#define SHA512_BLOCK_CAN_MANAGE_UNALIGNED_DATA
#endif
-int SHA384_Init (SHA512_CTX *c)
+fips_md_init_ctx(SHA384, SHA512)
{
c->h[0]=U64(0xcbbb9d5dc1059ed8);
c->h[1]=U64(0x629a292a367cd507);
c->h[5]=U64(0x8eb44a8768581511);
c->h[6]=U64(0xdb0c2e0d64f98fa7);
c->h[7]=U64(0x47b5481dbefa4fa4);
+
c->Nl=0; c->Nh=0;
c->num=0; c->md_len=SHA384_DIGEST_LENGTH;
return 1;
}
-int SHA512_Init (SHA512_CTX *c)
+fips_md_init(SHA512)
{
c->h[0]=U64(0x6a09e667f3bcc908);
c->h[1]=U64(0xbb67ae8584caa73b);
c->h[5]=U64(0x9b05688c2b3e6c1f);
c->h[6]=U64(0x1f83d9abfb41bd6b);
c->h[7]=U64(0x5be0cd19137e2179);
+
c->Nl=0; c->Nh=0;
c->num=0; c->md_len=SHA512_DIGEST_LENGTH;
return 1;
if (len < n)
{
- memcpy (p+c->num,data,len), c->num += len;
+ memcpy (p+c->num,data,len), c->num += (unsigned int)len;
return 1;
}
else {
#ifndef PEDANTIC
# if defined(__GNUC__) && __GNUC__>=2 && !defined(OPENSSL_NO_ASM) && !defined(OPENSSL_NO_INLINE_ASM)
# if defined(__x86_64) || defined(__x86_64__)
-# define ROTR(a,n) ({ unsigned long ret; \
+# define ROTR(a,n) ({ SHA_LONG64 ret; \
asm ("rorq %1,%0" \
: "=r"(ret) \
: "J"(n),"0"(a) \
# elif (defined(__i386) || defined(__i386__)) && !defined(B_ENDIAN)
# if defined(I386_ONLY)
# define PULL64(x) ({ const unsigned int *p=(const unsigned int *)(&(x));\
- unsigned int hi,lo; \
+ unsigned int hi=p[0],lo=p[1]; \
asm("xchgb %%ah,%%al;xchgb %%dh,%%dl;"\
"roll $16,%%eax; roll $16,%%edx; "\
"xchgb %%ah,%%al;xchgb %%dh,%%dl;" \
: "=a"(lo),"=d"(hi) \
- : "0"(p[1]),"1"(p[0]) : "cc"); \
+ : "0"(lo),"1"(hi) : "cc"); \
((SHA_LONG64)hi)<<32|lo; })
# else
# define PULL64(x) ({ const unsigned int *p=(const unsigned int *)(&(x));\
- unsigned int hi,lo; \
+ unsigned int hi=p[0],lo=p[1]; \
asm ("bswapl %0; bswapl %1;" \
: "=r"(lo),"=r"(hi) \
- : "0"(p[1]),"1"(p[0])); \
+ : "0"(lo),"1"(hi)); \
((SHA_LONG64)hi)<<32|lo; })
# endif
# elif (defined(_ARCH_PPC) && defined(__64BIT__)) || defined(_ARCH_PPC64)
-# define ROTR(a,n) ({ unsigned long ret; \
+# define ROTR(a,n) ({ SHA_LONG64 ret; \
asm ("rotrdi %0,%1,%2" \
: "=r"(ret) \
: "r"(a),"K"(n)); ret; })
# endif
# elif defined(_MSC_VER)
# if defined(_WIN64) /* applies to both IA-64 and AMD64 */
+# pragma intrinsic(_rotr64)
# define ROTR(a,n) _rotr64((a),n)
# endif
# if defined(_M_IX86) && !defined(OPENSSL_NO_ASM) && !defined(OPENSSL_NO_INLINE_ASM)
#if defined(__i386) || defined(__i386__) || defined(_M_IX86)
-
-#if defined(OPENSSL_IA32_SSE2) && !defined(OPENSSL_NO_ASM) && !defined(I386_ONLY)
-#define GO_FOR_SSE2(ctx,in,num) do { \
- void sha512_block_sse2(void *,const void *,size_t); \
- if (!(OPENSSL_ia32cap_P & (1<<26))) break; \
- sha512_block_sse2(ctx->h,in,num); return; \
- } while (0)
-#endif
-
+/*
+ * This code should give better results on 32-bit CPU with less than
+ * ~24 registers, both size and performance wise...
+ */
static void sha512_block_data_order (SHA512_CTX *ctx, const void *in, size_t num)
{
const SHA_LONG64 *W=in;
- SHA_LONG64 T1;
+ SHA_LONG64 A,E,T;
SHA_LONG64 X[9+80],*F;
int i;
-#ifdef GO_FOR_SSE2
- GO_FOR_SSE2(ctx,in,num);
-#endif
-
while (num--) {
- F = X+80;
- F[0] = ctx->h[0]; F[1] = ctx->h[1];
+ F = X+80;
+ A = ctx->h[0]; F[1] = ctx->h[1];
F[2] = ctx->h[2]; F[3] = ctx->h[3];
- F[4] = ctx->h[4]; F[5] = ctx->h[5];
+ E = ctx->h[4]; F[5] = ctx->h[5];
F[6] = ctx->h[6]; F[7] = ctx->h[7];
for (i=0;i<16;i++,F--)
{
#ifdef B_ENDIAN
- T1 = W[i];
+ T = W[i];
#else
- T1 = PULL64(W[i]);
+ T = PULL64(W[i]);
#endif
- F[8] = T1;
- T1 += F[7] + Sigma1(F[4]) + Ch(F[4],F[5],F[6]) + K512[i];
- F[3] += T1;
- T1 += Sigma0(F[0]) + Maj(F[0],F[1],F[2]);
- F[-1] = T1;
+ F[0] = A;
+ F[4] = E;
+ F[8] = T;
+ T += F[7] + Sigma1(E) + Ch(E,F[5],F[6]) + K512[i];
+ E = F[3] + T;
+ A = T + Sigma0(A) + Maj(A,F[1],F[2]);
}
for (;i<80;i++,F--)
{
- T1 = sigma0(F[8+16-1]);
- T1 += sigma1(F[8+16-14]);
-
- T1 += F[8+16] + F[8+16-9];
- F[8] = T1;
- T1 += F[7] + Sigma1(F[4]) + Ch(F[4],F[5],F[6]) + K512[i];
- F[3] += T1;
- T1 += Sigma0(F[0]) + Maj(F[0],F[1],F[2]);
- F[-1] = T1;
+ T = sigma0(F[8+16-1]);
+ T += sigma1(F[8+16-14]);
+ T += F[8+16] + F[8+16-9];
+
+ F[0] = A;
+ F[4] = E;
+ F[8] = T;
+ T += F[7] + Sigma1(E) + Ch(E,F[5],F[6]) + K512[i];
+ E = F[3] + T;
+ A = T + Sigma0(A) + Maj(A,F[1],F[2]);
}
- ctx->h[0] += F[0]; ctx->h[1] += F[1];
+ ctx->h[0] += A; ctx->h[1] += F[1];
ctx->h[2] += F[2]; ctx->h[3] += F[3];
- ctx->h[4] += F[4]; ctx->h[5] += F[5];
+ ctx->h[4] += E; ctx->h[5] += F[5];
ctx->h[6] += F[6]; ctx->h[7] += F[7];
W+=SHA_LBLOCK;
#endif /* SHA512_ASM */
-#endif /* OPENSSL_NO_SHA512 */
+#else /* !OPENSSL_NO_SHA512 */
+
+#if defined(PEDANTIC) || defined(__DECC) || defined(OPENSSL_SYS_MACOSX)
+static void *dummy=&dummy;
+#endif
+
+#endif /* !OPENSSL_NO_SHA512 */