diff mbox series

[v3,16/18] efd: add checks for max SIMD bitwidth

Message ID 20200930130415.11211-17-ciara.power@intel.com (mailing list archive)
State Superseded, archived
Delegated to: David Marchand
Headers
Series add max SIMD bitwidth to EAL |

Checks

Context Check Description
ci/checkpatch success coding style OK

Commit Message

Power, Ciara Sept. 30, 2020, 1:04 p.m. UTC 
  When choosing a vector path to take, an extra condition must be
satisfied to ensure the max SIMD bitwidth allows for the CPU enabled
path.

Cc: Byron Marohn <byron.marohn@intel.com>
Cc: Yipeng Wang <yipeng1.wang@intel.com>

Signed-off-by: Ciara Power <ciara.power@intel.com>
---
 lib/librte_efd/rte_efd.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

Comments

Wang, Yipeng1 Oct. 7, 2020, 12:51 a.m. UTC | #1 
> -----Original Message-----
> From: Power, Ciara <ciara.power@intel.com>
> Sent: Wednesday, September 30, 2020 6:04 AM
> To: dev@dpdk.org
> Cc: Power, Ciara <ciara.power@intel.com>; Marohn, Byron
> <byron.marohn@intel.com>; Wang, Yipeng1 <yipeng1.wang@intel.com>
> Subject: [PATCH v3 16/18] efd: add checks for max SIMD bitwidth
> 
> When choosing a vector path to take, an extra condition must be satisfied to
> ensure the max SIMD bitwidth allows for the CPU enabled path.
> 
> Cc: Byron Marohn <byron.marohn@intel.com>
> Cc: Yipeng Wang <yipeng1.wang@intel.com>
> 
> Signed-off-by: Ciara Power <ciara.power@intel.com>
> ---
>  lib/librte_efd/rte_efd.c | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/lib/librte_efd/rte_efd.c b/lib/librte_efd/rte_efd.c index
> 6a799556d4..509ecc8256 100644
> --- a/lib/librte_efd/rte_efd.c
> +++ b/lib/librte_efd/rte_efd.c
> @@ -645,7 +645,9 @@ rte_efd_create(const char *name, uint32_t
> max_num_rules, uint32_t key_len,
>  	 * For less than 4 bits, scalar function performs better
>  	 * than vectorised version
>  	 */
> -	if (RTE_EFD_VALUE_NUM_BITS > 3 &&
> rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX2))
> +	if (RTE_EFD_VALUE_NUM_BITS > 3
> +			&& rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX2)
> +			&& rte_get_max_simd_bitwidth() >=
> RTE_MAX_256_SIMD)
>  		table->lookup_fn = EFD_LOOKUP_AVX2;
>  	else
>  #endif
> @@ -655,7 +657,8 @@ rte_efd_create(const char *name, uint32_t
> max_num_rules, uint32_t key_len,
>  	 * than vectorised version
>  	 */
>  	if (RTE_EFD_VALUE_NUM_BITS > 16 &&
> -	    rte_cpu_get_flag_enabled(RTE_CPUFLAG_NEON))
> +	    rte_cpu_get_flag_enabled(RTE_CPUFLAG_NEON) &&
> +			rte_get_max_simd_bitwidth() >=
> RTE_MAX_128_SIMD)
>  		table->lookup_fn = EFD_LOOKUP_NEON;
>  	else
>  #endif
> --
> 2.17.1
[Wang, Yipeng] 
Acked-by: Yipeng Wang <yipeng1.wang@intel.com>
diff mbox series

Patch

diff --git a/lib/librte_efd/rte_efd.c b/lib/librte_efd/rte_efd.c
index 6a799556d4..509ecc8256 100644
--- a/lib/librte_efd/rte_efd.c
+++ b/lib/librte_efd/rte_efd.c
@@ -645,7 +645,9 @@  rte_efd_create(const char *name, uint32_t max_num_rules, uint32_t key_len,
 	 * For less than 4 bits, scalar function performs better
 	 * than vectorised version
 	 */
-	if (RTE_EFD_VALUE_NUM_BITS > 3 && rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX2))
+	if (RTE_EFD_VALUE_NUM_BITS > 3
+			&& rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX2)
+			&& rte_get_max_simd_bitwidth() >= RTE_MAX_256_SIMD)
 		table->lookup_fn = EFD_LOOKUP_AVX2;
 	else
 #endif
@@ -655,7 +657,8 @@  rte_efd_create(const char *name, uint32_t max_num_rules, uint32_t key_len,
 	 * than vectorised version
 	 */
 	if (RTE_EFD_VALUE_NUM_BITS > 16 &&
-	    rte_cpu_get_flag_enabled(RTE_CPUFLAG_NEON))
+	    rte_cpu_get_flag_enabled(RTE_CPUFLAG_NEON) &&
+			rte_get_max_simd_bitwidth() >= RTE_MAX_128_SIMD)
 		table->lookup_fn = EFD_LOOKUP_NEON;
 	else
 #endif