[Swan-commit] Changes to ref refs/heads/master

Paul Wouters paul at vault.libreswan.fi
Fri Apr 18 07:48:58 EEST 2014


New commits:
commit 3a46b6966b756f47d1835ff59591ceb54911a09d
Author: Paul Wouters <pwouters at redhat.com>
Date:   Fri Apr 18 00:48:41 2014 -0400

    testing: new version of testing/lib/libswan/OUTPUT.algparse.txt

commit ea38662b8bfe5f80106287332969a5022cf6a1f3
Author: Paul Wouters <pwouters at redhat.com>
Date:   Fri Apr 18 00:44:49 2014 -0400

    testing: rewrote testing/lib/libswan/algparse.c
    
    - This now tests many more esp= and ah= lines than before.
    - Known good output updated and more readable
    - Old good output kept for reference for now
    - Seems a passert() hits when we hit an error. This seems a change in
      behaviour that needs to be investigated. It prevents the test from
      continuing when an string error is found, which should not be the case
      (this happened already before the rewrite, it must be due to actual
       code changes)
    
    new output looks like:
    
    [  3des-sha1;modp1024]    OK: alg=3DES(3)_000-SHA1(2)_000; pfsgroup=MODP1024(2)
    [  3des-sha1;modp1536]    OK: alg=3DES(3)_000-SHA1(2)_000; pfsgroup=MODP1536(5)
    [  3des-sha1;modp2048]    OK: alg=3DES(3)_000-SHA1(2)_000; pfsgroup=MODP2048(14)
    [      3des-sha1;dh22]    OK: alg=3DES(3)_000-SHA1(2)_000; pfsgroup=DH22(22)
    [      3des-sha1;dh23]    OK: alg=3DES(3)_000-SHA1(2)_000; pfsgroup=DH23(23)
    [      3des-sha1;dh24]    OK: alg=3DES(3)_000-SHA1(2)_000; pfsgroup=DH24(24)
    [           3des-sha1]    OK: alg=3DES(3)_000-SHA1(2)_000
    [           null-sha1]    OK: alg=NULL(11)_000-SHA1(2)_000
    [         aes256-sha1]    OK: alg=AES(12)_256-SHA1(2)_000
    [         aes128-sha1]    OK: alg=AES(12)_128-SHA1(2)_000
    [            aes-sha1]    OK: alg=AES(12)_000-SHA1(2)_000
    
    etc etc. See testing/lib/libswan/OUTPUT.algparse.txt

commit f814448cc1233f26cc71872c2ac8c684940856d6
Author: Paul Wouters <pwouters at redhat.com>
Date:   Fri Apr 18 00:43:41 2014 -0400

    testing: set LIBRESWANSRCDIR consistently

commit 2b6c596ca7066c8c4de99a7c60bfda6628dbfad2
Author: Paul Wouters <pwouters at redhat.com>
Date:   Fri Apr 18 00:41:06 2014 -0400

    testing: functions.sh fixups
    
    - Seems the MAKE macro isn't working fully. For now just set it to "make"
    - Add -DSWAN_TESTING for prng.c
    - Restored functions do_unittest() and unittest()
    
    Note using functions.sh in make check in lib/libswan gives a weird
    conflict with a modules() macro/alias elsewhere. I had to get rid of
    the rpm for "environment-modules". I could not find the modules()
    function to rename though.

commit e6965e5080eb2446667ac600d1b76c77e00876b6
Author: Paul Wouters <pwouters at redhat.com>
Date:   Fri Apr 18 00:40:17 2014 -0400

    prng.c: userland was not allowed to use it, but testing uses it
    
    Allow userland to use it if SWAN_TESTING is defined

commit 824e06f46edb97affd1c216f80a9cc25bca6412e
Author: Paul Wouters <pwouters at redhat.com>
Date:   Fri Apr 18 00:39:44 2014 -0400

    Makefile.inc: if GCC or MAKE is not set, set it to gcc/make

commit d67fda1c1eee95af9c9ddddf9ddb25edf54286ac
Author: Paul Wouters <pwouters at redhat.com>
Date:   Fri Apr 18 00:39:08 2014 -0400

    alg_info.c: removed debug line interfering with testing
    
    also added two comments to note we have test cases for these functions

commit ac49d0250aecbf40f63be517efb10ff499c06253
Author: Paul Wouters <pwouters at redhat.com>
Date:   Thu Apr 17 23:06:31 2014 -0400

    testing: Remove testing/packaging
    
    It consisted of obsoleted "make install" targets for various cases.

commit a0b0c13088b293b85ba9efc2da14d16e10d45b88
Author: Paul Wouters <pwouters at redhat.com>
Date:   Thu Apr 17 23:03:25 2014 -0400

    Remove ifdef on obsoleted CONFIG_KLIPS_BIGGATE



More information about the Swan-commit mailing list