1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
|
November 1, 2011:
o Make the netmask address family fix work without knowledge of the struct
ifaddr definition. This fixes a crash on Lion, where the layout of the
structure has been changed, but at the cost of the fix no longer working for
IPv6. I think this is OK though, since mDNSResponder has been fixed on
Leopard and beyond to no longer require the hack.
o Proper multicast address checking for tun; multicast should now work reliably
with IP and IPv6 on tun also.
o A quite comprehensive test suite has been added that allows for quick release
testing.
o PPC support has been dropped due to XCode 4 no longer supporting PPC arch.
September 13, 2009:
o Change linker options to produce 64 bit kext bundle for Snow Leopard.
o Switch from kmem_alloc and friends to OSAlloc for memory allocation and
avoid the delay() call. Respective symbols are not available on 64 bit
kernels anymore.
September 5, 2009:
o Initial Snow Leopard port. Thanks to various people contributing patches in
the bugtracker. The new official version can only be compiled on Snow
Leopard but has been tested to work on all Tiger, Leopard and Snow Leopard
o Clean up unused locking code and switch to rwlocks even for simple mutexes,
which avoids a symbol incompatibility for Tiger and Leopard.
o Clean up compilation flags in the Makefiles.
July 4, 2008:
o Adapt the former Leopard package to also be installable on Tiger systems.
This obsoletes the Tiger version, both Leopard and Tiger are now supported
by a single package.
June 7, 2008:
o Protect the selwakeup() call by the lock. This fixes incorrect select()
behaviour, thanks to Roland Wendelin for reporting this.
o Fix tuntap_mbuf_queue::size initialization
o Use a proper wait condition for synchronization when detaching the network
interface. The old code would crash if the if_detached() handler was called
from a different thread than unregister_interface().
January 21, 2008:
o Work around an issue in the Darwin kernel. When unregistering an interface,
addresses are not properly removed from the interface. This leads to
crashes and other problems when reusing the interface. Introduce an ugly
hack that tries to remove all interface addresses when shutting the
interface down.
o Fix a small mbuf leak that could occur when the output queue was full.
Thanks to Oleg Dolgov for reporting this.
December 21, 2007:
o Fix paths in the startup item postflight scripts
o Check if_ioctl arguments more defensively after a report of a panic after
receiving a NULL arg.
November 14, 2007:
o I have done a complete rework of the installer package generation. The
package is now edited in PackageMaker. The distribution package can still
be built from the commandline though.
o Fix incorrect permission & ownership of the installed files.
Oktober 11, 2007:
o Fix the permissions of the postflight scripts. Installer packages should work
again.
o Drop the kmod and kmodc++ in the linker command, they seem to be unneeded
with Leopard.
September 20, 2007:
o Initial Leopard port, it's basically the latest Tiger version with some
Leopard-related fixes and s/Tiger/Leopard/g
o I have switched to a proper version management system (git) and could
remove some of the CVS hacks subsequently.
o The installation packages have been reworked a bit, they now install into
/System/Extensions and /System/StartupItems directly by using
DestinationPaths.
May 13, 2006:
o This version is not stable, it may crash, sorry.
o Universal binaries that run on ppc and intel macs.
o Adds tap MAC address randomization
o Redesigned locking.
o Better multicast support
o mDNSResponder workaround, so that the tap interfaces should get picked up
now. Note that we are fixing ifconfig/kernel behaviour here.
o All tapX and tunX devices are visible in /dev at all times, network
interfaces still created dynamically, though.
o Startup items moved to /Library/StartupItems
May 17, 2005:
o Initial Tiger port. We now have KPI-style interfaces. I guess the Tiger
version is little slower than the Panther version because of all the
wrapping and hiding in the kernel.
o The kernel extensions moved to /Library/Extensions. That is the place where
non-Apple kexts are supposed to live.
April 21, 2005:
o I added support in tun for AF prepending like some BSDs do. Thanks to Dennis
kSchneider for mailing the initial patch. You can also set the value of
AF_INET6 to be used.
o I finally found that major bug causing crashes (especially on multiprocessor
machines). It also caused a memory leak (lost mbufs), and might have caused
performance/througput/data-loss problems. Everyone is recommended to upgrade.
April 6, 2005:
o I rewrote the common part concerning the tun and tap initialization and
cleanup. This should make the code more maintainable (less duplication).
o The devices now reinitialize to the state they were started in when they
are closed by an application. This concerns IP addresses for example.
o I changed the package building system to use PackageMaker.app in batch
mode. The packages also check for version 10.3 now, so nobody should be
able to install tun/tap on 10.2 using installer packages. Furthermore I
have sprinkled some warnings telling you not to use tun/tap on SMP machines
over the installation process ;-)
o Some minor locking fixes.
November 19, 2004:
o Jamie Wood reported that the packet queue in the driver could be considered
empty even if there were packets in it. This was probably caused by a
synchronization problem that should be fixed now. People encountering
timeouts etc. should try the new version.
o I finally implemented support for changing the interface MTU. The driver
enforces the MTU when writing packets to the character device now. However,
packets coming from the kernel are not checked.
September 9, 2004:
o Marcello Teodori told me that the tun driver wasn't working with openvpn.
The problem was the fcntl call, fixed that. Should work now. Thanks
Marcello!
o changed the tun driver not to prepend the address family field before each
and every packet (which is the behaviour of OpenBSD). As there is currently
only IPv4 and IPv6 support there is no problem with the standard tun
approach used on other OSes. This should make the driver much more
compatible.
o Did a script and makefile support so that the installer packages can now be
built from the command prompt. Unfortunately this might break things
someday as I am not using the 'official' way to build the packages
o Cleaned up installer packages a little.
August 24, 2004:
o initial version put online
o basic tun/tap support, tap working with qemu
|