#define PUTU32(p,v) ((p)[0]=(u8)((v)>>24),(p)[1]=(u8)((v)>>16),(p)[2]=(u8)((v)>>8),(p)[3]=(u8)(v))
#endif
-#define PACK(s) ((size_t)(s)<<(sizeof(size_t)*8-16))
-
-#if 0
+#define PACK(s) ((size_t)(s)<<(sizeof(size_t)*8-16))
+#ifdef TABLE_BITS
+#undef TABLE_BITS
+#endif
/*
- * Under ideal conditions 8-bit version should be twice as fast as
- * 4-bit one. But world is far from ideal. For gcc-generated x86 code,
- * 8-bit was observed to run "only" ~50% faster. On x86_64 observed
+ * Even though permitted values for TABLE_BITS are 8, 4 and 1, it should
+ * never be set to 8. 8 is effectively reserved for testing purposes.
+ * Under ideal conditions "8-bit" version should be twice as fast as
+ * "4-bit" one. But world is far from ideal. For gcc-generated x86 code,
+ * "8-bit" was observed to run only ~50% faster. On x86_64 observed
* improvement was ~75%, much closer to optimal, but the fact of
* deviation means that references to pre-computed tables end up on
* critical path and as tables are pretty big, 4KB per key+1KB shared,
- * execution time is sensitive to cache trashing. It's not actually
+ * execution time is sensitive to cache timing. It's not actually
* proven, but 4-bit procedure is believed to provide adequate
* all-round performance...
*/
+#define TABLE_BITS 4
+
+#if TABLE_BITS==8
+
static void gcm_init_8bit(u128 Htable[256], u64 H[2])
{
int i, j;
else {
u32 T = 0xe1000000U & (0-(u32)(V.lo&1));
V.lo = (V.hi<<63)|(V.lo>>1);
- V.hi = (V.hi>>1) ^((u64)T<<32);
+ V.hi = (V.hi>>1 )^((u64)T<<32);
}
Htable[i] = V;
}
Xi[1] = Z.lo;
}
}
-#endif
+#define GCM_MUL(ctx,Xi) gcm_gmult_8bit(ctx->Xi.u,ctx->Htable)
-#define _4BIT 1 /* change to 0 to switch to 1-bit multiplication */
+#elif TABLE_BITS==4
-#if _4BIT
static void gcm_init_4bit(u128 Htable[16], u64 H[2])
{
int i;
#endif
}
-#ifndef GMULT_ASM
+#ifndef GHASH_ASM
static const size_t rem_4bit[16] = {
PACK(0x0000), PACK(0x1C20), PACK(0x3840), PACK(0x2460),
PACK(0x7080), PACK(0x6CA0), PACK(0x48C0), PACK(0x54E0),
#if !defined(OPENSSL_SMALL_FOOTPRINT)
/*
* Streamed gcm_mult_4bit, see CRYPTO_gcm128_[en|de]crypt for
- * details... It doesn't give any performance improvement, at least
- * not on x86[_64]. It's here mostly as a placeholder for possible
- * future non-trivial optimization[s]...
+ * details... Compiler-generated code doesn't seem to give any
+ * performance improvement, at least not on x86[_64]. It's here
+ * mostly as reference and a placeholder for possible future
+ * non-trivial optimization[s]...
*/
static void gcm_ghash_4bit(const u8 *inp,size_t len,u64 Xi[2], u128 Htable[16])
{
#endif
#define GCM_MUL(ctx,Xi) gcm_gmult_4bit(ctx->Xi.u,ctx->Htable)
-#define GHASH(in,len,ctx) gcm_ghash_4bit(in,len,ctx->Xi.u,ctx->Htable)
+#if defined(GHASH_ASM) || !defined(OPENSSL_SMALL_FOOTPRINT)
+#define GHASH(in,len,ctx) gcm_ghash_4bit(in,len,(ctx)->Xi.u,(ctx)->Htable)
+/* GHASH_CHUNK is "stride parameter" missioned to mitigate cache
+ * trashing effect. In other words idea is to hash data while it's
+ * still in L1 cache after encryption pass... */
#define GHASH_CHUNK 1024
+#endif
-#else /* !_4BIT */
+#else /* TABLE_BITS */
static void gcm_gmult_1bit(u64 Xi[2],const u64 H[2])
{
}
}
#define GCM_MUL(ctx,Xi) gcm_gmult_1bit(ctx->Xi.u,ctx->H.u)
+
#endif
typedef struct {
union { u64 u[2]; u32 d[4]; u8 c[16]; } Yi,EKi,EK0,
Xi,H,
len;
- /* Pre-computed table used by gcm_gmult_4bit */
+ /* Pre-computed table used by gcm_gmult_* */
+#if TABLE_BITS==8
+ u128 Htable[256];
+#else
u128 Htable[16];
+#endif
unsigned int res, ctr;
block128_f block;
void *key;
#endif
}
+#if TABLE_BITS==8
+ gcm_init_8bit(ctx->Htable,ctx->H.u);
+#elif TABLE_BITS==4
gcm_init_4bit(ctx->Htable,ctx->H.u);
+#endif
}
void CRYPTO_gcm128_setiv(GCM128_CONTEXT *ctx,const unsigned char *iv,size_t len)
len -= 16;
}
#endif
-
if (len) {
for (i=0; i<len; ++i) ctx->Xi.c[i] ^= aad[i];
GCM_MUL(ctx,Xi);
if (((size_t)in|(size_t)out)%sizeof(size_t) != 0)
break;
#endif
-#ifdef GHASH
+#if defined(GHASH) && defined(GHASH_CHUNK)
while (len>=GHASH_CHUNK) {
size_t j=GHASH_CHUNK;
if (((size_t)in|(size_t)out)%sizeof(size_t) != 0)
break;
#endif
-#ifdef GHASH
+#if defined(GHASH) && defined(GHASH_CHUNK)
while (len>=GHASH_CHUNK) {
size_t j=GHASH_CHUNK;
IV1[12],
*C1=NULL,
T1[]= {0x58,0xe2,0xfc,0xce,0xfa,0x7e,0x30,0x61,0x36,0x7f,0x1d,0x57,0xa4,0xe7,0x45,0x5a};
+
/* Test Case 2 */
#define K2 K1
#define A2 A1
0x73,0x80,0x69,0x00,0xe4,0x9f,0x24,0xb2,0x2b,0x09,0x75,0x44,0xd4,0x89,0x6b,0x42,
0x49,0x89,0xb5,0xe1,0xeb,0xac,0x0f,0x07,0xc2,0x3f,0x45,0x98},
T5[]= {0x36,0x12,0xd2,0xe7,0x9e,0x3b,0x07,0x85,0x56,0x1b,0xe1,0x4a,0xac,0xa2,0xfc,0xcb};
+
/* Test Case 6 */
#define K6 K5
#define P6 P5
TEST_CASE(17);
TEST_CASE(18);
+#ifdef OPENSSL_CPUID_OBJ
{
size_t start,stop,gcm_t,ctr_t,OPENSSL_rdtsc();
union { u64 u; u8 c[1024]; } buf;
- int i;
AES_set_encrypt_key(K1,sizeof(K1)*8,&key);
CRYPTO_gcm128_init(&ctx,&key,(block128_f)AES_encrypt);
(block128_f)AES_encrypt);
start = OPENSSL_rdtsc();
CRYPTO_ctr128_encrypt(buf.c,buf.c,sizeof(buf),
- &key,ctx.Yi.c,ctx.EKi.c,&ctx.res,
- (block128_f)AES_encrypt);
+ &key,ctx.Yi.c,ctx.EKi.c,&ctx.res,
+ (block128_f)AES_encrypt);
ctr_t = OPENSSL_rdtsc() - start;
printf("%.2f-%.2f=%.2f\n",
gcm_t/(double)sizeof(buf),
ctr_t/(double)sizeof(buf),
(gcm_t-ctr_t)/(double)sizeof(buf));
+#ifdef GHASH
+ GHASH(buf.c,sizeof(buf),&ctx);
+ start = OPENSSL_rdtsc();
+ GHASH(buf.c,sizeof(buf),&ctx);
+ gcm_t = OPENSSL_rdtsc() - start;
+ printf("%.2f\n",gcm_t/(double)sizeof(buf));
+#endif
}
+#endif
return ret;
}
projects / openssl / commitdiff