[v9,4/6] test/ring: modify perf test cases to use rte_ring_xxx_elem APIs
diff mbox series

Message ID 20200116052511.8557-5-honnappa.nagarahalli@arm.com
State Superseded
Delegated to: David Marchand
Headers show
Series
  • lib/ring: APIs to support custom element size
Related show

Checks

Context Check Description
ci/Intel-compilation success Compilation OK
ci/checkpatch success coding style OK

Commit Message

Honnappa Nagarahalli Jan. 16, 2020, 5:25 a.m. UTC
Adjust the performance test cases to test rte_ring_xxx_elem APIs.

Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Reviewed-by: Gavin Hu <gavin.hu@arm.com>
---
 app/test/test_ring_perf.c | 454 +++++++++++++++++++++++---------------
 1 file changed, 273 insertions(+), 181 deletions(-)

Comments

Olivier Matz Jan. 17, 2020, 5:12 p.m. UTC | #1
On Wed, Jan 15, 2020 at 11:25:09PM -0600, Honnappa Nagarahalli wrote:
> Adjust the performance test cases to test rte_ring_xxx_elem APIs.
> 
> Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
> Reviewed-by: Gavin Hu <gavin.hu@arm.com>
> ---
>  app/test/test_ring_perf.c | 454 +++++++++++++++++++++++---------------
>  1 file changed, 273 insertions(+), 181 deletions(-)
> 
> diff --git a/app/test/test_ring_perf.c b/app/test/test_ring_perf.c
> index 6c2aca483..8d1217951 100644
> --- a/app/test/test_ring_perf.c
> +++ b/app/test/test_ring_perf.c

[...]

> -static int
> -test_ring_perf(void)
> +/* Run all tests for a given element size */
> +static __rte_always_inline int
> +test_ring_perf_esize(const int esize)
>  {
>  	struct lcore_pair cores;
>  	struct rte_ring *r = NULL;
>  
> -	r = rte_ring_create(RING_NAME, RING_SIZE, rte_socket_id(), 0);
> +	/*
> +	 * Performance test for legacy/_elem APIs
> +	 * SP-SC/MP-MC, single
> +	 */
> +	r = test_ring_create(RING_NAME, esize, RING_SIZE, rte_socket_id(), 0);
>  	if (r == NULL)
>  		return -1;
>  
> -	printf("### Testing single element and burst enq/deq ###\n");
> -	test_single_enqueue_dequeue(r);
> -	test_burst_enqueue_dequeue(r);
> +	printf("\n### Testing single element enq/deq ###\n");
> +	if (test_single_enqueue_dequeue(r, esize,
> +			TEST_RING_THREAD_SPSC | TEST_RING_ELEM_SINGLE) < 0)
> +		return -1;

the ring is not freed on error (same below)

> +	if (test_single_enqueue_dequeue(r, esize,
> +			TEST_RING_THREAD_MPMC | TEST_RING_ELEM_SINGLE) < 0)
> +		return -1;
> +
> +	printf("\n### Testing burst enq/deq ###\n");
> +	if (test_burst_bulk_enqueue_dequeue(r, esize,
> +			TEST_RING_THREAD_SPSC | TEST_RING_ELEM_BURST) < 0)
> +		return -1;
> +	if (test_burst_bulk_enqueue_dequeue(r, esize,
> +			TEST_RING_THREAD_MPMC | TEST_RING_ELEM_BURST) < 0)
> +		return -1;
>  
> -	printf("\n### Testing empty dequeue ###\n");
> -	test_empty_dequeue(r);
> +	printf("\n### Testing bulk enq/deq ###\n");
> +	if (test_burst_bulk_enqueue_dequeue(r, esize,
> +			TEST_RING_THREAD_SPSC | TEST_RING_ELEM_BULK) < 0)
> +		return -1;
> +	if (test_burst_bulk_enqueue_dequeue(r, esize,
> +			TEST_RING_THREAD_MPMC | TEST_RING_ELEM_BULK) < 0)
> +		return -1;
>  
> -	printf("\n### Testing using a single lcore ###\n");
> -	test_bulk_enqueue_dequeue(r);
> +	printf("\n### Testing empty bulk deq ###\n");
> +	test_empty_dequeue(r, esize,
> +			TEST_RING_THREAD_SPSC | TEST_RING_ELEM_BULK);
> +	test_empty_dequeue(r, esize,
> +			TEST_RING_THREAD_MPMC | TEST_RING_ELEM_BULK);
>  
>  	if (get_two_hyperthreads(&cores) == 0) {
>  		printf("\n### Testing using two hyperthreads ###\n");
> -		run_on_core_pair(&cores, r, enqueue_bulk, dequeue_bulk);
> +		if (run_on_core_pair(&cores, r, esize) < 0)
> +			return -1;
>  	}
>  	if (get_two_cores(&cores) == 0) {
>  		printf("\n### Testing using two physical cores ###\n");
> -		run_on_core_pair(&cores, r, enqueue_bulk, dequeue_bulk);
> +		if (run_on_core_pair(&cores, r, esize) < 0)
> +			return -1;
>  	}
>  	if (get_two_sockets(&cores) == 0) {
>  		printf("\n### Testing using two NUMA nodes ###\n");
> -		run_on_core_pair(&cores, r, enqueue_bulk, dequeue_bulk);
> +		if (run_on_core_pair(&cores, r, esize) < 0)
> +			return -1;
>  	}
>  
>  	printf("\n### Testing using all slave nodes ###\n");
> -	run_on_all_cores(r);
> +	if (run_on_all_cores(r, esize) < 0)
> +		return -1;
>  
>  	rte_ring_free(r);
> +
> +	return 0;
> +}
Honnappa Nagarahalli Jan. 18, 2020, 4:28 p.m. UTC | #2
<snip>

> 
> On Wed, Jan 15, 2020 at 11:25:09PM -0600, Honnappa Nagarahalli wrote:
> > Adjust the performance test cases to test rte_ring_xxx_elem APIs.
> >
> > Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
> > Reviewed-by: Gavin Hu <gavin.hu@arm.com>
> > ---
> >  app/test/test_ring_perf.c | 454
> > +++++++++++++++++++++++---------------
> >  1 file changed, 273 insertions(+), 181 deletions(-)
> >
> > diff --git a/app/test/test_ring_perf.c b/app/test/test_ring_perf.c
> > index 6c2aca483..8d1217951 100644
> > --- a/app/test/test_ring_perf.c
> > +++ b/app/test/test_ring_perf.c
> 
> [...]
> 
> > -static int
> > -test_ring_perf(void)
> > +/* Run all tests for a given element size */ static
> > +__rte_always_inline int test_ring_perf_esize(const int esize)
> >  {
> >  	struct lcore_pair cores;
> >  	struct rte_ring *r = NULL;
> >
> > -	r = rte_ring_create(RING_NAME, RING_SIZE, rte_socket_id(), 0);
> > +	/*
> > +	 * Performance test for legacy/_elem APIs
> > +	 * SP-SC/MP-MC, single
> > +	 */
> > +	r = test_ring_create(RING_NAME, esize, RING_SIZE, rte_socket_id(),
> > +0);
> >  	if (r == NULL)
> >  		return -1;
> >
> > -	printf("### Testing single element and burst enq/deq ###\n");
> > -	test_single_enqueue_dequeue(r);
> > -	test_burst_enqueue_dequeue(r);
> > +	printf("\n### Testing single element enq/deq ###\n");
> > +	if (test_single_enqueue_dequeue(r, esize,
> > +			TEST_RING_THREAD_SPSC |
> TEST_RING_ELEM_SINGLE) < 0)
> > +		return -1;
> 
> the ring is not freed on error (same below)
Will fix.

<snip>

Patch
diff mbox series

diff --git a/app/test/test_ring_perf.c b/app/test/test_ring_perf.c
index 6c2aca483..8d1217951 100644
--- a/app/test/test_ring_perf.c
+++ b/app/test/test_ring_perf.c
@@ -13,6 +13,7 @@ 
 #include <string.h>
 
 #include "test.h"
+#include "test_ring.h"
 
 /*
  * Ring
@@ -41,6 +42,35 @@  struct lcore_pair {
 
 static volatile unsigned lcore_count = 0;
 
+static void
+test_ring_print_test_string(unsigned int api_type, int esize,
+	unsigned int bsz, double value)
+{
+	if (esize == -1)
+		printf("legacy APIs");
+	else
+		printf("elem APIs: element size %dB", esize);
+
+	if (api_type == TEST_RING_IGNORE_API_TYPE)
+		return;
+
+	if ((api_type & TEST_RING_THREAD_DEF) == TEST_RING_THREAD_DEF)
+		printf(": default enqueue/dequeue: ");
+	else if ((api_type & TEST_RING_THREAD_SPSC) == TEST_RING_THREAD_SPSC)
+		printf(": SP/SC: ");
+	else if ((api_type & TEST_RING_THREAD_MPMC) == TEST_RING_THREAD_MPMC)
+		printf(": MP/MC: ");
+
+	if ((api_type & TEST_RING_ELEM_SINGLE) == TEST_RING_ELEM_SINGLE)
+		printf("single: ");
+	else if ((api_type & TEST_RING_ELEM_BULK) == TEST_RING_ELEM_BULK)
+		printf("bulk (size: %u): ", bsz);
+	else if ((api_type & TEST_RING_ELEM_BURST) == TEST_RING_ELEM_BURST)
+		printf("burst (size: %u): ", bsz);
+
+	printf("%.2F\n", value);
+}
+
 /**** Functions to analyse our core mask to get cores for different tests ***/
 
 static int
@@ -117,27 +147,21 @@  get_two_sockets(struct lcore_pair *lcp)
 
 /* Get cycle counts for dequeuing from an empty ring. Should be 2 or 3 cycles */
 static void
-test_empty_dequeue(struct rte_ring *r)
+test_empty_dequeue(struct rte_ring *r, const int esize,
+			const unsigned int api_type)
 {
-	const unsigned iter_shift = 26;
-	const unsigned iterations = 1<<iter_shift;
-	unsigned i = 0;
+	const unsigned int iter_shift = 26;
+	const unsigned int iterations = 1 << iter_shift;
+	unsigned int i = 0;
 	void *burst[MAX_BURST];
 
-	const uint64_t sc_start = rte_rdtsc();
+	const uint64_t start = rte_rdtsc();
 	for (i = 0; i < iterations; i++)
-		rte_ring_sc_dequeue_bulk(r, burst, bulk_sizes[0], NULL);
-	const uint64_t sc_end = rte_rdtsc();
+		test_ring_dequeue(r, burst, esize, bulk_sizes[0], api_type);
+	const uint64_t end = rte_rdtsc();
 
-	const uint64_t mc_start = rte_rdtsc();
-	for (i = 0; i < iterations; i++)
-		rte_ring_mc_dequeue_bulk(r, burst, bulk_sizes[0], NULL);
-	const uint64_t mc_end = rte_rdtsc();
-
-	printf("SC empty dequeue: %.2F\n",
-			(double)(sc_end-sc_start) / iterations);
-	printf("MC empty dequeue: %.2F\n",
-			(double)(mc_end-mc_start) / iterations);
+	test_ring_print_test_string(api_type, esize, bulk_sizes[0],
+					((double)(end - start)) / iterations);
 }
 
 /*
@@ -151,19 +175,21 @@  struct thread_params {
 };
 
 /*
- * Function that uses rdtsc to measure timing for ring enqueue. Needs pair
- * thread running dequeue_bulk function
+ * Helper function to call bulk SP/MP enqueue functions.
+ * flag == 0 -> enqueue
+ * flag == 1 -> dequeue
  */
-static int
-enqueue_bulk(void *p)
+static __rte_always_inline int
+enqueue_dequeue_bulk_helper(const unsigned int flag, const int esize,
+	struct thread_params *p)
 {
-	const unsigned iter_shift = 23;
-	const unsigned iterations = 1<<iter_shift;
-	struct thread_params *params = p;
-	struct rte_ring *r = params->r;
-	const unsigned size = params->size;
-	unsigned i;
-	void *burst[MAX_BURST] = {0};
+	int ret;
+	const unsigned int iter_shift = 23;
+	const unsigned int iterations = 1 << iter_shift;
+	struct rte_ring *r = p->r;
+	unsigned int bsize = p->size;
+	unsigned int i;
+	void *burst = NULL;
 
 #ifdef RTE_USE_C11_MEM_MODEL
 	if (__atomic_add_fetch(&lcore_count, 1, __ATOMIC_RELAXED) != 2)
@@ -173,23 +199,67 @@  enqueue_bulk(void *p)
 		while(lcore_count != 2)
 			rte_pause();
 
+	burst = test_ring_calloc(MAX_BURST, esize);
+	if (burst == NULL)
+		return -1;
+
 	const uint64_t sp_start = rte_rdtsc();
 	for (i = 0; i < iterations; i++)
-		while (rte_ring_sp_enqueue_bulk(r, burst, size, NULL) == 0)
-			rte_pause();
+		do {
+			if (flag == 0)
+				ret = test_ring_enqueue(r, burst, esize, bsize,
+						TEST_RING_THREAD_SPSC |
+						TEST_RING_ELEM_BULK);
+			else if (flag == 1)
+				ret = test_ring_dequeue(r, burst, esize, bsize,
+						TEST_RING_THREAD_SPSC |
+						TEST_RING_ELEM_BULK);
+			if (ret == 0)
+				rte_pause();
+		} while (!ret);
 	const uint64_t sp_end = rte_rdtsc();
 
 	const uint64_t mp_start = rte_rdtsc();
 	for (i = 0; i < iterations; i++)
-		while (rte_ring_mp_enqueue_bulk(r, burst, size, NULL) == 0)
-			rte_pause();
+		do {
+			if (flag == 0)
+				ret = test_ring_enqueue(r, burst, esize, bsize,
+						TEST_RING_THREAD_MPMC |
+						TEST_RING_ELEM_BULK);
+			else if (flag == 1)
+				ret = test_ring_dequeue(r, burst, esize, bsize,
+						TEST_RING_THREAD_MPMC |
+						TEST_RING_ELEM_BULK);
+			if (ret == 0)
+				rte_pause();
+		} while (!ret);
 	const uint64_t mp_end = rte_rdtsc();
 
-	params->spsc = ((double)(sp_end - sp_start))/(iterations*size);
-	params->mpmc = ((double)(mp_end - mp_start))/(iterations*size);
+	p->spsc = ((double)(sp_end - sp_start))/(iterations * bsize);
+	p->mpmc = ((double)(mp_end - mp_start))/(iterations * bsize);
 	return 0;
 }
 
+/*
+ * Function that uses rdtsc to measure timing for ring enqueue. Needs pair
+ * thread running dequeue_bulk function
+ */
+static int
+enqueue_bulk(void *p)
+{
+	struct thread_params *params = p;
+
+	return enqueue_dequeue_bulk_helper(0, -1, params);
+}
+
+static int
+enqueue_bulk_16B(void *p)
+{
+	struct thread_params *params = p;
+
+	return enqueue_dequeue_bulk_helper(0, 16, params);
+}
+
 /*
  * Function that uses rdtsc to measure timing for ring dequeue. Needs pair
  * thread running enqueue_bulk function
@@ -197,49 +267,38 @@  enqueue_bulk(void *p)
 static int
 dequeue_bulk(void *p)
 {
-	const unsigned iter_shift = 23;
-	const unsigned iterations = 1<<iter_shift;
 	struct thread_params *params = p;
-	struct rte_ring *r = params->r;
-	const unsigned size = params->size;
-	unsigned i;
-	void *burst[MAX_BURST] = {0};
-
-#ifdef RTE_USE_C11_MEM_MODEL
-	if (__atomic_add_fetch(&lcore_count, 1, __ATOMIC_RELAXED) != 2)
-#else
-	if (__sync_add_and_fetch(&lcore_count, 1) != 2)
-#endif
-		while(lcore_count != 2)
-			rte_pause();
 
-	const uint64_t sc_start = rte_rdtsc();
-	for (i = 0; i < iterations; i++)
-		while (rte_ring_sc_dequeue_bulk(r, burst, size, NULL) == 0)
-			rte_pause();
-	const uint64_t sc_end = rte_rdtsc();
+	return enqueue_dequeue_bulk_helper(1, -1, params);
+}
 
-	const uint64_t mc_start = rte_rdtsc();
-	for (i = 0; i < iterations; i++)
-		while (rte_ring_mc_dequeue_bulk(r, burst, size, NULL) == 0)
-			rte_pause();
-	const uint64_t mc_end = rte_rdtsc();
+static int
+dequeue_bulk_16B(void *p)
+{
+	struct thread_params *params = p;
 
-	params->spsc = ((double)(sc_end - sc_start))/(iterations*size);
-	params->mpmc = ((double)(mc_end - mc_start))/(iterations*size);
-	return 0;
+	return enqueue_dequeue_bulk_helper(1, 16, params);
 }
 
 /*
  * Function that calls the enqueue and dequeue bulk functions on pairs of cores.
  * used to measure ring perf between hyperthreads, cores and sockets.
  */
-static void
-run_on_core_pair(struct lcore_pair *cores, struct rte_ring *r,
-		lcore_function_t f1, lcore_function_t f2)
+static int
+run_on_core_pair(struct lcore_pair *cores, struct rte_ring *r, const int esize)
 {
+	lcore_function_t *f1, *f2;
 	struct thread_params param1 = {0}, param2 = {0};
 	unsigned i;
+
+	if (esize == -1) {
+		f1 = enqueue_bulk;
+		f2 = dequeue_bulk;
+	} else {
+		f1 = enqueue_bulk_16B;
+		f2 = dequeue_bulk_16B;
+	}
+
 	for (i = 0; i < sizeof(bulk_sizes)/sizeof(bulk_sizes[0]); i++) {
 		lcore_count = 0;
 		param1.size = param2.size = bulk_sizes[i];
@@ -251,14 +310,20 @@  run_on_core_pair(struct lcore_pair *cores, struct rte_ring *r,
 		} else {
 			rte_eal_remote_launch(f1, &param1, cores->c1);
 			rte_eal_remote_launch(f2, &param2, cores->c2);
-			rte_eal_wait_lcore(cores->c1);
-			rte_eal_wait_lcore(cores->c2);
+			if (rte_eal_wait_lcore(cores->c1) < 0)
+				return -1;
+			if (rte_eal_wait_lcore(cores->c2) < 0)
+				return -1;
 		}
-		printf("SP/SC bulk enq/dequeue (size: %u): %.2F\n", bulk_sizes[i],
-				param1.spsc + param2.spsc);
-		printf("MP/MC bulk enq/dequeue (size: %u): %.2F\n", bulk_sizes[i],
-				param1.mpmc + param2.mpmc);
+		test_ring_print_test_string(
+			TEST_RING_THREAD_SPSC | TEST_RING_ELEM_BULK,
+			esize, bulk_sizes[i], param1.spsc + param2.spsc);
+		test_ring_print_test_string(
+			TEST_RING_THREAD_MPMC | TEST_RING_ELEM_BULK,
+			esize, bulk_sizes[i], param1.mpmc + param2.mpmc);
 	}
+
+	return 0;
 }
 
 static rte_atomic32_t synchro;
@@ -267,7 +332,7 @@  static uint64_t queue_count[RTE_MAX_LCORE];
 #define TIME_MS 100
 
 static int
-load_loop_fn(void *p)
+load_loop_fn_helper(struct thread_params *p, const int esize)
 {
 	uint64_t time_diff = 0;
 	uint64_t begin = 0;
@@ -275,7 +340,11 @@  load_loop_fn(void *p)
 	uint64_t lcount = 0;
 	const unsigned int lcore = rte_lcore_id();
 	struct thread_params *params = p;
-	void *burst[MAX_BURST] = {0};
+	void *burst = NULL;
+
+	burst = test_ring_calloc(MAX_BURST, esize);
+	if (burst == NULL)
+		return -1;
 
 	/* wait synchro for slaves */
 	if (lcore != rte_get_master_lcore())
@@ -284,22 +353,49 @@  load_loop_fn(void *p)
 
 	begin = rte_get_timer_cycles();
 	while (time_diff < hz * TIME_MS / 1000) {
-		rte_ring_mp_enqueue_bulk(params->r, burst, params->size, NULL);
-		rte_ring_mc_dequeue_bulk(params->r, burst, params->size, NULL);
+		test_ring_enqueue(params->r, burst, esize, params->size,
+				TEST_RING_THREAD_MPMC | TEST_RING_ELEM_BULK);
+		test_ring_dequeue(params->r, burst, esize, params->size,
+				TEST_RING_THREAD_MPMC | TEST_RING_ELEM_BULK);
 		lcount++;
 		time_diff = rte_get_timer_cycles() - begin;
 	}
 	queue_count[lcore] = lcount;
+
+	rte_free(burst);
+
 	return 0;
 }
 
 static int
-run_on_all_cores(struct rte_ring *r)
+load_loop_fn(void *p)
+{
+	struct thread_params *params = p;
+
+	return load_loop_fn_helper(params, -1);
+}
+
+static int
+load_loop_fn_16B(void *p)
+{
+	struct thread_params *params = p;
+
+	return load_loop_fn_helper(params, 16);
+}
+
+static int
+run_on_all_cores(struct rte_ring *r, const int esize)
 {
 	uint64_t total = 0;
 	struct thread_params param;
+	lcore_function_t *lcore_f;
 	unsigned int i, c;
 
+	if (esize == -1)
+		lcore_f = load_loop_fn;
+	else
+		lcore_f = load_loop_fn_16B;
+
 	memset(&param, 0, sizeof(struct thread_params));
 	for (i = 0; i < RTE_DIM(bulk_sizes); i++) {
 		printf("\nBulk enq/dequeue count on size %u\n", bulk_sizes[i]);
@@ -308,13 +404,12 @@  run_on_all_cores(struct rte_ring *r)
 
 		/* clear synchro and start slaves */
 		rte_atomic32_set(&synchro, 0);
-		if (rte_eal_mp_remote_launch(load_loop_fn, &param,
-			SKIP_MASTER) < 0)
+		if (rte_eal_mp_remote_launch(lcore_f, &param, SKIP_MASTER) < 0)
 			return -1;
 
 		/* start synchro and launch test on master */
 		rte_atomic32_set(&synchro, 1);
-		load_loop_fn(&param);
+		lcore_f(&param);
 
 		rte_eal_mp_wait_lcore();
 
@@ -335,155 +430,152 @@  run_on_all_cores(struct rte_ring *r)
  * Test function that determines how long an enqueue + dequeue of a single item
  * takes on a single lcore. Result is for comparison with the bulk enq+deq.
  */
-static void
-test_single_enqueue_dequeue(struct rte_ring *r)
+static int
+test_single_enqueue_dequeue(struct rte_ring *r, const int esize,
+	const unsigned int api_type)
 {
-	const unsigned iter_shift = 24;
-	const unsigned iterations = 1<<iter_shift;
-	unsigned i = 0;
+	const unsigned int iter_shift = 24;
+	const unsigned int iterations = 1 << iter_shift;
+	unsigned int i = 0;
 	void *burst = NULL;
 
-	const uint64_t sc_start = rte_rdtsc();
-	for (i = 0; i < iterations; i++) {
-		rte_ring_sp_enqueue(r, burst);
-		rte_ring_sc_dequeue(r, &burst);
-	}
-	const uint64_t sc_end = rte_rdtsc();
+	/* alloc dummy object pointers */
+	burst = test_ring_calloc(1, esize);
+	if (burst == NULL)
+		return -1;
 
-	const uint64_t mc_start = rte_rdtsc();
+	const uint64_t start = rte_rdtsc();
 	for (i = 0; i < iterations; i++) {
-		rte_ring_mp_enqueue(r, burst);
-		rte_ring_mc_dequeue(r, &burst);
+		test_ring_enqueue(r, burst, esize, 1, api_type);
+		test_ring_dequeue(r, burst, esize, 1, api_type);
 	}
-	const uint64_t mc_end = rte_rdtsc();
+	const uint64_t end = rte_rdtsc();
+
+	test_ring_print_test_string(api_type, esize, 1,
+					((double)(end - start)) / iterations);
 
-	printf("SP/SC single enq/dequeue: %.2F\n",
-			((double)(sc_end-sc_start)) / iterations);
-	printf("MP/MC single enq/dequeue: %.2F\n",
-			((double)(mc_end-mc_start)) / iterations);
+	rte_free(burst);
+
+	return 0;
 }
 
 /*
- * Test that does both enqueue and dequeue on a core using the burst() API calls
- * instead of the bulk() calls used in other tests. Results should be the same
- * as for the bulk function called on a single lcore.
+ * Test that does both enqueue and dequeue on a core using the burst/bulk API
+ * calls Results should be the same as for the bulk function called on a
+ * single lcore.
  */
-static void
-test_burst_enqueue_dequeue(struct rte_ring *r)
+static int
+test_burst_bulk_enqueue_dequeue(struct rte_ring *r, const int esize,
+	const unsigned int api_type)
 {
-	const unsigned iter_shift = 23;
-	const unsigned iterations = 1<<iter_shift;
-	unsigned sz, i = 0;
-	void *burst[MAX_BURST] = {0};
+	const unsigned int iter_shift = 23;
+	const unsigned int iterations = 1 << iter_shift;
+	unsigned int sz, i = 0;
+	void **burst = NULL;
 
-	for (sz = 0; sz < sizeof(bulk_sizes)/sizeof(bulk_sizes[0]); sz++) {
-		const uint64_t sc_start = rte_rdtsc();
-		for (i = 0; i < iterations; i++) {
-			rte_ring_sp_enqueue_burst(r, burst,
-					bulk_sizes[sz], NULL);
-			rte_ring_sc_dequeue_burst(r, burst,
-					bulk_sizes[sz], NULL);
-		}
-		const uint64_t sc_end = rte_rdtsc();
+	burst = test_ring_calloc(MAX_BURST, esize);
+	if (burst == NULL)
+		return -1;
 
-		const uint64_t mc_start = rte_rdtsc();
+	for (sz = 0; sz < RTE_DIM(bulk_sizes); sz++) {
+		const uint64_t start = rte_rdtsc();
 		for (i = 0; i < iterations; i++) {
-			rte_ring_mp_enqueue_burst(r, burst,
-					bulk_sizes[sz], NULL);
-			rte_ring_mc_dequeue_burst(r, burst,
-					bulk_sizes[sz], NULL);
+			test_ring_enqueue(r, burst, esize, bulk_sizes[sz],
+						api_type);
+			test_ring_dequeue(r, burst, esize, bulk_sizes[sz],
+						api_type);
 		}
-		const uint64_t mc_end = rte_rdtsc();
-
-		double mc_avg = ((double)(mc_end-mc_start) / iterations) /
-					bulk_sizes[sz];
-		double sc_avg = ((double)(sc_end-sc_start) / iterations) /
-					bulk_sizes[sz];
+		const uint64_t end = rte_rdtsc();
 
-		printf("SP/SC burst enq/dequeue (size: %u): %.2F\n",
-				bulk_sizes[sz], sc_avg);
-		printf("MP/MC burst enq/dequeue (size: %u): %.2F\n",
-				bulk_sizes[sz], mc_avg);
+		test_ring_print_test_string(api_type, esize, bulk_sizes[sz],
+					((double)(end - start)) / iterations);
 	}
-}
 
-/* Times enqueue and dequeue on a single lcore */
-static void
-test_bulk_enqueue_dequeue(struct rte_ring *r)
-{
-	const unsigned iter_shift = 23;
-	const unsigned iterations = 1<<iter_shift;
-	unsigned sz, i = 0;
-	void *burst[MAX_BURST] = {0};
+	rte_free(burst);
 
-	for (sz = 0; sz < sizeof(bulk_sizes)/sizeof(bulk_sizes[0]); sz++) {
-		const uint64_t sc_start = rte_rdtsc();
-		for (i = 0; i < iterations; i++) {
-			rte_ring_sp_enqueue_bulk(r, burst,
-					bulk_sizes[sz], NULL);
-			rte_ring_sc_dequeue_bulk(r, burst,
-					bulk_sizes[sz], NULL);
-		}
-		const uint64_t sc_end = rte_rdtsc();
-
-		const uint64_t mc_start = rte_rdtsc();
-		for (i = 0; i < iterations; i++) {
-			rte_ring_mp_enqueue_bulk(r, burst,
-					bulk_sizes[sz], NULL);
-			rte_ring_mc_dequeue_bulk(r, burst,
-					bulk_sizes[sz], NULL);
-		}
-		const uint64_t mc_end = rte_rdtsc();
-
-		double sc_avg = ((double)(sc_end-sc_start) /
-				(iterations * bulk_sizes[sz]));
-		double mc_avg = ((double)(mc_end-mc_start) /
-				(iterations * bulk_sizes[sz]));
-
-		printf("SP/SC bulk enq/dequeue (size: %u): %.2F\n", bulk_sizes[sz],
-				sc_avg);
-		printf("MP/MC bulk enq/dequeue (size: %u): %.2F\n", bulk_sizes[sz],
-				mc_avg);
-	}
+	return 0;
 }
 
-static int
-test_ring_perf(void)
+/* Run all tests for a given element size */
+static __rte_always_inline int
+test_ring_perf_esize(const int esize)
 {
 	struct lcore_pair cores;
 	struct rte_ring *r = NULL;
 
-	r = rte_ring_create(RING_NAME, RING_SIZE, rte_socket_id(), 0);
+	/*
+	 * Performance test for legacy/_elem APIs
+	 * SP-SC/MP-MC, single
+	 */
+	r = test_ring_create(RING_NAME, esize, RING_SIZE, rte_socket_id(), 0);
 	if (r == NULL)
 		return -1;
 
-	printf("### Testing single element and burst enq/deq ###\n");
-	test_single_enqueue_dequeue(r);
-	test_burst_enqueue_dequeue(r);
+	printf("\n### Testing single element enq/deq ###\n");
+	if (test_single_enqueue_dequeue(r, esize,
+			TEST_RING_THREAD_SPSC | TEST_RING_ELEM_SINGLE) < 0)
+		return -1;
+	if (test_single_enqueue_dequeue(r, esize,
+			TEST_RING_THREAD_MPMC | TEST_RING_ELEM_SINGLE) < 0)
+		return -1;
+
+	printf("\n### Testing burst enq/deq ###\n");
+	if (test_burst_bulk_enqueue_dequeue(r, esize,
+			TEST_RING_THREAD_SPSC | TEST_RING_ELEM_BURST) < 0)
+		return -1;
+	if (test_burst_bulk_enqueue_dequeue(r, esize,
+			TEST_RING_THREAD_MPMC | TEST_RING_ELEM_BURST) < 0)
+		return -1;
 
-	printf("\n### Testing empty dequeue ###\n");
-	test_empty_dequeue(r);
+	printf("\n### Testing bulk enq/deq ###\n");
+	if (test_burst_bulk_enqueue_dequeue(r, esize,
+			TEST_RING_THREAD_SPSC | TEST_RING_ELEM_BULK) < 0)
+		return -1;
+	if (test_burst_bulk_enqueue_dequeue(r, esize,
+			TEST_RING_THREAD_MPMC | TEST_RING_ELEM_BULK) < 0)
+		return -1;
 
-	printf("\n### Testing using a single lcore ###\n");
-	test_bulk_enqueue_dequeue(r);
+	printf("\n### Testing empty bulk deq ###\n");
+	test_empty_dequeue(r, esize,
+			TEST_RING_THREAD_SPSC | TEST_RING_ELEM_BULK);
+	test_empty_dequeue(r, esize,
+			TEST_RING_THREAD_MPMC | TEST_RING_ELEM_BULK);
 
 	if (get_two_hyperthreads(&cores) == 0) {
 		printf("\n### Testing using two hyperthreads ###\n");
-		run_on_core_pair(&cores, r, enqueue_bulk, dequeue_bulk);
+		if (run_on_core_pair(&cores, r, esize) < 0)
+			return -1;
 	}
 	if (get_two_cores(&cores) == 0) {
 		printf("\n### Testing using two physical cores ###\n");
-		run_on_core_pair(&cores, r, enqueue_bulk, dequeue_bulk);
+		if (run_on_core_pair(&cores, r, esize) < 0)
+			return -1;
 	}
 	if (get_two_sockets(&cores) == 0) {
 		printf("\n### Testing using two NUMA nodes ###\n");
-		run_on_core_pair(&cores, r, enqueue_bulk, dequeue_bulk);
+		if (run_on_core_pair(&cores, r, esize) < 0)
+			return -1;
 	}
 
 	printf("\n### Testing using all slave nodes ###\n");
-	run_on_all_cores(r);
+	if (run_on_all_cores(r, esize) < 0)
+		return -1;
 
 	rte_ring_free(r);
+
+	return 0;
+}
+
+static int
+test_ring_perf(void)
+{
+	/* Run all the tests for different element sizes */
+	if (test_ring_perf_esize(-1) == -1)
+		return -1;
+
+	if (test_ring_perf_esize(16) == -1)
+		return -1;
+
 	return 0;
 }