X-Git-Url: https://git.exim.org/users/jgh/exim.git/blobdiff_plain/14a806d6c13afdfb2f44dce64e50bffa6cb6869c..bd95ffc2ba87fbd3c752df17bc8fd9c01586d45a:/doc/doc-docbook/spec.xfpt

diff --git a/doc/doc-docbook/spec.xfpt b/doc/doc-docbook/spec.xfpt
index eea304d64..0d22aaefa 100644
--- a/doc/doc-docbook/spec.xfpt
+++ b/doc/doc-docbook/spec.xfpt
@@ -3966,7 +3966,7 @@ user.
 .vitem &%-MG%&&~<&'queue&~name'&>&~<&'message&~id'&>&~<&'message&~id'&>&~...
 .oindex "&%-MG%&"
 .cindex queue named
-.cindex "named queues"
+.cindex "named queues" "moving messages"
 .cindex "queue" "moving messages"
 This option requests that each listed message be moved from its current
 queue to the given named queue.
@@ -4185,6 +4185,7 @@ forces queueing.
 .vitem &%-odqs%&
 .oindex "&%-odqs%&"
 .cindex "SMTP" "delaying delivery"
+.cindex "first pass routing"
 This option is a hybrid between &%-odb%&/&%-odi%& and &%-odq%&.
 However, like &%-odb%& and &%-odi%&, this option has no effect if
 &%queue_only_override%& is false and one of the queueing options in the
@@ -4503,11 +4504,16 @@ appear in the correct order. Each flag is described in a separate item below.
 .cindex "queue" "double scanning"
 .cindex "queue" "routing"
 .cindex "routing" "whole queue before delivery"
+.cindex "first pass routing"
 An option starting with &%-qq%& requests a two-stage queue run. In the first
 stage, the queue is scanned as if the &%queue_smtp_domains%& option matched
 every domain. Addresses are routed, local deliveries happen, but no remote
 transports are run.
 
+.new
+Performance will be best if the &%queue_run_in_order%& option is false.
+.wen
+
 .cindex "hints database" "remembering routing"
 The hints database that remembers which messages are waiting for specific hosts
 is updated, as if delivery to those hosts had been deferred. After this is
@@ -4553,7 +4559,7 @@ for later delivery.
 .vitem &%-q[q][i][f[f]][l][G<name>[/<time>]]]%&
 .oindex "&%-qG%&"
 .cindex queue named
-.cindex "named queues"
+.cindex "named queues"	"deliver from"
 .cindex "queue" "delivering specific messages"
 If the &'G'& flag and a name is present, the queue runner operates on the
 queue with the given name rather than the default queue.
@@ -6362,7 +6368,7 @@ All other options are defaulted.
 .code
 local_delivery:
   driver = appendfile
-  file = /var/mail/$local_part
+  file = /var/mail/$local_part_verified
   delivery_date_add
   envelope_to_add
   return_path_add
@@ -6370,7 +6376,17 @@ local_delivery:
 # mode = 0660
 .endd
 This &(appendfile)& transport is used for local delivery to user mailboxes in
-traditional BSD mailbox format. By default it runs under the uid and gid of the
+traditional BSD mailbox format.
+
+.new
+We prefer to avoid using &$local_part$& directly to define the mailbox filename,
+as it is provided by a potential bad actor.
+Instead we use &$local_part_verified$&,
+the result of looking up &$local_part$& in the user database
+(done by using &%check_local_user%& in the the router).
+.wen
+
+By default &(appendfile)& runs under the uid and gid of the
 local user, which requires the sticky bit to be set on the &_/var/mail_&
 directory. Some systems use the alternative approach of running mail deliveries
 under a particular group instead of using the sticky bit. The commented options
@@ -8290,6 +8306,19 @@ domainlist  dom2 = !a.b : *.b
 where &'x.y'& does not match. It's best to avoid negation altogether in
 referenced lists if you can.
 
+.new
+.cindex "hiding named list values"
+.cindex "named lists" "hiding value of"
+Some named list definitions may contain sensitive data, for example, passwords for
+accessing databases. To stop non-admin users from using the &%-bP%& command
+line option to read these values, you can precede the definition with the
+word &"hide"&. For example:
+.code
+hide domainlist filter_for_domains = ldap;PASS=secret ldap::/// ...
+.endd
+.wen
+
+
 Named lists may have a performance advantage. When Exim is routing an
 address or checking an incoming message, it caches the result of tests on named
 lists. So, if you have a setting such as
@@ -9709,7 +9738,7 @@ letters appear. For example:
        "&*$bheader_*&<&'header&~name'&>&*:*&&~or&~&&&
         &*$bh_*&<&'header&~name'&>&*:*&" &&&
        "&*$lheader_*&<&'header&~name'&>&*:*&&~or&~&&&
-        &*$lh_*&<&'header&~name'&>&*:*&"
+        &*$lh_*&<&'header&~name'&>&*:*&" &&&
        "&*$rheader_*&<&'header&~name'&>&*:*&&~or&~&&&
         &*$rh_*&<&'header&~name'&>&*:*&"
 .cindex "expansion" "header insertion"
@@ -9815,6 +9844,12 @@ newline at the very end. For the &%header%& and &%bheader%& expansion, for
 those headers that contain lists of addresses, a comma is also inserted at the
 junctions between headers. This does not happen for the &%rheader%& expansion.
 
+.new
+.cindex "tainted data"
+When the headers are from an incoming message,
+the result of expanding any of these variables is tainted.
+.wen
+
 
 .vitem &*${hmac{*&<&'hashname'&>&*}{*&<&'secret'&>&*}{*&<&'string'&>&*}}*&
 .cindex "expansion" "hmac hashing"
@@ -12192,6 +12227,13 @@ When the &%smtp_etrn_command%& option is being expanded, &$domain$& contains
 the complete argument of the ETRN command (see section &<<SECTETRN>>&).
 .endlist
 
+.new
+.cindex "tainted data"
+If the origin of the data is an incoming message,
+the result of expanding this variable is tainted.
+See also &$domain_verified$&.
+.wen
+
 
 .vitem &$domain_data$&
 .vindex "&$domain_data$&"
@@ -12385,6 +12427,25 @@ Global address rewriting happens when a message is received, so the value of
 because a message may have many recipients and the system filter is called just
 once.
 
+.new
+.cindex "tainted data"
+If the origin of the data is an incoming message,
+the result of expanding this variable is tainted.
+
+&*Warning*&: the content of this variable is usually provided by a potential
+attacker.
+Consider carefully the implications of using it unvalidated as a name
+for file access.
+This presents issues for users' &_.forward_& and filter files.
+For traditional full user accounts, use &%check_local_users%& and the
+&$local_part_verified$& variable rather than this one.
+For virtual users, store a suitable pathname component in the database
+which is used for account name validation, and use that retrieved value
+rather than this variable.
+If needed, use a router &%address_data%& or &%set%& option for
+the retrieved data.
+.wen
+
 .vindex "&$local_part_prefix$&"
 .vindex "&$local_part_suffix$&"
 .cindex affix variables
@@ -12451,6 +12512,14 @@ When an address is being routed or delivered, and a
 specific suffix for the local part was recognized, it is available in this
 variable, having been removed from &$local_part$&.
 
+.new
+.vitem &$local_part_verified$&
+.vindex "&$local_part_verified$&"
+If the router generic option &%check_local_part%& has run successfully,
+this variable has the user database version of &$local_part$&.
+Such values are not tainted and hence usable for building file names.
+.wen
+
 .vitem &$local_scan_data$&
 .vindex "&$local_scan_data$&"
 This variable contains the text returned by the &[local_scan()]& function when
@@ -12781,7 +12850,7 @@ or if not set, the value of &$qualify_domain$&.
 
 .vitem &$queue_name$&
 .vindex &$queue_name$&
-.cindex "named queues"
+.cindex "named queues" variable
 .cindex queues named
 The name of the spool queue in use; empty for the default queue.
 
@@ -16696,13 +16765,14 @@ run. If you do not want queue runs to occur, omit the &%-q%&&'xx'& setting on
 the daemon's command line.
 
 .cindex queues named
-.cindex "named queues"
+.cindex "named queues" "resource limit"
 To set limits for different named queues use
 an expansion depending on the &$queue_name$& variable.
 
 .option queue_smtp_domains main "domain list&!!" unset
 .cindex "queueing incoming messages"
 .cindex "message" "queueing remote deliveries"
+.cindex "first pass routing"
 When this option is set, a delivery process is started whenever a message is
 received, routing is performed, and local deliveries take place.
 However, if any SMTP deliveries are required for domains that match
@@ -18584,15 +18654,19 @@ avoided. The &%repeat_use%& option of the &%redirect%& router may be of help.
 This option specifies a list of text headers,
 colon-separated (by default, changeable in the usual way &<<SECTlistsepchange>>&),
 that is associated with any addresses that are accepted by the router.
-Each item is separately expanded, at routing time.  However, this
-option has no effect when an address is just being verified. The way in which
+However, the option has no effect when an address is just being verified.
+Each list item is separately expanded, at transport time.
+.new
+If an item ends in *, it will match any header with the given prefix.
+.wen
+The way in which
 the text is used to remove header lines at transport time is described in
 section &<<SECTheadersaddrem>>&. Header lines are not actually removed until
 the message is in the process of being transported. This means that references
 to header lines in string expansions in the transport's configuration still
 &"see"& the original header lines.
 
-The &%headers_remove%& option is expanded after &%errors_to%& and
+The &%headers_remove%& option is handled after &%errors_to%& and
 &%headers_add%&, but before &%transport%&. If an item expansion is forced to fail,
 the item has no effect. Other expansion failures are treated as configuration
 errors.
@@ -20528,6 +20602,19 @@ is not the case when the file contains syntactically valid items that happen to
 yield empty addresses, for example, items containing only RFC 2822 address
 comments.
 
+.new
+.cindex "tainted data" "in filenames"
+.cindex redirect "tainted data"
+Tainted data may not be used for a filename.
+
+&*Warning*&: It is unwise to use &$local_part$& or &$domain$&
+directly for redirection,
+as they are provided by a potential attacker.
+In the examples above, &$local_part$& is used for looking up data held locally
+on the system, and not used directly (the second example derives &$home$& via
+the passsword file or database, using &$local_part$&).
+.wen
+
 
 
 .section "Forward files and address verification" "SECID125"
@@ -20753,6 +20840,11 @@ It must be given as
 .code
 list1:   :include:/opt/lists/list1
 .endd
+.new
+.cindex "tainted data" "in filenames"
+.cindex redirect "tainted data"
+Tainted data may not be used for a filename.
+.wen
 .next
 .cindex "address redirection" "to black hole"
 .cindex "delivery" "discard"
@@ -21753,15 +21845,21 @@ checked, since this option does not automatically suppress them.
 .option headers_remove transports list&!! unset
 .cindex "header lines" "removing"
 .cindex "transport" "header lines; removing"
-This option specifies a list of header names,
-colon-separated (by default, changeable in the usual way &<<SECTlistsepchange>>&);
-these headers are omitted from the message as it is transported, as described
-in section &<<SECTheadersaddrem>>&. Header removal can also be specified by
-routers.
+This option specifies a list of text headers,
+colon-separated (by default, changeable in the usual way &<<SECTlistsepchange>>&),
+to be removed from the message.
+However, the option has no effect when an address is just being verified.
 Each list item is separately expanded.
 If the result of the expansion is an empty string, or if the expansion
 is forced to fail, no action is taken. Other expansion failures are treated as
 errors and cause the delivery to be deferred.
+.new
+If an item ends in *, it will match any header with the given prefix.
+.wen
+
+Matching headers are omitted from the message as it is transported, as described
+in section &<<SECTheadersaddrem>>&. Header removal can also be specified by
+routers.
 
 Unlike most options, &%headers_remove%& can be specified multiple times
 for a transport; all listed headers are removed.
@@ -22312,6 +22410,14 @@ If &%file%& or &%directory%& is set for a delivery from a redirection, it is
 used to determine the file or directory name for the delivery. Normally, the
 contents of &$address_file$& are used in some way in the string expansion.
 .endlist
+.new
+.cindex "tainted data" "in filenames"
+.cindex appendfile "tainted data"
+Tainted data may not be used for a file or directory name.
+This means that, for instance, &$local_part$& cannot be used directly
+as a component of a path.  It can however be used as the key for a lookup
+which returns a path (or component).
+.wen
 
 
 .cindex "Sieve filter" "configuring &(appendfile)&"
@@ -23749,6 +23855,12 @@ directories are also controllable. See chapter &<<CHAPenvironment>>& for
 details of the local delivery environment and chapter &<<CHAPbatching>>&
 for a discussion of local delivery batching.
 
+.new
+.cindex "tainted data" "in pipe command"
+.cindex pipe "tainted data"
+Tainted data may not be used for the command name.
+.wen
+
 
 .section "Concurrent delivery" "SECID140"
 If two messages arrive at almost the same time, and both are routed to a pipe
@@ -26942,7 +27054,7 @@ There are good and bad examples at the end of the next section.
 
 .section "The PLAIN authentication mechanism" "SECID172"
 .cindex "PLAIN authentication mechanism"
-.cindex "authentication" "PLAIN mechanism"
+.cindex authentication PLAIN
 .cindex "binary zero" "in &(plaintext)& authenticator"
 The PLAIN authentication mechanism (RFC 2595) specifies that three strings be
 sent as one item of data (that is, one combined string containing two NUL
@@ -27024,7 +27136,7 @@ writing the test makes the logic clearer.
 
 .section "The LOGIN authentication mechanism" "SECID173"
 .cindex "LOGIN authentication mechanism"
-.cindex "authentication" "LOGIN mechanism"
+.cindex authentication LOGIN
 The LOGIN authentication mechanism is not documented in any RFC, but is in use
 in a number of programs. No data is sent with the AUTH command. Instead, a
 user name and password are supplied separately, in response to prompts. The
@@ -27145,7 +27257,7 @@ prompts.
 .scindex IIDcramauth1 "&(cram_md5)& authenticator"
 .scindex IIDcramauth2 "authenticators" "&(cram_md5)&"
 .cindex "CRAM-MD5 authentication mechanism"
-.cindex "authentication" "CRAM-MD5 mechanism"
+.cindex authentication CRAM-MD5
 The CRAM-MD5 authentication mechanism is described in RFC 2195. The server
 sends a challenge string to the client, and the response consists of a user
 name and the CRAM-MD5 digest of the challenge string combined with a secret
@@ -27434,7 +27546,7 @@ auth_mechanisms = plain login ntlm
 .cindex "authentication" "LOGIN"
 .cindex "authentication" "DIGEST-MD5"
 .cindex "authentication" "CRAM-MD5"
-.cindex "authentication" "SCRAM-SHA-1"
+.cindex "authentication" "SCRAM family"
 The &(gsasl)& authenticator provides integration for the GNU SASL
 library and the mechanisms it provides.  This is new as of the 4.80 release
 and there are a few areas where the library does not let Exim smoothly
@@ -27442,17 +27554,22 @@ scale to handle future authentication mechanisms, so no guarantee can be
 made that any particular new authentication mechanism will be supported
 without code changes in Exim.
 
-
 .new
+The library is expected to add support in an upcoming
+realease for the SCRAM-SHA-256 method.
+The macro _HAVE_AUTH_GSASL_SCRAM_SHA_256 will be defined
+when this happens.
+
+
 .option client_authz gsasl string&!! unset
 This option can be used to supply an &'authorization id'&
 which is different to the &'authentication_id'& provided
-by $%client_username%& option.
+by &%client_username%& option.
 If unset or (after expansion) empty it is not used,
 which is the common case.
 
 .option client_channelbinding gsasl boolean false
-See $%server_channelbinding%& below.
+See &%server_channelbinding%& below.
 
 .option client_password gsasl string&!! unset
 This option is exapanded before use, and should result in
@@ -27463,6 +27580,19 @@ This option is exapanded before use, and should result in
 the account name to be used.
 .wen
 
+.new
+.option client_spassword gsasl string&!! unset
+If a SCRAM mechanism is being used and this option is set
+it is used in preference to &%client_password%&.
+The value after expansion should be
+a 40 (for SHA-1) or 64 (for SHA-256) character string
+with the PBKDF2-prepared password, hex-encoded.
+Note that this value will depend on the salt and iteration-count
+supplied by the server.
+.wen
+
+
+
 .option server_channelbinding gsasl boolean false
 Do not set this true and rely on the properties
 without consulting a cryptographic engineer.
@@ -27481,6 +27611,7 @@ server to see different identifiers and authentication will fail.
 This is
 only usable by mechanisms which support "channel binding"; at time of
 writing, that's the SCRAM family.
+When using this feature the "-PLUS" variants of the method names need to be used.
 .wen
 
 This defaults off to ensure smooth upgrade across Exim releases, in case
@@ -27535,16 +27666,60 @@ This specifies the SASL realm that the server claims to be in.
 Some mechanisms will use this data.
 
 
-.option server_scram_iter gsasl string&!! unset
+.option server_scram_iter gsasl string&!! 4096
 This option provides data for the SCRAM family of mechanisms.
-&$auth1$& is not available at evaluation time.
-(This may change, as we receive feedback on use)
-
+.new
+The &$auth1$&, &$auth2$& and &$auth3$& variables are available
+when this option is expanded.
+
+The result of expansion should be a decimal number,
+and represents both a lower-bound on the security, and
+a compute cost factor imposed on the client
+(if it does not cache results, or the server changes
+either the iteration count or the salt).
+A minimum value of 4096 is required by the standards
+for all current SCRAM mechanism variants.
+.wen
 
 .option server_scram_salt gsasl string&!! unset
 This option provides data for the SCRAM family of mechanisms.
-&$auth1$& is not available at evaluation time.
-(This may change, as we receive feedback on use)
+.new
+The &$auth1$&, &$auth2$& and &$auth3$& variables are available
+when this option is expanded.
+The value should be a base64-encoded string,
+of random data typically 4-to-16 bytes long.
+If unset or empty after expansion the library will provides a value for the
+protocol conversation.
+.wen
+
+
+.new
+.option server_key gsasl string&!! unset
+.option server_skey gsasl string&!! unset
+These options can be used for the SCRAM family of mechanisms
+to provide stored information related to a password,
+the storage of which is preferable to plaintext.
+
+&%server_key%& is the value defined in the SCRAM standards as ServerKey;
+&%server_skey%& is StoredKey.
+
+They are only available for version 1.9.0 (or later) of the gsasl library.
+When this is so, the macros
+_OPT_AUTHENTICATOR_GSASL_SERVER_KEY
+and _HAVE_AUTH_GSASL_SCRAM_S_KEY
+will be defined.
+
+The &$authN$& variables are available when these options are expanded.
+
+If set, the results of expansion should for each
+should be a 28 (for SHA-1) or 44 (for SHA-256) character string
+of base64-coded data, and will be used in preference to the
+&%server_password%& option.
+If unset or not of the right length, &%server_password%& will be used.
+
+The libgsasl library release includes a utility &'gsasl'& which can be used
+to generate these values.
+.wen
 
 
 .option server_service gsasl string &`smtp`&
@@ -29976,7 +30151,7 @@ in several different ways. For example:
 It can be at the end of an &%accept%& statement:
 .code
     accept  ...some conditions
-            control = queue_only
+            control = queue
 .endd
 In this case, the control is applied when this statement yields &"accept"&, in
 other words, when the conditions are all true.
@@ -29985,7 +30160,7 @@ other words, when the conditions are all true.
 It can be in the middle of an &%accept%& statement:
 .code
     accept  ...some conditions...
-            control = queue_only
+            control = queue
             ...some more conditions...
 .endd
 If the first set of conditions are true, the control is applied, even if the
@@ -30556,16 +30731,32 @@ response to an EHLO command. Therefore, it should normally appear in an ACL
 controlled by &%acl_smtp_connect%& or &%acl_smtp_helo%&. See also
 &%pipelining_advertise_hosts%&.
 
-.vitem &*control&~=&~queue_only*&
+.new
+.vitem &*control&~=&~queue/*&<&'options'&>* &&&
+       &*control&~=&~queue_only*&
+.oindex "&%queue%&"
 .oindex "&%queue_only%&"
 .cindex "queueing incoming messages"
+.cindex queueing "forcing in ACL"
+.cindex "first pass routing"
 This control is permitted only for the MAIL, RCPT, DATA, and non-SMTP ACLs, in
 other words, only when a message is being received. If the message is accepted,
 it is placed on Exim's queue and left there for delivery by a subsequent queue
-runner. No immediate delivery process is started. In other words, it has the
-effect as the &%queue_only%& global option. However, the control applies only
-to the current message, not to any subsequent ones that may be received in the
-same SMTP connection.
+runner.
+If used with no options set,
+no immediate delivery process is started. In other words, it has the
+effect as the &%queue_only%& global option or &'-odq'& command-line option.
+
+If the &'first_pass_route'& option is given then
+the behaviour is like the command-line &'-oqds'& option;
+a delivery process is started which stops short of making
+any SMTP delivery.  The benefit is that the hints database will be updated for
+the message being waiting for a specific host, and a later queue run will be
+able to send all such messages on a single connection.
+
+The control only applies to the current message, not to any subsequent ones that
+ may be received in the same SMTP connection.
+.wen
 
 .vitem &*control&~=&~submission/*&<&'options'&>
 .cindex "message" "submission"
@@ -33716,13 +33907,14 @@ in &_Local/Makefile_& (see section &<<SECTconoptloc>>& below).
 .cindex &%dlfunc%& "API description"
 You must include this line near the start of your code:
 .code
+#define LOCAL_SCAN
 #include "local_scan.h"
 .endd
 This header file defines a number of variables and other values, and the
 prototype for the function itself. Exim is coded to use unsigned char values
 almost exclusively, and one of the things this header defines is a shorthand
 for &`unsigned char`& called &`uschar`&.
-It also contains the following macro definitions, to simplify casting character
+It also makes available the following macro definitions, to simplify casting character
 strings and pointers to character strings:
 .code
 #define CS   (char *)
@@ -36698,6 +36890,7 @@ delivered immediately.
 .cindex "SMTP" "passed connection"
 .cindex "SMTP" "multiple deliveries"
 .cindex "multiple SMTP deliveries"
+.cindex "first pass routing"
 Mail waiting to be sent from an intermittently connected host will probably
 not have been routed, because without a connection DNS lookups are not
 possible. This means that if a normal queue run is done at connection time,
@@ -37355,7 +37548,7 @@ the following table:
 &`DKIM`&        domain verified in incoming message
 &`DN  `&        distinguished name from peer certificate
 &`DS  `&        DNSSEC secured lookups
-&`DT  `&        on &`=>`& lines: time taken for a delivery
+&`DT  `&        on &`=>`&, &'=='& and &'**'& lines: time taken for, or to attempt, a delivery
 &`F   `&        sender address (on delivery lines)
 &`H   `&        host name and IP address
 &`I   `&        local interface used
@@ -37453,7 +37646,7 @@ selection marked by asterisks:
 &` arguments                  `&  command line arguments
 &`*connection_reject          `&  connection rejections
 &`*delay_delivery             `&  immediate delivery delayed
-&` deliver_time               `&  time taken to perform delivery
+&` deliver_time               `&  time taken to attempt delivery
 &` delivery_size              `&  add &`S=`&&'nnn'& to => lines
 &`*dkim                       `&  DKIM verified domain on <= lines
 &` dkim_verbose               `&  separate full DKIM verification result line, per signature
@@ -40571,9 +40764,8 @@ defines the location of a text file of valid
 top level domains the opendmarc library uses
 during domain parsing. Maintained by Mozilla,
 the most current version can be downloaded
-from a link at &url(https://publicsuffix.org/list/, currently pointing
-at https://publicsuffix.org/list/public_suffix_list.dat)
-See also util/renew-opendmarc-tlds.sh script.
+from a link at &url(https://publicsuffix.org/list/public_suffix_list.dat).
+See also the util/renew-opendmarc-tlds.sh script.
 .new
 The default for the option is unset.
 If not set, DMARC processing is disabled.