pfSense

Ensure 127.0.0.2 points out 127.0.0.1.

Make CP multi-interface capable.

• Create two new functions lock($subsystem)/unlock() to have more reliable locking using semaphores.
  This function can sleep till the resource is free and can help find not well behaving code.
• Remove most of the config_lock/config_unlock logics on the whole scripts/pages it is an abuse of this....

Rename filter_translate_type_to_real_interface to interface_translate_type_to_real
Move this function to interfaces.inc where it belongs.

• Teach get_interface_subnet() about carp.
• Increase scope to all interfaces when searching for the carp parent since the later is used even on those scopes.
• General cleanup of unused vars and code.

• Convert from exec_command to `$command` to remove the only dependency of interfaces.inc.
• Remove the heavy require_once(functions.inc) from interfaces.inc it is really not neededo

Remove extra $

Separate delete down actions on ifconfig to fix issues. Add L2TP virtual interafce to a list.

Fix typo to correctly load interface groups on startup.

• Move other functions around to where it makes sense.
• Reduce somewhat require_once() overhead by removing the includes from gwlb.inc they are not needed.
  Some more analysis is needed on the include path

More functions moving from pfsense-utils to interfaces.inc.

• Make the carp ip fix for ipsec more general so other services that use the same methodology work.
  - Basically get_interface_ip() now knows how to handle carp(4).
• Move interface related function from pfsense-utils.inc to interfaces.inc that is their place....

fix typo, clean up text

Update my copyright.

Remove debug string.

• Convert the QinQ code to use a file where the commands are passed. This speeds up most of the vlan creation.
  Before you'd need 20min to create qinq with 1000 member now it adds 2000 members in just 1 minute
• Fix some cleanups when deleting QinQ

Bring in support for QinQ. At this time it is limited to only 2 levels.

• Keep only one function to return the mac address
• Bring the mac address function to interfaces.inc its the only consumer of those.

Actually merge the part that does the real work with FreeBSD groups.

Fix display of DHCP status on Status -> Interfaces

Fix http://forum.pfsense.org/index.php/topic,14361.0.html

Revert "Merge IPv6 changes"

This reverts commit 1f9f2a95b7b42cf33e730535092e56e214fdb848.

Conflicts:

etc/inc/filter.inc
        etc/inc/interfaces.inc
        etc/inc/pfsense-utils.inc

Make the rc.banner a lot more faster.

Remove this while loop now. It never ends and runs the cpu usage up to 100% forever.

Revert "Back out ipv6 changes"

This reverts commit 541b7c561912174b57927839ce39bb7ec9c07ee4.

Back out ipv6 changes

Make sure we do not return the interface passed as parameter rather NULL.

Remove wrong comment.

Fix delaying of creating composite interfaces like bridge/gre/gif. This fixes the problems reported http://forum.pfsense.org/index.php/topic,13735.0.html.

MPD4 moved the authentication on auth section coorect the settings on PPTP which were forgotten.

Make L2TP known in other parts of system, schedules and the function get_interface_ip.

Add more L2TP improvements.

Correctly add lagg members and create lagg interfaces before vlans to allow the later to be created on top of lag.

Merge branch 'master' of git@rcs.pfsense.org:pfsense/mainline

Correctly add lagg members and create lagg interfaces before vlans to allow the later to be created on top of lag.

Merge IPv6 changes

Move laggproto setting after adding the members and making sure they are up.
Otherwise we might panic as explained in ticket #1864

Fix the {interface}up checks.

Fix generation of pppoe reset cron entry. Not sure if this should stay here or go into inteface_pppoe_configure!?

fix CARP panics with VLANs

Don't use broadcast on CARP IPs.

Don't add CARP IPs on an interface that doesn't exist to prevent panic.

Add another parameter to interface bring down which can be used to destroy the interface if it is a virtual/cloned one.

Temporary Fix for vlan creation issue.

Use correct var.

Add some checks on interface_bring_down and remove unlinking of nameserver.conf it needs some more thinking on how to propperly do that.

Correct error message.

It is legitimate to be called with an empty vlanif. The webgui uses this to create new vlans.

Correct some checks and use propper whitespace.

Change functions to function nameoffunction() {

• Add new function to interfaces.php setup_pppoe_reset_file() which
  sets up the periodic reset file. Prior to this restoring a config.xml
  backup would end up with a cron entry with no file!
• More text cleanups to make the files readable

• Do not destory enc0
• Only run ifconfig if a real ip and netmask exists instead of ifconfig

Change order slightly.

Introduce 2 new functions
find_interface_subnet and get_interface_subnet
similar to their ip ones
find_interface_ip and get_interface_ip
to easy getting this info for dynamic interfaces.

Update list of altq suppported interfaces with some of our custom internal names wich actually are really ng_iface(4) type.

Obey the dnsallowoverride option even for secondary dns on pppoe interface.

Use product_name

Try to use where possible filter_configure so upcoming event modifications can be easily integrated.

Call a backtrace and log it if we have a failed interface variable.

• Add interfaces_bring_up()
• Fix log_error (missinge else)
• Remove blank line after ?> that ermal just added

Better check if the knob is defined or has no value.

Reorder the calling.

• Log an error when an ifconfig up command is missing its interface
• Cleanup formatting

• Use interface_bring_down to not duplicate code
• If we are reconfiguring check bridge/gif/gre interfaces if they need to be reloaded.
• Invalidate array cache to avoid stale information

• Rename link_int_to_bridge_interface to link_interface_to_bridge
• Create two new function link_interface_to_{gif,gre}

Surpress message when trying to remove alias address

Fix typo.

Make sure to have an interface as a bridge member after reconfiguring.

Be smarter at finding the smallest mtu.

Fix mtu handling on bridge and lagg interfaces. Reported-by: Seth

Some better checks to avoid errors.

Do some better checking before acting on the bridge.

Add each member to the bridge individually rather than in one single command.

Restore correct stub

Add stub functions for code that references old names(mostly pacakges).

Suggested-by: Scott

• Rename get_current_wan_address to get_interface_ip
• Rename get_real_wan_interface to get_real_interface
• Simplify get_interface_ip by using find_interface_ip which should help in speed to since its using caching. Pointed by billm@
• Cleanup some code when passing or remove some unused one.

Improve vlan interface code to try to activate as much as possible on the parent, this allows to not maintain a supported list.

Oops, use tunnel address not network.

Clenup some more when bringing down an interface.

Clear the arp entries for the interface going down.

Slightly change order of commands for dhcp.

Refactor some code so frontend does not need to know about details that might change
during development.

style(9)

Move from $debugging to $g["debug"].

When interface is configured for pppoe and mtu is also configured then adjust the pppoe mtu c
orrespondingly as to us this means pppoe mtu and not parent interface mtu.

Restore accidentally removed code with bigpond removal which is causing issues with pppoe/pptp.

• Remove bigpond support from backend.
• Use killbypid instead of sigkillbypid SIGUSR2 is not doing what is expected.
• Actually start pppoe/pptp when the routing is called for this.

Actually take a shorter route at saving the gateway.

Use correct interface to record gateway.

run hostap later in script, fixes ral(4) card difference in FreeBSD 7.0. Works with ath(4) also.

Ticket #1813

Take into consideration the subnet mask configured.

Use correct variable name for parent interface.

kill(1) does not need SIG in front of signals anymore. Though i do not think mpd4 responds to SIGUSR[1,2] the same as mpd3.

Seems that mtu and up parameters of ifconfig cannot be specified together, at least on lagg(4) case.

Disable the second call to the same script for configuring wireless interfaces it seems redundant.

Fix ppp(3G) backend code and make it possible to create more the one.
Although it needs to be though how would we want to setup something as failover with such setups since we cannot expect for them to be always connected.
Maybe dial on request or some such might be needed!?

• Rename $debug to $g['debug']
• Fix renamed functions call especially the wireless config
• Fix some ppp code backend

Mostly missed renaming and typos.

• Missed function call rename
• This needs to be gif not gre

Bring up to more sane config.

PPP forgotten knobs syncronization.