aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Finish get_cc_info()lloyd2006-09-041-21/+29
|
* Use get_cc_info to get the single-variable values from the compilerlloyd2006-09-041-17/+14
| | | | descriptions; lists are still processed in set_cc_defines()
* First pass at get_cc_info()lloyd2006-09-041-1/+20
|
* Implement get_arch_info, which handles reading a single architecturelloyd2006-09-041-19/+42
| | | | | file and getting the results, which is called by set_arch_defines as needed.
* Finish get_os_info; set_os_defines now uses it for gathering all informationlloyd2006-09-041-51/+33
| | | | from the file.
* Begin to make a distinction between reading the file and setting the valueslloyd2006-09-041-12/+56
| | | | | | | in hashes; split the reading portion of the OS config info into a new function, get_os_info, with set_os_defines iterating over each file in misc/config/os, calling get_os_info, and then pulling values out of the results, setting the values where the rest of the script expects.
* Add helpers for reading single-line settings in the config descriptionlloyd2006-09-041-56/+55
| | | | | | | and module info files. Get set_undef working for get_module_info, so we don't have to create as many (explicit) anonymous subroutines.
* Change the function that reads modinfo.txt files to use read_hash as welllloyd2006-09-041-118/+84
|
* Change read_hash to take a function reference, which it calls for eachlloyd2006-09-041-73/+43
| | | | | matching line. This makes handling the variations much simpler. Convert more of the applicable code to use read_hash.
* Add some helper functions for reading information from the config info fileslloyd2006-09-041-59/+59
|
* Clean up how machine-specific compiler options are setlloyd2006-09-041-16/+27
|
* Remove use of the DirHandle modulelloyd2006-09-041-42/+17
|
* Define a new function, make_reader, which returns an closure over thelloyd2006-09-041-69/+54
| | | | | | file being read that strips comments and whitespace and then skips over blank lines. This was previously something done in every single loop that called process() directly.
* Make a passphrase optional for the DSA and RSA key generation exampleslloyd2006-09-042-12/+17
|
* Use formatted I/O to make the output a little easier to readlloyd2006-09-041-3/+3
|
* Document --build-dir and --local-config in the help output.lloyd2006-09-031-52/+47
|
* Move the (commented out) gzip.cpp/gzip.h entries into the add file blocklloyd2006-09-031-4/+3
|
* Make some of the code a bit more Perl-ishlloyd2006-09-031-153/+133
|
* Remove some dead codelloyd2006-09-031-60/+45
|
* Place the add_file/replace_file/ignore_file markers in the module infolloyd2006-09-0327-114/+163
| | | | | | files into blocks; makes a bit more sense, since there are potentially many arguments to each, and the current system was making it difficult to write a generic reader for the files.
* More small cleanupslloyd2006-09-031-136/+48
|
* Small cleanups, remove some obsolete code, etclloyd2006-09-021-33/+29
|
* Move some variables that did not need to be global into main() itselflloyd2006-09-021-8/+9
|
* Simplify how the hashes that specify bits of information about thelloyd2006-09-021-112/+36
| | | | system are passed around.
* Put all of the code into a function of some sort (mostly by movinglloyd2006-09-021-210/+182
| | | | the code that formed the main part of the script into a new main function).
* A few small cleanupslloyd2006-09-021-16/+9
|
* Drop the now basically useless readme.txt file for the configuration codelloyd2006-09-021-14/+0
|
* Load the configuration information when the user runs configure.pl, ratherlloyd2006-09-0218-3038/+360
| | | | than baking it in with mkconfig.pl
* Instead of including the local config header into build.h, simply copy thelloyd2006-09-021-1/+4
| | | | | contents into build.h. This keeps the issues with getting the include paths right, handling name conflicts, etc to a minimum.
* Remove explicit alignment settings before the loops; the loop macrolloyd2006-09-021-5/+2
| | | | | | | already sets alignment. Change the core multiply/add macro a bit; probably not any faster, but a bit cleaner.
* Allow for implementing hash functions in Pythonlloyd2006-09-025-69/+165
| | | | Move some code that is used in several places into a common header
* Note some slightly more ambitious todos with regards to the configure codelloyd2006-09-021-0/+7
|
* Remove various small unnecessary things from the configure codelloyd2006-09-024-17/+1
|
* Allow for implementing block ciphers in Pythonlloyd2006-09-022-64/+170
|
* merge of '2e51bbe4473463af7f2603deb7c12a332a7e1ae6'lloyd2006-09-023-31/+210
|\ | | | | | | and '94c8c963566255088b729b3c2cc7fe788f474945'
| * Remove the object() accessors, since those objects might not exist iflloyd2006-08-311-8/+0
| | | | | | | | an algorithm were implemented in pure Python.
| * Merge the block cipher, stream cipher, hash, and MAC exports into algos.cpplloyd2006-08-316-244/+207
| | | | | | | | Add various useful conversion functions
| * Export stream cipherslloyd2006-08-312-0/+57
| |
| * Add hash and MAC wrapperslloyd2006-08-313-0/+107
| |
| * Also export the name() and clear() methods of the block cipherlloyd2006-08-311-8/+7
| |
| * Add simple wrappers for block cipher objectslloyd2006-08-313-31/+92
| |
* | Let the user specify an extra include file (included into build.h)lloyd2006-09-012-0/+8
| |
* | All the user to specify --build-dir to configure.pl, which will allow forlloyd2006-09-011-0/+8
|/ | | | multiple builds in a single directory.
* Add some comments to the NIST test suite results file with regards tolloyd2006-08-231-0/+6
| | | | results that have been changed from the official expected results.
* One of the tests had its expected result set wrong (was for testing)lloyd2006-08-231-1/+1
|
* Clean up and finish off the Python version of the NIST X.509 test suite.lloyd2006-08-232-23/+93
| | | | | | | | Expected results are read from a file (results.txt), and pass/fails are reported. Also change validate to return the result normally, rather than throwing an exception; while it did save some code in validate() itself, with the addition of all the code to support it, it was a bit longer, as well as being somewhat hard to follow.
* Remove an old version of the Filter() wrapper functionlloyd2006-08-231-3/+1
|
* Just export OctetString, and define the alternate names for that typelloyd2006-08-233-28/+18
| | | | in the Python portion of the module
* Drop this nonsense with the build/ directory, it was just getting in the waylloyd2006-08-231-8/+4
|
* Make SymmetricKey and InitializationVector simple typedefs of OctetString;lloyd2006-08-231-23/+4
| | | | | since the RNG merger in 1.5.0, they have been effectively the same type anyway.
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-21 12:35:32 +0000