Steven Herring homework 4, complete Thursday May 5th for extra credit of 4 points.
No other extra credit objectives were knowingly satisfied for this assignment.
************************************************
** How I tested my solution to the assignment **
************************************************
Pulled hw4 repository into a fresh, clean directory and did the following
-----DID THESE STEPS ONCE-----
aclocal && autoreconf -f -i && automake --add-missing
./configure && make dist && cp skey-2.0.tar.gz ~
mkdir ~/tmp && cd ~/tmp && tar xzvf ~/skey-2.0.tar.gz
-----DID THESE FOR EACH SYSTEM-----
mkdir sysname && cd sysname
../skey-2.0/configure --prefix=$HOME/skey-2.0/sysname --enable-debug --enable-shared
make all
make check
make install
************************************************
************************************************
Due to lack of my own time on previous assignment, most changes were done from the start of hw4
Removed PRE-ANSI declarations across source.
Cleaned up some code.
Moved most header declarations into skey.h
Made files try to only include skey.h where applicable (mostly everywhere)
Removed MS_DOS checks, unnecessary ifdefs, extends as requested.
Added new features:
-v flag to print version of package (for example "skey version 2.0")
-h to print usage (with all new options)
//-e to echo back the passphrase being entered (otherwise don't echo back)
-d to print debugging/tracing output
-l ARG: Set the logging device to ARG #
-v was done by having a hard version # in the source code and printing it
-h usage macro, with predefined help.
-d debugging i used a macro to check for the debugging level and depending on that would define the debug functions differently. my methods can be viewed in debug.h, debug.c and skey.h
-l logging also used macros to check if logging was set, if so check for a valid FD upon opening and port the output there
Fixed Debugging logic that was causing debug output to route to an unknown location (segfault)
Debugg functions were applied before and after functions inside skey.c, in some cases after and after (readpass as to avoid interfering with password gathering)
Created configure.ac, Makefile.am, /src/Makefile.am - currently creates a full make file
configures properly, makes, installs and runs correctly on b-freebsd91
Added config check using configure.ac with AC_C_BIGENDIAN. Which defines WORDS_BIGENDIAN
Added Makefile destinations to include man and scripts
Added man/Makefile.am and scripts/Makefile.am to help extract man pages into distribution and use scripts for testing
AC_CONFIG_FILES([Makefile src/Makefile man/Makefile scripts/Makefile])
Add and Update AUTHORS files
Fixed bug where output was different depending on the system being 32 or 64bit. Needed to use ints as opposed to longs for some variables in md4.h and md4.c
Added and tested -e flag to work as intended. Used a flag variable and modified the signature for readpass to accomplish this
Successfully built and installed w/o make check on 8/9 machines (OSX unavail.) getting consistent output across all 8
Completed testing script using example from Zadok emails.
Successfully got make check to work, modifying the script to look in the correct places of build directory
Modified version number
Modified usage to include -e
Got static library linking by modifying Makefile.am for each directory involved in the library process.. this required organizing my directory structure
Added lib and include folders that managed .C and .H files respectively
#TODO: LIBTOOLS STUFF, MINOR ORGANIZATIONAL THINGS
Attempted to use libtoolize to use dynamic linking, this didn't work, had to modify Makefile.am's on my own, not hard, very little translations a->la library->ltlibrary etc.
OSX was not compiling due to clashing symbols multiple definitions of .h file - added header guard into skey.h to help
also removed global variable definition of char* 'logFile'
Had a bug where I was not preserving variable 'n' at some point in skey.c, this was causing my output to be wildly different than professors. Fixed, output now matches professors.
Upon fix of this bug, OSX output was no longer consistent.
OSX output bug was due to localization of the optind and optarg variables in skey.c caused by removing 'extern' key, instead removed their skey.c definition completely.
Resolved setenv/getenv issue by doing a function check in configure.ac for getenv and setenv.
Had to implement the given setenv and getenv functions that were supplied with SKEY distribution.
Also in the cases we use setenv we needed to change the signature for when setenv is defined and when we used the given SKEY version
Removed all left over system respective ifdefs, externs, etc, etc.