Commit graph

64 commits

Author SHA1 Message Date
Richard Levitte
f29f610bf5 make update 2003-09-10 09:15:22 +00:00
Richard Levitte
27d63818e1 Include "e_os.h" instead of "../e_os.h", and trust the building
procedure to give the correct -I options to the compiler.  This is
*especially* true for test programs that appear in two places, with
different paths to e_os.h depending on where they are built.
2003-09-10 09:15:09 +00:00
Richard Levitte
fa68935f57 Include openssl/fips.h outside of the check for FIPS, so make depend
doesn't differ between FIPS and non-FIPS modes.
2003-09-10 09:06:01 +00:00
Richard Levitte
b2f94f81e1 We currently define FIPS, not OPENSSL_FIPS. The reason for this is
(probably) that FIPS is an entirely internal macro, and is not
accessible by third-party authors.
2003-09-10 09:05:06 +00:00
Dr. Stephen Henson
ad6eec30a3 Use BIO_snprintf() instead of snprintf().
Update hashes.
2003-09-10 00:44:53 +00:00
Dr. Stephen Henson
25c578f3ca Typo. 2003-09-10 00:16:42 +00:00
Dr. Stephen Henson
22f083815c Include e_os.h in a few cases (to pick up
str(n)icmp defs).

Disable a few tests if not FIPS.
2003-09-10 00:10:34 +00:00
Dr. Stephen Henson
c61e45fba0 Update hashes. Fix a few typos in o_str.c 2003-09-09 23:43:29 +00:00
Richard Levitte
d3446728f9 Move the FIPS check so make depend doesn't give different results
depending on FIPS mode.
2003-09-09 16:38:16 +00:00
Richard Levitte
c2cdb1a877 Test data files should not be part of the TEST value, or util/mk1mf.pl
gets confused...  The separate TESTDATA variable was inspired from
crypto/evp/Makefile.ssl.
2003-09-09 09:10:45 +00:00
Dr. Stephen Henson
31b28f95be Put #ifdef FIPS round FIPS DSA_generate_parameters .
#if 0 unimplemented ciphers so mkdef.pl doesn't pick
them up.
2003-09-08 17:01:48 +00:00
Richard Levitte
64a014a6d2 Make it builadable in non-FIPS mode.
The current solution is very Unix-bound, and there is probably a better way to do this.
2003-09-08 16:43:55 +00:00
Dr. Stephen Henson
a86eff4296 included <string.h> in fips.c to pick up
memcmp definition.

update fips_make_sha1 to use fips_err.h

update hashes.
2003-09-08 12:49:08 +00:00
Dr. Stephen Henson
4c4ef336a9 Update dependencies. 2003-09-08 12:39:13 +00:00
Richard Levitte
19c8d4a5f8 Ignore the directory rsp (introduced when running tests). 2003-09-08 11:36:32 +00:00
Richard Levitte
e7d580a924 Some more files to ignore. 2003-09-08 11:35:23 +00:00
Richard Levitte
1f8e046af9 Since fips_err is really just used as a header by fips_err_wrapper.c,
let's change it's suffix from .c to .h.  This also avoids some
otherwise very mysterious (and probably sensible from a historical
point of view :-)) name changes done by mk1mf.pl.
2003-09-08 11:33:07 +00:00
Richard Levitte
ccf61c87fe Remove some unneeded space. 2003-09-08 09:59:43 +00:00
Richard Levitte
6d02c53762 Use $(TOP) instead of ../.. as much as possible. 2003-09-08 09:59:11 +00:00
Richard Levitte
dac0d334ad Because of changes in Makefile.ssl, the files got reordered. 2003-09-08 09:57:57 +00:00
Richard Levitte
9ef37a8dc9 make update. 2003-09-08 09:17:36 +00:00
Richard Levitte
fbb40083c1 fips_err.c doesn't belong with the headers. 2003-09-08 09:17:13 +00:00
Ben Laurie
104f570c73 Missing file. 2003-09-07 11:13:54 +00:00
Ben Laurie
e1015c1fe7 Samples. 2003-09-07 10:59:34 +00:00
Ben Laurie
97e62d5554 Add samples. 2003-09-07 10:53:13 +00:00
Ben Laurie
b8b47f67d0 Add fingerprint chain and checking. 2003-09-06 13:31:40 +00:00
Ben Laurie
b5da126b21 Make the problem clearer. 2003-09-06 10:41:27 +00:00
Richard Levitte
a7d64957b9 Include e_os.h to get the proper definition of OPENSSL_UNISTD, and use
that macro.

It's possible that OPENSSL_UNISTD_IO should be used instead of
OPENSSL_UNISTD, for the MSDOS case...
2003-09-05 14:09:40 +00:00
Richard Levitte
f743ef233a make update 2003-09-05 13:41:04 +00:00
Richard Levitte
79176d6053 ALWAYS check the standalone source.
make update.
2003-09-05 13:37:28 +00:00
Richard Levitte
e726e5f170 Make sure the compilation of the FIPS stuff goes through even in
non-FIPS mode.

Update the appropriate fingerprints accordingly.
(something is weird, someone else was working on the same stuff, and
removed fips_sha1_selftest.c from fips/sha1/standalone.sha1...)
2003-09-05 13:26:52 +00:00
Dr. Stephen Henson
c1a32376db Fix signed/unsigned warnings and C++ comments. Update hashes 2003-09-05 13:00:34 +00:00
Richard Levitte
6b211d8cd2 Include string.h and stdlib.h where needed, to avoid warnings about
strlen(), memcmp(), exit() and others to be used without a proper
declaration.

Update the appropriate fingerprints accordingly.
2003-09-05 12:22:21 +00:00
Ben Laurie
f3bda010df Missing files. 2003-09-04 16:46:42 +00:00
Ben Laurie
4ccac96346 Automagically seed FIPS PRNG. Add OPENSSL_FIPS flag. 2003-09-04 10:22:13 +00:00
Ben Laurie
4ef3352608 -DFIPS may be the last thing on the line. 2003-09-04 09:04:24 +00:00
Ben Laurie
1f1bd3a51a Selftests. 2003-09-04 07:17:43 +00:00
Ben Laurie
b2293a6cc6 DSA stuff and tests. 2003-09-03 14:11:33 +00:00
Ben Laurie
766332ac11 More test vectors. 2003-08-31 09:50:11 +00:00
Ben Laurie
51f7c5a6ea Add test. 2003-08-31 08:52:39 +00:00
Ben Laurie
fc0376e257 Handle 3DES tests. 2003-08-30 17:28:08 +00:00
Ben Laurie
2b3784c235 Updated test vectors (probably incorrect, but who am I to question?). 2003-08-30 15:35:37 +00:00
Ben Laurie
aac3861773 Build the test program when needed. 2003-08-30 13:19:03 +00:00
Richard Levitte
4435349cc6 Make sure the order matches the command line in Makefile.ssl. 2003-08-11 10:31:21 +00:00
Richard Levitte
ec47bb7c81 - Add a configuration keyword "fips" to compile with FIPS
implementations.
- Reorder the build so the standalone FIPS SHA1 checker is built
  first.
- Add necessary defines to avoid symbol clashes between FIPS and
  non-FIPS implementations.
- Change necessary signatures.
- Correct bugs in FIPS build Makefiles.
- make update
2003-08-11 10:24:52 +00:00
Richard Levitte
94a1183fab Avoid clashing with the regular DES functions when not compiling with
-DFIPS.  This is basically only visible when building with shared
library supoort...
2003-08-08 10:08:14 +00:00
Richard Levitte
3659df7d15 make update
(I'm quite worried about what this will do to compatibility with
earlier 0.9.7 versions)
2003-08-04 13:26:14 +00:00
Richard Levitte
16f1fdbc8b Add an empty list of AES tests. At least, the test suite will pass,
and perhaps the conflict this generates on the person that hasn't yet
committed the real file will prompt him to do so :-).
2003-08-04 12:03:56 +00:00
Ben Laurie
2541f58309 Make tests work (CFB1 still doesn't produce the right answers, strangely). 2003-08-03 12:22:35 +00:00
Dr. Stephen Henson
693f2e3625 Replace C++ style comments. 2003-08-01 13:07:29 +00:00