powerpc64: Add optimized assembly for sha512-compress

powerpc64: Add optimized assembly for sha512-compress

This patch introduces an optimized powerpc64 assembly implementation for
sha512-compress, derived from the implementation for sha256-compress-n.

The following data was captured on a POWER 10 LPAR @ ~4.050GHz

Current C implementation:
         Algorithm         mode Mbyte/s
            sha512       update  474.00
        sha512_224       update  474.61
        sha512_256       update  474.15
       hmac-sha512     64 bytes  104.08
       hmac-sha512    256 bytes  220.42
       hmac-sha512   1024 bytes  368.58
       hmac-sha512   4096 bytes  436.27
       hmac-sha512   single msg  460.10

With optimized assembly:
         Algorithm         mode Mbyte/s
            sha512       update  746.96
        sha512_224       update  746.96
        sha512_256       update  746.93
       hmac-sha512     64 bytes  150.54
       hmac-sha512    256 bytes  327.58
       hmac-sha512   1024 bytes  562.49
       hmac-sha512   4096 bytes  677.38
       hmac-sha512   single msg  713.06

Signed-off-by: Eric Richter <erichte@linux.ibm.com>