>> So should all occurrences of SHA_ASM just be changed to SHA1_ASM for
>> consistency (or, for even more consistency, all occurrences of
>> SHA1_ASM to SHA_ASM)?
> I vote for more consistency. :)
It wouldn't really result in any consistency because, while there's an
assembler implementation of sha1_block in crypto/sha/asm/, we have
none of sha_block; so
#if !defined(B_ENDIAN) && defined(SHA_ASM)
is really just #ifndef B_ENDIAN, and
#if defined(B_ENDIAN) || defined(SHA_ASM)
is just #ifdef B_ENDIAN. Because of this, Endian_Reverse32 (which is
where the explicit bswapl occurs unless I386_ONLY is defined) is
always used, whereas SHA1 with its assember version of sha1_block
avoids it.
______________________________________________________________________
OpenSSL Project http://www.openssl.org
Development Mailing List [EMAIL PROTECTED]
Automated List Manager [EMAIL PROTECTED]
