4797 allows the sender and the authenticated sender to be set when
4798 submitting a message from within Exim. Since child_open_exim() is
4799 documented for local_scan(), the new function should be too.
4801 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4802 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4803 results in an empty string is now treated as unset.
4805 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4807 PH/31 Added sender_verify_fail logging option.
4809 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4810 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4811 tidied the source and removed it altogether.
4813 PH/33 When a queue run was abandoned because the load average was too high, a
4814 log line was always written; now it is written only if the queue_run log
4815 selector is set. In addition, the log line for abandonment now contains
4816 information about the queue run such as the pid. This is always present
4817 in "start" and "stop" lines but was omitted from the "abandon" line.
4819 PH/34 Omit spaces between a header name and the colon in the error message that
4820 is given when verify = headers_syntax fails (if there are lots of them,
4821 the message gets confusing).
4823 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4824 names, as there are now some PTR records that contain slashes. This check
4825 is only to protect against broken name servers that fall over on strange
4826 characters, so the fact that it applies to all lookups doesn't matter.
4828 PH/36 Now that the new test suite is complete, we can remove some of the
4829 special code in Exim that was needed for the old test suite. For example,
4830 sorting DNS records because real resolvers return them in an arbitrary
4831 order. The new test suite's fake resolver always returns records in the
4834 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4835 bounces) except when queue_only is set, to avoid logging races between
4836 the different processes.
4838 PH/38 Panic-die if .include specifies a non-absolute path.
4840 PH/39 A tweak to the "H" retry rule from its user.
4842 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4843 a label. They prevented compilation on older perls.
4845 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4846 a warning to be raised on newish perls.
4848 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4849 on queue. Changes to match documented behaviour of showing count of
4850 messages matching specified criteria.
4852 PH/40 Changed the default ident timeout from 30s to 5s.
4854 PH/41 Added support for the use of login_cap features, on those BSD systems
4855 that have them, for controlling the resources used by pipe deliveries.
4857 PH/42 The content-scanning code uses fopen() to create files in which to put
4858 message data. Previously it was not paying any attention to the mode of
4859 the files. Exim runs with umask(0) because the rest of the code creates
4860 files with open(), and sets the required mode explicitly. Thus, these
4861 files were ending up world-writeable. This was not a big issue, because,
4862 being within the spool directory, they were not world-accessible. I have
4863 created a function called modefopen, which takes an additional mode
4864 argument. It sets umask(777), creates the file, chmods it to the required
4865 mode, then resets the umask. All the relevant calls to fopen() in the
4866 content scanning code have been changed to use this function.
4868 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4869 to 24 hours. This avoids potential overflow problems when processing G
4870 and H retry rules. I suspect nobody ever tinkers with this value.
4872 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4874 PH/45 When the plaintext authenticator is running as a client, the server's
4875 challenges are checked to ensure they are valid base64 strings. By
4876 default, the authentication attempt is cancelled if an invalid string is
4877 received. Setting client_ignore_invalid_base64 true ignores these errors.
4878 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4879 they are received. Thus, the responses can be made to depend on the
4880 challenges. If an invalid string is ignored, an empty string is placed in
4883 PH/46 Messages that are created by the autoreply transport now contains a
4884 References: header, in accordance with RFCs 2822 and 3834.
4886 PH/47 Added authenticated_sender_force to the smtp transport.
4888 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4890 PH/49 Installed latest patch from the Sieve maintainer.
4892 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4893 was also set, the appendfile transport was unnecessarily scanning a
4894 directory of message files (e.g. for maildir delivery) to find the count
4895 of files (along with the size), even though it did not need this
4896 information. It now does the scan only if it needs to find either the
4897 size of the count of files.
4899 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4901 PH/52 Two bugs concerned with error handling when the smtp transport is
4904 (i) Exim was not creating retry information for temporary errors given
4905 for individual recipients after the DATA command when the smtp transport
4906 was used in LMTP mode. This meant that they could be retried too
4907 frequently, and not timed out correctly.
4909 (ii) Exim was setting the flag that allows error details to be returned
4910 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4911 recipients that were returned after the DATA command.
4913 PH/53 This is related to PH/52, but is more general: for any failing address,
4914 when detailed error information was permitted to be returned to the
4915 sender, but the error was temporary, then after the final timeout, only
4916 "retry timeout exceeded" was returned. Now it returns the full error as
4917 well as "retry timeout exceeded".
4919 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4920 do this, and (what is worse) MTAs that accept it.
4922 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4923 will now be deprecated.
4925 PH/56 New os.c-cygwin from the Cygwin maintainer.
4927 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4928 formats (previously only available in exim formats via -bpr, -bpru,
4929 and -bpra. Now also available in native and exiqgrep formats)
4931 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4932 with very large, slow to parse queues
4934 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4936 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4938 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4939 responses to authentication challenges, though it was showing the
4940 challenges; (ii) I've removed the CR characters from the debug output for
4943 PH/58 Allow for the insertion of a newline as well as a space when a string
4944 is turned into more than one encoded-word during RFC 2047 encoding. The
4945 Sieve code now uses this.
4947 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4948 data_4xx, lost_connection, tls_required.
4950 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4951 message was being sent as an SMTP response.
4953 PH/61 Add -l and -k options to exicyclog.
4955 PH/62 When verifying, if an address was redirected to one new address, so that
4956 verification continued, and the new address failed or deferred after
4957 having set something in $address_data, the value of $address_data was not
4958 passed back to the ACL. This was different to the case when no
4959 redirection occurred. The value is now passed back in both cases.
4961 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4962 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4963 use login_cap.h, so on its own it isn't the distinguishing feature. The
4964 new name refers directly to the setclassresources() function.
4966 PH/65 Added configuration files for NetBSD3.
4968 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4970 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4971 is preferred over IPv4.
4973 PH/68 The bounce_return_message and bounce_return_body options were not being
4974 honoured for bounces generated during the reception of non-SMTP messages.
4975 In particular, this applied to messages rejected by the ACL. This bug has
4976 been fixed. However, if bounce_return_message is true and bounce_return_
4977 body is false, the headers that are returned for a non-SMTP message
4978 include only those that have been read before the error was detected.
4979 (In the case of an ACL rejection, they have all been read.)
4981 PH/69 The HTML version of the specification is now built in a directory called
4982 spec_html instead of spec.html, because the latter looks like a path with
4983 a MIME-type, and this confuses some software.
4985 PH/70 Catch two compiler warnings in sieve.c.
4987 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4988 function verify_get_ident() calls ip_connect() to connect a socket, but
4989 if the "connect()" function timed out, ip_connect() used to close the
4990 socket. However, verify_get_ident() also closes the socket later, and in
4991 between Exim writes to the log, which may get opened at this point. When
4992 the socket was closed in ip_connect(), the log could get the same file
4993 descriptor number as the socket. This naturally causes chaos. The fix is
4994 not to close the socket in ip_connect(); the socket should be closed by
4995 the function that creates it. There was only one place in the code where
4996 this was missing, in the iplookup router, which I don't think anybody now
4997 uses, but I've fixed it anyway.
4999 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5000 well as to direct DNS lookups. Otherwise the handling of names in host
5001 lists is inconsistent and therefore confusing.
5007 PH/01 Two changes to the default runtime configuration:
5009 (1) Move the checks for relay_from_hosts and authenticated clients from
5010 after to before the (commented out) DNS black list checks.
5012 (2) Add control=submission to the relay_from_hosts and authenticated
5013 clients checks, on the grounds that messages accepted by these
5014 statements are most likely to be submissions.
5016 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5018 (1) Generate an error if the third argument for the ${prvs expansion is
5021 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5024 (3) Reset the variables that are obtained from the first argument of
5025 ${prvscheck and used in the second argument before leaving the code,
5026 because their memory is reclaimed, so using them afterwards may do
5029 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5030 one (it's much easier than Tom thought :-).
5032 (5) Because of (4), we can now allow for the use of $prvscheck_result
5033 inside the third argument.
5035 PH/03 For some reason, the default setting of PATH when running a command from
5036 a pipe transport was just "/usr/bin". I have changed it to
5039 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5040 anything to be listed in the output from -bV.
5042 PH/05 When a filter generated an autoreply, the entire To: header line was
5043 quoted in the delivery log line, like this:
5045 => >A.N.Other <ano@some.domain> <original@ddress> ...
5047 This has been changed so that it extracts the operative address. There
5048 may be more than one such address. If so, they are comma-separated, like
5051 => >ano@some.domain,ona@other.domain <original@ddress> ...
5053 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5054 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5055 not being looked up in the rDNS to get a host name, Exim was showing the
5056 IP address twice in Received: lines, even though the IP addresses were
5057 identical. For example:
5059 Received: from [1.2.3.4] (helo=[1.2.3.4])
5061 However, if the real host name was known, it was omitting the HELO data
5062 if it matched the actual IP address. This has been tidied up so that it
5063 doesn't show the same IP address twice.
5065 PH/07 When both +timestamp and +memory debugging was on, the value given by
5066 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5067 called by the debug printing, thereby overwriting the timestamp buffer.
5068 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5070 PH/08 When the original message was included in an autoreply transport, it
5071 always said "this is a copy of the message, including all the headers",
5072 even if body_only or headers_only was set. It now gives an appropriate
5075 PH/09 Applied a patch from the Sieve maintainer which:
5077 o fixes some comments
5078 o adds the (disabled) notify extension core
5079 o adds some debug output for the result of if/elsif tests
5080 o points to the current vacation draft in the documentation
5081 and documents the missing references header update
5085 o fixes a bug in processing the envelope test (when testing
5086 multiple envelope elements, the last element determined the
5089 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5090 Electronic Mail") by including:
5092 Auto-submitted: auto-generated
5094 in the messages that it generates (bounce messages and others, such as
5095 warnings). In the case of bounce messages for non-SMTP messages, there was
5096 also a typo: it was using "Auto_submitted" (underscore instead of
5097 hyphen). Since every message generated by Exim is necessarily in response
5098 to another message, thes have all been changed to:
5100 Auto-Submitted: auto-replied
5102 in accordance with these statements in the RFC:
5104 The auto-replied keyword:
5106 - SHOULD be used on messages sent in direct response to another
5107 message by an automatic process,
5109 - MUST NOT be used on manually-generated messages,
5111 - MAY be used on Delivery Status Notifications (DSNs) and Message
5112 Disposition Notifications (MDNs),
5114 - MUST NOT be used on messages generated by automatic or periodic
5115 processes, except for messages which are automatic responses to
5118 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5119 to the default Received: header definition.
5121 PH/12 Added log selector acl_warn_skipped (default on).
5123 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5124 variables because (a) they are in the wrong storage pool and (b) even if
5125 they were copied, it wouldn't work properly because of the caching.
5127 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5128 checking when decoding. Apparently there are clients that generate
5129 overlong encoded strings. Why am I not surprised?
5131 PH/15 If the first argument of "${if match_address" was not empty, but did not
5132 contain an "@" character, Exim crashed. Now it writes a panic log message
5133 and treats the condition as false.
5135 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5137 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5138 extension "envelope-auth". The code is finished and in agreement with
5139 other implementations, but there is no documentation so far and in fact,
5140 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5141 not changing the active code.
5143 Print executed "if" and "elsif" statements when debugging is used. This
5144 helps a great deal to understand what a filter does.
5146 Document more things not specified clearly in RFC3028. I had all this
5147 sorted out, when out of a sudden new issues came to my mind. Oops."
5149 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5152 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5153 canonical form (as documented). However, after a host name lookup from
5154 the IP address, check_host() was doing a simple string comparison with
5155 addresses acquired from the DNS when checking that the found name did
5156 have the original IP as one of its addresses. Since any found IPv6
5157 addresses are likely to be in abbreviated form, the comparison could
5158 fail. Luckily, there already exists a function for doing the comparison
5159 by converting both addresses to binary, so now that is used instead of
5160 the text comparison.
5162 PH/20 There was another similar case to PH/19, when a complete host name was
5163 given in a host list; looking up its IP address could give an abbreviated
5164 form, whereas the current host's name might or might not be abbreviated.
5165 The same fix has been applied.
5171 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5172 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5175 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5176 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5178 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5180 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5181 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5182 Also, for "Auto-Submitted", treat anything other than "no" as
5183 non-personal, in accordance with RFC 3834. (Previously it treated
5184 anything starting "auto-" as non-personal.)
5186 TF/01 The control=submission/name=... option had a problem with syntax
5187 errors if the name included a slash character. The /name= option
5188 now slurps the rest of the string, so it can include any characters
5189 but it must come last in the list of options (after /sender_retain
5192 PH/05 Some modifications to the interface to the fake nameserver for the new
5200 TK/01 Added the "success_on_redirect" address verification option. See
5201 NewStuff for rationale and an example.
5203 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5205 PH/02 Patch to exigrep to allow it to work on syslog lines.
5207 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5208 fread() to skip over the body file's header line, because in Cygwin the
5209 header line is locked and is inaccessible.
5211 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5212 co-exist for some time) to make it clear that it is the Exim ID that is
5213 referenced, not the Message-ID: header line.
5215 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5216 string_format() function, because snprintf() does not exist on all
5219 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5220 ${stat: expansion item.
5222 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5223 protocol synchronization error", to keep the pedants happy.
5225 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5226 well as for IRIX systems, when gcc is being used. See the host.c source
5229 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5231 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5234 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5235 transport and to the smtp transport in LMTP mode.
5237 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5239 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5240 run a filter in a subprocess. This could lead to confusion in subsequent
5241 lookups in the parent process. There should also be a search_tidyup() at
5242 the end of the subprocess.
5244 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5245 only if the host matched helo_try_verify_hosts, which caused the
5246 verification to occur when the EHLO/HELO command was issued. The ACL just
5247 tested the remembered result. Now, if a previous verification attempt has
5248 not happened, "verify = helo" does it there and then.
5250 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5252 TK/03 Fix log output including CR from clamd.
5254 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5255 exist provoked a memory error which could cause a segfault.
5257 PH/15 Installed PCRE 6.2
5259 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5261 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5262 of the problem. Specifically, suggested +O2 rather than +O1 for the
5265 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5267 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5268 record, the port was not correctly logged when the outgoing_port log
5269 selector was set (it logged the transort's default port).
5271 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5272 fallback_hosts, and "hosts" in the smtp transport.
5274 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5275 host errors such as "Connection refused".
5277 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5278 authentication with radiusclient 0.4.9:
5280 - Error returned from rc_read_config was caught wrongly
5281 - Username/password not passed on to radius server due to wrong length.
5283 The presumption is that some radiusclient API changes for 4.51/PH/17
5284 were not taken care of correctly. The code is still untested by me (my
5285 Linux distribution still has 0.3.2 of radiusclient), but it was
5286 contributed by a Radius user.
5288 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5289 expanding the "port" option of the smtp transport.
5291 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5292 while reading a MIME header. Thanks to Tom Hughes for a patch.
5294 PH/24 Include config.h inside local_scan.h so that configuration settings are
5297 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5298 that in an ACL for RCPT (for example), you can examine exactly what was
5301 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5302 commands, but it was not correctly comparing the address with the actual
5303 client host address. Thus, it would show the EHLO address in Received:
5304 header lines when this was not necessary.
5306 PH/27 Added the % operator to ${eval:}.
5308 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5309 it should be ignoring failures (because with -C, for example, it has lost
5310 privilege). It wasn't ignoring creation failures other than "already
5313 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5316 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5317 because an input line was too long, either on its own, or by virtue of
5318 too many continuations, the temporary file was not being removed, and the
5319 return code was incorrect.
5321 PH/31 Missing "BOOL" in function definition in filtertest.c.
5323 PH/32 Applied Sieve patches from the maintainer.
5325 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5327 PH/33 Added "verify = not_blind".
5329 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5330 Local/Makefile (with some defaults set). These are used in built scripts
5331 such as exicyclog, but they have never been used in the exim_install
5332 script (though there are many overriding facilities there). I have
5333 arranged that the exim_install script now takes note of these two
5336 PH/35 Installed configuration files for Dragonfly.
5338 PH/36 When a locally submitted message by a trusted user did not contain a
5339 From: header, and the sender address was obtained from -f or from an SMTP
5340 MAIL command, and the trusted user did not use -F to supply a sender
5341 name, $originator_name was incorrectly used when constructing a From:
5342 header. Furthermore, $originator_name was used for submission mode
5343 messages from external hosts without From: headers in a similar way,
5344 which is clearly wrong.
5346 PH/37 Added control=suppress_local_fixups.
5348 PH/38 When log_selector = +received_sender was set, and the addition of the
5349 sender made the log line's construction buffer exactly full, or one byte
5350 less than full, an overflow happened when the terminating "\n" was
5353 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5354 when the result of a list match is failure because a DNS lookup failed.
5356 PH/40 RM_COMMAND is now used in the building process.
5358 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5359 the "build-* directories that it finds.
5361 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5362 address in a domain literal was a prefix of an interface address.
5364 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5365 when verifying a sender address, unless rewrite_headers is false.
5367 PH/44 Wrote a long comment about why errors_to addresses are verified as
5368 recipients, not senders.
5370 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5371 the ratelimit ACL was added.
5373 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5375 PH/46 Added extra information about PostgreSQL errors to the error string.
5377 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5378 avoiding the need to install special zones in a real server. This is
5379 backwards compatible; if it can't find the fake resolver, it drops back.
5380 Thus, both old and new test suites can be run.
5382 TF/02 Added util/ratelimit.pl
5384 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5385 clock is set back in time.
5387 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5388 Candler <B.Candler@pobox.com>.
5390 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5391 OK for addresses that are the result of redirections.
5393 PH/48 A number of further additions for the benefit of the new test suite,
5394 including a fake gethostbyname() that interfaces to the fake DNS resolver
5397 TF/06 The fix for widen_domains has also been applied to qualify_single and
5398 search_parents which are the other dnslookup options that can cause
5401 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5404 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5406 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5407 service" if it encountered an MX record with an empty target hostname.
5408 The message is now "an MX or SRV record indicated no SMTP service".
5410 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5411 if the DNS of the sending site is misconfigured. This is quite a
5412 common situation. This change restores the behaviour of treating a
5413 helo verification defer as a failure.
5415 PH/51 If self=fail was set on a router, the bounce message did not include the
5416 actual error message.
5422 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5424 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5425 command itself ended in error, the underlying message about the transport
5426 filter timeout was being overwritten with the pipe command error. Now the
5427 underlying error message should be appended to the second error message.
5429 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5431 PH/02 Reduce the amount of output that "make" produces by default. Full output
5432 can still be requested.
5434 PH/03 The warning log line about a condition test deferring for a "warn" verb
5435 was being output only once per connection, rather than after each
5436 occurrence (because it was using the same function as for successful
5437 "warn" verbs). This seems wrong, so I have changed it.
5439 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5440 it should not have, which might have caused a crash in the right
5441 circumstances, but probably never did.
5443 PH/04 Installed a modified version of Tony Finch's patch to make submission
5444 mode fix the return path as well as the Sender: header line, and to
5445 add a /name= option so that you can make the user's friendly name appear
5448 TF/03 Added the control = fakedefer ACL modifier.
5450 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5451 Mark Lowes for thorough testing.
5453 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5455 TK/03 Merged latest SRS patch from Miles Wilton.
5457 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5458 with the definition in sysexits.h (which is #included earlier).
5459 Fortunately, Exim does not actually use EX_OK. The code used to try to
5460 preserve the sysexits.h value, by assuming that macro definitions were
5461 scanned for macro replacements. I have been disabused of this notion,
5462 so now the code just undefines EX_OK before #including unistd.h.
5464 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5465 in the smtp transport. When a block could not be written in a single
5466 write() function, the timeout was being re-applied to each part-write.
5467 This seems wrong - if the receiver was accepting one byte at a time it
5468 would take for ever. The timeout is now adjusted when this happens. It
5469 doesn't have to be particularly precise.
5471 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5472 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5474 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5475 without a domain if the check to <postmaster@domain> fails.
5477 SC/01 Eximstats: added -xls and the ability to specify output files
5478 (patch written by Frank Heydlauf).
5480 SC/02 Eximstats: use FileHandles for outputting results.
5482 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5484 SC/04 Eximstats: fixed display of large numbers with -nvr option
5486 SC/05 Eximstats: fixed merging of reports with empty tables.
5488 SC/06 Eximstats: added the -include_original_destination flag
5490 SC/07 Eximstats: removed tabs and trailing whitespace.
5492 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5494 TK/06 MBOX spool code: Add real "From " MBOX separator line
5495 so the .eml file is really in mbox format (even though
5496 most programs do not really care). Patch from Alex Miller.
5498 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5499 The latter is generated from $received_to and is only set if the
5500 message has one envelope recipient. SA can use these headers,
5501 obviously out-of-the-box. Patch from Alex Miller.
5503 PH/08 The ${def test on a variable was returning false if the variable's
5504 value was "0", contrary to what the specification has always said!
5505 The result should be true unless the variable is empty.
5507 PH/09 The syntax error of a character other than { following "${if
5508 def:variable_name" (after optional whitespace) was not being diagnosed.
5509 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5510 accidental colon was present, for example, could give incorrect results.
5512 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5513 result is used (not including appendfile, where other changes are about
5516 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5517 This involved changing a lot of size variables from int to off_t. It
5518 should work with maildirs and everything.
5520 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5521 spamd dying while we are connected to it.
5523 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5526 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5527 function for BDB 4.3.
5529 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5531 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5532 the sources for PCRE have been re-arranged and more files are now
5535 PH/15 The code I had for printing potentially long long variables in PH/11
5536 above was not the best (it lost precision). The length of off_t variables
5537 is now inspected at build time, and an appropriate printing format (%ld
5538 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5539 to be "long long int" or "long int". This is needed for the internal
5540 formatting function string_vformat().
5542 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5543 the configuration file to be ":syslog", then the script "guesses" where
5544 the logs files are, rather than using the compiled in default. In our
5545 case the guess is not the same as the compiled default, so the script
5546 suddenly stopped working when I started to use syslog. The patch checks
5547 to see if log_file_path is "". If so, it attempts to read it from exim
5548 with no configuration file to get the compiled in version, before it
5549 falls back to the previous guessing code."
5551 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5552 implementing BATV in an Exim configuration. See NewStuff for the gory
5555 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5556 Makefile that are specific to HP-UX.
5558 PH/18 If the "use_postmaster" option was set for a recipient callout together
5559 with the "random" option, the postmaster address was used as the MAIL
5560 FROM address for the random test, but not for the subsequent recipient
5561 test. It is now used for both.
5563 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5564 patch removes a few documentation additions to RFC 3028, because the
5565 latest draft now contains them. It adds the new en;ascii-case comparator
5566 and a new error check for 8bit text in MIME parts. Comparator and
5567 require names are now matched exactly. I enabled the subaddress
5568 extension, but it is not well tested yet (read: it works for me)."
5570 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5571 rework some of the code of TK/09 above to avoid the hardwired use of
5572 "%lld" and "long long". Replaced the call to snprintf() with a call to
5575 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5576 records point to non-existent hosts", "retry timeout exceeded", and
5577 "retry time not reached for any host after a long failure period".
5579 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5580 experimental DomainKeys support:
5582 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5583 (2) On an error such as an illegally used "control", the wrong name for
5584 the control was given.
5586 These problems did NOT occur unless DomainKeys support was compiled.
5588 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5590 PH/24 Added ${if match_ip condition.
5592 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5593 databases so that it will be absolutely obvious if a crash occurs in the
5594 DB library. This is a regular occurrence (often caused by mis-matched
5597 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5598 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5599 picked up on a user's system that detects such things. There doesn't seem
5600 to be a gcc warning option for this - only an attribute that has to be
5601 put on the function's prototype. It seems that in Fedora Core 4 they have
5602 set this on a number of new functions. No doubt there will be more in due
5605 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5606 specify a transport. However, if an address that was verified by such a
5607 router was the subject of a callout, Exim crashed because it tried to
5608 read the rcpt_include_affixes from the non-existent transport. Now it
5609 just assumes that the setting of that option is false. This bug was
5610 introduced by 4.51/PH/31.
5612 PH/28 Changed -d+all to exclude +memory, because that information is very
5613 rarely of interest, but it makes the output a lot bigger. People tend to
5614 do -d+all out of habit.
5616 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5617 code in os-type was giving problems when libc.so lives in lib64, like on
5620 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5621 aren't the modern standard, and it seems that some systems' include files
5622 don't always have them. Exim was already checking for some of the newer
5623 ones like T_AAAA, and defining it itself. I've added checks for all the
5624 record types that Exim uses.
5626 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5627 not automatically generating a new one, as it is supposed to. This
5628 prevented TLS from working. If the file did exist, but contained invalid
5629 data, a new version was generated, as expected. It was only the case of a
5630 non-existent file that was broken.
5632 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5633 with a change in libdomainkeys > 0.64.
5635 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5636 from DNS. If the selector record carries the flag, it now has
5637 precedence over the domain-wide flag.
5639 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5641 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5642 the use of a port name that isn't defined in /etc/services) occurred, the
5643 message was deferred as in a normal delivery, and thus remained on the
5644 spool, instead of being failed because of the mua_wrapper setting. This
5645 is now fixed, and I tidied up some of the mua_wrapper messages at the
5648 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5649 the messages in a hash of arrays rather than using individual hashes.
5650 This is a bit cleaner and results in dramatic memory savings, albeit
5651 at a slight CPU cost.
5653 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5654 as requested by Marc Sherman.
5656 SC/10 Eximstats: added histograms for user specified patterns as requested
5659 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5661 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5662 fopen() in the content-scanning modules that did not already have it.
5668 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5669 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5671 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5673 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5675 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5676 to test Sieve filters that use "vacation".
5678 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5679 that changes the way the GnuTLS parameters are stored in the cache file.
5680 The new format can be generated externally. For backward compatibility,
5681 if the data in the cache doesn't make sense, Exim assumes it has read an
5682 old-format file, and it generates new data and writes a new file. This
5683 means that you can't go back to an older release without removing the
5686 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5687 work if there are any delivery delays because "one_time" forces the
5688 parent to be marked "delivered", so its unseen clone is never tried
5689 again. For this reason, Exim now forbids the simultaneous setting of
5692 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5693 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5694 entry if you want to know the details. The fix, however, neglected to
5695 consider the case when local delivery batching is involved. The test for
5696 "previously delivered" was not happening when checking to see if an
5697 address could be batched with a previous (undelivered) one; under
5698 certain circumstances this could lead to multiple deliveries to the same
5701 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5702 in its include files, and this causes problems building Exim.
5704 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5705 header_syntax) but Exim was just ignoring anything given after a slash.
5706 In particular, this caused confusion with an attempt to use "verify =
5707 reverse_host_lookup/defer_ok". An error is now given when options are
5708 supplied for verify items that do not have them. (Maybe reverse_host_
5709 lookup should have a defer_ok option, but that's a different point.)
5711 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5712 defined by RFC 821) to 2048, because there were problems with some AUTH
5713 commands, and RFC 1869 says the size should be increased for extended
5714 SMTP commands that take arguments.
5716 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5719 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5720 "unknown" error; now it says that the functionality isn't in the binary.
5722 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5723 an address' error message when a string expansion fails (syntax or
5724 whatever). Otherwise the password may appear in the log. Following change
5725 PH/42 below, there is no longer a chance of it appearing in a bounce
5728 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5730 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5731 of its addresses was ever tried. (Bugzilla bug #2.)
5733 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5734 the result incorrectly in the debug output. (It correctly added a newline
5735 to what was transported.)
5737 TF/01 Added $received_time.
5739 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5740 commented out examples of how to interface to a virus scanner and to
5741 SpamAssassin. Also added commented examples of av_scanner and
5742 spamd_address settings.
5744 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5745 and controls are allowed in which ACLs. There were a couple of minor
5746 errors. Some of the entries in the conditions table (which is a table of
5747 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5748 negation of where the condition IS allowed.
5750 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5752 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5753 header file does not have a version number, so I've had to invent a new
5754 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5755 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5756 radiusclient), but it was contributed by a Radius user.
5758 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5759 files or directories, for OpenSSL.
5761 PH/19 When an Exim process that is running as root has to create an Exim log
5762 file, it does so in a subprocess that runs as exim:exim so as to get the
5763 ownership right at creation (otherwise, other Exim processes might see
5764 the file with the wrong ownership). There was no test for failure of this
5765 fork() call, which would lead to the process getting stuck as it waited
5766 for a non-existent subprocess. Forks do occasionally fail when resources
5767 run out. I reviewed all the other calls to fork(); they all seem to check
5770 PH/20 When checking for unexpected SMTP input at connect time (before writing
5771 the banner), Exim was not dealing correctly with a non-positive return
5772 from the read() function. If the client had disconnected by this time,
5773 the result was a log entry for a synchronization error with an empty
5774 string after "input=" when read() returned zero. If read() returned -1
5775 (an event I could not check), uninitialized data bytes were printed.
5776 There were reports of junk text (parts of files, etc) appearing after
5779 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5781 PH/22 Added support for macro redefinition, and (re)definition in between
5782 driver and ACL definitions.
5784 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5785 forgetting to use the resulting value; it was using the unexpanded value.
5787 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5788 hadn't been configured. The fix is from Juergen Kreileder, who
5789 understands it better than I do:
5791 "Here's what I see happening with three configured cyrus_sasl
5792 authenticators configured (plain, login, cram-md5):
5794 On startup auth_cyrus_sasl_init() gets called for each of these.
5795 This means three calls to sasl_listmech() without a specified mech_list.
5796 => SASL tests which mechs of all available mechs actually work
5797 => three warnings about OTP not working
5798 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5800 With the patch, sasl_listmech() also gets called three times. But now
5801 SASL's mech_list option is set to the server_mech specified in the the
5802 authenticator. Or in other words, the answer from sasl_listmech()
5803 gets limited to just the mech you're testing for (which is different
5805 => the return list contains just 'plain' or 'login', 'cram-md5' or
5806 nothing depending on the value of ob->server_mech.
5808 I've just tested the patch: Authentication still works fine,
5809 unavailable mechs specified in the exim configuration are still
5810 caught, and the auth.log warnings about OTP are gone."
5812 PH/25 When debugging is enabled, the contents of the command line are added
5813 to the debugging output, even when log_selector=+arguments is not
5816 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5817 answer is "GNU", and only if the return is "GNU/something" is the answer
5820 PH/27 $acl_verify_message is now set immediately after the failure of a
5821 verification in an ACL, and so is available in subsequent modifiers. In
5822 particular, the message can be preserved by coding like this:
5824 warn !verify = sender
5825 set acl_m0 = $acl_verify_message
5827 Previously, $acl_verify_message was set only while expanding "message"
5828 and "log_message" when a very denied access.
5830 PH/28 Modified OS/os.c-Linux with
5832 -#ifndef OS_LOAD_AVERAGE
5833 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5835 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5836 nomenclature these days.)
5838 PH/29 Installed patch from the Sieve maintainer that adds the options
5839 sieve_useraddress and sieve_subaddress to the redirect router.
5841 PH/30 In these circumstances:
5842 . Two addresses routed to the same list of hosts;
5843 . First host does not offer TLS;
5844 . First host accepts first address;
5845 . First host gives temporary error to second address;
5846 . Second host offers TLS and a TLS session is established;
5847 . Second host accepts second address.
5848 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5849 and peerdn, if requested) that were in fact used only for the second
5852 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5853 attention to any local part prefix or suffix that was matched by the
5854 router that accepted the address. It now behaves in the same way as it
5855 does for delivery: the affixes are removed from the local part unless
5856 rcpt_include_affixes is set on the transport.
5858 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5859 timeout during the DATA phase of an incoming message.
5861 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5862 applied a patch sanctioned by the Sieve maintainer.
5864 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5865 the uid or gid is negative. A case of a negative gid caused this to be
5866 noticed. The fix allows for either to be negative.
5868 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5869 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5872 PH/36 The change PH/12 above was broken. Fixed it.
5874 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5875 the grounds that routing the same address twice would always produce the
5876 same answer. This might have been true once, but it is certainly no
5877 longer true now. Routing a child address may depend on the previous
5878 routing that produced that child. Some complicated redirection strategies
5879 went wrong when messages had multiple recipients, and made Exim's
5880 behaviour dependent on the order in which the addresses were given.
5882 I have moved the duplicate checking until after the routing is complete.
5883 Exim scans the addresses that are assigned to local and remote
5884 transports, and removes any duplicates. This means that more work will be
5885 done, as duplicates will always all be routed, but duplicates are
5886 presumably rare, so I don't expect this is of any significance.
5888 For deliveries to pipes, files, and autoreplies, the duplicate checking
5889 still happens during the routing process, since they are not going to be
5892 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5893 It corrects a timeout issue with spamd. This is Ian's comment: "The
5894 background is that sometimes spamd either never reads data from a
5895 connection it has accepted, or it never writes response data. The exiscan
5896 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5897 blindly assumes that writes won't block so it may never time out."
5899 PH/39 Allow G after quota size as well as K and M.
5901 PH/40 The value set for $authenticated_id in an authenticator may not contain
5902 binary zeroes or newlines because the value is written to log lines and
5903 to spool files. There was no check on this. Now the value is run through
5904 the string_printing() function so that such characters are converted to
5905 printable escape sequences.
5907 PH/41 $message_linecount is a new variable that contains the total number of
5908 lines in the message. Compare $body_linecount, which is the count for the
5911 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5912 bounce and delay warning messages, except in certain special cases, which
5915 (a) An SMTP error message from a remote host;
5916 (b) A message specified in a :fail: redirection;
5917 (c) A message specified in a "fail" command in a system filter;
5918 (d) A message specified in a FAIL return from the queryprogram router;
5919 (e) A message specified by the cannot_route_message router option.
5921 In these cases only, Exim does include the error details in bounce and
5922 warning messages. There are also a few cases where bland messages such
5923 as "unrouteable address" or "local delivery error" are given.
5925 PH/43 $value is now also set for the "else" part of a ${run expansion.
5927 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5928 being worked on, but at least Exim now implements the latest version to
5931 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5932 process to complete was treated as a delivery failure, a timeout while
5933 writing the message to the pipe was logged, but erroneously treated as a
5934 successful delivery. Such timeouts include transport filter timeouts. For
5935 consistency with the overall process timeout, these timeouts are now
5936 treated as errors, giving rise to delivery failures by default. However,
5937 there is now a new Boolean option for the pipe transport called
5938 timeout_defer, which, if set TRUE, converts the failures into defers for
5939 both kinds of timeout. A transport filter timeout is now identified in
5942 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5943 systems where "make" and "gmake" are different, calling "gmake" at top
5944 level broke things. I've arranged for the value of $(MAKE) to be passed
5945 from the Makefile to this script so that it can call the same version of
5949 A note about Exim versions 4.44 and 4.50
5950 ----------------------------------------
5952 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5953 changes of various kinds. As a consequence, a big documentation update was
5954 needed. This delayed the release for rather longer than seemed good, especially
5955 in the light of a couple of (minor) security issues. Therefore, the changes
5956 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5957 release. So 4.44 and 4.50 are in effect two different branches that both start
5960 I have left the 4.50 change log unchanged; it contains all the changes since
5961 4.43. The change log for 4.44 is below; many of its items are identical to
5962 those for 4.50. This seems to be the most sensible way to preserve the
5963 historical information.
5969 1. Minor wording change to the doc/README.SIEVE file.
5971 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5972 computation of the current number of files was incorrect.
5974 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5975 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5978 4. Give more explanation in the error message when the command for a transport
5979 filter fails to execute.
5981 5. There are several places where Exim runs a non-Exim command in a
5982 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5983 was being done only for the command run by the queryprogram router. It is
5984 now done for all such subprocesses. The other cases are: ${run, transport
5985 filters, and the commands run by the lmtp and pipe transports.
5987 6. Added CONFIGURE_GROUP build-time option.
5989 7. Some older OS have a limit of 256 on the maximum number of file
5990 descriptors. Exim was using setrlimit() to set 1000 as a large value
5991 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5992 systems. I've change it so that if it can't get 1000, it tries for 256.
5994 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5995 was an oversight, and furthermore, ever since the addition of extra
5996 controls (e.g. 4.43/32), the checks on when to allow different forms of
5997 "control" were broken. There should now be diagnostics for all cases when a
5998 control that does not make sense is encountered.
6000 9. Added the /retain_sender option to "control=submission".
6002 10. $recipients is now available in the predata ACL (oversight).
6004 11. Tidy the search cache before the fork to do a delivery from a message
6005 received from the command line. Otherwise the child will trigger a lookup
6006 failure and thereby defer the delivery if it tries to use (for example) a
6007 cached ldap connection that the parent has called unbind on.
6009 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6010 of $address_data from the recipient verification was clobbered by the
6011 sender verification.
6013 13. The value of address_data from a sender verification is now available in
6014 $sender_address_data in subsequent conditions in the ACL statement.
6016 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6018 15. Added a new option "connect=<time>" to callout options, to set a different
6021 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6022 was its contents. (It was OK if the option was not defined at all.)
6024 17. A "Completed" log line is now written for messages that are removed from
6025 the spool by the -Mrm option.
6027 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6028 information about exactly what failed.
6030 19. Added -dd to debug only the daemon process.
6032 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6033 handles timeouts, both on the server side and network timeouts. Renamed the
6034 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6036 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6037 It is now set to "smtps".
6039 22. $host_address is now set to the target address during the checking of
6040 ignore_target_hosts.
6042 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6043 being passed; this would have caused $sender_host_name to have been used if
6044 matching the list had actually called for a host name (not very likely,
6045 since this list is usually IP addresses). A host name is now passed as
6048 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6049 code that specifies a non-restarting handler (typically sigaction() in
6050 modern systems) in an attempt to fix a rare and obscure crash bug.
6052 25. Narrowed the window for a race in the daemon that could cause it to ignore
6053 SIGCHLD signals. This is not a major problem, because they are used only to
6054 wake it up if nothing else does.
6056 26. A malformed maildirsize file could cause Exim to calculate negative values
6057 for the mailbox size or file count. Odd effects could occur as a result.
6058 The maildirsize information is now recalculated if the size or filecount
6061 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6062 support for a long time. Removed HAVE_SYS_VFS_H.
6064 28. Installed the latest version of exipick from John Jetmore.
6066 29. In an address list, if the pattern was not a regular expression, an empty
6067 subject address (from a bounce message) matched only if the pattern was an
6068 empty string. Non-empty patterns were not even tested. This was the wrong
6069 because it is perfectly reasonable to use an empty address as part of a
6070 database query. An empty address is now tested by patterns that are
6071 lookups. However, all the other forms of pattern expect the subject to
6072 contain a local part and a domain, and therefore, for them, an empty
6073 address still always fails if the pattern is not itself empty.
6075 30. Exim went into a mad DNS loop when attempting to do a callout where the
6076 host was specified on an smtp transport, and looking it up yielded more
6077 than one IP address.
6079 31. Re-factored the code for checking spool and log partition space into a
6080 function that finds that data and another that does the check. The former
6081 is then used to implement four new variables: $spool_space, $log_space,
6082 $spool_inodes, and $log_inodes.
6084 32. The RFC2047 encoding function was originally intended for short strings
6085 such as real names; it was not keeping to the 75-character limit for
6086 encoded words that the RFC imposes. It now respects the limit, and
6087 generates multiple encoded words if necessary. To be on the safe side, I
6088 have increased the buffer size for the ${rfc2047: expansion operator from
6091 33. It is now permitted to omit both strings after an "if" condition; if the
6092 condition is true, the result is "true". As before, when the second string
6093 is omitted, a false condition yields an empty string. This makes it less
6094 cumbersome to write custom ACL and router conditions.
6096 34. Failure to deliver a bounce message always caused it to be frozen, even if
6097 there was an errors_to setting on the router. The errors_to setting is now
6100 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6101 canonical form (fully expanded) before being placed in
6102 $sender_host_address.
6104 36. The table in the code that translates DNS record types into text (T_A to
6105 "A" for instance) was missing entries for NS and CNAME. It is just possible
6106 that this could have caused confusion if both these types were looked up
6107 for the same domain, because the text type is used as part of Exim's
6108 per-process caching. But the chance of anyone hitting this buglet seems
6111 37. The dnsdb lookup has been extended in a number of ways.
6113 (1) There is a new type, "zns", which walks up the domain tree until it
6114 finds some nameserver records. It should be used with care.
6116 (2) There is a new type, "mxh", which is like "mx" except that it returns
6117 just the host names, not the priorities.
6119 (3) It is now possible to give a list of domains (or IP addresses) to be
6120 looked up. The behaviour when one of the lookups defers can be
6121 controlled by a keyword.
6123 (4) It is now possible to specify the separator character for use when
6124 multiple records are returned.
6126 38. The dnslists ACL condition has been extended: it is now possible to supply
6127 a list of IP addresses and/or domains to be looked up in a particular DNS
6130 39. Added log_selector=+queue_time_overall.
6132 40. When running the queue in the test harness, wait just a tad after forking a
6133 delivery process, to get repeatability of debugging output.
6135 41. Include certificate and key file names in error message when GnuTLS fails
6136 to set them up, because the GnuTLS error message doesn't include the name
6137 of the failing file when there is a problem reading it.
6139 42. Allow both -bf and -bF in the same test run.
6141 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6143 44. The "Exiscan patch" is now merged into the mainline Exim source.
6145 45. Sometimes the final signoff response after QUIT could fail to get
6146 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6147 < 0 before doing a fflush(). This bug looks as though it goes back to the
6148 introduction of TLS in release 3.20, but "sometimes" must have been rare
6149 because the tests only now provoked it.
6151 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6152 (this can affect the format of dates).
6154 47. exim_tidydb, when checking for the continued existence of a message for
6155 which it has found a message-specific retry record, was not finding
6156 messages that were in split spool directories. Consequently, it was
6157 deleting retry records that should have stayed in existence.
6159 48. Steve fixed some bugs in eximstats.
6161 49. The SPA authentication driver was not abandoning authentication and moving
6162 on to the next authenticator when an expansion was forced to fail,
6163 contradicting the general specification for all authenticators. Instead it
6164 was generating a temporary error. It now behaves as specified.
6166 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6167 (the order specifies the preference for clients). The order is now AES256,
6168 AES128, 3DES, ARCFOUR128.
6170 51. Small patch to Sieve code - explicitly set From: when generating an
6173 52. Exim crashed if a remote delivery caused a very long error message to be
6174 recorded - for instance if somebody sent an entire SpamAssassin report back
6175 as a large number of 550 error lines. This bug was coincidentally fixed by
6176 increasing the size of one of Exim's internal buffers (big_buffer) that
6177 happened as part of the Exiscan merge. However, to be on the safe side, I
6178 have made the code more robust (and fixed the comments that describe what
6181 53. Now that there can be additional text after "Completed" in log lines (if
6182 the queue_time_overall log selector is set), a one-byte patch to exigrep
6183 was needed to allow it to recognize "Completed" as not the last thing in
6186 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6187 patch that reportedly fixes this has been added. I am not expert enough to
6188 create a test for it. This is what the patch creator wrote:
6190 "I found a little strange behaviour of ldap code when working with
6191 Windows 2003 AD Domain, where users was placed in more than one
6192 Organization Units. When I tried to give exim partial DN, the exit code
6193 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6194 But simultaneously result of request was absolutely normal ldap result,
6195 so I produce this patch..."
6197 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6198 REFERENCE, so I have modified the code to exclude the patch when that macro
6201 55. Some experimental protocols are using DNS PTR records for new purposes. The
6202 keys for these records are domain names, not reversed IP addresses. The
6203 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6204 leaves it alone. Component reversal etc. now happens only for IP addresses.
6207 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6209 57. Double the size of the debug message buffer (to 2048) so that more of very
6210 long debug lines gets shown.
6212 58. The exicyclog utility now does better if the number of log files to keep
6213 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6215 59. Two changes related to the smtp_active_hostname option:
6217 (1) $smtp_active_hostname is now available as a variable.
6218 (2) The default for smtp_banner uses $smtp_active_hostname instead
6219 of $primary_hostname.
6221 60. The host_aton() function is supposed to be passed a string that is known
6222 to be a valid IP address. However, in the case of IPv6 addresses, it was
6223 not checking this. This is a hostage to fortune. Exim now panics and dies
6224 if the condition is not met. A case was found where this could be provoked
6225 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6226 components; fortuitously, this particular loophole had already been fixed
6227 by change 4.50/55 above.
6229 If there are any other similar loopholes, the new check in host_aton()
6230 itself should stop them being exploited. The report I received stated that
6231 data on the command line could provoke the exploit when Exim was running as
6232 exim, but did not say which command line option was involved. All I could
6233 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6234 running as the user.
6237 61. There was a buffer overflow vulnerability in the SPA authentication code
6238 (which came originally from the Samba project). I have added a test to the
6239 spa_base64_to_bits() function which I hope fixes it.
6242 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6243 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6245 63. The daemon start-up calls getloadavg() while still root for those OS that
6246 need the first call to be done as root, but it missed one case: when
6247 deliver_queue_load_max is set with deliver_drop_privilege. This is
6248 necessary for the benefit of the queue runner, because there is no re-exec
6249 when deliver_drop_privilege is set.
6251 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6252 This has been fixed.
6254 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6255 was in use, was not putting the data itself into the right store pool;
6256 consequently, it could be overwritten for a subsequent message in the same
6257 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6260 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6262 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6263 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6264 the function were treating the return as a boolean value, which happened to
6265 work because 0=false and not-0=true, but is not correct code.
6267 68. The host_aton() function was not handling scoped IPv6 addresses (those
6268 with, for example, "%eth0" on the end) correctly.
6270 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6271 negated items (that is, ~something) in unsigned ints. Some compilers
6272 apparently mutter when there is no cast.
6274 70. If an address verification called from an ACL failed, and did not produce a
6275 user-specific message (i.e. there was only a "system" message), nothing was
6276 put in $acl_verify_message. In this situation, it now puts the system
6279 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6280 session; change 4.31/43 added the unwanted input to the log line - except
6281 that it did not do this in the start of session case. It now does.
6283 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6284 This is wrong and can cause the other end to generate a synchronization
6285 error if it is another Exim or anything else that does the synchronization
6286 check. A QUIT command is no longer sent after a timeout.
6288 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6289 during host lookups.
6291 74. The defer_ok option of callout verification was not working if it was used
6292 when verifying addresses in header lines, that is, for this case:
6294 verify = header_sender/callout=defer_ok
6296 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6297 those file descriptors could be used for SMTP connections. If anything
6298 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6299 could be sent to the SMTP client, causing chaos. The daemon now opens
6300 stdin, stdout, and stderr to /dev/null when it puts itself into the
6303 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6304 log by default. The user can override this with suitable Perl magic.
6306 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6307 the log message when discard triggers, was not working for the DATA ACL or
6308 for the non-SMTP ACL.
6310 78. Error message wording change in sieve.c.
6312 79. If smtp_accept_max_per_host was set, the number of connections could be
6313 restricted to fewer than expected, because the daemon was trying to set up
6314 a new connection before checking whether the processes handling previous
6315 connections had finished. The check for completed processes is now done
6316 earlier. On busy systems, this bug wouldn't be noticed because something
6317 else would have woken the daemon, and it would have reaped the completed
6320 80. If a message was submitted locally by a user whose login name contained one
6321 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6322 It caused a spool format error. I have fixed the spool reading code. A
6323 related problem was that the "from" clause in the Received: line became
6324 illegal because of the space(s). It is now covered by ${quote_local_part.
6326 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6329 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6331 83. Patch from Sieve maintainer to fix unterminated string problem in
6332 "vacation" handling.
6334 84. Some minor changes to the Linux configuration files to help with other
6335 OS variants using glibc.
6337 85. One more patch for Sieve to update vacation handling to latest spec.
6340 ----------------------------------------------------
6341 See the note above about the 4.44 and 4.50 releases.
6342 ----------------------------------------------------
6348 1. Change 4.43/35 introduced a bug that caused file counts to be
6349 incorrectly computed when quota_filecount was set in an appendfile
6352 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6353 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6356 3. Give more explanation in the error message when the command for a transport
6357 filter fails to execute.
6359 4. There are several places where Exim runs a non-Exim command in a
6360 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6361 was being done only for the command run by the queryprogram router. It is
6362 now done for all such subprocesses. The other cases are: ${run, transport
6363 filters, and the commands run by the lmtp and pipe transports.
6365 5. Some older OS have a limit of 256 on the maximum number of file
6366 descriptors. Exim was using setrlimit() to set 1000 as a large value
6367 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6368 systems. I've change it so that if it can't get 1000, it tries for 256.
6370 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6371 was an oversight, and furthermore, ever since the addition of extra
6372 controls (e.g. 4.43/32), the checks on when to allow different forms of
6373 "control" were broken. There should now be diagnostics for all cases when a
6374 control that does not make sense is encountered.
6376 7. $recipients is now available in the predata ACL (oversight).
6378 8. Tidy the search cache before the fork to do a delivery from a message
6379 received from the command line. Otherwise the child will trigger a lookup
6380 failure and thereby defer the delivery if it tries to use (for example) a
6381 cached ldap connection that the parent has called unbind on.
6383 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6384 of $address_data from the recipient verification was clobbered by the
6385 sender verification.
6387 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6388 was its contents. (It was OK if the option was not defined at all.)
6390 11. A "Completed" log line is now written for messages that are removed from
6391 the spool by the -Mrm option.
6393 12. $host_address is now set to the target address during the checking of
6394 ignore_target_hosts.
6396 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6397 being passed; this would have caused $sender_host_name to have been used if
6398 matching the list had actually called for a host name (not very likely,
6399 since this list is usually IP addresses). A host name is now passed as
6402 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6403 code that specifies a non-restarting handler (typically sigaction() in
6404 modern systems) in an attempt to fix a rare and obscure crash bug.
6406 15. Narrowed the window for a race in the daemon that could cause it to ignore
6407 SIGCHLD signals. This is not a major problem, because they are used only to
6408 wake it up if nothing else does.
6410 16. A malformed maildirsize file could cause Exim to calculate negative values
6411 for the mailbox size or file count. Odd effects could occur as a result.
6412 The maildirsize information is now recalculated if the size or filecount
6415 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6416 support for a long time. Removed HAVE_SYS_VFS_H.
6418 18. Updated exipick to current release from John Jetmore.
6420 19. Allow an empty sender to be matched against a lookup in an address list.
6421 Previously the only cases considered were a regular expression, or an
6424 20. Exim went into a mad DNS lookup loop when doing a callout where the
6425 host was specified on the transport, if the DNS lookup yielded more than
6428 21. The RFC2047 encoding function was originally intended for short strings
6429 such as real names; it was not keeping to the 75-character limit for
6430 encoded words that the RFC imposes. It now respects the limit, and
6431 generates multiple encoded words if necessary. To be on the safe side, I
6432 have increased the buffer size for the ${rfc2047: expansion operator from
6435 22. Failure to deliver a bounce message always caused it to be frozen, even if
6436 there was an errors_to setting on the router. The errors_to setting is now
6439 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6440 canonical form (fully expanded) before being placed in
6441 $sender_host_address.
6443 24. Updated eximstats to version 1.33
6445 25. Include certificate and key file names in error message when GnuTLS fails
6446 to set them up, because the GnuTLS error message doesn't include the name
6447 of the failing file when there is a problem reading it.
6449 26. Expand error message when OpenSSL has problems setting up cert/key files.
6452 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6453 (this can affect the format of dates).
6455 28. exim_tidydb, when checking for the continued existence of a message for
6456 which it has found a message-specific retry record, was not finding
6457 messages that were in split spool directories. Consequently, it was
6458 deleting retry records that should have stayed in existence.
6460 29. eximstats updated to version 1.35
6461 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6462 1.35 - bugfix such that pie charts by volume are generated correctly
6464 30. The SPA authentication driver was not abandoning authentication and moving
6465 on to the next authenticator when an expansion was forced to fail,
6466 contradicting the general specification for all authenticators. Instead it
6467 was generating a temporary error. It now behaves as specified.
6469 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6470 (the order specifies the preference for clients). The order is now AES256,
6471 AES128, 3DES, ARCFOUR128.
6473 31. Small patch to Sieve code - explicitly set From: when generating an
6476 32. Exim crashed if a remote delivery caused a very long error message to be
6477 recorded - for instance if somebody sent an entire SpamAssassin report back
6478 as a large number of 550 error lines. This bug was coincidentally fixed by
6479 increasing the size of one of Exim's internal buffers (big_buffer) that
6480 happened as part of the Exiscan merge. However, to be on the safe side, I
6481 have made the code more robust (and fixed the comments that describe what
6484 33. Some experimental protocols are using DNS PTR records for new purposes. The
6485 keys for these records are domain names, not reversed IP addresses. The
6486 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6487 leaves it alone. Component reversal etc. now happens only for IP addresses.
6490 34. The host_aton() function is supposed to be passed a string that is known
6491 to be a valid IP address. However, in the case of IPv6 addresses, it was
6492 not checking this. This is a hostage to fortune. Exim now panics and dies
6493 if the condition is not met. A case was found where this could be provoked
6494 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6495 components; fortuitously, this particular loophole had already been fixed
6496 by change 4.50/55 or 4.44/33 above.
6498 If there are any other similar loopholes, the new check in host_aton()
6499 itself should stop them being exploited. The report I received stated that
6500 data on the command line could provoke the exploit when Exim was running as
6501 exim, but did not say which command line option was involved. All I could
6502 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6503 running as the user.
6506 35. There was a buffer overflow vulnerability in the SPA authentication code
6507 (which came originally from the Samba project). I have added a test to the
6508 spa_base64_to_bits() function which I hope fixes it.
6511 36. The daemon start-up calls getloadavg() while still root for those OS that
6512 need the first call to be done as root, but it missed one case: when
6513 deliver_queue_load_max is set with deliver_drop_privilege. This is
6514 necessary for the benefit of the queue runner, because there is no re-exec
6515 when deliver_drop_privilege is set.
6517 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6518 was in use, was not putting the data itself into the right store pool;
6519 consequently, it could be overwritten for a subsequent message in the same
6520 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6523 38. Sometimes the final signoff response after QUIT could fail to get
6524 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6525 < 0 before doing a fflush(). This bug looks as though it goes back to the
6526 introduction of TLS in release 3.20, but "sometimes" must have been rare
6527 because the tests only now provoked it.
6533 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6534 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6535 Now it returns an integer. A number of calls were still expecting a T/F
6536 return. Fortuitously, in all cases, the tests worked in OK situations,
6537 which is the norm. However, things would have gone wrong on any write
6538 failures on the smtp file descriptor. This function is used when sending
6539 messages over SMTP and also when doing verify callouts.
6541 2. When Exim is called to do synchronous delivery of a locally submitted
6542 message (the -odf or -odi options), it no longer closes stderr before doing
6545 3. Implemented the mua_wrapper option.
6547 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6549 5. Implemented the functions header_remove(), header_testname(),
6550 header_add_at_position(), and receive_remove_recipient(), and exported them
6553 6. If an ACL "warn" statement specified the addition of headers, Exim already
6554 inserted X-ACL-Warn: at the start if there was no header name. However, it
6555 was not making this test for the second and subsequent header lines if
6556 there were newlines in the string. This meant that an invalid header could
6557 be inserted if Exim was badly configured.
6559 7. Allow an ACL "warn" statement to add header lines at the start or after all
6560 the Received: headers, as well as at the end.
6562 8. Added the rcpt_4xx retry error code.
6564 9. Added postmaster_mailfrom=xxx to callout verification option.
6566 10. Added mailfrom=xxxx to the callout verification option, for verify=
6569 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6570 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6572 12. Inserted some casts to stop certain compilers warning when using pointer
6573 differences as field lengths or precisions in printf-type calls (mostly
6574 affecting debugging statements).
6576 13. Added optional readline() support for -be (dynamically loaded).
6578 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6579 same clock tick as a message's arrival, so that its received time was the
6580 same as the "first fail" time on the retry record, and that message
6581 remained on the queue past the ultimate address timeout, every queue runner
6582 would try a delivery (because it was past the ultimate address timeout) but
6583 after another failure, the ultimate address timeout, which should have then
6584 bounced the address, did not kick in. This was a "< instead of <=" error;
6585 in most cases the first failure would have been in the next clock tick
6586 after the received time, and all would be well.
6588 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6589 being recognized when the domain list was tested by the match_domain
6590 condition in an expansion string.
6592 16. Added the ${str2b64: operator.
6594 17. Exim was always calling setrlimit() to set a large limit for the number of
6595 processes, without checking whether the existing limit was already
6596 adequate. (It did check for the limit on file descriptors.) Furthermore,
6597 errors from getrlimit() and setrlimit() were being ignored. Now they are
6598 logged to the main and panic logs, but Exim does carry on, to try to do its
6599 job under whatever limits there are.
6601 18. Imported PCRE 5.0.
6603 19. Trivial typo in log message " temporarily refused connection" (the leading
6606 20. If the log selector return_path_on_delivery was set and an address was
6607 redirected to /dev/null, the delivery process crashed because it assumed
6608 that a return path would always be set for a "successful" delivery. In this
6609 case, the whole delivery is bypassed as an optimization, and therefore no
6612 21. Internal re-arrangement: the function for sending a challenge and reading
6613 a response while authentication was assuming a zero-terminated challenge
6614 string. It's now changed to take a pointer and a length, to allow for
6615 binary data in such strings.
6617 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6619 23. Exim was not respecting finduser_retries when seeking the login of the
6620 uid under which it was called; it was always trying 10 times. (The default
6621 setting of finduser_retries is zero.) Also, it was sleeping after the final
6622 failure, which is pointless.
6624 24. Implemented tls_on_connect_ports.
6626 25. Implemented acl_smtp_predata.
6628 26. If the domain in control=submission is set empty, Exim assumes that the
6629 authenticated id is a complete email address when it generates From: or
6630 Sender: header lines.
6632 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6633 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6634 chown and chgrp in /bin and hostname in /usr/bin.
6636 28. Exim was keeping the "process log" file open after each use, just as it
6637 does for the main log. This opens the possibility of it remaining open for
6638 long periods when the USR1 signal hits a daemon. Occasional processlog
6639 errors were reported, that could have been caused by this. Anyway, it seems
6640 much more sensible not to leave this file open at all, so that is what now
6643 29. The long-running daemon process does not normally write to the log once it
6644 has entered its main loop, and it closes the log before doing so. This is
6645 so that log files can straightforwardly be renamed and moved. However,
6646 there are a couple of unusual error situations where the daemon does write
6647 log entries, and I had neglected to close the log afterwards.
6649 30. The text of an SMTP error response that was received during a remote
6650 delivery was being truncated at 512 bytes. This is too short for some of
6651 the long messages that one sometimes sees. I've increased the limit to
6654 31. It is now possible to make retry rules that apply only when a message has a
6655 specific sender, in particular, an empty sender.
6657 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6658 it possible to be selective about when SMTP synchronization is enforced.
6660 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6662 32. Implemented hosts_connection_nolog.
6664 33. Added an ACL for QUIT.
6666 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6669 35. Added mailbox_size and mailbox_filecount to appendfile.
6671 36. Added control = no_multiline_responses to ACLs.
6673 37. There was a bug in the logic of the code that waits for the clock to tick
6674 in the case where the clock went backwards by a substantial amount such
6675 that the microsecond fraction of "now" was more than the microsecond
6676 fraction of "then" (but the whole seconds number was less).
6678 38. Added support for the libradius Radius client library this is found on
6679 FreeBSD (previously only the radiusclient library was supported).
6685 1. When certain lookups returned multiple values in the form name=value, the
6686 quoting of the values was not always being done properly. Specifically:
6687 (a) If the value started with a double quote, but contained no whitespace,
6689 (b) If the value contained whitespace other than a space character (i.e.
6690 tabs or newlines or carriage returns) it was not quoted.
6691 This fix has been applied to the mysql and pgsql lookups by writing a
6692 separate quoting function and calling it from the lookup code. The fix
6693 should probably also be applied to nisplus, ibase and oracle lookups, but
6694 since I cannot test any of those, I have not disturbed their existing code.
6696 2. A hit in the callout cache for a specific address caused a log line with no
6697 reason for rejecting RCPT. Now it says "Previous (cached) callout
6698 verification failure".
6700 3. There was an off-by-one bug in the queryprogram router. An over-long
6701 return line was truncated at 256 instead of 255 characters, thereby
6702 overflowing its buffer with the terminating zero. As well as fixing this, I
6703 have increased the buffer size to 1024 (and made a note to document this).
6705 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6706 when Exim is waiting for an SMTP response from a remote server, Exim
6707 restarts its select() call on the socket, thereby resetting its timeout.
6708 This is not a problem when such interrupts are rare. Somebody set up a cron
6709 job to run exiwhat every 2 minutes, which is less than the normal select()
6710 timeout (5 or 10 minutes). This meant that the select() timeout never
6711 kicked in because it was always reset. I have fixed this by comparing the
6712 time when an interrupt arrives with the time at the start of the first call
6713 to select(). If more time than the timeout has elapsed, the interrupt is
6714 treated as a timeout.
6716 5. Some internal re-factoring in preparation for the addition of Sieve
6717 extensions (by MH). In particular, the "personal" test is moved to a
6718 separate function, and given an option for scanning Cc: and Bcc: (which is
6719 not set for Exim filters).
6721 6. When Exim created an email address using the login of the caller as the
6722 local part (e.g. when creating a From: or Sender: header line), it was not
6723 quoting the local part when it contained special characters such as @.
6725 7. Installed new OpenBSD configuration files.
6727 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6728 try to make them clearer.
6730 9. Callout options, other than the timeout value, were being ignored when
6731 verifying sender addresses in header lines. For example, when using
6733 verify = header_sender/callout=no_cache
6735 the cache was (incorrectly) being used.
6737 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6738 only the Cygwin environment.
6740 11. When return_path_on_delivery was set as a log selector, if different remote
6741 addresses in the same message used different return paths and parallel
6742 remote delivery occurred, the wrong values would sometimes be logged.
6743 (Whenever a remote delivery process finished, the return path value from
6744 the most recently started remote delivery process was logged.)
6746 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6747 lines when AUTH and/or TLS are in use. This is the "received protocol"
6748 field. Exim used to use "asmtp" for authenticated SMTP, without any
6749 indication (in the protocol name) for TLS use. Now it follows the RFC and
6750 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6751 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6752 names appear in log lines as well as in Received: header lines.
6754 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6755 extensions, and comparison tests, and to fix some bugs.
6757 14. Changes to the "personal" filter test:
6759 (1) The test was buggy in that it was just doing the equivalent of
6760 "contains" tests on header lines. For example, if a user's address was
6761 anne@some.where, the "personal" test would incorrectly be true for
6763 To: susanne@some.where
6765 This test is now done by extracting each address from the header in turn,
6766 and checking the entire address. Other tests that are part of "personal"
6767 are now done using regular expressions (for example, to check local parts
6768 of addresses in From: header lines).
6770 (2) The list of non-personal local parts in From: addresses has been
6771 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6772 taken from the Sieve specification recommendations.
6774 (3) If the message contains any header line starting with "List-" it is
6775 treated as non-personal.
6777 (4) The test for "circular" in the Subject: header line has been removed
6778 because it now seems ill-conceived.
6780 15. Minor typos in src/EDITME comments corrected.
6782 16. Installed latest exipick from John Jetmore.
6784 17. If headers_add on a router specified a text string that was too long for
6785 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6786 of string_sprintf() is now avoided.
6788 18. $message_body_size was not set (it was always zero) when running the DATA
6789 ACL and the local_scan() function.
6791 19. For the "mail" command in an Exim filter, no default was being set for
6792 the once_repeat time, causing a random time value to be used if "once" was
6793 specified. (If the value happened to be <= 0, no repeat happened.) The
6794 default is now 0s, meaning "never repeat". The "vacation" command was OK
6795 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6796 (I found it when inspecting the code).
6798 20. There is now an overall timeout for performing a callout verification. It
6799 defaults to 4 times the callout timeout, which applies to individual SMTP
6800 commands during the callout. The overall timeout applies when there is more
6801 than one host that can be tried. The timeout is checked before trying the
6802 next host. This prevents very long delays if there are a large number of
6803 hosts and all are timing out (e.g. when the network connections are timing
6804 out). The value of the overall timeout can be changed by specifying an
6805 additional sub-option for "callout", called "maxwait". For example:
6807 verify = sender/callout=5s,maxwait=20s
6809 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6810 to the end before writing, but this should make it even safer).
6812 22. Exim was forgetting that it had advertised PIPELINING for the second and
6813 subsequent messages on an SMTP connection. It was also not resetting its
6814 memory on STARTTLS and an internal HELO.
6816 23. When Exim logs an SMTP synchronization error within a session, it now
6817 records whether PIPELINING has been advertised or not.
6819 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6820 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6821 rather than long int.
6823 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6825 26. Added the never_mail option to autoreply.
6831 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6832 crash if the getsockname() call failed; this can happen if a connection is
6833 closed very soon after it is established. The problem was simply in the
6834 order in which certain operations were done, causing Exim to try to write
6835 to the SMTP stream before it had set up the file descriptor. The bug has
6836 been fixed by making things happen in the correct order.
6842 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6843 before the connection was closed, thus losing the rejection response.
6845 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6846 some early Solaris releases, but causes trouble in current releases where
6847 socklen_t is defined.
6849 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6852 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6855 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6856 that releasing the top of it at the end releases what was used for sub-
6857 expansions (unless the block got too big). However, discard this block if
6858 the first thing is a variable or header, so that we can use its block when
6859 it is dynamic (useful for very large $message_headers, for example).
6861 6. Lookups now cache *every* query, not just the most recent. A new, separate
6862 store pool is used for this. It can be recovered when all lookup caches are
6863 flushed. Lookups now release memory at the end of their result strings.
6864 This has involved some general refactoring of the lookup sources.
6866 7. Some code has been added to the store_xxx() functions to reduce the amount
6867 of flapping under certain conditions.
6869 8. log_incoming_interface used to affect only the <= reception log lines. Now
6870 it causes the local interface and port to be added to several more SMTP log
6871 lines, for example "SMTP connection from", and rejection lines.
6873 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6875 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6877 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6878 could be overwritten at the end of the current message (or the start of a
6879 new message if it was set in a HELO ACL). The value is now preserved for
6880 the duration of the SMTP connection.
6882 12. If a transport had a headers_rewrite setting, and a matching header line
6883 contained an unqualified address, that address was qualified, even if it
6884 did not match any rewriting rules. The underlying bug was that the values
6885 of the flags that permit the existence of unqualified sender and recipient
6886 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6887 non-local messages, and by -bnq for local messages) were not being
6888 preserved with the message after it was received.
6890 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6891 "next input=" as part of the text comprising the host identity instead of
6892 the correct text. The code was using the same buffer for two different
6893 strings. However, depending on which order the printing function evaluated
6894 its arguments, the bug did not always show up. Under Linux, for example, my
6895 test suite worked just fine.
6897 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6898 doesn't work with some older versions of Perl. It has been changed to "my",
6899 which in any case is probably the better facility to use.
6901 15. A really picky compiler found some instances of statements for creating
6902 error messages that either had too many or two few arguments for the format
6905 16. The size of the buffer for calls to the DNS resolver has been increased
6906 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6907 for addresses that have a lot of PTR records. This alleviates a problem; it
6908 does not fully solve it.
6910 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6911 buffer now truncates the list and logs the incident, which is the same
6912 action as happens when Exim is looking up a host name and its aliases.
6913 Previously in this situation something unpredictable would happen;
6914 sometimes it was "internal error: store_reset failed".
6916 18. If a server dropped the connection unexpectedly when an Exim client was
6917 using GnuTLS and trying to read a response, the client delivery process
6918 crashed while trying to generate an error log message.
6920 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6921 string, for example:
6923 warn message = H1: something\nH2: something
6925 the text was added as a single header line from Exim's point of view
6926 though it ended up OK in the delivered message. However, searching for the
6927 second and subsequent header lines using $h_h2: did not work. This has been
6928 fixed. Similarly, if a system filter added multiple headers in this way,
6929 the routers could not see them.
6931 20. Expanded the error message when iplsearch is called with an invalid key to
6932 suggest using net-iplsearch in a host list.
6934 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6935 ACLs are no longer actually imposed (and a message to that effect is
6938 22. If a "gecos" field in a passwd entry contained escaped characters, in
6939 particular, if it contained a \" sequence, Exim got it wrong when building
6940 a From: or a Sender: header from that name. A second bug also caused
6941 incorrect handling when an unquoted " was present following a character
6942 that needed quoting.
6944 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6945 was not being matched caselessly.
6947 24. Arranged for all hyphens in the exim.8 source to be escaped with
6950 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6951 properly. Recipient callout cache records were still being keyed to include
6952 the sender, even when use_sender was set false. This led to far more
6953 callouts that were necessary. The sender is no longer included in the key
6954 when use_sender is false.
6956 26. Added "control = submission" modifier to ACLs.
6958 27. Added the ${base62d: operator to decode base 62 numbers.
6960 28. dnsdb lookups can now access SRV records.
6962 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6963 the configuration file.
6965 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6966 (-v) mode. This makes the output for a verbose queue run more intelligible.
6968 31. Added a use_postmaster feature to recipient callouts.
6970 32. Added the $body_zerocount variable, containing the number of binary zero
6971 bytes in the message body.
6973 33. The time of last modification of the "new" subdirectory is now used as the
6974 "mailbox time last read" when there is a quota error for a maildir
6977 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6979 35. Added +ignore_unknown as a special item in host lists.
6981 36. Code for decoding IPv6 addresses in host lists is now included, even if
6982 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6983 address was recognized as an IP address, but was then not correctly decoded
6984 into binary, causing unexpected and incorrect effects when compared with
6991 1. Very minor rewording of debugging text in manualroute to say "list of
6992 hosts" instead of "hostlist".
6994 2. If verify=header_syntax was set, and a header line with an unqualified
6995 address (no domain) and a large number of spaces between the end of the
6996 name and the colon was received, the reception process suffered a buffer
6997 overflow, and (when I tested it) crashed. This was caused by some obsolete
6998 code that should have been removed. The fix is to remove it!
7000 3. When running in the test harness, delay a bit after writing a bounce
7001 message to get a bit more predictability in the log output.
7003 4. Added a call to search_tidyup() just before forking a reception process. In
7004 theory, someone could use a lookup in the expansion of smtp_accept_max_
7005 per_host which, without the tidyup, could leave open a database connection.
7007 5. Added the variables $recipient_data and $sender_data which get set from a
7008 lookup success in an ACL "recipients" or "senders" condition, or a router
7009 "senders" option, similar to $domain_data and $local_part_data.
7011 6. Moved the writing of debug_print from before to after the "senders" test
7014 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7015 problems for message scanning, either using a data ACL, or using
7016 local_scan() because the Received: header was not generated till after they
7017 were called (in order to set the time as the time of reception completion).
7018 I have revised the way this works. The header is now generated after the
7019 body is received, but before the ACL or local_scan() are called. After they
7020 are run, the timestamp in the header is updated.
7026 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7027 before starting a queue runner without re-exec. This happened only when
7028 deliver_drop_privilege was set or when the Exim user was set to root. The
7029 effect of the bug was that timeouts during subsequent deliveries caused
7030 crashes instead of being properly handled. The handler is now left at its
7031 default (and expected) setting.
7033 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7034 message, again when deliver_drop_privilege is set or Exim is run as root.
7035 The bug described in (1) was not present in this case, but the tidying up
7036 of the other signals was missing. I have made the two cases consistent.
7038 3. The ignore_target_hosts setting on a manualroute router was being ignored
7039 for hosts that were looked up using the /MX notation.
7041 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7044 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7045 operated on the sender address. After changing the $sender_address to <>
7046 for the sender address verify, Exim was re-instated it as the original
7047 (before rewriting) address, but remembering that it had rewritten it, so it
7048 wasn't rewriting it again. This bug also had the effect of breaking the
7049 sender address verification caching when the sender address was rewritten.
7051 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7052 This has been changed so that if the ip literal address matches
7053 ignore_target_hosts, the router declines.
7055 7. Added expansion conditions match_domain, match_address, and match_local_
7056 part (NOT match_host).
7058 8. The placeholder for the Received: header didn't have a length field set.
7060 9. Added code to Exim itself and to exim_lock to test for a specific race
7061 condition that could lead to file corruption when using MBX delivery. The
7062 issue is with the lockfile that is created in /tmp. If this file is removed
7063 after a process has opened it but before that process has acquired a lock,
7064 there is the potential for a second process to recreate the file and also
7065 acquire a lock. This could lead to two Exim processes writing to the file
7066 at the same time. The added code performs the same test as UW imapd; it
7067 checks after acquiring the lock that its file descriptor still refers to
7068 the same named file.
7070 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7071 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7074 11. Added the smtp_active_hostname option. If used, this will typically be made
7075 to depend on the incoming interface address. Because $interface_address is
7076 not set up until the daemon has forked a reception process, error responses
7077 that can happen earlier (such as "too many connections") no longer contain
7080 12. If an expansion in a condition on a "warn" statement fails because a lookup
7081 defers, the "warn" statement is abandoned, and the next ACL statement is
7082 processed. Previously this caused the whole ACL to be aborted.
7084 13. Added the iplsearch lookup type.
7086 14. Added ident_timeout as a log selector.
7088 15. Added tls_certificate_verified as a log selector.
7090 16. Added a global option tls_require_ciphers (compare the smtp transport
7091 option of the same name). This controls incoming TLS connections.
7093 17. I finally figured out how to make tls_require_ciphers do a similar thing
7094 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7095 before starting the TLS session.
7097 18. Tabs are now shown as \t in -bP output.
7099 19. If the log selector return_path_on_delivery was set, Exim crashed when
7100 bouncing a message because it had too many Received: header lines.
7102 20. If two routers both had headers_remove settings, and the first one included
7103 a superfluous trailing colon, the final name in the first list and the
7104 first name in the second list were incorrectly joined into one item (with a
7105 colon in the middle).
7111 1. Added -C and -D options to the exinext utility, mainly to make it easier
7112 to include in the automated testing, but these could be helpful when
7113 multiple configurations are in use.
7115 2. The exinext utility was not formatting the output nicely when there was
7116 an alternate port involved in the retry record key, nor when there was a
7117 message id as well (for retries that were specific to a specific message
7118 and a specific host). It was also confused by IPv6 addresses, because of
7119 the additional colons they contain. I have fixed the IPv4 problem, and
7120 patched it up to do a reasonable job for IPv6.
7122 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7123 delivery, the log line now contains "pipelined" if PIPELINING was used.
7125 4. An SMTP transport process used to panic and die if the bind() call to set
7126 an explicit outgoing interface failed. This has been changed; it is now
7127 treated in the same way as a connect() failure.
7129 5. A reference to $sender_host_name in the part of a conditional expansion
7130 that was being skipped was still causing a DNS lookup. This no longer
7133 6. The def: expansion condition was not recognizing references to header lines
7134 that used bh_ and bheader_.
7136 7. Added the _cache feature to named lists.
7138 8. The code for checking quota_filecount in the appendfile transport was
7139 allowing one more file than it should have been.
7141 9. For compatibility with Sendmail, the command line option
7149 and sets the incoming protocol and host name (for trusted callers). The
7150 host name and its colon can be omitted when only the protocol is to be set.
7151 Note the Exim already has two private options, -pd and -ps, that refer to
7152 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7153 "s", but I don't think that's a major issue.
7155 10. A number of refactoring changes to the code, none of which should affect
7158 (a) The number of logging options was getting close to filling up the
7159 32-bit word that was used as a bit map. I have split them into two classes:
7160 those that are passed in the argument to log_write(), and those that are
7161 only ever tested independently outside of that function. These are now in
7162 separate 32-bit words, so there is plenty of room for expansion again.
7163 There is no change in the user interface or the logging behaviour.
7165 (b) When building, for example, log lines, the code previously used a
7166 macro that called string_cat() twice, in order to add two strings. This is
7167 not really sufficiently general. Furthermore, there was one instance where
7168 it was actually wrong because one of the argument was used twice, and in
7169 one call a function was used. (As it happened, calling the function twice
7170 did not affect the overall behaviour.) The macro has been replaced by a
7171 function that can join an arbitrary number of extra strings onto a growing
7174 (c) The code for expansion conditions now uses a table and a binary chop
7175 instead of a serial search (which was left over from when there were very
7176 few conditions). Also, it now recognizes conditions like "pam" even when
7177 the relevant support is not compiled in: a suitably worded error message is
7178 given if an attempt is made to use such a condition.
7180 11. Added ${time_interval:xxxxx}.
7182 12. A bug was causing one of the ddress fields not to be passed back correctly
7183 from remote delivery subprocesses. The field in question was not being
7184 subsequently used, so this caused to problems in practice.
7186 13. Added new log selectors queue_time and deliver_time.
7188 14. Might have fixed a bug in maildirsizefile handling that threw up
7189 "unexpected character" debug warnings, and recalculated the data
7190 unnecessarily. In any case, I expanded the warning message to give more
7193 15. Added the message "Restricted characters in address" to the statements in
7194 the default ACL that block characters like @ and % in local parts.
7196 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7197 Three changes have been made:
7199 (a) There was a serious bug; a negative response to MAIL caused the whole
7200 recipient domain to be cached as invalid, thereby blocking all messages
7201 to all local parts at the same domain, from all senders. This bug has
7202 been fixed. The domain is no longer cached after a negative response to
7203 MAIL if the sender used is not empty.
7205 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7208 (c) A new callout option, "use_sender" has been added for people who want
7209 the modified behaviour.
7215 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7218 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7219 indeed breaks things for older releases.
7221 3. Added additional logging to the case where there is a problem reading data
7222 from a filter that is running in a subprocess using a pipe, in order to
7223 try to track down a specific problem.
7225 4. Testing facility fudge: when running in the test harness and attempting
7226 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7227 getting "No route to host". Convert this to a timeout.
7229 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7232 6. Some OS don't have socklen_t but use size_t instead. This affects the
7233 fifth argument of getsockopt() amongst other things. This is now
7234 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7235 can be set for individual OS. I have set it for SunOS5, OSF1, and
7236 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7237 some earlier ones do not.
7239 7. Change 4.30/15 was not doing the test caselessly.
7241 8. The standard form for an IPv6 address literal was being rejected by address
7242 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7243 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7244 this, as well as the form without the "IPv6" on the front (but only when
7245 address literals are enabled, of course).
7247 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7249 10. Exim crashed if a message with an empty sender address specified by -f
7250 encountered a router with an errors_to setting. This could be provoked only
7251 by a command such as
7255 where an empty string was supplied; "<>" did not hit this bug.
7257 11. Installed PCRE release 4.5.
7259 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7260 remained set. It is now erased.
7262 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7263 times from message ids (which are base 36 rather than the normal 62).
7265 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7266 were being counted as actual protocol errors, and logged if the log
7267 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7268 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7269 and DATA following a set of rejected RCPTs do not count as protocol errors.
7270 In other words, Exim assumes they were pipelined, though this may not
7271 actually be the case. Of course, in all cases the client gets an
7272 appropriate error code.
7274 15. If a lookup fails in an ACL condition, a message about the failure may
7275 be available; it is used if testing the ACL cannot continue, because most
7276 such messages specify what the cause of the deferral is. However, some
7277 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7278 that caused an old message to be retained and used if a later statement
7279 caused a defer, replacing the real cause of the deferral.
7281 16. If an IP address had so many PTR records that the DNS lookup buffer
7282 was not large enough to hold them, Exim could crash while trying to process
7283 the truncated data. It now detects and logs this case.
7285 17. Further to 4.21/58, another change has been made: if (and only if) the
7286 first line of a message (the first header line) ends with CRLF, a bare LF
7287 in a subsequent header line has a space inserted after it, so as not to
7288 terminate the header.
7290 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7291 unnecessarily, used atoi() instead of strtol(), and didn't check the
7292 termination when getting file sizes from file names by regex.
7294 19. Completely re-implemented the support for maildirsize files, in the light
7295 of a number of problems with the previous contributed implementation
7296 (4.30/29). In particular:
7298 . If the quota is zero, the maildirsize file is maintained, but no quota is
7301 . If the maildir directory does not exist, it is created before any attempt
7302 to write a maildirsize file.
7304 . The quota value in the file is just a cache; if the quota is changed in
7305 the transport, the new value overrides.
7307 . A regular expression is available for excluding directories from the
7310 20. The autoreply transport checks the characters in options that define the
7311 message's headers; it allows continued headers, but it was checking with
7312 isspace() after an embedded newline instead of explicitly looking for a
7315 21. If all the "regular" hosts to which an address was routed had passed their
7316 expiry times, and had not reached their retry times, the address was
7317 bounced, even if fallback hosts were defined. Now Exim should go on to try
7320 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7321 equivalent code in the SMTP transport. Some hosts send humungous responses
7322 to HELO/EHLO, more than 1024 it seems.
7324 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7325 gives compiler warnings in some environments. I've now done it "properly",
7328 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7329 (because of problems with the built-in one) was declared to return uschar *
7330 instead of char *, causing compiler failure.
7332 25. Fixed a file descriptor leak when processing alias/forward files.
7334 26. Fixed a minor format string issue in dbfn.c.
7336 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7338 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7339 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7342 29. When a sender address is verified, it is cached, to save repeating the test
7343 when there is more than one recipient in a message. However, when the
7344 verification involves a callout, it is possible for different callout
7345 options to be set for different recipients. It is too complicated to keep
7346 track of this in the cache, so now Exim always runs a verification when a
7347 callout is required, relying on the callout cache for the optimization.
7348 The overhead is duplication of the address routing, but this should not be
7351 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7352 to be verified with callout=postmaster, and the main callout worked but the
7353 postmaster check failed, the verification correctly failed. However, if a
7354 subsequent RCPT command asked for sender verification *without* the
7355 postmaster check, incorrect caching caused this verification also to fail,
7358 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7359 it was not caching the DNS options (qualify_single, search_parents) that
7360 were used when the lookup failed. A subsequent lookup with different
7361 options therefore always gave the same answer, though there were cases
7362 where it should not have. (Example: a "domains = !$mx_any" option on a
7363 dnslookup router: the "domains" option is always processed without any
7364 widening, but the router might have qualify_single set.) Now Exim uses the
7365 cached value only when the same options are set.
7367 32. Added John Jetmore's "exipick" utility to the distribution.
7369 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7370 than a timeout (e.g. a certificate is required, and is not provided), an
7371 Exim server now closes the connection immediately. Previously it waited for
7372 the client to close - but if the client is SSL, it seems that they each
7373 wait for each other, leading to a delay before one of them times out.
7375 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7376 maintained 0.8.x compatibility because I don't think many are using it, and
7377 it is clearly obsolete.
7379 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7382 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7383 client certificate was expired. A simple patch fixes this, though I don't
7384 understand the full logic of why the verify callback is called multiple
7387 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7388 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7389 which causes problems with some clients (such as the Certicom SSL Plus
7390 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7391 disables the coutermeasure allowing Eudora to connect."
7393 38. Exim was not checking that a write() to a log file succeeded. This could
7394 lead to Bad Things if a log got too big, in particular if it hit a file
7395 size limit. Exim now panics and dies if it cannot write to a log file, just
7396 as it does if it cannot open a log file.
7398 39. Modified OS/Makefile-Linux so that it now contains
7400 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7402 The two -D definitions ensure that Exim is compiled with large file
7403 support, which makes it possible to handle log files that are bigger than
7406 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7407 instance) a domain was checked against a named list that involved a lookup,
7408 causing $domain_data to be set, then another domain was checked against the
7409 same list, then the first domain was re-checked, the value of $domain_data
7410 after the final check could be wrong. In particular, if the second check
7411 failed, it could be set empty. This bug probably also applied to
7414 41. The strip_trailing_dot option was not being applied to the address given
7415 with the -f command-line option.
7417 42. The code for reading a message's header from the spool was incrementing
7418 $received_count, but never initializing it. This meant that the value was
7419 incorrect (doubled) while delivering a message in the same process in which
7420 it was received. In the most common configuration of Exim, this never
7421 happens - a fresh exec is done - but it can happen when
7422 deliver_drop_privilege is set.
7424 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7425 it now includes up to 150 characters of the unexpected data in the log
7428 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7429 and building data strings. The size of both of these buffers was 10 000
7430 bytes - far larger than anybody would *ever* want, thought I. Needless to
7431 say, somebody hit the limit. I have increased the maximum line length to
7432 20 000 and the maximum data length of concatenated lines to 100 000. I have
7433 also fixed two bugs, because there was no checking on these buffers. Tsk,
7434 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7435 buffer is too small.
7437 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7438 lsearch lookups. Now it does.
7440 46. When parsing a route_list item in a manualroute router, a fixed-length
7441 buffer was used for the list of hosts. I made this 1024 bytes long,
7442 thinking that nobody would ever have a list of hosts that long. Wrong.
7443 Somebody had a whole pile of complicated expansion conditions, and the
7444 string was silently truncated, leading to an expansion error. It turns out
7445 that it is easier to change to an unlimited length (owing to other changes
7446 that have happened since this code was originally written) than to build
7447 structure for giving a limitation error. The length of the item that
7448 expands into the list of hosts is now unlimited.
7450 47. The lsearch lookup could not handle data where the length of text line was
7451 more than 4095 characters. Such lines were truncated, leading to shortened
7452 data being returned. It should now handle lines of any length.
7454 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7455 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7458 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7459 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7460 no longer happens in comment lines. A list of replacements is now placed
7461 at the head of all of the source files, except those whose only change is
7462 to replace PERL_COMMAND in the very first #! line.
7464 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7465 messages on the queue, with a bottom-up merge sort, using code contributed
7466 by Michael Haardt. This should make operations like -bp somewhat faster on
7467 large queues. It won't affect queue runners, except when queue_run_in_order
7470 51. Installed eximstats 1.31 in the distribution.
7472 52. Added support for SRV lookups to the dnslookup router.
7474 53. If an ACL referred to $message_body or $message_body_end, the value was not
7475 reset for any messages that followed in the same SMTP session.
7477 54. The store-handling optimization for building very long strings was not
7478 differentiating between the different store pools. I don't think this
7479 actually made any difference in practice, but I've tidied it.
7481 55. While running the routers to verify a sender address, $sender_address
7482 was still set to the sender address. This is wrong, because when routing to
7483 send a bounce to the sender, it would be empty. Therefore, I have changed
7484 it so that, while verifying a sender address, $sender_address is set to <>.
7485 (There is no change to what happens when verifying a recipient address.)
7487 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7488 target A or AAAA records (if not already returned) without resetting the
7489 qualify_single or search_parents options of the DNS resolver. These are
7490 inappropriate in this case because the targets of MX and SRV records must
7491 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7492 target that, when qualified, matched something in the local domain. These
7493 two options are now turned off when doing these lookups.
7495 57. It seems that at least some releases of Reiserfs (which does not have the
7496 concept of a fixed number of inodes) returns zero and not -1 for the
7497 number of available inodes. This interacted badly with check_spool_inodes,
7498 which assumed that -1 was the "no such thing" setting. What I have done is
7499 to check that the total number of inodes is greater than zero before doing
7500 the test of how many are available.
7502 58. When a "warn" ACL statement has a log_message modifier, the message is
7503 remembered, and not repeated. This is to avoid a lot of repetition when a
7504 message has many recipients that cause the same warning to be written.
7505 However, Exim was preserving the list of already written lines for an
7506 entire SMTP session, which doesn't seem right. The memory is now reset if a
7507 new message is started.
7509 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7510 debugging output unless log_rewrite was also set.
7512 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7513 of (int)(handle) when we know that handle contains (void *)(-1).
7515 61. The Exim daemon panic-logs an error return when it closes the incoming
7516 connection. However "connection reset by peer" seems to be common, and
7517 isn't really an error worthy of noting specially, so that particular error
7520 62. When Exim is trying to find all the local interfaces, it used to panic and
7521 die if the ioctl to get the interface flags failed. However, it seems that
7522 on at least one OS (Solaris 9) it is possible to have an interface that is
7523 included in the list of interfaces, but for which you get a failure error
7524 for this call. This happens when the interface is not "plumbed" into a
7525 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7526 failure of the "get flags" call assumes that the interface is down.
7528 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7529 makes life easier for people who are doing arithmetic on fields extracted
7530 from dates, where you often get leading zeros that should not be
7531 interpreted as octal.
7533 64. Added qualify_domain to the redirect router, to override the global
7536 65. If a pathologically long header line contained very many addresses (the
7537 report of this problem mentioned 10 000) and each of them was rewritten,
7538 Exim could use up a very large amount of memory. (It kept on making new
7539 copies of the header line as it rewrote, and never released the old ones.)
7540 At the expense of a bit more processing, the header rewriting function has
7541 been changed so that it no longer eats memory in this way.
7543 66. The generation of the Received: header has been moved from the time that a
7544 message starts to be received, to the time that it finishes. The timestamp
7545 in the Received: header should now be very close to that of the <= log
7546 line. There are two side-effects of this change:
7548 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7549 logged header lines no longer include the local Received: line, because
7550 it has not yet been created. The same applies to a copy of the message
7551 that is returned to a non-SMTP sender when a message is rejected.
7553 (b) When a filter file is tested using -bf, no additional Received: header
7554 is added to the test message. After some thought, I decided that this
7557 This change does not affect the value of $received_for. It is still set
7558 after address rewriting, but before local_scan() is called.
7560 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7562 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7563 gave an unhelpful panic error message, and a defer error. I have managed to
7564 change this behaviour so that it now rejects any supplied certificate,
7565 which seems right, as the list of acceptable certificates is empty.
7567 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7568 gave an unhelpful defer error. I have not managed to make this reject any
7569 supplied certificates, but the error message it gives is "no certificate
7570 supplied", which is not helpful.
7572 70. exigrep's output now also includes lines that are not associated with any
7573 message, but which match the given pattern. Implemented by a patch from
7574 Martin Sluka, which also tidied up the Perl a bit.
7576 71. Recipient callout verification, like sender verification, was using <> in
7577 the MAIL FROM command. This isn't really the right thing, since the actual
7578 sender may affect whether the remote host accepts the recipient or not. I
7579 have changed it to use the actual sender in the callout; this means that
7580 the cache record is now keyed on a recipient/sender pair, not just the
7581 recipient address. There doesn't seem to be a real danger of callout loops,
7582 since a callout by the remote host to check the sender would use <>.
7583 [SEE ABOVE: changed after hitting problems.]
7585 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7586 temporary errors. However, in the case of such a code being given after
7587 the end of a data transmission (i.e. after ".") Exim was failing to write
7588 a retry record for the message. (Yes, there was some broken host that was
7589 actually sending 8xx at this point.)
7591 73. An unknown lookup type in a host list could cause Exim to panic-die when
7592 the list was checked. (An example that provoked this was putting <; in the
7593 middle of a list instead of at the start.) If this happened during a DATA
7594 ACL check, a -D file could be left lying around. This kind of configuration
7595 error no longer causes Exim to die; instead it causes a defer error. The
7596 incident is still logged to the main and panic logs.
7598 74. Buglet left over from Exim 3 conversion. The message "too many messages
7599 in one connection" was written to the rejectlog but not the mainlog, except
7600 when address rewriting (yes!) was being logged.
7602 75. Added write_rejectlog option.
7604 76. When a system filter was run not as root (that is, when system_filter_user
7605 was set), the values of the $n variables were not being returned to the
7606 main process; thus, they were not subsequently available in the $sn
7609 77. Added +return_path_on_delivery log selector.
7611 78. A connection timeout was being treated differently from recipients deferred
7612 when testing hosts_max_try with a message that was older than the host's
7613 retry timeout. (The host should not be counted, thus allowing all hosts to
7614 be tried at least once before bouncing.) This may have been the cause of an
7615 occasionally reported bug whereby a message would remain on the queue
7616 longer than the retry timeout, but would be bounced if a delivery was
7617 forced. I say "may" because I never totally pinned down the problem;
7618 setting up timeout/retry tests is difficult. See also the next item.
7620 79. The ultimate address timeout was not being applied to errors that involved
7621 a combination of host plus message (for example, a timeout on a MAIL
7622 command). When an address resolved to a number of possible hosts, and they
7623 were not all tried for each delivery (e.g. because of hosts_max_try), a
7624 message could remain on the queue longer than the retry timeout.
7626 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7629 81. Fixed an obscure SMTP outgoing bug which required at least the following
7630 conditions: (a) there was another message waiting for the same server;
7631 (b) the server returned 5xx to all RCPT commands in the first message so
7632 that the message was not completed; (c) the server dropped the connection
7633 or gave a negative response to the RSET that Exim sends to abort the
7634 transaction. The observed case was a dropped connection after DATA that had
7635 been sent in pipelining mode. That is, the server had advertised PIPELINING
7636 but was not implementing it correctly. The effect of the bug was incorrect
7637 behaviour, such as trying another host, and this could lead to a crash.
7643 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7644 and daemon.c were passed as pointers to ints; they should have been
7645 pointers to socklen_t variables (which are typically unsigned ints).
7647 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7650 3. Fixed a really odd bug that affected only the testing scheme; patching a
7651 certain fixed string in the binary changed the value of another string that
7652 happened to be identical to the end of the original first string.
7654 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7655 name", it returns that address as the IP address. On some operating
7656 systems (e.g. Solaris), it also passes back the IP address string as the
7657 "host name". However, on others (e.g. Linux), it passes back an empty
7658 string. Exim wasn't checking for this, and was changing the host name to an
7659 empty string, assuming it had been canonicalized.
7661 5. Although rare, it is permitted to have more than one PTR record for a given
7662 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7663 all the names associated with an address, because they do in Solaris.
7664 However, it seems that they do not in Linux for data that comes from the
7665 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7666 I found this out when I moved to a new Linux workstation and tried to run
7667 the Exim test suite.
7669 To get round this problem I have changed the code so that it now does its
7670 own call to the DNS to look up PTR records when searching for a host name.
7671 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7672 addresses that are only in /etc/hosts are still found.
7674 This behaviour is, however, controlled by an option called host_lookup_
7675 order, which defaults to "bydns:byaddr". If people want to use the other
7676 order, or indeed, just use one or the other means of lookup, they can
7677 specify it in this variable.
7679 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7680 some operating systems, this comes back from gethostbyaddr() as an empty
7681 string, and this is what Exim used to test for. However, it seems that in
7682 other systems, "." is yielded. Exim now tests for this case too.
7684 7. The values of check_spool_space and check_log_space are now held internally
7685 as a number of kilobytes instead of an absolute number of bytes. If a
7686 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7687 kilobyte. This means that much larger values can be stored.
7689 8. Exim monitor: an attempt to get the action menu when not actually pointing
7690 at a message produces an empty menu entitled "No message selected". This
7691 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7692 no entries in it ("Shell widget menu has zero width and/or height"). So I
7693 have added a single, blank menu entry in this case.
7695 9. Added ${quote_local_part.
7697 10. MIME decoding is now applied to the contents of Subject: header lines when
7700 11. Now that a reference to $sender_host_address automatically causes a reverse
7701 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7702 host lookup before query-style lookups in lists that might use this
7703 variable. This has therefore been abolished, and the "net-" prefix is no
7704 longer necessary for query-style lookups.
7706 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7707 have been a typo for LFLAGS, so it has been changed.
7709 13. The install script calls Exim with "-C /dev/null" in order to find the
7710 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7711 to be output. However, since Exim outputs its version number before the
7712 error, it didn't break the script. It just looked ugly. I fixed this by
7713 always allowing "-C /dev/null" if the caller is root.
7715 14. Ignore overlarge ACL variable number when reading spool file - insurance
7716 against a later release with more variables having written the file.
7718 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7719 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7720 this, as well as the form without the "IPv6" on the front.
7722 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7723 OS/Makefile-Darwin file.
7725 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7726 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7728 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7729 variable. However, some OS use ints for inodes. I've added cast to long int
7730 to get rid of the compiler warning.
7732 19. I had forgotten to lock out "/../" in configuration file names when
7733 ALT_CONFIG_PREFIX was set.
7735 20. Routers used for verification do not need to specify transports. However,
7736 if such a router generated a host list, and callout was configured, Exim
7737 crashed, because it could not find a port number from the (non-existent)
7738 transport. It now assumes port 25 in this circumstance.
7740 21. Added the -t option to exigrep.
7742 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7743 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7744 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7745 them is set, LOOKUP_LSEARCH is forced.
7747 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7749 24. Added sender and host information to the "rejected by local_scan()" log
7750 line; previously there was no indication of these.
7752 25. Added .include_if_exists.
7754 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7755 Linux. It turns out that not all file systems support this. Apparently some
7756 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7757 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7758 not supported on the file descriptor, is now ignored when Exim is trying to
7759 sync a directory. This applies only to Linux.
7761 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7763 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7764 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7767 29. Added support for maildirsize files from supplied patch (modified a bit).
7769 30. The use of :fail: followed by an empty string could lead Exim to respond to
7770 sender verification failures with (e.g.):
7772 550 Verification failed for <xxx>
7773 550 Sender verify failed
7775 where the first response line was missing the '-' that indicates it is not
7776 the final line of the response.
7778 31. The loop for finding the name of the user that called Exim had a hardwired
7779 limit of 10; it now uses the value of finduser_retries, which is used for
7780 all other user lookups.
7782 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7785 33. Exim was neglecting to zero errno before one call of strtol() when
7786 expanding a string and expecting an integer value. On some systems this
7787 resulted in spurious "integer overflow" errors. Also, it was casting the
7788 result into an int without checking.
7790 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7791 did not work. The code looks as if it has *never* worked, though it appears
7792 to have been documented since at least release 1.62. I have made it work.
7794 35. The "timeout_DNS" error in retry rules, also documented since at least
7795 1.62, also never worked. As it isn't clear exactly what this means, and
7796 clearly it isn't a major issue, I have abolished the feature by treating it
7797 as "timeout", and writing a warning to the main and panic logs.
7799 36. The display of retry rules for -brt wasn't always showing the error code
7802 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7803 timeout_connect_A, timeout_connect_MX.
7805 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7806 to the empty sender.
7808 39. The daemon was not analysing the content of -oX till after it had closed
7809 stderr and disconnected from the controlling terminal. This meant that any
7810 syntax errors were only noted on the panic log, and the return code from
7811 the command was 0. By re-arranging the code a little, I've made the
7812 decoding happen first, so such errors now appear on stderr, and the return
7813 code is 1. However, the actual setting up of the sockets still happens in
7814 the disconnected process, so errors there are still only recorded on the
7817 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7818 connections (as happens on some IP stacks) was logged at start up time as
7819 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7820 it from "IPv6 and IPv4", which means that two separate sockets are being
7823 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7824 says whether AF_INET or AF_INET6 was passed as an argument.
7826 42. Exiwhat output was messed up when time zones were included in log
7829 43. Exiwhat now gives more information about the daemon's listening ports,
7830 and whether -tls-on-connect was used.
7832 44. The "port" option of the smtp transport is now expanded.
7834 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7835 silently ignored. Now an error message is written to the main and panic
7838 46. There's a new ACL modifier called "logwrite" which writes to a log file
7839 as soon as it is encountered.
7841 47. Added $local_user_uid and $local_user_gid at routing time.
7843 48. Exim crashed when trying to verify a sender address that was being
7846 49. Exim was recognizing only a space character after ".include". It now also
7847 recognizes a tab character.
7849 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7850 extracting the relevant information from the specification. The man page no
7851 longer contains scrambled data for the -d option, and I've added a section
7852 at the front about calling Exim under different names.
7854 51. Added "extra_headers" argument to the "mail" command in filter files.
7856 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7859 53. Installed eximstats 1.29.
7861 54. Added transport_filter_timeout as a generic transport option.
7863 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7864 Cc: header lines. This was required by RFC 822, but it not required by RFC
7867 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7868 incoming messages that did not have them. Now it does so only if the
7869 message originates locally, that is, if there is no associated remote host
7870 address. When Resent- header lines are present, this applies to the Resent-
7871 lines rather than the non-Resent- lines.
7873 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7874 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7876 58. Messages for configuration errors now include the name of the main
7877 configuration file - useful now that there may be more than one file in a
7878 list (.included file names were always shown).
7880 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7881 for those rare installations that do not start the daemon as root or run it
7882 setuid root. I've cut out the call to initgroups() if the daemon is not
7885 60. The Exim user and group can now be bound into the binary as text strings
7886 that are looked up at the start of Exim's processing.
7888 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7890 62. Added $mailstore_basename variable.
7892 63. Installed patch to sieve.c from Michael Haardt.
7894 64. When Exim failed to open the panic log after failing to open the main log,
7895 the original message it was trying to log was written to stderr and debug
7896 output, but if they were not available (the usual case in production), it
7897 was lost. Now it is written to syslog before the two lines that record the
7898 failures to open the logs.
7900 65. Users' Exim filters run in subprocesses under the user's uid. It is
7901 possible for a "deliver" command or an alias in a "personal" command to
7902 provoke an address rewrite. If logging of address rewriting is configured,
7903 this fails because the process is not running as root or exim. There may be
7904 a better way of dealing with this, but for the moment (because 4.30 needs
7905 to be released), I have disabled address rewrite logging when running a
7906 filter in a non-root, non-exim process.
7912 1. The buildconfig auxiliary program wasn't quoting the value set for
7913 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7914 not defined. This bug was masked in 4.22 by the effect that was fixed in
7917 2. Some messages that were rejected after a message id was allocated were
7918 shown as "incomplete" by exigrep. It no longer does this for messages that
7919 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7921 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7922 have allow_domain_literals set, the ID did not get logged in the <= line.
7923 Domain literals are now always recognized in Message-ID: header lines.
7925 4. The first argument for a ${extract expansion item is the key name or field
7926 number. Leading and trailing spaces in this item were not being ignored,
7927 causing some misleading effects.
7929 5. When deliver_drop_privilege was set, single queue runner processes started
7930 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7931 same command in the process it spins off) were not dropping privilege.
7933 6. When the daemon running as "exim" started a queue runner, it always
7934 re-executed Exim in the spun-off process. This is a waste of effort when
7935 deliver_drop_privilege is set. The new process now just calls the
7936 queue-runner function directly.
7942 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7945 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7946 ignored. Though the use of -oP was forcing the writing of a pid file, it
7947 was always written to the default place.
7949 3. If the message "no IP address found for host xxxx" is generated during
7950 incoming verification, it is now followed by identification of the incoming
7951 connection (so you can more easily find what provoked it).
7953 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7955 5. Added some features to "harden" Exim a bit more against certain attacks:
7957 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7958 be put in Local/Makefile. This is like the never_users runtime option,
7959 but it cannot be overridden. The default setting is "root".
7961 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7962 prefix string with which any file named in a -C command line option
7965 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7966 is retained for -C and -D only if the caller of Exim is root. Without
7967 it, the exim user may also use -C and -D and retain privilege.
7969 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7970 command line option is disabled.
7972 6. Macro names set by the -D option must start with an upper case letter, just
7973 like macro names defined in the configuration file.
7975 7. Added "dereference=" facility to LDAP.
7977 8. Two instances of the typo "uknown" in the source files are fixed.
7979 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7980 the Configure-Makefile script screwed up while processing it.
7982 10. Incorporated PCRE 4.4.
7984 11. The SMTP synchronization check was not operating right at the start of an
7985 SMTP session. For example, it could not catch a HELO sent before the client
7986 waited for the greeting. There is now a check for outstanding input at the
7987 point when the greeting is written. Because of the duplex, asynchronous
7988 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7989 way, but not yet received, when the check is performed.
7991 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7992 on TCP/IP sockets, because this apparently causes some broken clients to
7995 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7996 unchanged) from the Cygwin maintainer.
7998 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7999 was supported instead of testing for mbx. Effectively a typo.
8001 15. The spa authenticator server code was not checking that the input it
8002 received was valid base64.
8004 16. The debug output line for the "set" modifier in ACLs was not showing the
8005 name of the variable that was being set.
8007 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8009 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8010 lookup on its own. Something else was needed to trigger the lookup. For
8011 example, a match in host_lookup or the need for a host name in a host list.
8012 Now, if $sender_host_name is referenced and the host name has not yet been
8013 looked up, a lookup is performed. If the lookup fails, the variable remains
8014 empty, and $host_lookup_failed is set to "1".
8016 19. Added "eqi" as a case-independent comparison operator.
8018 20. The saslauthd authentication condition could segfault if neither service
8019 nor realm was specified.
8021 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8022 error message that was logged was misleading, and incoming SMTP
8023 connections were dropped. The message is now more accurate, and temporary
8024 errors are given to SMTP connections.
8026 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8027 (except RSET) with a 421 error, until QUIT is received. However, it was
8028 failing to send a response to QUIT.
8030 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8031 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8032 ACL is now run after the helo_try_verify_hosts code.
8034 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8035 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8036 case-independent, so other case variants are also recognized). Apparently
8037 some systems use these upper case variants.
8039 25. If more than two messages were waiting for the same host, and a transport
8040 filter was specified for the transport, Exim sent two messages over the
8041 same TCP/IP connection, and then failed with "socket operation on non-
8042 socket" when it tried to send the third.
8044 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8046 27. The extern definition of crypt16() in expand.c was not being excluded when
8047 the OS had its own crypt16() function.
8049 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8050 as a preferred synonym for return_size_limit, both as an option and as an
8053 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8055 30. Changed the default configuration ACL to relax the local part checking rule
8056 for addresses that are not in any local domains. For these addresses,
8057 slashes and pipe symbols are allowed within local parts, but the sequence
8058 /../ is explicitly forbidden.
8060 31. SPA server authentication was not clearing the challenge buffer before
8063 32. log_message in a "warn" ACL statement was writing to the reject log as
8064 well as to the main log, which contradicts the documentation and doesn't
8065 seem right (because no rejection is happening). So I have stopped it.
8067 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8068 However, I am unable to do any testing of this.
8070 34. Fixed an infelicity in the appendfile transport. When checking directories
8071 for a mailbox, to see if any needed to be created, it was accidentally
8072 using path names with one or more superfluous leading slashes; tracing
8073 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8075 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8076 contents are added to the log line that is written for every discarded
8077 recipient. (Previously a log_message setting was ignored.)
8079 36. The ${quote: operator now quotes the string if it is empty.
8081 37. The install script runs exim in order to find its version number. If for
8082 some reason other than non-existence or emptiness, which it checks, it
8083 could not run './exim', it was installing it with an empty version number,
8084 i.e. as "exim-". This error state is now caught, and the installation is
8087 38. An argument was missing from the function that creates an error message
8088 when Exim fails to connect to the socket for saslauthd authentication.
8089 This could cause Exim to crash, or give a corrupted message.
8091 39. Added isip, isip4, and isip6 to ${if conditions.
8093 40. The ACL variables $acl_xx are now saved with the message, and can be
8094 accessed later in routers, transports, and filters.
8096 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8097 strings in the file are not string-expanded.
8099 42. If a MAIL command specified a SIZE value that was too large to fit into an
8100 int variable, the check against message_size_limit failed. Such values are
8101 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8102 day this will have to be increased, but I don't think I want to be around
8103 when emails are that large.
8110 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8111 iconv() is not standard in FreeBSD.
8113 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8114 IPv6 enabled. The observed symptom was a segmentation fault on return from
8115 the function os_common_find_running_interfaces() in src/os.c.
8117 3. In the check_special_case() function in daemon.c I had used "errno" as an
8118 argument name, which causes warnings on some systems. This was basically a
8119 typo, since it was named "eno" in the comments!
8121 4. The code that waits for the clock to tick (at a resolution of some fraction
8122 of a second) so as to ensure message-id uniqueness was always waiting for
8123 at least one whole tick, when it could have waited for less. [This is
8124 almost certainly not relevant at current processor speeds, where it is
8125 unlikely to ever wait at all. But we try to future-proof.]
8127 5. The function that sleeps for a time interval that includes fractions of a
8128 second contained a race. It did not block SIGALRM between setting the
8129 timer, and suspending (a couple of lines later). If the interval was short
8130 and the sigsuspend() was delayed until after it had expired, the suspension
8131 never ended. On busy systems this could lead to processes getting stuck for
8134 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8135 process, before it forks any delivery processes. The open lookup caching
8136 mechanism meant that the open file or database connection was passed into
8137 the delivery process. The problem was that delivery processes always tidy
8138 up cached lookup data. This could cause a problem for the next delivery
8139 process started by the queue runner, because the external queue runner
8140 process does not know about the closure. So the next delivery process
8141 still has data in the lookup cache. In the case of a file lookup, there was
8142 no problem because closing a file descriptor in a subprocess doesn't affect
8143 the parent. However, if the lookup was caching a connection to a database,
8144 the connection was closed, and the second delivery process was likely to
8145 see errors such as "PGSQL: query failed: server closed the connection
8146 unexpectedly". The problem has been fixed by closing all cached lookups
8147 in a queue runner before running a delivery process.
8149 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8150 seem to have the "const" qualifier which it has on other OS. I've
8153 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8154 *of the same type* (client or server) with the same public name that an
8155 error should be diagnosed.
8157 9. When Exim looked up a host name for an IP address, but failed to find the
8158 original IP address when looking up the host name (a safety check), it
8159 output the message "<ip address> does not match any IP for NULL", which was
8160 confusing, to say the least. The bug was that the host name should have
8161 appeared instead of "NULL".
8163 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8164 user that is built into the binary, and the -C or -D options is used, root
8165 privilege is dropped before the configuration file is read. In addition,
8166 logging is switched to stderr instead of the normal log files. If the
8167 configuration then re-defines the Exim user, the unprivileged environment
8168 is probably not what is expected, so Exim logs a panic warning message (but
8171 However, if deliver_drop_privilege is set, the unprivileged state may well
8172 be exactly what is intended, so the warning has been cut out in that case,
8173 and Exim is allowed to try to write to its normal log files.
8179 1. smtp_return_error_details was not giving details for temporary sender
8180 or receiver verification errors.
8182 2. Diagnose a configuration error if two authenticators have the same public
8185 3. Exim used not to create the message log file for a message until the first
8186 delivery attempt. This could be confusing when incoming messages were held
8187 for policy or load reasons. The message log file is now created at the time
8188 the message is received, and an initial "Received" line is written to it.
8190 4. The automatically generated man page for command line options had a minor
8191 bug that caused no ill effects; however, a more serious problem was that
8192 the procedure for building the man page automatically didn't always
8193 operate. Consequently, release 4.20 contains an out-of-date version. This
8194 shouldn't happen again.
8196 5. When building Exim with embedded Perl support, the script that builds the
8197 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8198 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8200 6. The freeze_tell option was not being used for messages that were frozen on
8201 arrival, either by an ACL or by local_scan().
8203 7. Added the smtp_incomplete_transaction log selector.
8205 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8206 was accepting AUTH without a new EHLO.
8208 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8209 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8212 10. Logging of TCP/IP connections (when configured) now happens in the main
8213 daemon process instead of the child process, so that the TCP/IP connection
8214 count is more accurate (but it can never be perfect).
8216 11. The use of "drop" in a nested ACL was not being handled correctly in the
8217 outer ACL. Now, if condition failure induced by the nested "drop" causes
8218 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8219 or "require"), the connection is dropped.
8221 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8222 that yield "discard" can now be used with an "accept" or a "discard" verb,
8223 but an error is generated for any others (because I can't see a useful way
8224 to define what should happen).
8226 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8227 are now processed in the same way as lines in the Exim configuration file.
8228 In particular, continuation lines are supported.
8230 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8232 15. Added -ti meaning -t -i.
8234 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8235 domains, and warn by logging if others are found.
8237 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8238 returned from GIFCONF when Exim is trying to find the list of interfaces on
8239 a host. The code in os.c has been modified to copy each ifreq to an aligned
8240 structure in all cases.
8242 Also, in some cases, the returned ifreq's were being copied to a 'struct
8243 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8244 means the last couple of bytes of an IPv6 address could be chopped if the
8245 ifreq contained only a normal sockaddr (14 bytes storage).
8247 18. Named domain lists were not supported in the hosts_treat_as_local option.
8248 An entry such as +xxxx was not recognized, and was treated as a literal
8251 19. Ensure that header lines added by a DATA ACL are included in the reject log
8252 if the ACL subsequently rejects the message.
8254 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8255 MD5 (which is deprecated).
8257 21. When testing a filter file using -bf, Exim was writing a message when it
8258 took the sender from a "From " line in the message, but it was not doing so
8259 when it took $return_path from a Return-Path: header line. It now does.
8261 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8262 with a valid header line field name (a series of printing characters
8263 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8265 23. Changed "disc" in the source to "disk" to conform to the documentation and
8266 the book and for uniformity.
8268 24. Ignore Sendmail's -Ooption=value command line item.
8270 25. When execve() failed while trying to run a command in a pipe transport,
8271 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8272 could be confused with a return value of 69 from the command itself. This
8273 has been changed to 127, the value the shell returns if it is asked to run
8274 a non-existent command. The wording for the related log line suggests a
8275 non-existent command as the problem.
8277 26. If received_header_text expands to an empty string, do not add a Received:
8278 header line to the message. (Well, it adds a token one on the spool, but
8279 marks it "old" so that it doesn't get used or transmitted.)
8281 27. Installed eximstats 1.28 (addition of -nt option).
8283 28. There was no check for failure on the call to getsockname() in the daemon
8284 code. This can fail if there is a shortage of resources on the system, with
8285 ENOMEM, for example. A temporary error is now given on failure.
8287 29. Contrary to the C standard, it seems that in some environments, the
8288 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8289 program. Exim now does this explicitly; it affects the formatting of
8290 timestamps using strftime().
8292 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8293 complaints. I've now initialized all the variables, to avoid this.
8295 32. Header lines added by a system filter were not being "seen" during
8296 transport-time rewrites.
8298 33. The info_callback() function passed to OpenSSL is set up with type void
8299 (*)(SSL *, int, int), as described somewhere. However, when calling the
8300 function (actually a macro) that sets it up, the type void(*)() is
8301 expected. I've put in a cast to prevent warnings from picky compilers.
8303 34. If a DNS black list lookup found a CNAME record, but there were no A
8304 records associated with the domain it pointed at, Exim crashed.
8306 35. If a DNS black list lookup returned more than one A record, Exim ignored
8307 all but the first. It now scans all returned addresses if a particular IP
8308 value is being sought. In this situation, the contents of the
8309 $dnslist_value variable are a list of all the addresses, separated by a
8312 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8313 to accept a host name and all its aliases if the forward lookup for any of
8314 them yielded the IP address of the incoming connection. Now it accepts only
8315 those names whose forward lookup yields the correct IP address. Any other
8316 names are discarded. This closes a loophole whereby a rogue DNS
8317 administrator could create reverse DNS records to break through a
8318 wildcarded host restriction in an ACL.
8320 37. If a user filter or a system filter that ran in a subprocess used any of
8321 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8322 the wrong values were passed to the pipe command ($thisaddress had the
8323 value of $0, $0 had the value of $1, etc). This bug was introduced by
8324 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8326 38. Improved the line breaking for long SMTP error messages from ACLs.
8327 Previously, if there was no break point between 40 and 75 characters, Exim
8328 left the rest of the message alone. Two changes have been made: (a) I've
8329 reduced the minimum length to 35 characters; (b) if it can't find a break
8330 point between 35 and 75 characters, it looks ahead and uses the first one
8331 that it finds. This may give the occasional overlong line, but at least the
8332 remaining text gets split now.
8334 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8335 file descriptors might be low, and that setting 1000 would always raise it.
8336 It turns out that in some environments, the limit is already over 1000 and
8337 that lowering it causes trouble. So now Exim takes care not to decrease it.
8339 40. When delivering a message, the value of $return_path is set to $sender_
8340 address at the start of routing (routers may change the value). By an
8341 oversight, this default was not being set up when an address was tested by
8342 -bt or -bv, which affected the outcome if any router or filter referred to
8345 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8346 log only when "message" or "log_message" are set. However, if one of the
8347 conditions was an address verification, or a call to a nested ACL, the
8348 messages generated by the underlying test were being passed through. This
8349 no longer happens. The underlying message is available in $acl_verify_
8350 message for both "message" and "log_message" expansions, so it can be
8351 passed through if needed.
8353 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8354 new expansion $bh_ to give the encoded byte string without charset
8355 translation. Translation happens only if iconv() is available; HAVE_ICONV
8356 indicates this at build time. HEADERS_CHARSET gives the charset to
8357 translate to; headers_charset can change it in the configuration, and
8358 "headers charset" can change it in an individual filter file.
8360 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8361 that creates RFC 2047 encoded "words" labels them as that charset instead
8362 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8363 expansion operator; (ii) when Exim creates a From: line for a local
8364 message; (iii) when a header line is rewritten to include a "phrase" part.
8366 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8367 buggy, causing it to skip the first lines of messages whose message ID
8368 ended in 'D'. This would not have bitten before Exim release 4.14, because
8369 message IDs were unlikely to end in 'D' before then. The effect was to have
8370 incorrect size information for certain domains.
8372 45. #include "config.h" was missing at the start of the crypt16.c module. This
8373 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8376 46. If there was a timeout during a "random" callout check, Exim treated it as
8377 a failure of the random address, and carried on sending RSET and the real
8378 address. If the delay was just some slowness somewhere, the response to the
8379 original RCPT would be taken as a response to RSET and so on, causing
8380 mayhem of various kinds.
8382 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8383 when I implemented it. It didn't allow for the fact that some option values
8384 may legitimately be negative (e.g. size_addition), and it didn't even do
8385 the right test for positive values.
8387 48. Domain names in DNS records are case-independent. Exim always looks them up
8388 in lower case. Some resolvers return domain names in exactly the case they
8389 appear in the zone file, that is, they may contain uppercase letters. Not
8390 all resolvers do this - some return always lower case. Exim was treating a
8391 change of case by a resolver as a change of domain, similar to a widening
8392 of a domain abbreviation. This triggered its re-routing code and so it was
8393 trying to route what was effectively the same domain again. This normally
8394 caused routing to fail (because the router wouldn't handle the domain
8395 twice). Now Exim checks for this case specially, and just changes the
8396 casing of the domain that it ultimately uses when it transmits the message
8399 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8402 50. If a filter generated a file delivery with a non-absolute name (possible if
8403 no home directory exists for the router), the forbid_file option was not
8406 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8407 the existing equality matching.
8409 52. Exim was using ints instead of ino_t variables in some places where it was
8410 dealing with inode numbers.
8412 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8413 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8414 called TMPDIR, and if it finds it is different, it changes its value.
8416 54. The smtp_printf() function is now made available to local_scan() so
8417 additional output lines can be written before returning. There is also an
8418 smtp_fflush() function to enable the detection of a dropped connection.
8419 The variables smtp_input and smtp_batched_input are exported to
8422 55. Changed the default runtime configuration: the message "Unknown user"
8423 has been removed from the ACL, and instead placed on the localuser router,
8424 using the cannot_route_message feature. This means that any verification
8425 failures that generate their own messages won't get overridden. Similarly,
8426 the "Unrouteable address" message that was in the ACL for unverifiable
8427 relay addresses has also been removed.
8429 56. Added hosts_avoid_esmtp to the smtp transport.
8431 57. The exicyclog script was not checking for the esoteric option
8432 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8433 will work only if exicyclog is run under the appropriate euid.
8435 58. Following a discussion on the list, the rules by which Exim recognises line
8436 endings on incoming messages have been changed. The -dropcr and drop_cr
8437 options are now no-ops, retained only for backwards compatibility. The
8438 following line terminators are recognized: LF CRLF CR. However, special
8439 processing applies to CR:
8441 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8442 nor a local message in the state where . is a terminator.
8444 (ii) If a bare CR is encountered in a header line, an extra space is added
8445 after the line terminator so as not to end the header. The reasoning
8446 behind this is that bare CRs in header lines are most likely either
8447 to be mistakes, or people trying to play silly games.
8449 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8450 was being incorrectly given as 18 bytes larger than it should have been.
8451 This is a VOB (very old bug).
8453 60. This may never have affected anything current, but just in case it has:
8454 When the local host is found other than at the start of a list of hosts,
8455 the local host, those with the same MX, and any that follow, are discarded.
8456 When the list in question was part of a longer list of hosts, the following
8457 hosts (not currently being processed) were also being discarded. This no
8458 longer happens. I'm not sure if this situation could ever has previously
8461 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8464 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8465 header. This is something that is recommended in a new Internet Draft, and
8466 is something that is documented as being done by Sendmail. There are two
8467 possible values. For messages generated by the autoreply transport, Exim
8470 Auto-Submitted: auto-replied
8472 whereas for all other generated messages (e.g. bounces) it adds
8474 Auto-Submitted: auto-generated
8476 63. The "personal" condition in filters now includes a test for the
8477 Auto-Submitted: header. If it contains the string "auto-" the message it
8478 not considered personal.
8480 64. Added rcpt_include_affixes as a generic transport option.
8482 65. Added queue_only_override (default true).
8484 66. Added the syslog_duplication option.
8486 67. If what should have been the first header line of a message consisted of
8487 a space followed by a colon, Exim was mis-interpreting it as a header line.
8488 It isn't of course - it is syntactically invalid and should therefore be
8489 treated as the start of the message body. The misbehaviour could have
8490 caused a number of strange effects, including loss of data in subsequent
8491 header lines, and spool format errors.
8493 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8494 client host had authenticated. This control can now be exercised by an ACL
8495 for more flexibility.
8497 69. By default, callouts do not happen when testing with -bh. There is now a
8498 variant, -bhc, which does actually run the callout code, including
8499 consulting and updating the callout cache.
8501 70. Added support for saslauthd authentication, courtesy of Alexander
8504 71. If statvfs() failed on the spool or log directories while checking their
8505 size for availability, Exim confusingly gave the error "space shortage".
8506 Furthermore, in debugging mode it crashed with a floating point exception.
8507 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8508 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8509 serious problem, Exim now writes to the main and panic logs when this
8510 happens, with details of the failure. It then refuses to accept the
8511 incoming message, giving the message "spool directory problem" or "log
8512 directory problem" with a 421 code for SMTP messages.
8514 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8515 0, 1, and 2 exist, because some OS complain for execs without them (see
8516 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8517 descriptors. However, the code omitted to check that the open succeeded,
8518 causing mysterious errors if for some reason the permissions on /dev/null
8519 got screwed. Now Exim writes a message to the main and panic logs, and
8520 bombs out if it can't open /dev/null.
8522 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8523 interact so that it is all more flexible. It is supposed to remain
8524 backwards compatible. Also added extra_local_interfaces.
8526 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8527 to bomb out with an assertion failure - to the client this appears as a
8528 connection drop. This problem occurs in the part of the code that was taken
8529 from the Samba project. Fortunately, the assertion is in a very simple
8530 function, so I have fixed this by reproducing the function inline in the
8531 one place where it is called, and arranging for authentication to fail
8532 instead of killing the process with assert().
8534 75. The SPA client code was not working when the server requested OEM rather
8535 than Unicode encoding.
8537 76. Added code to make require_files with a specific uid setting more usable in
8538 the case where statting the file as root fails - usually a non-root-mounted
8539 NFS file system. When this happens and the failure is EACCES, Exim now
8540 forks a subprocess and does the per-uid checking as the relevant uid.
8542 77. Added process_log_path.
8544 78. If log_file_path was not explicitly set, a setting of check_log_space or
8545 check_log_inodes was ignored.
8547 79. If a space check for the spool or log partitions fails, the incident is now
8548 logged. Of course, in the latter case the data may get lost...
8550 80. Added the %p formatting code to string_format() so that it can be used to
8551 print addresses in debug_print(). Adjusted all the address printing in the
8552 debugging in store.c to use %p rather than %d.
8554 81. There was a concern that a line of code in smtp_in.c could overflow a
8555 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8556 initially expressed, the concern was not well-founded, because trailing
8557 spaces are removed early. However, if the trailing spaces were followed by
8558 a NULL, they did not get removed, so the overflow was possible. Two fixes
8561 (a) I re-wrote the offending code in a cleaner fashion.
8562 (b) If an incoming SMTP command contains a NULL character, it is rejected
8565 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8566 runs initgroups(), so that if the Exim user is in any additional groups,
8567 they will be used during message reception.
8573 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob