pfSense

Echo done after configuring carp

MFC 6906 6907
Use /var/etc/ for storing configuration files MFC: ASAP (By sullrich)
Use /var/etc/ for storing configuration files MFC: ASAP (By sullrich)

MFC 6899
Allow bridging on ath0 + wpa

MFC 6897
Write out to /tmp/ file commands used when adjusting mtu during bridging operations

MFC 6894
Force all members of a bridge to the bridging from interface mtu.

MFC 6892
Setup span tree on all ports correctly.

MFC 6871
If WAN is pppoe, hold down boot until connection is established This will help anyone with PPPOE on WAN and DHCP on OPTx not get incorrect gateways assigned.

MFC 1.178
successfully kill hostapd if changing settings back from WPA to WEP (fix by lsf, noticed by marcusah)

Correct sysctl location

• pfsync0 should already exist, do not create
• assign syncdev and bring up at same time to match FreeBSD's pfsync rc.d script

• Major rewrite of CARP/PFSYNC init code. Remove a bunch of if() and for() to make the code 100* more readable
• When bringing up CARP, set its advskew to 200 and when we are finally up at the end of the boot sequence assign the normal ADVSKEW

Patch-tested-by: BillM

• Various safety checks
• Return next available slot correctly

Make get_number_of_bridged_interfaces() a little more industrial strength by actually looking at the bridge numbers instead of the quantity of bridges

Missing }

Do not call ifconfig on non-existant interfaces when testing

• Correctly bring up a new bridge
• Correctly destroy a bridge when no longer used

Remove debugging line

Start at 0

Use consistent variable names

Trim whitespace

Change get_number_of_bridged_interfaces() behavior

Correctly return # of bridges

Set $x = 0 before for()

Sync with 1.2

Sleep between down, delete and destroy. Race conditions are bad, mmmmkay?

Return a numerical value instead of "bridge{$x}"

Return a numerical value instead of "bridge{$x}"

++ bridge count

Move wan gateway to interfaces->wan->gateway

Correctly define () brackets

Test for active bridge better

Down bridge before deleting and destroying

• Add destroy_bridge()
• Add discover_bridge()

Use get_next_available_bridge_interface() when creating new bridged interface

• Add get_number_of_bridged_interfaces()
• Add get_next_available_bridge_interface()

Added hidden option for Roaming.

Assign $wancfg array before resolving real interface

Translate friendly name into real interface name

Split dhcp out of get_current_wan_address into its own logic. Do not use get_real_wan_interface() when using dhcp.

• Extend get_current_wan_address() to accept a interface.
• Use "wan" if no interface supplied
• Pass interface when creating nat port forwards

Run dhclient without -d

Ticket #565

Do not run dhclient in the background with mwexec_bg()

Ticket #565

Complete rewrite of wireless. Everything should be working now

Let ifconfig changes happen after hostapd is killed.
No more sighup hostapd as this seems to create problems.

If debugging is enabled, do not mute the console when bringing up the optional interfacs. we're trying to chase down a panic/freeze.

When debugging, alert while assigning syncdev

Do not use syncdev and syncif... Use syncdev.

• Add new /debugging check which enables $debugging. We should use this in the future to output as much additional information that we need to in diagnosing problems. in this case we're adding data to carp on boot to deterimine why its locking up / rebooting...

Do not process items if it's not an array.

Work around the fact that our traffic shaper configuration takes time to create by leaving the carp interfaces in a down state until the very end of bootup.

Only sleep for 1 second

Remove function stub. No traces of it being called now.

Adjust carp bootup sequence. Do not down the interfaces and later up them.

• Sync master password correctly
• Do not down carp interfaces on bootup

Backout. This wasn't meant to be commited just yet.

Reboot after restore

Correct carp bringup sequence

How about a space. Insightful.

Adding (EXPERIMENTAL) WPA Support.
Fixing up ifconfig settings not beeing removed when they are disabled after first time enabling.

Move is_jumbo_capable to pfsense-utils

Adding (EXPERIMENTAL) WPA Support.
Still celaning ifconfig bugs. Settings are not done correctly

Adding (EXPERIMENTAL) WPA Support.
Fixed a few bugs with setting turbo mode and hidessid and Pure G mode.

Move is_jump_capable() to pfsense-utils

Adding (EXPERIMENTAL) WPA Support.
Adding ifconfig down / up in order to activate channel changes and other ifconfig changes.

Adding (EXPERIMENTAL) WPA Support.

Do not destroy carp interfaces. FreeBSD no likie.

hush little pfsync, don't say a word. GeekGod's gonna buy you a mocking bird.

Add is_altq_capable() - will return true if the interface is altq capable

Add is_jumbo_capable() function
Tests to see if an interface is jumbo frame capable

Kill mpd and restart on wan changes if mpd is running.

Ticket #477

If hostapd is already running, -HUP it

Use full path to hostapd (/usr/sbin/hostapd)

Not need to require secondary dns on pppoe connections

Handle 11g only correctly

-f is not even an option. Bad dhclient web page giving out false info.

• Launch DHCLIENT in foreground mode.
• use mwexec_bg to launch item in the background

Confusing, eh?

Do not setup supplicant if running hostap

Add HOSTAP WPA support

Add WPA backend support

Even if a user doesn't fill in a hostname we really should run dhclient-script.

Doh! This is a bug on m0n0wall too.

$bridges_total should be a global variable. Treat as such.

More fixes submitted by databeestje

Do not echo done in interfaces_optional_configure_if()

Use correct interface

Use correct interface

Correct grab the optional configuration profile

Correctly bring up optional interface dhclient

Reported-by: databeestje

Down the interface in question. DHCLIENT should exit gracefully afterwards and if not, kill it!

Fixup Renew/Release

Allow multi interface Renew/Release

Add interfaces_dhcp_down($interface);

Update location of dhclient-script

Remove no longer used dhclient -r option.

Tell dhclient to go into background and don't call mwexec_bg()

Back out last commit.

++ counter before processing

Remove any dangling carp occurances. This bug showed its head during the hackathon.

Do not write to /tmp/sbin/ifconfig, write to /tmp/ifconfig

Correctly write out /tmp/ file

correct path to ifconfig