1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.456 2007/01/22 15:56:47 steve Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
10 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
13 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
16 PH/02 In an ACL statement such as
18 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
20 if a client was not listed at all, or was listed with a value other than
21 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
22 the condition was not true (as it should be), so access was not denied.
23 The bug was that the ! inversion was incorrectly passed on to the second
24 item. This has been fixed.
26 PH/03 Added additional dnslists conditions == and =& which are different from
27 = and & when the dns lookup returns more than one IP address.
29 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
30 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
32 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
38 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
39 fixed by 4.65/MH/01 (is this a record?) are fixed:
41 (i) An empty string was always treated as zero by the numeric comparison
42 operators. This behaviour has been restored.
44 (ii) It is documented that the numeric comparison operators always treat
45 their arguments as decimal numbers. This was broken in that numbers
46 starting with 0 were being interpreted as octal.
48 While fixing these problems I realized that there was another issue that
49 hadn't been noticed. Values of message_size_limit (both the global option
50 and the transport option) were treated as octal if they started with 0.
51 The documentation was vague. These values are now always treated as
52 decimal, and I will make that clear in the documentation.
58 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
59 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
62 MH/01 Don't check that the operands of numeric comparison operators are
63 integers when their expansion is in "skipping" mode (fixes bug
64 introduced by 4.64-PH/07).
66 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
67 child addresses, Exim now panics and dies. Previously, because the count
68 is held in a short int, deliveries were likely to be lost. As such a
69 large number of recipients for a single message is ridiculous
70 (performance will be very, very poor), I have chosen to impose a limit
71 rather than extend the field.
77 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
78 leftover -K file (the existence of which was triggered by #402).
79 While we were at it, introduced process PID as part of the -K
80 filename. This should rule out race conditions when creating
83 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
84 processing considerably. Previous code took too long for large mails,
85 triggering a timeout which in turn triggers #401.
87 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
88 in the DK code in transports.c. sendfile() is not really portable,
89 hence the _LINUX specificness.
91 TF/01 In the add_headers option to the mail command in an Exim filter,
92 there was a bug that Exim would claim a syntax error in any
93 header after the first one which had an odd number of characters
96 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
97 callout verification, Exim cached a "reject" for the entire domain. This
98 is correct for most verifications, but it is not correct for a recipient
99 verification with use_sender or use_postmaster set, because in that case
100 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
101 case of MAIL FROM:<> rejection from other early rejections (e.g.
102 rejection of HELO). When verifying a recipient using a non-null MAIL
103 address, the cache is ignored if it shows MAIL FROM:<> rejection.
104 Whatever the result of the callout, the value of the domain cache is
105 left unchanged (for any other kind of callout, getting as far as trying
106 RCPT means that the domain itself is ok).
108 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
111 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
112 manifest itself as EPIPE rather than ECONNECT. When tidying away a
113 session, the daemon ignores ECONNECT errors and logs others; it now
114 ignores EPIPE as well.
116 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
117 (quoted-printable decoding).
119 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
120 later the small subsequent patch to fix an introduced bug.
122 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
124 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
126 PH/08 An error is now given if message_size_limit is specified negative.
128 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
129 to be given (somewhat) arbitrary names.
131 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
134 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
135 miscellaneous code fixes
137 PH/10 Added the log_reject_target ACL modifier to specify where to log
140 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
141 hostname. This is wrong, because it relates to the incoming message (and
142 probably the interface on which it is arriving) and not to the outgoing
143 callout (which could be using a different interface). This has been
144 changed to use the value of the helo_data option from the smtp transport
145 instead - this is what is used when a message is actually being sent. If
146 there is no remote transport (possible with a router that sets up host
147 addresses), $smtp_active_hostname is used.
149 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
150 tweaks were necessary in order to get it to work (see also 21 below):
151 (a) The code assumed that strncpy() returns a negative number on buffer
152 overflow, which isn't the case. Replaced with Exim's string_format()
154 (b) There were several signed/unsigned issues. I just did the minimum
155 hacking in of casts. There is scope for a larger refactoring.
156 (c) The code used strcasecmp() which is not a standard C function.
157 Replaced with Exim's strcmpic() function.
158 (d) The code set only $1; it now sets $auth1 as well.
159 (e) A simple test gave the error "authentication client didn't specify
160 service in request". It would seem that Dovecot has changed its
161 interface. Fortunately there's a specification; I followed it and
162 changed what the client sends and it appears to be working now.
164 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
167 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
168 address A is aliased to B and C, where B exists and C does not. Without
169 -v the output is "A verified" because verification stops after a
170 successful redirection if more than one address is generated. However,
171 with -v the child addresses are also verified. Exim was outputting "A
172 failed to verify" and then showing the successful verification for C,
173 with its parentage. It now outputs "B failed to verify", showing B's
174 parentage before showing the successful verification of C.
176 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
177 look up a TXT record in a specific list after matching in a combined
180 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
181 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
182 they consult the DNS. I had assumed they would set it the way they
183 wanted; and indeed my experiments on Linux seem to show that in some
184 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
185 To be on the safe side, however, I have now made the interface to
186 host_find_byname() similar to host_find_bydns(), with an argument
187 containing the DNS resolver options. The host_find_byname() function now
188 sets these options at its start, just as host_find_bydns() does. The smtp
189 transport options dns_qualify_single and dns_search_parents are passed to
190 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
191 of host_find_byname() use the default settings of RES_DEFNAMES
192 (qualify_single) but not RES_DNSRCH (search_parents).
194 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
195 spool_read_header() do less string testing, by means of a preliminary
196 switch on the second character of optional "-foo" lines. (This is
197 overdue, caused by the large number of possibilities that now exist.
198 Originally there were few.) While I was there, I also converted the
199 str(n)cmp tests so they don't re-test the leading "-" and the first
200 character, in the hope this might squeeze out yet more improvement.
202 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
203 flag allowing group syntax was set by the header_syntax check but not
204 turned off, possible causing trouble later; (2) The flag was not being
205 set at all for the header_verify test, causing "group"-style headers to
206 be rejected. I have now set it in this case, and also caused header_
207 verify to ignore an empty address taken from a group. While doing this, I
208 came across some other cases where the code for allowing group syntax
209 while scanning a header line wasn't quite right (mostly, not resetting
210 the flag correctly in the right place). These bugs could have caused
211 trouble for malformed header lines. I hope it is now all correct.
213 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
214 with the "reply" argument non-NULL. The code, however (which originally
215 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
216 but it didn't always do it. This confused somebody who was copying the
217 code for some other use. I have removed all the tests.
219 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
220 feature that was used to support insecure browsers during the U.S. crypto
221 embargo. It requires special client support, and Exim is probably the
222 only MTA that supported it -- and would never use it because real RSA is
223 always available. This code has been removed, because it had the bad
224 effect of slowing Exim down by computing (never used) parameters for the
225 RSA_EXPORT functionality.
227 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
228 authenticator to fail if there's a tab character in the incoming data
229 (there should never be unless someone is messing about, as it's supposed
230 to be base64-encoded). Also added, on Timo's advice, the "secured" option
231 if the connection is using TLS or if the remote IP is the same as the
232 local IP, and the "valid-client-cert option" if a client certificate has
235 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
236 authenticators. This can be used for authorization after authentication
237 succeeds. (In the case of plaintext, it servers for both authentication
240 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
241 if any retry times were supplied.
243 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
244 connection, where there is no client IP address to check. In this
245 situation, the verify now always succeeds.
247 PH/25 Applied John Jetmore's -Mset patch.
249 PH/26 Added -bem to be like -Mset, but loading a message from a file.
251 PH/27 In a string expansion for a processed (not raw) header when multiple
252 headers of the same name were present, leading whitespace was being
253 removed from all of them, but trailing whitespace was being removed only
254 from the last one. Now trailing whitespace is removed from each header
255 before concatenation. Completely empty headers in a concatenation (as
258 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
259 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
261 PH/29 [Removed. This was a change that I later backed out, and forgot to
262 correct the ChangeLog entry (that I had efficiently created) before
263 committing the later change.]
265 PH/30 Exim was sometimes attempting to deliver messages that had suffered
266 address errors (4xx response to RCPT) over the same connection as other
267 messages routed to the same hosts. Such deliveries are always "forced",
268 so retry times are not inspected. This resulted in far too many retries
269 for the affected addresses. The effect occurred only when there were more
270 hosts than the hosts_max_try setting in the smtp transport when it had
271 the 4xx errors. Those hosts that it had tried were not added to the list
272 of hosts for which the message was waiting, so if all were tried, there
273 was no problem. Two fixes have been applied:
275 (i) If there are any address or message errors in an SMTP delivery, none
276 of the hosts (tried or untried) are now added to the list of hosts
277 for which the message is waiting, so the message should not be a
278 candidate for sending over the same connection that was used for a
279 successful delivery of some other message. This seems entirely
280 reasonable: after all the message is NOT "waiting for some host".
281 This is so "obvious" that I'm not sure why it wasn't done
282 previously. Hope I haven't missed anything, but it can't do any
283 harm, as the worst effect is to miss an optimization.
285 (ii) If, despite (i), such a delivery is accidentally attempted, the
286 routing retry time is respected, so at least it doesn't keep
287 hammering the server.
289 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
290 in ${readsocket because some servers need this prod.
292 PH/32 Added some extra debug output when updating a wait-xxx database.
294 PH/33 The hint "could be header name not terminated by colon", which has been
295 given for certain expansion errors for a long time, was not being given
296 for the ${if def:h_colon_omitted{... case.
298 PH/34 The spec says: "With one important exception, whenever a domain list is
299 being scanned, $domain contains the subject domain." There was at least
300 one case where this was not true.
302 PH/35 The error "getsockname() failed: connection reset by peer" was being
303 written to the panic log as well as the main log, but it isn't really
304 panic-worthy as it just means the connection died rather early on. I have
305 removed the panic log writing for the ECONNRESET error when getsockname()
308 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
309 runs only) independently of the message's sender address. This meant
310 that, if the 4xx error was in fact related to the sender, a different
311 message to the same recipient with a different sender could confuse
312 things. In particualar, this can happen when sending to a greylisting
313 server, but other circumstances could also provoke similar problems.
314 I have changed the default so that the retry time for these errors is now
315 based a combination of the sender and recipient addresses. This change
316 can be overridden by setting address_retry_include_sender=false in the
319 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
320 remote server are returned as part of bounce messages. This was not
321 happening for LMTP over a pipe (the lmtp transport), but now it is the
322 same for both kinds of LMTP.
324 PH/38 Despite being documented as not happening, Exim was rewriting addresses
325 in header lines that were in fact CNAMEs. This is no longer the case.
327 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
328 and queue runs started by the daemon processed all messages. This has
329 been fixed so that -R and -S can now usefully be given with -q<time>.
331 PH/40 Import PCRE release 6.7 (fixes some bugs).
333 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
335 PH/42 Give an error if -q is specified more than once.
337 PH/43 Renamed the variables $interface_address and $interface_port as
338 $received_ip_address and $received_port, to make it clear that these
339 values apply to message reception, and not to the outgoing interface when
340 a message is delivered. (The old names remain recognized, of course.)
342 PH/44 There was no timeout on the connect() call when using a Unix domain
343 socket in the ${readsocket expansion. There now is.
345 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
346 be meaningful with "accept".
348 SC/01 Eximstats V1.43
349 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
351 SC/02 Eximstats V1.44
352 Use a glob alias rather than an array ref in the generated
353 parser. This improves both readability and performance.
355 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
356 Collect SpamAssassin and rejection statistics.
357 Don't display local sender or destination tables unless
358 there is data to show.
359 Added average volumes into the top table text output.
361 SC/04 Eximstats V1.46
362 Collect data on the number of addresses (recipients)
363 as well as the number of messages.
365 SC/05 Eximstats V1.47
366 Added 'Message too big' to the list of mail rejection
367 reasons (thanks to Marco Gaiarin).
369 SC/06 Eximstats V1.48
370 Mainlog lines which have GMT offsets and are too short to
371 have a flag are now skipped.
373 SC/07 Eximstats V1.49 (Alain Williams)
374 Added the -emptyok flag.
376 SC/08 Eximstats V1.50
377 Fixes for obtaining the IP address from reject messages.
379 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
380 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
381 whitesspace changes from 4.64-PH/27
383 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
386 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
387 are found, allow negative numbers in numeric criteria)
389 JJ/06 exipick.20061117.2, added new $message_body_missing variable
391 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
392 to match changes made in 4.64-PH/43
394 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
396 PH/47 Put in an explicit test for a DNS lookup of an address record where the
397 "domain" is actually an IP address, and force a failure. This locks out
398 those revolvers/nameservers that support "A-for-A" lookups, in
399 contravention of the specifications.
401 PH/48 When a host name was looked up from an IP address, and the subsequent
402 forward lookup of the name timed out, the host name was left in
403 $sender_host_name, contrary to the specification.
405 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
406 restricted to single-key lookups, Exim was not diagnosing an error if
407 * or *@ was used with a query-style lookup.
409 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
411 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
412 long ago, but noone interested enough thought of it. Let's just say that
413 the "1.1" means that there are some new functions that weren't there at
414 some point in the past.
416 PH/51 Error processing for expansion failure of helo_data from an smtp
417 transport during callout processing was broken.
419 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
420 tested/used via the -bh/-bhc/-bs options.
422 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
423 bug, fixed in subsequent PCRE releases).
425 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
426 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
428 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
434 SC/01 Use a glob alias rather than an array ref in eximstats generated
435 parser. This improves both readability and performance.
437 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
438 Don't display local sender or destination tables in eximstats unless
439 there is data to show.
440 Added average volumes into the eximstats top table text output.
442 SC/03 Collect data on the number of addresses (recipients) as well
443 as the number of messages in eximstats.
445 TF/01 Correct an error in the documentation for the redirect router. Exim
446 does (usually) call initgroups() when daemonizing.
448 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
449 with consistent privilege compared to when running as a daemon.
451 TF/03 Note in the spec that $authenticated_id is not set for local
452 submissions from trusted users.
454 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
455 Thanks to Dean Brooks <dean@iglou.com> for the patch.
457 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
458 by adding some example configuration directives to the default
459 configuration file. A little bit of work is required to uncomment the
460 directives and define how usernames and passwords are checked, but
461 there is now a framework to start from.
463 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
464 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
465 without this. I don't know how relevant this is to other LDAP libraries.
467 PH/02 Add the verb name to the "unknown ACL verb" error.
469 PH/03 Magnus Holmgren's patch for filter_prepend_home.
471 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
473 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
474 directory not expanded when it should be if an expanded home directory
475 was set for the address (which is overridden by the transport).
477 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
480 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
481 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
482 because it is too late at that time, and has no effect.
484 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
485 security issue with \' (bugzilla #107). I could not use the
486 PQescapeStringConn() function, because it needs a PGconn value as one of
489 PH/08 When testing addresses using -bt, indicate those final addresses that
490 are duplicates that would not cause an additional delivery. At least one
491 person was confused, thinking that -bt output corresponded to deliveries.
492 (Suppressing duplicates isn't a good idea as you lose the information
493 about possibly different redirections that led to the duplicates.)
495 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
496 systems where poll() doesn't work, in particular OS X.
498 PH/10 Added more information to debugging output for retry time not reached.
500 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
501 operations in malware.c.
503 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
506 PH/13 If write_rejectlog was set false when logging was sent to syslog with
507 syslog_duplication set false, log lines that would normally be written
508 both the the main log and to the reject log were not written to syslog at
511 PH/14 In the default configuration, change the use of "message" in ACL warn
512 statements to "add_header".
514 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
515 not followed by a command (e.g. "seen endif").
517 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
518 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
521 PH/17 Added extra conditions to the default value of delay_warning_condition
525 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
526 { match{$h_precedence:}{(?i)bulk|list|junk} } \
527 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
530 The Auto-Submitted: and various List- headers are standardised, whereas I
531 don't think Precedence: ever was.
533 PH/18 Refactored debugging code in route_finduser() to show more information,
534 in particular, the error code if getpwnam() issues one.
536 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
537 This is apparently needed in addition to the PH/07 change above to avoid
538 any possible encoding problems.
540 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
541 but not after initializing Perl.
543 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
544 output them only if debugging. By default they are written stderr,
545 apparently, which is not desirable.
547 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
550 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
553 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
555 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
556 authenticated or an ident call has been made. Suppress the default
557 values for $authenticated_id and $authenticated_sender (but permit -oMai
558 and -oMas) when testing with -bh.
560 PH/24 Re-jigged the order of the tests in the default configuration so that the
561 tests for valid domains and recipients precede the DNS black list and CSA
562 tests, on the grounds that those ones are more expensive.
564 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
565 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
566 This bug exists in every version of Exim that I still have, right back to
569 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
570 However, an attempt to turn on case-sensitivity in a regex key by
571 including (?-i) didn't work because the subject string was already
572 lowercased, and the effects were non-intuitive. It turns out that a
573 one-line patch can be used to allow (?-i) to work as expected.
579 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
580 other effects) broke the use of negated acl sub-conditions.
582 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
585 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
586 "Deny" causes Exim to reject the incoming connection with a 554 error.
587 Unfortunately, if there is a major crisis, such as a disk failure,
588 tcp-wrappers gives "deny", whereas what one would like would be some
589 kind of temporary error. A kludge has been added to help with this.
590 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
591 554 error is used if errno is still zero or contains ENOENT (which occurs
592 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
595 PH/03 Add -lutil to the default FreeBSD LIBS setting.
597 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
598 errors. Otherwise a message that provokes a temporary error (when other
599 messages do not) can cause a whole host to time out.
601 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
602 addresses were routed directly to files or pipes from a redirect router.
603 File deliveries just didn't batch; pipe deliveries might have suffered
606 PH/06 A failure to get a lock for a hints database would erroneously always say
607 "Failed to get write lock", even when it was really a read lock.
609 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
610 of 0600. This has been changed to use the value of the lockfile_mode
611 option (which defaults to 0600).
613 PH/08 Applied small patch from the Sieve maintainer.
615 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
616 folder from quota calculations, a direct delivery into this folder messed
617 up the contents of the maildirsize file. This was because the regex was
618 used only to exclude .Trash (or whatever) when the size of the mailbox
619 was calculated. There was no check that a delivery was happening into an
620 excluded directory. This bug has been fixed by ignoring all quota
621 processing for deliveries into excluded directories.
623 PH/10 Added the maildirfolder_create_regex option to appendfile.
629 PH/01 The code for finding all the local interface addresses on a FreeBSD
630 system running IPv6 was broken. This may well have applied to all BSD
631 systems, as well as to others that have similar system calls. The broken
632 code found IPv4 interfaces correctly, but gave incorrect values for the
633 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
634 that it would not match correctly against @[] and not recognize the IPv6
637 PH/02 The ipliteral router was not recognizing addresses of the form user@
638 [ipv6:....] because it didn't know about the "ipv6:" prefix.
640 PH/03 Added disable_ipv6.
642 PH/04 Changed $reply_address to use the raw form of the headers instead of the
643 decoded form, because it is most often used to construct To: headers
644 lines in autoreplies, and the decoded form may well be syntactically
645 invalid. However, $reply_address has leading white space removed, and all
646 newlines turned into spaces so that the autoreply transport does not
649 PH/05 If group was specified without a user on a router, and no group or user
650 was specified on a transport, the group from the router was ignored.
652 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
653 for visible compile-time settings that can be used to change these
654 numbers, for those that want even more. Backwards compatibility with old
655 spool files has been maintained. However, going back to a previous Exim
656 release will lost any variables that are in spool files.
658 PH/07 Two small changes when running in the test harness: increase delay when
659 passing a TCP/IP connection to a new process, in case the original
660 process has to generate a bounce, and remove special handling of
661 127.0.0.2 (sic), which is no longer necessary.
663 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
664 be the same on different OS.
666 PH/09 Moved a debug statement in filter processing to avoid a race problem when
669 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
670 whether --show-vars was specified or not
672 JJ/02 exipick: Added support for new ACL variable spool format introduced
675 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
676 syntactically invalid From: or Reply-to: line, and a filter used this to
677 generate an autoreply, and therefore failed to obtain an address for the
678 autoreply, Exim could try to deliver to a non-existent relative file
679 name, causing unrelated and misleading errors. What now happens is that
680 it logs this as a hard delivery error, but does not attempt to create a
683 PH/11 The exinext utility has a -C option for testing purposes, but although
684 the given file was scanned by exinext itself; it wasn't being passed on
685 when Exim was called.
687 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
688 an end-of-file indication when reading a command response.
690 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
691 compiled. In many other places in Exim, IPv6 addresses are always
692 recognized, so I have changed this. It also means that IPv4 domain
693 literals of the form [IPV4:n.n.n.n] are now always recognized.
695 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
696 used if the router is not running as root, for example, when verifying at
697 ACL time, or when using -bh. The debugging output from this situation was
698 non-existent - all you got was a failure to exec. I have made two
701 (a) Failures to set uid/gid, the current directory, or a process leader
702 in a subprocess such as that created by queryprogram now generate
703 suitable debugging ouput when -d is set.
705 (b) The queryprogram router detects when it is not running as root,
706 outputs suitable debugging information if -d is set, and then runs
707 the subprocess without attempting to change uid/gid.
709 PH/15 Minor change to Makefile for building test_host (undocumented testing
712 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
713 additional section of a DNS packet that returns MX or SRV records.
714 Instead, it always explicitly searches for A/AAAA records. This avoids
715 major problems that occur when a DNS server includes only records of one
716 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
717 fixed another bug: if SRV records were looked up and the corresponding
718 address records were *not* found in the additional section, the port
719 values from the SRV records were lost.
721 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
722 using the correct key (the original address) when searching the retry
723 rules in order to find which one to use for generating the retry hint.
725 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
726 adding the default one. Similarly, if it contains a Reply-To: header, the
727 errors_reply_to option, if set, is not used.
729 PH/19 When calculating a retry time, Exim used to measure the "time since
730 failure" by looking at the "first failed" field in the retry record. Now
731 it does not use this if it is later than than the arrival time of the
732 message. Instead it uses the arrival time. This makes for better
733 behaviour in cases where some deliveries succeed, thus re-setting the
734 "first failed" field. An example is a quota failure for a huge message
735 when small messages continue to be delivered. Without this change, the
736 "time since failure" will always be short, possible causing more frequent
737 delivery attempts for the huge message than are intended.
738 [Note: This change was subsequently modified - see PH/04 for 4.62.]
740 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
741 $1, $2, $3) because the numerical variables can be reset during some
742 expansion items (e.g. "match"), thereby losing the authentication data.
744 PH/21 Make -bV show the size of off_t variables so that the test suite can
745 decide whether to run tests for quotas > 2G.
747 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
748 mailbox_size, and mailbox_filecount in the appendfile transport. If a
749 filecount value is greater than 2G or if a quota value is greater than 2G
750 on a system where the size of off_t is not greater than 4, a panic error
753 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
754 never match. The debug and -bh output now contains an explicit error
755 message indicating a malformed IPv4 address or mask.
757 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
758 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
761 PH/25 Do not write to syslog when running in the test harness. The only
762 occasion when this arises is a failure to open the main or panic logs
763 (for which there is an explicit test).
765 PH/26 Added the /no_tell option to "control=freeze".
767 PH/27 If a host name lookup failed very early in a connection, for example, if
768 the IP address matched host_lookup and the reverse lookup yielded a name
769 that did not have a forward lookup, an error message of the form "no IP
770 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
771 could be logged. Now it outputs the IP address instead of "NULL".
773 PH/28 An enabling patch from MH: add new function child_open_exim2() which
774 allows the sender and the authenticated sender to be set when
775 submitting a message from within Exim. Since child_open_exim() is
776 documented for local_scan(), the new function should be too.
778 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
779 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
780 results in an empty string is now treated as unset.
782 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
784 PH/31 Added sender_verify_fail logging option.
786 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
787 needed by RFC 822 but not by RFC 2822 was commented out. I have now
788 tidied the source and removed it altogether.
790 PH/33 When a queue run was abandoned because the load average was too high, a
791 log line was always written; now it is written only if the queue_run log
792 selector is set. In addition, the log line for abandonment now contains
793 information about the queue run such as the pid. This is always present
794 in "start" and "stop" lines but was omitted from the "abandon" line.
796 PH/34 Omit spaces between a header name and the colon in the error message that
797 is given when verify = headers_syntax fails (if there are lots of them,
798 the message gets confusing).
800 PH/35 Change the default for dns_check_names_pattern to allow slashes within
801 names, as there are now some PTR records that contain slashes. This check
802 is only to protect against broken name servers that fall over on strange
803 characters, so the fact that it applies to all lookups doesn't matter.
805 PH/36 Now that the new test suite is complete, we can remove some of the
806 special code in Exim that was needed for the old test suite. For example,
807 sorting DNS records because real resolvers return them in an arbitrary
808 order. The new test suite's fake resolver always returns records in the
811 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
812 bounces) except when queue_only is set, to avoid logging races between
813 the different processes.
815 PH/38 Panic-die if .include specifies a non-absolute path.
817 PH/39 A tweak to the "H" retry rule from its user.
819 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
820 a label. They prevented compilation on older perls.
822 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
823 a warning to be raised on newish perls.
825 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
826 on queue. Changes to match documented behaviour of showing count of
827 messages matching specified criteria.
829 PH/40 Changed the default ident timeout from 30s to 5s.
831 PH/41 Added support for the use of login_cap features, on those BSD systems
832 that have them, for controlling the resources used by pipe deliveries.
834 PH/42 The content-scanning code uses fopen() to create files in which to put
835 message data. Previously it was not paying any attention to the mode of
836 the files. Exim runs with umask(0) because the rest of the code creates
837 files with open(), and sets the required mode explicitly. Thus, these
838 files were ending up world-writeable. This was not a big issue, because,
839 being within the spool directory, they were not world-accessible. I have
840 created a function called modefopen, which takes an additional mode
841 argument. It sets umask(777), creates the file, chmods it to the required
842 mode, then resets the umask. All the relevant calls to fopen() in the
843 content scanning code have been changed to use this function.
845 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
846 to 24 hours. This avoids potential overflow problems when processing G
847 and H retry rules. I suspect nobody ever tinkers with this value.
849 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
851 PH/45 When the plaintext authenticator is running as a client, the server's
852 challenges are checked to ensure they are valid base64 strings. By
853 default, the authentication attempt is cancelled if an invalid string is
854 received. Setting client_ignore_invalid_base64 true ignores these errors.
855 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
856 they are received. Thus, the responses can be made to depend on the
857 challenges. If an invalid string is ignored, an empty string is placed in
860 PH/46 Messages that are created by the autoreply transport now contains a
861 References: header, in accordance with RFCs 2822 and 3834.
863 PH/47 Added authenticated_sender_force to the smtp transport.
865 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
867 PH/49 Installed latest patch from the Sieve maintainer.
869 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
870 was also set, the appendfile transport was unnecessarily scanning a
871 directory of message files (e.g. for maildir delivery) to find the count
872 of files (along with the size), even though it did not need this
873 information. It now does the scan only if it needs to find either the
874 size of the count of files.
876 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
878 PH/52 Two bugs concerned with error handling when the smtp transport is
881 (i) Exim was not creating retry information for temporary errors given
882 for individual recipients after the DATA command when the smtp transport
883 was used in LMTP mode. This meant that they could be retried too
884 frequently, and not timed out correctly.
886 (ii) Exim was setting the flag that allows error details to be returned
887 for LMTP errors on RCPT commands, but not for LMTP errors for individual
888 recipients that were returned after the DATA command.
890 PH/53 This is related to PH/52, but is more general: for any failing address,
891 when detailed error information was permitted to be returned to the
892 sender, but the error was temporary, then after the final timeout, only
893 "retry timeout exceeded" was returned. Now it returns the full error as
894 well as "retry timeout exceeded".
896 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
897 do this, and (what is worse) MTAs that accept it.
899 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
900 will now be deprecated.
902 PH/56 New os.c-cygwin from the Cygwin maintainer.
904 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
905 formats (previously only available in exim formats via -bpr, -bpru,
906 and -bpra. Now also available in native and exiqgrep formats)
908 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
909 with very large, slow to parse queues
911 JJ/08 exipick: added ! as generic prefix to negate any criteria format
913 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
915 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
916 responses to authentication challenges, though it was showing the
917 challenges; (ii) I've removed the CR characters from the debug output for
920 PH/58 Allow for the insertion of a newline as well as a space when a string
921 is turned into more than one encoded-word during RFC 2047 encoding. The
922 Sieve code now uses this.
924 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
925 data_4xx, lost_connection, tls_required.
927 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
928 message was being sent as an SMTP response.
930 PH/61 Add -l and -k options to exicyclog.
932 PH/62 When verifying, if an address was redirected to one new address, so that
933 verification continued, and the new address failed or deferred after
934 having set something in $address_data, the value of $address_data was not
935 passed back to the ACL. This was different to the case when no
936 redirection occurred. The value is now passed back in both cases.
938 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
939 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
940 use login_cap.h, so on its own it isn't the distinguishing feature. The
941 new name refers directly to the setclassresources() function.
943 PH/65 Added configuration files for NetBSD3.
945 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
947 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
948 is preferred over IPv4.
950 PH/68 The bounce_return_message and bounce_return_body options were not being
951 honoured for bounces generated during the reception of non-SMTP messages.
952 In particular, this applied to messages rejected by the ACL. This bug has
953 been fixed. However, if bounce_return_message is true and bounce_return_
954 body is false, the headers that are returned for a non-SMTP message
955 include only those that have been read before the error was detected.
956 (In the case of an ACL rejection, they have all been read.)
958 PH/69 The HTML version of the specification is now built in a directory called
959 spec_html instead of spec.html, because the latter looks like a path with
960 a MIME-type, and this confuses some software.
962 PH/70 Catch two compiler warnings in sieve.c.
964 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
965 function verify_get_ident() calls ip_connect() to connect a socket, but
966 if the "connect()" function timed out, ip_connect() used to close the
967 socket. However, verify_get_ident() also closes the socket later, and in
968 between Exim writes to the log, which may get opened at this point. When
969 the socket was closed in ip_connect(), the log could get the same file
970 descriptor number as the socket. This naturally causes chaos. The fix is
971 not to close the socket in ip_connect(); the socket should be closed by
972 the function that creates it. There was only one place in the code where
973 this was missing, in the iplookup router, which I don't think anybody now
974 uses, but I've fixed it anyway.
976 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
977 well as to direct DNS lookups. Otherwise the handling of names in host
978 lists is inconsistent and therefore confusing.
984 PH/01 Two changes to the default runtime configuration:
986 (1) Move the checks for relay_from_hosts and authenticated clients from
987 after to before the (commented out) DNS black list checks.
989 (2) Add control=submission to the relay_from_hosts and authenticated
990 clients checks, on the grounds that messages accepted by these
991 statements are most likely to be submissions.
993 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
995 (1) Generate an error if the third argument for the ${prvs expansion is
998 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1001 (3) Reset the variables that are obtained from the first argument of
1002 ${prvscheck and used in the second argument before leaving the code,
1003 because their memory is reclaimed, so using them afterwards may do
1006 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1007 one (it's much easier than Tom thought :-).
1009 (5) Because of (4), we can now allow for the use of $prvscheck_result
1010 inside the third argument.
1012 PH/03 For some reason, the default setting of PATH when running a command from
1013 a pipe transport was just "/usr/bin". I have changed it to
1016 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1017 anything to be listed in the output from -bV.
1019 PH/05 When a filter generated an autoreply, the entire To: header line was
1020 quoted in the delivery log line, like this:
1022 => >A.N.Other <ano@some.domain> <original@ddress> ...
1024 This has been changed so that it extracts the operative address. There
1025 may be more than one such address. If so, they are comma-separated, like
1028 => >ano@some.domain,ona@other.domain <original@ddress> ...
1030 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1031 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1032 not being looked up in the rDNS to get a host name, Exim was showing the
1033 IP address twice in Received: lines, even though the IP addresses were
1034 identical. For example:
1036 Received: from [1.2.3.4] (helo=[1.2.3.4])
1038 However, if the real host name was known, it was omitting the HELO data
1039 if it matched the actual IP address. This has been tidied up so that it
1040 doesn't show the same IP address twice.
1042 PH/07 When both +timestamp and +memory debugging was on, the value given by
1043 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1044 called by the debug printing, thereby overwriting the timestamp buffer.
1045 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1047 PH/08 When the original message was included in an autoreply transport, it
1048 always said "this is a copy of the message, including all the headers",
1049 even if body_only or headers_only was set. It now gives an appropriate
1052 PH/09 Applied a patch from the Sieve maintainer which:
1054 o fixes some comments
1055 o adds the (disabled) notify extension core
1056 o adds some debug output for the result of if/elsif tests
1057 o points to the current vacation draft in the documentation
1058 and documents the missing references header update
1062 o fixes a bug in processing the envelope test (when testing
1063 multiple envelope elements, the last element determinted the
1066 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1067 Electronic Mail") by including:
1069 Auto-submitted: auto-generated
1071 in the messages that it generates (bounce messages and others, such as
1072 warnings). In the case of bounce messages for non-SMTP mesages, there was
1073 also a typo: it was using "Auto_submitted" (underscore instead of
1074 hyphen). Since every message generated by Exim is necessarily in response
1075 to another message, thes have all been changed to:
1077 Auto-Submitted: auto-replied
1079 in accordance with these statements in the RFC:
1081 The auto-replied keyword:
1083 - SHOULD be used on messages sent in direct response to another
1084 message by an automatic process,
1086 - MUST NOT be used on manually-generated messages,
1088 - MAY be used on Delivery Status Notifications (DSNs) and Message
1089 Disposition Notifications (MDNs),
1091 - MUST NOT be used on messages generated by automatic or periodic
1092 processes, except for messages which are automatic responses to
1095 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1096 to the default Received: header definition.
1098 PH/12 Added log selector acl_warn_skipped (default on).
1100 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1101 variables because (a) they are in the wrong storage pool and (b) even if
1102 they were copied, it wouldn't work properly because of the caching.
1104 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1105 checking when decoding. Apparently there are clients that generate
1106 overlong encoded strings. Why am I not surprised?
1108 PH/15 If the first argument of "${if match_address" was not empty, but did not
1109 contain an "@" character, Exim crashed. Now it writes a panic log message
1110 and treats the condition as false.
1112 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1114 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1115 extension "envelope-auth". The code is finished and in agreement with
1116 other implementations, but there is no documentation so far and in fact,
1117 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1118 not changing the active code.
1120 Print executed "if" and "elsif" statements when debugging is used. This
1121 helps a great deal to understand what a filter does.
1123 Document more things not specified clearly in RFC3028. I had all this
1124 sorted out, when out of a sudden new issues came to my mind. Oops."
1126 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1129 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1130 canonical form (as documented). However, after a host name lookup from
1131 the IP address, check_host() was doing a simple string comparison with
1132 addresses acquired from the DNS when checking that the found name did
1133 have the original IP as one of its addresses. Since any found IPv6
1134 addresses are likely to be in abbreviated form, the comparison could
1135 fail. Luckily, there already exists a function for doing the comparison
1136 by converting both addresses to binary, so now that is used instead of
1137 the text comparison.
1139 PH/20 There was another similar case to PH/19, when a complete host name was
1140 given in a host list; looking up its IP address could give an abbreviated
1141 form, whereas the current host's name might or might not be abbreviated.
1142 The same fix has been applied.
1148 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1149 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1152 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1153 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1155 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1157 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1158 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1159 Also, for "Auto-Submitted", treat anything other than "no" as
1160 non-personal, in accordance with RFC 3834. (Previously it treated
1161 anything starting "auto-" as non-personal.)
1163 TF/01 The control=submission/name=... option had a problem with syntax
1164 errors if the name included a slash character. The /name= option
1165 now slurps the rest of the string, so it can include any characters
1166 but it must come last in the list of options (after /sender_retain
1169 PH/05 Some modifications to the interface to the fake nameserver for the new
1177 TK/01 Added the "success_on_redirect" address verification option. See
1178 NewStuff for rationale and an example.
1180 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1182 PH/02 Patch to exigrep to allow it to work on syslog lines.
1184 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1185 fread() to skip over the body file's header line, because in Cygwin the
1186 header line is locked and is inaccessible.
1188 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1189 co-exist for some time) to make it clear that it is the Exim ID that is
1190 referenced, not the Message-ID: header line.
1192 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1193 string_format() function, because snprintf() does not exist on all
1196 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1197 ${stat: expansion item.
1199 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1200 protocol synchronization error", to keep the pedants happy.
1202 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1203 well as for IRIX systems, when gcc is being used. See the host.c source
1206 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1208 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1211 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1212 transport and to the smtp transport in LMTP mode.
1214 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1216 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1217 run a filter in a subprocess. This could lead to confusion in subsequent
1218 lookups in the parent process. There should also be a search_tidyup() at
1219 the end of the subprocess.
1221 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1222 only if the host matched helo_try_verify_hosts, which caused the
1223 verification to occur when the EHLO/HELO command was issued. The ACL just
1224 tested the remembered result. Now, if a previous verification attempt has
1225 not happened, "verify = helo" does it there and then.
1227 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1229 TK/03 Fix log output including CR from clamd.
1231 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1232 exist provoked a memory error which could cause a segfault.
1234 PH/15 Installed PCRE 6.2
1236 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1238 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1239 of the problem. Specifically, suggested +O2 rather than +O1 for the
1242 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1244 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1245 record, the port was not correctly logged when the outgoing_port log
1246 selector was set (it logged the transort's default port).
1248 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1249 fallback_hosts, and "hosts" in the smtp transport.
1251 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1252 host errors such as "Connection refused".
1254 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1255 authentication with radiusclient 0.4.9:
1257 - Error returned from rc_read_config was caught wrongly
1258 - Username/password not passed on to radius server due to wrong length.
1260 The presumption is that some radiusclient API changes for 4.51/PH/17
1261 were not taken care of correctly. The code is still untested by me (my
1262 Linux distribution still has 0.3.2 of radiusclient), but it was
1263 contributed by a Radius user.
1265 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1266 expanding the "port" option of the smtp transport.
1268 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1269 while reading a MIME header. Thanks to Tom Hughes for a patch.
1271 PH/24 Include config.h inside local_scan.h so that configuration settings are
1274 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1275 that in an ACL for RCPT (for example), you can examine exactly what was
1278 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1279 commands, but it was not correctly comparing the address with the actual
1280 client host address. Thus, it would show the EHLO address in Received:
1281 header lines when this was not necessary.
1283 PH/27 Added the % operator to ${eval:}.
1285 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1286 it should be ignoring failures (because with -C, for example, it has lost
1287 privilege). It wasn't ignoring creation failures other than "already
1290 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1293 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1294 because an input line was too long, either on its own, or by virtue of
1295 too many continuations, the temporary file was not being removed, and the
1296 return code was incorrect.
1298 PH/31 Missing "BOOL" in function definition in filtertest.c.
1300 PH/32 Applied Sieve patches from the maintainer.
1302 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1304 PH/33 Added "verify = not_blind".
1306 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1307 Local/Makefile (with some defaults set). These are used in built scripts
1308 such as exicyclog, but they have never been used in the exim_install
1309 script (though there are many overriding facilities there). I have
1310 arranged that the exim_install script now takes note of these two
1313 PH/35 Installed configuration files for Dragonfly.
1315 PH/36 When a locally submitted message by a trusted user did not contain a
1316 From: header, and the sender address was obtained from -f or from an SMTP
1317 MAIL command, and the trusted user did not use -F to supply a sender
1318 name, $originator_name was incorrectly used when constructing a From:
1319 header. Furthermore, $originator_name was used for submission mode
1320 messages from external hosts without From: headers in a similar way,
1321 which is clearly wrong.
1323 PH/37 Added control=suppress_local_fixups.
1325 PH/38 When log_selector = +received_sender was set, and the addition of the
1326 sender made the log line's construction buffer exactly full, or one byte
1327 less than full, an overflow happened when the terminating "\n" was
1330 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1331 when the result of a list match is failure because a DNS lookup failed.
1333 PH/40 RM_COMMAND is now used in the building process.
1335 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1336 the "build-* directories that it finds.
1338 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1339 address in a domain literal was a prefix of an interface address.
1341 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1342 when verifying a sender address, unless rewrite_headers is false.
1344 PH/44 Wrote a long comment about why errors_to addresses are verified as
1345 recipients, not senders.
1347 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1348 the ratelimit ACL was added.
1350 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1352 PH/46 Added extra information about PostgreSQL errors to the error string.
1354 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1355 avoiding the need to install special zones in a real server. This is
1356 backwards compatible; if it can't find the fake resolver, it drops back.
1357 Thus, both old and new test suites can be run.
1359 TF/02 Added util/ratelimit.pl
1361 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1362 clock is set back in time.
1364 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1365 Candler <B.Candler@pobox.com>.
1367 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1368 OK for addresses that are the result of redirections.
1370 PH/48 A number of further additions for the benefit of the new test suite,
1371 including a fake gethostbyname() that interfaces to the fake DNS resolver
1374 TF/06 The fix for widen_domains has also been applied to qualify_single and
1375 search_parents which are the other dnslookup options that can cause
1378 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1381 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1383 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1384 service" if it encountered an MX record with an empty target hostname.
1385 The message is now "an MX or SRV record indicated no SMTP service".
1387 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1388 if the DNS of the sending site is misconfigured. This is quite a
1389 common situation. This change restores the behaviour of treating a
1390 helo verification defer as a failure.
1392 PH/51 If self=fail was set on a router, the bounce message did not include the
1393 actual error message.
1399 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1401 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1402 command itself ended in error, the underlying message about the transport
1403 filter timeout was being overwritten with the pipe command error. Now the
1404 underlying error message should be appended to the second error message.
1406 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1408 PH/02 Reduce the amount of output that "make" produces by default. Full output
1409 can still be requested.
1411 PH/03 The warning log line about a condition test deferring for a "warn" verb
1412 was being output only once per connection, rather than after each
1413 occurrence (because it was using the same function as for successful
1414 "warn" verbs). This seems wrong, so I have changed it.
1416 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1417 it should not have, which might have caused a crash in the right
1418 circumstances, but probably never did.
1420 PH/04 Installed a modified version of Tony Finch's patch to make submission
1421 mode fix the return path as well as the Sender: header line, and to
1422 add a /name= option so that you can make the user's friendly name appear
1425 TF/03 Added the control = fakedefer ACL modifier.
1427 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1428 Mark Lowes for thorough testing.
1430 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1432 TK/03 Merged latest SRS patch from Miles Wilton.
1434 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1435 with the definition in sysexits.h (which is #included earlier).
1436 Fortunately, Exim does not actually use EX_OK. The code used to try to
1437 preserve the sysexits.h value, by assumimg that macro definitions were
1438 scanned for macro replacements. I have been disabused of this notion,
1439 so now the code just undefines EX_OK before #including unistd.h.
1441 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1442 in the smtp transport. When a block could not be written in a single
1443 write() function, the timeout was being re-applied to each part-write.
1444 This seems wrong - if the receiver was accepting one byte at a time it
1445 would take for ever. The timeout is now adjusted when this happens. It
1446 doesn't have to be particularly precise.
1448 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1449 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1451 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1452 without a domain if the check to <postmaster@domain> fails.
1454 SC/01 Eximstats: added -xls and the ability to specify output files
1455 (patch written by Frank Heydlauf).
1457 SC/02 Eximstats: use FileHandles for outputing results.
1459 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1461 SC/04 Eximstats: fixed display of large numbers with -nvr option
1463 SC/05 Eximstats: fixed merging of reports with empty tables.
1465 SC/06 Eximstats: added the -include_original_destination flag
1467 SC/07 Eximstats: removed tabs and trailing whitespace.
1469 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1471 TK/06 MBOX spool code: Add real "From " MBOX separator line
1472 so the .eml file is really in mbox format (even though
1473 most programs do not really care). Patch from Alex Miller.
1475 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1476 The latter is generated from $received_to and is only set if the
1477 message has one envelope recipient. SA can use these headers,
1478 obviously out-of-the-box. Patch from Alex Miller.
1480 PH/08 The ${def test on a variable was returning false if the variable's
1481 value was "0", contrary to what the specification has always said!
1482 The result should be true unless the variable is empty.
1484 PH/09 The syntax error of a character other than { following "${if
1485 def:variable_name" (after optional whitespace) was not being diagnosed.
1486 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1487 accidental colon was present, for example, could give incorrect results.
1489 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1490 result is used (not including appendfile, where other changes are about
1493 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1494 This involved changing a lot of size variables from int to off_t. It
1495 should work with maildirs and everything.
1497 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1498 spamd dying while we are connected to it.
1500 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1503 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1504 function for BDB 4.3.
1506 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1508 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1509 the sources for PCRE have been re-arranged and more files are now
1512 PH/15 The code I had for printing potentially long long variables in PH/11
1513 above was not the best (it lost precision). The length of off_t variables
1514 is now inspected at build time, and an appropriate printing format (%ld
1515 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1516 to be "long long int" or "long int". This is needed for the internal
1517 formatting function string_vformat().
1519 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1520 the configuration file to be ":syslog", then the script "guesses" where
1521 the logs files are, rather than using the compiled in default. In our
1522 case the guess is not the same as the compiled default, so the script
1523 suddenly stopped working when I started to use syslog. The patch checks
1524 to see if log_file_path is "". If so, it attempts to read it from exim
1525 with no configuration file to get the compiled in version, before it
1526 falls back to the previous guessing code."
1528 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1529 implementing BATV in an Exim configuration. See NewStuff for the gory
1532 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1533 Makefile that are specific to HP-UX.
1535 PH/18 If the "use_postmaster" option was set for a recipient callout together
1536 with the "random" option, the postmaster address was used as the MAIL
1537 FROM address for the random test, but not for the subsequent recipient
1538 test. It is now used for both.
1540 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1541 patch removes a few documentation additions to RFC 3028, because the
1542 latest draft now contains them. It adds the new en;ascii-case comparator
1543 and a new error check for 8bit text in MIME parts. Comparator and
1544 require names are now matched exactly. I enabled the subaddress
1545 extension, but it is not well tested yet (read: it works for me)."
1547 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1548 rework some of the code of TK/09 above to avoid the hardwired use of
1549 "%lld" and "long long". Replaced the call to snprintf() with a call to
1552 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1553 records point to non-existent hosts", "retry timeout exceeded", and
1554 "retry time not reached for any host after a long failure period".
1556 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1557 experimental DomainKeys support:
1559 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1560 (2) On an error such as an illegally used "control", the wrong name for
1561 the control was given.
1563 These problems did NOT occur unless DomainKeys support was compiled.
1565 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1567 PH/24 Added ${if match_ip condition.
1569 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1570 databases so that it will be absolutely obvious if a crash occurs in the
1571 DB library. This is a regular occurrence (often caused by mis-matched
1574 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1575 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1576 picked up on a user's system that detects such things. There doesn't seem
1577 to be a gcc warning option for this - only an attribute that has to be
1578 put on the function's prototype. It seems that in Fedora Core 4 they have
1579 set this on a number of new functions. No doubt there will be more in due
1582 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1583 specify a transport. However, if an address that was verified by such a
1584 router was the subject of a callout, Exim crashed because it tried to
1585 read the rcpt_include_affixes from the non-existent transport. Now it
1586 just assumes that the setting of that option is false. This bug was
1587 introduced by 4.51/PH/31.
1589 PH/28 Changed -d+all to exclude +memory, because that information is very
1590 rarely of interest, but it makes the output a lot bigger. People tend to
1591 do -d+all out of habit.
1593 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1594 code in os-type was giving problems when libc.so lives in lib64, like on
1597 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1598 aren't the modern standard, and it seems that some systems' include files
1599 don't always have them. Exim was already checking for some of the newer
1600 ones like T_AAAA, and defining it itself. I've added checks for all the
1601 record types that Exim uses.
1603 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1604 not automatically generating a new one, as it is supposed to. This
1605 prevented TLS from working. If the file did exist, but contained invalid
1606 data, a new version was generated, as expected. It was only the case of a
1607 non-existent file that was broken.
1609 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1610 with a change in libdomainkeys > 0.64.
1612 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1613 from DNS. If the selector record carries the flag, it now has
1614 precedence over the domain-wide flag.
1616 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1618 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1619 the use of a port name that isn't defined in /etc/services) occurred, the
1620 message was deferred as in a normal delivery, and thus remained on the
1621 spool, instead of being failed because of the mua_wrapper setting. This
1622 is now fixed, and I tidied up some of the mua_wrapper messages at the
1625 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1626 the messages in a hash of arrays rather than using individual hashes.
1627 This is a bit cleaner and results in dramatic memory savings, albeit
1628 at a slight CPU cost.
1630 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1631 as requested by Marc Sherman.
1633 SC/10 Eximstats: added histograms for user specified patterns as requested
1636 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1638 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1639 fopen() in the content-scanning modules that did not already have it.
1645 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1646 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1648 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1650 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1652 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1653 to test Sieve filters that use "vacation".
1655 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1656 that changes the way the GnuTLS parameters are stored in the cache file.
1657 The new format can be generated externally. For backward compatibility,
1658 if the data in the cache doesn't make sense, Exim assumes it has read an
1659 old-format file, and it generates new data and writes a new file. This
1660 means that you can't go back to an older release without removing the
1663 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1664 work if there are any delivery delays because "one_time" forces the
1665 parent to be marked "delivered", so its unseen clone is never tried
1666 again. For this reason, Exim now forbids the simultaneous setting of
1669 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1670 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1671 entry if you want to know the details. The fix, however, neglected to
1672 consider the case when local delivery batching is involved. The test for
1673 "previously delivered" was not happening when checking to see if an
1674 address could be batched with a previous (undelivered) one; under
1675 certain circumstances this could lead to multiple deliveries to the same
1678 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1679 in its include files, and this causes problems building Exim.
1681 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1682 header_syntax) but Exim was just ignoring anything given after a slash.
1683 In particular, this caused confusion with an attempt to use "verify =
1684 reverse_host_lookup/defer_ok". An error is now given when options are
1685 supplied for verify items that do not have them. (Maybe reverse_host_
1686 lookup should have a defer_ok option, but that's a different point.)
1688 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1689 defined by RFC 821) to 2048, because there were problems with some AUTH
1690 commands, and RFC 1869 says the size should be increased for extended
1691 SMTP commands that take arguments.
1693 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1696 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1697 "unknown" error; now it says that the functionality isn't in the binary.
1699 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1700 an address' error message when a string expansion fails (syntax or
1701 whatever). Otherwise the password may appear in the log. Following change
1702 PH/42 below, there is no longer a chance of it appearing in a bounce
1705 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1707 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1708 of its addresses was ever tried. (Bugzilla bug #2.)
1710 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1711 the result incorrectly in the debug output. (It correctly added a newline
1712 to what was transported.)
1714 TF/01 Added $received_time.
1716 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1717 commented out examples of how to interface to a virus scanner and to
1718 SpamAssassin. Also added commented examples of av_scanner and
1719 spamd_address settings.
1721 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1722 and controls are allowed in which ACLs. There were a couple of minor
1723 errors. Some of the entries in the conditions table (which is a table of
1724 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1725 negation of where the condition IS allowed.
1727 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1729 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1730 header file does not have a version number, so I've had to invent a new
1731 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1732 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1733 radiusclient), but it was contributed by a Radius user.
1735 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1736 files or directories, for OpenSSL.
1738 PH/19 When an Exim process that is running as root has to create an Exim log
1739 file, it does so in a subprocess that runs as exim:exim so as to get the
1740 ownership right at creation (otherwise, other Exim processes might see
1741 the file with the wrong ownership). There was no test for failure of this
1742 fork() call, which would lead to the process getting stuck as it waited
1743 for a non-existent subprocess. Forks do occasionally fail when resources
1744 run out. I reviewed all the other calls to fork(); they all seem to check
1747 PH/20 When checking for unexpected SMTP input at connect time (before writing
1748 the banner), Exim was not dealing correctly with a non-positive return
1749 from the read() function. If the client had disconnected by this time,
1750 the result was a log entry for a synchronization error with an empty
1751 string after "input=" when read() returned zero. If read() returned -1
1752 (an event I could not check), uninitialized data bytes were printed.
1753 There were reports of junk text (parts of files, etc) appearing after
1756 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1758 PH/22 Added support for macro redefinition, and (re)definition in between
1759 driver and ACL definitions.
1761 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1762 forgetting to use the resulting value; it was using the unexpanded value.
1764 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1765 hadn't been configured. The fix is from Juergen Kreileder, who
1766 understands it better than I do:
1768 "Here's what I see happening with three configured cyrus_sasl
1769 authenticators configured (plain, login, cram-md5):
1771 On startup auth_cyrus_sasl_init() gets called for each of these.
1772 This means three calls to sasl_listmech() without a specified mech_list.
1773 => SASL tests which mechs of all available mechs actually work
1774 => three warnings about OTP not working
1775 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1777 With the patch, sasl_listmech() also gets called three times. But now
1778 SASL's mech_list option is set to the server_mech specified in the the
1779 authenticator. Or in other words, the answer from sasl_listmech()
1780 gets limited to just the mech you're testing for (which is different
1782 => the return list contains just 'plain' or 'login', 'cram-md5' or
1783 nothing depending on the value of ob->server_mech.
1785 I've just tested the patch: Authentication still works fine,
1786 unavailable mechs specified in the exim configuration are still
1787 caught, and the auth.log warnings about OTP are gone."
1789 PH/25 When debugging is enabled, the contents of the command line are added
1790 to the debugging output, even when log_selector=+arguments is not
1793 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1794 answer is "GNU", and only if the return is "GNU/something" is the answer
1797 PH/27 $acl_verify_message is now set immediately after the failure of a
1798 verification in an ACL, and so is available in subsequent modifiers. In
1799 particular, the message can be preserved by coding like this:
1801 warn !verify = sender
1802 set acl_m0 = $acl_verify_message
1804 Previously, $acl_verify_message was set only while expanding "message"
1805 and "log_message" when a very denied access.
1807 PH/28 Modified OS/os.c-Linux with
1809 -#ifndef OS_LOAD_AVERAGE
1810 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1812 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1813 nomenclature these days.)
1815 PH/29 Installed patch from the Sieve maintainer that adds the options
1816 sieve_useraddress and sieve_subaddress to the redirect router.
1818 PH/30 In these circumstances:
1819 . Two addresses routed to the same list of hosts;
1820 . First host does not offer TLS;
1821 . First host accepts first address;
1822 . First host gives temporary error to second address;
1823 . Second host offers TLS and a TLS session is established;
1824 . Second host accepts second address.
1825 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1826 and peerdn, if requested) that were in fact used only for the second
1829 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1830 attention to any local part prefix or suffix that was matched by the
1831 router that accepted the address. It now behaves in the same way as it
1832 does for delivery: the affixes are removed from the local part unless
1833 rcpt_include_affixes is set on the transport.
1835 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1836 timeout during the DATA phase of an incoming message.
1838 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1839 applied a patch sanctioned by the Sieve maintainer.
1841 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1842 the uid or gid is negative. A case of a negative gid caused this to be
1843 noticed. The fix allows for either to be negative.
1845 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1846 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1849 PH/36 The change PH/12 above was broken. Fixed it.
1851 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1852 the grounds that routing the same address twice would always produce the
1853 same answer. This might have been true once, but it is certainly no
1854 longer true now. Routing a child address may depend on the previous
1855 routing that produced that child. Some complicated redirection strategies
1856 went wrong when messages had multiple recipients, and made Exim's
1857 behaviour dependent on the order in which the addresses were given.
1859 I have moved the duplicate checking until after the routing is complete.
1860 Exim scans the addresses that are assigned to local and remote
1861 transports, and removes any duplicates. This means that more work will be
1862 done, as duplicates will always all be routed, but duplicates are
1863 presumably rare, so I don't expect this is of any significance.
1865 For deliveries to pipes, files, and autoreplies, the duplicate checking
1866 still happens during the routing process, since they are not going to be
1869 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1870 It corrects a timeout issue with spamd. This is Ian's comment: "The
1871 background is that sometimes spamd either never reads data from a
1872 connection it has accepted, or it never writes response data. The exiscan
1873 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1874 blindly assumes that writes won't block so it may never time out."
1876 PH/39 Allow G after quota size as well as K and M.
1878 PH/40 The value set for $authenticated_id in an authenticator may not contain
1879 binary zeroes or newlines because the value is written to log lines and
1880 to spool files. There was no check on this. Now the value is run through
1881 the string_printing() function so that such characters are converted to
1882 printable escape sequences.
1884 PH/41 $message_linecount is a new variable that contains the total number of
1885 lines in the message. Compare $body_linecount, which is the count for the
1888 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1889 bounce and delay warning messages, except in certain special cases, which
1892 (a) An SMTP error message from a remote host;
1893 (b) A message specified in a :fail: redirection;
1894 (c) A message specified in a "fail" command in a system filter;
1895 (d) A message specified in a FAIL return from the queryprogram router;
1896 (e) A message specified by the cannot_route_message router option.
1898 In these cases only, Exim does include the error details in bounce and
1899 warning messages. There are also a few cases where bland messages such
1900 as "unrouteable address" or "local delivery error" are given.
1902 PH/43 $value is now also set for the "else" part of a ${run expansion.
1904 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1905 being worked on, but at least Exim now implements the latest version to
1908 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1909 process to complete was treated as a delivery failure, a timeout while
1910 writing the message to the pipe was logged, but erroneously treated as a
1911 successful delivery. Such timeouts include transport filter timeouts. For
1912 consistency with the overall process timeout, these timeouts are now
1913 treated as errors, giving rise to delivery failures by default. However,
1914 there is now a new Boolean option for the pipe transport called
1915 timeout_defer, which, if set TRUE, converts the failures into defers for
1916 both kinds of timeout. A transport filter timeout is now identified in
1919 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1920 systems where "make" and "gmake" are different, calling "gmake" at top
1921 level broke things. I've arranged for the value of $(MAKE) to be passed
1922 from the Makefile to this script so that it can call the same version of
1926 A note about Exim versions 4.44 and 4.50
1927 ----------------------------------------
1929 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1930 changes of various kinds. As a consequence, a big documentation update was
1931 needed. This delayed the release for rather longer than seemed good, especially
1932 in the light of a couple of (minor) security issues. Therefore, the changes
1933 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1934 release. So 4.44 and 4.50 are in effect two different branches that both start
1937 I have left the 4.50 change log unchanged; it contains all the changes since
1938 4.43. The change log for 4.44 is below; many of its items are identical to
1939 those for 4.50. This seems to be the most sensible way to preserve the
1940 historical information.
1946 1. Minor wording change to the doc/README.SIEVE file.
1948 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1949 computation of the current number of files was incorrect.
1951 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1952 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1955 4. Give more explanation in the error message when the command for a transport
1956 filter fails to execute.
1958 5. There are several places where Exim runs a non-Exim command in a
1959 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1960 was being done only for the command run by the queryprogram router. It is
1961 now done for all such subprocesses. The other cases are: ${run, transport
1962 filters, and the commands run by the lmtp and pipe transports.
1964 6. Added CONFIGURE_GROUP build-time option.
1966 7. Some older OS have a limit of 256 on the maximum number of file
1967 descriptors. Exim was using setrlimit() to set 1000 as a large value
1968 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1969 systems. I've change it so that if it can't get 1000, it tries for 256.
1971 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1972 was an oversight, and furthermore, ever since the addition of extra
1973 controls (e.g. 4.43/32), the checks on when to allow different forms of
1974 "control" were broken. There should now be diagnostics for all cases when a
1975 control that does not make sense is encountered.
1977 9. Added the /retain_sender option to "control=submission".
1979 10. $recipients is now available in the predata ACL (oversight).
1981 11. Tidy the search cache before the fork to do a delivery from a message
1982 received from the command line. Otherwise the child will trigger a lookup
1983 failure and thereby defer the delivery if it tries to use (for example) a
1984 cached ldap connection that the parent has called unbind on.
1986 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1987 of $address_data from the recipient verification was clobbered by the
1988 sender verification.
1990 13. The value of address_data from a sender verification is now available in
1991 $sender_address_data in subsequent conditions in the ACL statement.
1993 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1995 15. Added a new option "connect=<time>" to callout options, to set a different
1998 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1999 was its contents. (It was OK if the option was not defined at all.)
2001 17. A "Completed" log line is now written for messages that are removed from
2002 the spool by the -Mrm option.
2004 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2005 information about exactly what failed.
2007 19. Added -dd to debug only the daemon process.
2009 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2010 handles timeouts, both on the server side and network timeouts. Renamed the
2011 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2013 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2014 It is now set to "smtps".
2016 22. $host_address is now set to the target address during the checking of
2017 ignore_target_hosts.
2019 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2020 being passed; this would have caused $sender_host_name to have been used if
2021 matching the list had actually called for a host name (not very likely,
2022 since this list is usually IP addresses). A host name is now passed as
2025 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2026 code that specifies a non-restarting handler (typically sigaction() in
2027 modern systems) in an attempt to fix a rare and obscure crash bug.
2029 25. Narrowed the window for a race in the daemon that could cause it to ignore
2030 SIGCHLD signals. This is not a major problem, because they are used only to
2031 wake it up if nothing else does.
2033 26. A malformed maildirsize file could cause Exim to calculate negative values
2034 for the mailbox size or file count. Odd effects could occur as a result.
2035 The maildirsize information is now recalculated if the size or filecount
2038 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2039 support for a long time. Removed HAVE_SYS_VFS_H.
2041 28. Installed the latest version of exipick from John Jetmore.
2043 29. In an address list, if the pattern was not a regular expression, an empty
2044 subject address (from a bounce message) matched only if the pattern was an
2045 empty string. Non-empty patterns were not even tested. This was the wrong
2046 because it is perfectly reasonable to use an empty address as part of a
2047 database query. An empty address is now tested by patterns that are
2048 lookups. However, all the other forms of pattern expect the subject to
2049 contain a local part and a domain, and therefore, for them, an empty
2050 address still always fails if the pattern is not itself empty.
2052 30. Exim went into a mad DNS loop when attempting to do a callout where the
2053 host was specified on an smtp transport, and looking it up yielded more
2054 than one IP address.
2056 31. Re-factored the code for checking spool and log partition space into a
2057 function that finds that data and another that does the check. The former
2058 is then used to implement four new variables: $spool_space, $log_space,
2059 $spool_inodes, and $log_inodes.
2061 32. The RFC2047 encoding function was originally intended for short strings
2062 such as real names; it was not keeping to the 75-character limit for
2063 encoded words that the RFC imposes. It now respects the limit, and
2064 generates multiple encoded words if necessary. To be on the safe side, I
2065 have increased the buffer size for the ${rfc2047: expansion operator from
2068 33. It is now permitted to omit both strings after an "if" condition; if the
2069 condition is true, the result is "true". As before, when the second string
2070 is omitted, a false condition yields an empty string. This makes it less
2071 cumbersome to write custom ACL and router conditions.
2073 34. Failure to deliver a bounce message always caused it to be frozen, even if
2074 there was an errors_to setting on the router. The errors_to setting is now
2077 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2078 canonical form (fully expanded) before being placed in
2079 $sender_host_address.
2081 36. The table in the code that translates DNS record types into text (T_A to
2082 "A" for instance) was missing entries for NS and CNAME. It is just possible
2083 that this could have caused confusion if both these types were looked up
2084 for the same domain, because the text type is used as part of Exim's
2085 per-process caching. But the chance of anyone hitting this buglet seems
2088 37. The dnsdb lookup has been extended in a number of ways.
2090 (1) There is a new type, "zns", which walks up the domain tree until it
2091 finds some nameserver records. It should be used with care.
2093 (2) There is a new type, "mxh", which is like "mx" except that it returns
2094 just the host names, not the priorities.
2096 (3) It is now possible to give a list of domains (or IP addresses) to be
2097 looked up. The behaviour when one of the lookups defers can be
2098 controlled by a keyword.
2100 (4) It is now possible to specify the separator character for use when
2101 multiple records are returned.
2103 38. The dnslists ACL condition has been extended: it is now possible to supply
2104 a list of IP addresses and/or domains to be looked up in a particular DNS
2107 39. Added log_selector=+queue_time_overall.
2109 40. When running the queue in the test harness, wait just a tad after forking a
2110 delivery process, to get repeatability of debugging output.
2112 41. Include certificate and key file names in error message when GnuTLS fails
2113 to set them up, because the GnuTLS error message doesn't include the name
2114 of the failing file when there is a problem reading it.
2116 42. Allow both -bf and -bF in the same test run.
2118 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2120 44. The "Exiscan patch" is now merged into the mainline Exim source.
2122 45. Sometimes the final signoff response after QUIT could fail to get
2123 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2124 < 0 before doing a fflush(). This bug looks as though it goes back to the
2125 introduction of TLS in release 3.20, but "sometimes" must have been rare
2126 because the tests only now provoked it.
2128 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2129 (this can affect the format of dates).
2131 47. exim_tidydb, when checking for the continued existence of a message for
2132 which it has found a message-specific retry record, was not finding
2133 messages that were in split spool directories. Consequently, it was
2134 deleting retry records that should have stayed in existence.
2136 48. Steve fixed some bugs in eximstats.
2138 49. The SPA authentication driver was not abandoning authentication and moving
2139 on to the next authenticator when an expansion was forced to fail,
2140 contradicting the general specification for all authenticators. Instead it
2141 was generating a temporary error. It now behaves as specified.
2143 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2144 (the order specifies the preference for clients). The order is now AES256,
2145 AES128, 3DES, ARCFOUR128.
2147 51. Small patch to Sieve code - explicitly set From: when generating an
2150 52. Exim crashed if a remote delivery caused a very long error message to be
2151 recorded - for instance if somebody sent an entire SpamAssassin report back
2152 as a large number of 550 error lines. This bug was coincidentally fixed by
2153 increasing the size of one of Exim's internal buffers (big_buffer) that
2154 happened as part of the Exiscan merge. However, to be on the safe side, I
2155 have made the code more robust (and fixed the comments that describe what
2158 53. Now that there can be additional text after "Completed" in log lines (if
2159 the queue_time_overall log selector is set), a one-byte patch to exigrep
2160 was needed to allow it to recognize "Completed" as not the last thing in
2163 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2164 patch that reportedly fixes this has been added. I am not expert enough to
2165 create a test for it. This is what the patch creator wrote:
2167 "I found a little strange behaviour of ldap code when working with
2168 Windows 2003 AD Domain, where users was placed in more than one
2169 Organization Units. When I tried to give exim partial DN, the exit code
2170 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2171 But simultaneously result of request was absolutely normal ldap result,
2172 so I produce this patch..."
2174 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2175 REFERENCE, so I have modified the code to exclude the patch when that macro
2178 55. Some experimental protocols are using DNS PTR records for new purposes. The
2179 keys for these records are domain names, not reversed IP addresses. The
2180 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2181 leaves it alone. Component reversal etc. now happens only for IP addresses.
2184 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2186 57. Double the size of the debug message buffer (to 2048) so that more of very
2187 long debug lines gets shown.
2189 58. The exicyclog utility now does better if the number of log files to keep
2190 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2192 59. Two changes related to the smtp_active_hostname option:
2194 (1) $smtp_active_hostname is now available as a variable.
2195 (2) The default for smtp_banner uses $smtp_active_hostname instead
2196 of $primary_hostname.
2198 60. The host_aton() function is supposed to be passed a string that is known
2199 to be a valid IP address. However, in the case of IPv6 addresses, it was
2200 not checking this. This is a hostage to fortune. Exim now panics and dies
2201 if the condition is not met. A case was found where this could be provoked
2202 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2203 components; fortuitously, this particular loophole had already been fixed
2204 by change 4.50/55 above.
2206 If there are any other similar loopholes, the new check in host_aton()
2207 itself should stop them being exploited. The report I received stated that
2208 data on the command line could provoke the exploit when Exim was running as
2209 exim, but did not say which command line option was involved. All I could
2210 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2211 running as the user.
2214 61. There was a buffer overflow vulnerability in the SPA authentication code
2215 (which came originally from the Samba project). I have added a test to the
2216 spa_base64_to_bits() function which I hope fixes it.
2219 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2220 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2222 63. The daemon start-up calls getloadavg() while still root for those OS that
2223 need the first call to be done as root, but it missed one case: when
2224 deliver_queue_load_max is set with deliver_drop_privilege. This is
2225 necessary for the benefit of the queue runner, because there is no re-exec
2226 when deliver_drop_privilege is set.
2228 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2229 This has been fixed.
2231 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2232 was in use, was not putting the data itself into the right store pool;
2233 consequently, it could be overwritten for a subsequent message in the same
2234 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2237 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2239 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2240 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2241 the function were treating the return as a boolean value, which happened to
2242 work because 0=false and not-0=true, but is not correct code.
2244 68. The host_aton() function was not handling scoped IPv6 addresses (those
2245 with, for example, "%eth0" on the end) correctly.
2247 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2248 negated items (that is, ~something) in unsigned ints. Some compilers
2249 apparently mutter when there is no cast.
2251 70. If an address verification called from an ACL failed, and did not produce a
2252 user-specific message (i.e. there was only a "system" message), nothing was
2253 put in $acl_verify_message. In this situation, it now puts the system
2256 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2257 session; change 4.31/43 added the unwanted input to the log line - except
2258 that it did not do this in the start of session case. It now does.
2260 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2261 This is wrong and can cause the other end to generate a synchronization
2262 error if it is another Exim or anything else that does the synchronization
2263 check. A QUIT command is no longer sent after a timeout.
2265 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2266 during host lookups.
2268 74. The defer_ok option of callout verification was not working if it was used
2269 when verifying addresses in header lines, that is, for this case:
2271 verify = header_sender/callout=defer_ok
2273 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2274 those file descriptors could be used for SMTP connections. If anything
2275 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2276 could be sent to the SMTP client, causing chaos. The daemon now opens
2277 stdin, stdout, and stderr to /dev/null when it puts itself into the
2280 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2281 log by default. The user can override this with suitable Perl magic.
2283 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2284 the log message when discard triggers, was not working for the DATA ACL or
2285 for the non-SMTP ACL.
2287 78. Error message wording change in sieve.c.
2289 79. If smtp_accept_max_per_host was set, the number of connections could be
2290 restricted to fewer than expected, because the daemon was trying to set up
2291 a new connection before checking whether the processes handling previous
2292 connections had finished. The check for completed processes is now done
2293 earlier. On busy systems, this bug wouldn't be noticed because something
2294 else would have woken the daemon, and it would have reaped the completed
2297 80. If a message was submitted locally by a user whose login name contained one
2298 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2299 It caused a spool format error. I have fixed the spool reading code. A
2300 related problem was that the "from" clause in the Received: line became
2301 illegal because of the space(s). It is now covered by ${quote_local_part.
2303 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2306 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2308 83. Patch from Sieve maintainer to fix unterminated string problem in
2309 "vacation" handling.
2311 84. Some minor changes to the Linux configuration files to help with other
2312 OS variants using glibc.
2314 85. One more patch for Sieve to update vacation handling to latest spec.
2317 ----------------------------------------------------
2318 See the note above about the 4.44 and 4.50 releases.
2319 ----------------------------------------------------
2325 1. Change 4.43/35 introduced a bug that caused file counts to be
2326 incorrectly computed when quota_filecount was set in an appendfile
2329 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2330 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2333 3. Give more explanation in the error message when the command for a transport
2334 filter fails to execute.
2336 4. There are several places where Exim runs a non-Exim command in a
2337 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2338 was being done only for the command run by the queryprogram router. It is
2339 now done for all such subprocesses. The other cases are: ${run, transport
2340 filters, and the commands run by the lmtp and pipe transports.
2342 5. Some older OS have a limit of 256 on the maximum number of file
2343 descriptors. Exim was using setrlimit() to set 1000 as a large value
2344 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2345 systems. I've change it so that if it can't get 1000, it tries for 256.
2347 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2348 was an oversight, and furthermore, ever since the addition of extra
2349 controls (e.g. 4.43/32), the checks on when to allow different forms of
2350 "control" were broken. There should now be diagnostics for all cases when a
2351 control that does not make sense is encountered.
2353 7. $recipients is now available in the predata ACL (oversight).
2355 8. Tidy the search cache before the fork to do a delivery from a message
2356 received from the command line. Otherwise the child will trigger a lookup
2357 failure and thereby defer the delivery if it tries to use (for example) a
2358 cached ldap connection that the parent has called unbind on.
2360 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2361 of $address_data from the recipient verification was clobbered by the
2362 sender verification.
2364 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2365 was its contents. (It was OK if the option was not defined at all.)
2367 11. A "Completed" log line is now written for messages that are removed from
2368 the spool by the -Mrm option.
2370 12. $host_address is now set to the target address during the checking of
2371 ignore_target_hosts.
2373 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2374 being passed; this would have caused $sender_host_name to have been used if
2375 matching the list had actually called for a host name (not very likely,
2376 since this list is usually IP addresses). A host name is now passed as
2379 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2380 code that specifies a non-restarting handler (typically sigaction() in
2381 modern systems) in an attempt to fix a rare and obscure crash bug.
2383 15. Narrowed the window for a race in the daemon that could cause it to ignore
2384 SIGCHLD signals. This is not a major problem, because they are used only to
2385 wake it up if nothing else does.
2387 16. A malformed maildirsize file could cause Exim to calculate negative values
2388 for the mailbox size or file count. Odd effects could occur as a result.
2389 The maildirsize information is now recalculated if the size or filecount
2392 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2393 support for a long time. Removed HAVE_SYS_VFS_H.
2395 18. Updated exipick to current release from John Jetmore.
2397 19. Allow an empty sender to be matched against a lookup in an address list.
2398 Previously the only cases considered were a regular expression, or an
2401 20. Exim went into a mad DNS lookup loop when doing a callout where the
2402 host was specified on the transport, if the DNS lookup yielded more than
2405 21. The RFC2047 encoding function was originally intended for short strings
2406 such as real names; it was not keeping to the 75-character limit for
2407 encoded words that the RFC imposes. It now respects the limit, and
2408 generates multiple encoded words if necessary. To be on the safe side, I
2409 have increased the buffer size for the ${rfc2047: expansion operator from
2412 22. Failure to deliver a bounce message always caused it to be frozen, even if
2413 there was an errors_to setting on the router. The errors_to setting is now
2416 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2417 canonical form (fully expanded) before being placed in
2418 $sender_host_address.
2420 24. Updated eximstats to version 1.33
2422 25. Include certificate and key file names in error message when GnuTLS fails
2423 to set them up, because the GnuTLS error message doesn't include the name
2424 of the failing file when there is a problem reading it.
2426 26. Expand error message when OpenSSL has problems setting up cert/key files.
2429 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2430 (this can affect the format of dates).
2432 28. exim_tidydb, when checking for the continued existence of a message for
2433 which it has found a message-specific retry record, was not finding
2434 messages that were in split spool directories. Consequently, it was
2435 deleting retry records that should have stayed in existence.
2437 29. eximstats updated to version 1.35
2438 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2439 1.35 - bugfix such that pie charts by volume are generated correctly
2441 30. The SPA authentication driver was not abandoning authentication and moving
2442 on to the next authenticator when an expansion was forced to fail,
2443 contradicting the general specification for all authenticators. Instead it
2444 was generating a temporary error. It now behaves as specified.
2446 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2447 (the order specifies the preference for clients). The order is now AES256,
2448 AES128, 3DES, ARCFOUR128.
2450 31. Small patch to Sieve code - explicitly set From: when generating an
2453 32. Exim crashed if a remote delivery caused a very long error message to be
2454 recorded - for instance if somebody sent an entire SpamAssassin report back
2455 as a large number of 550 error lines. This bug was coincidentally fixed by
2456 increasing the size of one of Exim's internal buffers (big_buffer) that
2457 happened as part of the Exiscan merge. However, to be on the safe side, I
2458 have made the code more robust (and fixed the comments that describe what
2461 33. Some experimental protocols are using DNS PTR records for new purposes. The
2462 keys for these records are domain names, not reversed IP addresses. The
2463 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2464 leaves it alone. Component reversal etc. now happens only for IP addresses.
2467 34. The host_aton() function is supposed to be passed a string that is known
2468 to be a valid IP address. However, in the case of IPv6 addresses, it was
2469 not checking this. This is a hostage to fortune. Exim now panics and dies
2470 if the condition is not met. A case was found where this could be provoked
2471 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2472 components; fortuitously, this particular loophole had already been fixed
2473 by change 4.50/55 or 4.44/33 above.
2475 If there are any other similar loopholes, the new check in host_aton()
2476 itself should stop them being exploited. The report I received stated that
2477 data on the command line could provoke the exploit when Exim was running as
2478 exim, but did not say which command line option was involved. All I could
2479 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2480 running as the user.
2483 35. There was a buffer overflow vulnerability in the SPA authentication code
2484 (which came originally from the Samba project). I have added a test to the
2485 spa_base64_to_bits() function which I hope fixes it.
2488 36. The daemon start-up calls getloadavg() while still root for those OS that
2489 need the first call to be done as root, but it missed one case: when
2490 deliver_queue_load_max is set with deliver_drop_privilege. This is
2491 necessary for the benefit of the queue runner, because there is no re-exec
2492 when deliver_drop_privilege is set.
2494 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2495 was in use, was not putting the data itself into the right store pool;
2496 consequently, it could be overwritten for a subsequent message in the same
2497 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2500 38. Sometimes the final signoff response after QUIT could fail to get
2501 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2502 < 0 before doing a fflush(). This bug looks as though it goes back to the
2503 introduction of TLS in release 3.20, but "sometimes" must have been rare
2504 because the tests only now provoked it.
2510 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2511 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2512 Now it returns an integer. A number of calls were still expecting a T/F
2513 return. Fortuitously, in all cases, the tests worked in OK situations,
2514 which is the norm. However, things would have gone wrong on any write
2515 failures on the smtp file descriptor. This function is used when sending
2516 messages over SMTP and also when doing verify callouts.
2518 2. When Exim is called to do synchronous delivery of a locally submitted
2519 message (the -odf or -odi options), it no longer closes stderr before doing
2522 3. Implemented the mua_wrapper option.
2524 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2526 5. Implemented the functions header_remove(), header_testname(),
2527 header_add_at_position(), and receive_remove_recipient(), and exported them
2530 6. If an ACL "warn" statement specified the addition of headers, Exim already
2531 inserted X-ACL-Warn: at the start if there was no header name. However, it
2532 was not making this test for the second and subsequent header lines if
2533 there were newlines in the string. This meant that an invalid header could
2534 be inserted if Exim was badly configured.
2536 7. Allow an ACL "warn" statement to add header lines at the start or after all
2537 the Received: headers, as well as at the end.
2539 8. Added the rcpt_4xx retry error code.
2541 9. Added postmaster_mailfrom=xxx to callout verification option.
2543 10. Added mailfrom=xxxx to the callout verification option, for verify=
2546 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2547 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2549 12. Inserted some casts to stop certain compilers warning when using pointer
2550 differences as field lengths or precisions in printf-type calls (mostly
2551 affecting debugging statements).
2553 13. Added optional readline() support for -be (dynamically loaded).
2555 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2556 same clock tick as a message's arrival, so that its received time was the
2557 same as the "first fail" time on the retry record, and that message
2558 remained on the queue past the ultimate address timeout, every queue runner
2559 would try a delivery (because it was past the ultimate address timeout) but
2560 after another failure, the ultimate address timeout, which should have then
2561 bounced the address, did not kick in. This was a "< instead of <=" error;
2562 in most cases the first failure would have been in the next clock tick
2563 after the received time, and all would be well.
2565 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2566 being recognized when the domain list was tested by the match_domain
2567 condition in an expansion string.
2569 16. Added the ${str2b64: operator.
2571 17. Exim was always calling setrlimit() to set a large limit for the number of
2572 processes, without checking whether the existing limit was already
2573 adequate. (It did check for the limit on file descriptors.) Furthermore,
2574 errors from getrlimit() and setrlimit() were being ignored. Now they are
2575 logged to the main and panic logs, but Exim does carry on, to try to do its
2576 job under whatever limits there are.
2578 18. Imported PCRE 5.0.
2580 19. Trivial typo in log message " temporarily refused connection" (the leading
2583 20. If the log selector return_path_on_delivery was set and an address was
2584 redirected to /dev/null, the delivery process crashed because it assumed
2585 that a return path would always be set for a "successful" delivery. In this
2586 case, the whole delivery is bypassed as an optimization, and therefore no
2589 21. Internal re-arrangement: the function for sending a challenge and reading
2590 a response while authentication was assuming a zero-terminated challenge
2591 string. It's now changed to take a pointer and a length, to allow for
2592 binary data in such strings.
2594 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2596 23. Exim was not respecting finduser_retries when seeking the login of the
2597 uid under which it was called; it was always trying 10 times. (The default
2598 setting of finduser_retries is zero.) Also, it was sleeping after the final
2599 failure, which is pointless.
2601 24. Implemented tls_on_connect_ports.
2603 25. Implemented acl_smtp_predata.
2605 26. If the domain in control=submission is set empty, Exim assumes that the
2606 authenticated id is a complete email address when it generates From: or
2607 Sender: header lines.
2609 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2610 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2611 chown and chgrp in /bin and hostname in /usr/bin.
2613 28. Exim was keeping the "process log" file open after each use, just as it
2614 does for the main log. This opens the possibility of it remaining open for
2615 long periods when the USR1 signal hits a daemon. Occasional processlog
2616 errors were reported, that could have been caused by this. Anyway, it seems
2617 much more sensible not to leave this file open at all, so that is what now
2620 29. The long-running daemon process does not normally write to the log once it
2621 has entered its main loop, and it closes the log before doing so. This is
2622 so that log files can straightforwardly be renamed and moved. However,
2623 there are a couple of unusual error situations where the daemon does write
2624 log entries, and I had neglected to close the log afterwards.
2626 30. The text of an SMTP error response that was received during a remote
2627 delivery was being truncated at 512 bytes. This is too short for some of
2628 the long messages that one sometimes sees. I've increased the limit to
2631 31. It is now possible to make retry rules that apply only when a message has a
2632 specific sender, in particular, an empty sender.
2634 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2635 it possible to be selective about when SMTP synchronization is enforced.
2637 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2639 32. Implemented hosts_connection_nolog.
2641 33. Added an ACL for QUIT.
2643 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2646 35. Added mailbox_size and mailbox_filecount to appendfile.
2648 36. Added control = no_multiline_responses to ACLs.
2650 37. There was a bug in the logic of the code that waits for the clock to tick
2651 in the case where the clock went backwards by a substantial amount such
2652 that the microsecond fraction of "now" was more than the microsecond
2653 fraction of "then" (but the whole seconds number was less).
2655 38. Added support for the libradius Radius client library this is found on
2656 FreeBSD (previously only the radiusclient library was supported).
2662 1. When certain lookups returned multiple values in the form name=value, the
2663 quoting of the values was not always being done properly. Specifically:
2664 (a) If the value started with a double quote, but contained no whitespace,
2666 (b) If the value contained whitespace other than a space character (i.e.
2667 tabs or newlines or carriage returns) it was not quoted.
2668 This fix has been applied to the mysql and pgsql lookups by writing a
2669 separate quoting function and calling it from the lookup code. The fix
2670 should probably also be applied to nisplus, ibase and oracle lookups, but
2671 since I cannot test any of those, I have not disturbed their existing code.
2673 2. A hit in the callout cache for a specific address caused a log line with no
2674 reason for rejecting RCPT. Now it says "Previous (cached) callout
2675 verification failure".
2677 3. There was an off-by-one bug in the queryprogram router. An over-long
2678 return line was truncated at 256 instead of 255 characters, thereby
2679 overflowing its buffer with the terminating zero. As well as fixing this, I
2680 have increased the buffer size to 1024 (and made a note to document this).
2682 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2683 when Exim is waiting for an SMTP response from a remote server, Exim
2684 restarts its select() call on the socket, thereby resetting its timeout.
2685 This is not a problem when such interrupts are rare. Somebody set up a cron
2686 job to run exiwhat every 2 minutes, which is less than the normal select()
2687 timeout (5 or 10 minutes). This meant that the select() timeout never
2688 kicked in because it was always reset. I have fixed this by comparing the
2689 time when an interrupt arrives with the time at the start of the first call
2690 to select(). If more time than the timeout has elapsed, the interrupt is
2691 treated as a timeout.
2693 5. Some internal re-factoring in preparation for the addition of Sieve
2694 extensions (by MH). In particular, the "personal" test is moved to a
2695 separate function, and given an option for scanning Cc: and Bcc: (which is
2696 not set for Exim filters).
2698 6. When Exim created an email address using the login of the caller as the
2699 local part (e.g. when creating a From: or Sender: header line), it was not
2700 quoting the local part when it contained special characters such as @.
2702 7. Installed new OpenBSD configuration files.
2704 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2705 try to make them clearer.
2707 9. Callout options, other than the timeout value, were being ignored when
2708 verifying sender addresses in header lines. For example, when using
2710 verify = header_sender/callout=no_cache
2712 the cache was (incorrectly) being used.
2714 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2715 only the Cygwin environment.
2717 11. When return_path_on_delivery was set as a log selector, if different remote
2718 addresses in the same message used different return paths and parallel
2719 remote delivery occurred, the wrong values would sometimes be logged.
2720 (Whenever a remote delivery process finished, the return path value from
2721 the most recently started remote delivery process was logged.)
2723 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2724 lines when AUTH and/or TLS are in use. This is the "received protocol"
2725 field. Exim used to use "asmtp" for authenticated SMTP, without any
2726 indication (in the protocol name) for TLS use. Now it follows the RFC and
2727 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2728 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2729 names appear in log lines as well as in Received: header lines.
2731 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2732 extensions, and comparison tests, and to fix some bugs.
2734 14. Changes to the "personal" filter test:
2736 (1) The test was buggy in that it was just doing the equivalent of
2737 "contains" tests on header lines. For example, if a user's address was
2738 anne@some.where, the "personal" test would incorrectly be true for
2740 To: susanne@some.where
2742 This test is now done by extracting each address from the header in turn,
2743 and checking the entire address. Other tests that are part of "personal"
2744 are now done using regular expressions (for example, to check local parts
2745 of addresses in From: header lines).
2747 (2) The list of non-personal local parts in From: addresses has been
2748 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2749 taken from the Sieve specification recommendations.
2751 (3) If the message contains any header line starting with "List-" it is
2752 treated as non-personal.
2754 (4) The test for "circular" in the Subject: header line has been removed
2755 because it now seems ill-conceived.
2757 15. Minor typos in src/EDITME comments corrected.
2759 16. Installed latest exipick from John Jetmore.
2761 17. If headers_add on a router specified a text string that was too long for
2762 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2763 of string_sprintf() is now avoided.
2765 18. $message_body_size was not set (it was always zero) when running the DATA
2766 ACL and the local_scan() function.
2768 19. For the "mail" command in an Exim filter, no default was being set for
2769 the once_repeat time, causing a random time value to be used if "once" was
2770 specified. (If the value happened to be <= 0, no repeat happened.) The
2771 default is now 0s, meaning "never repeat". The "vacation" command was OK
2772 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2773 (I found it when inspecting the code).
2775 20. There is now an overall timeout for performing a callout verification. It
2776 defaults to 4 times the callout timeout, which applies to individual SMTP
2777 commands during the callout. The overall timeout applies when there is more
2778 than one host that can be tried. The timeout is checked before trying the
2779 next host. This prevents very long delays if there are a large number of
2780 hosts and all are timing out (e.g. when the network connections are timing
2781 out). The value of the overall timeout can be changed by specifying an
2782 additional sub-option for "callout", called "maxwait". For example:
2784 verify = sender/callout=5s,maxwait=20s
2786 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2787 to the end before writing, but this should make it even safer).
2789 22. Exim was forgetting that it had advertised PIPELINING for the second and
2790 subsequent messages on an SMTP connection. It was also not resetting its
2791 memory on STARTTLS and an internal HELO.
2793 23. When Exim logs an SMTP synchronization error within a session, it now
2794 records whether PIPELINING has been advertised or not.
2796 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2797 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2798 rather than long int.
2800 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2802 26. Added the never_mail option to autoreply.
2808 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2809 crash if the getsockname() call failed; this can happen if a connection is
2810 closed very soon after it is established. The problem was simply in the
2811 order in which certain operations were done, causing Exim to try to write
2812 to the SMTP stream before it had set up the file descriptor. The bug has
2813 been fixed by making things happen in the correct order.
2819 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2820 before the connection was closed, thus losing the rejection response.
2822 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2823 some early Solaris releases, but causes trouble in current releases where
2824 socklen_t is defined.
2826 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2829 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2832 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2833 that releasing the top of it at the end releases what was used for sub-
2834 expansions (unless the block got too big). However, discard this block if
2835 the first thing is a variable or header, so that we can use its block when
2836 it is dynamic (useful for very large $message_headers, for example).
2838 6. Lookups now cache *every* query, not just the most recent. A new, separate
2839 store pool is used for this. It can be recovered when all lookup caches are
2840 flushed. Lookups now release memory at the end of their result strings.
2841 This has involved some general refactoring of the lookup sources.
2843 7. Some code has been added to the store_xxx() functions to reduce the amount
2844 of flapping under certain conditions.
2846 8. log_incoming_interface used to affect only the <= reception log lines. Now
2847 it causes the local interface and port to be added to several more SMTP log
2848 lines, for example "SMTP connection from", and rejection lines.
2850 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2852 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2854 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2855 could be overwritten at the end of the current message (or the start of a
2856 new message if it was set in a HELO ACL). The value is now preserved for
2857 the duration of the SMTP connection.
2859 12. If a transport had a headers_rewrite setting, and a matching header line
2860 contained an unqualified address, that address was qualified, even if it
2861 did not match any rewriting rules. The underlying bug was that the values
2862 of the flags that permit the existence of unqualified sender and recipient
2863 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2864 non-local messages, and by -bnq for local messages) were not being
2865 preserved with the message after it was received.
2867 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2868 "next input=" as part of the text comprising the host identity instead of
2869 the correct text. The code was using the same buffer for two different
2870 strings. However, depending on which order the printing function evaluated
2871 its arguments, the bug did not always show up. Under Linux, for example, my
2872 test suite worked just fine.
2874 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2875 doesn't work with some older versions of Perl. It has been changed to "my",
2876 which in any case is probably the better facility to use.
2878 15. A really picky compiler found some instances of statements for creating
2879 error messages that either had too many or two few arguments for the format
2882 16. The size of the buffer for calls to the DNS resolver has been increased
2883 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2884 for addresses that have a lot of PTR records. This alleviates a problem; it
2885 does not fully solve it.
2887 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2888 buffer now truncates the list and logs the incident, which is the same
2889 action as happens when Exim is looking up a host name and its aliases.
2890 Previously in this situation something unpredictable would happen;
2891 sometimes it was "internal error: store_reset failed".
2893 18. If a server dropped the connection unexpectedly when an Exim client was
2894 using GnuTLS and trying to read a response, the client delivery process
2895 crashed while trying to generate an error log message.
2897 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2898 string, for example:
2900 warn message = H1: something\nH2: something
2902 the text was added as a single header line from Exim's point of view
2903 though it ended up OK in the delivered message. However, searching for the
2904 second and subsequent header lines using $h_h2: did not work. This has been
2905 fixed. Similarly, if a system filter added multiple headers in this way,
2906 the routers could not see them.
2908 20. Expanded the error message when iplsearch is called with an invalid key to
2909 suggest using net-iplsearch in a host list.
2911 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2912 ACLs are no longer actually imposed (and a message to that effect is
2915 22. If a "gecos" field in a passwd entry contained escaped characters, in
2916 particular, if it contained a \" sequence, Exim got it wrong when building
2917 a From: or a Sender: header from that name. A second bug also caused
2918 incorrect handling when an unquoted " was present following a character
2919 that needed quoting.
2921 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2922 was not being matched caselessly.
2924 24. Arranged for all hyphens in the exim.8 source to be escaped with
2927 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2928 properly. Recipient callout cache records were still being keyed to include
2929 the sender, even when use_sender was set false. This led to far more
2930 callouts that were necessary. The sender is no longer included in the key
2931 when use_sender is false.
2933 26. Added "control = submission" modifier to ACLs.
2935 27. Added the ${base62d: operator to decode base 62 numbers.
2937 28. dnsdb lookups can now access SRV records.
2939 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2940 the configuration file.
2942 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2943 (-v) mode. This makes the output for a verbose queue run more intelligible.
2945 31. Added a use_postmaster feature to recipient callouts.
2947 32. Added the $body_zerocount variable, containing the number of binary zero
2948 bytes in the message body.
2950 33. The time of last modification of the "new" subdirectory is now used as the
2951 "mailbox time last read" when there is a quota error for a maildir
2954 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2956 35. Added +ignore_unknown as a special item in host lists.
2958 36. Code for decoding IPv6 addresses in host lists is now included, even if
2959 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2960 address was recognized as an IP address, but was then not correctly decoded
2961 into binary, causing unexpected and incorrect effects when compared with
2968 1. Very minor rewording of debugging text in manualroute to say "list of
2969 hosts" instead of "hostlist".
2971 2. If verify=header_syntax was set, and a header line with an unqualified
2972 address (no domain) and a large number of spaces between the end of the
2973 name and the colon was received, the reception process suffered a buffer
2974 overflow, and (when I tested it) crashed. This was caused by some obsolete
2975 code that should have been removed. The fix is to remove it!
2977 3. When running in the test harness, delay a bit after writing a bounce
2978 message to get a bit more predictability in the log output.
2980 4. Added a call to search_tidyup() just before forking a reception process. In
2981 theory, someone could use a lookup in the expansion of smtp_accept_max_
2982 per_host which, without the tidyup, could leave open a database connection.
2984 5. Added the variables $recipient_data and $sender_data which get set from a
2985 lookup success in an ACL "recipients" or "senders" condition, or a router
2986 "senders" option, similar to $domain_data and $local_part_data.
2988 6. Moved the writing of debug_print from before to after the "senders" test
2991 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2992 problems for message scanning, either using a data ACL, or using
2993 local_scan() because the Received: header was not generated till after they
2994 were called (in order to set the time as the time of reception completion).
2995 I have revised the way this works. The header is now generated after the
2996 body is received, but before the ACL or local_scan() are called. After they
2997 are run, the timestamp in the header is updated.
3003 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3004 before starting a queue runner without re-exec. This happened only when
3005 deliver_drop_privilege was set or when the Exim user was set to root. The
3006 effect of the bug was that timeouts during subsequent deliveries caused
3007 crashes instead of being properly handled. The handler is now left at its
3008 default (and expected) setting.
3010 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3011 message, again when deliver_drop_privilege is set or Exim is run as root.
3012 The bug described in (1) was not present in this case, but the tidying up
3013 of the other signals was missing. I have made the two cases consistent.
3015 3. The ignore_target_hosts setting on a manualroute router was being ignored
3016 for hosts that were looked up using the /MX notation.
3018 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3021 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3022 operated on the sender address. After changing the $sender_address to <>
3023 for the sender address verify, Exim was re-instated it as the original
3024 (before rewriting) address, but remembering that it had rewritten it, so it
3025 wasn't rewriting it again. This bug also had the effect of breaking the
3026 sender address verification caching when the sender address was rewritten.
3028 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3029 This has been changed so that if the ip literal address matches
3030 ignore_target_hosts, the router declines.
3032 7. Added expansion conditions match_domain, match_address, and match_local_
3033 part (NOT match_host).
3035 8. The placeholder for the Received: header didn't have a length field set.
3037 9. Added code to Exim itself and to exim_lock to test for a specific race
3038 condition that could lead to file corruption when using MBX delivery. The
3039 issue is with the lockfile that is created in /tmp. If this file is removed
3040 after a process has opened it but before that process has acquired a lock,
3041 there is the potential for a second process to recreate the file and also
3042 acquire a lock. This could lead to two Exim processes writing to the file
3043 at the same time. The added code performs the same test as UW imapd; it
3044 checks after acquiring the lock that its file descriptor still refers to
3045 the same named file.
3047 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3048 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3051 11. Added the smtp_active_hostname option. If used, this will typically be made
3052 to depend on the incoming interface address. Because $interface_address is
3053 not set up until the daemon has forked a reception process, error responses
3054 that can happen earlier (such as "too many connections") no longer contain
3057 12. If an expansion in a condition on a "warn" statement fails because a lookup
3058 defers, the "warn" statement is abandoned, and the next ACL statement is
3059 processed. Previously this caused the whole ACL to be aborted.
3061 13. Added the iplsearch lookup type.
3063 14. Added ident_timeout as a log selector.
3065 15. Added tls_certificate_verified as a log selector.
3067 16. Added a global option tls_require_ciphers (compare the smtp transport
3068 option of the same name). This controls incoming TLS connections.
3070 17. I finally figured out how to make tls_require_ciphers do a similar thing
3071 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3072 before starting the TLS session.
3074 18. Tabs are now shown as \t in -bP output.
3076 19. If the log selector return_path_on_delivery was set, Exim crashed when
3077 bouncing a message because it had too many Received: header lines.
3079 20. If two routers both had headers_remove settings, and the first one included
3080 a superfluous trailing colon, the final name in the first list and the
3081 first name in the second list were incorrectly joined into one item (with a
3082 colon in the middle).
3088 1. Added -C and -D options to the exinext utility, mainly to make it easier
3089 to include in the automated testing, but these could be helpful when
3090 multiple configurations are in use.
3092 2. The exinext utility was not formatting the output nicely when there was
3093 an alternate port involved in the retry record key, nor when there was a
3094 message id as well (for retries that were specific to a specific message
3095 and a specific host). It was also confused by IPv6 addresses, because of
3096 the additional colons they contain. I have fixed the IPv4 problem, and
3097 patched it up to do a reasonable job for IPv6.
3099 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3100 delivery, the log line now contains "pipelined" if PIPELINING was used.
3102 4. An SMTP transport process used to panic and die if the bind() call to set
3103 an explicit outgoing interface failed. This has been changed; it is now
3104 treated in the same way as a connect() failure.
3106 5. A reference to $sender_host_name in the part of a conditional expansion
3107 that was being skipped was still causing a DNS lookup. This no longer
3110 6. The def: expansion condition was not recognizing references to header lines
3111 that used bh_ and bheader_.
3113 7. Added the _cache feature to named lists.
3115 8. The code for checking quota_filecount in the appendfile transport was
3116 allowing one more file than it should have been.
3118 9. For compatibility with Sendmail, the command line option
3126 and sets the incoming protocol and host name (for trusted callers). The
3127 host name and its colon can be omitted when only the protocol is to be set.
3128 Note the Exim already has two private options, -pd and -ps, that refer to
3129 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3130 "s", but I don't think that's a major issue.
3132 10. A number of refactoring changes to the code, none of which should affect
3135 (a) The number of logging options was getting close to filling up the
3136 32-bit word that was used as a bit map. I have split them into two classes:
3137 those that are passed in the argument to log_write(), and those that are
3138 only ever tested independently outside of that function. These are now in
3139 separate 32-bit words, so there is plenty of room for expansion again.
3140 There is no change in the user interface or the logging behaviour.
3142 (b) When building, for example, log lines, the code previously used a
3143 macro that called string_cat() twice, in order to add two strings. This is
3144 not really sufficiently general. Furthermore, there was one instance where
3145 it was actually wrong because one of the argument was used twice, and in
3146 one call a function was used. (As it happened, calling the function twice
3147 did not affect the overall behaviour.) The macro has been replaced by a
3148 function that can join an arbitrary number of extra strings onto a growing
3151 (c) The code for expansion conditions now uses a table and a binary chop
3152 instead of a serial search (which was left over from when there were very
3153 few conditions). Also, it now recognizes conditions like "pam" even when
3154 the relevant support is not compiled in: a suitably worded error message is
3155 given if an attempt is made to use such a condition.
3157 11. Added ${time_interval:xxxxx}.
3159 12. A bug was causing one of the ddress fields not to be passed back correctly
3160 from remote delivery subprocesses. The field in question was not being
3161 subsequently used, so this caused to problems in practice.
3163 13. Added new log selectors queue_time and deliver_time.
3165 14. Might have fixed a bug in maildirsizefile handling that threw up
3166 "unexpected character" debug warnings, and recalculated the data
3167 unnecessarily. In any case, I expanded the warning message to give more
3170 15. Added the message "Restricted characters in address" to the statements in
3171 the default ACL that block characters like @ and % in local parts.
3173 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3174 Three changes have been made:
3176 (a) There was a serious bug; a negative response to MAIL caused the whole
3177 recipient domain to be cached as invalid, thereby blocking all messages
3178 to all local parts at the same domain, from all senders. This bug has
3179 been fixed. The domain is no longer cached after a negative response to
3180 MAIL if the sender used is not empty.
3182 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3185 (c) A new callout option, "use_sender" has been added for people who want
3186 the modified behaviour.
3192 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3195 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3196 indeed breaks things for older releases.
3198 3. Added additional logging to the case where there is a problem reading data
3199 from a filter that is running in a subprocess using a pipe, in order to
3200 try to track down a specific problem.
3202 4. Testing facility fudge: when running in the test harness and attempting
3203 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3204 getting "No route to host". Convert this to a timeout.
3206 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3209 6. Some OS don't have socklen_t but use size_t instead. This affects the
3210 fifth argument of getsockopt() amongst other things. This is now
3211 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3212 can be set for individual OS. I have set it for SunOS5, OSF1, and
3213 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3214 some earlier ones do not.
3216 7. Change 4.30/15 was not doing the test caselessly.
3218 8. The standard form for an IPv6 address literal was being rejected by address
3219 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3220 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3221 this, as well as the form without the "IPv6" on the front (but only when
3222 address literals are enabled, of course).
3224 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3226 10. Exim crashed if a message with an empty sender address specified by -f
3227 encountered a router with an errors_to setting. This could be provoked only
3228 by a command such as
3232 where an empty string was supplied; "<>" did not hit this bug.
3234 11. Installed PCRE release 4.5.
3236 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3237 remained set. It is now erased.
3239 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3240 times from message ids (which are base 36 rather than the normal 62).
3242 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3243 were being counted as actual protocol errors, and logged if the log
3244 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3245 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3246 and DATA following a set of rejected RCPTs do not count as protocol errors.
3247 In other words, Exim assumes they were pipelined, though this may not
3248 actually be the case. Of course, in all cases the client gets an
3249 appropriate error code.
3251 15. If a lookup fails in an ACL condition, a message about the failure may
3252 be available; it is used if testing the ACL cannot continue, because most
3253 such messages specify what the cause of the deferral is. However, some
3254 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3255 that caused an old message to be retained and used if a later statement
3256 caused a defer, replacing the real cause of the deferral.
3258 16. If an IP address had so many PTR records that the DNS lookup buffer
3259 was not large enough to hold them, Exim could crash while trying to process
3260 the truncated data. It now detects and logs this case.
3262 17. Further to 4.21/58, another change has been made: if (and only if) the
3263 first line of a message (the first header line) ends with CRLF, a bare LF
3264 in a subsequent header line has a space inserted after it, so as not to
3265 terminate the header.
3267 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3268 unnecessarily, used atoi() instead of strtol(), and didn't check the
3269 termination when getting file sizes from file names by regex.
3271 19. Completely re-implemented the support for maildirsize files, in the light
3272 of a number of problems with the previous contributed implementation
3273 (4.30/29). In particular:
3275 . If the quota is zero, the maildirsize file is maintained, but no quota is
3278 . If the maildir directory does not exist, it is created before any attempt
3279 to write a maildirsize file.
3281 . The quota value in the file is just a cache; if the quota is changed in
3282 the transport, the new value overrides.
3284 . A regular expression is available for excluding directories from the
3287 20. The autoreply transport checks the characters in options that define the
3288 message's headers; it allows continued headers, but it was checking with
3289 isspace() after an embedded newline instead of explicitly looking for a
3292 21. If all the "regular" hosts to which an address was routed had passed their
3293 expiry times, and had not reached their retry times, the address was
3294 bounced, even if fallback hosts were defined. Now Exim should go on to try
3297 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3298 equivalent code in the SMTP transport. Some hosts send humungous responses
3299 to HELO/EHLO, more than 1024 it seems.
3301 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3302 gives compiler warnings in some environments. I've now done it "properly",
3305 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3306 (because of problems with the built-in one) was declared to return uschar *
3307 instead of char *, causing compiler failure.
3309 25. Fixed a file descriptor leak when processing alias/forward files.
3311 26. Fixed a minor format string issue in dbfn.c.
3313 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3315 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3316 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3319 29. When a sender address is verified, it is cached, to save repeating the test
3320 when there is more than one recipient in a message. However, when the
3321 verification involves a callout, it is possible for different callout
3322 options to be set for different recipients. It is too complicated to keep
3323 track of this in the cache, so now Exim always runs a verification when a
3324 callout is required, relying on the callout cache for the optimization.
3325 The overhead is duplication of the address routing, but this should not be
3328 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3329 to be verified with callout=postmaster, and the main callout worked but the
3330 postmaster check failed, the verification correctly failed. However, if a
3331 subsequent RCPT command asked for sender verification *without* the
3332 postmaster check, incorrect caching caused this verification also to fail,
3335 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3336 it was not caching the DNS options (qualify_single, search_parents) that
3337 were used when the lookup failed. A subsequent lookup with different
3338 options therefore always gave the same answer, though there were cases
3339 where it should not have. (Example: a "domains = !$mx_any" option on a
3340 dnslookup router: the "domains" option is always processed without any
3341 widening, but the router might have qualify_single set.) Now Exim uses the
3342 cached value only when the same options are set.
3344 32. Added John Jetmore's "exipick" utility to the distribution.
3346 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3347 than a timeout (e.g. a certificate is required, and is not provided), an
3348 Exim server now closes the connection immediately. Previously it waited for
3349 the client to close - but if the client is SSL, it seems that they each
3350 wait for each other, leading to a delay before one of them times out.
3352 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3353 maintained 0.8.x compatibility because I don't think many are using it, and
3354 it is clearly obsolete.
3356 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3359 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3360 client certificate was expired. A simple patch fixes this, though I don't
3361 understand the full logic of why the verify callback is called multiple
3364 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3365 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3366 which causes problems with some clients (such as the Certicom SSL Plus
3367 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3368 disables the coutermeasure allowing Eudora to connect."
3370 38. Exim was not checking that a write() to a log file succeeded. This could
3371 lead to Bad Things if a log got too big, in particular if it hit a file
3372 size limit. Exim now panics and dies if it cannot write to a log file, just
3373 as it does if it cannot open a log file.
3375 39. Modified OS/Makefile-Linux so that it now contains
3377 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3379 The two -D definitions ensure that Exim is compiled with large file
3380 support, which makes it possible to handle log files that are bigger than
3383 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3384 instance) a domain was checked against a named list that involved a lookup,
3385 causing $domain_data to be set, then another domain was checked against the
3386 same list, then the first domain was re-checked, the value of $domain_data
3387 after the final check could be wrong. In particular, if the second check
3388 failed, it could be set empty. This bug probably also applied to
3391 41. The strip_trailing_dot option was not being applied to the address given
3392 with the -f command-line option.
3394 42. The code for reading a message's header from the spool was incrementing
3395 $received_count, but never initializing it. This meant that the value was
3396 incorrect (doubled) while delivering a message in the same process in which
3397 it was received. In the most common configuration of Exim, this never
3398 happens - a fresh exec is done - but it can happen when
3399 deliver_drop_privilege is set.
3401 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3402 it now includes up to 150 characters of the unexpected data in the log
3405 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3406 and building data strings. The size of both of these buffers was 10 000
3407 bytes - far larger than anybody would *ever* want, thought I. Needless to
3408 say, somebody hit the limit. I have increased the maximum line length to
3409 20 000 and the maximum data length of concatenated lines to 100 000. I have
3410 also fixed two bugs, because there was no checking on these buffers. Tsk,
3411 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3412 buffer is too small.
3414 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3415 lsearch lookups. Now it does.
3417 46. When parsing a route_list item in a manualroute router, a fixed-length
3418 buffer was used for the list of hosts. I made this 1024 bytes long,
3419 thinking that nobody would ever have a list of hosts that long. Wrong.
3420 Somebody had a whole pile of complicated expansion conditions, and the
3421 string was silently truncated, leading to an expansion error. It turns out
3422 that it is easier to change to an unlimited length (owing to other changes
3423 that have happened since this code was originally written) than to build
3424 structure for giving a limitation error. The length of the item that
3425 expands into the list of hosts is now unlimited.
3427 47. The lsearch lookup could not handle data where the length of text line was
3428 more than 4095 characters. Such lines were truncated, leading to shortened
3429 data being returned. It should now handle lines of any length.
3431 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3432 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3435 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3436 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3437 no longer happens in comment lines. A list of replacements is now placed
3438 at the head of all of the source files, except those whose only change is
3439 to replace PERL_COMMAND in the very first #! line.
3441 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3442 messages on the queue, with a bottom-up merge sort, using code contributed
3443 by Michael Haardt. This should make operations like -bp somewhat faster on
3444 large queues. It won't affect queue runners, except when queue_run_in_order
3447 51. Installed eximstats 1.31 in the distribution.
3449 52. Added support for SRV lookups to the dnslookup router.
3451 53. If an ACL referred to $message_body or $message_body_end, the value was not
3452 reset for any messages that followed in the same SMTP session.
3454 54. The store-handling optimization for building very long strings was not
3455 differentiating between the different store pools. I don't think this
3456 actually made any difference in practice, but I've tidied it.
3458 55. While running the routers to verify a sender address, $sender_address
3459 was still set to the sender address. This is wrong, because when routing to
3460 send a bounce to the sender, it would be empty. Therefore, I have changed
3461 it so that, while verifying a sender address, $sender_address is set to <>.
3462 (There is no change to what happens when verifying a recipient address.)
3464 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3465 target A or AAAA records (if not already returned) without resetting the
3466 qualify_single or search_parents options of the DNS resolver. These are
3467 inappropriate in this case because the targets of MX and SRV records must
3468 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3469 target that, when qualified, matched something in the local domain. These
3470 two options are now turned off when doing these lookups.
3472 57. It seems that at least some releases of Reiserfs (which does not have the
3473 concept of a fixed number of inodes) returns zero and not -1 for the
3474 number of available inodes. This interacted badly with check_spool_inodes,
3475 which assumed that -1 was the "no such thing" setting. What I have done is
3476 to check that the total number of inodes is greater than zero before doing
3477 the test of how many are available.
3479 58. When a "warn" ACL statement has a log_message modifier, the message is
3480 remembered, and not repeated. This is to avoid a lot of repetition when a
3481 message has many recipients that cause the same warning to be written.
3482 Howewer, Exim was preserving the list of already written lines for an
3483 entire SMTP session, which doesn't seem right. The memory is now reset if a
3484 new message is started.
3486 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3487 debugging output unless log_rewrite was also set.
3489 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3490 of (int)(handle) when we know that handle contains (void *)(-1).
3492 61. The Exim daemon panic-logs an error return when it closes the incoming
3493 connection. However "connection reset by peer" seems to be common, and
3494 isn't really an error worthy of noting specially, so that particular error
3497 62. When Exim is trying to find all the local interfaces, it used to panic and
3498 die if the ioctl to get the interface flags failed. However, it seems that
3499 on at least one OS (Solaris 9) it is possible to have an interface that is
3500 included in the list of interfaces, but for which you get a failure error
3501 for this call. This happens when the interface is not "plumbed" into a
3502 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3503 failure of the "get flags" call assumes that the interface is down.
3505 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3506 makes life easier for people who are doing arithmetic on fields extracted
3507 from dates, where you often get leading zeros that should not be
3508 interpreted as octal.
3510 64. Added qualify_domain to the redirect router, to override the global
3513 65. If a pathologically long header line contained very many addresses (the
3514 report of this problem mentioned 10 000) and each of them was rewritten,
3515 Exim could use up a very large amount of memory. (It kept on making new
3516 copies of the header line as it rewrote, and never released the old ones.)
3517 At the expense of a bit more processing, the header rewriting function has
3518 been changed so that it no longer eats memory in this way.
3520 66. The generation of the Received: header has been moved from the time that a
3521 message starts to be received, to the time that it finishes. The timestamp
3522 in the Received: header should now be very close to that of the <= log
3523 line. There are two side-effects of this change:
3525 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3526 logged header lines no longer include the local Received: line, because
3527 it has not yet been created. The same applies to a copy of the message
3528 that is returned to a non-SMTP sender when a message is rejected.
3530 (b) When a filter file is tested using -bf, no additional Received: header
3531 is added to the test message. After some thought, I decided that this
3534 This change does not affect the value of $received_for. It is still set
3535 after address rewriting, but before local_scan() is called.
3537 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3539 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3540 gave an unhelpful panic error message, and a defer error. I have managed to
3541 change this behaviour so that it now rejects any supplied certificate,
3542 which seems right, as the list of acceptable certificates is empty.
3544 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3545 gave an unhelpful defer error. I have not managed to make this reject any
3546 supplied certificates, but the error message it gives is "no certificate
3547 supplied", which is not helpful.
3549 70. exigrep's output now also includes lines that are not associated with any
3550 message, but which match the given pattern. Implemented by a patch from
3551 Martin Sluka, which also tidied up the Perl a bit.
3553 71. Recipient callout verification, like sender verification, was using <> in
3554 the MAIL FROM command. This isn't really the right thing, since the actual
3555 sender may affect whether the remote host accepts the recipient or not. I
3556 have changed it to use the actual sender in the callout; this means that
3557 the cache record is now keyed on a recipient/sender pair, not just the
3558 recipient address. There doesn't seem to be a real danger of callout loops,
3559 since a callout by the remote host to check the sender would use <>.
3560 [SEE ABOVE: changed after hitting problems.]
3562 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3563 temporary errors. However, in the case of such a code being given after
3564 the end of a data transmission (i.e. after ".") Exim was failing to write
3565 a retry record for the message. (Yes, there was some broken host that was
3566 actually sending 8xx at this point.)
3568 73. An unknown lookup type in a host list could cause Exim to panic-die when
3569 the list was checked. (An example that provoked this was putting <; in the
3570 middle of a list instead of at the start.) If this happened during a DATA
3571 ACL check, a -D file could be left lying around. This kind of configuration
3572 error no longer causes Exim to die; instead it causes a defer errror. The
3573 incident is still logged to the main and panic logs.
3575 74. Buglet left over from Exim 3 conversion. The message "too many messages
3576 in one connection" was written to the rejectlog but not the mainlog, except
3577 when address rewriting (yes!) was being logged.
3579 75. Added write_rejectlog option.
3581 76. When a system filter was run not as root (that is, when system_filter_user
3582 was set), the values of the $n variables were not being returned to the
3583 main process; thus, they were not subsequently available in the $sn
3586 77. Added +return_path_on_delivery log selector.
3588 78. A connection timeout was being treated differently from recipients deferred
3589 when testing hosts_max_try with a message that was older than the host's
3590 retry timeout. (The host should not be counted, thus allowing all hosts to
3591 be tried at least once before bouncing.) This may have been the cause of an
3592 occasionally reported bug whereby a message would remain on the queue
3593 longer than the retry timeout, but would be bounced if a delivery was
3594 forced. I say "may" because I never totally pinned down the problem;
3595 setting up timeout/retry tests is difficult. See also the next item.
3597 79. The ultimate address timeout was not being applied to errors that involved
3598 a combination of host plus message (for example, a timeout on a MAIL
3599 command). When an address resolved to a number of possible hosts, and they
3600 were not all tried for each delivery (e.g. because of hosts_max_try), a
3601 message could remain on the queue longer than the retry timeout.
3603 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3606 81. Fixed an obscure SMTP outgoing bug which required at least the following
3607 conditions: (a) there was another message waiting for the same server;
3608 (b) the server returned 5xx to all RCPT commands in the first message so
3609 that the message was not completed; (c) the server dropped the connection
3610 or gave a negative response to the RSET that Exim sends to abort the
3611 transaction. The observed case was a dropped connection after DATA that had
3612 been sent in pipelining mode. That is, the server had advertised PIPELINING
3613 but was not implementing it correctly. The effect of the bug was incorrect
3614 behaviour, such as trying another host, and this could lead to a crash.
3620 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3621 and daemon.c were passed as pointers to ints; they should have been
3622 pointers to socklen_t variables (which are typically unsigned ints).
3624 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3627 3. Fixed a really odd bug that affected only the testing scheme; patching a
3628 certain fixed string in the binary changed the value of another string that
3629 happened to be identical to the end of the original first string.
3631 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3632 name", it returns that address as the IP address. On some operating
3633 systems (e.g. Solaris), it also passes back the IP address string as the
3634 "host name". However, on others (e.g. Linux), it passes back an empty
3635 string. Exim wasn't checking for this, and was changing the host name to an
3636 empty string, assuming it had been canonicized.
3638 5. Although rare, it is permitted to have more than one PTR record for a given
3639 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3640 all the names associated with an address, because they do in Solaris.
3641 However, it seems that they do not in Linux for data that comes from the
3642 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3643 I found this out when I moved to a new Linux workstation and tried to run
3644 the Exim test suite.
3646 To get round this problem I have changed the code so that it now does its
3647 own call to the DNS to look up PTR records when searching for a host name.
3648 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3649 addresses that are only in /etc/hosts are still found.
3651 This behaviour is, however, controlled by an option called host_lookup_
3652 order, which defaults to "bydns:byaddr". If people want to use the other
3653 order, or indeed, just use one or the other means of lookup, they can
3654 specify it in this variable.
3656 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3657 some operating systems, this comes back from gethostbyaddr() as an empty
3658 string, and this is what Exim used to test for. However, it seems that in
3659 other systems, "." is yielded. Exim now tests for this case too.
3661 7. The values of check_spool_space and check_log_space are now held internally
3662 as a number of kilobytes instead of an absolute number of bytes. If a
3663 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3664 kilobyte. This means that much larger values can be stored.
3666 8. Exim monitor: an attempt to get the action menu when not actually pointing
3667 at a message produces an empty menu entitled "No message selected". This
3668 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3669 no entries in it ("Shell widget menu has zero width and/or height"). So I
3670 have added a single, blank menu entry in this case.
3672 9. Added ${quote_local_part.
3674 10. MIME decoding is now applied to the contents of Subject: header lines when
3677 11. Now that a reference to $sender_host_address automatically causes a reverse
3678 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3679 host lookup before query-style lookups in lists that might use this
3680 variable. This has therefore been abolished, and the "net-" prefix is no
3681 longer necessary for query-style lookups.
3683 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3684 have been a typo for LFLAGS, so it has been changed.
3686 13. The install script calls Exim with "-C /dev/null" in order to find the
3687 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3688 to be output. Howeve, since Exim outputs its version number before the
3689 error, it didn't break the script. It just looked ugly. I fixed this by
3690 always allowing "-C /dev/null" if the caller is root.
3692 14. Ignore overlarge ACL variable number when reading spool file - insurance
3693 against a later release with more variables having written the file.
3695 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3696 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3697 this, as well as the form without the "IPv6" on the front.
3699 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3700 OS/Makefile-Darwin file.
3702 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3703 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3705 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3706 variable. However, some OS use ints for inodes. I've added cast to long int
3707 to get rid of the compiler warning.
3709 19. I had forgotten to lock out "/../" in configuration file names when
3710 ALT_CONFIG_PREFIX was set.
3712 20. Routers used for verification do not need to specify transports. However,
3713 if such a router generated a host list, and callout was configured, Exim
3714 crashed, because it could not find a port number from the (non-existent)
3715 transport. It now assumes port 25 in this circumstance.
3717 21. Added the -t option to exigrep.
3719 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3720 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3721 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3722 them is set, LOOKUP_LSEARCH is forced.
3724 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3726 24. Added sender and host information to the "rejected by local_scan()" log
3727 line; previously there was no indication of these.
3729 25. Added .include_if_exists.
3731 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3732 Linux. It turns out that not all file systems support this. Apparently some
3733 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3734 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3735 not supported on the file descriptor, is now ignored when Exim is trying to
3736 sync a directory. This applies only to Linux.
3738 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3740 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3741 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3744 29. Added support for maildirsize files from supplied patch (modified a bit).
3746 30. The use of :fail: followed by an empty string could lead Exim to respond to
3747 sender verification failures with (e.g.):
3749 550 Verification failed for <xxx>
3750 550 Sender verify failed
3752 where the first response line was missing the '-' that indicates it is not
3753 the final line of the response.
3755 31. The loop for finding the name of the user that called Exim had a hardwired
3756 limit of 10; it now uses the value of finduser_retries, which is used for
3757 all other user lookups.
3759 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3762 33. Exim was neglecting to zero errno before one call of strtol() when
3763 expanding a string and expecting an integer value. On some systems this
3764 resulted in spurious "integer overflow" errors. Also, it was casting the
3765 result into an int without checking.
3767 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3768 did not work. The code looks as if it has *never* worked, though it appears
3769 to have been documented since at least releast 1.62. I have made it work.
3771 35. The "timeout_DNS" error in retry rules, also documented since at least
3772 1.62, also never worked. As it isn't clear exactly what this means, and
3773 clearly it isn't a major issue, I have abolished the feature by treating it
3774 as "timeout", and writing a warning to the main and panic logs.
3776 36. The display of retry rules for -brt wasn't always showing the error code
3779 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3780 timeout_connect_A, timeout_connect_MX.
3782 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3783 to the empty sender.
3785 39. The daemon was not analysing the content of -oX till after it had closed
3786 stderr and disconnected from the controlling terminal. This meant that any
3787 syntax errors were only noted on the panic log, and the return code from
3788 the command was 0. By re-arranging the code a little, I've made the
3789 decoding happen first, so such errors now appear on stderr, and the return
3790 code is 1. However, the actual setting up of the sockets still happens in
3791 the disconnected process, so errors there are still only recorded on the
3794 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3795 connections (as happens on some IP stacks) was logged at start up time as
3796 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3797 it from "IPv6 and IPv4", which means that two separate sockets are being
3800 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3801 says whether AF_INET or AF_INET6 was passed as an argument.
3803 42. Exiwhat output was messed up when time zones were included in log
3806 43. Exiwhat now gives more information about the daemon's listening ports,
3807 and whether -tls-on-connect was used.
3809 44. The "port" option of the smtp transport is now expanded.
3811 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3812 silently ignored. Now an error message is written to the main and panic
3815 46. There's a new ACL modifier called "logwrite" which writes to a log file
3816 as soon as it is encountered.
3818 47. Added $local_user_uid and $local_user_gid at routing time.
3820 48. Exim crashed when trying to verify a sender address that was being
3823 49. Exim was recognizing only a space character after ".include". It now also
3824 recognizes a tab character.
3826 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3827 extracting the relevant information from the specification. The man page no
3828 longer contains scrambled data for the -d option, and I've added a section
3829 at the front about calling Exim under different names.
3831 51. Added "extra_headers" argument to the "mail" command in filter files.
3833 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3836 53. Installed eximstats 1.29.
3838 54. Added transport_filter_timeout as a generic transport option.
3840 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3841 Cc: header lines. This was required by RFC 822, but it not required by RFC
3844 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3845 incoming messages that did not have them. Now it does so only if the
3846 message originates locally, that is, if there is no associated remote host
3847 address. When Resent- header lines are present, this applies to the Resent-
3848 lines rather than the non-Resent- lines.
3850 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3851 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3853 58. Messages for configuration errors now include the name of the main
3854 configuration file - useful now that there may be more than one file in a
3855 list (.included file names were always shown).
3857 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3858 for those rare installations that do not start the daemon as root or run it
3859 setuid root. I've cut out the call to initgroups() if the daemon is not
3862 60. The Exim user and group can now be bound into the binary as text strings
3863 that are looked up at the start of Exim's processing.
3865 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3867 62. Added $mailstore_basename variable.
3869 63. Installed patch to sieve.c from Michael Haardt.
3871 64. When Exim failed to open the panic log after failing to open the main log,
3872 the original message it was trying to log was written to stderr and debug
3873 output, but if they were not available (the usual case in production), it
3874 was lost. Now it is written to syslog before the two lines that record the
3875 failures to open the logs.
3877 65. Users' Exim filters run in subprocesses under the user's uid. It is
3878 possible for a "deliver" command or an alias in a "personal" command to
3879 provoke an address rewrite. If logging of address rewriting is configured,
3880 this fails because the process is not running as root or exim. There may be
3881 a better way of dealing with this, but for the moment (because 4.30 needs
3882 to be released), I have disabled address rewrite logging when running a
3883 filter in a non-root, non-exim process.
3889 1. The buildconfig auxiliary program wasn't quoting the value set for
3890 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3891 not defined. This bug was masked in 4.22 by the effect that was fixed in
3894 2. Some messages that were rejected after a message id was allocated were
3895 shown as "incomplete" by exigrep. It no longer does this for messages that
3896 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3898 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3899 have allow_domain_literals set, the ID did not get logged in the <= line.
3900 Domain literals are now always recognized in Message-ID: header lines.
3902 4. The first argument for a ${extract expansion item is the key name or field
3903 number. Leading and trailing spaces in this item were not being ignored,
3904 causing some misleading effects.
3906 5. When deliver_drop_privilege was set, single queue runner processes started
3907 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3908 same command in the process it spins off) were not dropping privilege.
3910 6. When the daemon running as "exim" started a queue runner, it always
3911 re-executed Exim in the spun-off process. This is a waste of effort when
3912 deliver_drop_privilege is set. The new process now just calls the
3913 queue-runner function directly.
3919 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3922 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3923 ignored. Though the use of -oP was forcing the writing of a pid file, it
3924 was always written to the default place.
3926 3. If the message "no IP address found for host xxxx" is generated during
3927 incoming verification, it is now followed by identification of the incoming
3928 connection (so you can more easily find what provoked it).
3930 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3932 5. Added some features to "harden" Exim a bit more against certain attacks:
3934 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3935 be put in Local/Makefile. This is like the never_users runtime option,
3936 but it cannot be overridden. The default setting is "root".
3938 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3939 prefix string with which any file named in a -C command line option
3942 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3943 is retained for -C and -D only if the caller of Exim is root. Without
3944 it, the exim user may also use -C and -D and retain privilege.
3946 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3947 command line option is disabled.
3949 6. Macro names set by the -D option must start with an upper case letter, just
3950 like macro names defined in the configuration file.
3952 7. Added "dereference=" facility to LDAP.
3954 8. Two instances of the typo "uknown" in the source files are fixed.
3956 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3957 the Configure-Makefile script screwed up while processing it.
3959 10. Incorporated PCRE 4.4.
3961 11. The SMTP synchronization check was not operating right at the start of an
3962 SMTP session. For example, it could not catch a HELO sent before the client
3963 waited for the greeting. There is now a check for outstanding input at the
3964 point when the greeting is written. Because of the duplex, asynchronous
3965 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3966 way, but not yet received, when the check is performed.
3968 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3969 on TCP/IP sockets, because this apparently causes some broken clients to
3972 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3973 unchanged) from the Cygwin maintainer.
3975 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3976 was supported instead of testing for mbx. Effectively a typo.
3978 15. The spa authenticator server code was not checking that the input it
3979 received was valid base64.
3981 16. The debug output line for the "set" modifier in ACLs was not showing the
3982 name of the variable that was being set.
3984 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3986 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3987 lookup on its own. Something else was needed to trigger the lookup. For
3988 example, a match in host_lookup or the need for a host name in a host list.
3989 Now, if $sender_host_name is referenced and the host name has not yet been
3990 looked up, a lookup is performed. If the lookup fails, the variable remains
3991 empty, and $host_lookup_failed is set to "1".
3993 19. Added "eqi" as a case-independent comparison operator.
3995 20. The saslauthd authentication condition could segfault if neither service
3996 nor realm was specified.
3998 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3999 error message that was logged was misleading, and incoming SMTP
4000 connections were dropped. The message is now more accurate, and temporary
4001 errors are given to SMTP connections.
4003 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4004 (except RSET) with a 421 error, until QUIT is received. However, it was
4005 failing to send a response to QUIT.
4007 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4008 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4009 ACL is now run after the helo_try_verify_hosts code.
4011 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4012 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4013 case-independent, so other case variants are also recognized). Apparently
4014 some systems use these upper case variants.
4016 25. If more than two messages were waiting for the same host, and a transport
4017 filter was specified for the transport, Exim sent two messages over the
4018 same TCP/IP connection, and then failed with "socket operation on non-
4019 socket" when it tried to send the third.
4021 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4023 27. The extern definition of crypt16() in expand.c was not being excluded when
4024 the OS had its own crypt16() function.
4026 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4027 as a preferred synonym for return_size_limit, both as an option and as an
4030 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4032 30. Changed the default configuration ACL to relax the local part checking rule
4033 for addresses that are not in any local domains. For these addresses,
4034 slashes and pipe symbols are allowed within local parts, but the sequence
4035 /../ is explicitly forbidden.
4037 31. SPA server authentication was not clearing the challenge buffer before
4040 32. log_message in a "warn" ACL statement was writing to the reject log as
4041 well as to the main log, which contradicts the documentation and doesn't
4042 seem right (because no rejection is happening). So I have stopped it.
4044 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4045 However, I am unable to do any testing of this.
4047 34. Fixed an infelicity in the appendfile transport. When checking directories
4048 for a mailbox, to see if any needed to be created, it was accidentally
4049 using path names with one or more superfluous leading slashes; tracing
4050 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4052 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4053 contents are added to the log line that is written for every discarded
4054 recipient. (Previously a log_message setting was ignored.)
4056 36. The ${quote: operator now quotes the string if it is empty.
4058 37. The install script runs exim in order to find its version number. If for
4059 some reason other than non-existence or emptiness, which it checks, it
4060 could not run './exim', it was installing it with an empty version number,
4061 i.e. as "exim-". This error state is now caught, and the installation is
4064 38. An argument was missing from the function that creates an error message
4065 when Exim fails to connect to the socket for saslauthd authentication.
4066 This could cause Exim to crash, or give a corrupted message.
4068 39. Added isip, isip4, and isip6 to ${if conditions.
4070 40. The ACL variables $acl_xx are now saved with the message, and can be
4071 accessed later in routers, transports, and filters.
4073 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4074 strings in the file are not string-expanded.
4076 42. If a MAIL command specified a SIZE value that was too large to fit into an
4077 int variable, the check against message_size_limit failed. Such values are
4078 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4079 day this will have to be increased, but I don't think I want to be around
4080 when emails are that large.
4087 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4088 iconv() is not standard in FreeBSD.
4090 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4091 IPv6 enabled. The observed symptom was a segmentation fault on return from
4092 the function os_common_find_running_interfaces() in src/os.c.
4094 3. In the check_special_case() function in daemon.c I had used "errno" as an
4095 argument name, which causes warnings on some systems. This was basically a
4096 typo, since it was named "eno" in the comments!
4098 4. The code that waits for the clock to tick (at a resolution of some fraction
4099 of a second) so as to ensure message-id uniqueness was always waiting for
4100 at least one whole tick, when it could have waited for less. [This is
4101 almost certainly not relevant at current processor speeds, where it is
4102 unlikely to ever wait at all. But we try to future-proof.]
4104 5. The function that sleeps for a time interval that includes fractions of a
4105 second contained a race. It did not block SIGALRM between setting the
4106 timer, and suspending (a couple of lines later). If the interval was short
4107 and the sigsuspend() was delayed until after it had expired, the suspension
4108 never ended. On busy systems this could lead to processes getting stuck for
projects / exim.git / blob