@@ -230,6 +230,8 @@ enum {
SCALAR_DATA_BUF_2_HASH_IDX,
GFNI_DATA_BUF_1_HASH_IDX,
GFNI_DATA_BUF_2_HASH_IDX,
+ GFNI_BULK_DATA_BUF_1_HASH_IDX,
+ GFNI_BULK_DATA_BUF_2_HASH_IDX,
HASH_IDXES
};
@@ -241,6 +243,7 @@ test_toeplitz_hash_rand_data(void)
uint32_t hash[HASH_IDXES] = { 0 };
uint64_t rss_key_matrixes[RTE_DIM(default_rss_key)];
int i, j;
+ uint8_t *bulk_data[2];
if (!rte_thash_gfni_supported())
return TEST_SKIPPED;
@@ -248,6 +251,9 @@ test_toeplitz_hash_rand_data(void)
rte_thash_complete_matrix(rss_key_matrixes, default_rss_key,
RTE_DIM(default_rss_key));
+ for (i = 0; i < 2; i++)
+ bulk_data[i] = (uint8_t *)data[i];
+
for (i = 0; i < ITER; i++) {
for (j = 0; j < DATA_SZ; j++) {
data[0][j] = rte_rand();
@@ -266,11 +272,18 @@ test_toeplitz_hash_rand_data(void)
hash[GFNI_DATA_BUF_2_HASH_IDX] = rte_thash_gfni(
rss_key_matrixes, (uint8_t *)data[1],
DATA_SZ * sizeof(uint32_t));
+ rte_thash_gfni_bulk(rss_key_matrixes,
+ DATA_SZ * sizeof(uint32_t), bulk_data,
+ &hash[GFNI_BULK_DATA_BUF_1_HASH_IDX], 2);
if ((hash[SCALAR_DATA_BUF_1_HASH_IDX] !=
hash[GFNI_DATA_BUF_1_HASH_IDX]) ||
+ (hash[SCALAR_DATA_BUF_1_HASH_IDX] !=
+ hash[GFNI_BULK_DATA_BUF_1_HASH_IDX]) ||
(hash[SCALAR_DATA_BUF_2_HASH_IDX] !=
- hash[GFNI_DATA_BUF_2_HASH_IDX]))
+ hash[GFNI_DATA_BUF_2_HASH_IDX]) ||
+ (hash[SCALAR_DATA_BUF_2_HASH_IDX] !=
+ hash[GFNI_BULK_DATA_BUF_2_HASH_IDX]))
return -TEST_FAILED;
}
@@ -284,6 +297,57 @@ enum {
};
static int
+test_toeplitz_hash_gfni_bulk(void)
+{
+ uint32_t i, j;
+ union rte_thash_tuple tuple[2];
+ uint8_t *tuples[2];
+ uint32_t rss[2] = { 0 };
+ uint64_t rss_key_matrixes[RTE_DIM(default_rss_key)];
+
+ if (!rte_thash_gfni_supported())
+ return TEST_SKIPPED;
+
+ /* Convert RSS key into matrixes */
+ rte_thash_complete_matrix(rss_key_matrixes, default_rss_key,
+ RTE_DIM(default_rss_key));
+
+ for (i = 0; i < RTE_DIM(tuples); i++) {
+ /* allocate memory enough for a biggest tuple */
+ tuples[i] = rte_zmalloc(NULL, RTE_THASH_V6_L4_LEN * 4, 0);
+ if (tuples[i] == NULL)
+ return -TEST_FAILED;
+ }
+
+ for (i = 0; i < RTE_MIN(RTE_DIM(v4_tbl), RTE_DIM(v6_tbl)); i++) {
+ /*Load IPv4 headers and copy it into the corresponding tuple*/
+ tuple[0].v4.src_addr = rte_cpu_to_be_32(v4_tbl[i].src_ip);
+ tuple[0].v4.dst_addr = rte_cpu_to_be_32(v4_tbl[i].dst_ip);
+ tuple[0].v4.sport = rte_cpu_to_be_16(v4_tbl[i].dst_port);
+ tuple[0].v4.dport = rte_cpu_to_be_16(v4_tbl[i].src_port);
+ rte_memcpy(tuples[0], &tuple[0], RTE_THASH_V4_L4_LEN * 4);
+
+ /*Load IPv6 headers and copy it into the corresponding tuple*/
+ for (j = 0; j < RTE_DIM(tuple[1].v6.src_addr); j++)
+ tuple[1].v6.src_addr[j] = v6_tbl[i].src_ip[j];
+ for (j = 0; j < RTE_DIM(tuple[1].v6.dst_addr); j++)
+ tuple[1].v6.dst_addr[j] = v6_tbl[i].dst_ip[j];
+ tuple[1].v6.sport = rte_cpu_to_be_16(v6_tbl[i].dst_port);
+ tuple[1].v6.dport = rte_cpu_to_be_16(v6_tbl[i].src_port);
+ rte_memcpy(tuples[1], &tuple[1], RTE_THASH_V6_L4_LEN * 4);
+
+ rte_thash_gfni_bulk(rss_key_matrixes, RTE_THASH_V6_L4_LEN * 4,
+ tuples, rss, 2);
+
+ if ((rss[RSS_V4_IDX] != v4_tbl[i].hash_l3l4) ||
+ (rss[RSS_V6_IDX] != v6_tbl[i].hash_l3l4))
+ return -TEST_FAILED;
+ }
+
+ return TEST_SUCCESS;
+}
+
+static int
test_big_tuple_gfni(void)
{
uint32_t arr[16];
@@ -748,6 +812,7 @@ static struct unit_test_suite thash_tests = {
TEST_CASE(test_toeplitz_hash_calc),
TEST_CASE(test_toeplitz_hash_gfni),
TEST_CASE(test_toeplitz_hash_rand_data),
+ TEST_CASE(test_toeplitz_hash_gfni_bulk),
TEST_CASE(test_big_tuple_gfni),
TEST_CASE(test_create_invalid),
TEST_CASE(test_multiple_create),
@@ -19,11 +19,12 @@ to calculate the RSS hash sum to spread the traffic among the queues.
Toeplitz hash function API
--------------------------
-There are three functions that provide calculation of the Toeplitz hash sum:
+There are four functions that provide calculation of the Toeplitz hash sum:
* ``rte_softrss()``
* ``rte_softrss_be()``
* ``rte_thash_gfni()``
+* ``rte_thash_gfni_bulk()``
First two functions are scalar implementation and take the parameters:
@@ -38,11 +39,12 @@ to be exactly the same as the one installed on the NIC.
The ``rte_softrss_be`` function is a faster implementation,
but it expects ``rss_key`` to be converted to the host byte order.
-The last function is vectorized implementation using
-Galois Fields New Instructions. Could be used if ``rte_thash_gfni_supported`` returns true.
-It expects the tuple to be in network byte order.
+The last two functions are vectorized implementations using
+Galois Fields New Instructions. Could be used if ``rte_thash_gfni_supported`` is true.
+They expect the tuple to be in network byte order.
-``rte_thash_gfni()`` calculates the hash value for a single tuple
+``rte_thash_gfni()`` calculates the hash value for a single tuple, and
+``rte_thash_gfni_bulk()`` bulk implementation of the rte_thash_gfni().
``rte_thash_gfni()`` takes the parameters:
@@ -50,6 +52,14 @@ It expects the tuple to be in network byte order.
* A pointer to the tuple.
* A length of the tuple in bytes.
+``rte_thash_gfni_bulk()`` takes the parameters:
+
+* A pointer to the matrices derived from the RSS hash key using ``rte_thash_complete_matrix()``.
+* A length of the longest tuple in bytes.
+* Array of the pointers on data to be hashed.
+* Array of ``uint32_t`` where to put calculated Toeplitz hash values
+* Number of tuples in a bulk.
+
``rte_thash_complete_matrix()`` is a function that calculates matrices required by
GFNI implementations from the RSS hash key. It takes the parameters:
@@ -45,6 +45,39 @@ rte_thash_gfni(const uint64_t *mtrx __rte_unused,
return 0;
}
+/**
+ * Bulk implementation for Toeplitz hash.
+ * Dummy implementation.
+ *
+ * @warning
+ * @b EXPERIMENTAL: this API may change without prior notice.
+ *
+ * @param m
+ * Pointer to the matrices generated from the corresponding
+ * RSS hash key using rte_thash_complete_matrix().
+ * @param len
+ * Length of the largest data buffer to be hashed.
+ * @param tuple
+ * Array of the pointers on data to be hashed.
+ * Data must be in network byte order.
+ * @param val
+ * Array of uint32_t where to put calculated Toeplitz hash values
+ * @param num
+ * Number of tuples to hash.
+ */
+__rte_experimental
+static inline void
+rte_thash_gfni_bulk(const uint64_t *mtrx __rte_unused,
+ int len __rte_unused, uint8_t *tuple[] __rte_unused,
+ uint32_t val[], uint32_t num)
+{
+ unsigned int i;
+
+ RTE_LOG(ERR, HASH, "%s is undefined under given arch\n", __func__);
+ for (i = 0; i < num; i++)
+ val[i] = 0;
+}
+
#endif /* RTE_THASH_GFNI_DEFINED */
#ifdef __cplusplus
@@ -173,6 +173,45 @@ rte_thash_gfni(const uint64_t *m, const uint8_t *tuple, int len)
return val;
}
+/**
+ * Bulk implementation for Toeplitz hash.
+ *
+ * @warning
+ * @b EXPERIMENTAL: this API may change without prior notice.
+ *
+ * @param m
+ * Pointer to the matrices generated from the corresponding
+ * RSS hash key using rte_thash_complete_matrix().
+ * @param len
+ * Length of the largest data buffer to be hashed.
+ * @param tuple
+ * Array of the pointers on data to be hashed.
+ * Data must be in network byte order.
+ * @param val
+ * Array of uint32_t where to put calculated Toeplitz hash values
+ * @param num
+ * Number of tuples to hash.
+ */
+__rte_experimental
+static inline void
+rte_thash_gfni_bulk(const uint64_t *mtrx, int len, uint8_t *tuple[],
+ uint32_t val[], uint32_t num)
+{
+ uint32_t i;
+ uint32_t val_zero;
+ __m512i xor_acc;
+
+ for (i = 0; i != (num & ~1); i += 2) {
+ xor_acc = __rte_thash_gfni(mtrx, tuple[i], tuple[i + 1], len);
+ __rte_thash_xor_reduce(xor_acc, val + i, val + i + 1);
+ }
+
+ if (num & 1) {
+ xor_acc = __rte_thash_gfni(mtrx, tuple[i], NULL, len);
+ __rte_thash_xor_reduce(xor_acc, val + i, &val_zero);
+ }
+}
+
#endif /* _GFNI_ */
#ifdef __cplusplus