des_setparity() cuts keysize to 48 bits; how much do we care?

[Florian Weimer]

On 05/19/2018 10:56 PM, Zack Weinberg wrote:
> des_setparity() is documented to fix up a packed DES key so that each
> byte has odd parity, as required by the DES specification.  The manual
> doesn't do a very good job explaining how it does that, and when I
> went to look at the source code to find out more details, I discovered
> this:
> 
> /*
>   * Table giving odd parity in the low bit for ASCII characters
>   */
> static const char partab[128] =
> { 0x01, 0x01, 0x02, 0x02, 0x04, 0x04, 0x07, 0x07,
>    ... };
> 
> void
> des_setparity (char *p)
> {
>    int i;
> 
>    for (i = 0; i < 8; i++)
>      {
>        *p = partab[*p & 0x7f];
>        p++;
>      }
> }
> 
> So it overwrites the low bit of each byte as necessary to make each
> byte have odd parity, as documented, but it also forces the *high* bit
> of each byte to be zero, which cuts the keyspace down even further - a
> DES key that's passed through des_setparity() has only 48 bits of
> entropy.

I would have understood 40 bit, but 48 bit looks like a bug.

I found this:

Patch-ID# T104620-01 (Preliminary Patch - Not Yet Released)
Keywords: security des_setparity DES RPC
Synopsis: SunOS 5.4: POINT PATCH: 1149767 - des_setparity() fix
SunOS release: 5.4
Relevant Architectures: sparc
BugId's fixed with this patch: 1149767

Apparently, this was when des_setparity_g was introduced.  des_setparity 
was not changed.  So I don't think there is anything left to do.  I'm 
not even sure if these functions should still be documented in the manual.

libtirpc supports RPCSEC_GSS, and I think that's the de-facto successor 
for RPC with cryptography.

Thanks,
Florian