In the previous implementation, the use case for autossl was assumed to
be STARTTLS with POP3 or SMTP. But there are users who use autossl with
HTTP too. The split mode was fine, but the divert mode was broken. This
change makes autossl a generic upgrade mechanism.
Also fix sslproxy line in autossl, change p to s if upgraded.
Add e2e tests for autossl in divert and split mode.
Setting srvdst.bev to NULL can be used as a flag to indicate that we
have reused the srvdst of the parent as the dst of the first child conn.
This also prevents access to srvdst.bev after we do the xfer, without
any extra flag.
The ReconnectSSL option allows rule developers to write struct filtering
rules using SNI and CN SSL specifications to override the SSL
configuration of a connection.
Otherwise, without this new option, filtering rules cannot change SSL
options using SSL filtering fields to match connections (the SSL config
in the rule would not have any effect on the server side of the matching
connection). Without ReconnectSSL, only DstIP and DstPort fields can be
used to override the SSL config of a connection.
If the ReconnectSSL option in a struct filtering rule is set, we
disconnect and free the server side of the matching SSL connection, and
reconnect it with the SSL options in the matching struct filtering rule.
This enforces the SSL config in the rule.
Do not use the ReconnectSSL option if server disconnect is not desirable
or acceptable in your case.
We have to apply the DstHost filter both (1) as early as in
pxy_conn_connect() and also (2) after user owner of the conn is
determined in srvdst connected callback functions for tcp and ssl.
Otherwise, we cannot override SSL options of conns if we don't apply it
before SSL establishment (1), and we cannot apply user auth filtering
rules if we don't apply after determining the user owner of conn (2).
This commit actually adds the same calls in the same places as they were
before the structured filtering rules were introduced.
So for example, now we have to apply filters 4x for an HTTPS conn: 2x
dsthost, 1x ssl, and 1x http.
So that we can replace the SSL/TLS configuration of the conn with the
one in the matching filtering rule. Otherwise, once the server conn is
established, we cannot change the SSL config, or would risk confusing
the SSL routines.
Now all connection oriented proxy options possible to specify
per-proxyspec or globally can be specified in structured filtering rules
to be selectively applied to connections too. One line filtering rules
can specify filter and log actions only.
For example, we can enable/disable user authentication, protocol
validation, server ssl verification, and many other options
per-connection, or configure SSL/TLS connection options per-connection.
So, now we replace the conn_opts struct of a connection's ctx with the
conn_opts struct of the matching structured filtering rule. (One line
filtering rules have a NULL conn_opts, so we first check if the
conn_opts is not NULL.)
Now, all of the 'to' site fields in filtering rules can specify a port,
not just the dstip sites.
Fix the precedence of sites in the same type of rules. For example, if
we find a match with an sni site, we should not stop searching for a
match in cn, because a matching cn site may have a higher precedence
than the matching sni site. We should apply the action of the cn site,
although sni rules have precedence over cn. The same applies to http
host and uri rules too.
Fix the precedence of dstip rules.
Improve and update unit and e2e tests accordingly.
So, for 'to' fields too, we use two separate data structures: binary
search trees (BST) for exact match and linked lists for substring match.
Now all 'from' and 'to' fields in filtering rules use these two data
structures.
To repeat, filtering rules should be written with exact matches instead
of substring matches, as much as possible. Because BST search must be
much faster than substring search over linked lists.
To repeat, we have modifed kbtree to support complex data structures in
from fields.
Also, update the unit tests accordingly.
Now the target IP address filters can use port specs too.
Refactor for code reuse, create filter_action struct used by rules,
sites, and ports.
Also, improve code and documentation.
Do not call protopassthrough_engage() twice. If we call it to apply the
deferred pass action and also set the ctx->pass flag, it will be called
again in protossl_setup_src_ssl(). So, just set the flag and leave the
rest to protossl_setup_src_ssl().
Filtering rules can enable/disable or don't change logging. If a rule
does not mention a log action, its logging should not change. So, binary
log action fields were not enough to represent those 3 possibilities,
hence we have increased the size of those fields to 2-bits.
We should obey the order of rules as they are written in the conf file,
because latter rules should be able to override the log actions of
earlier rules. So, keep the order.
We should defer pass and/or block actions as long as possible, because a
higher precedence rule in SSL filter should be able to override (cancel)
deferred pass and block actions taken by a lower precedence rule in Dst
Host filter. And in HTTP filter the same applies to deferred block
actions taken by Dst Host and SSL filters.
Also, thanks to this new deferred actions, now HTTP filter can keep
enabled divert and split modes. In other words, a higher precedence HTTP
filter rule can cancel a deferred block action set by a lower precedence
rule earlier, which was not possible before without deferred actions and
rule precedence.
And other improvements.
Now we assign precedence to each filtering rule. More specific rules
have higher precedence. So, filtering rules at lower precedence cannot
override the actions applied to a conn by filtering rules at higher
precedence.
The other precedence rules still apply.
Log actions specified in HTTP filter rules can never enable disabled
logging, because their loggers would not be initialized.
Perhaps we should initialize them in the log submit function, if they
are initialized yet.
- Match action is added to be used with log actions only, the other
filter actions can specify log actions too
- Log actions do not configure any loggers. Global loggers for
respective log actions should have been configured for those log actions
to have any effect.
- If no filter rules are defined for a proxyspec, all log actions are
enabled. Otherwise, all log actions are disabled, and filtering rules
should enable them specifically.
- Fix max number of tokens in proxyspec and filter parsers
- Fix issues with rejecting unknown args in filter rule parser
- Do not use filter_rules field of proxyspec after config finished, it
is used for filter configuration and freed afterwards
- SSL and Dst Host filters can take all of the actions.
- HTTP filter can only take block action, not divert, split, or pass.
Because, we cannot tear a conn down and reconnect its src, after the
processing of HTTP request header is complete, e.g. SSLproxy line has
already been added to its dst buffer. Also, any change in child conns
would affect listening programs too.
- The precedence of filters is as Dst Host > SSL > HTTP.
- The precedence of actions is as Divert > Split > Pass > Block. This is
only for the same type of filter.
- The precedence of match sites is as sni > cn for ssl filter and host >
uri for http filter.
For example, pass action of dst host filter is taken before split action
of ssl filter, due to the precedence order of filters.
For example, pass action of sni site is taken before split action of cn,
due to the precedence order of sites.
We now create src ssl before enabling src to be able to take divert or
split actions of SSL filter. Otherwise, we wouldn't be able to switch
between divert and split while enabling src, only pass or block action
could be taken at that stage.
Also, refactor and clean up.
Now the site field in PassSite option can have an '*' suffix to search
for a match anywhere in sni or common names. Note that this is not a
regex or wildcard search.
Previously, we only supported exact matches in sni and between slashes
in common names. This change makes it possible to cover multiple sites
in one PassSite option. In fact, without this change, certain sites
could not be added as passsite, because it was impossible to know their
subdomain names beforehand, for example *.fbcdn.net, which may have many
subdomain names in place of asterisk.
So to use substring match, append an '*' to a site name in PassSite
option (the asterisk is removed before substring search). For example,
use ".fbcdn.net*" to match all subdomains of fbcdn.net, notice the
asterisk at the end.
We also add a warning log starting with "Closing on ssl error without
passsite match" to report sites that can be added as passsite, which is
expected to help in writing PassSite rules.
Also, we now set dstaddr_str earlier in conn handling, so we can print
it in debug logs. This also helps in IDLE and EXPIRED conn logs.
This prevents unnecessary malloc and memmove calls in non-debug mode.
This change is for correctness not for speed, because it improves
conn handling only of the first packet and for non-http protos.
- Fix segfault introduced in previous commit to prevent extra eof event.
We should NULL srvdst.bev after terminating child dst xferred from
srvdst of parent, so that we don't try to access srvdst.bev. This
happens if child conn with dst xferred from parent srvdst is terminated
before parent conn.
- Fix autossl crash trying to engage passthrough mode. We cannot engage
passthrough mode in autossl, because src is already enabled. But we
shouldn't crash either. These changes are expected to fix other possible
segfaults if passthrough is engaged on eventcb of a child conn.
Otherwise, we cannot classify user if we need to issue identify_user
events, in case database is busy or locked. We should call classify_user
callback right after the user is identified.
So we introduce classify_user callback to achieve that, which fixes the
classify_user behavior for autssl proto too.
Return void in pxy_userauth
Fix typo in clasify
Add testproxy e2e tests for POP3 and SMTP protocol validation.
We have detected that POP3 and SMTP protocol validation was broken
thanks to these new testproxy e2e tests. This is yet another example why
e2e tests are important.
We don't need parent or child ids unless debugging. IDLE and EXPIRED
conn logs do not need to report ids either. Ids are useful only in
detailed debug logs.
This is necessary to uniquely identify child conns. The src fd of child
conns was possibly not unique. We use this id in debug logs only.
Also relocate the update code related with this id.
We don't need a privsep call to open a socket for child listener,
because listener port of child conns are assigned by the system, hence
are from non-privileged range above 1024.
So the open privsep socket is used only to update user atime now.
We have carried almost all conn init tasks from thrmgr to conn handling
thread. So we immediately add the conn to the conn list of its thr,
which renders both pending ssl conns list and in_thr_conns flag useless.
The only time we go over the linked list is to check idle or expired
connections, or to print debug info. Otherwise, mostly what we need is
to add and remove list nodes. Removing a list node becomes a very simple
task if we keep track of the previous node too. So now we also keep
record of prev node, and update prev node as we add and remove nodes.
All three linked lists we use benefit from this data structure
improvement, making it very fast to remove a list node.
Another benefit of this change is that we don't need to identify conns
with their id numbers or child conns with their src fds. So now we
directly delete them, without needing to check their ids or fds.
Soner Tari