The word &"set"& at the start of a line, followed by a single space,
is recognised specially as defining a value for a variable.
+.new
+.cindex "tainted data" "expansion testing"
+If the sequence &",t"& is inserted before the space,
+the value is marked as tainted.
+.wen
The syntax is otherwise the same as the ACL modifier &"set ="&.
.cmdopt -bem <&'filename'&>
For the string-expansion kind of lookups, the query is given in the first
bracketed argument of the &${lookup ...}$& expansion.
-For the list-argument kind of lookup the quury is given by the remainder of the
+For the list-argument kind of lookup the query is given by the remainder of the
list item after the first semicolon.
.cindex "tainted data" "quoting for lookups"
-If tainted data is used in the query then it should be quuted by
+If tainted data is used in the query then it should be quoted by
using the &*${quote_*&<&'lookup-type'&>&*:*&<&'string'&>&*}*& expansion operator
appropriate for the lookup.
.endlist
lookup types support only literal keys.
&*Warning 2*&: In a host list, you must always use &(net-iplsearch)& so that
-the implicit key is the host's IP address rather than its name (see section
-&<<SECThoslispatsikey>>&).
+the implicit key is the host's IP address rather than its name
+(see section &<<SECThoslispatsikey>>&).
&*Warning 3*&: Do not use an IPv4-mapped IPv6 address for a key; use the
IPv4, in dotted-quad form. (Exim converts IPv4-mapped IPv6 addresses to this
${lookup pgsql,servers=master/db/name/pw {UPDATE ...} }
.endd
-An older syntax places the servers specification before the query,
+.new
+A now-deprecated syntax places the servers specification before the query,
semicolon separated:
.code
${lookup mysql{servers=master; UPDATE ...} }
.endd
-The new version avoids potential issues with tainted
-arguments in the query, for explicit expansion.
+The new version avoids issues with tainted
+arguments explicitly expanded as part of the query.
+The entire string within the braces becomes tainted,
+including the server sepcification - which is not permissible.
+If the older sytax is used, a warning message will be logged.
+This syntax will be removed in a future release.
+.wen
+
&*Note*&: server specifications in list-style lookups are still problematic.
.section "Expansion of lists" "SECTlistexpand"
.cindex "expansion" "of lists"
Each list is expanded as a single string before it is used.
+.cindex "tainted data" tracking
+&*Note*&: As a result, if any componend was tainted then the
+entire result string becomes tainted.
&'Exception: the router headers_remove option, where list-item
splitting is done before string-expansion.'&
&*Reminder*&: With this kind of pattern, you must have host &'names'& as
keys in the file, not IP addresses. If you want to do lookups based on IP
-addresses, you must precede the search type with &"net-"& (see section
-&<<SECThoslispatsikey>>&). There is, however, no reason why you could not use
+addresses, you must precede the search type with &"net-"&
+(see section &<<SECThoslispatsikey>>&).
+There is, however, no reason why you could not use
two items in the same list, one doing an address lookup and one doing a name
lookup, both using the same file.
below in section &<<SECTexpansionitems>>& onwards. Backslash is used as an
escape character, as described in the following section.
+.cindex "tainted data" tracking
+If any porttion of the result string is tainted, the entire result is.
+
Whether a string is expanded depends upon the context. Usually this is solely
dependent upon the option for which a value is sought; in this documentation,
options for which string expansion is performed are marked with † after
.next
The item @[] matches any of the local host's interface addresses.
.next
-Single-key lookups are assumed to be like &"net-"& style lookups in host lists,
+Single-key lookups are assumed to be like &"net-"& style lookups in host lists
+(see section &<<SECThoslispatsikey>>&),
even if &`net-`& is not specified. There is never any attempt to turn the IP
address into a host name. The most common type of linear search for
&*match_ip*& is likely to be &*iplsearch*&, in which the file can contain CIDR
.row &%dsn_advertise_hosts%& "advertise DSN extensions to these hosts"
.row &%ignore_fromline_hosts%& "allow &""From ""& from these hosts"
.row &%ignore_fromline_local%& "allow &""From ""& from local SMTP"
+.row &%limits_advertise_hosts%& "advertise LIMITS to these hosts"
.row &%pipelining_advertise_hosts%& "advertise pipelining to these hosts"
.row &%pipelining_connect_advertise_hosts%& "advertise pipelining to these hosts"
.row &%prdr_enable%& "advertise PRDR to all hosts"
See section &<<SECDKIMVFY>>&.
-.option dmarc_forensic_sender main string&!! unset
-.option dmarc_history_file main string unset
-.option dmarc_tld_file main string unset
+.option dmarc_forensic_sender main string&!! unset &&&
+ dmarc_history_file main string unset &&&
+ dmarc_tld_file main string unset
.cindex DMARC "main section options"
These options control DMARC processing.
See section &<<SECDMARC>>& for details.
+.new
+.option limits_advertise_hosts main "host list&!!" *
+.cindex LIMITS "suppressing advertising"
+.cindex "ESMTP extensions" LIMITS
+This option can be used to suppress the advertisement of the SMTP
+LIMITS extension (RFC 9422) to specific hosts.
+If permitted, Exim as a servier will advertise in the EHLO response
+the limit for RCPT commands set by the &%recipients_max%& option (if it is set)
+and the limit for MAIL commands set by the &%smtp_accept_max_per_connection%&
+option.
+,wen
+
.option local_from_check main boolean true
.cindex "&'Sender:'& header line" "disabling addition of"
.cindex "&'From:'& header line" "disabling checking of"
.option localhost_number main string&!! unset
.cindex "host" "locally unique number for"
.cindex "message ids" "with multiple hosts"
+.cindex multiple "systems sharing a spool"
+.cindex "multiple hosts" "sharing a spool"
+.cindex "shared spool directory"
+.cindex "spool directory" sharing
.vindex "&$localhost_number$&"
Exim's message ids are normally unique only within the local host. If
-uniqueness among a set of hosts is required, each host must set a different
+uniqueness among a set of hosts is required
+(eg. because they share a spool directory),
+each host must set a different
value for the &%localhost_number%& option. The string is expanded immediately
after reading the configuration file (so that a number can be computed from the
host name, for example) and the result of the expansion must be a number in the
range 0&--16 (or 0&--10 on operating systems with case-insensitive file
systems). This is available in subsequent string expansions via the variable
-&$localhost_number$&. When &%localhost_number is set%&, the final two
+&$localhost_number$&. When &%localhost_number%& is set, the final four
characters of the message id, instead of just being a fractional part of the
time, are computed from the time and the local host number as described in
section &<<SECTmessiden>>&.
option was not set.
-.option recipients_max main integer 50000
+.option recipients_max main integer&!! 50000
.cindex "limit" "number of recipients"
.cindex "recipient" "maximum number"
-If this option is set greater than zero, it specifies the maximum number of
+If the value resulting from expanding this option
+is set greater than zero, it specifies the maximum number of
original recipients for any message. Additional recipients that are generated
by aliasing or forwarding do not count. SMTP messages get a 452 response for
all recipients over the limit; earlier recipients are delivered as normal.
Non-SMTP messages with too many recipients are failed, and no deliveries are
done.
+.new
+For SMTP message the expansion is done after the connection is
+accepted (but before any SMTP conversation) and may depend on
+the IP addresses and port numbers of the connection.
+&*Note*&: If an expansion is used for the option,
+care should be taken that a resonable value results for
+non-SMTP messages.
+.wen
+
.cindex "RCPT" "maximum number of incoming"
&*Note*&: The RFCs specify that an SMTP server should accept at least 100
RCPT commands in a single message.
For testing purposes, this value can be overridden by the &%-oB%& command line
option.
+.new
+.cindex "ESMTP extensions" LIMITS
+If the peer advertises a LIMITS extension with a MAILMAX value,
+and either TLSS is in use or was not advertised,
+that value also constrains the result of this option.
+.wen
+
.option dane_require_tls_ciphers smtp string&!! unset
.cindex "TLS" "requiring specific ciphers for DANE"
so can cause parallel connections to the same host if &%remote_max_parallel%&
permits this.
+.new
+.cindex "ESMTP extensions" LIMITS
+If the peer advertises a LIMITS extension with a RCPTMAX value,
+and either TLSS is in use or was not advertised,
+that value also constrains the result of this option
+and no parallel connections will be caused on meeting the RCPTMAX limit.
+.wen
+
.option message_linelength_limit smtp integer 998
.cindex "line length" limit
This option sets the maximum line length, in bytes, that the transport
will send. Any messages with lines exceeding the given value
+(before a transport filter, if any)
will fail and a failure-DSN ("bounce") message will if possible be returned
to the sender.
The default value is that defined by the SMTP standards.
only messages having the domain used for the DANE TLSA lookup are
sent on the connection.
+.new
+.cindex "ESMTP extensions" LIMITS
+If the peer advertises a LIMITS extension with a RCPTDOMAINMAX value,
+and either TLSS is in use or was not advertised,
+this option is regarded as being false.
+.wen
+
+
.option port smtp string&!! "see below"
.cindex "port" "sending TCP/IP"
.cindex "TCP/IP" "setting outgoing port"
.subsection General
Under GnuTLS, DANE is only supported from version 3.0.0 onwards.
-DANE is specified in published RFCs and decouples certificate authority trust
+DANE is specified in RFC 6698. It decouples certificate authority trust
selection from a "race to the bottom" of "you must trust everything for mail
to get through".
-There is an alternative technology called MTA-STS, which
-instead publishes MX trust anchor information on an HTTPS website. At the
-time this text was last updated, MTA-STS was still a draft, not yet an RFC.
+It does retain the need to trust the assurances provided by the DNSSEC tree.
+
+There is an alternative technology called MTA-STS (RFC 8461), which
+instead publishes MX trust anchor information on an HTTPS website.
+The discovery of the address for that website does not (per standard)
+require DNSSEC, and could be regarded as being less secure than DANE
+as a result.
+
Exim has no support for MTA-STS as a client, but Exim mail server operators
can choose to publish information describing their TLS configuration using
MTA-STS to let those clients who do use that protocol derive trust
The &%acl_smtp_dkim%& ACL is available only when Exim is compiled with DKIM support
enabled (which is the default).
+If, for a specific message, an ACL control
+&*dkim_disable_verify*&
+has been set, this &%acl_smtp_dkim%& ACL is not called.
+
The ACL test specified by &%acl_smtp_dkim%& happens after a message has been
received, and is executed for each DKIM signature found in a message. If not
otherwise specified, the default action is to accept.
.section "Format of an ACL" "SECID199"
.cindex "&ACL;" "format of"
.cindex "&ACL;" "verbs, definition of"
-An individual ACL consists of a number of statements. Each statement starts
+An individual ACL definition consists of a number of statements.
+Each statement starts
with a verb, optionally followed by a number of conditions and &"modifiers"&.
Modifiers can change the way the verb operates, define error and log messages,
set variables, insert delays, and vary the processing of accepted messages.
all the conditions make sense at every testing point. For example, you cannot
test a sender address in the ACL that is run for a VRFY command.
+The definition of an ACL ends where another starts,
+or a different configuration section starts.
+
.section "ACL verbs" "SECID200"
The ACL verbs are as follows:
follows:
.ilist
-LF not preceded by CR is treated as a line ending.
-.next
CR is treated as a line ending; if it is immediately followed by LF, the LF
is ignored.
.next
.next
If the first header line received in a message ends with CRLF, a subsequent
bare LF in a header line is treated in the same way as a bare CR in a header
-line.
+line and a bare LF in a body line is replaced with a space.
+.next
+If the first header line received in a message does not end with CRLF, a subsequent
+LF not preceded by CR is treated as a line ending.
.endlist
.irow &`deliver_time`& "time taken to attempt delivery"
.irow &`delivery_size`& "add &`S=`&&'nnn'& to => lines"
.irow &`dkim`& * "DKIM verified domain on <= lines"
-.irow &`dkim_verbose`& "separate full DKIM verification result line, per signature"
+.irow &`dkim_verbose`& "separate full DKIM verification result line, per signature; DKIM signing"
.irow &`dnslist_defer`& * "defers of DNS list (aka RBL) lookups"
.irow &`dnssec`& "DNSSEC secured lookups"
.irow &`etrn`& * "ETRN commands"
.irow &`tls_peerdn`& "TLS peer DN on <= and => lines"
.irow &`tls_resumption`& "append * to cipher field"
.irow &`tls_sni`& "TLS SNI on <= lines"
-.irow &`unknown_in_list`& "DNS lookup failed in list match"
+.irow &`unknown_in_list`& "lookup failed in list match"
.irow &`all`& "&*all of the above*&"
.endtable
See also the &%slow_lookup_log%& main configuration option,
.cindex log "DKIM verification"
.cindex DKIM "verification logging"
&%dkim_verbose%&: A log entry is written for each attempted DKIM verification.
+.new
+Also, on message delivery lines signing information (domain and selector)
+is added, tagged with DKIM=.
+.wen
.next
.cindex "log" "dnslist defer"
.cindex "DNS list" "logging defer"
.next
.cindex "log" "DNS failure in list"
&%unknown_in_list%&: This setting causes a log entry to be written when the
-result of a list match is failure because a DNS lookup failed.
+result of a list match is failure because a DNS lookup failed, or because
+a bad IP address was in the list.
.endlist
.olist
Signing outgoing messages: This function is implemented in the SMTP transport.
It can co-exist with all other Exim features
-(including transport filters)
-except cutthrough delivery.
+(including transport filters) except cutthrough delivery.
+.new
+However, signing options may not depend on headers modified by
+routers, the transport or a transport filter.
+.wen
.next
Verifying signatures in incoming messages: This is implemented by an additional
ACL (acl_smtp_dkim), which can be called several times per message, with
.option dkim_sign_headers smtp string&!! "see below"
If set, this option must expand to a colon-separated
list of header names.
-Headers with these names, or the absence or such a header, will be included
+Headers with these names, or the absence of such a header, will be included
in the message signature.
When unspecified, the header names listed in RFC4871 will be used,
whether or not each header is present in the message.
.option dkim_timestamps smtp integer&!! unset
This option controls the inclusion of timestamp information in the signature.
If not set, no such information will be included.
-Otherwise, must be an unsigned number giving an offset in seconds from the current time
-for the expiry tag
-(eg. 1209600 for two weeks);
-both creation (t=) and expiry (x=) tags will be included.
+.new
+Otherwise, must be an unsigned number giving an offset in seconds from the
+current time for the expiry tag (e.g. 1209600 for two weeks); both creation
+(t=) and expiry (x=) tags will be included unless the offset is 0 (no expiry).
+.wen
RFC 6376 lists these tags as RECOMMENDED.
.subsection "SRS (Sender Rewriting Scheme)" SECTSRS
.cindex SRS "sender rewriting scheme"
+.cindex VERP "variable envelope return path"
SRS can be used to modify sender addresses when forwarding so that
SPF verification does not object to them.
-It operates by encoding the original envelope sender in a new
+It can also be used to identify a received bounce message as
+likely (or not) having been trigged by a message from the
+local system, and for identifying dead addresses in mailing lists.
+It is one implementation of a VERP (Variable Envelope Return Path) method.
+
+SRS operates by encoding the original envelope sender in a new
sender local part and using a domain run by the forwarding site
as the new domain for the sender. Any DSN message should be returned
to this new sender at the forwarding site, which can extract the
allow_fail
data = :fail: Invalid SRS recipient address
- #... further routers here
+ #... further routers here get inbound_srs-redirected recipients
+ # and any that were not SRS'd
# transport; should look like the non-forward outbound
The name is placed in the variable &$event_name$& and the event action
expansion must check this, as it will be called for every possible event type.
+.new
The current list of events is:
+.wen
.itable all 0 0 4 25* left 10* center 15* center 50* left
.row auth:fail after both "per driver per authentication attempt"
.row dane:fail after transport "per connection"
+.row dns:fail after both "per lookup"
.row msg:complete after main "per message"
.row msg:defer after transport "per message per delivery try"
.row msg:delivery after transport "per recipient"
.itable all 0 0 2 20* left 80* left
.row auth:fail "smtp response"
.row dane:fail "failure reason"
+.row dns:fail "failure reason, key and lookup-type"
.row msg:defer "error string"
.row msg:delivery "smtp confirmation message"
.row msg:fail:internal "failure reason"
For OpenSSL it will trigger for every chain element including those
loaded locally.
+.new
+For dns:fail events from dnsdb lookups, a &"defer_never"& option does not
+affect the reporting of DNS_AGAIN.
+.wen
+
. ////////////////////////////////////////////////////////////////////////////
. ////////////////////////////////////////////////////////////////////////////
git://git.exim.org / exim.git / blobdiff