Summary Period: 2004-03-21 to 2004-12-29
18 (0.0%)
85 (0.0%)
Directory | Changes | Lines of Code | Lines per Change |
---|---|---|---|
Totals | 18 (100.0%) | 85 (100.0%) | 4.7 |
/ | 5 (27.8%) | 21 (24.7%) | 4.2 |
programs/_realsetup/ | 3 (16.7%) | 20 (23.5%) | 6.6 |
programs/barf/ | 1 (5.6%) | 19 (22.4%) | 19.0 |
programs/showhostkey/ | 3 (16.7%) | 8 (9.4%) | 2.6 |
programs/setup/ | 1 (5.6%) | 7 (8.2%) | 7.0 |
programs/pluto/ | 3 (16.7%) | 4 (4.7%) | 1.3 |
programs/ipsec/ | 1 (5.6%) | 4 (4.7%) | 4.0 |
programs/_startklips/ | 1 (5.6%) | 2 (2.4%) | 2.0 |
Some regexp had striped many references to /proc of the leading
slash, causing one outright failure and lots of output nits.
19 lines of code changed in:
The macro AES_CBC_BLOCK_SIZE was defined in two different ways, which
happy for us yeiled the same results. The local copy in this file was
removed thus letting the global copy 'win'. One less warning.
0 lines of code changed in:
By adding const to the second argument of Decode we get rid GCC
warning messages.
2 lines of code changed in:
Add structure names A & B so that modern GCC's don't return a size of
0 and thus a union size of 0.
2 lines of code changed in:
It's critical to install the _right_ file as the module, and for 2.6
kernels thats the .ko file and not the .o file!
2 lines of code changed in:
This added test gets rid of an annoying message when there is no ipsec.o
or ipsec.ko module on 2.6 systems. Tested on 2.4.
2 lines of code changed in:
This set of changes makes it so that we --stop the program we unload
the kame'ish modules the same as the KLIPS ipsec module.
There is also a change to test for KLIPS to be running when killing
KLIPS so that we don't bitch about missing files.
8 lines of code changed in:
We forgot to clean up the pluto control socket file, so now it gets
removed when we stop if it's around.
3 lines of code changed in:
This patch cleans up the output of the status subcommand in cases where
KLIPS and/or Pluto are (or should not) be running. What to print logic
was fixed, errent error messages were suppressed and the number of
dangeling eroutes is now printed when there should be none.
9 lines of code changed in:
A group of changes for clean compiling on linux-2.6 host.
A symlink to get a missing assember file in the right place, a couple
of changes to get make clean working, change dummy target to help as the
2.6 kernel has no dummy targer and make sure our dummy is PHONY.
7 lines of code changed in:
For shipping system we want to have the default USE_LWRES be false.
2 lines of code changed in:
Minor cleanups: better white spacing, don't reference missing makefiles,
use ='s rather then *'s which are most often used to find error reports.
8 lines of code changed in:
The makefiles were complaining of missing arch//Makefile* files,
so we export ARCH.
2 lines of code changed in:
Left over FreeS/WAN to Openswan conversion.
4 lines of code changed in:
This shows the dangers of coding up a fix too quickly, the first two
times I 'fixed' this I did NOT read the code closely enough, and presumed
that each comment insertation was a different line of code and I could
just change the comment character.
On use I figured out that I needed to give the code a closer look, and
make more substansave change to get the comment characters right for the
different target programs.
dhclient.conf wants '#' and named/BIND zone files want ';'.
I need to go find a grill for my hat...
4 lines of code changed in:
Seems we had lots of the comment characters wrong.
2 lines of code changed in:
We were generating the wrong comment character for zone flies still.
2 lines of code changed in:
Some versions of the program logger(1) seem to miss the EOF on stdin or
stdout and hang forever when getting piped input from a subshell. So I
rearanged the code such that the _realsetup output is buffered throug
a tmp file rather then a subshell and a pipe.
This should fix 'startup hangs' bugs seen on the Zaurus and Debian
distros. Obviously logger still needs fixing...
7 lines of code changed in:
Generated by StatCVS 0.2.2