Chengwen, as the maintainer of this driver, what do you think?
28/03/2022 16:40, Varghese, Vipin:
> [AMD Official Use Only]
>
> Verified on ` AMD EPYC 7713 64-Core Processor` platform with argument `
> --vdev=dma_skeleton0,lcore=4 --vdev=dma_skeleton1,lcore=5`.
>
> Tested-by: Vipin Varghese <vipin.vargh...@amd.com>
>
> -----Original Message-----
> From: Sivaprasad Tummala <sivaprasad.tumm...@amd.com>
> Sent: Monday, March 28, 2022 7:53 PM
> To: fengcheng...@huawei.com
> Cc: dev@dpdk.org; Tummala, Sivaprasad <sivaprasad.tumm...@amd.com>
> Subject: [PATCH] dma/skeleton: support multiple instances
>
> [CAUTION: External Email]
>
> dpdk app can support multiple hardware dma instances.
> with dma skeleton, only a single instance can be configured.
>
> This patch supports multiple driver instances per device.
>
> Signed-off-by: Sivaprasad Tummala <sivaprasad.tumm...@amd.com>
> ---
> drivers/dma/skeleton/skeleton_dmadev.c | 37 +++++++++++++++-----------
> 1 file changed, 22 insertions(+), 15 deletions(-)
>
> diff --git a/drivers/dma/skeleton/skeleton_dmadev.c
> b/drivers/dma/skeleton/skeleton_dmadev.c
> index 81cbdd286e..6b1c01d89c 100644
> --- a/drivers/dma/skeleton/skeleton_dmadev.c
> +++ b/drivers/dma/skeleton/skeleton_dmadev.c
> @@ -100,6 +100,7 @@ static int
> skeldma_start(struct rte_dma_dev *dev)
> {
> struct skeldma_hw *hw = dev->data->dev_private;
> + char name[RTE_MAX_THREAD_NAME_LEN];
> rte_cpuset_t cpuset;
> int ret;
>
> @@ -125,7 +126,8 @@ skeldma_start(struct rte_dma_dev *dev)
>
> rte_mb();
>
> - ret = rte_ctrl_thread_create(&hw->thread, "dma_skeleton", NULL,
> + snprintf(name, sizeof(name), "dma_skel_%d", dev->data->dev_id);
> + ret = rte_ctrl_thread_create(&hw->thread, name, NULL,
> cpucopy_thread, dev);
> if (ret) {
> SKELDMA_LOG(ERR, "Start cpucopy thread fail!"); @@ -160,7
> +162,7 @@ skeldma_stop(struct rte_dma_dev *dev) }
>
> static int
> -vchan_setup(struct skeldma_hw *hw, uint16_t nb_desc)
> +vchan_setup(struct skeldma_hw *hw, int16_t dev_id, uint16_t nb_desc)
> {
> struct skeldma_desc *desc;
> struct rte_ring *empty;
> @@ -168,8 +170,12 @@ vchan_setup(struct skeldma_hw *hw, uint16_t nb_desc)
> struct rte_ring *running;
> struct rte_ring *completed;
> uint16_t i;
> + char pool_name[RTE_RING_NAMESIZE];
> + char ring_name[RTE_RING_NAMESIZE];
>
> - desc = rte_zmalloc_socket("dma_skeleton_desc",
> + snprintf(pool_name, RTE_RING_NAMESIZE, "dma_skel_desc_pool_%d",
> + dev_id);
> + desc = rte_zmalloc_socket(pool_name,
> nb_desc * sizeof(struct skeldma_desc),
> RTE_CACHE_LINE_SIZE, hw->socket_id);
> if (desc == NULL) {
> @@ -177,13 +183,21 @@ vchan_setup(struct skeldma_hw *hw, uint16_t nb_desc)
> return -ENOMEM;
> }
>
> - empty = rte_ring_create("dma_skeleton_desc_empty", nb_desc,
> + snprintf(ring_name, RTE_RING_NAMESIZE, "dma_skel_desc_empty_%d",
> + dev_id);
> + empty = rte_ring_create(ring_name, nb_desc,
> hw->socket_id, RING_F_SP_ENQ | RING_F_SC_DEQ);
> - pending = rte_ring_create("dma_skeleton_desc_pending", nb_desc,
> + snprintf(ring_name, RTE_RING_NAMESIZE, "dma_skel_desc_pend_%d",
> + dev_id);
> + pending = rte_ring_create(ring_name, nb_desc,
> hw->socket_id, RING_F_SP_ENQ |
> RING_F_SC_DEQ);
> - running = rte_ring_create("dma_skeleton_desc_running", nb_desc,
> + snprintf(ring_name, RTE_RING_NAMESIZE, "dma_skel_desc_run_%d",
> + dev_id);
> + running = rte_ring_create(ring_name, nb_desc,
> hw->socket_id, RING_F_SP_ENQ |
> RING_F_SC_DEQ);
> - completed = rte_ring_create("dma_skeleton_desc_completed", nb_desc,
> + snprintf(ring_name, RTE_RING_NAMESIZE, "dma_skel_desc_comp_%d",
> + dev_id);
> + completed = rte_ring_create(ring_name, nb_desc,
> hw->socket_id, RING_F_SP_ENQ |
> RING_F_SC_DEQ);
> if (empty == NULL || pending == NULL || running == NULL ||
> completed == NULL) {
> @@ -254,7 +268,7 @@ skeldma_vchan_setup(struct rte_dma_dev *dev, uint16_t
> vchan,
> }
>
> vchan_release(hw);
> - return vchan_setup(hw, conf->nb_desc);
> + return vchan_setup(hw, dev->data->dev_id, conf->nb_desc);
> }
>
> static int
> @@ -548,13 +562,6 @@ skeldma_probe(struct rte_vdev_device *vdev)
> return -EINVAL;
> }
>
> - /* More than one instance is not supported */
> - if (skeldma_count > 0) {
> - SKELDMA_LOG(ERR, "Multiple instance not supported for %s",
> - name);
> - return -EINVAL;
> - }
> -
> skeldma_parse_vdev_args(vdev, &lcore_id);
>
> ret = skeldma_create(name, vdev, lcore_id);
> --
> 2.17.1
>
