tests: Move benchmarks into a separate folder

Make it clear that these files are related to benchmarks by moving
them into a new folder called "bench".

Message-Id: <20210312092238.79509-1-thuth@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

7 files changed, 1311 insertions, 0 deletions

diff --git a/tests/bench/atomic64-bench.c b/tests/bench/atomic64-bench.c
new file mode 100644
index 0000000000..e474753d34
--- /dev/null
+++ b/tests/bench/atomic64-bench.c
@@ -0,0 +1,169 @@
+/*
+ * Copyright (C) 2018, Emilio G. Cota <cota@braap.org>
+ *
+ * License: GNU GPL, version 2 or later.
+ *   See the COPYING file in the top-level directory.
+ */
+#include "qemu/osdep.h"
+#include "qemu/thread.h"
+#include "qemu/host-utils.h"
+#include "qemu/processor.h"
+
+struct thread_info {
+    uint64_t r;
+    uint64_t accesses;
+} QEMU_ALIGNED(64);
+
+struct count {
+    int64_t i64;
+} QEMU_ALIGNED(64);
+
+static QemuThread *threads;
+static struct thread_info *th_info;
+static unsigned int n_threads = 1;
+static unsigned int n_ready_threads;
+static struct count *counts;
+static unsigned int duration = 1;
+static unsigned int range = 1024;
+static bool test_start;
+static bool test_stop;
+
+static const char commands_string[] =
+    " -d = duration in seconds\n"
+    " -n = number of threads\n"
+    " -r = range (will be rounded up to pow2)";
+
+static void usage_complete(char *argv[])
+{
+    fprintf(stderr, "Usage: %s [options]\n", argv[0]);
+    fprintf(stderr, "options:\n%s\n", commands_string);
+}
+
+/*
+ * From: https://en.wikipedia.org/wiki/Xorshift
+ * This is faster than rand_r(), and gives us a wider range (RAND_MAX is only
+ * guaranteed to be >= INT_MAX).
+ */
+static uint64_t xorshift64star(uint64_t x)
+{
+    x ^= x >> 12; /* a */
+    x ^= x << 25; /* b */
+    x ^= x >> 27; /* c */
+    return x * UINT64_C(2685821657736338717);
+}
+
+static void *thread_func(void *arg)
+{
+    struct thread_info *info = arg;
+
+    qatomic_inc(&n_ready_threads);
+    while (!qatomic_read(&test_start)) {
+        cpu_relax();
+    }
+
+    while (!qatomic_read(&test_stop)) {
+        unsigned int index;
+
+        info->r = xorshift64star(info->r);
+        index = info->r & (range - 1);
+        qatomic_read_i64(&counts[index].i64);
+        info->accesses++;
+    }
+    return NULL;
+}
+
+static void run_test(void)
+{
+    unsigned int i;
+
+    while (qatomic_read(&n_ready_threads) != n_threads) {
+        cpu_relax();
+    }
+
+    qatomic_set(&test_start, true);
+    g_usleep(duration * G_USEC_PER_SEC);
+    qatomic_set(&test_stop, true);
+
+    for (i = 0; i < n_threads; i++) {
+        qemu_thread_join(&threads[i]);
+    }
+}
+
+static void create_threads(void)
+{
+    unsigned int i;
+
+    threads = g_new(QemuThread, n_threads);
+    th_info = g_new(struct thread_info, n_threads);
+    counts = g_malloc0_n(range, sizeof(*counts));
+
+    for (i = 0; i < n_threads; i++) {
+        struct thread_info *info = &th_info[i];
+
+        info->r = (i + 1) ^ time(NULL);
+        info->accesses = 0;
+        qemu_thread_create(&threads[i], NULL, thread_func, info,
+                           QEMU_THREAD_JOINABLE);
+    }
+}
+
+static void pr_params(void)
+{
+    printf("Parameters:\n");
+    printf(" # of threads:      %u\n", n_threads);
+    printf(" duration:          %u\n", duration);
+    printf(" ops' range:        %u\n", range);
+}
+
+static void pr_stats(void)
+{
+    unsigned long long val = 0;
+    double tx;
+    int i;
+
+    for (i = 0; i < n_threads; i++) {
+        val += th_info[i].accesses;
+    }
+    tx = val / duration / 1e6;
+
+    printf("Results:\n");
+    printf("Duration:            %u s\n", duration);
+    printf(" Throughput:         %.2f Mops/s\n", tx);
+    printf(" Throughput/thread:  %.2f Mops/s/thread\n", tx / n_threads);
+}
+
+static void parse_args(int argc, char *argv[])
+{
+    int c;
+
+    for (;;) {
+        c = getopt(argc, argv, "hd:n:r:");
+        if (c < 0) {
+            break;
+        }
+        switch (c) {
+        case 'h':
+            usage_complete(argv);
+            exit(0);
+        case 'd':
+            duration = atoi(optarg);
+            break;
+        case 'n':
+            n_threads = atoi(optarg);
+            break;
+        case 'r':
+            range = pow2ceil(atoi(optarg));
+            break;
+        }
+    }
+}
+
+int main(int argc, char *argv[])
+{
+    parse_args(argc, argv);
+    pr_params();
+    create_threads();
+    run_test();
+    pr_stats();
+    return 0;
+}
diff --git a/tests/bench/atomic_add-bench.c b/tests/bench/atomic_add-bench.c
new file mode 100644
index 0000000000..f05471ab45
--- /dev/null
+++ b/tests/bench/atomic_add-bench.c
@@ -0,0 +1,180 @@
+#include "qemu/osdep.h"
+#include "qemu/thread.h"
+#include "qemu/host-utils.h"
+#include "qemu/processor.h"
+
+struct thread_info {
+    uint64_t r;
+} QEMU_ALIGNED(64);
+
+struct count {
+    QemuMutex lock;
+    unsigned long val;
+} QEMU_ALIGNED(64);
+
+static QemuThread *threads;
+static struct thread_info *th_info;
+static unsigned int n_threads = 1;
+static unsigned int n_ready_threads;
+static struct count *counts;
+static unsigned int duration = 1;
+static unsigned int range = 1024;
+static bool use_mutex;
+static bool test_start;
+static bool test_stop;
+
+static const char commands_string[] =
+    " -n = number of threads\n"
+    " -m = use mutexes instead of atomic increments\n"
+    " -p = enable sync profiler\n"
+    " -d = duration in seconds\n"
+    " -r = range (will be rounded up to pow2)";
+
+static void usage_complete(char *argv[])
+{
+    fprintf(stderr, "Usage: %s [options]\n", argv[0]);
+    fprintf(stderr, "options:\n%s\n", commands_string);
+}
+
+/*
+ * From: https://en.wikipedia.org/wiki/Xorshift
+ * This is faster than rand_r(), and gives us a wider range (RAND_MAX is only
+ * guaranteed to be >= INT_MAX).
+ */
+static uint64_t xorshift64star(uint64_t x)
+{
+    x ^= x >> 12; /* a */
+    x ^= x << 25; /* b */
+    x ^= x >> 27; /* c */
+    return x * UINT64_C(2685821657736338717);
+}
+
+static void *thread_func(void *arg)
+{
+    struct thread_info *info = arg;
+
+    qatomic_inc(&n_ready_threads);
+    while (!qatomic_read(&test_start)) {
+        cpu_relax();
+    }
+
+    while (!qatomic_read(&test_stop)) {
+        unsigned int index;
+
+        info->r = xorshift64star(info->r);
+        index = info->r & (range - 1);
+        if (use_mutex) {
+            qemu_mutex_lock(&counts[index].lock);
+            counts[index].val += 1;
+            qemu_mutex_unlock(&counts[index].lock);
+        } else {
+            qatomic_inc(&counts[index].val);
+        }
+    }
+    return NULL;
+}
+
+static void run_test(void)
+{
+    unsigned int i;
+
+    while (qatomic_read(&n_ready_threads) != n_threads) {
+        cpu_relax();
+    }
+
+    qatomic_set(&test_start, true);
+    g_usleep(duration * G_USEC_PER_SEC);
+    qatomic_set(&test_stop, true);
+
+    for (i = 0; i < n_threads; i++) {
+        qemu_thread_join(&threads[i]);
+    }
+}
+
+static void create_threads(void)
+{
+    unsigned int i;
+
+    threads = g_new(QemuThread, n_threads);
+    th_info = g_new(struct thread_info, n_threads);
+    counts = qemu_memalign(64, sizeof(*counts) * range);
+    memset(counts, 0, sizeof(*counts) * range);
+    for (i = 0; i < range; i++) {
+        qemu_mutex_init(&counts[i].lock);
+    }
+
+    for (i = 0; i < n_threads; i++) {
+        struct thread_info *info = &th_info[i];
+
+        info->r = (i + 1) ^ time(NULL);
+        qemu_thread_create(&threads[i], NULL, thread_func, info,
+                           QEMU_THREAD_JOINABLE);
+    }
+}
+
+static void pr_params(void)
+{
+    printf("Parameters:\n");
+    printf(" # of threads:      %u\n", n_threads);
+    printf(" duration:          %u\n", duration);
+    printf(" ops' range:        %u\n", range);
+}
+
+static void pr_stats(void)
+{
+    unsigned long long val = 0;
+    unsigned int i;
+    double tx;
+
+    for (i = 0; i < range; i++) {
+        val += counts[i].val;
+    }
+    tx = val / duration / 1e6;
+
+    printf("Results:\n");
+    printf("Duration:            %u s\n", duration);
+    printf(" Throughput:         %.2f Mops/s\n", tx);
+    printf(" Throughput/thread:  %.2f Mops/s/thread\n", tx / n_threads);
+}
+
+static void parse_args(int argc, char *argv[])
+{
+    int c;
+
+    for (;;) {
+        c = getopt(argc, argv, "hd:n:mpr:");
+        if (c < 0) {
+            break;
+        }
+        switch (c) {
+        case 'h':
+            usage_complete(argv);
+            exit(0);
+        case 'd':
+            duration = atoi(optarg);
+            break;
+        case 'n':
+            n_threads = atoi(optarg);
+            break;
+        case 'm':
+            use_mutex = true;
+            break;
+        case 'p':
+            qsp_enable();
+            break;
+        case 'r':
+            range = pow2ceil(atoi(optarg));
+            break;
+        }
+    }
+}
+
+int main(int argc, char *argv[])
+{
+    parse_args(argc, argv);
+    pr_params();
+    create_threads();
+    run_test();
+    pr_stats();
+    return 0;
+}
diff --git a/tests/bench/benchmark-crypto-cipher.c b/tests/bench/benchmark-crypto-cipher.c
new file mode 100644
index 0000000000..c04f0a0fba
--- /dev/null
+++ b/tests/bench/benchmark-crypto-cipher.c
@@ -0,0 +1,208 @@
+/*
+ * QEMU Crypto cipher speed benchmark
+ *
+ * Copyright (c) 2017 HUAWEI TECHNOLOGIES CO., LTD.
+ *
+ * Authors:
+ *    Longpeng(Mike) <longpeng2@huawei.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or
+ * (at your option) any later version.  See the COPYING file in the
+ * top-level directory.
+ */
+#include "qemu/osdep.h"
+#include "qemu/units.h"
+#include "crypto/init.h"
+#include "crypto/cipher.h"
+
+static void test_cipher_speed(size_t chunk_size,
+                              QCryptoCipherMode mode,
+                              QCryptoCipherAlgorithm alg)
+{
+    QCryptoCipher *cipher;
+    Error *err = NULL;
+    uint8_t *key = NULL, *iv = NULL;
+    uint8_t *plaintext = NULL, *ciphertext = NULL;
+    size_t nkey;
+    size_t niv;
+    const size_t total = 2 * GiB;
+    size_t remain;
+
+    if (!qcrypto_cipher_supports(alg, mode)) {
+        return;
+    }
+
+    nkey = qcrypto_cipher_get_key_len(alg);
+    niv = qcrypto_cipher_get_iv_len(alg, mode);
+    if (mode == QCRYPTO_CIPHER_MODE_XTS) {
+        nkey *= 2;
+    }
+
+    key = g_new0(uint8_t, nkey);
+    memset(key, g_test_rand_int(), nkey);
+
+    iv = g_new0(uint8_t, niv);
+    memset(iv, g_test_rand_int(), niv);
+
+    ciphertext = g_new0(uint8_t, chunk_size);
+
+    plaintext = g_new0(uint8_t, chunk_size);
+    memset(plaintext, g_test_rand_int(), chunk_size);
+
+    cipher = qcrypto_cipher_new(alg, mode,
+                                key, nkey, &err);
+    g_assert(cipher != NULL);
+
+    if (mode != QCRYPTO_CIPHER_MODE_ECB)
+        g_assert(qcrypto_cipher_setiv(cipher,
+                                      iv, niv,
+                                      &err) == 0);
+
+    g_test_timer_start();
+    remain = total;
+    while (remain) {
+        g_assert(qcrypto_cipher_encrypt(cipher,
+                                        plaintext,
+                                        ciphertext,
+                                        chunk_size,
+                                        &err) == 0);
+        remain -= chunk_size;
+    }
+    g_test_timer_elapsed();
+
+    g_test_message("enc(%s-%s) chunk %zu bytes %.2f MB/sec ",
+                   QCryptoCipherAlgorithm_str(alg),
+                   QCryptoCipherMode_str(mode),
+                   chunk_size, (double)total / MiB / g_test_timer_last());
+
+    g_test_timer_start();
+    remain = total;
+    while (remain) {
+        g_assert(qcrypto_cipher_decrypt(cipher,
+                                        plaintext,
+                                        ciphertext,
+                                        chunk_size,
+                                        &err) == 0);
+        remain -= chunk_size;
+    }
+    g_test_timer_elapsed();
+
+    g_test_message("dec(%s-%s) chunk %zu bytes %.2f MB/sec ",
+                   QCryptoCipherAlgorithm_str(alg),
+                   QCryptoCipherMode_str(mode),
+                   chunk_size, (double)total / MiB / g_test_timer_last());
+
+    qcrypto_cipher_free(cipher);
+    g_free(plaintext);
+    g_free(ciphertext);
+    g_free(iv);
+    g_free(key);
+}
+
+
+static void test_cipher_speed_ecb_aes_128(const void *opaque)
+{
+    size_t chunk_size = (size_t)opaque;
+    test_cipher_speed(chunk_size,
+                      QCRYPTO_CIPHER_MODE_ECB,
+                      QCRYPTO_CIPHER_ALG_AES_128);
+}
+
+static void test_cipher_speed_ecb_aes_256(const void *opaque)
+{
+    size_t chunk_size = (size_t)opaque;
+    test_cipher_speed(chunk_size,
+                      QCRYPTO_CIPHER_MODE_ECB,
+                      QCRYPTO_CIPHER_ALG_AES_256);
+}
+
+static void test_cipher_speed_cbc_aes_128(const void *opaque)
+{
+    size_t chunk_size = (size_t)opaque;
+    test_cipher_speed(chunk_size,
+                      QCRYPTO_CIPHER_MODE_CBC,
+                      QCRYPTO_CIPHER_ALG_AES_128);
+}
+
+static void test_cipher_speed_cbc_aes_256(const void *opaque)
+{
+    size_t chunk_size = (size_t)opaque;
+    test_cipher_speed(chunk_size,
+                      QCRYPTO_CIPHER_MODE_CBC,
+                      QCRYPTO_CIPHER_ALG_AES_256);
+}
+
+static void test_cipher_speed_ctr_aes_128(const void *opaque)
+{
+    size_t chunk_size = (size_t)opaque;
+    test_cipher_speed(chunk_size,
+                      QCRYPTO_CIPHER_MODE_CTR,
+                      QCRYPTO_CIPHER_ALG_AES_128);
+}
+
+static void test_cipher_speed_ctr_aes_256(const void *opaque)
+{
+    size_t chunk_size = (size_t)opaque;
+    test_cipher_speed(chunk_size,
+                      QCRYPTO_CIPHER_MODE_CTR,
+                      QCRYPTO_CIPHER_ALG_AES_256);
+}
+
+static void test_cipher_speed_xts_aes_128(const void *opaque)
+{
+    size_t chunk_size = (size_t)opaque;
+    test_cipher_speed(chunk_size,
+                      QCRYPTO_CIPHER_MODE_XTS,
+                      QCRYPTO_CIPHER_ALG_AES_128);
+}
+
+static void test_cipher_speed_xts_aes_256(const void *opaque)
+{
+    size_t chunk_size = (size_t)opaque;
+    test_cipher_speed(chunk_size,
+                      QCRYPTO_CIPHER_MODE_XTS,
+                      QCRYPTO_CIPHER_ALG_AES_256);
+}
+
+
+int main(int argc, char **argv)
+{
+    char *alg = NULL;
+    char *size = NULL;
+    g_test_init(&argc, &argv, NULL);
+    g_assert(qcrypto_init(NULL) == 0);
+
+#define ADD_TEST(mode, cipher, keysize, chunk)                          \
+    if ((!alg || g_str_equal(alg, #mode)) &&                            \
+        (!size || g_str_equal(size, #chunk)))                           \
+        g_test_add_data_func(                                           \
+        "/crypto/cipher/" #mode "-" #cipher "-" #keysize "/chunk-" #chunk, \
+        (void *)chunk,                                                  \
+        test_cipher_speed_ ## mode ## _ ## cipher ## _ ## keysize)
+
+    if (argc >= 2) {
+        alg = argv[1];
+    }
+    if (argc >= 3) {
+        size = argv[2];
+    }
+
+#define ADD_TESTS(chunk)                        \
+    do {                                        \
+        ADD_TEST(ecb, aes, 128, chunk);         \
+        ADD_TEST(ecb, aes, 256, chunk);         \
+        ADD_TEST(cbc, aes, 128, chunk);         \
+        ADD_TEST(cbc, aes, 256, chunk);         \
+        ADD_TEST(ctr, aes, 128, chunk);         \
+        ADD_TEST(ctr, aes, 256, chunk);         \
+        ADD_TEST(xts, aes, 128, chunk);         \
+        ADD_TEST(xts, aes, 256, chunk);         \
+    } while (0)
+
+    ADD_TESTS(512);
+    ADD_TESTS(4096);
+    ADD_TESTS(16384);
+    ADD_TESTS(65536);
+
+    return g_test_run();
+}
diff --git a/tests/bench/benchmark-crypto-hash.c b/tests/bench/benchmark-crypto-hash.c
new file mode 100644
index 0000000000..927b00bb4d
--- /dev/null
+++ b/tests/bench/benchmark-crypto-hash.c
@@ -0,0 +1,116 @@
+/*
+ * QEMU Crypto hash speed benchmark
+ *
+ * Copyright (c) 2017 HUAWEI TECHNOLOGIES CO., LTD.
+ *
+ * Authors:
+ *    Longpeng(Mike) <longpeng2@huawei.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or
+ * (at your option) any later version.  See the COPYING file in the
+ * top-level directory.
+ */
+#include "qemu/osdep.h"
+#include "qemu/units.h"
+#include "crypto/init.h"
+#include "crypto/hash.h"
+
+typedef struct QCryptoHashOpts {
+    size_t chunk_size;
+    QCryptoHashAlgorithm alg;
+} QCryptoHashOpts;
+
+static void test_hash_speed(const void *opaque)
+{
+    const QCryptoHashOpts *opts = opaque;
+    uint8_t *in = NULL, *out = NULL;
+    size_t out_len = 0;
+    const size_t total = 2 * GiB;
+    size_t remain;
+    struct iovec iov;
+    int ret;
+
+    in = g_new0(uint8_t, opts->chunk_size);
+    memset(in, g_test_rand_int(), opts->chunk_size);
+
+    iov.iov_base = (char *)in;
+    iov.iov_len = opts->chunk_size;
+
+    g_test_timer_start();
+    remain = total;
+    while (remain) {
+        ret = qcrypto_hash_bytesv(opts->alg,
+                                  &iov, 1, &out, &out_len,
+                                  NULL);
+        g_assert(ret == 0);
+
+        remain -= opts->chunk_size;
+    }
+    g_test_timer_elapsed();
+
+    g_test_message("hash(%s): chunk %zu bytes %.2f MB/sec",
+                   QCryptoHashAlgorithm_str(opts->alg),
+                   opts->chunk_size, total / g_test_timer_last());
+
+    g_free(out);
+    g_free(in);
+}
+
+int main(int argc, char **argv)
+{
+    char name[64];
+
+    g_test_init(&argc, &argv, NULL);
+    g_assert(qcrypto_init(NULL) == 0);
+
+#define TEST_ONE(a, c)                                          \
+    QCryptoHashOpts opts ## a ## c = {                          \
+        .alg = QCRYPTO_HASH_ALG_ ## a, .chunk_size = c,         \
+    };                                                          \
+    memset(name, 0 , sizeof(name));                             \
+    snprintf(name, sizeof(name),                                \
+             "/crypto/benchmark/hash/%s/bufsize-%d",            \
+             QCryptoHashAlgorithm_str(QCRYPTO_HASH_ALG_ ## a),  \
+             c);                                                \
+    if (qcrypto_hash_supports(QCRYPTO_HASH_ALG_ ## a))          \
+        g_test_add_data_func(name,                              \
+                             &opts ## a ## c,                   \
+                             test_hash_speed);
+
+    TEST_ONE(MD5, 512);
+    TEST_ONE(MD5, 1024);
+    TEST_ONE(MD5, 4096);
+    TEST_ONE(MD5, 16384);
+
+    TEST_ONE(SHA1, 512);
+    TEST_ONE(SHA1, 1024);
+    TEST_ONE(SHA1, 4096);
+    TEST_ONE(SHA1, 16384);
+
+    TEST_ONE(SHA224, 512);
+    TEST_ONE(SHA224, 1024);
+    TEST_ONE(SHA224, 4096);
+    TEST_ONE(SHA224, 16384);
+
+    TEST_ONE(SHA384, 512);
+    TEST_ONE(SHA384, 1024);
+    TEST_ONE(SHA384, 4096);
+    TEST_ONE(SHA384, 16384);
+
+    TEST_ONE(SHA256, 512);
+    TEST_ONE(SHA256, 1024);
+    TEST_ONE(SHA256, 4096);
+    TEST_ONE(SHA256, 16384);
+
+    TEST_ONE(SHA512, 512);
+    TEST_ONE(SHA512, 1024);
+    TEST_ONE(SHA512, 4096);
+    TEST_ONE(SHA512, 16384);
+
+    TEST_ONE(RIPEMD160, 512);
+    TEST_ONE(RIPEMD160, 1024);
+    TEST_ONE(RIPEMD160, 4096);
+    TEST_ONE(RIPEMD160, 16384);
+
+    return g_test_run();
+}
diff --git a/tests/bench/benchmark-crypto-hmac.c b/tests/bench/benchmark-crypto-hmac.c
new file mode 100644
index 0000000000..5cca636789
--- /dev/null
+++ b/tests/bench/benchmark-crypto-hmac.c
@@ -0,0 +1,81 @@
+/*
+ * QEMU Crypto hmac speed benchmark
+ *
+ * Copyright (c) 2017 HUAWEI TECHNOLOGIES CO., LTD.
+ *
+ * Authors:
+ *    Longpeng(Mike) <longpeng2@huawei.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or
+ * (at your option) any later version.  See the COPYING file in the
+ * top-level directory.
+ */
+#include "qemu/osdep.h"
+#include "qemu/units.h"
+#include "crypto/init.h"
+#include "crypto/hmac.h"
+
+#define KEY "monkey monkey monkey monkey"
+
+static void test_hmac_speed(const void *opaque)
+{
+    size_t chunk_size = (size_t)opaque;
+    QCryptoHmac *hmac = NULL;
+    uint8_t *in = NULL, *out = NULL;
+    size_t out_len = 0;
+    double total = 0.0;
+    struct iovec iov;
+    Error *err = NULL;
+    int ret;
+
+    if (!qcrypto_hmac_supports(QCRYPTO_HASH_ALG_SHA256)) {
+        return;
+    }
+
+    in = g_new0(uint8_t, chunk_size);
+    memset(in, g_test_rand_int(), chunk_size);
+
+    iov.iov_base = (char *)in;
+    iov.iov_len = chunk_size;
+
+    g_test_timer_start();
+    do {
+        hmac = qcrypto_hmac_new(QCRYPTO_HASH_ALG_SHA256,
+                                (const uint8_t *)KEY, strlen(KEY), &err);
+        g_assert(err == NULL);
+        g_assert(hmac != NULL);
+
+        ret = qcrypto_hmac_bytesv(hmac, &iov, 1, &out, &out_len, &err);
+        g_assert(ret == 0);
+        g_assert(err == NULL);
+
+        qcrypto_hmac_free(hmac);
+
+        total += chunk_size;
+    } while (g_test_timer_elapsed() < 5.0);
+
+    total /= MiB;
+    g_test_message("hmac(%s): chunk %zu bytes %.2f MB/sec",
+                   QCryptoHashAlgorithm_str(QCRYPTO_HASH_ALG_SHA256),
+                   chunk_size, total / g_test_timer_last());
+
+    g_free(out);
+    g_free(in);
+}
+
+int main(int argc, char **argv)
+{
+    size_t i;
+    char name[64];
+
+    g_test_init(&argc, &argv, NULL);
+    g_assert(qcrypto_init(NULL) == 0);
+
+    for (i = 512; i <= 64 * KiB; i *= 2) {
+        memset(name, 0 , sizeof(name));
+        snprintf(name, sizeof(name), "/crypto/hmac/speed-%zu", i);
+        g_test_add_data_func(name, (void *)i, test_hmac_speed);
+    }
+
+    return g_test_run();
+}
diff --git a/tests/bench/meson.build b/tests/bench/meson.build
new file mode 100644
index 0000000000..00b3c209dc
--- /dev/null
+++ b/tests/bench/meson.build
@@ -0,0 +1,34 @@
+
+qht_bench = executable('qht-bench',
+                       sources: 'qht-bench.c',
+                       dependencies: [qemuutil])
+
+executable('atomic_add-bench',
+           sources: files('atomic_add-bench.c'),
+           dependencies: [qemuutil],
+           build_by_default: false)
+
+executable('atomic64-bench',
+           sources: files('atomic64-bench.c'),
+           dependencies: [qemuutil],
+           build_by_default: false)
+
+benchs = {}
+
+if have_block
+  benchs += {
+     'benchmark-crypto-hash': [crypto],
+     'benchmark-crypto-hmac': [crypto],
+     'benchmark-crypto-cipher': [crypto],
+  }
+endif
+
+foreach bench_name, deps: benchs
+  exe = executable(bench_name, bench_name + '.c',
+                   dependencies: [qemuutil] + deps)
+  benchmark(bench_name, exe,
+            args: ['--tap', '-k'],
+            protocol: 'tap',
+            timeout: 0,
+            suite: ['speed'])
+endforeach
diff --git a/tests/bench/qht-bench.c b/tests/bench/qht-bench.c
new file mode 100644
index 0000000000..2e5b70ccd0
--- /dev/null
+++ b/tests/bench/qht-bench.c
@@ -0,0 +1,523 @@
+/*
+ * Copyright (C) 2016, Emilio G. Cota <cota@braap.org>
+ *
+ * License: GNU GPL, version 2 or later.
+ *   See the COPYING file in the top-level directory.
+ */
+#include "qemu/osdep.h"
+#include "qemu/processor.h"
+#include "qemu/atomic.h"
+#include "qemu/qht.h"
+#include "qemu/rcu.h"
+#include "qemu/xxhash.h"
+
+struct thread_stats {
+    size_t rd;
+    size_t not_rd;
+    size_t in;
+    size_t not_in;
+    size_t rm;
+    size_t not_rm;
+    size_t rz;
+    size_t not_rz;
+};
+
+struct thread_info {
+    void (*func)(struct thread_info *);
+    struct thread_stats stats;
+    /*
+     * Seed is in the range [1..UINT64_MAX], because the RNG requires
+     * a non-zero seed.  To use, subtract 1 and compare against the
+     * threshold with </>=.  This lets threshold = 0 never match (0% hit),
+     * and threshold = UINT64_MAX always match (100% hit).
+     */
+    uint64_t seed;
+    bool write_op; /* writes alternate between insertions and removals */
+    bool resize_down;
+} QEMU_ALIGNED(64); /* avoid false sharing among threads */
+
+static struct qht ht;
+static QemuThread *rw_threads;
+
+#define DEFAULT_RANGE (4096)
+#define DEFAULT_QHT_N_ELEMS DEFAULT_RANGE
+
+static unsigned int duration = 1;
+static unsigned int n_rw_threads = 1;
+static unsigned long lookup_range = DEFAULT_RANGE;
+static unsigned long update_range = DEFAULT_RANGE;
+static size_t init_range = DEFAULT_RANGE;
+static size_t init_size = DEFAULT_RANGE;
+static size_t n_ready_threads;
+static long populate_offset;
+static long *keys;
+
+static size_t resize_min;
+static size_t resize_max;
+static struct thread_info *rz_info;
+static unsigned long resize_delay = 1000;
+static double resize_rate; /* 0.0 to 1.0 */
+static unsigned int n_rz_threads = 1;
+static QemuThread *rz_threads;
+static bool precompute_hash;
+
+static double update_rate; /* 0.0 to 1.0 */
+static uint64_t update_threshold;
+static uint64_t resize_threshold;
+
+static size_t qht_n_elems = DEFAULT_QHT_N_ELEMS;
+static int qht_mode;
+
+static bool test_start;
+static bool test_stop;
+
+static struct thread_info *rw_info;
+
+static const char commands_string[] =
+    " -d = duration, in seconds\n"
+    " -n = number of threads\n"
+    "\n"
+    " -o = offset at which keys start\n"
+    " -p = precompute hashes\n"
+    "\n"
+    " -g = set -s,-k,-K,-l,-r to the same value\n"
+    " -s = initial size hint\n"
+    " -k = initial number of keys\n"
+    " -K = initial range of keys (will be rounded up to pow2)\n"
+    " -l = lookup range of keys (will be rounded up to pow2)\n"
+    " -r = update range of keys (will be rounded up to pow2)\n"
+    "\n"
+    " -u = update rate (0.0 to 100.0), 50/50 split of insertions/removals\n"
+    "\n"
+    " -R = enable auto-resize\n"
+    " -S = resize rate (0.0 to 100.0)\n"
+    " -D = delay (in us) between potential resizes\n"
+    " -N = number of resize threads";
+
+static void usage_complete(int argc, char *argv[])
+{
+    fprintf(stderr, "Usage: %s [options]\n", argv[0]);
+    fprintf(stderr, "options:\n%s\n", commands_string);
+    exit(-1);
+}
+
+static bool is_equal(const void *ap, const void *bp)
+{
+    const long *a = ap;
+    const long *b = bp;
+
+    return *a == *b;
+}
+
+static uint32_t h(unsigned long v)
+{
+    return qemu_xxhash2(v);
+}
+
+static uint32_t hval(unsigned long v)
+{
+    return v;
+}
+
+static uint32_t (*hfunc)(unsigned long v) = h;
+
+/*
+ * From: https://en.wikipedia.org/wiki/Xorshift
+ * This is faster than rand_r(), and gives us a wider range (RAND_MAX is only
+ * guaranteed to be >= INT_MAX).
+ */
+static uint64_t xorshift64star(uint64_t x)
+{
+    x ^= x >> 12; /* a */
+    x ^= x << 25; /* b */
+    x ^= x >> 27; /* c */
+    return x * UINT64_C(2685821657736338717);
+}
+
+static void do_rz(struct thread_info *info)
+{
+    struct thread_stats *stats = &info->stats;
+    uint64_t r = info->seed - 1;
+
+    if (r < resize_threshold) {
+        size_t size = info->resize_down ? resize_min : resize_max;
+        bool resized;
+
+        resized = qht_resize(&ht, size);
+        info->resize_down = !info->resize_down;
+
+        if (resized) {
+            stats->rz++;
+        } else {
+            stats->not_rz++;
+        }
+    }
+    g_usleep(resize_delay);
+}
+
+static void do_rw(struct thread_info *info)
+{
+    struct thread_stats *stats = &info->stats;
+    uint64_t r = info->seed - 1;
+    uint32_t hash;
+    long *p;
+
+    if (r >= update_threshold) {
+        bool read;
+
+        p = &keys[r & (lookup_range - 1)];
+        hash = hfunc(*p);
+        read = qht_lookup(&ht, p, hash);
+        if (read) {
+            stats->rd++;
+        } else {
+            stats->not_rd++;
+        }
+    } else {
+        p = &keys[r & (update_range - 1)];
+        hash = hfunc(*p);
+        if (info->write_op) {
+            bool written = false;
+
+            if (qht_lookup(&ht, p, hash) == NULL) {
+                written = qht_insert(&ht, p, hash, NULL);
+            }
+            if (written) {
+                stats->in++;
+            } else {
+                stats->not_in++;
+            }
+        } else {
+            bool removed = false;
+
+            if (qht_lookup(&ht, p, hash)) {
+                removed = qht_remove(&ht, p, hash);
+            }
+            if (removed) {
+                stats->rm++;
+            } else {
+                stats->not_rm++;
+            }
+        }
+        info->write_op = !info->write_op;
+    }
+}
+
+static void *thread_func(void *p)
+{
+    struct thread_info *info = p;
+
+    rcu_register_thread();
+
+    qatomic_inc(&n_ready_threads);
+    while (!qatomic_read(&test_start)) {
+        cpu_relax();
+    }
+
+    rcu_read_lock();
+    while (!qatomic_read(&test_stop)) {
+        info->seed = xorshift64star(info->seed);
+        info->func(info);
+    }
+    rcu_read_unlock();
+
+    rcu_unregister_thread();
+    return NULL;
+}
+
+/* sets everything except info->func */
+static void prepare_thread_info(struct thread_info *info, int i)
+{
+    /* seed for the RNG; each thread should have a different one */
+    info->seed = (i + 1) ^ time(NULL);
+    /* the first update will be a write */
+    info->write_op = true;
+    /* the first resize will be down */
+    info->resize_down = true;
+
+    memset(&info->stats, 0, sizeof(info->stats));
+}
+
+static void
+th_create_n(QemuThread **threads, struct thread_info **infos, const char *name,
+            void (*func)(struct thread_info *), int offset, int n)
+{
+    struct thread_info *info;
+    QemuThread *th;
+    int i;
+
+    th = g_malloc(sizeof(*th) * n);
+    *threads = th;
+
+    info = qemu_memalign(64, sizeof(*info) * n);
+    *infos = info;
+
+    for (i = 0; i < n; i++) {
+        prepare_thread_info(&info[i], offset + i);
+        info[i].func = func;
+        qemu_thread_create(&th[i], name, thread_func, &info[i],
+                           QEMU_THREAD_JOINABLE);
+    }
+}
+
+static void create_threads(void)
+{
+    th_create_n(&rw_threads, &rw_info, "rw", do_rw, 0, n_rw_threads);
+    th_create_n(&rz_threads, &rz_info, "rz", do_rz, n_rw_threads, n_rz_threads);
+}
+
+static void pr_params(void)
+{
+    printf("Parameters:\n");
+    printf(" duration:          %d s\n", duration);
+    printf(" # of threads:      %u\n", n_rw_threads);
+    printf(" initial # of keys: %zu\n", init_size);
+    printf(" initial size hint: %zu\n", qht_n_elems);
+    printf(" auto-resize:       %s\n",
+           qht_mode & QHT_MODE_AUTO_RESIZE ? "on" : "off");
+    if (resize_rate) {
+        printf(" resize_rate:       %f%%\n", resize_rate * 100.0);
+        printf(" resize range:      %zu-%zu\n", resize_min, resize_max);
+        printf(" # resize threads   %u\n", n_rz_threads);
+    }
+    printf(" update rate:       %f%%\n", update_rate * 100.0);
+    printf(" offset:            %ld\n", populate_offset);
+    printf(" initial key range: %zu\n", init_range);
+    printf(" lookup range:      %lu\n", lookup_range);
+    printf(" update range:      %lu\n", update_range);
+}
+
+static void do_threshold(double rate, uint64_t *threshold)
+{
+    /*
+     * For 0 <= rate <= 1, scale to fit in a uint64_t.
+     *
+     * Scale by 2**64, with a special case for 1.0.
+     * The remainder of the possible values are scattered between 0
+     * and 0xfffffffffffff800 (nextafter(0x1p64, 0)).
+     *
+     * Note that we cannot simply scale by UINT64_MAX, because that
+     * value is not representable as an IEEE double value.
+     *
+     * If we scale by the next largest value, nextafter(0x1p64, 0),
+     * then the remainder of the possible values are scattered between
+     * 0 and 0xfffffffffffff000.  Which leaves us with a gap between
+     * the final two inputs that is twice as large as any other.
+     */
+    if (rate == 1.0) {
+        *threshold = UINT64_MAX;
+    } else {
+        *threshold = rate * 0x1p64;
+    }
+}
+
+static void htable_init(void)
+{
+    unsigned long n = MAX(init_range, update_range);
+    uint64_t r = time(NULL);
+    size_t retries = 0;
+    size_t i;
+
+    /* avoid allocating memory later by allocating all the keys now */
+    keys = g_malloc(sizeof(*keys) * n);
+    for (i = 0; i < n; i++) {
+        long val = populate_offset + i;
+
+        keys[i] = precompute_hash ? h(val) : hval(val);
+    }
+
+    /* some sanity checks */
+    g_assert_cmpuint(lookup_range, <=, n);
+
+    /* compute thresholds */
+    do_threshold(update_rate, &update_threshold);
+    do_threshold(resize_rate, &resize_threshold);
+
+    if (resize_rate) {
+        resize_min = n / 2;
+        resize_max = n;
+        assert(resize_min < resize_max);
+    } else {
+        n_rz_threads = 0;
+    }
+
+    /* initialize the hash table */
+    qht_init(&ht, is_equal, qht_n_elems, qht_mode);
+    assert(init_size <= init_range);
+
+    pr_params();
+
+    fprintf(stderr, "Initialization: populating %zu items...", init_size);
+    for (i = 0; i < init_size; i++) {
+        for (;;) {
+            uint32_t hash;
+            long *p;
+
+            r = xorshift64star(r);
+            p = &keys[r & (init_range - 1)];
+            hash = hfunc(*p);
+            if (qht_insert(&ht, p, hash, NULL)) {
+                break;
+            }
+            retries++;
+        }
+    }
+    fprintf(stderr, " populated after %zu retries\n", retries);
+}
+
+static void add_stats(struct thread_stats *s, struct thread_info *info, int n)
+{
+    int i;
+
+    for (i = 0; i < n; i++) {
+        struct thread_stats *stats = &info[i].stats;
+
+        s->rd += stats->rd;
+        s->not_rd += stats->not_rd;
+
+        s->in += stats->in;
+        s->not_in += stats->not_in;
+
+        s->rm += stats->rm;
+        s->not_rm += stats->not_rm;
+
+        s->rz += stats->rz;
+        s->not_rz += stats->not_rz;
+    }
+}
+
+static void pr_stats(void)
+{
+    struct thread_stats s = {};
+    double tx;
+
+    add_stats(&s, rw_info, n_rw_threads);
+    add_stats(&s, rz_info, n_rz_threads);
+
+    printf("Results:\n");
+
+    if (resize_rate) {
+        printf(" Resizes:           %zu (%.2f%% of %zu)\n",
+               s.rz, (double)s.rz / (s.rz + s.not_rz) * 100, s.rz + s.not_rz);
+    }
+
+    printf(" Read:              %.2f M (%.2f%% of %.2fM)\n",
+           (double)s.rd / 1e6,
+           (double)s.rd / (s.rd + s.not_rd) * 100,
+           (double)(s.rd + s.not_rd) / 1e6);
+    printf(" Inserted:          %.2f M (%.2f%% of %.2fM)\n",
+           (double)s.in / 1e6,
+           (double)s.in / (s.in + s.not_in) * 100,
+           (double)(s.in + s.not_in) / 1e6);
+    printf(" Removed:           %.2f M (%.2f%% of %.2fM)\n",
+           (double)s.rm / 1e6,
+           (double)s.rm / (s.rm + s.not_rm) * 100,
+           (double)(s.rm + s.not_rm) / 1e6);
+
+    tx = (s.rd + s.not_rd + s.in + s.not_in + s.rm + s.not_rm) / 1e6 / duration;
+    printf(" Throughput:        %.2f MT/s\n", tx);
+    printf(" Throughput/thread: %.2f MT/s/thread\n", tx / n_rw_threads);
+}
+
+static void run_test(void)
+{
+    int i;
+
+    while (qatomic_read(&n_ready_threads) != n_rw_threads + n_rz_threads) {
+        cpu_relax();
+    }
+
+    qatomic_set(&test_start, true);
+    g_usleep(duration * G_USEC_PER_SEC);
+    qatomic_set(&test_stop, true);
+
+    for (i = 0; i < n_rw_threads; i++) {
+        qemu_thread_join(&rw_threads[i]);
+    }
+    for (i = 0; i < n_rz_threads; i++) {
+        qemu_thread_join(&rz_threads[i]);
+    }
+}
+
+static void parse_args(int argc, char *argv[])
+{
+    int c;
+
+    for (;;) {
+        c = getopt(argc, argv, "d:D:g:k:K:l:hn:N:o:pr:Rs:S:u:");
+        if (c < 0) {
+            break;
+        }
+        switch (c) {
+        case 'd':
+            duration = atoi(optarg);
+            break;
+        case 'D':
+            resize_delay = atol(optarg);
+            break;
+        case 'g':
+            init_range = pow2ceil(atol(optarg));
+            lookup_range = pow2ceil(atol(optarg));
+            update_range = pow2ceil(atol(optarg));
+            qht_n_elems = atol(optarg);
+            init_size = atol(optarg);
+            break;
+        case 'h':
+            usage_complete(argc, argv);
+            exit(0);
+        case 'k':
+            init_size = atol(optarg);
+            break;
+        case 'K':
+            init_range = pow2ceil(atol(optarg));
+            break;
+        case 'l':
+            lookup_range = pow2ceil(atol(optarg));
+            break;
+        case 'n':
+            n_rw_threads = atoi(optarg);
+            break;
+        case 'N':
+            n_rz_threads = atoi(optarg);
+            break;
+        case 'o':
+            populate_offset = atol(optarg);
+            break;
+        case 'p':
+            precompute_hash = true;
+            hfunc = hval;
+            break;
+        case 'r':
+            update_range = pow2ceil(atol(optarg));
+            break;
+        case 'R':
+            qht_mode |= QHT_MODE_AUTO_RESIZE;
+            break;
+        case 's':
+            qht_n_elems = atol(optarg);
+            break;
+        case 'S':
+            resize_rate = atof(optarg) / 100.0;
+            if (resize_rate > 1.0) {
+                resize_rate = 1.0;
+            }
+            break;
+        case 'u':
+            update_rate = atof(optarg) / 100.0;
+            if (update_rate > 1.0) {
+                update_rate = 1.0;
+            }
+            break;
+        }
+    }
+}
+
+int main(int argc, char *argv[])
+{
+    parse_args(argc, argv);
+    htable_init();
+    create_threads();
+    run_test();
+    pr_stats();
+    return 0;
+}