* [dpdk-dev] [PATCH] app/crypto-perf: use rte_mempool_put_bulk
@ 2017-06-30 7:56 Sergio Gonzalez Monroy
2017-06-30 16:13 ` Trahe, Fiona
0 siblings, 1 reply; 3+ messages in thread
From: Sergio Gonzalez Monroy @ 2017-06-30 7:56 UTC (permalink / raw)
To: dev; +Cc: pablo.de.lara.guarch
Use rte_mempool_put_bulk for both latency and throughput tests instead
of rte_crypto_op_free to improve application performance.
Signed-off-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
---
app/test-crypto-perf/cperf_test_latency.c | 14 ++++++++------
app/test-crypto-perf/cperf_test_throughput.c | 8 ++++----
2 files changed, 12 insertions(+), 10 deletions(-)
diff --git a/app/test-crypto-perf/cperf_test_latency.c b/app/test-crypto-perf/cperf_test_latency.c
index 0c85ac8..4fb7a9a 100644
--- a/app/test-crypto-perf/cperf_test_latency.c
+++ b/app/test-crypto-perf/cperf_test_latency.c
@@ -393,8 +393,10 @@ cperf_latency_test_runner(void *arg)
tsc_end = rte_rdtsc_precise();
/* Free memory for not enqueued operations */
- for (i = ops_enqd; i < burst_size; i++)
- rte_crypto_op_free(ops[i]);
+ if (ops_enqd != burst_size)
+ rte_mempool_put_bulk(ctx->crypto_op_pool,
+ (void **)&ops_processed[ops_enqd],
+ burst_size - ops_enqd);
for (i = 0; i < ops_enqd; i++) {
ctx->res[tsc_idx].tsc_start = tsc_start;
@@ -414,9 +416,9 @@ cperf_latency_test_runner(void *arg)
(ops_processed[i]->opaque_data);
pres->status = ops_processed[i]->status;
pres->tsc_end = tsc_end;
-
- rte_crypto_op_free(ops_processed[i]);
}
+ rte_mempool_put_bulk(ctx->crypto_op_pool,
+ (void **)ops_processed, ops_deqd);
deqd_tot += ops_deqd;
deqd_max = max(ops_deqd, deqd_max);
@@ -450,9 +452,9 @@ cperf_latency_test_runner(void *arg)
(ops_processed[i]->opaque_data);
pres->status = ops_processed[i]->status;
pres->tsc_end = tsc_end;
-
- rte_crypto_op_free(ops_processed[i]);
}
+ rte_mempool_put_bulk(ctx->crypto_op_pool,
+ (void **)ops_processed, ops_deqd);
deqd_tot += ops_deqd;
deqd_max = max(ops_deqd, deqd_max);
diff --git a/app/test-crypto-perf/cperf_test_throughput.c b/app/test-crypto-perf/cperf_test_throughput.c
index 053dc66..85947a5 100644
--- a/app/test-crypto-perf/cperf_test_throughput.c
+++ b/app/test-crypto-perf/cperf_test_throughput.c
@@ -396,8 +396,8 @@ cperf_throughput_test_runner(void *test_ctx)
* the crypto operation will change the data and cause
* failures.
*/
- for (i = 0; i < ops_deqd; i++)
- rte_crypto_op_free(ops_processed[i]);
+ rte_mempool_put_bulk(ctx->crypto_op_pool,
+ (void **)ops_processed, ops_deqd);
ops_deqd_total += ops_deqd;
} else {
@@ -426,8 +426,8 @@ cperf_throughput_test_runner(void *test_ctx)
if (ops_deqd == 0)
ops_deqd_failed++;
else {
- for (i = 0; i < ops_deqd; i++)
- rte_crypto_op_free(ops_processed[i]);
+ rte_mempool_put_bulk(ctx->crypto_op_pool,
+ (void **)ops_processed, ops_deqd);
ops_deqd_total += ops_deqd;
}
--
2.9.3
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [dpdk-dev] [PATCH] app/crypto-perf: use rte_mempool_put_bulk
2017-06-30 7:56 [dpdk-dev] [PATCH] app/crypto-perf: use rte_mempool_put_bulk Sergio Gonzalez Monroy
@ 2017-06-30 16:13 ` Trahe, Fiona
2017-07-02 11:06 ` De Lara Guarch, Pablo
0 siblings, 1 reply; 3+ messages in thread
From: Trahe, Fiona @ 2017-06-30 16:13 UTC (permalink / raw)
To: Gonzalez Monroy, Sergio, dev; +Cc: De Lara Guarch, Pablo
> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Sergio Gonzalez Monroy
> Sent: Friday, June 30, 2017 8:57 AM
> To: dev@dpdk.org
> Cc: De Lara Guarch, Pablo <pablo.de.lara.guarch@intel.com>
> Subject: [dpdk-dev] [PATCH] app/crypto-perf: use rte_mempool_put_bulk
>
> Use rte_mempool_put_bulk for both latency and throughput tests instead
> of rte_crypto_op_free to improve application performance.
>
> Signed-off-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
> ---
> app/test-crypto-perf/cperf_test_latency.c | 14 ++++++++------
> app/test-crypto-perf/cperf_test_throughput.c | 8 ++++----
> 2 files changed, 12 insertions(+), 10 deletions(-)
>
> diff --git a/app/test-crypto-perf/cperf_test_latency.c b/app/test-crypto-perf/cperf_test_latency.c
> index 0c85ac8..4fb7a9a 100644
> --- a/app/test-crypto-perf/cperf_test_latency.c
> +++ b/app/test-crypto-perf/cperf_test_latency.c
> @@ -393,8 +393,10 @@ cperf_latency_test_runner(void *arg)
> tsc_end = rte_rdtsc_precise();
>
> /* Free memory for not enqueued operations */
> - for (i = ops_enqd; i < burst_size; i++)
> - rte_crypto_op_free(ops[i]);
> + if (ops_enqd != burst_size)
> + rte_mempool_put_bulk(ctx->crypto_op_pool,
> + (void **)&ops_processed[ops_enqd],
> + burst_size - ops_enqd);
>
> for (i = 0; i < ops_enqd; i++) {
> ctx->res[tsc_idx].tsc_start = tsc_start;
> @@ -414,9 +416,9 @@ cperf_latency_test_runner(void *arg)
> (ops_processed[i]->opaque_data);
> pres->status = ops_processed[i]->status;
> pres->tsc_end = tsc_end;
> -
> - rte_crypto_op_free(ops_processed[i]);
> }
> + rte_mempool_put_bulk(ctx->crypto_op_pool,
> + (void **)ops_processed, ops_deqd);
>
> deqd_tot += ops_deqd;
> deqd_max = max(ops_deqd, deqd_max);
> @@ -450,9 +452,9 @@ cperf_latency_test_runner(void *arg)
> (ops_processed[i]->opaque_data);
> pres->status = ops_processed[i]->status;
> pres->tsc_end = tsc_end;
> -
> - rte_crypto_op_free(ops_processed[i]);
> }
> + rte_mempool_put_bulk(ctx->crypto_op_pool,
> + (void **)ops_processed, ops_deqd);
>
> deqd_tot += ops_deqd;
> deqd_max = max(ops_deqd, deqd_max);
> diff --git a/app/test-crypto-perf/cperf_test_throughput.c b/app/test-crypto-
> perf/cperf_test_throughput.c
> index 053dc66..85947a5 100644
> --- a/app/test-crypto-perf/cperf_test_throughput.c
> +++ b/app/test-crypto-perf/cperf_test_throughput.c
> @@ -396,8 +396,8 @@ cperf_throughput_test_runner(void *test_ctx)
> * the crypto operation will change the data and cause
> * failures.
> */
> - for (i = 0; i < ops_deqd; i++)
> - rte_crypto_op_free(ops_processed[i]);
> + rte_mempool_put_bulk(ctx->crypto_op_pool,
> + (void **)ops_processed, ops_deqd);
>
> ops_deqd_total += ops_deqd;
> } else {
> @@ -426,8 +426,8 @@ cperf_throughput_test_runner(void *test_ctx)
> if (ops_deqd == 0)
> ops_deqd_failed++;
> else {
> - for (i = 0; i < ops_deqd; i++)
> - rte_crypto_op_free(ops_processed[i]);
> + rte_mempool_put_bulk(ctx->crypto_op_pool,
> + (void **)ops_processed, ops_deqd);
>
> ops_deqd_total += ops_deqd;
> }
> --
> 2.9.3
Acked-by: Fiona Trahe <fiona.trahe@intel.com>
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [dpdk-dev] [PATCH] app/crypto-perf: use rte_mempool_put_bulk
2017-06-30 16:13 ` Trahe, Fiona
@ 2017-07-02 11:06 ` De Lara Guarch, Pablo
0 siblings, 0 replies; 3+ messages in thread
From: De Lara Guarch, Pablo @ 2017-07-02 11:06 UTC (permalink / raw)
To: Trahe, Fiona, Gonzalez Monroy, Sergio, dev
> -----Original Message-----
> From: Trahe, Fiona
> Sent: Friday, June 30, 2017 5:13 PM
> To: Gonzalez Monroy, Sergio <sergio.gonzalez.monroy@intel.com>;
> dev@dpdk.org
> Cc: De Lara Guarch, Pablo <pablo.de.lara.guarch@intel.com>
> Subject: RE: [dpdk-dev] [PATCH] app/crypto-perf: use
> rte_mempool_put_bulk
>
>
>
> > -----Original Message-----
> > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Sergio Gonzalez
> > Monroy
> > Sent: Friday, June 30, 2017 8:57 AM
> > To: dev@dpdk.org
> > Cc: De Lara Guarch, Pablo <pablo.de.lara.guarch@intel.com>
> > Subject: [dpdk-dev] [PATCH] app/crypto-perf: use rte_mempool_put_bulk
> >
> > Use rte_mempool_put_bulk for both latency and throughput tests
> instead
> > of rte_crypto_op_free to improve application performance.
> >
> > Signed-off-by: Sergio Gonzalez Monroy
> > <sergio.gonzalez.monroy@intel.com>
> > ---
> Acked-by: Fiona Trahe <fiona.trahe@intel.com>
Applied to dpdk-next-crypto.
Thanks,
Pablo
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2017-07-02 11:07 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-06-30 7:56 [dpdk-dev] [PATCH] app/crypto-perf: use rte_mempool_put_bulk Sergio Gonzalez Monroy
2017-06-30 16:13 ` Trahe, Fiona
2017-07-02 11:06 ` De Lara Guarch, Pablo
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).