On Wed, Oct 31, 2007 at 02:20:49PM +0000, Stuart Henderson wrote:
> On 2007/10/29 22:54, Claudio Jeker wrote:
> > Once again a port thinks that the world is only built of i386 systems.
> > pgpring uses a md5 and sha1 implementation which do not work on big
> > endian systems. The results are wrong key IDs and so many operations where
> > keys are selected fail. pgpring is only used by pgp (gpg does not need
> > it).
> 
> It doesn't break anything for me, but I already see the same
> output on i386, sparc64 and amd64 before patching.
> 
> To test, I just installed pgp (tried 2 and 5), generated a
> key and ran pgpring, should I be doing more than this?
> 

pgpring -5 unpatched:
pub:f:1024:1:9D55D0454D27B591:2000-10-25::::Claudio Jeker <[EMAIL 
PROTECTED]>::es:
pub:q:1024:17:E4F0C5C364FC763E:1997-06-29::::Claudio Jeker <[EMAIL 
PROTECTED]>::s:
sub:q:2047:16:03D82080E3A50272:1997-06-29::::Claudio Jeker <[EMAIL 
PROTECTED]>::e:
pub:f:1022:17:D073122E5B6E1CE5:2007-10-29::::Claudio Jeker <[EMAIL 
PROTECTED]>::s:
sub:f:2047:16:6606E7A2C6CCEE1E:2007-10-29::::Claudio Jeker <[EMAIL 
PROTECTED]>::e:

pgpring -5 patched:
pub:f:1024:1:9D55D0454D27B591:2000-10-25::::Claudio Jeker <[EMAIL 
PROTECTED]>::es:
pub:q:1024:17:833785CBCC18475F:1997-06-29::::Claudio Jeker <[EMAIL 
PROTECTED]>::s:
sub:q:2047:16:0CF922B5B39D33DD:1997-06-29::::Claudio Jeker <[EMAIL 
PROTECTED]>::e:
pub:f:1022:17:619DA11B9D06801F:2007-10-29::::Claudio Jeker <[EMAIL 
PROTECTED]>::s:
sub:f:2047:16:C21600F66FE49665:2007-10-29::::Claudio Jeker <[EMAIL 
PROTECTED]>::e:

pgpk -l output
Type Bits KeyID      Created    Expires    Algorithm       Use
sec+ 1024 0x4D27B591 2000-10-25 ---------- RSA             Sign & Encrypt 
uid  Claudio Jeker <[EMAIL PROTECTED]>

sec  1024 0xCC18475F 1997-06-29 ---------- DSS             Sign & Encrypt 
sub  2048 0xB39D33DD 1997-06-29 ---------- Diffie-Hellman                 
uid  Claudio Jeker <[EMAIL PROTECTED]>

sec+ 1024 0x9D06801F 2007-10-29 ---------- DSS             Sign & Encrypt 
sub  2048 0x6FE49665 2007-10-29 ---------- Diffie-Hellman                 
uid  Claudio Jeker <[EMAIL PROTECTED]>

As you can see the first RSA key is calculated correctly but the IDs of the
DSA keys change between patched and unpatched.

-- 
:wq Claudio

Reply via email to