crypto/virtio: allocate private data by NUMA affinity

It's better to allocate device private data on the same NUMA node with
device, rather than with the main thread. This helps avoid cross-NUMA
access for worker thread.

Signed-off-by: Xiao Wang <xiao.w.wang@intel.com>
Reviewed-by: Jay Zhou <jianjay.zhou@huawei.com>
This commit is contained in:
Xiao Wang 2019-09-04 21:21:51 +08:00 committed by Akhil Goyal
parent 407ca9a72d
commit 933f42eac9

View File

@ -1441,7 +1441,7 @@ crypto_virtio_pci_probe(
{ {
struct rte_cryptodev_pmd_init_params init_params = { struct rte_cryptodev_pmd_init_params init_params = {
.name = "", .name = "",
.socket_id = rte_socket_id(), .socket_id = pci_dev->device.numa_node,
.private_data_size = sizeof(struct virtio_crypto_hw) .private_data_size = sizeof(struct virtio_crypto_hw)
}; };
char name[RTE_CRYPTODEV_NAME_MAX_LEN]; char name[RTE_CRYPTODEV_NAME_MAX_LEN];