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
80 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
81 SMTP connection" log lines.
83 JH/02 Option default value updates:
84 - queue_fast_ramp (main) true (was false)
85 - remote_max_parallel (main) 4 (was 2)
87 JH/03 Cache static regex pattern compilations, for use by ACLs.
89 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
90 Make the rewrite never match and keep the logging. Trust the
91 admin to be using verify=header-syntax (to actually reject the message).
93 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
94 (under Linux) we watch the dir containing the final file; previously
95 it would be the dir with the first symlink. We still do not monitor
98 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
99 dn_expand() is happy to pass them through.
101 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
102 the file with the proof had an unchanged name, the new proof(s) were
103 loaded on top of the old ones (and nover used; the old ones were stapled).
105 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
106 more than one message arrived in a single connection a reference from
107 the earlier message could be re-used. Often a sigsegv resulted.
108 These variables were introduced in Exim 4.87.
109 Debug help from Graeme Fowler.
111 JH/09 Fix ${filter } for conditions that modify $value. Previously the
112 modified version would be used in construction the result, and a memory
115 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
116 Find and fix by Jasen Betts.
118 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
119 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
120 the systemwide configuration to override the Exim config.
122 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
123 API changes in libopendmarc.
125 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
126 pid 1, in the normal "background daemon" mode, having to drop process-
127 group leadership also lost track of needing to create listener sockets.
129 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
130 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
132 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
133 a capture group which obtained no text (eg. "(abc)*" matching zero
134 occurrences) could cause a segfault if the corresponding $<n> was
137 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
138 included a close-brace character (eg. it itself used an expansion) an
141 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
142 starting TLS. Previously it was after, meaning that attackers on such
143 ports had to be screened using the host_reject_connection main config
144 option. The new sequence aligns better with the STARTTLS behaviour, and
145 permits defences against crypto-processing load attacks, even though it
146 is strictly an incompatible change.
147 Also, avoid sending any SMTP fail response for either the connect ACL
148 or host_reject_connection, for TLS-on-connect ports.
150 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
151 Previously this was not permitted, but it makes reasonable sense.
152 While there, restore a restriction on using it from a connect ACL; given
153 the change JH/16 it could only return false (and before 4.91 was not
156 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
157 was exactly sized compared to the log buffer, a crash occurred with the
158 misleading message "bad memory reference; pool not found".
159 Found and traced by Jasen Betts.
161 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
162 dns_again_means_nonexist included an element causing a DNS lookup which
163 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
164 included (though probably not limited to) a process crash from stack
165 memory limit, or from excessive open files. Replace this with a paniclog
166 whine (as this is likely a configuration error), and returning
169 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
170 this always failed, probably leading to the usual downgrade to in-clear
173 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
174 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
175 to not checking that list for these lookups.
177 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
180 JH/23 Fix crash in string expansions. Previously, if an empty variable was
181 immediately followed by an expansion operator, a null-indirection read
182 was done, killing the process.
184 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
185 include an SMTP response string which is longer than that supported
186 by the delivering transport. Alleviate by wrapping such lines before
189 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
190 chars (RFC limit). Previously a limit of 12 items was made, which with
191 a not-impossible References: in the message being bounced could still
192 be over-large and get stopped in the transport.
194 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
195 close. Previously a bare socket close was done.
197 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
200 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
201 message being considered for delivery was already being handled by
202 another process, and having an SMTP connection already open, the function
203 to close it tried to use an uninitialized variable. This would afftect
204 high-volume sites more, especially when running mailing-list-style loads.
205 Pollution of logs was the major effect, as the other process delivered
206 the message. Found and partly investigated by Graeme Fowler.
208 JH/29 Change format of the internal ID used for message identification. The old
209 version only supported 31 bits for a PID element; the new 64 (on systems
210 which can use Base-62 encoding, which is all currently supported ones
211 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
212 and must use Base-36). The new ID is 23 characters rather than 16, and is
213 visible in various places - notably logs, message headers, and spool file
214 names. Various of the ancillary utilities also have to know the format.
215 As well as the expanded PID portion, the sub-second part of the time
216 recorded in the ID is expanded to support finer precision. Theoretically
217 this permits a receive rate from a single comms channel of better than the
219 The major timestamp part of the ID is not changed; at 6 characters it is
220 usable until about year 3700.
221 Updating from previously releases is fully supported: old-format spool
222 files are still usable, and the utilities support both formats. New
223 message will use the new format. The one hints-DB file type which uses
224 message-IDs (the transport wait- DB) will be discarded if an old-format ID
225 is seen; new ones will be built with only new-format IDs.
226 Optionally, a utility can be used to convert spool files from old to new,
227 but this is only an efficiency measure not a requirement for operation
228 Downgrading from new to old requires running a provided utility, having
229 first stopped all operations. This will convert any spool files from new
230 back to old (losing time-precision and PID information) and remove any
231 wait- hints databases.
233 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
234 we treated them as item separators when parsing for a list item, but they
235 need to be protected by the doublequotes. While there, add handling for
238 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
239 Found and fixed by Jasen Betts. No testcase for this as my usual text
240 editor insists on emitting only valid UTF-8.
242 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
243 a null-indirection SIGSEGV for the receive process.
245 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
246 -bh use. Internal checking would spot this and take a panic.
248 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
249 In 4.96 this would expand to empty.
251 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
252 certificate. Find and fix by Andreas Metzler.
254 JH/36 Add ARC info to DMARC hostory records.
256 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
257 or fakedefer. Previously the sender could discover that the message
258 had in fact been accepted.
260 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
261 sequences. Previously the input was not noted as being tainted; notably
262 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
263 bad coding of authenticators.
265 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
266 and ${tr...}. Found and diagnosed by Heiko Schlichting.
268 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
269 could be triggered by externally-supplied input. Found by Trend Micro.
272 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
273 be triggered by externally-controlled input. Found by Trend Micro.
276 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
277 be triggered by externally-controlled input. Found by Trend Micro.
280 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
281 Make the rewrite never match and keep the logging. Trust the
282 admin to be using verify=header-syntax (to actually reject the message).
284 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
287 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
293 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
294 after reception to before a subsequent reception. This should
295 mean slightly faster delivery, and also confirmation of reception
298 JH/02 Move from using the pcre library to pcre2. The former is no longer
299 being developed or supported (by the original developer).
301 JH/03 Constification work in the filters module required a major version
302 bump for the local-scan API. Specifically, the "headers_charset"
303 global which is visible via the API is now const and may therefore
304 not be modified by local-scan code.
306 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
307 sendfile() didi not account for the way the ClamAV driver code called it.
309 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
310 time check was being done for every character; replace that with one
313 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
314 sent was prefixed with a length byte.
316 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
317 RFC 5321. Previously Dovecot (at least) would log errors during
320 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
321 in 4.61 without a reason logged. Binaries will be bigger, which might
322 matter on diskspace-constrained systems, but debug is easier.
324 JH/09 Fix macro-definition during "-be" expansion testing. The move to
325 write-protected store for macros had not accounted for these runtime
326 additions; fix by removing this protection for "-be" mode.
328 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
329 handing out large-numbered file descriptors, violating the usual Unix
330 assumption (and required by Posix) that the lowest possible number will be
331 allocated by the kernel when a new one is needed. In the daemon, and any
332 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
333 are not useable for FD_SET() [and hence select()] and overwrite the stack.
334 Assorted crashes happen.
336 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
337 main-section options or in a connect ACL, the value from the first ever
338 connection was never replaced for subsequent connections. Found by
341 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
342 though only once PCRE2 was introduced: the memory accounting used under
343 debug offset allocations by an int, giving a hard trap in early startup.
344 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
346 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
347 with underbars is given. The write-protection of configuration introduced
348 in 4.95 trapped when normalisation was applied to an option not needing
351 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
353 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
354 failing to close the certificates directory, every hour or any time it
357 JH/16 Debugging initiated by an ACL control now continues through into routing
358 and transport processes. Previously debugging stopped any time Exim
359 re-execs, or for processing a queued message.
361 JH/17 The "expand" debug selector now gives more detail, specifically on the
362 result of expansion operators and items.
364 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
365 bad comparison between the option value and the name of the file to
366 be included was done, and a mismatch was wrongly identified.
367 4.88 to 4.95 are affected.
369 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
371 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
372 supplied as the db file-pair basename. Previously, if a directory
373 path was given, for example via the autoreply "once" option, the DB
374 file.pag and file.dir files would be created in that directory's
377 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
378 "taint" log_selector. These were previously deprecated.
380 JH/22 Fix static address-list lookups to properly return the matched item.
381 Previously only the domain part was returned.
383 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
384 the call into OpenSSL to send a TLS Close was being repeated; this
385 resulted in the library waiting for the peer's Close. If that was never
386 sent we waited forever. Fix by tracking send calls.
388 JH/24 The ${run} expansion item now expands its command string elements after
389 splitting. Previously it was before; the new ordering makes handling
390 zero-length arguments simpler. The old ordering can be obtained by
391 appending a new option "preexpand", after a comma, to the "run".
393 JH/25 Taint-check exec arguments for transport-initiated external processes.
394 Previously, tainted values could be used. This affects "pipe", "lmtp" and
395 "queryprogram" transport, transport-filter, and ETRN commands.
396 The ${run} expansion is also affected: in "preexpand" mode no part of
397 the command line may be tainted, in default mode the executable name
400 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
401 the the facility was not passed across execs, and only the first message
402 passed over a connection could use BDAT; any further ones using DATA.
404 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
405 uses $sending_ip_address and an interface is specified.
406 Previously any use of the local address in the EHLO name disabled
407 PIPECONNECT, the common case being to use the rDNS of it.
409 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
410 resumption. Previously verify failed because no certificate status is
411 passed on the wire for the restarted session. Fix by using the recorded
412 ocsp status of the stored session for the new connection.
414 JH/29 TLS resumption: the key for session lookup in the client now includes
415 more info that a server could potentially use in configuring a TLS
416 session, avoiding oferring mismatching sessions to such a server.
417 Previously only the server IP was used.
419 JH/30 Fix string_copyn() for limit greater than actual string length.
420 Previously the copied amount was the limit, which could result in a
421 overlapping memcpy for newly allocated destination soon after a
422 source string shorter than the limit. Found/investigated by KM.
424 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
425 close; it may be needed for a subsequent connection. This caused a
426 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
428 JH/32 Fix CHUNKING for a second message on a connection when the first was
429 rejected. Previously we did not reset the chunking-offered state, and
430 erroneously rejected the BDAT command. Investigation help from
433 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
434 an empty address. Previously the expansion returned an error.
436 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
437 proxy. Previously these were misparsed, leading to paniclog entries.
443 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
444 related applications. Previously an "H" was used where available info
445 says that "M" should be, so change to match.
447 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
448 as arguments, so an implementation trying to copy these into a local
449 buffer was taking a taint-enforcement trap. Fix by using dynamically
450 created buffers. Similar fix for radius expansion condition.
452 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
453 reasonable, eg. to count headers. Fix by using dynamically created
454 buffers rather than a local. Do similar fixes for ACL actions "dcc",
455 "log_reject_target", "malware" and "spam"; the arguments are expanded
456 so could be handling tainted values.
458 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
459 broken the (no-op) support for this sendmail command. Restore it
460 to doing nothing, silently, and returning good status.
462 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
463 record path was given (or the default used) without a leading directory
464 path, an error occurred on trying to open it. Use the transport's working
467 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
468 transport. Previously it was the name on the DNS A-record; use instead
469 the head of the CNAME chain leading there (if there is one). This seems
470 to align better with RFC 6125.
472 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
473 smtp_accept_max_per_host allocated resources which were not released
474 when the limit was exceeded. This eventually crashed the daemon. Fix
475 by adding a release action in that path.
477 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
478 expanded; previously using tainted values was rejected. Fix by using
479 dynamically-created buffers.
481 JH/09 Relax restrictions on ACL verify condition needing access to message
482 headers. Previously they were only permitted in data and non-smtp ACLs;
483 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
484 not_blind, header_sender and header_names_ascii verification.
486 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
487 Previously a macro used one argument twice; when called with the
488 argument as an expression having side-effects, incorrect operation
489 resulted. Use an inlineable function.
491 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
492 held open for a verify callout. Previously this wan not accounted for
493 and a corrupt onward SMTP conversation resulted.
495 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
496 Previously they were embedded naively in the constructed address; when
497 needed, strip the quoting and quote the entire local_part.
498 Also make the inbound_srs expansion condition handle quoting.
500 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
501 excluded, not matching the documentation.
503 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
504 was given for the sqlite_dbfile a trap resulted.
506 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
507 "name" argument resulted in a trap. There is no reason to disallow such;
508 this was a coding error.
510 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
511 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
512 spent suspended, ignoring the POSIX definition. Previously we assumed
513 it did and a constant offset from real time could be used as a correction.
514 Change to using the same clock source for the start-of-message and the
515 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
516 exists, just to get a clock slightly more aligned to reality.
518 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
519 RFC says it is optional some validators care. The missing char was not
520 intended but triggered by a line-wrap alignment. Discovery and fix by
521 Guillaume Outters, hacked on by JH.
523 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
524 name being quoted was tainted a trap would be taken. Fix by using
525 dynamically created buffers. The routine could have been called by a
526 rewrite with the "h" flag, by using the "-F" command-line option, or
527 by using a "name=" option on a control=submission ACL modifier.
529 JH/19 SPF: change the Authentication-Results expansion component to give
530 smtp.helo when the sender domain is empty. Previously it gave
533 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
534 not in 127.0.0.0/8 to help in spotting list domains taken over by a
535 domain-parking registrar.
537 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
538 Previously when a whitespace character was specified it was not inserted
539 after removing the newline.
541 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
542 the domain part of the recipient address. This overrides any tls_sni
543 option set, which was previously used.
545 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
548 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
549 is_tainted() had an off-by-one error in the overenthusiastic direction.
550 Find and fix by Gavan. Although NetBSD is not a supported platform for
551 4.94 this bug could affect other platforms.
553 PP/01 Fix default prime selection to be consistent.
554 One path used ike23 still, instead of exim.dev.20160529.3; now both
555 execution flows will use the same DH primes (currently
556 exim.dev.20160529.3).
558 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
559 option in smtp transport, to match the documentation. Previously
560 verification was not being done in some cases where it should have been.
562 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
563 than one server was defined and depending on the platform memory layout
564 details, an internal consistency trap could be hit while walking the list
567 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
568 files. The value is used by the authresults expansion item. Previously
569 if this was used in a router or transport, a crash could result.
571 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
572 have a sufficient incidence of EINTR returns from select that an
573 interaction with other operations done by the main daemon loop exposed
574 a bug in the error-handling. This was benign apart from the log
577 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
578 for consistency with delivered "=>" and failed "**" lines. While we're
579 there, handle PRX and TFO.
581 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
582 applied. This resulted, if any header-line rewrite rules were configured,
583 in a panic-log triggerable by sending a message with a long address in
584 a header. Fix by increasing the arbitrary limit to larger than a single
585 (dewrapped) 5322 header line maximum size.
587 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
588 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
589 RFC 6648 which deprecates X- options in protocols as a general practice.
590 Changeover between the implementations is handled by the mechanisms
593 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
594 a remote which uses callout/hold. Previously the local was queued.
596 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
599 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
600 log line. The asterisk is now consistently appended to the remote IP
601 (and port, if given), and will also be provided on defer and fail log
602 lines. Previously it could be placed on the local IP if that was being
603 logged, and was only provided on delivery lines.
605 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
607 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
608 authenticator client_send option. Previously the next char, after a pair
609 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
610 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
611 documentation. There is still no way to get a leading ^ immediately
612 after a NUL (ie. for the password of a PLAIN method authenticator.
614 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
615 files. For bad sizes read, delete the record and whine to paniclog.
617 JH/38 When logging an AUTH failure, as server, do not include sensitive
618 information. Previously, the credentials would be included if given
619 as part of the AUTH command line and an ACL denied authentication.
621 JH/39 Bug 2691: fix $local_part_data. When the matching list element
622 referred to a file, bad data was returned. This likely also affected
625 JH/40 The gsasl authenticator now supports caching of the salted password
626 generated by the client-side implementation. This required the addition
627 of a new variable: $auth4.
629 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
630 left undeleted; the attempt to re-create it then failed - resulting in
631 the usual "SIGHUP tp have daemon reload configuration" to not work.
632 This affected any platform not supporting "abstract" Unix-domain
633 sockets (i.e. not Linux).
635 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
636 recipients" response to RCPT in a later response, with a 250. The
637 previous coding assumed this would not happen, and under PIPELINING
638 would result in both lost and duplicate recipients for a message.
640 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
641 Previously the weighting was incorrectly applied. Similar fix for socks
642 proxies. Found and fixed by Heiko Schlichting.
644 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
645 not handle sub-lists included using the +namedlist syntax. While
646 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
647 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
650 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
651 Previously the permanent pool was used, so the sore could not be freed.
652 This meant a connection with many messages would use continually-growing
655 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
656 per-pool so as not to waste too much space. Previously a constant size
657 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
658 DOS attacks harder. The cost is wasted memory use in the larger blocks.
660 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
661 same space repeatedly, and a smaller process footprint.
663 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
664 for GnuTLS. Previously the name started "%s" which, while not a bug,
665 looked as if if might be one.
667 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
668 "continued connection") make the $proxy_* variables available. Previously
669 the information was not passed across the exec() call for subsequent
670 transport executions. This also mean that the log lines for the
671 messages can show the proxy information.
673 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
674 receive time. With modern systems the difference is significant.
675 The historical behaviour can be restored by disabling (a new) log_selector
676 "queue_time_exclusive".
678 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
679 content) were specifically tested for. Now, also cover expansions
680 resulting in ACL names and inline ACL content.
682 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
683 making it unusable in complex expressions.
685 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
686 fail with a taint-check on the spoolfile name, and leave the message
689 HS/01 Enforce absolute PID file path name.
691 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
693 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
695 PP/02 Bug 2643: Correct TLS DH constants.
696 A missing NUL termination in our code-generation tool had led to some
697 incorrect Diffie-Hellman constants in the Exim source.
698 Reported by kylon94, code-gen tool fix by Simon Arlott.
700 PP/03 Impose security length checks on various command-line options.
701 Fixes CVE-2020-SPRSS reported by Qualys.
703 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
704 better. Reported by Qualys.
706 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
707 providing a particularly obnoxious sender full name.
710 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
712 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
715 PP/08 Change default for recipients_max from unlimited to 50,000.
717 PP/09 Fix security issue with too many recipients on a message (to remove a
718 known security problem if someone does set recipients_max to unlimited,
719 or if local additions add to the recipient list).
720 Fixes CVE-2020-RCPTL reported by Qualys.
722 PP/10 Fix security issue in SMTP verb option parsing
723 Fixes CVE-2020-EXOPT reported by Qualys.
725 PP/11 Fix security issue in BDAT state confusion.
726 Ensure we reset known-good where we know we need to not be reading BDAT
727 data, as a general case fix, and move the places where we switch to BDAT
728 mode until after various protocol state checks.
729 Fixes CVE-2020-BDATA reported by Qualys.
731 HS/03 Die on "/../" in msglog file names
733 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
734 the Exim runtime user are allowed to create files.
736 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
739 QS/03 When reading the output from interpreted forward files we do not
740 pass the pipe between the parent and the interpreting process to
741 executed child processes (if any).
743 QS/04 Always die if requested from internal logging, even is logging is
746 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
747 the API; compilation noo longer completes with DMARC support included.
748 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
749 been reported on other platforms.
751 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
753 JH/56 Make the majority of info read from config files readonly, for defence-in-
754 depth against exploits. Suggestion by Qualys.
755 Not supported on Solaris 10.
757 JH/57 Fix control=fakreject for a custom message containing tainted data.
758 Previously this resulted in a log complaint, due to a re-expansion present
759 since fakereject was originally introduced.
761 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
762 modifier was given, a loop resulted.
764 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
765 pipelined SMTP command followed the BDAT LAST then it would be
766 incorrectly treated as part of the message body, causing a verification
769 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
770 require looser validation rules than those for 821-level addresses,
771 which only permit IP addresses.
777 JH/01 Avoid costly startup code when not strictly needed. This reduces time
778 for some exim process initialisations. It does mean that the logging
779 of TLS configuration problems is only done for the daemon startup.
781 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
783 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
784 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
787 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
788 DKIM signing is being done. Previously a transport_filter would always
789 disable CHUNKING, falling back to traditional DATA.
791 JH/05 Regard command-line recipients as tainted.
793 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
795 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
796 PAM library frees one of the arguments given to it, despite the
797 documentation. Therefore a plain malloc must be used.
799 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
800 on-stack buffers were used, resulting in a taint trap when DSN information
801 copied from a received message was written into the buffer.
803 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
804 the ordering of its ARC headers. This caused a crash.
806 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
807 a new record was being constructed with information from the peer, a trap
810 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
811 installation would get error messages from DMARC verify, when it hit the
812 nonexistent file indicated by the default. Distros wanting DMARC enabled
813 should both provide the file and set the option.
814 Also enforce no DMARC verification for command-line sourced messages.
816 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
817 could, depending on the platform, hang at the STARTTLS response.
819 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
820 message on a connection. Previously if one message had ARC headers and
821 the following one did not, a crash could result when adding an
822 Authentication-Results: header.
824 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
825 Exim main code and Exim-related utities. The introduction of taint
826 tracking also did many adjustments to string handling. Since then, eximon
827 frequently terminated with an assert failure.
829 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
830 check for 452 responses. This slightly helps the inefficieny of doing
831 a large alias-expansion into a recipient-limited target. The max_rcpt
832 transport option still applies (and at the current default, will override
833 the new feature). The check is done for either cause of synch, and forces
834 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
835 connection. The new facility is not tunable at this time.
837 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
838 library live data was being used, so the results became garbage. Make
839 copies while it is still usable.
841 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
842 on delivery deferred (==) and failed (**) lines (if a delivery was
843 attemtped). Previously it was only on completion (=>) lines.
845 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
846 for the expansion of the server_scram_iter and server_scram_salt options.
848 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
849 are now specifically given a NO_DATA response without hitting the system
850 resolver. The library goes on to do the now-standard TXT lookup.
851 Use of dnsdb lookups is not affected.
853 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
854 only retrieve the errormessage once. Previously two calls to dlerror()
855 were used, and the second one (for mainlog/paniclog) retrieved null
858 JH/20 Taint checking: disallow use of tainted data for
859 - the appendfile transport file and directory options
860 - the pipe transport command
861 - the autoreply transport file, log and once options
862 - file names used by the redirect router (including filter files)
864 - paths used by single-key lookups
865 Previously this was permitted.
867 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
868 adjusted the size of a major service buffer; this failed because the
869 buffer was in use at the time. Change to a compile-time increase in the
870 buffer size, when this authenticator is compiled into exim.
872 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
873 previous fast-mode was untenable in the face of glibs using mmap to
874 support larger malloc requests.
876 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
877 New values supported, if defined on system where compiled:
878 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
879 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
881 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
882 running a limited number of proceses in parallel, a benefit is gained. The
883 amount varies with the platform hardware and load. The use of the option
884 queue_run_in_order means we cannot do this, as ordering becomes
887 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
888 had introduced a string-copy (for ensuring NUL-termination) which was not
889 appropriate for that case, which can include embedded NUL bytes in the
890 block of data. Investigation showed the copy to actually be needless, the
891 data being length-specified.
893 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
894 done during a receiving connection, and both used TLS, global info was
895 used rather than per-connection info for tracking the state of data
896 queued for transmission. This could result in a connection hang.
898 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
899 Previously, when delivering serveral messages down a single connection
900 only the first would provide a SIZE. This was due to the size information
901 not being properly tracked.
903 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
904 TAI (at 37 seconds currently), pretend to be in UTC for time-related
905 expansion and logging. Previously, spurious values such as a future
906 minute could be seen.
908 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
909 it could crash from a null-deref. This could also affect the
910 ${addresses: } operator and ${readsock } item.
912 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
913 message following a mime one, the variable was not reset.
915 JH/30 When an pipelined-connect fails at the first response, assume incorrect
916 cached capability (perhaps the peer reneged?) and immediately retry in
919 JH/31 Fix spurious detection of timeout while writing to transport filter.
921 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
922 an attempt to copy the string was made before checking it.
924 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
925 taint of the lookup key was maintained; we now regard the presence in the
926 filesystem as sufficient validation.
928 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
929 argument is supplied.
931 JH/35 The dsearch lookup now requires that the directory is an absolute path.
932 Previously this was not checked, and nonempty relative paths made an
933 access under Exim's current working directory.
935 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
936 Previously no event was raised.
938 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
939 parameter supplied by the sender MAIL FROM command. Previously it was
940 ignored, and only the check_spool_space option value for the required
943 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
944 the size of the signing public-key. Previously it was instead giving
945 the size of the signature hash.
947 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
948 the default. See the (new) dkim_verify_min_keysizes option.
950 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
951 and an ACL use a lookup for checking either the local_part or domain,
952 stale data could be accessed. Ensure that variable references are
953 dropped between messages.
955 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
956 by the client was not checked as pointing within response data before
957 being used. A malicious client could thus cause an out-of-bounds read and
958 possibly gain authentication. Fix by adding the check.
960 JH/42 Internationalisation: change the default for downconversion in the smtp
961 transport to be "if needed". Previously it was "as previously set" for
962 the message, which usually meant "if needed" for message-submission but
963 "no" for everything else. However, MTAs have been seen using SMTPUTF8
964 even when the envelope addresses did not need it, resulting in forwarding
965 failures to non-supporting MTAs. A downconvert in such cases will be
966 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
967 transport. The change does mean that addresses needing conversion will
968 be converted when previously a delivery failure would occur.
970 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
971 response was received it would be used unchecked in a fail-DSN, violating
972 standards on line-length limits. Truncate if needed.
974 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
975 doesn't work. (Additionally add a new main config option to configure the
982 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
983 side, to decode a TLS 1.3 packet capture.
985 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
986 Previously the default library behaviour applied, sending two, each in
989 JH/03 Debug output for ACL now gives the config file name and line number for
992 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
994 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
996 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
997 buffer overrun for (non-chunking) other transports.
999 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1000 TLS1.3, means that a server rejecting a client certificate is not visible
1001 to the client until the first read of encrypted data (typically the
1002 response to EHLO). Add detection for that case and treat it as a failed
1003 TLS connection attempt, so that the normal retry-in-clear can work (if
1004 suitably configured).
1006 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1007 and/or domain. Found and fixed by Jason Betts.
1009 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1010 configuration). If a CNAME target was not a wellformed name pattern, a
1013 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1014 the OS reports them interleaved with other addresses.
1016 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1017 used both for input and for a verify callout, both encrypted, SMTP
1018 responses being sent by the server could be lost. This resulted in
1019 dropped connections and sometimes bounces generated by a peer sending
1022 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1023 string. Previously it was possible to cause undefined behaviour in a
1024 library routine (usually a crash). Found by "zerons".
1026 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1029 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1030 API was removed, so update to use the newer ones.
1032 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1033 any timeout set, is taking a long time. Previously we would hang on to a
1034 rotated logfile "forever" if the input was arriving with long gaps
1035 (a previous attempt to fix addressed lack, for a long time, of initial
1038 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1039 shared (NFS) environment. The length of the tempfile name is now
1040 4 + 16 ("hdr.$message_exim_id") which might break on file
1041 systems which restrict the file name length to lower values.
1042 (It was "hdr.$pid".)
1044 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1045 shared (NFS) environment.
1047 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1048 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1051 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1052 on some platforms for bit 31.
1054 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1055 to changes apparently associated with TLS1.3 handling some of the APIs
1056 previously used were either nonfunctional or inappropriate. Strings
1057 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1058 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1059 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1060 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1061 and the use of specific cipher names in the encrypted= ACL condition.
1063 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1065 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1066 verification result was not updated unless hosts_require_ocsp applied.
1068 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1069 queue_list_requires_admin set to false, non-admin users were denied the
1072 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1073 directory-of-certs mode. Previously they were advertised despite the
1076 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1077 A single TCP connection by a client will now hold a TLS connection open
1078 for multiple message deliveries, by default. Previously the default was to
1081 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1082 default. If built with the facility, DANE will be used. The facility
1083 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1085 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1086 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1087 must be defined and you must still, unless you define DISABLE_TLS, manage
1088 the the include-dir and library-file requirements that go with that
1089 choice. Non-TLS builds are still supported.
1091 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1094 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1095 default. If the platform supports and has the facility enabled, it will
1096 be requested on all coneections.
1098 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1099 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1101 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1103 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1104 success-DSN messages. Previously the From: header was always the default
1105 one for these; the option was ignored.
1107 JH/28 Fix the timeout on smtp response to apply to the whole response.
1108 Previously it was reset for every read, so a teergrubing peer sending
1109 single bytes within the time limit could extend the connection for a
1110 long time. Credit to Qualsys Security Advisory Team for the discovery.
1112 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1113 delivery address, which leaked information of the results of local
1114 forwarding. Change to the original envelope recipient address, per
1117 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1118 requested. Previously not bounce was generated and a log entry of
1119 error ignored was made.
1121 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1123 JH/32 Introduce a general tainting mechanism for values read from the input
1124 channel, and values derived from them. Refuse to expand any tainted
1125 values, to catch one form of exploit.
1127 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1128 was unused and the unexpanded text used for the test. Found and
1129 fixed by Ruben Jenster.
1131 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1132 an attempt to use a TLS library read routine dereffed a nul pointer,
1135 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1136 them in our smtp response.
1138 JH/36 Have the general router option retry_use_local_part default to true when
1139 any of the restrictive preconditions are set (to anything). Previously it
1140 was only for check_local user. The change removes one item of manual
1141 configuration which is required for proper retries when a remote router
1142 handles a subset of addresses for a domain.
1144 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1145 link count into consideration.
1147 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1148 caused the extension of big_buffer, the following lines were ignored.
1150 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1151 accordance with RFC 2308. Previously there was no expiry, so a longlived
1152 receive process (eg. due to ACL delays) versus a short SOA value could
1155 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1157 JH/39 Promote DMARC support to mainline.
1159 JH/40 Bug 2452: Add a References: header to DSNs.
1161 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1162 parameters. The relevant library call is documented as "Deprecated: This
1163 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1164 3.6.0, DH parameters are negotiated following RFC7919."
1166 HS/06 Change the default of dnssec_request_domains to "*"
1168 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1169 carried on and emitted a BDAT command, even when PIPELINING was not
1172 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1173 buffer was used for the filename, resulting in a trap when tainted
1174 arguments (eg. $domain) were used.
1176 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1177 recommended to avoid a possible server-load attack. The feature can be
1178 re-enabled via the openssl_options main cofiguration option.
1180 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1181 for version 4.90 - adding a "more data" boolean to the arguments.
1182 Bumped the ABI version number also, this having been missed previously;
1183 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1184 respect of usage of smtp_printf() by either local_scan code or libraries
1185 accessed via the ${dlfunc } expansion item. Both will need coding
1186 adjustment for any calls to smtp_printf() to match the new function
1187 signature; a FALSE value for the new argument is always safe.
1189 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1190 the file-offset (which the Linux syscall does, and exim expects); this
1191 resulted in an indefinite loop.
1193 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1194 to do ARC verification. The Authentication-Results: header line added
1195 by the configuration then had no ARC item.
1197 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1198 an incorrect mode was used for reading the result, resulting in it being
1205 JH/01 Remove code calling the customisable local_scan function, unless a new
1206 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1208 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1209 non-signal-safe functions being used.
1211 JH/03 Bug 2269: When presented with a received message having a stupidly large
1212 number of DKIM-Signature headers, disable DKIM verification to avoid
1213 a resource-consumption attack. The limit is set at twenty.
1215 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1216 report of oldest_pass in ${authres } in consequence, and separate out
1217 some descriptions of reasons for verification fail.
1219 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1220 files in the spool were present and unlocked. A queue-runner could spot
1221 them, resulting in a duplicate delivery. Fix that by doing the unlock
1222 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1223 add more error-checking on spoolfile handling while that code is being
1226 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1227 No known attacks, no CVE, this is defensive hardening.
1229 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1230 a queue-runner could start a delivery while other operations were ongoing.
1231 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1232 Found and investigated by Tim Stewart. Fix by using the open message data
1233 file handle rather than opening another, and not locally closing it (which
1234 releases a lock) for that case, while creating the temporary .eml format
1235 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1237 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1238 $sender_verify_failure/$recipient_verify_failure to "random".
1240 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1243 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1244 Previously this would segfault.
1246 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1249 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1250 like zero, since the resolver should be doing this for us, But we need one
1251 as a CNAME but no MX presence gets the CNAME returned; we need to check
1252 that doesn't point to an MX to declare it "no MX returned" rather than
1253 "error, loop". A new main option is added so the older capability of
1254 following some limited number of chain links is maintained.
1256 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1258 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1259 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1260 now-deprecated earlier definitions used only specified the range up to TLS
1261 1.2 (in the older-version library docs).
1263 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1265 JH/15 Rework TLS client-side context management. Stop using a global, and
1266 explicitly pass a context around. This enables future use of TLS for
1267 connections to service-daemons (eg. malware scanning) while a client smtp
1268 connection is using TLS; with cutthrough connections this is quite likely.
1270 JH/16 Fix ARC verification to do AS checks in reverse order.
1272 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1274 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1275 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1276 Previously the "utf8" would be re-prepended for every additional message.
1278 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1279 Previously thery were accepted, resulting in issues when attempting to
1280 forward messages to a non-supporting MTA.
1282 PP/02 Let -n work with printing macros too, not just options.
1284 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1285 one parent address was copied, and bogus data was used at delivery-logging
1286 time. Either a crash (after delivery) or bogus log data could result.
1287 Discovery and analysis by Tim Stewart.
1289 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1290 Previously if the string ended mid-character, we did not insert the
1291 promised '?' replacement.
1293 PP/04 Documentation: current string operators work on bytes, not codepoints.
1295 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1296 should pack well giving a smaller memory footprint so better caching and
1297 therefore performance. Group the declarations where this can't be done so
1298 that the byte-sized flag variables are not interspersed among pointer
1299 variables, giving a better chance of good packing by the compiler.
1301 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1302 non-null, to avoid issues with sites running BATV. Previously reports were
1303 sent with an empty envelope sender so looked like bounces.
1305 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1306 The ignore_error flag wasn't being returned from the filter subprocess so
1307 was not set for later routers. Investigation and fix by Matthias Kurz.
1309 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1310 and a msg:complete for the whole, when a message is manually removed using
1311 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1313 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1314 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1315 hope that is portable enough.
1317 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1318 requiring a known-CA anchor certificate; make it now rely entirely on the
1319 TLSA as an anchor. Checking the name on the leaf cert against the name
1320 on the A-record for the host is still done for TA (but not for EE mode).
1322 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1323 included in delivery lines for non-proxied connections, when compiled with
1324 SUPPORT_SOCKS and running with proxy logging enabled.
1326 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1327 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1328 move the existing event to fire before the normal logging of message
1329 failure so that custom logging is bracketed by normal logging.
1331 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1332 msg:fail:internal event. Developement by Matthias Kurz.
1334 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1335 far too small for todays use of crypto signatures stored there. Go all
1336 the way to the max DNS message size of 64kB, even though this might be
1337 overmuch for IOT constrained device use.
1339 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1340 copy a string over itself. The library routine is documented as not
1341 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1343 JH/32 For main options check_spool_space and check_inode_space, where the
1344 platform supports 64b integers, support more than the previous 2^31 kB
1345 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1346 the previous G, M, k.
1348 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1349 $authenticated_fail_id variable on authentication failure. Previously
1352 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1353 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1354 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1355 GNUTLS_SEC_PARAM_MEDIUM.
1357 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1358 side. Previously we would continue as if no SNI had been received.
1360 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1361 "<" character, which should be a list-separator specification, we walked
1362 off past the nul-terimation.
1364 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1365 causes) even when the retry time is not yet met. Previously they were
1366 not, meaning that when (say) an account was over-quota and temp-rejecting,
1367 and multiple senders' messages were queued, only one sender would get
1368 notified on each configured delay_warning cycle.
1370 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1372 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1373 cipher-suites, an error can be left on the stack even for a succeeding
1374 accept; this results in impossible error messages when a later operation
1377 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1378 return error codes indicating retry. Under TLS1.3 this becomes required.
1380 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1381 it only wrote the new authenticators, resulting in a lack of tracking of
1382 peer changes of ESMTP extensions until the next cache flush.
1384 JH/41 Fix the loop reading a message header line to check for integer overflow,
1385 and more-often against header_maxsize. Previously a crafted message could
1386 induce a crash of the recive process; now the message is cleanly rejected.
1388 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1389 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1395 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1396 When redis_servers is set to a list of > 1 element, and the Redis servers
1397 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1398 case of MOVED into a DEFER case instead, thus moving the query onto the
1399 next server in the list. For a cluster of N elements, all N servers must
1400 be defined in redis_servers.
1402 GF/02 Catch and remove uninitialized value warning in exiqsumm
1403 Check for existence of @ARGV before looking at $ARGV[0]
1405 JH/01 Replace the store_release() internal interface with store_newblock(),
1406 which internalises the check required to safely use the old one, plus
1407 the allocate and data copy operations duplicated in both (!) of the
1408 extant use locations.
1410 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1411 modifier. This matches the restriction on the commandline.
1413 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1414 Previously only the last row was returned.
1416 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1417 we assumed that tags in the header were well-formed, and parsed the
1418 element content after inspecting only the first char of the tag.
1419 Assumptions at that stage could crash the receive process on malformed
1422 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1423 While running the DKIM ACL we operate on the Permanent memory pool so that
1424 variables created with "set" persist to the DATA ACL. Also (at any time)
1425 DNS lookups that fail create cache records using the Permanent pool. But
1426 expansions release any allocations made on the current pool - so a dnsdb
1427 lookup expansion done in the DKIM ACL releases the memory used for the
1428 DNS negative-cache, and bad things result. Solution is to switch to the
1429 Main pool for expansions.
1430 While we're in that code, add checks on the DNS cache during store_reset,
1431 active in the testsuite.
1432 Problem spotted, and debugging aided, by Wolfgang Breyha.
1434 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1435 When none of the hosts presented to a transport match an already-open
1436 connection, close it and proceed with the list. Previously we would
1437 queue the message. Spotted by Lena with Yahoo, probably involving
1440 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1441 Previously a spurious "250 OK id=" response was appended to the proper
1444 JH/08 The "support for" informational output now, which built with Content
1445 Scanning support, has a line for the malware scanner interfaces compiled
1446 in. Interface can be individually included or not at build time.
1448 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1449 by the template makefile "src/EDITME". The "STREAM" support for an older
1450 ClamAV interface method is removed.
1452 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1453 rows affected is given instead).
1455 JH/11 The runtime Berkeley DB library version is now additionally output by
1456 "exim -d -bV". Previously only the compile-time version was shown.
1458 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1459 SMTP connection. Previously, when one had more recipients than the
1460 first, an abortive onward connection was made. Move to full support for
1461 multiple onward connections in sequence, handling cutthrough connection
1462 for all multi-message initiating connections.
1464 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1465 routers. Previously, a multi-recipient message would fail to match the
1466 onward-connection opened for the first recipient, and cause its closure.
1468 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1469 a timeout on read on a GnuTLS initiating connection, resulting in the
1470 initiating connection being dropped. This mattered most when the callout
1471 was marked defer_ok. Fix to keep the two timeout-detection methods
1474 JH/15 Relax results from ACL control request to enable cutthrough, in
1475 unsupported situations, from error to silently (except under debug)
1476 ignoring. This covers use with PRDR, frozen messages, queue-only and
1479 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1481 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1482 metadata, resulting in a crash in free().
1484 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1485 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1486 Broken also in d185889f4, with init system revamp.
1488 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1489 Previously we abruptly closed the connection after reading a malware-
1490 found indication; now we go on to read the "scan ok" response line,
1493 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1494 ACL. Previously, a crash would result.
1496 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1497 macro text after a replacement (previously it was only once per line) and
1498 by skipping builtin macros when searching for an uppercase lead character.
1500 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1501 for the build is renamed.
1503 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1504 was allocated for every new TLS startup, meaning one per message. Fix
1505 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1507 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1508 reported the original. Fix to report (as far as possible) the ACL
1509 result replacing the original.
1511 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1512 OpenSSL. Certificate information is loaded for every new TLS startup,
1513 and the resources needed to be freed.
1515 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1517 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1520 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1521 DATA response info to the (existing) per-recipient response info for
1522 the "C=" log element. It can have useful tracking info from the
1523 destination system. Patch from Simon Arlott.
1525 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1526 length value. Previously this would segfault.
1528 HS/02 Support Avast multiline protoocol, this allows passing flags to
1529 newer versions of the scanner.
1531 JH/28 Ensure that variables possibly set during message acceptance are marked
1532 dead before release of memory in the daemon loop. This stops complaints
1533 about them when the debug_store option is enabled. Discovered specifically
1534 for sender_rate_period, but applies to a whole set of variables.
1535 Do the same for the queue-runner and queue-list loops, for variables set
1536 from spool message files. Do the same for the SMTP per-message loop, for
1537 certain variables indirectly set in ACL operations.
1539 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1540 as a multi-recipient message from a mailinglist manager). The coding had
1541 an arbitrary cutoff number of characters while checking for more input;
1542 enforced by writing a NUL into the buffer. This corrupted long / fast
1543 input. The problem was exposed more widely when more pipelineing of SMTP
1544 responses was introduced, and one Exim system was feeding another.
1545 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1546 receiving system, and refused recipients seen by the sending system
1547 (propating to people being dropped from mailing lists).
1548 Discovered and pinpointed by David Carter.
1550 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1551 replaced by the ${authresults } expansion.
1553 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1555 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1556 allows proper process termination in container environments.
1558 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1559 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1561 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1562 and "err_perm", deprecated since 4.83 when the RFC-defined words
1563 "temperror" and "permerror" were introduced.
1565 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1566 transport-filters or DKIM-signing. The restriction was lost in the
1567 consolidation of verify-callout and delivery SMTP handling.
1568 Extend the restriction to also cover ARC-signing.
1570 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1571 in defer=pass mode supply a 450 to the initiator. Previously the message
1574 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1575 tls_require_ciphers is used as before.
1577 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1578 "pass_unscanned". Only tmpfails from the scanner are written to
1579 the paniclog, as they may require admin intervention (permission
1580 denied, license issues). Other scanner errors (like decompression
1581 bombs) do not cause a paniclog entry.
1583 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1584 Previously it was possible to log spurious information in receive log
1587 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1588 triggered odd behaviour from Outlook Express clients.
1590 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1593 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1594 since the IETF WG has not yet settled on that versus the original
1595 "bare" representation.
1597 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1598 Previously the millisecond value corrupted the output.
1599 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1600 corrupted the output.
1606 JH/01 Rework error string handling in TLS interface so that the caller in
1607 more cases is responsible for logging. This permits library-sourced
1608 string to be attached to addresses during delivery, and collapses
1609 pairs of long lines into single ones.
1611 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1612 during configuration. Wildcards are allowed and expanded.
1614 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1615 This permits better logging.
1617 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1618 do not close it down and have the child transport start it up again on
1619 the passed-on TCP connection. Instead, proxy the child (and any
1620 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1621 affected: the continued delivery log lines do not have any DNSSEC, TLS
1622 Certificate or OCSP information. TLS cipher information is still logged.
1624 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1625 identical IP addresses on different listening ports. Will also affect
1628 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1629 add noisy ifdef guards to special-case this sillyness.
1630 Patch from Bernd Kuhls.
1632 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1633 than 255 are no longer allowed.
1635 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1636 Disable the session-cache too, which might reduce our load. Since we
1637 currrectly use a new context for every connection, both as server and
1638 client, there is no benefit for these.
1639 GnuTLS appears to not support tickets server-side by default (we don't
1640 call gnutls_session_ticket_enable_server()) but client side is enabled
1641 by default on recent versions (3.1.3 +) unless the PFS priority string
1644 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1645 <https://reproducible-builds.org/specs/source-date-epoch/>.
1647 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1648 the check for any unsuccessful recipients did not notice the limit, and
1649 erroneously found still-pending ones.
1651 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1652 MSG_MORE. Only in-clear (not on TLS connections).
1654 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1655 transport-filter is involved we can buffer the headers in memory for
1656 creating the signature, and read the spool data file once for the
1657 signature and again for transmission.
1659 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1660 4.77 as the kernel support then wasn't solid, having issues in 64bit
1661 mode. Now, it's been long enough. Add support for FreeBSD also.
1663 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1664 case where the routing stage had gathered several addresses to send to
1665 a host before calling the transport for the first, we previously failed
1666 to close down TLS in the old transport process before passing the TCP
1667 connection to the new process. The new one sent a STARTTLS command
1668 which naturally failed, giving a failed delivery and bloating the retry
1669 database. Investigation and fix prototype from Wolfgang Breyha.
1671 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1672 false-negatives in the check that the sender had not preempted a response
1673 or prompt from Exim (running as a server), due to that code's lack of
1674 awareness of the SMTP input buffering.
1676 PP/04 Add commandline_checks_require_admin option.
1677 Exim drops privileges sanely, various checks such as -be aren't a
1678 security problem, as long as you trust local users with access to their
1679 own account. When invoked by services which pass untrusted data to
1680 Exim, this might be an issue. Set this option in main configuration
1681 AND make fixes to the calling application, such as using `--` to stop
1684 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1685 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1686 responses to those, into a single TLS record each way (this usually means
1687 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1690 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1691 affects you only if you're dancing at the edge of the param size limits.
1692 If you are, and this message makes sense to you, then: raise the
1693 configured limit or use OpenSSL 1.1. Nothing we can do for older
1696 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1697 cmdline element to get the documented default one. Previously it was
1700 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1701 are unsuable for non-message-specific reasons - eg. connection timeout,
1702 banner-time rejection.
1704 JH/16 Fix logging of delivery remote port, when specified by router, under
1707 PP/06 Repair manualroute's ability to take options in any order, even if one
1708 is the name of a transport.
1711 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1713 JH/17 Change the list-building routines interface to use the expanding-string
1714 triplet model, for better allocation and copying behaviour.
1716 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1717 Previously it was constructed the first time a possibly-matching string
1718 was met in the configuration file input during startup; now it is done
1721 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1722 compatible one, to avoid the (poorly documented) possibility of a config
1723 file in the working directory redirecting the DB files, possibly correpting
1724 some existing file. CVE-2017-10140 assigned for BDB.
1726 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1727 cache-hot. Previously, although the result was properly cached, the
1728 initial verify call returned a defer.
1730 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1731 the main verify for receipient in uncached-mode.
1733 JH/22 Retire historical build files to an "unsupported" subdir. These are
1734 defined as "ones for which we have no current evidence of testing".
1736 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1737 if present. Previously it was ignored.
1739 JH/24 Start using specified-initialisers in C structure init coding. This is
1740 a C99 feature (it's 2017, so now considered safe).
1742 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1743 if was a fixed-sized field and bitmask ops via macros; it is now more
1746 PP/07 GitHub PR 56: Apply MariaDB build fix.
1747 Patch provided by Jaroslav Škarvada.
1749 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1750 during Coverity cleanups [4.87 JH/47]
1751 Diagnosis and fix provided by Michael Fischer v. Mollard.
1753 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1754 the right size to place the terminating semicolon on its own folded
1755 line, the header hash was calculated to an incorrect value thanks to
1756 the (relaxed) space the fold became.
1758 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1759 and confused the parent.
1761 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1762 which could crash as a result. This could lead to undeliverable messages.
1764 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1767 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1768 matters on fast-turnover and PID-randomising systems, which were getting
1769 out-of-order delivery.
1771 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1772 a possibly-overlapping copy. The symptom was that "Remote host closed
1773 connection in response to HELO" was logged instead of the actual 4xx
1776 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1777 Previously only that bufferd was discarded, resulting in SYMTP command
1780 JH/32 DKIM: when a message has multiple signatures matching an identity given
1781 in dkim_verify_signers, run the dkim acl once for each. Previously only
1782 one run was done. Bug 2189.
1784 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1785 "panic the current process" to "deliberately defer". The panic log is
1786 still written with the problem list name; the mail and reject logs now
1787 get a temp-reject line for the message that was being handled, saying
1788 something like "domains check lookup or other defer". The SMTP 451
1789 message is still "Temporary local problem".
1791 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1792 A crafted sequence of BDAT commands could result in in-use memory beeing
1793 freed. CVE-2017-16943.
1795 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1796 from SMTP input. Previously it was always done; now only done for DATA
1797 and not BDAT commands. CVE-2017-16944.
1799 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1800 to the message (such as an overlong header line). Previously this was
1801 not done and we did not exit BDAT mode. Followon from the previous item
1802 though a different problem.
1808 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1809 than -2003 did; needs libidn2 in addition to libidn.
1811 JH/02 The path option on a pipe transport is now expanded before use.
1813 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1814 Patch provided by "Björn", documentation fix added too.
1816 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1817 missing a wire-to-host endian conversion.
1819 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1820 close after a BDAT command line could be taken as a following command,
1821 giving a synch failure. Fix by only checking for synch immediately
1822 before acknowledging the chunk.
1824 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1825 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1826 macro. Patches provided by Josh Soref.
1828 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1829 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1830 by IANA as a service extension. However, John Klensin suggests that we
1833 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1834 the dkim code may be unix-mode line endings rather than smtp wire-format
1835 CRLF, so prepend a CR to any bare LF.
1837 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1838 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1840 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1841 the first were themselves being wrongly included in the feed into dkim
1842 processing; with most chunk sizes in use this resulted in an incorrect
1843 body hash calculated value.
1845 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1846 DKIM signature block, for verification. Although advised against by
1847 standards it is specifically not ruled illegal.
1849 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1851 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1852 missing a body hash (the bh= tag).
1854 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1855 It seems that HAProxy sends the Proxy Protocol information in clear and
1856 only then does a TLS startup, so do the same.
1858 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1859 TCP connections (such as for Spamd) unless the daemon successfully set
1860 Fast Open mode on its listening sockets. This fixes breakage seen on
1861 too-old kernels or those not configured for Fast Open, at the cost of
1862 requiring both directions being enabled for TFO, and TFO never being used
1863 by non-daemon-related Exim processes.
1865 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1866 endings, at least on the first header line. Try to canonify any that get
1867 past that check, despite the cost.
1869 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1870 now limited to an arbitrary five deep, while parsing addresses with the
1871 strip_excess_angle_brackets option enabled.
1873 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1874 instead leave the unprompted TLS handshake in socket buffer for the
1875 TLS library to consume.
1877 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1879 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1881 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1882 region in one operation, for speed, and this leaves assigned pointers
1883 dangling. Add checks run only under the testsuite which checks all
1884 variables at a store-reset and panics on a dangling pointer; add code
1885 explicitly nulling out all the variables discovered. Fixes one known
1886 bug: a transport crash, where a dangling pointer for $sending_ip_address
1887 originally assigned in a verify callout, is re-used.
1889 PP/06 Drop '.' from @INC in various Perl scripts.
1891 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1893 PP/08 Reduce a number of compilation warnings under clang; building with
1894 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1895 should be warning-free.
1897 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1899 HS/01 Fix portability problems introduced by PP/08 for platforms where
1900 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1902 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1903 chunk. This allows us to accept broken chunked messages. We need a more
1904 general solution here.
1906 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1907 already-broken messages in the queue.
1909 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1911 JH/19 Fix reference counting bug in routing-generated-address tracking.
1917 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1918 supports it and a size is available (ie. the sending peer gave us one).
1920 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1921 years of being deprecated). The replacements are the ACLs
1922 acl_smtp_mime and acl_not_smtp_mime.
1924 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1925 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1926 or even an in-clear connection were permitted. Now, if the host lookup
1927 was dnssec and dane was requested then the host is only used if the
1928 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1929 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1930 if one fails this test.
1931 This means that a poorly-configured remote DNS will make it incommunicado;
1932 but it protects against a DNS-interception attack on it.
1934 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1935 non-noisy when a race steals the message being considered.
1937 JH/05 If main configuration option tls_certificate is unset, generate a
1938 self-signed certificate for inbound TLS connections.
1940 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1941 in rewrites and routers.
1943 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1944 and logged a warning sing 4.83; now they are a configuration file error.
1946 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1947 (lacking @domain). Apply the same qualification processing as RCPT.
1949 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1951 JH/10 Support ${sha256:} applied to a string (as well as the previous
1954 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1955 a cutthrough deliver is pending, as we always want to make a connection.
1956 This also avoids re-routing the message when later placing the cutthrough
1957 connection after a verify cache hit.
1958 Do not update it with the verify result either.
1960 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1961 when routing results in more than one destination address.
1963 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1964 signing (which inhibits the cutthrough capability). Previously only
1965 the presence of an option was tested; now an expansion evaluating as
1966 empty is permissible (obviously it should depend only on data available
1967 when the cutthrough connection is made).
1969 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1970 the relevant preceding SMTP command did not note the pipelining mode.
1972 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1973 Previously they were not counted.
1975 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1976 as one having no matching records. Previously we deferred the message
1977 that needed the lookup.
1979 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1980 distinguished as "(=".
1982 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1983 for missing MX records. Previously it only worked for missing A records.
1985 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1987 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1988 after the data-go-ahead and data-ack. Patch from Jason Betts.
1990 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1991 even for a "none" policy. Patch from Tony Meyer.
1993 JH/22 Fix continued use of a connection for further deliveries. If a port was
1994 specified by a router, it must also match for the delivery to be
1997 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1998 When one of the recipients of a message was unsuitable for the connection
1999 (has no matching addresses), we lost track of needing to mark it
2000 deferred. As a result mail would be lost.
2002 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2004 JH/25 Decoding ACL controls is now done using a binary search; the source code
2005 takes up less space and should be simpler to maintain. Merge the ACL
2006 condition decode tables also, with similar effect.
2008 JH/26 Fix problem with one_time used on a redirect router which returned the
2009 parent address unchanged. A retry would see the parent address marked as
2010 delivered, so not attempt the (identical) child. As a result mail would
2013 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2014 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2015 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2018 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2019 Main config options check_{log,spool}_{inodes,space} are now
2020 100 inodes, 10MB unless set otherwise in the configuration.
2022 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2023 Previously it only applied to the main-section connection policy
2026 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2028 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2029 by me. Added RFC7919 DH primes as an alternative.
2031 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2032 are not in the system include path.
2034 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2035 GnuTLS, when a session startup failed (eg because the client disconnected)
2036 Exim did stdio operations after fclose. This was exposed by a recent
2037 change which nulled out the file handle after the fclose.
2039 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2040 signed directly by the cert-signing cert, rather than an intermediate
2041 OCSP-signing cert. This is the model used by LetsEncrypt.
2043 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2045 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2046 an incoming connection.
2048 HS/02 Bug 1802: Do not half-close the connection after sending a request
2051 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2052 fallback to "prime256v1".
2054 JH/34 SECURITY: Use proper copy of DATA command in error message.
2055 Could leak key material. Remotely exploitable. CVE-2016-9963.
2061 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2062 and 3.4.4 - once the server is enabled to respond to an OCSP request
2063 it does even when not requested, resulting in a stapling non-aware
2064 client dropping the TLS connection.
2066 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2067 support variable-length bit vectors. No functional change.
2069 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2070 The I= interface field on outgoing lines is now after the H= remote
2071 host field, same as incoming lines. There is a separate
2072 outgoing_interface log selector which allows you to disable the
2075 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2076 If not running log_selector +smtp_connection the mainlog would be held
2077 open indefinitely after a "too many connections" event, including to a
2078 deleted file after a log rotate. Leave the per net connection logging
2079 leaving it open for efficiency as that will be quickly detected by the
2080 check on the next write.
2082 HS/01 Bug 1671: Fix post transport crash.
2083 Processing the wait-<transport> messages could crash the delivery
2084 process if the message IDs didn't exist for some reason. When
2085 using 'split_spool_directory=yes' the construction of the spool
2086 file name failed already, exposing the same netto behaviour.
2088 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2089 mime_regex ACL conditions.
2091 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2092 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2093 to HELO, local diagnostic string.
2095 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2096 log line to debug. Even when configured with a tls authenticator many
2097 client connections are expected to not authenticate in this way, so
2098 an authenticate fail is not an error.
2100 HS/02 Add the Exim version string to the process info. This way exiwhat
2101 gives some more detail about the running daemon.
2103 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2104 matter for fast-change records such as DNSBLs.
2106 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2107 retry record, even if constant. There may be multiple transports with
2108 different interface settings and the retry behaviour needs to be kept
2111 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2113 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2115 JH/10 Bug 840: fix log_defer_output option of pipe transport
2117 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2118 hosts_randomize. This matters a lot when combined with mua_wrapper.
2120 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2121 ${quote_pgsql:<string>} operator.
2123 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2125 JH/14 Change header file name for freeradius-client. Relevant if compiling
2126 with Radius support; from the Gentoo tree and checked under Fedora.
2128 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2130 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2131 option transport_filter, meaning no filtering.
2133 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2135 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2136 defaults to "*" (all hosts). The variable is now available when not built
2137 with TLS, default unset, mainly to enable keeping the testsuite sane.
2138 If a server certificate is not supplied (via tls_certificate) an error is
2139 logged, and clients will find TLS connections fail on startup. Presumably
2140 they will retry in-clear.
2141 Packagers of Exim are strongly encouraged to create a server certificate
2142 at installation time.
2144 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2145 with the $config_file variable.
2147 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2148 in transport context, after the attempt, and per-recipient. The latter type
2149 is per host attempted. The event data is the error message, and the errno
2150 information encodes the lookup type (A vs. MX) used for the (first) host,
2151 and the trailing two digits of the smtp 4xx response.
2153 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2154 to write to mainlog (or rejectlog, paniclog) in the window between file
2155 creation and permissions/ownership being changed. Particularly affects
2156 installations where exicyclog is run as root, rather than exim user;
2157 result is that the running daemon panics and dies.
2159 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2161 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2162 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2163 "pri" and "weight". Note that the previous implicit priority given by the
2164 list order is no longer honoured.
2166 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2167 for DKIM processing.
2169 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2170 by defining SUPPORT_SOCKS.
2172 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2173 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2174 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2175 variables are renamed to proxy_{local,external}_{address,port}.
2177 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2178 for a build by defining SUPPORT_I18N
2180 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2181 of the query string, and make ${quote_redis:} do that quoting.
2183 JH/29 Move Events support from Experimental to mainline, enabled by default
2184 and removable for a build by defining DISABLE_EVENT.
2186 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2188 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2189 cached by the daemon.
2191 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2192 by defining LOOKUP_REDIS. The libhiredis library is required.
2194 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2195 keys are given for lookup.
2197 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2198 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2199 only supported when built with TLS support. The PolarSSL SHA routines
2200 are still used when the TLS library is too old for convenient support.
2202 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2203 openssl_options), for security. OpenSSL forces this from version 1.1.0
2204 server-side so match that on older versions.
2206 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2207 allocation for $value could be released as the expansion processing
2208 concluded, but leaving the global pointer active for it.
2210 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2211 and to use the domains and local_parts ACL conditions.
2213 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2214 incorrectly not doubled on cutthrough transmission, hence seen as a
2215 body-termination at the receiving system - resulting in truncated mails.
2216 Commonly the sender saw a TCP-level error, and retransmitted the message
2217 via the normal store-and-forward channel. This could result in duplicates
2218 received - but deduplicating mailstores were liable to retain only the
2219 initial truncated version.
2221 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2223 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2225 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2226 we're in there, support oversigning also; bug 1309.
2228 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2230 HS/04 Add support for keep_environment and add_environment options.
2232 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2233 either intentional arithmetic overflow during PRNG, or testing config-
2236 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2237 delivery resulted in actual delivery. Cancel cutthrough before DATA
2240 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2241 rejects a recipient: pass the reject to the originator.
2243 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2244 Many were false-positives and ignorable, but it's worth fixing the
2247 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2248 for the new environment-manipulation done at startup. Move the routines
2249 from being local to tls.c to being global via the os.c file.
2251 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2252 an extract embedded as result-arg for a map, the first arg for extract
2253 is unavailable so we cannot tell if this is a numbered or keyed
2254 extraction. Accept either.
2260 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2263 JH/02 The smtp transport option "multi_domain" is now expanded.
2265 JH/03 The smtp transport now requests PRDR by default, if the server offers
2268 JH/04 Certificate name checking on server certificates, when exim is a client,
2269 is now done by default. The transport option tls_verify_cert_hostnames
2270 can be used to disable this per-host. The build option
2271 EXPERIMENTAL_CERTNAMES is withdrawn.
2273 JH/05 The value of the tls_verify_certificates smtp transport and main options
2274 default to the word "system" to access the system default CA bundle.
2275 For GnuTLS, only version 3.0.20 or later.
2277 JH/06 Verification of the server certificate for a TLS connection is now tried
2278 (but not required) by default. The verification status is now logged by
2279 default, for both outbound TLS and client-certificate supplying inbound
2282 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2285 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2286 Status Notification (bounce) messages are now MIME format per RFC 3464.
2287 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2288 under the control of the dsn_advertise_hosts option, and routers may
2289 have a dsn_lasthop option.
2291 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2292 default, modifiable by a malware= option. The list separator for
2293 the options can now be changed in the usual way. Bug 68.
2295 JH/10 The smtp_receive_timeout main option is now expanded before use.
2297 JH/11 The incoming_interface log option now also enables logging of the
2298 local interface on delivery outgoing connections.
2300 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2301 if the interface and destination host and port all match.
2303 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2306 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2307 Patch from Andrew Lewis.
2309 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2310 now supports optional time-restrictions, weighting, and priority
2311 modifiers per server. Patch originally by <rommer@active.by>.
2313 JH/16 The spamd_address main option now supports a mixed list of local
2314 and remote servers. Remote servers can be IPv6 addresses, and
2315 specify a port-range.
2317 JH/17 Bug 68: The spamd_address main option now supports an optional
2318 timeout value per server.
2320 JH/18 Bug 1581: Router and transport options headers_add/remove can
2321 now have the list separator specified.
2323 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2326 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2329 JH/21 Support for the A6 type of dns record is withdrawn.
2331 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2332 rather than the verbs used.
2334 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2335 from 255 to 1024 chars.
2337 JH/24 Verification callouts now attempt to use TLS by default.
2339 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2340 are generic router options now. The defaults didn't change.
2342 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2343 Original patch from Alexander Shikoff, worked over by JH.
2345 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2346 files if ZCAT_COMMAND is not executable.
2348 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2350 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2352 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2353 Normally benign, it bites when the pair was led to by a CNAME;
2354 modern usage is to not canonicalize the domain to a CNAME target
2355 (and we were inconsistent anyway for A-only vs AAAA+A).
2357 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2359 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2360 when evaluating $sender_host_dnssec.
2362 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2363 $sender_helo_dnssec variable.
2365 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2367 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2369 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2371 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2372 documented as working, but never had. Support all but $spam_report.
2374 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2375 added for tls authenticator.
2377 HS/03 Add perl_taintmode main config option
2383 TL/01 When running the test suite, the README says that variables such as
2384 no_msglog_check are global and can be placed anywhere in a specific
2385 test's script, however it was observed that placement needed to be near
2386 the beginning for it to behave that way. Changed the runtest perl
2387 script to read through the entire script once to detect and set these
2388 variables, reset to the beginning of the script, and then run through
2389 the script parsing/test process like normal.
2391 TL/02 The BSD's have an arc4random API. One of the functions to induce
2392 adding randomness was arc4random_stir(), but it has been removed in
2393 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2394 function when detected.
2396 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2397 cause callback expansion.
2399 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2400 syntax errors in an expansion can be treated as a string instead of
2401 logging or causing an error, due to the internal use of bool_lax
2402 instead of bool when processing it.
2404 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2405 server certificates when making smtp deliveries.
2407 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2409 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2411 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2413 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2414 Merged patch from Sebastian Wiedenroth.
2416 JH/05 Fix results-pipe from transport process. Several recipients, combined
2417 with certificate use, exposed issues where response data items split
2418 over buffer boundaries were not parsed properly. This eventually
2419 resulted in duplicates being sent. This issue only became common enough
2420 to notice due to the introduction of connection certificate information,
2421 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2423 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2424 size buffer was used, resulting in syntax errors when an expansion
2427 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2428 version 3.3.6 or later.
2430 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2431 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2432 both become "event_action", the variables become $event_name, $event_data
2433 and $event_defer_errno. There is a new variable $verify_mode, usable in
2434 routers, transports and related events. The tls:cert event is now also
2435 raised for inbound connections, if the main configuration event_action
2438 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2439 early OCSP support, but no stapling (appears to be less than 1.0.0).
2441 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2442 server certificate names available under the smtp transport option
2443 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2446 JH/10 Time-related extraction expansions from certificates now use the main
2447 option "timezone" setting for output formatting, and are consistent
2448 between OpenSSL and GnuTLS compilations. Bug 1541.
2450 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2451 encoded parameter in the incoming message. Bug 1558.
2453 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2454 include certificate info, eximon was claiming there were spoolfile
2457 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2459 JH/14 Log delivery-related information more consistently, using the sequence
2460 "H=<name> [<ip>]" wherever possible.
2462 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2463 are problematic for Debian distribution, omit them from the release
2466 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2468 JH/16 Fix string representation of time values on 64bit time_t architectures.
2471 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2472 output list separator was used.
2477 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2478 checkers that were complaining about end of non-void function with no
2481 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2482 This was a regression introduced in 4.83 by another bugfix.
2484 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2486 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2487 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2493 TF/01 Correctly close the server side of TLS when forking for delivery.
2495 When a message was received over SMTP with TLS, Exim failed to clear up
2496 the incoming connection properly after forking off the child process to
2497 deliver the message. In some situations the subsequent outgoing
2498 delivery connection happened to have the same fd number as the incoming
2499 connection previously had. Exim would try to use TLS and fail, logging
2500 a "Bad file descriptor" error.
2502 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2503 utilities have not been installed.
2505 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2506 temporary space as the ACL may create new global variables.
2508 TL/01 LDAP support uses per connection or global context settings, depending
2509 upon the detected version of the libraries at build time.
2511 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2512 to extract and use the src ip:port in logging and expansions as if it
2513 were a direct connection from the outside internet. PPv2 support was
2514 updated based on HAProxy spec change in May 2014.
2516 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2518 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2519 Properly escape header and check for NULL return.
2521 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2524 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2526 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2527 characters in header names, implemented as a verify condition.
2528 Contributed by Michael Fischer v. Mollard.
2530 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2531 results permerror and temperror. Previous values are deprecated but
2532 still accepted. In a future release, err_perm and err_temp will be
2533 completely removed, which will be a backward incompatibility if the
2534 ACL tests for either of these two old results. Patch contributed by
2535 user bes-internal on the mailing list.
2537 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2539 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2540 selectors, in both main and reject logs.
2542 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2545 JH/07 Add malware type "sock" for talking to simple daemon.
2547 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2549 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2550 routers/transports under cutthrough routing.
2552 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2553 numbers. Touch up "bool" conditional to keep the same definition.
2555 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2557 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2559 JH/12 Expand items in router/transport headers_add or headers_remove lists
2560 individually rather than the list as a whole. Bug 1452.
2562 Required for reasonable handling of multiple headers_ options when
2563 they may be empty; requires that headers_remove items with embedded
2564 colons must have them doubled (or the list-separator changed).
2566 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2567 view the policy declared in the DMARC record. Currently, $dmarc_status
2568 is a combined value of both the record presence and the result of the
2571 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2573 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2574 dnslookup router and the smtp transport (applying to the forward
2577 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2578 of ldap servers used for a specific lookup. Patch provided by Heiko
2581 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2582 New variable $lookup_dnssec_authenticated for observability.
2584 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2585 Patch submitted by Lars Timman.
2587 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2589 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2590 Requires trusted mode and valid format message id, aborts otherwise.
2591 Patch contributed by Heiko Schlichting.
2593 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2594 certextract with support for various fields. Bug 1358.
2596 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2597 is requested by default, modifiable by smtp transport option
2600 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2601 operate on certificate variables to give certificate fingerprints
2602 Also new ${sha256:cert_variable}.
2604 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2606 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2607 Christian Aistleitner.
2609 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2611 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2612 file. Patch from Wolfgang Breyha.
2614 JH/25 Expand the coverage of the delivery $host and $host_address to
2615 client authenticators run in verify callout. Bug 1476.
2617 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2618 align with daemon_smtp_ports. Bug 72.
2620 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2621 support and error reporting did not work properly.
2623 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2624 and is readable. Patch from Andrew Colin Kissa.
2626 TL/14 Enhance documentation of ${run expansion and how it parses the
2627 commandline after expansion, particularly in the case when an
2628 unquoted variable expansion results in an empty value.
2630 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2632 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2633 by renaming away from it. Addresses GH issue 15, reported by
2636 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2638 TL/15 SECURITY: prevent double expansion in math comparison functions
2639 (can expand unsanitized data). Not remotely exploitable.
2646 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2648 PP/02 Make -n do something, by making it not do something.
2649 When combined with -bP, the name of an option is not output.
2651 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2654 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2655 $sender_host_name and config options to manage this, and basic check
2658 PP/05 DSCP support for outbound connections and control modifier for inbound.
2660 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2661 (Only plugin which currently uses this is kerberos4, which nobody should
2662 be using, but we should make it available and other future plugins might
2663 conceivably use it, even though it would break NAT; stuff *should* be
2664 using channel bindings instead).
2666 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2667 name; added for Sendmail compatibility; requires admin caller.
2668 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2669 just ignore it); requires trusted caller.
2670 Also parse but ignore: -Ac -Am -X<logfile>
2673 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2675 TL/02 Add +smtp_confirmation as a default logging option.
2677 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2678 Patch by Magnus Holmgren from 2007-02-20.
2680 TL/04 Bugzilla 1281 - Spec typo.
2681 Bugzilla 1283 - Spec typo.
2682 Bugzilla 1290 - Spec grammar fixes.
2684 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2686 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2688 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2689 mailing list by by Dmitry Isaikin.
2691 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2693 JH/02 Support "G" suffix to numbers in ${if comparisons.
2695 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2697 NM/01 Bugzilla 1197 - Spec typo
2698 Bugzilla 1196 - Spec examples corrections
2700 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2702 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2703 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2706 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2707 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2709 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2710 "acl {{name}{arg}...}", and optional args on acl condition
2713 JH/05 Permit multiple router/transport headers_add/remove lines.
2715 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2717 JH/07 Avoid using a waiting database for a single-message-only transport.
2718 Performance patch from Paul Fisher. Bugzilla 1262.
2720 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2723 JH/09 Add $headers_added variable, with content from use of ACL modifier
2724 add_header (but not yet added to the message). Bugzilla 199.
2726 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2727 Pulled from Bugzilla 817 by Wolfgang Breyha.
2729 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2731 (nb: this is the same fix as in Exim 4.80.1)
2733 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2736 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2739 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2741 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2742 advertise SMTP AUTH mechanism to us, instead of a generic
2743 protocol violation error. Also, make Exim more robust to bad
2744 data from the Dovecot auth socket.
2746 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2748 When a queue runner is handling a message, Exim first routes the
2749 recipient addresses, during which it prunes them based on the retry
2750 hints database. After that it attempts to deliver the message to
2751 any remaining recipients. It then updates the hints database using
2754 So if a recipient address works intermittently, it can get repeatedly
2755 deferred at routing time. The retry hints record remains fresh so the
2756 address never reaches the final cutoff time.
2758 This is a fairly common occurrence when a user is bumping up against
2759 their storage quota. Exim had some logic in its local delivery code
2760 to deal with this. However it did not apply to per-recipient defers
2761 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2763 This change adds a proper retry rule check during routing so that the
2764 final cutoff time is checked against the message's age. We only do
2765 this check if there is an address retry record and there is not a
2766 domain retry record; this implies that previous attempts to handle
2767 the address had the retry_use_local_parts option turned on. We use
2768 this as an approximation for the destination being like a local
2769 delivery, as in LMTP.
2771 I suspect this new check makes the old local delivery cutoff check
2772 redundant, but I have not verified this so I left the code in place.
2774 TF/02 Correct gecos expansion when From: is a prefix of the username.
2776 Test 0254 submits a message to Exim with the header
2780 When I ran the test suite under the user fanf2, Exim expanded
2781 the header to contain my full name, whereas it should have added
2782 a Resent-Sender: header. It erroneously treats any prefix of the
2783 username as equal to the username.
2785 This change corrects that bug.
2787 GF/01 DCC debug and logging tidyup
2788 Error conditions log to paniclog rather than rejectlog.
2789 Debug lines prefixed by "DCC: " to remove any ambiguity.
2791 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2793 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2794 Bug spotted by Jeremy Harris; was flawed since initial commit.
2795 Would have resulted in OCSP responses post-SNI triggering an Exim
2796 NULL dereference and crash.
2798 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2800 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2801 Bug detection, analysis and fix by Samuel Thibault.
2802 Bugzilla 1331, Debian bug #698092.
2804 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2806 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2807 Server implementation by Todd Lyons, client by JH.
2808 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2809 config variable "prdr_enable" controls whether the server
2810 advertises the facility. If the client requests PRDR a new
2811 acl_data_smtp_prdr ACL is called once for each recipient, after
2812 the body content is received and before the acl_smtp_data ACL.
2813 The client is controlled by both of: a hosts_try_prdr option
2814 on the smtp transport, and the server advertisement.
2815 Default client logging of deliveries and rejections involving
2816 PRDR are flagged with the string "PRDR".
2818 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2819 fclose(). Diagnosis by Todd Lyons.
2821 PP/17 Update configure.default to handle IPv6 localhost better.
2822 Patch by Alain Williams (plus minor tweaks).
2825 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2826 This is now consistent with GnuTLS, and is now documented: the
2827 previous undocumented portable approach to treating the option as
2828 unset was to force an expansion failure. That still works, and
2829 an empty string is now equivalent.
2831 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2832 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2833 not performing validation itself.
2835 PP/20 Added force_command boolean option to pipe transport.
2836 Patch from Nick Koston, of cPanel Inc.
2838 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2841 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2843 PP/21 Fix eximon continuous updating with timestamped log-files.
2844 Broken in a format-string cleanup in 4.80, missed when I repaired the
2845 other false fix of the same issue.
2846 Report and fix from Heiko Schlichting.
2849 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2850 Report from Prashanth Katuri.
2852 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2853 It's SecureTransport, so affects any MacOS clients which use the
2854 system-integrated TLS libraries, including email clients.
2856 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2857 using a MIME ACL for non-SMTP local injection.
2858 Report and assistance in diagnosis by Warren Baker.
2860 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2862 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2864 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2865 last id that failed so it may be referenced in subsequent ACL's.
2867 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2870 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2871 ldap library initialization, allowing self-signed CA's to be
2872 used. Also properly sets require_cert option later in code by
2873 using NULL (global ldap config) instead of ldap handle (per
2874 session). Bug diagnosis and testing by alxgomz.
2876 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2877 the src/util/ subdirectory.
2879 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2880 renamed to Transport Post Delivery Action by Jeremy Harris, as
2883 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2884 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2885 redis_servers = needs to be configured which will be used by the redis
2886 lookup. Patch from Warren Baker, of The Packet Hub.
2888 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2890 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2891 hostname or reverse DNS when processing a host list. Used suggestions
2892 from multiple comments on this bug.
2894 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2896 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2897 Missed a few lines, added it to make the runtest require no keyboard
2900 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2901 contains upper case chars. Make router use caseful_local_part.
2903 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2904 support when GnuTLS has been built with p11-kit.
2910 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2912 This, or similar/improved, will also be change PP/11 of 4.82.
2918 PP/01 Handle short writes when writing local log-files.
2919 In practice, only affects FreeBSD (8 onwards).
2920 Bugzilla 1053, with thanks to Dmitry Isaikin.
2922 NM/01 Bugzilla 949 - Documentation tweak
2924 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2927 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2929 PP/02 Implemented gsasl authenticator.
2931 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2933 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2934 `pkg-config foo` for cflags/libs.
2936 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2937 with rest of GSASL and with heimdal_gssapi.
2939 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2940 `pkg-config foo` for cflags/libs for the TLS implementation.
2942 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2943 properties get this fed in as external SSF. A number of robustness
2944 and debugging improvements to the cyrus_sasl authenticator.
2946 PP/08 cyrus_sasl server now expands the server_realm option.
2948 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2949 Patch by Jeremy Harris.
2951 PP/10 Added dbmjz lookup type.
2953 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2955 PP/12 MAIL args handles TAB as well as SP, for better interop with
2956 non-compliant senders.
2957 Analysis and variant patch by Todd Lyons.
2959 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2960 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2961 Patch from Dirk Mueller <dmueller@suse.com>
2963 PP/13 tls_peerdn now print-escaped for spool files.
2964 Observed some $tls_peerdn in wild which contained \n, which resulted
2965 in spool file corruption.
2967 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2968 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2969 or write after TLS renegotiation, which otherwise led to messages
2972 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2973 as a tracking header (ie: a signed header comes before the signature).
2974 Patch from Wolfgang Breyha.
2976 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2977 comma-sep list; embedded commas doubled.
2979 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2981 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2983 Report and patch from Dmitry Banschikov.
2985 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2986 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2987 failures. We appear to now support TLS1.1+ with Exim.
2989 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2990 lets Exim select keys and certificates based upon TLS SNI from client.
2991 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2992 before an outbound SMTP session. New log_selector, +tls_sni.
2994 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2995 NULL dereference. Report and patch from Alun Jones.
2997 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2998 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2999 Not seeing resolver debug output on NetBSD, but suspect this is a
3000 resolver implementation change.
3002 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3003 Left warnings. Added "eximon gdb" invocation mode.
3005 PP/21 Defaulting "accept_8bitmime" to true, not false.
3007 PP/22 Added -bw for inetd wait mode support.
3009 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3010 locate the relevant includes and libraries. Made this the default.
3012 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3013 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3015 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3016 This may cause build issues on older platforms.
3018 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3019 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3020 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3021 Added SNI support via GnuTLS too.
3022 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3024 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3026 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3027 Applied second patch from Janne, implementing suggestion to default
3028 multiple-strings-in-record handling to match SPF spec.
3030 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3032 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3033 read-only, out of scope).
3034 Patch from Wolfgang Breyha, report from Stuart Northfield.
3036 PP/29 Fix three issues highlighted by clang analyser static analysis.
3037 Only crash-plausible issue would require the Cambridge-specific
3038 iplookup router and a misconfiguration.
3039 Report from Marcin Mirosław.
3041 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3043 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3044 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3045 As part of this, removing so much warning spew let me fix some minor
3046 real issues in debug logging.
3048 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3049 assignment on my part. Fixed.
3051 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3052 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3053 Janne Snabb (who went above and beyond: thank you).
3055 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3056 string otherwise requires a connection and a bunch more work and it's
3057 relatively easy to get wrong. Should also expose TLS library linkage
3060 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3061 64-bit ${eval} (JH/03).
3063 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3064 GNU libc to support some of the 64-bit stuff, should not lead to
3065 conflicts. Defined before os.h is pulled in, so if a given platform
3066 needs to override this, it can.
3068 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3069 protection layer was required, which is not implemented.
3070 Bugzilla 1254, patch from Wolfgang Breyha.
3072 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3073 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3074 tls_dhparam take prime identifiers. Also unbreak combination of
3075 OpenSSL+DH_params+TLSSNI.
3077 PP/39 Disable SSLv2 by default in OpenSSL support.
3083 PP/01 Solaris build fix for Oracle's LDAP libraries.
3084 Bugzilla 1109, patch from Stephen Usher.
3086 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3088 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3091 TF/02 Fix a couple more cases where we did not log the error message
3092 when unlink() failed. See also change 4.74-TF/03.
3094 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3095 lock up or crash if it happened to be inside a call to libc when it
3096 got a SIGUSR1 from exiwhat.
3098 The SIGUSR1 handler appends the current process status to the process
3099 log which is later printed by exiwhat. It used to use the general
3100 purpose logging code to do this, but several functions it calls are
3101 not safe for signals.
3103 The new output code in the SIGUSR1 handler is specific to the process
3104 log, and simple enough that it's easy to inspect for signal safety.
3105 Removing some special cases also simplifies the general logging code.
3106 Removing the spurious timestamps from the process log simplifies
3109 TF/04 Improved ratelimit ACL condition.
3111 The /noupdate option has been deprecated in favour of /readonly which
3112 has clearer semantics. The /leaky, /strict, and /readonly update modes
3113 are mutually exclusive. The update mode is no longer included in the
3114 database key; it just determines when the database is updated. (This
3115 means that when you upgrade Exim will forget old rate measurements.)
3117 Exim now checks that the per_* options are used with an update mode that
3118 makes sense for the current ACL. For example, when Exim is processing a
3119 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3120 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3121 must specify per_mail/readonly. If you omit the update mode it defaults to
3122 /leaky where that makes sense (as before) or /readonly where required.
3124 The /noupdate option is now undocumented but still supported for
3125 backwards compatibility. It is equivalent to /readonly except that in
3126 ACLs where /readonly is required you may specify /leaky/noupdate or
3127 /strict/noupdate which are treated the same as /readonly.
3129 A useful new feature is the /count= option. This is a generalization
3130 of the per_byte option, so that you can measure the throughput of other
3131 aggregate values. For example, the per_byte option is now equivalent
3132 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3134 The per_rcpt option has been generalized using the /count= mechanism
3135 (though it's more complicated than the per_byte equivalence). When it is
3136 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3137 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3138 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3139 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3140 Note that using per_rcpt with a non-readonly update mode in more than
3141 one ACL will cause the recipients to be double-counted. (The per_mail
3142 and per_byte options don't have this problem.)
3144 The handling of very low rates has changed slightly. If the computed rate
3145 is less than the event's count (usually one) then this event is the first
3146 after a long gap. In this case the rate is set to the same as this event's
3147 count, so that the first message of a spam run is counted properly.
3149 The major new feature is a mechanism for counting the rate of unique
3150 events. The new per_addr option counts the number of different
3151 recipients that someone has sent messages to in the last time period. It
3152 behaves like per_rcpt if all the recipient addresses are different, but
3153 duplicate recipient addresses do not increase the measured rate. Like
3154 the /count= option this is a general mechanism, so the per_addr option
3155 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3156 example, measure the rate that a client uses different sender addresses
3157 with the options per_mail/unique=$sender_address. There are further
3158 details in the main documentation.
3160 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3162 TF/06 Removed a few PCRE remnants.
3164 TF/07 Automatically extract Exim's version number from tags in the git
3165 repository when doing development or release builds.
3167 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3168 Bugzilla 879. Patch from Paul Fisher.
3170 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3171 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3174 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3176 PP/05 Variable $av_failed, true if the AV scanner deferred.
3177 Bugzilla 1078. Patch from John Horne.
3179 PP/06 Stop make process more reliably on build failure.
3180 Bugzilla 1087. Patch from Heiko Schlittermann.
3182 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3183 Bugzilla 1089. Patch from Heiko Schlittermann.
3185 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3186 Bugzilla 1131. Patch from Holger Weiß.
3188 PP/09 Handle IPv6 addresses with SPF.
3189 Bugzilla 860. Patch from Wolfgang Breyha.
3191 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3193 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3196 PP/11 match_* no longer expand right-hand-side by default.
3197 New compile-time build option, EXPAND_LISTMATCH_RHS.
3198 New expansion conditions, "inlist", "inlisti".
3200 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3202 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3204 PP/14 fix log_write() format string regression from TF/03.
3205 Bugzilla 1152. Patch from Dmitry Isaikin.
3211 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3213 PP/02 Harmonised TLS library version reporting; only show if debugging.
3214 Layout now matches that introduced for other libraries in 4.74 PP/03.
3216 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3218 PP/04 New "dns_use_edns0" global option.
3220 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3223 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3224 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3226 TK/01 Updated PolarSSL code to 0.14.2.
3227 Bugzilla 1097. Patch from Andreas Metzler.
3229 PP/07 Catch divide-by-zero in ${eval:...}.
3230 Fixes bugzilla 1102.
3232 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3235 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3236 format-string attack -- SECURITY: remote arbitrary code execution.
3238 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3239 time unintentionally subject to list matching rules, letting the header
3240 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3241 arbitrary string expansion). This allowed for information disclosure.
3243 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3244 INT_MIN/-1 -- value coerced to INT_MAX.
3250 NM/01 Workaround for PCRE version dependency in version reporting
3253 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3254 This fixes portability to compilers other than gcc, notably
3255 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3257 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3258 makefiles for portability to HP-UX and POSIX correctness.
3260 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3261 Also via indented variable definition in the Makefile.
3262 (Debugging by Oliver Heesakkers).
3264 PP/02 Restore caching of spamd results with expanded spamd_address.
3265 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3267 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3268 Improves build reliability. Fix from: Frank Elsner
3270 NM/02 Fix wide character breakage in the rfc2047 coding
3271 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3273 NM/03 Allow underscore in dnslist lookups
3274 Fixes bug 1026. Patch from Graeme Fowler
3276 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3277 Code patches from Adam Ciarcinski of NetBSD.
3279 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3282 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3283 is logged, to avoid truncation. Patch from John Horne.
3285 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3286 Patch from Jakob Hirsch.
3288 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3289 SQL string expansion failure details.
3290 Patch from Andrey Oktyabrski.
3292 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3293 Patch from Simon Arlott.
3295 PP/09 New lookups functionality failed to compile on old gcc which rejects
3296 extern declarations in function scope.
3297 Patch from Oliver Fleischmann
3299 PP/10 Use sig_atomic_t for flags set from signal handlers.
3300 Check getgroups() return and improve debugging.
3301 Fixed developed for diagnosis in bug 927 (which turned out to be
3304 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3305 Patch from Mark Zealey.
3307 PP/12 Bugzilla 1056: Improved spamd server selection.
3308 Patch from Mark Zealey.
3310 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3311 Based on patch from Heiko Schlittermann.
3313 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3314 Patch from Uwe Doering, sign-off by Michael Haardt.
3316 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3317 variable declaration deep within a block. Bug and patch from
3320 PP/15 lookups-Makefile IRIX compatibility coercion.
3322 PP/16 Make DISABLE_DKIM build knob functional.
3324 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3325 Patch by Simon Arlott
3327 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3328 variable argument macros. Our copy now differs from upstream.
3334 TF/01 Failure to get a lock on a hints database can have serious
3335 consequences so log it to the panic log.
3337 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3338 controlled using the smtp_confirmation log selector.
3340 TF/03 Include the error message when we fail to unlink a spool file.
3342 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3343 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3344 for maintaining out-of-tree patches for some time.
3346 PP/01 Bugzilla 139: Documentation and portability issues.
3347 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3348 Handle per-OS dynamic-module compilation flags.
3350 PP/02 Let /dev/null have normal permissions.
3351 The 4.73 fixes were a little too stringent and complained about the
3352 permissions on /dev/null. Exempt it from some checks.
3353 Reported by Andreas M. Kirchwitz.
3355 PP/03 Report version information for many libraries, including
3356 Exim version information for dynamically loaded libraries. Created
3357 version.h, now support a version extension string for distributors
3358 who patch heavily. Dynamic module ABI change.
3360 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3361 privilege escalation vulnerability whereby the Exim run-time user
3362 can cause root to append content of the attacker's choosing to
3365 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3368 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3369 If dropping privileges for untrusted macros, we disabled normal logging
3370 on the basis that it would fail; for the Exim run-time user, this is not
3371 the case, and it resulted in successful deliveries going unlogged.
3372 Fixed. Reported by Andreas Metzler.
3378 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3379 only prepend for the Resent-* case. Fixes regression introduced in
3380 Exim 4.70 by NM/22 for Bugzilla 607.
3382 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3383 increasing numbers of administrators be bitten by this.
3385 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3387 PP/03 Bugzilla 994: added openssl_options main configuration option.
3389 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3391 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3393 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3395 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3396 without a peer certificate, leading to a segfault because of an
3397 assumption that peers always have certificates. Be a little more
3398 paranoid. Problem reported by Martin Tscholak.
3400 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3401 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3402 NB: ClamAV planning to remove STREAM in "middle of 2010".
3403 CL also introduces -bmalware, various -d+acl logging additions and
3404 more caution in buffer sizes.
3406 PP/09 Implemented reverse_ip expansion operator.
3408 PP/10 Bugzilla 937: provide a "debug" ACL control.
3410 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3412 PP/12 Bugzilla 973: Implement --version.
3414 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3416 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3418 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3420 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3421 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3422 ignore trailing whitespace.
3424 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3426 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3429 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3430 Notification from Dr Andrew Aitchison.
3432 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3433 ExtendedDetectionInfo response format.
3434 Notification from John Horne.
3436 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3439 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3440 XSL and documented dependency on system catalogs, with examples of how
3443 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3446 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3447 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3448 configuration file which is writeable by the Exim user or group.
3450 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3451 of configuration files to cover files specified with the -C option if
3452 they are going to be used with root privileges, not just the default
3455 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3456 option (effectively making it always true).
3458 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3459 files to be used while preserving root privileges.
3461 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3462 that rogue child processes cannot use them.
3464 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3465 run-time user, instead of root.
3467 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3468 Exim run-time user without dropping privileges.
3470 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3471 result string, instead of calling string_vformat() twice with the same
3474 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3475 for other users. Others should always drop root privileges if they use
3476 -C on the command line, even for a whitelisted configure file.
3478 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3480 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3486 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3487 $data_path, and $header_path variables; fixed documentation bugs and
3490 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3491 exipick to access non-standard spools, including the "frozen" queue
3494 NM/01 Bugzilla 965: Support mysql stored procedures.
3495 Patch from Alain Williams
3497 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3499 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3500 Patch from Andreas Metzler
3502 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3503 Patch from Kirill Miazine
3505 NM/05 Bugzilla 671: Added umask to procmail example.
3507 JJ/03 installed exipick 20100323.0, fixing doc bug
3509 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3510 directory. Notification and patch from Dan Rosenberg.
3512 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3514 TK/02 Improve log output when DKIM signing operation fails.
3516 MH/01 Treat the transport option dkim_domain as a colon separated
3517 list, not as a single string, and sign the message with each element,
3518 omitting multiple occurences of the same signer.
3520 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3521 Bugzilla 985, 986. Patch by Simon Arlott
3523 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3524 Patch by Simon Arlott
3526 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3527 MBX locking. Notification from Dan Rosenberg.
3533 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3535 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3537 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3539 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3541 NM/04 Bugzilla 915: Flag broken perl installation during build.
3547 TK/01 Added patch by Johannes Berg that expands the main option
3548 "spamd_address" if it starts with a dollar sign.
3550 TK/02 Write list of recipients to X-Envelope-Sender header when building
3551 the mbox-format spool file for content scanning (suggested by Jakob
3554 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3555 (http://www.dcc-servers.net/) support via dccifd. Activated by
3556 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3558 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3559 by Mark Daniel Reidel <mr@df.eu>.
3561 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3562 When building exim an external PCRE library is now needed -
3563 PCRE is a system library on the majority of modern systems.
3564 See entry on PCRE_LIBS in EDITME file.
3566 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3567 conversation. Added nologin parameter to request.
3568 Patch contributed by Kirill Miazine.
3570 TF/01 Do not log submission mode rewrites if they do not change the address.
3572 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3574 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3575 log files in place. Contributed by Roberto Lima.
3577 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3579 TF/03 Bugzilla 615: When checking the local_parts router precondition
3580 after a local_part_suffix or local_part_prefix option, Exim now
3581 does not use the address's named list lookup cache, since this
3582 contains cached lookups for the whole local part.
3584 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3585 Robert Millan. Documentation is in experimental-spec.txt.
3587 TF/04 Bugzilla 668: Fix parallel build (make -j).
3589 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3591 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3592 Patch provided by Jan Srzednicki.
3594 TF/05 Leading white space used to be stripped from $spam_report which
3595 wrecked the formatting. Now it is preserved.
3597 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3598 that they are available at delivery time.
3600 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3602 TF/08 TLS error reporting now respects the incoming_interface and
3603 incoming_port log selectors.
3605 TF/09 Produce a more useful error message if an SMTP transport's hosts
3606 setting expands to an empty string.
3608 NM/06 Bugzilla 744: EXPN did not work under TLS.
3609 Patch provided by Phil Pennock.
3611 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3612 Patch provided by Richard Godbee.
3614 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3615 acl_smtp_notquit, added index entry.
3617 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3618 Patch provided by Eugene Bujak.
3620 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3621 accept(). Patch provided by Maxim Dounin.
3623 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3624 Patch provided by Phil Pennock.
3626 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3628 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3629 Patch provided by Brad "anomie" Jorsch.
3631 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3632 Patch provided by Dean Brooks.
3634 TK/05 Add native DKIM support (does not depend on external libraries).
3636 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3637 Patch provided by Graeme Fowler.
3639 NM/16 Bugzilla 851: Documentation example syntax fix.
3641 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3643 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3646 NM/19 Bugzilla 745: TLS version reporting.
3647 Patch provided by Phil Pennock.
3649 NM/20 Bugzilla 167: bool: condition support.
3650 Patch provided by Phil Pennock.
3652 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3653 clients. Patch provided by Phil Pennock.
3655 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3656 Patch provided by Brad "anomie" Jorsch.
3658 NM/23 Bugzilla 687: Fix misparses in eximstats.
3659 Patch provided by Heiko Schlittermann.
3661 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3662 Patch provided by Heiko Schlittermann.
3664 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3665 plus update to original patch.
3667 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3669 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3670 Patch provided by David Brownlee.
3672 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3674 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3676 NM/30 Bugzilla 888: TLS documentation bugfixes.
3678 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3680 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3681 Unlike the original bugzilla I have changed all shell scripts in src tree.
3683 NM/33 Bugzilla 898: Transport filter timeout fix.
3684 Patch by Todd Rinaldo.
3686 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3687 Patch by Serge Demonchaux.
3689 NM/35 Bugzilla 39: Base64 decode bug fixes.
3690 Patch by Jakob Hirsch.
3692 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3694 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3696 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3698 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3704 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3705 ALT-N's libdkim that I have put here:
3706 http://duncanthrax.net/exim-experimental/
3708 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3709 were called 'true' and it seems that C99 defines that as a reserved
3710 keyword to be used with 'bool' variable types. That means you could
3711 not include C99-style headers which use bools without triggering
3712 build errors in sieve.c.
3714 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3715 as mailq or other aliases. Changed the --help handling significantly
3716 to do whats expected. exim_usage() emits usage/help information.
3718 SC/01 Added the -bylocaldomain option to eximstats.
3720 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3722 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3724 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3730 PH/01 Another patch from the Sieve maintainer.
3732 PH/02 When an IPv6 address is converted to a string for single-key lookup
3733 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3734 dots are used instead of colons so that keys in lsearch files need not
3735 contain colons. This was done some time before quoting was made available
3736 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3737 (notated using the quote facility) so as to distinguish them from IPv4
3738 keys. This meant that lookups for IP addresses in host lists did not work
3739 for iplsearch lookups.
3741 This has been fixed by arranging for IPv6 addresses to be expressed with
3742 colons if the lookup type is iplsearch. This is not incompatible, because
3743 previously such lookups could never work.
3745 The situation is now rather anomalous, since one *can* have colons in
3746 ordinary lsearch keys. However, making the change in all cases is
3747 incompatible and would probably break a number of configurations.
3749 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3752 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3753 conversion specification without a maximum field width, thereby enabling
3754 a rogue spamd server to cause a buffer overflow. While nobody in their
3755 right mind would setup Exim to query an untrusted spamd server, an
3756 attacker that gains access to a server running spamd could potentially
3757 exploit this vulnerability to run arbitrary code as the Exim user.
3759 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3760 $primary_hostname instead of what libspf2 thinks the hosts name is.
3762 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3763 a directory entry by the name of the lookup key. Previously, if a
3764 symlink pointed to a non-existing file or a file in a directory that
3765 Exim lacked permissions to read, a lookup for a key matching that
3766 symlink would fail. Now it is enough that a matching directory entry
3767 exists, symlink or not. (Bugzilla 503.)
3769 PH/03 The body_linecount and body_zerocount variables are now exported in the
3772 PH/04 Added the $dnslist_matched variable.
3774 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3775 This means they are set thereafter only if the connection becomes
3778 PH/06 Added the client_condition to authenticators so that some can be skipped
3779 by clients under certain conditions.
3781 PH/07 The error message for a badly-placed control=no_multiline_responses left
3782 "_responses" off the end of the name.
3784 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3786 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3787 (without spaces) instead of just copying the configuration text.
3789 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3791 PH/11 Added $max_received_linelength.
3793 PH/12 Added +ignore_defer and +include_defer to host lists.
3795 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3796 way in which PCRE > 7.0 is built.
3798 PH/14 Implemented queue_only_load_latch.
3800 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3801 MAIL command. The effect was to mangle the value on 64-bit systems.
3803 PH/16 Another patch from the Sieve maintainer.
3805 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3807 PH/18 If a system quota error occurred while trying to create the file for
3808 a maildir delivery, the message "Mailbox is full" was not appended to the
3809 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3810 only to a quota excession during the actual writing of the file.
3812 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3813 characters?) which causes problems in log lines. The DN values are now
3814 passed through string_printing() before being added to log lines.
3816 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3817 and InterBase are left for another time.)
3819 PH/21 Added message_body_newlines option.
3821 PH/22 Guard against possible overflow in moan_check_errorcopy().
3823 PH/23 POSIX allows open() to be a macro; guard against that.
3825 PH/24 If the recipient of an error message contained an @ in the local part
3826 (suitably quoted, of course), incorrect values were put in $domain and
3827 $local_part during the evaluation of errors_copy.
3833 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3834 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3837 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3838 issue a MAIL command.
3840 PH/02 In an ACL statement such as
3842 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3844 if a client was not listed at all, or was listed with a value other than
3845 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3846 the condition was not true (as it should be), so access was not denied.
3847 The bug was that the ! inversion was incorrectly passed on to the second
3848 item. This has been fixed.
3850 PH/03 Added additional dnslists conditions == and =& which are different from
3851 = and & when the dns lookup returns more than one IP address.
3853 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3854 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3856 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3857 FSYNC, which compiles an option called disable_fsync that allows for
3858 bypassing fsync(). The documentation is heavily laced with warnings.
3860 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3862 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3863 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3864 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3865 including adding "make clean"; (3) Added -fPIC when compiling the test
3866 dynamically loaded module, to get rid of a warning.
3868 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3869 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3870 = 0s. The bug is otherwise harmless.
3872 PH/07 There was a bug in the dovecot authenticator such that the value of
3873 $auth1 could be overwritten, and so not correctly preserved, after a
3874 successful authentication. This usually meant that the value preserved by
3875 the server_setid option was incorrect.
3877 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3879 PH/09 Installed PCRE release 7.0.
3881 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3882 run for batched SMTP input. It is now run at the start of every message
3883 in the batch. While fixing this I discovered that the process information
3884 (output by running exiwhat) was not always getting set for -bs and -bS
3885 input. This is fixed, and it now also says "batched" for BSMTP.
3887 PH/11 Added control=no_pipelining.
3889 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3890 patch, slightly modified), and move the expansion of helo_data till after
3891 the connection is made in the smtp transport (so it can use these
3894 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3896 PH/14 Added log_selector = +pid.
3898 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3900 PH/16 Add ${if forany and ${if forall.
3902 PH/17 Added dsn_from option to vary the From: line in DSNs.
3904 PH/18 Flush SMTP output before performing a callout, unless control =
3905 no_callout_flush is set.
3907 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3908 was true (the default) a successful delivery failed to delete the retry
3909 item, thus causing premature timeout of the address. The bug is now
3912 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3914 PH/21 Long custom messages for fakedefer and fakereject are now split up
3915 into multiline responses in the same way that messages for "deny" and
3916 other ACL rejections are.
3918 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3919 with slight modification.
3921 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3922 draft, changing the syntax and factoring some duplicate code".
3924 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3925 for deliveries of the second and subsequent messages over the same SMTP
3928 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3929 ${reduce, with only minor "tidies".
3931 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3933 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3934 expansion side effects.
3936 PH/27 When a message times out after an over-quota error from an Exim-imposed
3937 quota, the bounce message says "mailbox is full". This message was not
3938 being given when it was a system quota that was exceeded. It now should
3941 MH/03 Made $recipients available in local_scan(). local_scan() already has
3942 better access to the recipient list through recipients_list[], but
3943 $recipients can be useful in postmaster-provided expansion strings.
3945 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3946 in the case of a MAIL command with additional options following the
3947 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3948 were accidentally chopped off.
3950 PH/29 SMTP synchronization checks are implemented when a command is read -
3951 there is a check that no more input is waiting when there shouldn't be
3952 any. However, for some commands, a delay in an ACL can mean that it is
3953 some time before the response is written. In this time, more input might
3954 arrive, invalidly. So now there are extra checks after an ACL has run for
3955 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3956 pipelining has not been advertised.
3958 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3960 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3961 returned in the SMTP response when smtp_return_error_details was false.
3962 This has been fixed.
3964 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3965 instead of the C I/O that was originally supplied, because problems were
3966 reported on Solaris.
3968 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3969 Exim which did not show up earlier: it was assuming that a call to
3970 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3971 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3972 was a macro that became an assignment, so it seemed to work. This has
3973 changed to a proper function call with a void return, hence the compile
3974 error. Exim's code has been fixed.
3976 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3979 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3981 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3983 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3984 because it turns out to be too big or there is a timeout) there is no
3985 "Completed" line in the log. When some messages of this type were
3986 selected by exigrep, they were listed as "not completed". Others were
3987 picked up by some special patterns. I have improved the selection
3988 criteria to be more general.
3990 PH/38 The host_find_failed option in the manualroute router can now be set
3991 to "ignore", to completely ignore a host whose IP address cannot be
3992 found. If all hosts are ignored, the behaviour is controlled by the new
3993 host_all_ignored option.
3995 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3996 homing or because of multiple MX records with /mx) generated more than
3997 one IP address, and the following item turned out to be the local host,
3998 all the secondary addresses of the first item were incorrectly removed
3999 from the list, along with the local host and any following hosts (which
4000 is what is supposed to happen).
4002 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4003 whoever called Exim into the -H file. In the case of the daemon it was
4004 behaving confusingly. When first started, it used values for whoever
4005 started the daemon, but after a SIGHUP it used the Exim user (because it
4006 calls itself on a restart). I have changed the code so that it now always
4009 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4010 message are rejected with the same error (e.g. no authentication or bad
4011 sender address), and a DATA command is nevertheless sent (as can happen
4012 with PIPELINING or a stupid MUA), the error message that was given to the
4013 RCPT commands is included in the rejection of the DATA command. This is
4014 intended to be helpful for MUAs that show only the final error to their
4017 PH/42 Another patch from the Sieve maintainer.
4019 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4020 Eximstats - Fixed some broken HTML links and added missing column headers
4022 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4023 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4025 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4027 PH/43 Yet another patch from the Sieve maintainer.
4029 PH/44 I found a way to check for a TCP/IP connection going away before sending
4030 the response to the final '.' that terminates a message, but only in the
4031 case where the client has not sent further data following the '.'
4032 (unfortunately, this is allowed). However, in many cases there won't be
4033 any further data because there won't be any more messages to send. A call
4034 to select() can be used: if it shows that the input is "ready", there is
4035 either input waiting, or the socket has been closed. An attempt to read
4036 the next input character can distinguish the two cases. Previously, Exim
4037 would have sent an OK response which the client would never have see.
4038 This could lead to message repetition. This fix should cure that, at
4039 least in a lot of common cases.
4041 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4042 advertised in response to EHLO.
4048 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4049 fixed by 4.65/MH/01 (is this a record?) are fixed:
4051 (i) An empty string was always treated as zero by the numeric comparison
4052 operators. This behaviour has been restored.
4054 (ii) It is documented that the numeric comparison operators always treat
4055 their arguments as decimal numbers. This was broken in that numbers
4056 starting with 0 were being interpreted as octal.
4058 While fixing these problems I realized that there was another issue that
4059 hadn't been noticed. Values of message_size_limit (both the global option
4060 and the transport option) were treated as octal if they started with 0.
4061 The documentation was vague. These values are now always treated as
4062 decimal, and I will make that clear in the documentation.
4068 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4069 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4072 MH/01 Don't check that the operands of numeric comparison operators are
4073 integers when their expansion is in "skipping" mode (fixes bug
4074 introduced by 4.64-PH/07).
4076 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4077 child addresses, Exim now panics and dies. Previously, because the count
4078 is held in a short int, deliveries were likely to be lost. As such a
4079 large number of recipients for a single message is ridiculous
4080 (performance will be very, very poor), I have chosen to impose a limit
4081 rather than extend the field.
4087 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4088 leftover -K file (the existence of which was triggered by #402).
4089 While we were at it, introduced process PID as part of the -K
4090 filename. This should rule out race conditions when creating
4093 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4094 processing considerably. Previous code took too long for large mails,
4095 triggering a timeout which in turn triggers #401.
4097 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4098 in the DK code in transports.c. sendfile() is not really portable,
4099 hence the _LINUX specificness.
4101 TF/01 In the add_headers option to the mail command in an Exim filter,
4102 there was a bug that Exim would claim a syntax error in any
4103 header after the first one which had an odd number of characters
4106 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4107 callout verification, Exim cached a "reject" for the entire domain. This
4108 is correct for most verifications, but it is not correct for a recipient
4109 verification with use_sender or use_postmaster set, because in that case
4110 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4111 case of MAIL FROM:<> rejection from other early rejections (e.g.
4112 rejection of HELO). When verifying a recipient using a non-null MAIL
4113 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4114 Whatever the result of the callout, the value of the domain cache is
4115 left unchanged (for any other kind of callout, getting as far as trying
4116 RCPT means that the domain itself is ok).
4118 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4121 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4122 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4123 session, the daemon ignores ECONNECT errors and logs others; it now
4124 ignores EPIPE as well.
4126 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4127 (quoted-printable decoding).
4129 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4130 later the small subsequent patch to fix an introduced bug.
4132 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4134 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4136 PH/08 An error is now given if message_size_limit is specified negative.
4138 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4139 to be given (somewhat) arbitrary names.
4141 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4144 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4145 miscellaneous code fixes
4147 PH/10 Added the log_reject_target ACL modifier to specify where to log
4150 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4151 hostname. This is wrong, because it relates to the incoming message (and
4152 probably the interface on which it is arriving) and not to the outgoing
4153 callout (which could be using a different interface). This has been
4154 changed to use the value of the helo_data option from the smtp transport
4155 instead - this is what is used when a message is actually being sent. If
4156 there is no remote transport (possible with a router that sets up host
4157 addresses), $smtp_active_hostname is used.
4159 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4160 tweaks were necessary in order to get it to work (see also 21 below):
4161 (a) The code assumed that strncpy() returns a negative number on buffer
4162 overflow, which isn't the case. Replaced with Exim's string_format()
4164 (b) There were several signed/unsigned issues. I just did the minimum
4165 hacking in of casts. There is scope for a larger refactoring.
4166 (c) The code used strcasecmp() which is not a standard C function.
4167 Replaced with Exim's strcmpic() function.
4168 (d) The code set only $1; it now sets $auth1 as well.
4169 (e) A simple test gave the error "authentication client didn't specify
4170 service in request". It would seem that Dovecot has changed its
4171 interface. Fortunately there's a specification; I followed it and
4172 changed what the client sends and it appears to be working now.
4174 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4177 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4178 address A is aliased to B and C, where B exists and C does not. Without
4179 -v the output is "A verified" because verification stops after a
4180 successful redirection if more than one address is generated. However,
4181 with -v the child addresses are also verified. Exim was outputting "A
4182 failed to verify" and then showing the successful verification for C,
4183 with its parentage. It now outputs "B failed to verify", showing B's
4184 parentage before showing the successful verification of C.
4186 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4187 look up a TXT record in a specific list after matching in a combined
4190 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4191 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4192 they consult the DNS. I had assumed they would set it the way they
4193 wanted; and indeed my experiments on Linux seem to show that in some
4194 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4195 To be on the safe side, however, I have now made the interface to
4196 host_find_byname() similar to host_find_bydns(), with an argument
4197 containing the DNS resolver options. The host_find_byname() function now
4198 sets these options at its start, just as host_find_bydns() does. The smtp
4199 transport options dns_qualify_single and dns_search_parents are passed to
4200 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4201 of host_find_byname() use the default settings of RES_DEFNAMES
4202 (qualify_single) but not RES_DNSRCH (search_parents).
4204 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4205 spool_read_header() do less string testing, by means of a preliminary
4206 switch on the second character of optional "-foo" lines. (This is
4207 overdue, caused by the large number of possibilities that now exist.
4208 Originally there were few.) While I was there, I also converted the
4209 str(n)cmp tests so they don't re-test the leading "-" and the first
4210 character, in the hope this might squeeze out yet more improvement.
4212 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4213 flag allowing group syntax was set by the header_syntax check but not
4214 turned off, possible causing trouble later; (2) The flag was not being
4215 set at all for the header_verify test, causing "group"-style headers to
4216 be rejected. I have now set it in this case, and also caused header_
4217 verify to ignore an empty address taken from a group. While doing this, I
4218 came across some other cases where the code for allowing group syntax
4219 while scanning a header line wasn't quite right (mostly, not resetting
4220 the flag correctly in the right place). These bugs could have caused
4221 trouble for malformed header lines. I hope it is now all correct.
4223 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4224 with the "reply" argument non-NULL. The code, however (which originally
4225 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4226 but it didn't always do it. This confused somebody who was copying the
4227 code for some other use. I have removed all the tests.
4229 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4230 feature that was used to support insecure browsers during the U.S. crypto
4231 embargo. It requires special client support, and Exim is probably the
4232 only MTA that supported it -- and would never use it because real RSA is
4233 always available. This code has been removed, because it had the bad
4234 effect of slowing Exim down by computing (never used) parameters for the
4235 RSA_EXPORT functionality.
4237 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4238 authenticator to fail if there's a tab character in the incoming data
4239 (there should never be unless someone is messing about, as it's supposed
4240 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4241 if the connection is using TLS or if the remote IP is the same as the
4242 local IP, and the "valid-client-cert option" if a client certificate has
4245 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4246 authenticators. This can be used for authorization after authentication
4247 succeeds. (In the case of plaintext, it servers for both authentication
4250 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4251 if any retry times were supplied.
4253 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4254 connection, where there is no client IP address to check. In this
4255 situation, the verify now always succeeds.
4257 PH/25 Applied John Jetmore's -Mset patch.
4259 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4261 PH/27 In a string expansion for a processed (not raw) header when multiple
4262 headers of the same name were present, leading whitespace was being
4263 removed from all of them, but trailing whitespace was being removed only
4264 from the last one. Now trailing whitespace is removed from each header
4265 before concatenation. Completely empty headers in a concatenation (as
4266 before) are ignored.
4268 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4269 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4271 PH/29 [Removed. This was a change that I later backed out, and forgot to
4272 correct the ChangeLog entry (that I had efficiently created) before
4273 committing the later change.]
4275 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4276 address errors (4xx response to RCPT) over the same connection as other
4277 messages routed to the same hosts. Such deliveries are always "forced",
4278 so retry times are not inspected. This resulted in far too many retries
4279 for the affected addresses. The effect occurred only when there were more
4280 hosts than the hosts_max_try setting in the smtp transport when it had
4281 the 4xx errors. Those hosts that it had tried were not added to the list
4282 of hosts for which the message was waiting, so if all were tried, there
4283 was no problem. Two fixes have been applied:
4285 (i) If there are any address or message errors in an SMTP delivery, none
4286 of the hosts (tried or untried) are now added to the list of hosts
4287 for which the message is waiting, so the message should not be a
4288 candidate for sending over the same connection that was used for a
4289 successful delivery of some other message. This seems entirely
4290 reasonable: after all the message is NOT "waiting for some host".
4291 This is so "obvious" that I'm not sure why it wasn't done
4292 previously. Hope I haven't missed anything, but it can't do any
4293 harm, as the worst effect is to miss an optimization.
4295 (ii) If, despite (i), such a delivery is accidentally attempted, the
4296 routing retry time is respected, so at least it doesn't keep
4297 hammering the server.
4299 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4300 in ${readsocket because some servers need this prod.
4302 PH/32 Added some extra debug output when updating a wait-xxx database.
4304 PH/33 The hint "could be header name not terminated by colon", which has been
4305 given for certain expansion errors for a long time, was not being given
4306 for the ${if def:h_colon_omitted{... case.
4308 PH/34 The spec says: "With one important exception, whenever a domain list is
4309 being scanned, $domain contains the subject domain." There was at least
4310 one case where this was not true.
4312 PH/35 The error "getsockname() failed: connection reset by peer" was being
4313 written to the panic log as well as the main log, but it isn't really
4314 panic-worthy as it just means the connection died rather early on. I have
4315 removed the panic log writing for the ECONNRESET error when getsockname()
4318 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4319 runs only) independently of the message's sender address. This meant
4320 that, if the 4xx error was in fact related to the sender, a different
4321 message to the same recipient with a different sender could confuse
4322 things. In particular, this can happen when sending to a greylisting
4323 server, but other circumstances could also provoke similar problems.
4324 I have changed the default so that the retry time for these errors is now
4325 based a combination of the sender and recipient addresses. This change
4326 can be overridden by setting address_retry_include_sender=false in the
4329 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4330 remote server are returned as part of bounce messages. This was not
4331 happening for LMTP over a pipe (the lmtp transport), but now it is the
4332 same for both kinds of LMTP.
4334 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4335 in header lines that were in fact CNAMEs. This is no longer the case.
4337 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4338 and queue runs started by the daemon processed all messages. This has
4339 been fixed so that -R and -S can now usefully be given with -q<time>.
4341 PH/40 Import PCRE release 6.7 (fixes some bugs).
4343 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4345 PH/42 Give an error if -q is specified more than once.
4347 PH/43 Renamed the variables $interface_address and $interface_port as
4348 $received_ip_address and $received_port, to make it clear that these
4349 values apply to message reception, and not to the outgoing interface when
4350 a message is delivered. (The old names remain recognized, of course.)
4352 PH/44 There was no timeout on the connect() call when using a Unix domain
4353 socket in the ${readsocket expansion. There now is.
4355 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4356 be meaningful with "accept".
4358 SC/01 Eximstats V1.43
4359 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4361 SC/02 Eximstats V1.44
4362 Use a glob alias rather than an array ref in the generated
4363 parser. This improves both readability and performance.
4365 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4366 Collect SpamAssassin and rejection statistics.
4367 Don't display local sender or destination tables unless
4368 there is data to show.
4369 Added average volumes into the top table text output.
4371 SC/04 Eximstats V1.46
4372 Collect data on the number of addresses (recipients)
4373 as well as the number of messages.
4375 SC/05 Eximstats V1.47
4376 Added 'Message too big' to the list of mail rejection
4377 reasons (thanks to Marco Gaiarin).
4379 SC/06 Eximstats V1.48
4380 Mainlog lines which have GMT offsets and are too short to
4381 have a flag are now skipped.
4383 SC/07 Eximstats V1.49 (Alain Williams)
4384 Added the -emptyok flag.
4386 SC/08 Eximstats V1.50
4387 Fixes for obtaining the IP address from reject messages.
4389 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4390 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4391 whitespace changes from 4.64-PH/27
4393 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4396 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4397 are found, allow negative numbers in numeric criteria)
4399 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4401 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4402 to match changes made in 4.64-PH/43
4404 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4406 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4407 "domain" is actually an IP address, and force a failure. This locks out
4408 those revolvers/nameservers that support "A-for-A" lookups, in
4409 contravention of the specifications.
4411 PH/48 When a host name was looked up from an IP address, and the subsequent
4412 forward lookup of the name timed out, the host name was left in
4413 $sender_host_name, contrary to the specification.
4415 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4416 restricted to single-key lookups, Exim was not diagnosing an error if
4417 * or *@ was used with a query-style lookup.
4419 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4421 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4422 long ago, but noone interested enough thought of it. Let's just say that
4423 the "1.1" means that there are some new functions that weren't there at
4424 some point in the past.
4426 PH/51 Error processing for expansion failure of helo_data from an smtp
4427 transport during callout processing was broken.
4429 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4430 tested/used via the -bh/-bhc/-bs options.
4432 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4433 bug, fixed in subsequent PCRE releases).
4435 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4436 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4438 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4444 SC/01 Use a glob alias rather than an array ref in eximstats generated
4445 parser. This improves both readability and performance.
4447 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4448 Don't display local sender or destination tables in eximstats unless
4449 there is data to show.
4450 Added average volumes into the eximstats top table text output.
4452 SC/03 Collect data on the number of addresses (recipients) as well
4453 as the number of messages in eximstats.
4455 TF/01 Correct an error in the documentation for the redirect router. Exim
4456 does (usually) call initgroups() when daemonizing.
4458 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4459 with consistent privilege compared to when running as a daemon.
4461 TF/03 Note in the spec that $authenticated_id is not set for local
4462 submissions from trusted users.
4464 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4465 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4467 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4468 by adding some example configuration directives to the default
4469 configuration file. A little bit of work is required to uncomment the
4470 directives and define how usernames and passwords are checked, but
4471 there is now a framework to start from.
4473 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4474 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4475 without this. I don't know how relevant this is to other LDAP libraries.
4477 PH/02 Add the verb name to the "unknown ACL verb" error.
4479 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4481 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4483 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4484 directory not expanded when it should be if an expanded home directory
4485 was set for the address (which is overridden by the transport).
4487 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4490 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4491 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4492 because it is too late at that time, and has no effect.
4494 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4495 security issue with \' (bugzilla #107). I could not use the
4496 PQescapeStringConn() function, because it needs a PGconn value as one of
4499 PH/08 When testing addresses using -bt, indicate those final addresses that
4500 are duplicates that would not cause an additional delivery. At least one
4501 person was confused, thinking that -bt output corresponded to deliveries.
4502 (Suppressing duplicates isn't a good idea as you lose the information
4503 about possibly different redirections that led to the duplicates.)
4505 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4506 systems where poll() doesn't work, in particular OS X.
4508 PH/10 Added more information to debugging output for retry time not reached.
4510 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4511 operations in malware.c.
4513 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4516 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4517 syslog_duplication set false, log lines that would normally be written
4518 both the the main log and to the reject log were not written to syslog at
4521 PH/14 In the default configuration, change the use of "message" in ACL warn
4522 statements to "add_header".
4524 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4525 not followed by a command (e.g. "seen endif").
4527 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4528 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4531 PH/17 Added extra conditions to the default value of delay_warning_condition
4535 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4536 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4537 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4540 The Auto-Submitted: and various List- headers are standardised, whereas I
4541 don't think Precedence: ever was.
4543 PH/18 Refactored debugging code in route_finduser() to show more information,
4544 in particular, the error code if getpwnam() issues one.
4546 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4547 This is apparently needed in addition to the PH/07 change above to avoid
4548 any possible encoding problems.
4550 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4551 but not after initializing Perl.
4553 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4554 output them only if debugging. By default they are written stderr,
4555 apparently, which is not desirable.
4557 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4560 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4563 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4565 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4566 authenticated or an ident call has been made. Suppress the default
4567 values for $authenticated_id and $authenticated_sender (but permit -oMai
4568 and -oMas) when testing with -bh.
4570 PH/24 Re-jigged the order of the tests in the default configuration so that the
4571 tests for valid domains and recipients precede the DNS black list and CSA
4572 tests, on the grounds that those ones are more expensive.
4574 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4575 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4576 This bug exists in every version of Exim that I still have, right back to
4579 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4580 However, an attempt to turn on case-sensitivity in a regex key by
4581 including (?-i) didn't work because the subject string was already
4582 lowercased, and the effects were non-intuitive. It turns out that a
4583 one-line patch can be used to allow (?-i) to work as expected.
4589 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4590 other effects) broke the use of negated acl sub-conditions.
4592 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4595 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4596 "Deny" causes Exim to reject the incoming connection with a 554 error.
4597 Unfortunately, if there is a major crisis, such as a disk failure,
4598 tcp-wrappers gives "deny", whereas what one would like would be some
4599 kind of temporary error. A kludge has been added to help with this.
4600 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4601 554 error is used if errno is still zero or contains ENOENT (which occurs
4602 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4605 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4607 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4608 errors. Otherwise a message that provokes a temporary error (when other
4609 messages do not) can cause a whole host to time out.
4611 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4612 addresses were routed directly to files or pipes from a redirect router.
4613 File deliveries just didn't batch; pipe deliveries might have suffered
4616 PH/06 A failure to get a lock for a hints database would erroneously always say
4617 "Failed to get write lock", even when it was really a read lock.
4619 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4620 of 0600. This has been changed to use the value of the lockfile_mode
4621 option (which defaults to 0600).
4623 PH/08 Applied small patch from the Sieve maintainer.
4625 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4626 folder from quota calculations, a direct delivery into this folder messed
4627 up the contents of the maildirsize file. This was because the regex was
4628 used only to exclude .Trash (or whatever) when the size of the mailbox
4629 was calculated. There was no check that a delivery was happening into an
4630 excluded directory. This bug has been fixed by ignoring all quota
4631 processing for deliveries into excluded directories.
4633 PH/10 Added the maildirfolder_create_regex option to appendfile.
4639 PH/01 The code for finding all the local interface addresses on a FreeBSD
4640 system running IPv6 was broken. This may well have applied to all BSD
4641 systems, as well as to others that have similar system calls. The broken
4642 code found IPv4 interfaces correctly, but gave incorrect values for the
4643 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4644 that it would not match correctly against @[] and not recognize the IPv6
4647 PH/02 The ipliteral router was not recognizing addresses of the form user@
4648 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4650 PH/03 Added disable_ipv6.
4652 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4653 decoded form, because it is most often used to construct To: headers
4654 lines in autoreplies, and the decoded form may well be syntactically
4655 invalid. However, $reply_address has leading white space removed, and all
4656 newlines turned into spaces so that the autoreply transport does not
4659 PH/05 If group was specified without a user on a router, and no group or user
4660 was specified on a transport, the group from the router was ignored.
4662 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4663 for visible compile-time settings that can be used to change these
4664 numbers, for those that want even more. Backwards compatibility with old
4665 spool files has been maintained. However, going back to a previous Exim
4666 release will lost any variables that are in spool files.
4668 PH/07 Two small changes when running in the test harness: increase delay when
4669 passing a TCP/IP connection to a new process, in case the original
4670 process has to generate a bounce, and remove special handling of
4671 127.0.0.2 (sic), which is no longer necessary.
4673 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4674 be the same on different OS.
4676 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4679 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4680 whether --show-vars was specified or not
4682 JJ/02 exipick: Added support for new ACL variable spool format introduced
4685 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4686 syntactically invalid From: or Reply-to: line, and a filter used this to
4687 generate an autoreply, and therefore failed to obtain an address for the
4688 autoreply, Exim could try to deliver to a non-existent relative file
4689 name, causing unrelated and misleading errors. What now happens is that
4690 it logs this as a hard delivery error, but does not attempt to create a
4693 PH/11 The exinext utility has a -C option for testing purposes, but although
4694 the given file was scanned by exinext itself; it wasn't being passed on
4695 when Exim was called.
4697 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4698 an end-of-file indication when reading a command response.
4700 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4701 compiled. In many other places in Exim, IPv6 addresses are always
4702 recognized, so I have changed this. It also means that IPv4 domain
4703 literals of the form [IPV4:n.n.n.n] are now always recognized.
4705 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4706 used if the router is not running as root, for example, when verifying at
4707 ACL time, or when using -bh. The debugging output from this situation was
4708 non-existent - all you got was a failure to exec. I have made two
4711 (a) Failures to set uid/gid, the current directory, or a process leader
4712 in a subprocess such as that created by queryprogram now generate
4713 suitable debugging output when -d is set.
4715 (b) The queryprogram router detects when it is not running as root,
4716 outputs suitable debugging information if -d is set, and then runs
4717 the subprocess without attempting to change uid/gid.
4719 PH/15 Minor change to Makefile for building test_host (undocumented testing
4722 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4723 additional section of a DNS packet that returns MX or SRV records.
4724 Instead, it always explicitly searches for A/AAAA records. This avoids
4725 major problems that occur when a DNS server includes only records of one
4726 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4727 fixed another bug: if SRV records were looked up and the corresponding
4728 address records were *not* found in the additional section, the port
4729 values from the SRV records were lost.
4731 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4732 using the correct key (the original address) when searching the retry
4733 rules in order to find which one to use for generating the retry hint.
4735 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4736 adding the default one. Similarly, if it contains a Reply-To: header, the
4737 errors_reply_to option, if set, is not used.
4739 PH/19 When calculating a retry time, Exim used to measure the "time since
4740 failure" by looking at the "first failed" field in the retry record. Now
4741 it does not use this if it is later than than the arrival time of the
4742 message. Instead it uses the arrival time. This makes for better
4743 behaviour in cases where some deliveries succeed, thus re-setting the
4744 "first failed" field. An example is a quota failure for a huge message
4745 when small messages continue to be delivered. Without this change, the
4746 "time since failure" will always be short, possible causing more frequent
4747 delivery attempts for the huge message than are intended.
4748 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4750 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4751 $1, $2, $3) because the numerical variables can be reset during some
4752 expansion items (e.g. "match"), thereby losing the authentication data.
4754 PH/21 Make -bV show the size of off_t variables so that the test suite can
4755 decide whether to run tests for quotas > 2G.
4757 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4758 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4759 filecount value is greater than 2G or if a quota value is greater than 2G
4760 on a system where the size of off_t is not greater than 4, a panic error
4763 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4764 never match. The debug and -bh output now contains an explicit error
4765 message indicating a malformed IPv4 address or mask.
4767 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4768 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4769 PH/23 above applies.
4771 PH/25 Do not write to syslog when running in the test harness. The only
4772 occasion when this arises is a failure to open the main or panic logs
4773 (for which there is an explicit test).
4775 PH/26 Added the /no_tell option to "control=freeze".
4777 PH/27 If a host name lookup failed very early in a connection, for example, if
4778 the IP address matched host_lookup and the reverse lookup yielded a name
4779 that did not have a forward lookup, an error message of the form "no IP
4780 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4781 could be logged. Now it outputs the IP address instead of "NULL".
4783 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4784 allows the sender and the authenticated sender to be set when
4785 submitting a message from within Exim. Since child_open_exim() is
4786 documented for local_scan(), the new function should be too.
4788 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4789 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4790 results in an empty string is now treated as unset.
4792 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4794 PH/31 Added sender_verify_fail logging option.
4796 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4797 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4798 tidied the source and removed it altogether.
4800 PH/33 When a queue run was abandoned because the load average was too high, a
4801 log line was always written; now it is written only if the queue_run log
4802 selector is set. In addition, the log line for abandonment now contains
4803 information about the queue run such as the pid. This is always present
4804 in "start" and "stop" lines but was omitted from the "abandon" line.
4806 PH/34 Omit spaces between a header name and the colon in the error message that
4807 is given when verify = headers_syntax fails (if there are lots of them,
4808 the message gets confusing).
4810 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4811 names, as there are now some PTR records that contain slashes. This check
4812 is only to protect against broken name servers that fall over on strange
4813 characters, so the fact that it applies to all lookups doesn't matter.
4815 PH/36 Now that the new test suite is complete, we can remove some of the
4816 special code in Exim that was needed for the old test suite. For example,
4817 sorting DNS records because real resolvers return them in an arbitrary
4818 order. The new test suite's fake resolver always returns records in the
4821 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4822 bounces) except when queue_only is set, to avoid logging races between
4823 the different processes.
4825 PH/38 Panic-die if .include specifies a non-absolute path.
4827 PH/39 A tweak to the "H" retry rule from its user.
4829 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4830 a label. They prevented compilation on older perls.
4832 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4833 a warning to be raised on newish perls.
4835 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4836 on queue. Changes to match documented behaviour of showing count of
4837 messages matching specified criteria.
4839 PH/40 Changed the default ident timeout from 30s to 5s.
4841 PH/41 Added support for the use of login_cap features, on those BSD systems
4842 that have them, for controlling the resources used by pipe deliveries.
4844 PH/42 The content-scanning code uses fopen() to create files in which to put
4845 message data. Previously it was not paying any attention to the mode of
4846 the files. Exim runs with umask(0) because the rest of the code creates
4847 files with open(), and sets the required mode explicitly. Thus, these
4848 files were ending up world-writeable. This was not a big issue, because,
4849 being within the spool directory, they were not world-accessible. I have
4850 created a function called modefopen, which takes an additional mode
4851 argument. It sets umask(777), creates the file, chmods it to the required
4852 mode, then resets the umask. All the relevant calls to fopen() in the
4853 content scanning code have been changed to use this function.
4855 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4856 to 24 hours. This avoids potential overflow problems when processing G
4857 and H retry rules. I suspect nobody ever tinkers with this value.
4859 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4861 PH/45 When the plaintext authenticator is running as a client, the server's
4862 challenges are checked to ensure they are valid base64 strings. By
4863 default, the authentication attempt is cancelled if an invalid string is
4864 received. Setting client_ignore_invalid_base64 true ignores these errors.
4865 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4866 they are received. Thus, the responses can be made to depend on the
4867 challenges. If an invalid string is ignored, an empty string is placed in
4870 PH/46 Messages that are created by the autoreply transport now contains a
4871 References: header, in accordance with RFCs 2822 and 3834.
4873 PH/47 Added authenticated_sender_force to the smtp transport.
4875 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4877 PH/49 Installed latest patch from the Sieve maintainer.
4879 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4880 was also set, the appendfile transport was unnecessarily scanning a
4881 directory of message files (e.g. for maildir delivery) to find the count
4882 of files (along with the size), even though it did not need this
4883 information. It now does the scan only if it needs to find either the
4884 size of the count of files.
4886 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4888 PH/52 Two bugs concerned with error handling when the smtp transport is
4891 (i) Exim was not creating retry information for temporary errors given
4892 for individual recipients after the DATA command when the smtp transport
4893 was used in LMTP mode. This meant that they could be retried too
4894 frequently, and not timed out correctly.
4896 (ii) Exim was setting the flag that allows error details to be returned
4897 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4898 recipients that were returned after the DATA command.
4900 PH/53 This is related to PH/52, but is more general: for any failing address,
4901 when detailed error information was permitted to be returned to the
4902 sender, but the error was temporary, then after the final timeout, only
4903 "retry timeout exceeded" was returned. Now it returns the full error as
4904 well as "retry timeout exceeded".
4906 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4907 do this, and (what is worse) MTAs that accept it.
4909 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4910 will now be deprecated.
4912 PH/56 New os.c-cygwin from the Cygwin maintainer.
4914 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4915 formats (previously only available in exim formats via -bpr, -bpru,
4916 and -bpra. Now also available in native and exiqgrep formats)
4918 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4919 with very large, slow to parse queues
4921 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4923 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4925 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4926 responses to authentication challenges, though it was showing the
4927 challenges; (ii) I've removed the CR characters from the debug output for
4930 PH/58 Allow for the insertion of a newline as well as a space when a string
4931 is turned into more than one encoded-word during RFC 2047 encoding. The
4932 Sieve code now uses this.
4934 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4935 data_4xx, lost_connection, tls_required.
4937 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4938 message was being sent as an SMTP response.
4940 PH/61 Add -l and -k options to exicyclog.
4942 PH/62 When verifying, if an address was redirected to one new address, so that
4943 verification continued, and the new address failed or deferred after
4944 having set something in $address_data, the value of $address_data was not
4945 passed back to the ACL. This was different to the case when no
4946 redirection occurred. The value is now passed back in both cases.
4948 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4949 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4950 use login_cap.h, so on its own it isn't the distinguishing feature. The
4951 new name refers directly to the setclassresources() function.
4953 PH/65 Added configuration files for NetBSD3.
4955 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4957 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4958 is preferred over IPv4.
4960 PH/68 The bounce_return_message and bounce_return_body options were not being
4961 honoured for bounces generated during the reception of non-SMTP messages.
4962 In particular, this applied to messages rejected by the ACL. This bug has
4963 been fixed. However, if bounce_return_message is true and bounce_return_
4964 body is false, the headers that are returned for a non-SMTP message
4965 include only those that have been read before the error was detected.
4966 (In the case of an ACL rejection, they have all been read.)
4968 PH/69 The HTML version of the specification is now built in a directory called
4969 spec_html instead of spec.html, because the latter looks like a path with
4970 a MIME-type, and this confuses some software.
4972 PH/70 Catch two compiler warnings in sieve.c.
4974 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4975 function verify_get_ident() calls ip_connect() to connect a socket, but
4976 if the "connect()" function timed out, ip_connect() used to close the
4977 socket. However, verify_get_ident() also closes the socket later, and in
4978 between Exim writes to the log, which may get opened at this point. When
4979 the socket was closed in ip_connect(), the log could get the same file
4980 descriptor number as the socket. This naturally causes chaos. The fix is
4981 not to close the socket in ip_connect(); the socket should be closed by
4982 the function that creates it. There was only one place in the code where
4983 this was missing, in the iplookup router, which I don't think anybody now
4984 uses, but I've fixed it anyway.
4986 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4987 well as to direct DNS lookups. Otherwise the handling of names in host
4988 lists is inconsistent and therefore confusing.
4994 PH/01 Two changes to the default runtime configuration:
4996 (1) Move the checks for relay_from_hosts and authenticated clients from
4997 after to before the (commented out) DNS black list checks.
4999 (2) Add control=submission to the relay_from_hosts and authenticated
5000 clients checks, on the grounds that messages accepted by these
5001 statements are most likely to be submissions.
5003 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5005 (1) Generate an error if the third argument for the ${prvs expansion is
5008 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5011 (3) Reset the variables that are obtained from the first argument of
5012 ${prvscheck and used in the second argument before leaving the code,
5013 because their memory is reclaimed, so using them afterwards may do
5016 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5017 one (it's much easier than Tom thought :-).
5019 (5) Because of (4), we can now allow for the use of $prvscheck_result
5020 inside the third argument.
5022 PH/03 For some reason, the default setting of PATH when running a command from
5023 a pipe transport was just "/usr/bin". I have changed it to
5026 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5027 anything to be listed in the output from -bV.
5029 PH/05 When a filter generated an autoreply, the entire To: header line was
5030 quoted in the delivery log line, like this:
5032 => >A.N.Other <ano@some.domain> <original@ddress> ...
5034 This has been changed so that it extracts the operative address. There
5035 may be more than one such address. If so, they are comma-separated, like
5038 => >ano@some.domain,ona@other.domain <original@ddress> ...
5040 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5041 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5042 not being looked up in the rDNS to get a host name, Exim was showing the
5043 IP address twice in Received: lines, even though the IP addresses were
5044 identical. For example:
5046 Received: from [1.2.3.4] (helo=[1.2.3.4])
5048 However, if the real host name was known, it was omitting the HELO data
5049 if it matched the actual IP address. This has been tidied up so that it
5050 doesn't show the same IP address twice.
5052 PH/07 When both +timestamp and +memory debugging was on, the value given by
5053 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5054 called by the debug printing, thereby overwriting the timestamp buffer.
5055 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5057 PH/08 When the original message was included in an autoreply transport, it
5058 always said "this is a copy of the message, including all the headers",
5059 even if body_only or headers_only was set. It now gives an appropriate
5062 PH/09 Applied a patch from the Sieve maintainer which:
5064 o fixes some comments
5065 o adds the (disabled) notify extension core
5066 o adds some debug output for the result of if/elsif tests
5067 o points to the current vacation draft in the documentation
5068 and documents the missing references header update
5072 o fixes a bug in processing the envelope test (when testing
5073 multiple envelope elements, the last element determined the
5076 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5077 Electronic Mail") by including:
5079 Auto-submitted: auto-generated
5081 in the messages that it generates (bounce messages and others, such as
5082 warnings). In the case of bounce messages for non-SMTP messages, there was
5083 also a typo: it was using "Auto_submitted" (underscore instead of
5084 hyphen). Since every message generated by Exim is necessarily in response
5085 to another message, thes have all been changed to:
5087 Auto-Submitted: auto-replied
5089 in accordance with these statements in the RFC:
5091 The auto-replied keyword:
5093 - SHOULD be used on messages sent in direct response to another
5094 message by an automatic process,
5096 - MUST NOT be used on manually-generated messages,
5098 - MAY be used on Delivery Status Notifications (DSNs) and Message
5099 Disposition Notifications (MDNs),
5101 - MUST NOT be used on messages generated by automatic or periodic
5102 processes, except for messages which are automatic responses to
5105 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5106 to the default Received: header definition.
5108 PH/12 Added log selector acl_warn_skipped (default on).
5110 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5111 variables because (a) they are in the wrong storage pool and (b) even if
5112 they were copied, it wouldn't work properly because of the caching.
5114 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5115 checking when decoding. Apparently there are clients that generate
5116 overlong encoded strings. Why am I not surprised?
5118 PH/15 If the first argument of "${if match_address" was not empty, but did not
5119 contain an "@" character, Exim crashed. Now it writes a panic log message
5120 and treats the condition as false.
5122 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5124 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5125 extension "envelope-auth". The code is finished and in agreement with
5126 other implementations, but there is no documentation so far and in fact,
5127 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5128 not changing the active code.
5130 Print executed "if" and "elsif" statements when debugging is used. This
5131 helps a great deal to understand what a filter does.
5133 Document more things not specified clearly in RFC3028. I had all this
5134 sorted out, when out of a sudden new issues came to my mind. Oops."
5136 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5139 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5140 canonical form (as documented). However, after a host name lookup from
5141 the IP address, check_host() was doing a simple string comparison with
5142 addresses acquired from the DNS when checking that the found name did
5143 have the original IP as one of its addresses. Since any found IPv6
5144 addresses are likely to be in abbreviated form, the comparison could
5145 fail. Luckily, there already exists a function for doing the comparison
5146 by converting both addresses to binary, so now that is used instead of
5147 the text comparison.
5149 PH/20 There was another similar case to PH/19, when a complete host name was
5150 given in a host list; looking up its IP address could give an abbreviated
5151 form, whereas the current host's name might or might not be abbreviated.
5152 The same fix has been applied.
5158 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5159 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5162 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5163 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5165 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5167 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5168 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5169 Also, for "Auto-Submitted", treat anything other than "no" as
5170 non-personal, in accordance with RFC 3834. (Previously it treated
5171 anything starting "auto-" as non-personal.)
5173 TF/01 The control=submission/name=... option had a problem with syntax
5174 errors if the name included a slash character. The /name= option
5175 now slurps the rest of the string, so it can include any characters
5176 but it must come last in the list of options (after /sender_retain
5179 PH/05 Some modifications to the interface to the fake nameserver for the new
5187 TK/01 Added the "success_on_redirect" address verification option. See
5188 NewStuff for rationale and an example.
5190 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5192 PH/02 Patch to exigrep to allow it to work on syslog lines.
5194 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5195 fread() to skip over the body file's header line, because in Cygwin the
5196 header line is locked and is inaccessible.
5198 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5199 co-exist for some time) to make it clear that it is the Exim ID that is
5200 referenced, not the Message-ID: header line.
5202 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5203 string_format() function, because snprintf() does not exist on all
5206 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5207 ${stat: expansion item.
5209 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5210 protocol synchronization error", to keep the pedants happy.
5212 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5213 well as for IRIX systems, when gcc is being used. See the host.c source
5216 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5218 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5221 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5222 transport and to the smtp transport in LMTP mode.
5224 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5226 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5227 run a filter in a subprocess. This could lead to confusion in subsequent
5228 lookups in the parent process. There should also be a search_tidyup() at
5229 the end of the subprocess.
5231 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5232 only if the host matched helo_try_verify_hosts, which caused the
5233 verification to occur when the EHLO/HELO command was issued. The ACL just
5234 tested the remembered result. Now, if a previous verification attempt has
5235 not happened, "verify = helo" does it there and then.
5237 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5239 TK/03 Fix log output including CR from clamd.
5241 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5242 exist provoked a memory error which could cause a segfault.
5244 PH/15 Installed PCRE 6.2
5246 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5248 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5249 of the problem. Specifically, suggested +O2 rather than +O1 for the
5252 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5254 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5255 record, the port was not correctly logged when the outgoing_port log
5256 selector was set (it logged the transort's default port).
5258 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5259 fallback_hosts, and "hosts" in the smtp transport.
5261 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5262 host errors such as "Connection refused".
5264 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5265 authentication with radiusclient 0.4.9:
5267 - Error returned from rc_read_config was caught wrongly
5268 - Username/password not passed on to radius server due to wrong length.
5270 The presumption is that some radiusclient API changes for 4.51/PH/17
5271 were not taken care of correctly. The code is still untested by me (my
5272 Linux distribution still has 0.3.2 of radiusclient), but it was
5273 contributed by a Radius user.
5275 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5276 expanding the "port" option of the smtp transport.
5278 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5279 while reading a MIME header. Thanks to Tom Hughes for a patch.
5281 PH/24 Include config.h inside local_scan.h so that configuration settings are
5284 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5285 that in an ACL for RCPT (for example), you can examine exactly what was
5288 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5289 commands, but it was not correctly comparing the address with the actual
5290 client host address. Thus, it would show the EHLO address in Received:
5291 header lines when this was not necessary.
5293 PH/27 Added the % operator to ${eval:}.
5295 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5296 it should be ignoring failures (because with -C, for example, it has lost
5297 privilege). It wasn't ignoring creation failures other than "already
5300 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5303 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5304 because an input line was too long, either on its own, or by virtue of
5305 too many continuations, the temporary file was not being removed, and the
5306 return code was incorrect.
5308 PH/31 Missing "BOOL" in function definition in filtertest.c.
5310 PH/32 Applied Sieve patches from the maintainer.
5312 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5314 PH/33 Added "verify = not_blind".
5316 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5317 Local/Makefile (with some defaults set). These are used in built scripts
5318 such as exicyclog, but they have never been used in the exim_install
5319 script (though there are many overriding facilities there). I have
5320 arranged that the exim_install script now takes note of these two
5323 PH/35 Installed configuration files for Dragonfly.
5325 PH/36 When a locally submitted message by a trusted user did not contain a
5326 From: header, and the sender address was obtained from -f or from an SMTP
5327 MAIL command, and the trusted user did not use -F to supply a sender
5328 name, $originator_name was incorrectly used when constructing a From:
5329 header. Furthermore, $originator_name was used for submission mode
5330 messages from external hosts without From: headers in a similar way,
5331 which is clearly wrong.
5333 PH/37 Added control=suppress_local_fixups.
5335 PH/38 When log_selector = +received_sender was set, and the addition of the
5336 sender made the log line's construction buffer exactly full, or one byte
5337 less than full, an overflow happened when the terminating "\n" was
5340 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5341 when the result of a list match is failure because a DNS lookup failed.
5343 PH/40 RM_COMMAND is now used in the building process.
5345 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5346 the "build-* directories that it finds.
5348 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5349 address in a domain literal was a prefix of an interface address.
5351 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5352 when verifying a sender address, unless rewrite_headers is false.
5354 PH/44 Wrote a long comment about why errors_to addresses are verified as
5355 recipients, not senders.
5357 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5358 the ratelimit ACL was added.
5360 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5362 PH/46 Added extra information about PostgreSQL errors to the error string.
5364 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5365 avoiding the need to install special zones in a real server. This is
5366 backwards compatible; if it can't find the fake resolver, it drops back.
5367 Thus, both old and new test suites can be run.
5369 TF/02 Added util/ratelimit.pl
5371 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5372 clock is set back in time.
5374 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5375 Candler <B.Candler@pobox.com>.
5377 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5378 OK for addresses that are the result of redirections.
5380 PH/48 A number of further additions for the benefit of the new test suite,
5381 including a fake gethostbyname() that interfaces to the fake DNS resolver
5384 TF/06 The fix for widen_domains has also been applied to qualify_single and
5385 search_parents which are the other dnslookup options that can cause
5388 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5391 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5393 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5394 service" if it encountered an MX record with an empty target hostname.
5395 The message is now "an MX or SRV record indicated no SMTP service".
5397 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5398 if the DNS of the sending site is misconfigured. This is quite a
5399 common situation. This change restores the behaviour of treating a
5400 helo verification defer as a failure.
5402 PH/51 If self=fail was set on a router, the bounce message did not include the
5403 actual error message.
5409 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5411 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5412 command itself ended in error, the underlying message about the transport
5413 filter timeout was being overwritten with the pipe command error. Now the
5414 underlying error message should be appended to the second error message.
5416 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5418 PH/02 Reduce the amount of output that "make" produces by default. Full output
5419 can still be requested.
5421 PH/03 The warning log line about a condition test deferring for a "warn" verb
5422 was being output only once per connection, rather than after each
5423 occurrence (because it was using the same function as for successful
5424 "warn" verbs). This seems wrong, so I have changed it.
5426 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5427 it should not have, which might have caused a crash in the right
5428 circumstances, but probably never did.
5430 PH/04 Installed a modified version of Tony Finch's patch to make submission
5431 mode fix the return path as well as the Sender: header line, and to
5432 add a /name= option so that you can make the user's friendly name appear
5435 TF/03 Added the control = fakedefer ACL modifier.
5437 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5438 Mark Lowes for thorough testing.
5440 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5442 TK/03 Merged latest SRS patch from Miles Wilton.
5444 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5445 with the definition in sysexits.h (which is #included earlier).
5446 Fortunately, Exim does not actually use EX_OK. The code used to try to
5447 preserve the sysexits.h value, by assuming that macro definitions were
5448 scanned for macro replacements. I have been disabused of this notion,
5449 so now the code just undefines EX_OK before #including unistd.h.
5451 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5452 in the smtp transport. When a block could not be written in a single
5453 write() function, the timeout was being re-applied to each part-write.
5454 This seems wrong - if the receiver was accepting one byte at a time it
5455 would take for ever. The timeout is now adjusted when this happens. It
5456 doesn't have to be particularly precise.
5458 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5459 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5461 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5462 without a domain if the check to <postmaster@domain> fails.
5464 SC/01 Eximstats: added -xls and the ability to specify output files
5465 (patch written by Frank Heydlauf).
5467 SC/02 Eximstats: use FileHandles for outputting results.
5469 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5471 SC/04 Eximstats: fixed display of large numbers with -nvr option
5473 SC/05 Eximstats: fixed merging of reports with empty tables.
5475 SC/06 Eximstats: added the -include_original_destination flag
5477 SC/07 Eximstats: removed tabs and trailing whitespace.
5479 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5481 TK/06 MBOX spool code: Add real "From " MBOX separator line
5482 so the .eml file is really in mbox format (even though
5483 most programs do not really care). Patch from Alex Miller.
5485 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5486 The latter is generated from $received_to and is only set if the
5487 message has one envelope recipient. SA can use these headers,
5488 obviously out-of-the-box. Patch from Alex Miller.
5490 PH/08 The ${def test on a variable was returning false if the variable's
5491 value was "0", contrary to what the specification has always said!
5492 The result should be true unless the variable is empty.
5494 PH/09 The syntax error of a character other than { following "${if
5495 def:variable_name" (after optional whitespace) was not being diagnosed.
5496 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5497 accidental colon was present, for example, could give incorrect results.
5499 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5500 result is used (not including appendfile, where other changes are about
5503 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5504 This involved changing a lot of size variables from int to off_t. It
5505 should work with maildirs and everything.
5507 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5508 spamd dying while we are connected to it.
5510 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5513 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5514 function for BDB 4.3.
5516 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5518 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5519 the sources for PCRE have been re-arranged and more files are now
5522 PH/15 The code I had for printing potentially long long variables in PH/11
5523 above was not the best (it lost precision). The length of off_t variables
5524 is now inspected at build time, and an appropriate printing format (%ld
5525 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5526 to be "long long int" or "long int". This is needed for the internal
5527 formatting function string_vformat().
5529 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5530 the configuration file to be ":syslog", then the script "guesses" where
5531 the logs files are, rather than using the compiled in default. In our
5532 case the guess is not the same as the compiled default, so the script
5533 suddenly stopped working when I started to use syslog. The patch checks
5534 to see if log_file_path is "". If so, it attempts to read it from exim
5535 with no configuration file to get the compiled in version, before it
5536 falls back to the previous guessing code."
5538 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5539 implementing BATV in an Exim configuration. See NewStuff for the gory
5542 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5543 Makefile that are specific to HP-UX.
5545 PH/18 If the "use_postmaster" option was set for a recipient callout together
5546 with the "random" option, the postmaster address was used as the MAIL
5547 FROM address for the random test, but not for the subsequent recipient
5548 test. It is now used for both.
5550 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5551 patch removes a few documentation additions to RFC 3028, because the
5552 latest draft now contains them. It adds the new en;ascii-case comparator
5553 and a new error check for 8bit text in MIME parts. Comparator and
5554 require names are now matched exactly. I enabled the subaddress
5555 extension, but it is not well tested yet (read: it works for me)."
5557 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5558 rework some of the code of TK/09 above to avoid the hardwired use of
5559 "%lld" and "long long". Replaced the call to snprintf() with a call to
5562 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5563 records point to non-existent hosts", "retry timeout exceeded", and
5564 "retry time not reached for any host after a long failure period".
5566 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5567 experimental DomainKeys support:
5569 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5570 (2) On an error such as an illegally used "control", the wrong name for
5571 the control was given.
5573 These problems did NOT occur unless DomainKeys support was compiled.
5575 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5577 PH/24 Added ${if match_ip condition.
5579 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5580 databases so that it will be absolutely obvious if a crash occurs in the
5581 DB library. This is a regular occurrence (often caused by mis-matched
5584 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5585 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5586 picked up on a user's system that detects such things. There doesn't seem
5587 to be a gcc warning option for this - only an attribute that has to be
5588 put on the function's prototype. It seems that in Fedora Core 4 they have
5589 set this on a number of new functions. No doubt there will be more in due
5592 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5593 specify a transport. However, if an address that was verified by such a
5594 router was the subject of a callout, Exim crashed because it tried to
5595 read the rcpt_include_affixes from the non-existent transport. Now it
5596 just assumes that the setting of that option is false. This bug was
5597 introduced by 4.51/PH/31.
5599 PH/28 Changed -d+all to exclude +memory, because that information is very
5600 rarely of interest, but it makes the output a lot bigger. People tend to
5601 do -d+all out of habit.
5603 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5604 code in os-type was giving problems when libc.so lives in lib64, like on
5607 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5608 aren't the modern standard, and it seems that some systems' include files
5609 don't always have them. Exim was already checking for some of the newer
5610 ones like T_AAAA, and defining it itself. I've added checks for all the
5611 record types that Exim uses.
5613 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5614 not automatically generating a new one, as it is supposed to. This
5615 prevented TLS from working. If the file did exist, but contained invalid
5616 data, a new version was generated, as expected. It was only the case of a
5617 non-existent file that was broken.
5619 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5620 with a change in libdomainkeys > 0.64.
5622 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5623 from DNS. If the selector record carries the flag, it now has
5624 precedence over the domain-wide flag.
5626 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5628 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5629 the use of a port name that isn't defined in /etc/services) occurred, the
5630 message was deferred as in a normal delivery, and thus remained on the
5631 spool, instead of being failed because of the mua_wrapper setting. This
5632 is now fixed, and I tidied up some of the mua_wrapper messages at the
5635 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5636 the messages in a hash of arrays rather than using individual hashes.
5637 This is a bit cleaner and results in dramatic memory savings, albeit
5638 at a slight CPU cost.
5640 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5641 as requested by Marc Sherman.
5643 SC/10 Eximstats: added histograms for user specified patterns as requested
5646 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5648 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5649 fopen() in the content-scanning modules that did not already have it.
5655 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5656 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5658 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5660 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5662 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5663 to test Sieve filters that use "vacation".
5665 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5666 that changes the way the GnuTLS parameters are stored in the cache file.
5667 The new format can be generated externally. For backward compatibility,
5668 if the data in the cache doesn't make sense, Exim assumes it has read an
5669 old-format file, and it generates new data and writes a new file. This
5670 means that you can't go back to an older release without removing the
5673 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5674 work if there are any delivery delays because "one_time" forces the
5675 parent to be marked "delivered", so its unseen clone is never tried
5676 again. For this reason, Exim now forbids the simultaneous setting of
5679 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5680 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5681 entry if you want to know the details. The fix, however, neglected to
5682 consider the case when local delivery batching is involved. The test for
5683 "previously delivered" was not happening when checking to see if an
5684 address could be batched with a previous (undelivered) one; under
5685 certain circumstances this could lead to multiple deliveries to the same
5688 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5689 in its include files, and this causes problems building Exim.
5691 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5692 header_syntax) but Exim was just ignoring anything given after a slash.
5693 In particular, this caused confusion with an attempt to use "verify =
5694 reverse_host_lookup/defer_ok". An error is now given when options are
5695 supplied for verify items that do not have them. (Maybe reverse_host_
5696 lookup should have a defer_ok option, but that's a different point.)
5698 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5699 defined by RFC 821) to 2048, because there were problems with some AUTH
5700 commands, and RFC 1869 says the size should be increased for extended
5701 SMTP commands that take arguments.
5703 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5706 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5707 "unknown" error; now it says that the functionality isn't in the binary.
5709 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5710 an address' error message when a string expansion fails (syntax or
5711 whatever). Otherwise the password may appear in the log. Following change
5712 PH/42 below, there is no longer a chance of it appearing in a bounce
5715 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5717 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5718 of its addresses was ever tried. (Bugzilla bug #2.)
5720 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5721 the result incorrectly in the debug output. (It correctly added a newline
5722 to what was transported.)
5724 TF/01 Added $received_time.
5726 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5727 commented out examples of how to interface to a virus scanner and to
5728 SpamAssassin. Also added commented examples of av_scanner and
5729 spamd_address settings.
5731 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5732 and controls are allowed in which ACLs. There were a couple of minor
5733 errors. Some of the entries in the conditions table (which is a table of
5734 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5735 negation of where the condition IS allowed.
5737 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5739 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5740 header file does not have a version number, so I've had to invent a new
5741 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5742 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5743 radiusclient), but it was contributed by a Radius user.
5745 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5746 files or directories, for OpenSSL.
5748 PH/19 When an Exim process that is running as root has to create an Exim log
5749 file, it does so in a subprocess that runs as exim:exim so as to get the
5750 ownership right at creation (otherwise, other Exim processes might see
5751 the file with the wrong ownership). There was no test for failure of this
5752 fork() call, which would lead to the process getting stuck as it waited
5753 for a non-existent subprocess. Forks do occasionally fail when resources
5754 run out. I reviewed all the other calls to fork(); they all seem to check
5757 PH/20 When checking for unexpected SMTP input at connect time (before writing
5758 the banner), Exim was not dealing correctly with a non-positive return
5759 from the read() function. If the client had disconnected by this time,
5760 the result was a log entry for a synchronization error with an empty
5761 string after "input=" when read() returned zero. If read() returned -1
5762 (an event I could not check), uninitialized data bytes were printed.
5763 There were reports of junk text (parts of files, etc) appearing after
5766 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5768 PH/22 Added support for macro redefinition, and (re)definition in between
5769 driver and ACL definitions.
5771 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5772 forgetting to use the resulting value; it was using the unexpanded value.
5774 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5775 hadn't been configured. The fix is from Juergen Kreileder, who
5776 understands it better than I do:
5778 "Here's what I see happening with three configured cyrus_sasl
5779 authenticators configured (plain, login, cram-md5):
5781 On startup auth_cyrus_sasl_init() gets called for each of these.
5782 This means three calls to sasl_listmech() without a specified mech_list.
5783 => SASL tests which mechs of all available mechs actually work
5784 => three warnings about OTP not working
5785 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5787 With the patch, sasl_listmech() also gets called three times. But now
5788 SASL's mech_list option is set to the server_mech specified in the the
5789 authenticator. Or in other words, the answer from sasl_listmech()
5790 gets limited to just the mech you're testing for (which is different
5792 => the return list contains just 'plain' or 'login', 'cram-md5' or
5793 nothing depending on the value of ob->server_mech.
5795 I've just tested the patch: Authentication still works fine,
5796 unavailable mechs specified in the exim configuration are still
5797 caught, and the auth.log warnings about OTP are gone."
5799 PH/25 When debugging is enabled, the contents of the command line are added
5800 to the debugging output, even when log_selector=+arguments is not
5803 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5804 answer is "GNU", and only if the return is "GNU/something" is the answer
5807 PH/27 $acl_verify_message is now set immediately after the failure of a
5808 verification in an ACL, and so is available in subsequent modifiers. In
5809 particular, the message can be preserved by coding like this:
5811 warn !verify = sender
5812 set acl_m0 = $acl_verify_message
5814 Previously, $acl_verify_message was set only while expanding "message"
5815 and "log_message" when a very denied access.
5817 PH/28 Modified OS/os.c-Linux with
5819 -#ifndef OS_LOAD_AVERAGE
5820 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5822 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5823 nomenclature these days.)
5825 PH/29 Installed patch from the Sieve maintainer that adds the options
5826 sieve_useraddress and sieve_subaddress to the redirect router.
5828 PH/30 In these circumstances:
5829 . Two addresses routed to the same list of hosts;
5830 . First host does not offer TLS;
5831 . First host accepts first address;
5832 . First host gives temporary error to second address;
5833 . Second host offers TLS and a TLS session is established;
5834 . Second host accepts second address.
5835 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5836 and peerdn, if requested) that were in fact used only for the second
5839 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5840 attention to any local part prefix or suffix that was matched by the
5841 router that accepted the address. It now behaves in the same way as it
5842 does for delivery: the affixes are removed from the local part unless
5843 rcpt_include_affixes is set on the transport.
5845 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5846 timeout during the DATA phase of an incoming message.
5848 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5849 applied a patch sanctioned by the Sieve maintainer.
5851 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5852 the uid or gid is negative. A case of a negative gid caused this to be
5853 noticed. The fix allows for either to be negative.
5855 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5856 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5859 PH/36 The change PH/12 above was broken. Fixed it.
5861 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5862 the grounds that routing the same address twice would always produce the
5863 same answer. This might have been true once, but it is certainly no
5864 longer true now. Routing a child address may depend on the previous
5865 routing that produced that child. Some complicated redirection strategies
5866 went wrong when messages had multiple recipients, and made Exim's
5867 behaviour dependent on the order in which the addresses were given.
5869 I have moved the duplicate checking until after the routing is complete.
5870 Exim scans the addresses that are assigned to local and remote
5871 transports, and removes any duplicates. This means that more work will be
5872 done, as duplicates will always all be routed, but duplicates are
5873 presumably rare, so I don't expect this is of any significance.
5875 For deliveries to pipes, files, and autoreplies, the duplicate checking
5876 still happens during the routing process, since they are not going to be
5879 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5880 It corrects a timeout issue with spamd. This is Ian's comment: "The
5881 background is that sometimes spamd either never reads data from a
5882 connection it has accepted, or it never writes response data. The exiscan
5883 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5884 blindly assumes that writes won't block so it may never time out."
5886 PH/39 Allow G after quota size as well as K and M.
5888 PH/40 The value set for $authenticated_id in an authenticator may not contain
5889 binary zeroes or newlines because the value is written to log lines and
5890 to spool files. There was no check on this. Now the value is run through
5891 the string_printing() function so that such characters are converted to
5892 printable escape sequences.
5894 PH/41 $message_linecount is a new variable that contains the total number of
5895 lines in the message. Compare $body_linecount, which is the count for the
5898 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5899 bounce and delay warning messages, except in certain special cases, which
5902 (a) An SMTP error message from a remote host;
5903 (b) A message specified in a :fail: redirection;
5904 (c) A message specified in a "fail" command in a system filter;
5905 (d) A message specified in a FAIL return from the queryprogram router;
5906 (e) A message specified by the cannot_route_message router option.
5908 In these cases only, Exim does include the error details in bounce and
5909 warning messages. There are also a few cases where bland messages such
5910 as "unrouteable address" or "local delivery error" are given.
5912 PH/43 $value is now also set for the "else" part of a ${run expansion.
5914 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5915 being worked on, but at least Exim now implements the latest version to
5918 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5919 process to complete was treated as a delivery failure, a timeout while
5920 writing the message to the pipe was logged, but erroneously treated as a
5921 successful delivery. Such timeouts include transport filter timeouts. For
5922 consistency with the overall process timeout, these timeouts are now
5923 treated as errors, giving rise to delivery failures by default. However,
5924 there is now a new Boolean option for the pipe transport called
5925 timeout_defer, which, if set TRUE, converts the failures into defers for
5926 both kinds of timeout. A transport filter timeout is now identified in
5929 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5930 systems where "make" and "gmake" are different, calling "gmake" at top
5931 level broke things. I've arranged for the value of $(MAKE) to be passed
5932 from the Makefile to this script so that it can call the same version of
5936 A note about Exim versions 4.44 and 4.50
5937 ----------------------------------------
5939 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5940 changes of various kinds. As a consequence, a big documentation update was
5941 needed. This delayed the release for rather longer than seemed good, especially
5942 in the light of a couple of (minor) security issues. Therefore, the changes
5943 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5944 release. So 4.44 and 4.50 are in effect two different branches that both start
5947 I have left the 4.50 change log unchanged; it contains all the changes since
5948 4.43. The change log for 4.44 is below; many of its items are identical to
5949 those for 4.50. This seems to be the most sensible way to preserve the
5950 historical information.
5956 1. Minor wording change to the doc/README.SIEVE file.
5958 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5959 computation of the current number of files was incorrect.
5961 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5962 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5965 4. Give more explanation in the error message when the command for a transport
5966 filter fails to execute.
5968 5. There are several places where Exim runs a non-Exim command in a
5969 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5970 was being done only for the command run by the queryprogram router. It is
5971 now done for all such subprocesses. The other cases are: ${run, transport
5972 filters, and the commands run by the lmtp and pipe transports.
5974 6. Added CONFIGURE_GROUP build-time option.
5976 7. Some older OS have a limit of 256 on the maximum number of file
5977 descriptors. Exim was using setrlimit() to set 1000 as a large value
5978 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5979 systems. I've change it so that if it can't get 1000, it tries for 256.
5981 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5982 was an oversight, and furthermore, ever since the addition of extra
5983 controls (e.g. 4.43/32), the checks on when to allow different forms of
5984 "control" were broken. There should now be diagnostics for all cases when a
5985 control that does not make sense is encountered.
5987 9. Added the /retain_sender option to "control=submission".
5989 10. $recipients is now available in the predata ACL (oversight).
5991 11. Tidy the search cache before the fork to do a delivery from a message
5992 received from the command line. Otherwise the child will trigger a lookup
5993 failure and thereby defer the delivery if it tries to use (for example) a
5994 cached ldap connection that the parent has called unbind on.
5996 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5997 of $address_data from the recipient verification was clobbered by the
5998 sender verification.
6000 13. The value of address_data from a sender verification is now available in
6001 $sender_address_data in subsequent conditions in the ACL statement.
6003 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6005 15. Added a new option "connect=<time>" to callout options, to set a different
6008 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6009 was its contents. (It was OK if the option was not defined at all.)
6011 17. A "Completed" log line is now written for messages that are removed from
6012 the spool by the -Mrm option.
6014 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6015 information about exactly what failed.
6017 19. Added -dd to debug only the daemon process.
6019 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6020 handles timeouts, both on the server side and network timeouts. Renamed the
6021 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6023 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6024 It is now set to "smtps".
6026 22. $host_address is now set to the target address during the checking of
6027 ignore_target_hosts.
6029 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6030 being passed; this would have caused $sender_host_name to have been used if
6031 matching the list had actually called for a host name (not very likely,
6032 since this list is usually IP addresses). A host name is now passed as
6035 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6036 code that specifies a non-restarting handler (typically sigaction() in
6037 modern systems) in an attempt to fix a rare and obscure crash bug.
6039 25. Narrowed the window for a race in the daemon that could cause it to ignore
6040 SIGCHLD signals. This is not a major problem, because they are used only to
6041 wake it up if nothing else does.
6043 26. A malformed maildirsize file could cause Exim to calculate negative values
6044 for the mailbox size or file count. Odd effects could occur as a result.
6045 The maildirsize information is now recalculated if the size or filecount
6048 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6049 support for a long time. Removed HAVE_SYS_VFS_H.
6051 28. Installed the latest version of exipick from John Jetmore.
6053 29. In an address list, if the pattern was not a regular expression, an empty
6054 subject address (from a bounce message) matched only if the pattern was an
6055 empty string. Non-empty patterns were not even tested. This was the wrong
6056 because it is perfectly reasonable to use an empty address as part of a
6057 database query. An empty address is now tested by patterns that are
6058 lookups. However, all the other forms of pattern expect the subject to
6059 contain a local part and a domain, and therefore, for them, an empty
6060 address still always fails if the pattern is not itself empty.
6062 30. Exim went into a mad DNS loop when attempting to do a callout where the
6063 host was specified on an smtp transport, and looking it up yielded more
6064 than one IP address.
6066 31. Re-factored the code for checking spool and log partition space into a
6067 function that finds that data and another that does the check. The former
6068 is then used to implement four new variables: $spool_space, $log_space,
6069 $spool_inodes, and $log_inodes.
6071 32. The RFC2047 encoding function was originally intended for short strings
6072 such as real names; it was not keeping to the 75-character limit for
6073 encoded words that the RFC imposes. It now respects the limit, and
6074 generates multiple encoded words if necessary. To be on the safe side, I
6075 have increased the buffer size for the ${rfc2047: expansion operator from
6078 33. It is now permitted to omit both strings after an "if" condition; if the
6079 condition is true, the result is "true". As before, when the second string
6080 is omitted, a false condition yields an empty string. This makes it less
6081 cumbersome to write custom ACL and router conditions.
6083 34. Failure to deliver a bounce message always caused it to be frozen, even if
6084 there was an errors_to setting on the router. The errors_to setting is now
6087 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6088 canonical form (fully expanded) before being placed in
6089 $sender_host_address.
6091 36. The table in the code that translates DNS record types into text (T_A to
6092 "A" for instance) was missing entries for NS and CNAME. It is just possible
6093 that this could have caused confusion if both these types were looked up
6094 for the same domain, because the text type is used as part of Exim's
6095 per-process caching. But the chance of anyone hitting this buglet seems
6098 37. The dnsdb lookup has been extended in a number of ways.
6100 (1) There is a new type, "zns", which walks up the domain tree until it
6101 finds some nameserver records. It should be used with care.
6103 (2) There is a new type, "mxh", which is like "mx" except that it returns
6104 just the host names, not the priorities.
6106 (3) It is now possible to give a list of domains (or IP addresses) to be
6107 looked up. The behaviour when one of the lookups defers can be
6108 controlled by a keyword.
6110 (4) It is now possible to specify the separator character for use when
6111 multiple records are returned.
6113 38. The dnslists ACL condition has been extended: it is now possible to supply
6114 a list of IP addresses and/or domains to be looked up in a particular DNS
6117 39. Added log_selector=+queue_time_overall.
6119 40. When running the queue in the test harness, wait just a tad after forking a
6120 delivery process, to get repeatability of debugging output.
6122 41. Include certificate and key file names in error message when GnuTLS fails
6123 to set them up, because the GnuTLS error message doesn't include the name
6124 of the failing file when there is a problem reading it.
6126 42. Allow both -bf and -bF in the same test run.
6128 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6130 44. The "Exiscan patch" is now merged into the mainline Exim source.
6132 45. Sometimes the final signoff response after QUIT could fail to get
6133 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6134 < 0 before doing a fflush(). This bug looks as though it goes back to the
6135 introduction of TLS in release 3.20, but "sometimes" must have been rare
6136 because the tests only now provoked it.
6138 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6139 (this can affect the format of dates).
6141 47. exim_tidydb, when checking for the continued existence of a message for
6142 which it has found a message-specific retry record, was not finding
6143 messages that were in split spool directories. Consequently, it was
6144 deleting retry records that should have stayed in existence.
6146 48. Steve fixed some bugs in eximstats.
6148 49. The SPA authentication driver was not abandoning authentication and moving
6149 on to the next authenticator when an expansion was forced to fail,
6150 contradicting the general specification for all authenticators. Instead it
6151 was generating a temporary error. It now behaves as specified.
6153 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6154 (the order specifies the preference for clients). The order is now AES256,
6155 AES128, 3DES, ARCFOUR128.
6157 51. Small patch to Sieve code - explicitly set From: when generating an
6160 52. Exim crashed if a remote delivery caused a very long error message to be
6161 recorded - for instance if somebody sent an entire SpamAssassin report back
6162 as a large number of 550 error lines. This bug was coincidentally fixed by
6163 increasing the size of one of Exim's internal buffers (big_buffer) that
6164 happened as part of the Exiscan merge. However, to be on the safe side, I
6165 have made the code more robust (and fixed the comments that describe what
6168 53. Now that there can be additional text after "Completed" in log lines (if
6169 the queue_time_overall log selector is set), a one-byte patch to exigrep
6170 was needed to allow it to recognize "Completed" as not the last thing in
6173 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6174 patch that reportedly fixes this has been added. I am not expert enough to
6175 create a test for it. This is what the patch creator wrote:
6177 "I found a little strange behaviour of ldap code when working with
6178 Windows 2003 AD Domain, where users was placed in more than one
6179 Organization Units. When I tried to give exim partial DN, the exit code
6180 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6181 But simultaneously result of request was absolutely normal ldap result,
6182 so I produce this patch..."
6184 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6185 REFERENCE, so I have modified the code to exclude the patch when that macro
6188 55. Some experimental protocols are using DNS PTR records for new purposes. The
6189 keys for these records are domain names, not reversed IP addresses. The
6190 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6191 leaves it alone. Component reversal etc. now happens only for IP addresses.
6194 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6196 57. Double the size of the debug message buffer (to 2048) so that more of very
6197 long debug lines gets shown.
6199 58. The exicyclog utility now does better if the number of log files to keep
6200 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6202 59. Two changes related to the smtp_active_hostname option:
6204 (1) $smtp_active_hostname is now available as a variable.
6205 (2) The default for smtp_banner uses $smtp_active_hostname instead
6206 of $primary_hostname.
6208 60. The host_aton() function is supposed to be passed a string that is known
6209 to be a valid IP address. However, in the case of IPv6 addresses, it was
6210 not checking this. This is a hostage to fortune. Exim now panics and dies
6211 if the condition is not met. A case was found where this could be provoked
6212 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6213 components; fortuitously, this particular loophole had already been fixed
6214 by change 4.50/55 above.
6216 If there are any other similar loopholes, the new check in host_aton()
6217 itself should stop them being exploited. The report I received stated that
6218 data on the command line could provoke the exploit when Exim was running as
6219 exim, but did not say which command line option was involved. All I could
6220 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6221 running as the user.
6224 61. There was a buffer overflow vulnerability in the SPA authentication code
6225 (which came originally from the Samba project). I have added a test to the
6226 spa_base64_to_bits() function which I hope fixes it.
6229 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6230 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6232 63. The daemon start-up calls getloadavg() while still root for those OS that
6233 need the first call to be done as root, but it missed one case: when
6234 deliver_queue_load_max is set with deliver_drop_privilege. This is
6235 necessary for the benefit of the queue runner, because there is no re-exec
6236 when deliver_drop_privilege is set.
6238 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6239 This has been fixed.
6241 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6242 was in use, was not putting the data itself into the right store pool;
6243 consequently, it could be overwritten for a subsequent message in the same
6244 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6247 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6249 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6250 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6251 the function were treating the return as a boolean value, which happened to
6252 work because 0=false and not-0=true, but is not correct code.
6254 68. The host_aton() function was not handling scoped IPv6 addresses (those
6255 with, for example, "%eth0" on the end) correctly.
6257 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6258 negated items (that is, ~something) in unsigned ints. Some compilers
6259 apparently mutter when there is no cast.
6261 70. If an address verification called from an ACL failed, and did not produce a
6262 user-specific message (i.e. there was only a "system" message), nothing was
6263 put in $acl_verify_message. In this situation, it now puts the system
6266 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6267 session; change 4.31/43 added the unwanted input to the log line - except
6268 that it did not do this in the start of session case. It now does.
6270 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6271 This is wrong and can cause the other end to generate a synchronization
6272 error if it is another Exim or anything else that does the synchronization
6273 check. A QUIT command is no longer sent after a timeout.
6275 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6276 during host lookups.
6278 74. The defer_ok option of callout verification was not working if it was used
6279 when verifying addresses in header lines, that is, for this case:
6281 verify = header_sender/callout=defer_ok
6283 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6284 those file descriptors could be used for SMTP connections. If anything
6285 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6286 could be sent to the SMTP client, causing chaos. The daemon now opens
6287 stdin, stdout, and stderr to /dev/null when it puts itself into the
6290 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6291 log by default. The user can override this with suitable Perl magic.
6293 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6294 the log message when discard triggers, was not working for the DATA ACL or
6295 for the non-SMTP ACL.
6297 78. Error message wording change in sieve.c.
6299 79. If smtp_accept_max_per_host was set, the number of connections could be
6300 restricted to fewer than expected, because the daemon was trying to set up
6301 a new connection before checking whether the processes handling previous
6302 connections had finished. The check for completed processes is now done
6303 earlier. On busy systems, this bug wouldn't be noticed because something
6304 else would have woken the daemon, and it would have reaped the completed
6307 80. If a message was submitted locally by a user whose login name contained one
6308 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6309 It caused a spool format error. I have fixed the spool reading code. A
6310 related problem was that the "from" clause in the Received: line became
6311 illegal because of the space(s). It is now covered by ${quote_local_part.
6313 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6316 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6318 83. Patch from Sieve maintainer to fix unterminated string problem in
6319 "vacation" handling.
6321 84. Some minor changes to the Linux configuration files to help with other
6322 OS variants using glibc.
6324 85. One more patch for Sieve to update vacation handling to latest spec.
6327 ----------------------------------------------------
6328 See the note above about the 4.44 and 4.50 releases.
6329 ----------------------------------------------------
6335 1. Change 4.43/35 introduced a bug that caused file counts to be
6336 incorrectly computed when quota_filecount was set in an appendfile
6339 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6340 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6343 3. Give more explanation in the error message when the command for a transport
6344 filter fails to execute.
6346 4. There are several places where Exim runs a non-Exim command in a
6347 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6348 was being done only for the command run by the queryprogram router. It is
6349 now done for all such subprocesses. The other cases are: ${run, transport
6350 filters, and the commands run by the lmtp and pipe transports.
6352 5. Some older OS have a limit of 256 on the maximum number of file
6353 descriptors. Exim was using setrlimit() to set 1000 as a large value
6354 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6355 systems. I've change it so that if it can't get 1000, it tries for 256.
6357 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6358 was an oversight, and furthermore, ever since the addition of extra
6359 controls (e.g. 4.43/32), the checks on when to allow different forms of
6360 "control" were broken. There should now be diagnostics for all cases when a
6361 control that does not make sense is encountered.
6363 7. $recipients is now available in the predata ACL (oversight).
6365 8. Tidy the search cache before the fork to do a delivery from a message
6366 received from the command line. Otherwise the child will trigger a lookup
6367 failure and thereby defer the delivery if it tries to use (for example) a
6368 cached ldap connection that the parent has called unbind on.
6370 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6371 of $address_data from the recipient verification was clobbered by the
6372 sender verification.
6374 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6375 was its contents. (It was OK if the option was not defined at all.)
6377 11. A "Completed" log line is now written for messages that are removed from
6378 the spool by the -Mrm option.
6380 12. $host_address is now set to the target address during the checking of
6381 ignore_target_hosts.
6383 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6384 being passed; this would have caused $sender_host_name to have been used if
6385 matching the list had actually called for a host name (not very likely,
6386 since this list is usually IP addresses). A host name is now passed as
6389 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6390 code that specifies a non-restarting handler (typically sigaction() in
6391 modern systems) in an attempt to fix a rare and obscure crash bug.
6393 15. Narrowed the window for a race in the daemon that could cause it to ignore
6394 SIGCHLD signals. This is not a major problem, because they are used only to
6395 wake it up if nothing else does.
6397 16. A malformed maildirsize file could cause Exim to calculate negative values
6398 for the mailbox size or file count. Odd effects could occur as a result.
6399 The maildirsize information is now recalculated if the size or filecount
6402 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6403 support for a long time. Removed HAVE_SYS_VFS_H.
6405 18. Updated exipick to current release from John Jetmore.
6407 19. Allow an empty sender to be matched against a lookup in an address list.
6408 Previously the only cases considered were a regular expression, or an
6411 20. Exim went into a mad DNS lookup loop when doing a callout where the
6412 host was specified on the transport, if the DNS lookup yielded more than
6415 21. The RFC2047 encoding function was originally intended for short strings
6416 such as real names; it was not keeping to the 75-character limit for
6417 encoded words that the RFC imposes. It now respects the limit, and
6418 generates multiple encoded words if necessary. To be on the safe side, I
6419 have increased the buffer size for the ${rfc2047: expansion operator from
6422 22. Failure to deliver a bounce message always caused it to be frozen, even if
6423 there was an errors_to setting on the router. The errors_to setting is now
6426 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6427 canonical form (fully expanded) before being placed in
6428 $sender_host_address.
6430 24. Updated eximstats to version 1.33
6432 25. Include certificate and key file names in error message when GnuTLS fails
6433 to set them up, because the GnuTLS error message doesn't include the name
6434 of the failing file when there is a problem reading it.
6436 26. Expand error message when OpenSSL has problems setting up cert/key files.
6439 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6440 (this can affect the format of dates).
6442 28. exim_tidydb, when checking for the continued existence of a message for
6443 which it has found a message-specific retry record, was not finding
6444 messages that were in split spool directories. Consequently, it was
6445 deleting retry records that should have stayed in existence.
6447 29. eximstats updated to version 1.35
6448 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6449 1.35 - bugfix such that pie charts by volume are generated correctly
6451 30. The SPA authentication driver was not abandoning authentication and moving
6452 on to the next authenticator when an expansion was forced to fail,
6453 contradicting the general specification for all authenticators. Instead it
6454 was generating a temporary error. It now behaves as specified.
6456 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6457 (the order specifies the preference for clients). The order is now AES256,
6458 AES128, 3DES, ARCFOUR128.
6460 31. Small patch to Sieve code - explicitly set From: when generating an
6463 32. Exim crashed if a remote delivery caused a very long error message to be
6464 recorded - for instance if somebody sent an entire SpamAssassin report back
6465 as a large number of 550 error lines. This bug was coincidentally fixed by
6466 increasing the size of one of Exim's internal buffers (big_buffer) that
6467 happened as part of the Exiscan merge. However, to be on the safe side, I
6468 have made the code more robust (and fixed the comments that describe what
6471 33. Some experimental protocols are using DNS PTR records for new purposes. The
6472 keys for these records are domain names, not reversed IP addresses. The
6473 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6474 leaves it alone. Component reversal etc. now happens only for IP addresses.
6477 34. The host_aton() function is supposed to be passed a string that is known
6478 to be a valid IP address. However, in the case of IPv6 addresses, it was
6479 not checking this. This is a hostage to fortune. Exim now panics and dies
6480 if the condition is not met. A case was found where this could be provoked
6481 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6482 components; fortuitously, this particular loophole had already been fixed
6483 by change 4.50/55 or 4.44/33 above.
6485 If there are any other similar loopholes, the new check in host_aton()
6486 itself should stop them being exploited. The report I received stated that
6487 data on the command line could provoke the exploit when Exim was running as
6488 exim, but did not say which command line option was involved. All I could
6489 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6490 running as the user.
6493 35. There was a buffer overflow vulnerability in the SPA authentication code
6494 (which came originally from the Samba project). I have added a test to the
6495 spa_base64_to_bits() function which I hope fixes it.
6498 36. The daemon start-up calls getloadavg() while still root for those OS that
6499 need the first call to be done as root, but it missed one case: when
6500 deliver_queue_load_max is set with deliver_drop_privilege. This is
6501 necessary for the benefit of the queue runner, because there is no re-exec
6502 when deliver_drop_privilege is set.
6504 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6505 was in use, was not putting the data itself into the right store pool;
6506 consequently, it could be overwritten for a subsequent message in the same
6507 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6510 38. Sometimes the final signoff response after QUIT could fail to get
6511 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6512 < 0 before doing a fflush(). This bug looks as though it goes back to the
6513 introduction of TLS in release 3.20, but "sometimes" must have been rare
6514 because the tests only now provoked it.
6520 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6521 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6522 Now it returns an integer. A number of calls were still expecting a T/F
6523 return. Fortuitously, in all cases, the tests worked in OK situations,
6524 which is the norm. However, things would have gone wrong on any write
6525 failures on the smtp file descriptor. This function is used when sending
6526 messages over SMTP and also when doing verify callouts.
6528 2. When Exim is called to do synchronous delivery of a locally submitted
6529 message (the -odf or -odi options), it no longer closes stderr before doing
6532 3. Implemented the mua_wrapper option.
6534 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6536 5. Implemented the functions header_remove(), header_testname(),
6537 header_add_at_position(), and receive_remove_recipient(), and exported them
6540 6. If an ACL "warn" statement specified the addition of headers, Exim already
6541 inserted X-ACL-Warn: at the start if there was no header name. However, it
6542 was not making this test for the second and subsequent header lines if
6543 there were newlines in the string. This meant that an invalid header could
6544 be inserted if Exim was badly configured.
6546 7. Allow an ACL "warn" statement to add header lines at the start or after all
6547 the Received: headers, as well as at the end.
6549 8. Added the rcpt_4xx retry error code.
6551 9. Added postmaster_mailfrom=xxx to callout verification option.
6553 10. Added mailfrom=xxxx to the callout verification option, for verify=
6556 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6557 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6559 12. Inserted some casts to stop certain compilers warning when using pointer
6560 differences as field lengths or precisions in printf-type calls (mostly
6561 affecting debugging statements).
6563 13. Added optional readline() support for -be (dynamically loaded).
6565 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6566 same clock tick as a message's arrival, so that its received time was the
6567 same as the "first fail" time on the retry record, and that message
6568 remained on the queue past the ultimate address timeout, every queue runner
6569 would try a delivery (because it was past the ultimate address timeout) but
6570 after another failure, the ultimate address timeout, which should have then
6571 bounced the address, did not kick in. This was a "< instead of <=" error;
6572 in most cases the first failure would have been in the next clock tick
6573 after the received time, and all would be well.
6575 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6576 being recognized when the domain list was tested by the match_domain
6577 condition in an expansion string.
6579 16. Added the ${str2b64: operator.
6581 17. Exim was always calling setrlimit() to set a large limit for the number of
6582 processes, without checking whether the existing limit was already
6583 adequate. (It did check for the limit on file descriptors.) Furthermore,
6584 errors from getrlimit() and setrlimit() were being ignored. Now they are
6585 logged to the main and panic logs, but Exim does carry on, to try to do its
6586 job under whatever limits there are.
6588 18. Imported PCRE 5.0.
6590 19. Trivial typo in log message " temporarily refused connection" (the leading
6593 20. If the log selector return_path_on_delivery was set and an address was
6594 redirected to /dev/null, the delivery process crashed because it assumed
6595 that a return path would always be set for a "successful" delivery. In this
6596 case, the whole delivery is bypassed as an optimization, and therefore no
6599 21. Internal re-arrangement: the function for sending a challenge and reading
6600 a response while authentication was assuming a zero-terminated challenge
6601 string. It's now changed to take a pointer and a length, to allow for
6602 binary data in such strings.
6604 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6606 23. Exim was not respecting finduser_retries when seeking the login of the
6607 uid under which it was called; it was always trying 10 times. (The default
6608 setting of finduser_retries is zero.) Also, it was sleeping after the final
6609 failure, which is pointless.
6611 24. Implemented tls_on_connect_ports.
6613 25. Implemented acl_smtp_predata.
6615 26. If the domain in control=submission is set empty, Exim assumes that the
6616 authenticated id is a complete email address when it generates From: or
6617 Sender: header lines.
6619 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6620 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6621 chown and chgrp in /bin and hostname in /usr/bin.
6623 28. Exim was keeping the "process log" file open after each use, just as it
6624 does for the main log. This opens the possibility of it remaining open for
6625 long periods when the USR1 signal hits a daemon. Occasional processlog
6626 errors were reported, that could have been caused by this. Anyway, it seems
6627 much more sensible not to leave this file open at all, so that is what now
6630 29. The long-running daemon process does not normally write to the log once it
6631 has entered its main loop, and it closes the log before doing so. This is
6632 so that log files can straightforwardly be renamed and moved. However,
6633 there are a couple of unusual error situations where the daemon does write
6634 log entries, and I had neglected to close the log afterwards.
6636 30. The text of an SMTP error response that was received during a remote
6637 delivery was being truncated at 512 bytes. This is too short for some of
6638 the long messages that one sometimes sees. I've increased the limit to
6641 31. It is now possible to make retry rules that apply only when a message has a
6642 specific sender, in particular, an empty sender.
6644 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6645 it possible to be selective about when SMTP synchronization is enforced.
6647 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6649 32. Implemented hosts_connection_nolog.
6651 33. Added an ACL for QUIT.
6653 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6656 35. Added mailbox_size and mailbox_filecount to appendfile.
6658 36. Added control = no_multiline_responses to ACLs.
6660 37. There was a bug in the logic of the code that waits for the clock to tick
6661 in the case where the clock went backwards by a substantial amount such
6662 that the microsecond fraction of "now" was more than the microsecond
6663 fraction of "then" (but the whole seconds number was less).
6665 38. Added support for the libradius Radius client library this is found on
6666 FreeBSD (previously only the radiusclient library was supported).
6672 1. When certain lookups returned multiple values in the form name=value, the
6673 quoting of the values was not always being done properly. Specifically:
6674 (a) If the value started with a double quote, but contained no whitespace,
6676 (b) If the value contained whitespace other than a space character (i.e.
6677 tabs or newlines or carriage returns) it was not quoted.
6678 This fix has been applied to the mysql and pgsql lookups by writing a
6679 separate quoting function and calling it from the lookup code. The fix
6680 should probably also be applied to nisplus, ibase and oracle lookups, but
6681 since I cannot test any of those, I have not disturbed their existing code.
6683 2. A hit in the callout cache for a specific address caused a log line with no
6684 reason for rejecting RCPT. Now it says "Previous (cached) callout
6685 verification failure".
6687 3. There was an off-by-one bug in the queryprogram router. An over-long
6688 return line was truncated at 256 instead of 255 characters, thereby
6689 overflowing its buffer with the terminating zero. As well as fixing this, I
6690 have increased the buffer size to 1024 (and made a note to document this).
6692 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6693 when Exim is waiting for an SMTP response from a remote server, Exim
6694 restarts its select() call on the socket, thereby resetting its timeout.
6695 This is not a problem when such interrupts are rare. Somebody set up a cron
6696 job to run exiwhat every 2 minutes, which is less than the normal select()
6697 timeout (5 or 10 minutes). This meant that the select() timeout never
6698 kicked in because it was always reset. I have fixed this by comparing the
6699 time when an interrupt arrives with the time at the start of the first call
6700 to select(). If more time than the timeout has elapsed, the interrupt is
6701 treated as a timeout.
6703 5. Some internal re-factoring in preparation for the addition of Sieve
6704 extensions (by MH). In particular, the "personal" test is moved to a
6705 separate function, and given an option for scanning Cc: and Bcc: (which is
6706 not set for Exim filters).
6708 6. When Exim created an email address using the login of the caller as the
6709 local part (e.g. when creating a From: or Sender: header line), it was not
6710 quoting the local part when it contained special characters such as @.
6712 7. Installed new OpenBSD configuration files.
6714 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6715 try to make them clearer.
6717 9. Callout options, other than the timeout value, were being ignored when
6718 verifying sender addresses in header lines. For example, when using
6720 verify = header_sender/callout=no_cache
6722 the cache was (incorrectly) being used.
6724 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6725 only the Cygwin environment.
6727 11. When return_path_on_delivery was set as a log selector, if different remote
6728 addresses in the same message used different return paths and parallel
6729 remote delivery occurred, the wrong values would sometimes be logged.
6730 (Whenever a remote delivery process finished, the return path value from
6731 the most recently started remote delivery process was logged.)
6733 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6734 lines when AUTH and/or TLS are in use. This is the "received protocol"
6735 field. Exim used to use "asmtp" for authenticated SMTP, without any
6736 indication (in the protocol name) for TLS use. Now it follows the RFC and
6737 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6738 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6739 names appear in log lines as well as in Received: header lines.
6741 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6742 extensions, and comparison tests, and to fix some bugs.
6744 14. Changes to the "personal" filter test:
6746 (1) The test was buggy in that it was just doing the equivalent of
6747 "contains" tests on header lines. For example, if a user's address was
6748 anne@some.where, the "personal" test would incorrectly be true for
6750 To: susanne@some.where
6752 This test is now done by extracting each address from the header in turn,
6753 and checking the entire address. Other tests that are part of "personal"
6754 are now done using regular expressions (for example, to check local parts
6755 of addresses in From: header lines).
6757 (2) The list of non-personal local parts in From: addresses has been
6758 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6759 taken from the Sieve specification recommendations.
6761 (3) If the message contains any header line starting with "List-" it is
6762 treated as non-personal.
6764 (4) The test for "circular" in the Subject: header line has been removed
6765 because it now seems ill-conceived.
6767 15. Minor typos in src/EDITME comments corrected.
6769 16. Installed latest exipick from John Jetmore.
6771 17. If headers_add on a router specified a text string that was too long for
6772 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6773 of string_sprintf() is now avoided.
6775 18. $message_body_size was not set (it was always zero) when running the DATA
6776 ACL and the local_scan() function.
6778 19. For the "mail" command in an Exim filter, no default was being set for
6779 the once_repeat time, causing a random time value to be used if "once" was
6780 specified. (If the value happened to be <= 0, no repeat happened.) The
6781 default is now 0s, meaning "never repeat". The "vacation" command was OK
6782 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6783 (I found it when inspecting the code).
6785 20. There is now an overall timeout for performing a callout verification. It
6786 defaults to 4 times the callout timeout, which applies to individual SMTP
6787 commands during the callout. The overall timeout applies when there is more
6788 than one host that can be tried. The timeout is checked before trying the
6789 next host. This prevents very long delays if there are a large number of
6790 hosts and all are timing out (e.g. when the network connections are timing
6791 out). The value of the overall timeout can be changed by specifying an
6792 additional sub-option for "callout", called "maxwait". For example:
6794 verify = sender/callout=5s,maxwait=20s
6796 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6797 to the end before writing, but this should make it even safer).
6799 22. Exim was forgetting that it had advertised PIPELINING for the second and
6800 subsequent messages on an SMTP connection. It was also not resetting its
6801 memory on STARTTLS and an internal HELO.
6803 23. When Exim logs an SMTP synchronization error within a session, it now
6804 records whether PIPELINING has been advertised or not.
6806 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6807 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6808 rather than long int.
6810 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6812 26. Added the never_mail option to autoreply.
6818 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6819 crash if the getsockname() call failed; this can happen if a connection is
6820 closed very soon after it is established. The problem was simply in the
6821 order in which certain operations were done, causing Exim to try to write
6822 to the SMTP stream before it had set up the file descriptor. The bug has
6823 been fixed by making things happen in the correct order.
6829 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6830 before the connection was closed, thus losing the rejection response.
6832 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6833 some early Solaris releases, but causes trouble in current releases where
6834 socklen_t is defined.
6836 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6839 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6842 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6843 that releasing the top of it at the end releases what was used for sub-
6844 expansions (unless the block got too big). However, discard this block if
6845 the first thing is a variable or header, so that we can use its block when
6846 it is dynamic (useful for very large $message_headers, for example).
6848 6. Lookups now cache *every* query, not just the most recent. A new, separate
6849 store pool is used for this. It can be recovered when all lookup caches are
6850 flushed. Lookups now release memory at the end of their result strings.
6851 This has involved some general refactoring of the lookup sources.
6853 7. Some code has been added to the store_xxx() functions to reduce the amount
6854 of flapping under certain conditions.
6856 8. log_incoming_interface used to affect only the <= reception log lines. Now
6857 it causes the local interface and port to be added to several more SMTP log
6858 lines, for example "SMTP connection from", and rejection lines.
6860 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6862 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6864 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6865 could be overwritten at the end of the current message (or the start of a
6866 new message if it was set in a HELO ACL). The value is now preserved for
6867 the duration of the SMTP connection.
6869 12. If a transport had a headers_rewrite setting, and a matching header line
6870 contained an unqualified address, that address was qualified, even if it
6871 did not match any rewriting rules. The underlying bug was that the values
6872 of the flags that permit the existence of unqualified sender and recipient
6873 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6874 non-local messages, and by -bnq for local messages) were not being
6875 preserved with the message after it was received.
6877 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6878 "next input=" as part of the text comprising the host identity instead of
6879 the correct text. The code was using the same buffer for two different
6880 strings. However, depending on which order the printing function evaluated
6881 its arguments, the bug did not always show up. Under Linux, for example, my
6882 test suite worked just fine.
6884 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6885 doesn't work with some older versions of Perl. It has been changed to "my",
6886 which in any case is probably the better facility to use.
6888 15. A really picky compiler found some instances of statements for creating
6889 error messages that either had too many or two few arguments for the format
6892 16. The size of the buffer for calls to the DNS resolver has been increased
6893 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6894 for addresses that have a lot of PTR records. This alleviates a problem; it
6895 does not fully solve it.
6897 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6898 buffer now truncates the list and logs the incident, which is the same
6899 action as happens when Exim is looking up a host name and its aliases.
6900 Previously in this situation something unpredictable would happen;
6901 sometimes it was "internal error: store_reset failed".
6903 18. If a server dropped the connection unexpectedly when an Exim client was
6904 using GnuTLS and trying to read a response, the client delivery process
6905 crashed while trying to generate an error log message.
6907 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6908 string, for example:
6910 warn message = H1: something\nH2: something
6912 the text was added as a single header line from Exim's point of view
6913 though it ended up OK in the delivered message. However, searching for the
6914 second and subsequent header lines using $h_h2: did not work. This has been
6915 fixed. Similarly, if a system filter added multiple headers in this way,
6916 the routers could not see them.
6918 20. Expanded the error message when iplsearch is called with an invalid key to
6919 suggest using net-iplsearch in a host list.
6921 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6922 ACLs are no longer actually imposed (and a message to that effect is
6925 22. If a "gecos" field in a passwd entry contained escaped characters, in
6926 particular, if it contained a \" sequence, Exim got it wrong when building
6927 a From: or a Sender: header from that name. A second bug also caused
6928 incorrect handling when an unquoted " was present following a character
6929 that needed quoting.
6931 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6932 was not being matched caselessly.
6934 24. Arranged for all hyphens in the exim.8 source to be escaped with
6937 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6938 properly. Recipient callout cache records were still being keyed to include
6939 the sender, even when use_sender was set false. This led to far more
6940 callouts that were necessary. The sender is no longer included in the key
6941 when use_sender is false.
6943 26. Added "control = submission" modifier to ACLs.
6945 27. Added the ${base62d: operator to decode base 62 numbers.
6947 28. dnsdb lookups can now access SRV records.
6949 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6950 the configuration file.
6952 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6953 (-v) mode. This makes the output for a verbose queue run more intelligible.
6955 31. Added a use_postmaster feature to recipient callouts.
6957 32. Added the $body_zerocount variable, containing the number of binary zero
6958 bytes in the message body.
6960 33. The time of last modification of the "new" subdirectory is now used as the
6961 "mailbox time last read" when there is a quota error for a maildir
6964 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6966 35. Added +ignore_unknown as a special item in host lists.
6968 36. Code for decoding IPv6 addresses in host lists is now included, even if
6969 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6970 address was recognized as an IP address, but was then not correctly decoded
6971 into binary, causing unexpected and incorrect effects when compared with
6978 1. Very minor rewording of debugging text in manualroute to say "list of
6979 hosts" instead of "hostlist".
6981 2. If verify=header_syntax was set, and a header line with an unqualified
6982 address (no domain) and a large number of spaces between the end of the
6983 name and the colon was received, the reception process suffered a buffer
6984 overflow, and (when I tested it) crashed. This was caused by some obsolete
6985 code that should have been removed. The fix is to remove it!
6987 3. When running in the test harness, delay a bit after writing a bounce
6988 message to get a bit more predictability in the log output.
6990 4. Added a call to search_tidyup() just before forking a reception process. In
6991 theory, someone could use a lookup in the expansion of smtp_accept_max_
6992 per_host which, without the tidyup, could leave open a database connection.
6994 5. Added the variables $recipient_data and $sender_data which get set from a
6995 lookup success in an ACL "recipients" or "senders" condition, or a router
6996 "senders" option, similar to $domain_data and $local_part_data.
6998 6. Moved the writing of debug_print from before to after the "senders" test
7001 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7002 problems for message scanning, either using a data ACL, or using
7003 local_scan() because the Received: header was not generated till after they
7004 were called (in order to set the time as the time of reception completion).
7005 I have revised the way this works. The header is now generated after the
7006 body is received, but before the ACL or local_scan() are called. After they
7007 are run, the timestamp in the header is updated.
7013 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7014 before starting a queue runner without re-exec. This happened only when
7015 deliver_drop_privilege was set or when the Exim user was set to root. The
7016 effect of the bug was that timeouts during subsequent deliveries caused
7017 crashes instead of being properly handled. The handler is now left at its
7018 default (and expected) setting.
7020 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7021 message, again when deliver_drop_privilege is set or Exim is run as root.
7022 The bug described in (1) was not present in this case, but the tidying up
7023 of the other signals was missing. I have made the two cases consistent.
7025 3. The ignore_target_hosts setting on a manualroute router was being ignored
7026 for hosts that were looked up using the /MX notation.
7028 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7031 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7032 operated on the sender address. After changing the $sender_address to <>
7033 for the sender address verify, Exim was re-instated it as the original
7034 (before rewriting) address, but remembering that it had rewritten it, so it
7035 wasn't rewriting it again. This bug also had the effect of breaking the
7036 sender address verification caching when the sender address was rewritten.
7038 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7039 This has been changed so that if the ip literal address matches
7040 ignore_target_hosts, the router declines.
7042 7. Added expansion conditions match_domain, match_address, and match_local_
7043 part (NOT match_host).
7045 8. The placeholder for the Received: header didn't have a length field set.
7047 9. Added code to Exim itself and to exim_lock to test for a specific race
7048 condition that could lead to file corruption when using MBX delivery. The
7049 issue is with the lockfile that is created in /tmp. If this file is removed
7050 after a process has opened it but before that process has acquired a lock,
7051 there is the potential for a second process to recreate the file and also
7052 acquire a lock. This could lead to two Exim processes writing to the file
7053 at the same time. The added code performs the same test as UW imapd; it
7054 checks after acquiring the lock that its file descriptor still refers to
7055 the same named file.
7057 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7058 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7061 11. Added the smtp_active_hostname option. If used, this will typically be made
7062 to depend on the incoming interface address. Because $interface_address is
7063 not set up until the daemon has forked a reception process, error responses
7064 that can happen earlier (such as "too many connections") no longer contain
7067 12. If an expansion in a condition on a "warn" statement fails because a lookup
7068 defers, the "warn" statement is abandoned, and the next ACL statement is
7069 processed. Previously this caused the whole ACL to be aborted.
7071 13. Added the iplsearch lookup type.
7073 14. Added ident_timeout as a log selector.
7075 15. Added tls_certificate_verified as a log selector.
7077 16. Added a global option tls_require_ciphers (compare the smtp transport
7078 option of the same name). This controls incoming TLS connections.
7080 17. I finally figured out how to make tls_require_ciphers do a similar thing
7081 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7082 before starting the TLS session.
7084 18. Tabs are now shown as \t in -bP output.
7086 19. If the log selector return_path_on_delivery was set, Exim crashed when
7087 bouncing a message because it had too many Received: header lines.
7089 20. If two routers both had headers_remove settings, and the first one included
7090 a superfluous trailing colon, the final name in the first list and the
7091 first name in the second list were incorrectly joined into one item (with a
7092 colon in the middle).
7098 1. Added -C and -D options to the exinext utility, mainly to make it easier
7099 to include in the automated testing, but these could be helpful when
7100 multiple configurations are in use.
7102 2. The exinext utility was not formatting the output nicely when there was
7103 an alternate port involved in the retry record key, nor when there was a
7104 message id as well (for retries that were specific to a specific message
7105 and a specific host). It was also confused by IPv6 addresses, because of
7106 the additional colons they contain. I have fixed the IPv4 problem, and
7107 patched it up to do a reasonable job for IPv6.
7109 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7110 delivery, the log line now contains "pipelined" if PIPELINING was used.
7112 4. An SMTP transport process used to panic and die if the bind() call to set
7113 an explicit outgoing interface failed. This has been changed; it is now
7114 treated in the same way as a connect() failure.
7116 5. A reference to $sender_host_name in the part of a conditional expansion
7117 that was being skipped was still causing a DNS lookup. This no longer
7120 6. The def: expansion condition was not recognizing references to header lines
7121 that used bh_ and bheader_.
7123 7. Added the _cache feature to named lists.
7125 8. The code for checking quota_filecount in the appendfile transport was
7126 allowing one more file than it should have been.
7128 9. For compatibility with Sendmail, the command line option
7136 and sets the incoming protocol and host name (for trusted callers). The
7137 host name and its colon can be omitted when only the protocol is to be set.
7138 Note the Exim already has two private options, -pd and -ps, that refer to
7139 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7140 "s", but I don't think that's a major issue.
7142 10. A number of refactoring changes to the code, none of which should affect
7145 (a) The number of logging options was getting close to filling up the
7146 32-bit word that was used as a bit map. I have split them into two classes:
7147 those that are passed in the argument to log_write(), and those that are
7148 only ever tested independently outside of that function. These are now in
7149 separate 32-bit words, so there is plenty of room for expansion again.
7150 There is no change in the user interface or the logging behaviour.
7152 (b) When building, for example, log lines, the code previously used a
7153 macro that called string_cat() twice, in order to add two strings. This is
7154 not really sufficiently general. Furthermore, there was one instance where
7155 it was actually wrong because one of the argument was used twice, and in
7156 one call a function was used. (As it happened, calling the function twice
7157 did not affect the overall behaviour.) The macro has been replaced by a
7158 function that can join an arbitrary number of extra strings onto a growing
7161 (c) The code for expansion conditions now uses a table and a binary chop
7162 instead of a serial search (which was left over from when there were very
7163 few conditions). Also, it now recognizes conditions like "pam" even when
7164 the relevant support is not compiled in: a suitably worded error message is
7165 given if an attempt is made to use such a condition.
7167 11. Added ${time_interval:xxxxx}.
7169 12. A bug was causing one of the ddress fields not to be passed back correctly
7170 from remote delivery subprocesses. The field in question was not being
7171 subsequently used, so this caused to problems in practice.
7173 13. Added new log selectors queue_time and deliver_time.
7175 14. Might have fixed a bug in maildirsizefile handling that threw up
7176 "unexpected character" debug warnings, and recalculated the data
7177 unnecessarily. In any case, I expanded the warning message to give more
7180 15. Added the message "Restricted characters in address" to the statements in
7181 the default ACL that block characters like @ and % in local parts.
7183 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7184 Three changes have been made:
7186 (a) There was a serious bug; a negative response to MAIL caused the whole
7187 recipient domain to be cached as invalid, thereby blocking all messages
7188 to all local parts at the same domain, from all senders. This bug has
7189 been fixed. The domain is no longer cached after a negative response to
7190 MAIL if the sender used is not empty.
7192 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7195 (c) A new callout option, "use_sender" has been added for people who want
7196 the modified behaviour.
7202 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7205 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7206 indeed breaks things for older releases.
7208 3. Added additional logging to the case where there is a problem reading data
7209 from a filter that is running in a subprocess using a pipe, in order to
7210 try to track down a specific problem.
7212 4. Testing facility fudge: when running in the test harness and attempting
7213 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7214 getting "No route to host". Convert this to a timeout.
7216 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7219 6. Some OS don't have socklen_t but use size_t instead. This affects the
7220 fifth argument of getsockopt() amongst other things. This is now
7221 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7222 can be set for individual OS. I have set it for SunOS5, OSF1, and
7223 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7224 some earlier ones do not.
7226 7. Change 4.30/15 was not doing the test caselessly.
7228 8. The standard form for an IPv6 address literal was being rejected by address
7229 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7230 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7231 this, as well as the form without the "IPv6" on the front (but only when
7232 address literals are enabled, of course).
7234 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7236 10. Exim crashed if a message with an empty sender address specified by -f
7237 encountered a router with an errors_to setting. This could be provoked only
7238 by a command such as
7242 where an empty string was supplied; "<>" did not hit this bug.
7244 11. Installed PCRE release 4.5.
7246 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7247 remained set. It is now erased.
7249 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7250 times from message ids (which are base 36 rather than the normal 62).
7252 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7253 were being counted as actual protocol errors, and logged if the log
7254 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7255 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7256 and DATA following a set of rejected RCPTs do not count as protocol errors.
7257 In other words, Exim assumes they were pipelined, though this may not
7258 actually be the case. Of course, in all cases the client gets an
7259 appropriate error code.
7261 15. If a lookup fails in an ACL condition, a message about the failure may
7262 be available; it is used if testing the ACL cannot continue, because most
7263 such messages specify what the cause of the deferral is. However, some
7264 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7265 that caused an old message to be retained and used if a later statement
7266 caused a defer, replacing the real cause of the deferral.
7268 16. If an IP address had so many PTR records that the DNS lookup buffer
7269 was not large enough to hold them, Exim could crash while trying to process
7270 the truncated data. It now detects and logs this case.
7272 17. Further to 4.21/58, another change has been made: if (and only if) the
7273 first line of a message (the first header line) ends with CRLF, a bare LF
7274 in a subsequent header line has a space inserted after it, so as not to
7275 terminate the header.
7277 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7278 unnecessarily, used atoi() instead of strtol(), and didn't check the
7279 termination when getting file sizes from file names by regex.
7281 19. Completely re-implemented the support for maildirsize files, in the light
7282 of a number of problems with the previous contributed implementation
7283 (4.30/29). In particular:
7285 . If the quota is zero, the maildirsize file is maintained, but no quota is
7288 . If the maildir directory does not exist, it is created before any attempt
7289 to write a maildirsize file.
7291 . The quota value in the file is just a cache; if the quota is changed in
7292 the transport, the new value overrides.
7294 . A regular expression is available for excluding directories from the
7297 20. The autoreply transport checks the characters in options that define the
7298 message's headers; it allows continued headers, but it was checking with
7299 isspace() after an embedded newline instead of explicitly looking for a
7302 21. If all the "regular" hosts to which an address was routed had passed their
7303 expiry times, and had not reached their retry times, the address was
7304 bounced, even if fallback hosts were defined. Now Exim should go on to try
7307 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7308 equivalent code in the SMTP transport. Some hosts send humungous responses
7309 to HELO/EHLO, more than 1024 it seems.
7311 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7312 gives compiler warnings in some environments. I've now done it "properly",
7315 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7316 (because of problems with the built-in one) was declared to return uschar *
7317 instead of char *, causing compiler failure.
7319 25. Fixed a file descriptor leak when processing alias/forward files.
7321 26. Fixed a minor format string issue in dbfn.c.
7323 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7325 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7326 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7329 29. When a sender address is verified, it is cached, to save repeating the test
7330 when there is more than one recipient in a message. However, when the
7331 verification involves a callout, it is possible for different callout
7332 options to be set for different recipients. It is too complicated to keep
7333 track of this in the cache, so now Exim always runs a verification when a
7334 callout is required, relying on the callout cache for the optimization.
7335 The overhead is duplication of the address routing, but this should not be
7338 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7339 to be verified with callout=postmaster, and the main callout worked but the
7340 postmaster check failed, the verification correctly failed. However, if a
7341 subsequent RCPT command asked for sender verification *without* the
7342 postmaster check, incorrect caching caused this verification also to fail,
7345 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7346 it was not caching the DNS options (qualify_single, search_parents) that
7347 were used when the lookup failed. A subsequent lookup with different
7348 options therefore always gave the same answer, though there were cases
7349 where it should not have. (Example: a "domains = !$mx_any" option on a
7350 dnslookup router: the "domains" option is always processed without any
7351 widening, but the router might have qualify_single set.) Now Exim uses the
7352 cached value only when the same options are set.
7354 32. Added John Jetmore's "exipick" utility to the distribution.
7356 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7357 than a timeout (e.g. a certificate is required, and is not provided), an
7358 Exim server now closes the connection immediately. Previously it waited for
7359 the client to close - but if the client is SSL, it seems that they each
7360 wait for each other, leading to a delay before one of them times out.
7362 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7363 maintained 0.8.x compatibility because I don't think many are using it, and
7364 it is clearly obsolete.
7366 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7369 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7370 client certificate was expired. A simple patch fixes this, though I don't
7371 understand the full logic of why the verify callback is called multiple
7374 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7375 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7376 which causes problems with some clients (such as the Certicom SSL Plus
7377 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7378 disables the coutermeasure allowing Eudora to connect."
7380 38. Exim was not checking that a write() to a log file succeeded. This could
7381 lead to Bad Things if a log got too big, in particular if it hit a file
7382 size limit. Exim now panics and dies if it cannot write to a log file, just
7383 as it does if it cannot open a log file.
7385 39. Modified OS/Makefile-Linux so that it now contains
7387 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7389 The two -D definitions ensure that Exim is compiled with large file
7390 support, which makes it possible to handle log files that are bigger than
7393 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7394 instance) a domain was checked against a named list that involved a lookup,
7395 causing $domain_data to be set, then another domain was checked against the
7396 same list, then the first domain was re-checked, the value of $domain_data
7397 after the final check could be wrong. In particular, if the second check
7398 failed, it could be set empty. This bug probably also applied to
7401 41. The strip_trailing_dot option was not being applied to the address given
7402 with the -f command-line option.
7404 42. The code for reading a message's header from the spool was incrementing
7405 $received_count, but never initializing it. This meant that the value was
7406 incorrect (doubled) while delivering a message in the same process in which
7407 it was received. In the most common configuration of Exim, this never
7408 happens - a fresh exec is done - but it can happen when
7409 deliver_drop_privilege is set.
7411 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7412 it now includes up to 150 characters of the unexpected data in the log
7415 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7416 and building data strings. The size of both of these buffers was 10 000
7417 bytes - far larger than anybody would *ever* want, thought I. Needless to
7418 say, somebody hit the limit. I have increased the maximum line length to
7419 20 000 and the maximum data length of concatenated lines to 100 000. I have
7420 also fixed two bugs, because there was no checking on these buffers. Tsk,
7421 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7422 buffer is too small.
7424 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7425 lsearch lookups. Now it does.
7427 46. When parsing a route_list item in a manualroute router, a fixed-length
7428 buffer was used for the list of hosts. I made this 1024 bytes long,
7429 thinking that nobody would ever have a list of hosts that long. Wrong.
7430 Somebody had a whole pile of complicated expansion conditions, and the
7431 string was silently truncated, leading to an expansion error. It turns out
7432 that it is easier to change to an unlimited length (owing to other changes
7433 that have happened since this code was originally written) than to build
7434 structure for giving a limitation error. The length of the item that
7435 expands into the list of hosts is now unlimited.
7437 47. The lsearch lookup could not handle data where the length of text line was
7438 more than 4095 characters. Such lines were truncated, leading to shortened
7439 data being returned. It should now handle lines of any length.
7441 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7442 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7445 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7446 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7447 no longer happens in comment lines. A list of replacements is now placed
7448 at the head of all of the source files, except those whose only change is
7449 to replace PERL_COMMAND in the very first #! line.
7451 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7452 messages on the queue, with a bottom-up merge sort, using code contributed
7453 by Michael Haardt. This should make operations like -bp somewhat faster on
7454 large queues. It won't affect queue runners, except when queue_run_in_order
7457 51. Installed eximstats 1.31 in the distribution.
7459 52. Added support for SRV lookups to the dnslookup router.
7461 53. If an ACL referred to $message_body or $message_body_end, the value was not
7462 reset for any messages that followed in the same SMTP session.
7464 54. The store-handling optimization for building very long strings was not
7465 differentiating between the different store pools. I don't think this
7466 actually made any difference in practice, but I've tidied it.
7468 55. While running the routers to verify a sender address, $sender_address
7469 was still set to the sender address. This is wrong, because when routing to
7470 send a bounce to the sender, it would be empty. Therefore, I have changed
7471 it so that, while verifying a sender address, $sender_address is set to <>.
7472 (There is no change to what happens when verifying a recipient address.)
7474 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7475 target A or AAAA records (if not already returned) without resetting the
7476 qualify_single or search_parents options of the DNS resolver. These are
7477 inappropriate in this case because the targets of MX and SRV records must
7478 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7479 target that, when qualified, matched something in the local domain. These
7480 two options are now turned off when doing these lookups.
7482 57. It seems that at least some releases of Reiserfs (which does not have the
7483 concept of a fixed number of inodes) returns zero and not -1 for the
7484 number of available inodes. This interacted badly with check_spool_inodes,
7485 which assumed that -1 was the "no such thing" setting. What I have done is
7486 to check that the total number of inodes is greater than zero before doing
7487 the test of how many are available.
7489 58. When a "warn" ACL statement has a log_message modifier, the message is
7490 remembered, and not repeated. This is to avoid a lot of repetition when a
7491 message has many recipients that cause the same warning to be written.
7492 However, Exim was preserving the list of already written lines for an
7493 entire SMTP session, which doesn't seem right. The memory is now reset if a
7494 new message is started.
7496 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7497 debugging output unless log_rewrite was also set.
7499 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7500 of (int)(handle) when we know that handle contains (void *)(-1).
7502 61. The Exim daemon panic-logs an error return when it closes the incoming
7503 connection. However "connection reset by peer" seems to be common, and
7504 isn't really an error worthy of noting specially, so that particular error
7507 62. When Exim is trying to find all the local interfaces, it used to panic and
7508 die if the ioctl to get the interface flags failed. However, it seems that
7509 on at least one OS (Solaris 9) it is possible to have an interface that is
7510 included in the list of interfaces, but for which you get a failure error
7511 for this call. This happens when the interface is not "plumbed" into a
7512 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7513 failure of the "get flags" call assumes that the interface is down.
7515 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7516 makes life easier for people who are doing arithmetic on fields extracted
7517 from dates, where you often get leading zeros that should not be
7518 interpreted as octal.
7520 64. Added qualify_domain to the redirect router, to override the global
7523 65. If a pathologically long header line contained very many addresses (the
7524 report of this problem mentioned 10 000) and each of them was rewritten,
7525 Exim could use up a very large amount of memory. (It kept on making new
7526 copies of the header line as it rewrote, and never released the old ones.)
7527 At the expense of a bit more processing, the header rewriting function has
7528 been changed so that it no longer eats memory in this way.
7530 66. The generation of the Received: header has been moved from the time that a
7531 message starts to be received, to the time that it finishes. The timestamp
7532 in the Received: header should now be very close to that of the <= log
7533 line. There are two side-effects of this change:
7535 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7536 logged header lines no longer include the local Received: line, because
7537 it has not yet been created. The same applies to a copy of the message
7538 that is returned to a non-SMTP sender when a message is rejected.
7540 (b) When a filter file is tested using -bf, no additional Received: header
7541 is added to the test message. After some thought, I decided that this
7544 This change does not affect the value of $received_for. It is still set
7545 after address rewriting, but before local_scan() is called.
7547 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7549 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7550 gave an unhelpful panic error message, and a defer error. I have managed to
7551 change this behaviour so that it now rejects any supplied certificate,
7552 which seems right, as the list of acceptable certificates is empty.
7554 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7555 gave an unhelpful defer error. I have not managed to make this reject any
7556 supplied certificates, but the error message it gives is "no certificate
7557 supplied", which is not helpful.
7559 70. exigrep's output now also includes lines that are not associated with any
7560 message, but which match the given pattern. Implemented by a patch from
7561 Martin Sluka, which also tidied up the Perl a bit.
7563 71. Recipient callout verification, like sender verification, was using <> in
7564 the MAIL FROM command. This isn't really the right thing, since the actual
7565 sender may affect whether the remote host accepts the recipient or not. I
7566 have changed it to use the actual sender in the callout; this means that
7567 the cache record is now keyed on a recipient/sender pair, not just the
7568 recipient address. There doesn't seem to be a real danger of callout loops,
7569 since a callout by the remote host to check the sender would use <>.
7570 [SEE ABOVE: changed after hitting problems.]
7572 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7573 temporary errors. However, in the case of such a code being given after
7574 the end of a data transmission (i.e. after ".") Exim was failing to write
7575 a retry record for the message. (Yes, there was some broken host that was
7576 actually sending 8xx at this point.)
7578 73. An unknown lookup type in a host list could cause Exim to panic-die when
7579 the list was checked. (An example that provoked this was putting <; in the
7580 middle of a list instead of at the start.) If this happened during a DATA
7581 ACL check, a -D file could be left lying around. This kind of configuration
7582 error no longer causes Exim to die; instead it causes a defer error. The
7583 incident is still logged to the main and panic logs.
7585 74. Buglet left over from Exim 3 conversion. The message "too many messages
7586 in one connection" was written to the rejectlog but not the mainlog, except
7587 when address rewriting (yes!) was being logged.
7589 75. Added write_rejectlog option.
7591 76. When a system filter was run not as root (that is, when system_filter_user
7592 was set), the values of the $n variables were not being returned to the
7593 main process; thus, they were not subsequently available in the $sn
7596 77. Added +return_path_on_delivery log selector.
7598 78. A connection timeout was being treated differently from recipients deferred
7599 when testing hosts_max_try with a message that was older than the host's
7600 retry timeout. (The host should not be counted, thus allowing all hosts to
7601 be tried at least once before bouncing.) This may have been the cause of an
7602 occasionally reported bug whereby a message would remain on the queue
7603 longer than the retry timeout, but would be bounced if a delivery was
7604 forced. I say "may" because I never totally pinned down the problem;
7605 setting up timeout/retry tests is difficult. See also the next item.
7607 79. The ultimate address timeout was not being applied to errors that involved
7608 a combination of host plus message (for example, a timeout on a MAIL
7609 command). When an address resolved to a number of possible hosts, and they
7610 were not all tried for each delivery (e.g. because of hosts_max_try), a
7611 message could remain on the queue longer than the retry timeout.
7613 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7616 81. Fixed an obscure SMTP outgoing bug which required at least the following
7617 conditions: (a) there was another message waiting for the same server;
7618 (b) the server returned 5xx to all RCPT commands in the first message so
7619 that the message was not completed; (c) the server dropped the connection
7620 or gave a negative response to the RSET that Exim sends to abort the
7621 transaction. The observed case was a dropped connection after DATA that had
7622 been sent in pipelining mode. That is, the server had advertised PIPELINING
7623 but was not implementing it correctly. The effect of the bug was incorrect
7624 behaviour, such as trying another host, and this could lead to a crash.
7630 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7631 and daemon.c were passed as pointers to ints; they should have been
7632 pointers to socklen_t variables (which are typically unsigned ints).
7634 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7637 3. Fixed a really odd bug that affected only the testing scheme; patching a
7638 certain fixed string in the binary changed the value of another string that
7639 happened to be identical to the end of the original first string.
7641 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7642 name", it returns that address as the IP address. On some operating
7643 systems (e.g. Solaris), it also passes back the IP address string as the
7644 "host name". However, on others (e.g. Linux), it passes back an empty
7645 string. Exim wasn't checking for this, and was changing the host name to an
7646 empty string, assuming it had been canonicalized.
7648 5. Although rare, it is permitted to have more than one PTR record for a given
7649 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7650 all the names associated with an address, because they do in Solaris.
7651 However, it seems that they do not in Linux for data that comes from the
7652 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7653 I found this out when I moved to a new Linux workstation and tried to run
7654 the Exim test suite.
7656 To get round this problem I have changed the code so that it now does its
7657 own call to the DNS to look up PTR records when searching for a host name.
7658 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7659 addresses that are only in /etc/hosts are still found.
7661 This behaviour is, however, controlled by an option called host_lookup_
7662 order, which defaults to "bydns:byaddr". If people want to use the other
7663 order, or indeed, just use one or the other means of lookup, they can
7664 specify it in this variable.
7666 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7667 some operating systems, this comes back from gethostbyaddr() as an empty
7668 string, and this is what Exim used to test for. However, it seems that in
7669 other systems, "." is yielded. Exim now tests for this case too.
7671 7. The values of check_spool_space and check_log_space are now held internally
7672 as a number of kilobytes instead of an absolute number of bytes. If a
7673 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7674 kilobyte. This means that much larger values can be stored.
7676 8. Exim monitor: an attempt to get the action menu when not actually pointing
7677 at a message produces an empty menu entitled "No message selected". This
7678 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7679 no entries in it ("Shell widget menu has zero width and/or height"). So I
7680 have added a single, blank menu entry in this case.
7682 9. Added ${quote_local_part.
7684 10. MIME decoding is now applied to the contents of Subject: header lines when
7687 11. Now that a reference to $sender_host_address automatically causes a reverse
7688 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7689 host lookup before query-style lookups in lists that might use this
7690 variable. This has therefore been abolished, and the "net-" prefix is no
7691 longer necessary for query-style lookups.
7693 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7694 have been a typo for LFLAGS, so it has been changed.
7696 13. The install script calls Exim with "-C /dev/null" in order to find the
7697 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7698 to be output. However, since Exim outputs its version number before the
7699 error, it didn't break the script. It just looked ugly. I fixed this by
7700 always allowing "-C /dev/null" if the caller is root.
7702 14. Ignore overlarge ACL variable number when reading spool file - insurance
7703 against a later release with more variables having written the file.
7705 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7706 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7707 this, as well as the form without the "IPv6" on the front.
7709 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7710 OS/Makefile-Darwin file.
7712 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7713 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7715 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7716 variable. However, some OS use ints for inodes. I've added cast to long int
7717 to get rid of the compiler warning.
7719 19. I had forgotten to lock out "/../" in configuration file names when
7720 ALT_CONFIG_PREFIX was set.
7722 20. Routers used for verification do not need to specify transports. However,
7723 if such a router generated a host list, and callout was configured, Exim
7724 crashed, because it could not find a port number from the (non-existent)
7725 transport. It now assumes port 25 in this circumstance.
7727 21. Added the -t option to exigrep.
7729 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7730 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7731 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7732 them is set, LOOKUP_LSEARCH is forced.
7734 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7736 24. Added sender and host information to the "rejected by local_scan()" log
7737 line; previously there was no indication of these.
7739 25. Added .include_if_exists.
7741 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7742 Linux. It turns out that not all file systems support this. Apparently some
7743 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7744 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7745 not supported on the file descriptor, is now ignored when Exim is trying to
7746 sync a directory. This applies only to Linux.
7748 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7750 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7751 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7754 29. Added support for maildirsize files from supplied patch (modified a bit).
7756 30. The use of :fail: followed by an empty string could lead Exim to respond to
7757 sender verification failures with (e.g.):
7759 550 Verification failed for <xxx>
7760 550 Sender verify failed
7762 where the first response line was missing the '-' that indicates it is not
7763 the final line of the response.
7765 31. The loop for finding the name of the user that called Exim had a hardwired
7766 limit of 10; it now uses the value of finduser_retries, which is used for
7767 all other user lookups.
7769 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7772 33. Exim was neglecting to zero errno before one call of strtol() when
7773 expanding a string and expecting an integer value. On some systems this
7774 resulted in spurious "integer overflow" errors. Also, it was casting the
7775 result into an int without checking.
7777 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7778 did not work. The code looks as if it has *never* worked, though it appears
7779 to have been documented since at least release 1.62. I have made it work.
7781 35. The "timeout_DNS" error in retry rules, also documented since at least
7782 1.62, also never worked. As it isn't clear exactly what this means, and
7783 clearly it isn't a major issue, I have abolished the feature by treating it
7784 as "timeout", and writing a warning to the main and panic logs.
7786 36. The display of retry rules for -brt wasn't always showing the error code
7789 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7790 timeout_connect_A, timeout_connect_MX.
7792 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7793 to the empty sender.
7795 39. The daemon was not analysing the content of -oX till after it had closed
7796 stderr and disconnected from the controlling terminal. This meant that any
7797 syntax errors were only noted on the panic log, and the return code from
7798 the command was 0. By re-arranging the code a little, I've made the
7799 decoding happen first, so such errors now appear on stderr, and the return
7800 code is 1. However, the actual setting up of the sockets still happens in
7801 the disconnected process, so errors there are still only recorded on the
7804 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7805 connections (as happens on some IP stacks) was logged at start up time as
7806 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7807 it from "IPv6 and IPv4", which means that two separate sockets are being
7810 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7811 says whether AF_INET or AF_INET6 was passed as an argument.
7813 42. Exiwhat output was messed up when time zones were included in log
7816 43. Exiwhat now gives more information about the daemon's listening ports,
7817 and whether -tls-on-connect was used.
7819 44. The "port" option of the smtp transport is now expanded.
7821 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7822 silently ignored. Now an error message is written to the main and panic
7825 46. There's a new ACL modifier called "logwrite" which writes to a log file
7826 as soon as it is encountered.
7828 47. Added $local_user_uid and $local_user_gid at routing time.
7830 48. Exim crashed when trying to verify a sender address that was being
7833 49. Exim was recognizing only a space character after ".include". It now also
7834 recognizes a tab character.
7836 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7837 extracting the relevant information from the specification. The man page no
7838 longer contains scrambled data for the -d option, and I've added a section
7839 at the front about calling Exim under different names.
7841 51. Added "extra_headers" argument to the "mail" command in filter files.
7843 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7846 53. Installed eximstats 1.29.
7848 54. Added transport_filter_timeout as a generic transport option.
7850 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7851 Cc: header lines. This was required by RFC 822, but it not required by RFC
7854 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7855 incoming messages that did not have them. Now it does so only if the
7856 message originates locally, that is, if there is no associated remote host
7857 address. When Resent- header lines are present, this applies to the Resent-
7858 lines rather than the non-Resent- lines.
7860 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7861 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7863 58. Messages for configuration errors now include the name of the main
7864 configuration file - useful now that there may be more than one file in a
7865 list (.included file names were always shown).
7867 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7868 for those rare installations that do not start the daemon as root or run it
7869 setuid root. I've cut out the call to initgroups() if the daemon is not
7872 60. The Exim user and group can now be bound into the binary as text strings
7873 that are looked up at the start of Exim's processing.
7875 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7877 62. Added $mailstore_basename variable.
7879 63. Installed patch to sieve.c from Michael Haardt.
7881 64. When Exim failed to open the panic log after failing to open the main log,
7882 the original message it was trying to log was written to stderr and debug
7883 output, but if they were not available (the usual case in production), it
7884 was lost. Now it is written to syslog before the two lines that record the
7885 failures to open the logs.
7887 65. Users' Exim filters run in subprocesses under the user's uid. It is
7888 possible for a "deliver" command or an alias in a "personal" command to
7889 provoke an address rewrite. If logging of address rewriting is configured,
7890 this fails because the process is not running as root or exim. There may be
7891 a better way of dealing with this, but for the moment (because 4.30 needs
7892 to be released), I have disabled address rewrite logging when running a
7893 filter in a non-root, non-exim process.
7899 1. The buildconfig auxiliary program wasn't quoting the value set for
7900 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7901 not defined. This bug was masked in 4.22 by the effect that was fixed in
7904 2. Some messages that were rejected after a message id was allocated were
7905 shown as "incomplete" by exigrep. It no longer does this for messages that
7906 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7908 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7909 have allow_domain_literals set, the ID did not get logged in the <= line.
7910 Domain literals are now always recognized in Message-ID: header lines.
7912 4. The first argument for a ${extract expansion item is the key name or field
7913 number. Leading and trailing spaces in this item were not being ignored,
7914 causing some misleading effects.
7916 5. When deliver_drop_privilege was set, single queue runner processes started
7917 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7918 same command in the process it spins off) were not dropping privilege.
7920 6. When the daemon running as "exim" started a queue runner, it always
7921 re-executed Exim in the spun-off process. This is a waste of effort when
7922 deliver_drop_privilege is set. The new process now just calls the
7923 queue-runner function directly.
7929 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7932 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7933 ignored. Though the use of -oP was forcing the writing of a pid file, it
7934 was always written to the default place.
7936 3. If the message "no IP address found for host xxxx" is generated during
7937 incoming verification, it is now followed by identification of the incoming
7938 connection (so you can more easily find what provoked it).
7940 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7942 5. Added some features to "harden" Exim a bit more against certain attacks:
7944 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7945 be put in Local/Makefile. This is like the never_users runtime option,
7946 but it cannot be overridden. The default setting is "root".
7948 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7949 prefix string with which any file named in a -C command line option
7952 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7953 is retained for -C and -D only if the caller of Exim is root. Without
7954 it, the exim user may also use -C and -D and retain privilege.
7956 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7957 command line option is disabled.
7959 6. Macro names set by the -D option must start with an upper case letter, just
7960 like macro names defined in the configuration file.
7962 7. Added "dereference=" facility to LDAP.
7964 8. Two instances of the typo "uknown" in the source files are fixed.
7966 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7967 the Configure-Makefile script screwed up while processing it.
7969 10. Incorporated PCRE 4.4.
7971 11. The SMTP synchronization check was not operating right at the start of an
7972 SMTP session. For example, it could not catch a HELO sent before the client
7973 waited for the greeting. There is now a check for outstanding input at the
7974 point when the greeting is written. Because of the duplex, asynchronous
7975 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7976 way, but not yet received, when the check is performed.
7978 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7979 on TCP/IP sockets, because this apparently causes some broken clients to
7982 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7983 unchanged) from the Cygwin maintainer.
7985 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7986 was supported instead of testing for mbx. Effectively a typo.
7988 15. The spa authenticator server code was not checking that the input it
7989 received was valid base64.
7991 16. The debug output line for the "set" modifier in ACLs was not showing the
7992 name of the variable that was being set.
7994 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7996 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7997 lookup on its own. Something else was needed to trigger the lookup. For
7998 example, a match in host_lookup or the need for a host name in a host list.
7999 Now, if $sender_host_name is referenced and the host name has not yet been
8000 looked up, a lookup is performed. If the lookup fails, the variable remains
8001 empty, and $host_lookup_failed is set to "1".
8003 19. Added "eqi" as a case-independent comparison operator.
8005 20. The saslauthd authentication condition could segfault if neither service
8006 nor realm was specified.
8008 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8009 error message that was logged was misleading, and incoming SMTP
8010 connections were dropped. The message is now more accurate, and temporary
8011 errors are given to SMTP connections.
8013 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8014 (except RSET) with a 421 error, until QUIT is received. However, it was
8015 failing to send a response to QUIT.
8017 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8018 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8019 ACL is now run after the helo_try_verify_hosts code.
8021 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8022 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8023 case-independent, so other case variants are also recognized). Apparently
8024 some systems use these upper case variants.
8026 25. If more than two messages were waiting for the same host, and a transport
8027 filter was specified for the transport, Exim sent two messages over the
8028 same TCP/IP connection, and then failed with "socket operation on non-
8029 socket" when it tried to send the third.
8031 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8033 27. The extern definition of crypt16() in expand.c was not being excluded when
8034 the OS had its own crypt16() function.
8036 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8037 as a preferred synonym for return_size_limit, both as an option and as an
8040 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8042 30. Changed the default configuration ACL to relax the local part checking rule
8043 for addresses that are not in any local domains. For these addresses,
8044 slashes and pipe symbols are allowed within local parts, but the sequence
8045 /../ is explicitly forbidden.
8047 31. SPA server authentication was not clearing the challenge buffer before
8050 32. log_message in a "warn" ACL statement was writing to the reject log as
8051 well as to the main log, which contradicts the documentation and doesn't
8052 seem right (because no rejection is happening). So I have stopped it.
8054 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8055 However, I am unable to do any testing of this.
8057 34. Fixed an infelicity in the appendfile transport. When checking directories
8058 for a mailbox, to see if any needed to be created, it was accidentally
8059 using path names with one or more superfluous leading slashes; tracing
8060 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8062 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8063 contents are added to the log line that is written for every discarded
8064 recipient. (Previously a log_message setting was ignored.)
8066 36. The ${quote: operator now quotes the string if it is empty.
8068 37. The install script runs exim in order to find its version number. If for
8069 some reason other than non-existence or emptiness, which it checks, it
8070 could not run './exim', it was installing it with an empty version number,
8071 i.e. as "exim-". This error state is now caught, and the installation is
8074 38. An argument was missing from the function that creates an error message
8075 when Exim fails to connect to the socket for saslauthd authentication.
8076 This could cause Exim to crash, or give a corrupted message.
8078 39. Added isip, isip4, and isip6 to ${if conditions.
8080 40. The ACL variables $acl_xx are now saved with the message, and can be
8081 accessed later in routers, transports, and filters.
8083 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8084 strings in the file are not string-expanded.
8086 42. If a MAIL command specified a SIZE value that was too large to fit into an
8087 int variable, the check against message_size_limit failed. Such values are
8088 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8089 day this will have to be increased, but I don't think I want to be around
8090 when emails are that large.
8097 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8098 iconv() is not standard in FreeBSD.
8100 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8101 IPv6 enabled. The observed symptom was a segmentation fault on return from
8102 the function os_common_find_running_interfaces() in src/os.c.
8104 3. In the check_special_case() function in daemon.c I had used "errno" as an
8105 argument name, which causes warnings on some systems. This was basically a
8106 typo, since it was named "eno" in the comments!
8108 4. The code that waits for the clock to tick (at a resolution of some fraction
8109 of a second) so as to ensure message-id uniqueness was always waiting for
8110 at least one whole tick, when it could have waited for less. [This is
8111 almost certainly not relevant at current processor speeds, where it is
8112 unlikely to ever wait at all. But we try to future-proof.]
8114 5. The function that sleeps for a time interval that includes fractions of a
8115 second contained a race. It did not block SIGALRM between setting the
8116 timer, and suspending (a couple of lines later). If the interval was short
8117 and the sigsuspend() was delayed until after it had expired, the suspension
8118 never ended. On busy systems this could lead to processes getting stuck for
8121 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8122 process, before it forks any delivery processes. The open lookup caching
8123 mechanism meant that the open file or database connection was passed into
8124 the delivery process. The problem was that delivery processes always tidy
8125 up cached lookup data. This could cause a problem for the next delivery
8126 process started by the queue runner, because the external queue runner
8127 process does not know about the closure. So the next delivery process
8128 still has data in the lookup cache. In the case of a file lookup, there was
8129 no problem because closing a file descriptor in a subprocess doesn't affect
8130 the parent. However, if the lookup was caching a connection to a database,
8131 the connection was closed, and the second delivery process was likely to
8132 see errors such as "PGSQL: query failed: server closed the connection
8133 unexpectedly". The problem has been fixed by closing all cached lookups
8134 in a queue runner before running a delivery process.
8136 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8137 seem to have the "const" qualifier which it has on other OS. I've
8140 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8141 *of the same type* (client or server) with the same public name that an
8142 error should be diagnosed.
8144 9. When Exim looked up a host name for an IP address, but failed to find the
8145 original IP address when looking up the host name (a safety check), it
8146 output the message "<ip address> does not match any IP for NULL", which was
8147 confusing, to say the least. The bug was that the host name should have
8148 appeared instead of "NULL".
8150 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8151 user that is built into the binary, and the -C or -D options is used, root
8152 privilege is dropped before the configuration file is read. In addition,
8153 logging is switched to stderr instead of the normal log files. If the
8154 configuration then re-defines the Exim user, the unprivileged environment
8155 is probably not what is expected, so Exim logs a panic warning message (but
8158 However, if deliver_drop_privilege is set, the unprivileged state may well
8159 be exactly what is intended, so the warning has been cut out in that case,
8160 and Exim is allowed to try to write to its normal log files.
8166 1. smtp_return_error_details was not giving details for temporary sender
8167 or receiver verification errors.
8169 2. Diagnose a configuration error if two authenticators have the same public
8172 3. Exim used not to create the message log file for a message until the first
8173 delivery attempt. This could be confusing when incoming messages were held
8174 for policy or load reasons. The message log file is now created at the time
8175 the message is received, and an initial "Received" line is written to it.
8177 4. The automatically generated man page for command line options had a minor
8178 bug that caused no ill effects; however, a more serious problem was that
8179 the procedure for building the man page automatically didn't always
8180 operate. Consequently, release 4.20 contains an out-of-date version. This
8181 shouldn't happen again.
8183 5. When building Exim with embedded Perl support, the script that builds the
8184 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8185 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8187 6. The freeze_tell option was not being used for messages that were frozen on
8188 arrival, either by an ACL or by local_scan().
8190 7. Added the smtp_incomplete_transaction log selector.
8192 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8193 was accepting AUTH without a new EHLO.
8195 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8196 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8199 10. Logging of TCP/IP connections (when configured) now happens in the main
8200 daemon process instead of the child process, so that the TCP/IP connection
8201 count is more accurate (but it can never be perfect).
8203 11. The use of "drop" in a nested ACL was not being handled correctly in the
8204 outer ACL. Now, if condition failure induced by the nested "drop" causes
8205 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8206 or "require"), the connection is dropped.
8208 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8209 that yield "discard" can now be used with an "accept" or a "discard" verb,
8210 but an error is generated for any others (because I can't see a useful way
8211 to define what should happen).
8213 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8214 are now processed in the same way as lines in the Exim configuration file.
8215 In particular, continuation lines are supported.
8217 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8219 15. Added -ti meaning -t -i.
8221 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8222 domains, and warn by logging if others are found.
8224 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8225 returned from GIFCONF when Exim is trying to find the list of interfaces on
8226 a host. The code in os.c has been modified to copy each ifreq to an aligned
8227 structure in all cases.
8229 Also, in some cases, the returned ifreq's were being copied to a 'struct
8230 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8231 means the last couple of bytes of an IPv6 address could be chopped if the
8232 ifreq contained only a normal sockaddr (14 bytes storage).
8234 18. Named domain lists were not supported in the hosts_treat_as_local option.
8235 An entry such as +xxxx was not recognized, and was treated as a literal
8238 19. Ensure that header lines added by a DATA ACL are included in the reject log
8239 if the ACL subsequently rejects the message.
8241 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8242 MD5 (which is deprecated).
8244 21. When testing a filter file using -bf, Exim was writing a message when it
8245 took the sender from a "From " line in the message, but it was not doing so
8246 when it took $return_path from a Return-Path: header line. It now does.
8248 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8249 with a valid header line field name (a series of printing characters
8250 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8252 23. Changed "disc" in the source to "disk" to conform to the documentation and
8253 the book and for uniformity.
8255 24. Ignore Sendmail's -Ooption=value command line item.
8257 25. When execve() failed while trying to run a command in a pipe transport,
8258 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8259 could be confused with a return value of 69 from the command itself. This
8260 has been changed to 127, the value the shell returns if it is asked to run
8261 a non-existent command. The wording for the related log line suggests a
8262 non-existent command as the problem.
8264 26. If received_header_text expands to an empty string, do not add a Received:
8265 header line to the message. (Well, it adds a token one on the spool, but
8266 marks it "old" so that it doesn't get used or transmitted.)
8268 27. Installed eximstats 1.28 (addition of -nt option).
8270 28. There was no check for failure on the call to getsockname() in the daemon
8271 code. This can fail if there is a shortage of resources on the system, with
8272 ENOMEM, for example. A temporary error is now given on failure.
8274 29. Contrary to the C standard, it seems that in some environments, the
8275 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8276 program. Exim now does this explicitly; it affects the formatting of
8277 timestamps using strftime().
8279 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8280 complaints. I've now initialized all the variables, to avoid this.
8282 32. Header lines added by a system filter were not being "seen" during
8283 transport-time rewrites.
8285 33. The info_callback() function passed to OpenSSL is set up with type void
8286 (*)(SSL *, int, int), as described somewhere. However, when calling the
8287 function (actually a macro) that sets it up, the type void(*)() is
8288 expected. I've put in a cast to prevent warnings from picky compilers.
8290 34. If a DNS black list lookup found a CNAME record, but there were no A
8291 records associated with the domain it pointed at, Exim crashed.
8293 35. If a DNS black list lookup returned more than one A record, Exim ignored
8294 all but the first. It now scans all returned addresses if a particular IP
8295 value is being sought. In this situation, the contents of the
8296 $dnslist_value variable are a list of all the addresses, separated by a
8299 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8300 to accept a host name and all its aliases if the forward lookup for any of
8301 them yielded the IP address of the incoming connection. Now it accepts only
8302 those names whose forward lookup yields the correct IP address. Any other
8303 names are discarded. This closes a loophole whereby a rogue DNS
8304 administrator could create reverse DNS records to break through a
8305 wildcarded host restriction in an ACL.
8307 37. If a user filter or a system filter that ran in a subprocess used any of
8308 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8309 the wrong values were passed to the pipe command ($thisaddress had the
8310 value of $0, $0 had the value of $1, etc). This bug was introduced by
8311 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8313 38. Improved the line breaking for long SMTP error messages from ACLs.
8314 Previously, if there was no break point between 40 and 75 characters, Exim
8315 left the rest of the message alone. Two changes have been made: (a) I've
8316 reduced the minimum length to 35 characters; (b) if it can't find a break
8317 point between 35 and 75 characters, it looks ahead and uses the first one
8318 that it finds. This may give the occasional overlong line, but at least the
8319 remaining text gets split now.
8321 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8322 file descriptors might be low, and that setting 1000 would always raise it.
8323 It turns out that in some environments, the limit is already over 1000 and
8324 that lowering it causes trouble. So now Exim takes care not to decrease it.
8326 40. When delivering a message, the value of $return_path is set to $sender_
8327 address at the start of routing (routers may change the value). By an
8328 oversight, this default was not being set up when an address was tested by
8329 -bt or -bv, which affected the outcome if any router or filter referred to
8332 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8333 log only when "message" or "log_message" are set. However, if one of the
8334 conditions was an address verification, or a call to a nested ACL, the
8335 messages generated by the underlying test were being passed through. This
8336 no longer happens. The underlying message is available in $acl_verify_
8337 message for both "message" and "log_message" expansions, so it can be
8338 passed through if needed.
8340 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8341 new expansion $bh_ to give the encoded byte string without charset
8342 translation. Translation happens only if iconv() is available; HAVE_ICONV
8343 indicates this at build time. HEADERS_CHARSET gives the charset to
8344 translate to; headers_charset can change it in the configuration, and
8345 "headers charset" can change it in an individual filter file.
8347 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8348 that creates RFC 2047 encoded "words" labels them as that charset instead
8349 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8350 expansion operator; (ii) when Exim creates a From: line for a local
8351 message; (iii) when a header line is rewritten to include a "phrase" part.
8353 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8354 buggy, causing it to skip the first lines of messages whose message ID
8355 ended in 'D'. This would not have bitten before Exim release 4.14, because
8356 message IDs were unlikely to end in 'D' before then. The effect was to have
8357 incorrect size information for certain domains.
8359 45. #include "config.h" was missing at the start of the crypt16.c module. This
8360 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8363 46. If there was a timeout during a "random" callout check, Exim treated it as
8364 a failure of the random address, and carried on sending RSET and the real
8365 address. If the delay was just some slowness somewhere, the response to the
8366 original RCPT would be taken as a response to RSET and so on, causing
8367 mayhem of various kinds.
8369 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8370 when I implemented it. It didn't allow for the fact that some option values
8371 may legitimately be negative (e.g. size_addition), and it didn't even do
8372 the right test for positive values.
8374 48. Domain names in DNS records are case-independent. Exim always looks them up
8375 in lower case. Some resolvers return domain names in exactly the case they
8376 appear in the zone file, that is, they may contain uppercase letters. Not
8377 all resolvers do this - some return always lower case. Exim was treating a
8378 change of case by a resolver as a change of domain, similar to a widening
8379 of a domain abbreviation. This triggered its re-routing code and so it was
8380 trying to route what was effectively the same domain again. This normally
8381 caused routing to fail (because the router wouldn't handle the domain
8382 twice). Now Exim checks for this case specially, and just changes the
8383 casing of the domain that it ultimately uses when it transmits the message
8386 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8389 50. If a filter generated a file delivery with a non-absolute name (possible if
8390 no home directory exists for the router), the forbid_file option was not
8393 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8394 the existing equality matching.
8396 52. Exim was using ints instead of ino_t variables in some places where it was
8397 dealing with inode numbers.
8399 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8400 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8401 called TMPDIR, and if it finds it is different, it changes its value.
8403 54. The smtp_printf() function is now made available to local_scan() so
8404 additional output lines can be written before returning. There is also an
8405 smtp_fflush() function to enable the detection of a dropped connection.
8406 The variables smtp_input and smtp_batched_input are exported to
8409 55. Changed the default runtime configuration: the message "Unknown user"
8410 has been removed from the ACL, and instead placed on the localuser router,
8411 using the cannot_route_message feature. This means that any verification
8412 failures that generate their own messages won't get overridden. Similarly,
8413 the "Unrouteable address" message that was in the ACL for unverifiable
8414 relay addresses has also been removed.
8416 56. Added hosts_avoid_esmtp to the smtp transport.
8418 57. The exicyclog script was not checking for the esoteric option
8419 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8420 will work only if exicyclog is run under the appropriate euid.
8422 58. Following a discussion on the list, the rules by which Exim recognises line
8423 endings on incoming messages have been changed. The -dropcr and drop_cr
8424 options are now no-ops, retained only for backwards compatibility. The
8425 following line terminators are recognized: LF CRLF CR. However, special
8426 processing applies to CR:
8428 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8429 nor a local message in the state where . is a terminator.
8431 (ii) If a bare CR is encountered in a header line, an extra space is added
8432 after the line terminator so as not to end the header. The reasoning
8433 behind this is that bare CRs in header lines are most likely either
8434 to be mistakes, or people trying to play silly games.
8436 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8437 was being incorrectly given as 18 bytes larger than it should have been.
8438 This is a VOB (very old bug).
8440 60. This may never have affected anything current, but just in case it has:
8441 When the local host is found other than at the start of a list of hosts,
8442 the local host, those with the same MX, and any that follow, are discarded.
8443 When the list in question was part of a longer list of hosts, the following
8444 hosts (not currently being processed) were also being discarded. This no
8445 longer happens. I'm not sure if this situation could ever has previously
8448 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8451 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8452 header. This is something that is recommended in a new Internet Draft, and
8453 is something that is documented as being done by Sendmail. There are two
8454 possible values. For messages generated by the autoreply transport, Exim
8457 Auto-Submitted: auto-replied
8459 whereas for all other generated messages (e.g. bounces) it adds
8461 Auto-Submitted: auto-generated
8463 63. The "personal" condition in filters now includes a test for the
8464 Auto-Submitted: header. If it contains the string "auto-" the message it
8465 not considered personal.
8467 64. Added rcpt_include_affixes as a generic transport option.
8469 65. Added queue_only_override (default true).
8471 66. Added the syslog_duplication option.
8473 67. If what should have been the first header line of a message consisted of
8474 a space followed by a colon, Exim was mis-interpreting it as a header line.
8475 It isn't of course - it is syntactically invalid and should therefore be
8476 treated as the start of the message body. The misbehaviour could have
8477 caused a number of strange effects, including loss of data in subsequent
8478 header lines, and spool format errors.
8480 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8481 client host had authenticated. This control can now be exercised by an ACL
8482 for more flexibility.
8484 69. By default, callouts do not happen when testing with -bh. There is now a
8485 variant, -bhc, which does actually run the callout code, including
8486 consulting and updating the callout cache.
8488 70. Added support for saslauthd authentication, courtesy of Alexander
8491 71. If statvfs() failed on the spool or log directories while checking their
8492 size for availability, Exim confusingly gave the error "space shortage".
8493 Furthermore, in debugging mode it crashed with a floating point exception.
8494 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8495 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8496 serious problem, Exim now writes to the main and panic logs when this
8497 happens, with details of the failure. It then refuses to accept the
8498 incoming message, giving the message "spool directory problem" or "log
8499 directory problem" with a 421 code for SMTP messages.
8501 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8502 0, 1, and 2 exist, because some OS complain for execs without them (see
8503 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8504 descriptors. However, the code omitted to check that the open succeeded,
8505 causing mysterious errors if for some reason the permissions on /dev/null
8506 got screwed. Now Exim writes a message to the main and panic logs, and
8507 bombs out if it can't open /dev/null.
8509 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8510 interact so that it is all more flexible. It is supposed to remain
8511 backwards compatible. Also added extra_local_interfaces.
8513 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8514 to bomb out with an assertion failure - to the client this appears as a
8515 connection drop. This problem occurs in the part of the code that was taken
8516 from the Samba project. Fortunately, the assertion is in a very simple
8517 function, so I have fixed this by reproducing the function inline in the
8518 one place where it is called, and arranging for authentication to fail
8519 instead of killing the process with assert().
8521 75. The SPA client code was not working when the server requested OEM rather
8522 than Unicode encoding.
8524 76. Added code to make require_files with a specific uid setting more usable in
8525 the case where statting the file as root fails - usually a non-root-mounted
8526 NFS file system. When this happens and the failure is EACCES, Exim now
8527 forks a subprocess and does the per-uid checking as the relevant uid.
8529 77. Added process_log_path.
8531 78. If log_file_path was not explicitly set, a setting of check_log_space or
8532 check_log_inodes was ignored.
8534 79. If a space check for the spool or log partitions fails, the incident is now
8535 logged. Of course, in the latter case the data may get lost...
8537 80. Added the %p formatting code to string_format() so that it can be used to
8538 print addresses in debug_print(). Adjusted all the address printing in the
8539 debugging in store.c to use %p rather than %d.
8541 81. There was a concern that a line of code in smtp_in.c could overflow a
8542 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8543 initially expressed, the concern was not well-founded, because trailing
8544 spaces are removed early. However, if the trailing spaces were followed by
8545 a NULL, they did not get removed, so the overflow was possible. Two fixes
8548 (a) I re-wrote the offending code in a cleaner fashion.
8549 (b) If an incoming SMTP command contains a NULL character, it is rejected
8552 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8553 runs initgroups(), so that if the Exim user is in any additional groups,
8554 they will be used during message reception.
8560 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob