-/*-
- * BSD LICENSE
- *
- * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
- * All rights reserved.
- *
- * Redistribution and use in source and binary forms, with or without
- * modification, are permitted provided that the following conditions
- * are met:
- *
- * * Redistributions of source code must retain the above copyright
- * notice, this list of conditions and the following disclaimer.
- * * Redistributions in binary form must reproduce the above copyright
- * notice, this list of conditions and the following disclaimer in
- * the documentation and/or other materials provided with the
- * distribution.
- * * Neither the name of Intel Corporation nor the names of its
- * contributors may be used to endorse or promote products derived
- * from this software without specific prior written permission.
- *
- * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2010-2014 Intel Corporation
+ * Copyright(c) 2019 Arm Limited
*/
#include <stdio.h>
#include <stdint.h>
#include <unistd.h>
+#include <inttypes.h>
#include <sys/queue.h>
#include <rte_memory.h>
-#include <rte_memzone.h>
#include <rte_per_lcore.h>
#include <rte_launch.h>
#include <rte_atomic.h>
#include <rte_eal.h>
-#include <rte_per_lcore.h>
#include <rte_lcore.h>
+#include <rte_random.h>
+#include <rte_hash_crc.h>
#include "test.h"
* Atomic Variables
* ================
*
- * - The main test function performs three subtests. The first test
+ * - The main test function performs several subtests. The first
* checks that the usual inc/dec/add/sub functions are working
* correctly:
*
* - The function checks that once all lcores finish their function,
* the value of the atomic variables are still the same.
*
- * - The second test verifies the behavior of "test and set" functions.
+ * - Test "test and set" functions.
*
* - Initialize 16-bit, 32-bit and 64-bit atomic variables to zero.
*
* - The main function checks that the atomic counter was incremented
* twice only (one for 16-bit, one for 32-bit and one for 64-bit values).
*
- * - Test "add/sub and return"
+ * - Test "add/sub and return" functions
*
* - Initialize 16-bit, 32-bit and 64-bit atomic variables to zero.
*
* atomic_sub(&count, tmp+1);
*
* - At the end of the test, the *count* value must be 0.
+ *
+ * - Test "128-bit compare and swap" (aarch64 and x86_64 only)
+ *
+ * - Initialize 128-bit atomic variables to zero.
+ *
+ * - Invoke ``test_atomic128_cmp_exchange()`` on each lcore. Before doing
+ * anything else, the cores are waiting a synchro. Each lcore does
+ * these compare and swap (CAS) operations several times::
+ *
+ * Acquired CAS update counter.val[0] + 2; counter.val[1] + 1;
+ * Released CAS update counter.val[0] + 2; counter.val[1] + 1;
+ * Acquired_Released CAS update counter.val[0] + 2; counter.val[1] + 1;
+ * Relaxed CAS update counter.val[0] + 2; counter.val[1] + 1;
+ *
+ * - At the end of the test, the *count128* first 64-bit value and
+ * second 64-bit value differ by the total iterations.
+ *
+ * - Test "atomic exchange" functions
+ *
+ * - Create a 64 bit token that can be tested for data integrity
+ *
+ * - Invoke ``test_atomic_exchange`` on each lcore. Before doing
+ * anything else, the cores wait for a synchronization event.
+ * Each core then does the follwoing for N iterations:
+ *
+ * Generate a new token with a data integrity check
+ * Exchange the new token for previously generated token
+ * Increment a counter if a corrupt token was received
+ *
+ * - At the end of the test, the number of corrupted tokens must be 0.
*/
#define NUM_ATOMIC_TYPES 3
-#define N 10000
+#define N 1000000
static rte_atomic16_t a16;
static rte_atomic32_t a32;
static rte_atomic32_t synchro;
static int
-test_atomic_usual(__attribute__((unused)) void *arg)
+test_atomic_usual(__rte_unused void *arg)
{
unsigned i;
}
static int
-test_atomic_tas(__attribute__((unused)) void *arg)
+test_atomic_tas(__rte_unused void *arg)
{
while (rte_atomic32_read(&synchro) == 0)
;
}
static int
-test_atomic_addsub_and_return(__attribute__((unused)) void *arg)
+test_atomic_addsub_and_return(__rte_unused void *arg)
{
uint32_t tmp16;
uint32_t tmp32;
*
*/
static int
-test_atomic_inc_and_test(__attribute__((unused)) void *arg)
+test_atomic_inc_and_test(__rte_unused void *arg)
{
while (rte_atomic32_read(&synchro) == 0)
;
* be checked as the result later.
*/
static int
-test_atomic_dec_and_test(__attribute__((unused)) void *arg)
+test_atomic_dec_and_test(__rte_unused void *arg)
{
while (rte_atomic32_read(&synchro) == 0)
;
return 0;
}
+#if defined(RTE_ARCH_X86_64) || defined(RTE_ARCH_ARM64)
+static rte_int128_t count128;
+
+/*
+ * rte_atomic128_cmp_exchange() should update a 128 bits counter's first 64
+ * bits by 2 and the second 64 bits by 1 in this test. It should return true
+ * if the compare exchange operation is successful.
+ * This test repeats 128 bits compare and swap operations N rounds. In each
+ * iteration it runs compare and swap operation with different memory models.
+ */
+static int
+test_atomic128_cmp_exchange(__rte_unused void *arg)
+{
+ rte_int128_t expected;
+ int success;
+ unsigned int i;
+
+ while (rte_atomic32_read(&synchro) == 0)
+ ;
+
+ expected = count128;
+
+ for (i = 0; i < N; i++) {
+ do {
+ rte_int128_t desired;
+
+ desired.val[0] = expected.val[0] + 2;
+ desired.val[1] = expected.val[1] + 1;
+
+ success = rte_atomic128_cmp_exchange(&count128,
+ &expected, &desired, 1,
+ __ATOMIC_ACQUIRE, __ATOMIC_RELAXED);
+ } while (success == 0);
+
+ do {
+ rte_int128_t desired;
+
+ desired.val[0] = expected.val[0] + 2;
+ desired.val[1] = expected.val[1] + 1;
+
+ success = rte_atomic128_cmp_exchange(&count128,
+ &expected, &desired, 1,
+ __ATOMIC_RELEASE, __ATOMIC_RELAXED);
+ } while (success == 0);
+
+ do {
+ rte_int128_t desired;
+
+ desired.val[0] = expected.val[0] + 2;
+ desired.val[1] = expected.val[1] + 1;
+
+ success = rte_atomic128_cmp_exchange(&count128,
+ &expected, &desired, 1,
+ __ATOMIC_ACQ_REL, __ATOMIC_RELAXED);
+ } while (success == 0);
+
+ do {
+ rte_int128_t desired;
+
+ desired.val[0] = expected.val[0] + 2;
+ desired.val[1] = expected.val[1] + 1;
+
+ success = rte_atomic128_cmp_exchange(&count128,
+ &expected, &desired, 1,
+ __ATOMIC_RELAXED, __ATOMIC_RELAXED);
+ } while (success == 0);
+ }
+
+ return 0;
+}
+#endif
+
+/*
+ * Helper definitions/variables/functions for
+ * atomic exchange tests
+ */
+typedef union {
+ uint16_t u16;
+ uint8_t u8[2];
+} test16_t;
+
+typedef union {
+ uint32_t u32;
+ uint16_t u16[2];
+ uint8_t u8[4];
+} test32_t;
+
+typedef union {
+ uint64_t u64;
+ uint32_t u32[2];
+ uint16_t u16[4];
+ uint8_t u8[8];
+} test64_t;
+
+const uint8_t CRC8_POLY = 0x91;
+uint8_t crc8_table[256];
+
+volatile uint16_t token16;
+volatile uint32_t token32;
+volatile uint64_t token64;
+
+static void
+build_crc8_table(void)
+{
+ uint8_t val;
+ int i, j;
+
+ for (i = 0; i < 256; i++) {
+ val = i;
+ for (j = 0; j < 8; j++) {
+ if (val & 1)
+ val ^= CRC8_POLY;
+ val >>= 1;
+ }
+ crc8_table[i] = val;
+ }
+}
+
+static uint8_t
+get_crc8(uint8_t *message, int length)
+{
+ uint8_t crc = 0;
+ int i;
+
+ for (i = 0; i < length; i++)
+ crc = crc8_table[crc ^ message[i]];
+ return crc;
+}
+
+/*
+ * The atomic exchange test sets up a token in memory and
+ * then spins up multiple lcores whose job is to generate
+ * new tokens, exchange that new token for the old one held
+ * in memory, and then verify that the old token is still
+ * valid (i.e. the exchange did not corrupt the token).
+ *
+ * A token is made up of random data and 8 bits of crc
+ * covering that random data. The following is an example
+ * of a 64bit token.
+ *
+ * +------------+------------+
+ * | 63 56 | 55 0 |
+ * +------------+------------+
+ * | CRC8 | Data |
+ * +------------+------------+
+ */
+static int
+test_atomic_exchange(__rte_unused void *arg)
+{
+ int i;
+ test16_t nt16, ot16; /* new token, old token */
+ test32_t nt32, ot32;
+ test64_t nt64, ot64;
+
+ /* Wait until all of the other threads have been dispatched */
+ while (rte_atomic32_read(&synchro) == 0)
+ ;
+
+ /*
+ * Let the battle begin! Every thread attempts to steal the current
+ * token with an atomic exchange operation and install its own newly
+ * generated token. If the old token is valid (i.e. it has the
+ * appropriate crc32 hash for the data) then the test iteration has
+ * passed. If the token is invalid, increment the counter.
+ */
+ for (i = 0; i < N; i++) {
+
+ /* Test 64bit Atomic Exchange */
+ nt64.u64 = rte_rand();
+ nt64.u8[7] = get_crc8(&nt64.u8[0], sizeof(nt64) - 1);
+ ot64.u64 = rte_atomic64_exchange(&token64, nt64.u64);
+ if (ot64.u8[7] != get_crc8(&ot64.u8[0], sizeof(ot64) - 1))
+ rte_atomic64_inc(&count);
+
+ /* Test 32bit Atomic Exchange */
+ nt32.u32 = (uint32_t)rte_rand();
+ nt32.u8[3] = get_crc8(&nt32.u8[0], sizeof(nt32) - 1);
+ ot32.u32 = rte_atomic32_exchange(&token32, nt32.u32);
+ if (ot32.u8[3] != get_crc8(&ot32.u8[0], sizeof(ot32) - 1))
+ rte_atomic64_inc(&count);
+
+ /* Test 16bit Atomic Exchange */
+ nt16.u16 = (uint16_t)rte_rand();
+ nt16.u8[1] = get_crc8(&nt16.u8[0], sizeof(nt16) - 1);
+ ot16.u16 = rte_atomic16_exchange(&token16, nt16.u16);
+ if (ot16.u8[1] != get_crc8(&ot16.u8[0], sizeof(ot16) - 1))
+ rte_atomic64_inc(&count);
+ }
+
+ return 0;
+}
static int
test_atomic(void)
{
return -1;
}
+#if defined(RTE_ARCH_X86_64) || defined(RTE_ARCH_ARM64)
+ /*
+ * This case tests the functionality of rte_atomic128_cmp_exchange
+ * API. It calls rte_atomic128_cmp_exchange with four kinds of memory
+ * models successively on each slave core. Once each 128-bit atomic
+ * compare and swap operation is successful, it updates the global
+ * 128-bit counter by 2 for the first 64-bit and 1 for the second
+ * 64-bit. Each slave core iterates this test N times.
+ * At the end of test, verify whether the first 64-bits of the 128-bit
+ * counter and the second 64bits is differ by the total iterations. If
+ * it is, the test passes.
+ */
+ printf("128-bit compare and swap test\n");
+ uint64_t iterations = 0;
+
+ rte_atomic32_clear(&synchro);
+ count128.val[0] = 0;
+ count128.val[1] = 0;
+
+ rte_eal_mp_remote_launch(test_atomic128_cmp_exchange, NULL,
+ SKIP_MASTER);
+ rte_atomic32_set(&synchro, 1);
+ rte_eal_mp_wait_lcore();
+ rte_atomic32_clear(&synchro);
+
+ iterations = count128.val[0] - count128.val[1];
+ if (iterations != 4*N*(rte_lcore_count()-1)) {
+ printf("128-bit compare and swap failed\n");
+ return -1;
+ }
+#endif
+
+ /*
+ * Test 16/32/64bit atomic exchange.
+ */
+ test64_t t;
+
+ printf("exchange test\n");
+
+ rte_atomic32_clear(&synchro);
+ rte_atomic64_clear(&count);
+
+ /* Generate the CRC8 lookup table */
+ build_crc8_table();
+
+ /* Create the initial tokens used by the test */
+ t.u64 = rte_rand();
+ token16 = (get_crc8(&t.u8[0], sizeof(token16) - 1) << 8)
+ | (t.u16[0] & 0x00ff);
+ token32 = ((uint32_t)get_crc8(&t.u8[0], sizeof(token32) - 1) << 24)
+ | (t.u32[0] & 0x00ffffff);
+ token64 = ((uint64_t)get_crc8(&t.u8[0], sizeof(token64) - 1) << 56)
+ | (t.u64 & 0x00ffffffffffffff);
+
+ rte_eal_mp_remote_launch(test_atomic_exchange, NULL, SKIP_MASTER);
+ rte_atomic32_set(&synchro, 1);
+ rte_eal_mp_wait_lcore();
+ rte_atomic32_clear(&synchro);
+
+ if (rte_atomic64_read(&count) > 0) {
+ printf("Atomic exchange test failed\n");
+ return -1;
+ }
+
return 0;
}
-
-static struct test_command atomic_cmd = {
- .command = "atomic_autotest",
- .callback = test_atomic,
-};
-REGISTER_TEST_COMMAND(atomic_cmd);
+REGISTER_TEST_COMMAND(atomic_autotest, test_atomic);