From: David Marchand <david.marchand@redhat.com>
To: Michal Krawczyk <mk@semihalf.com>
Cc: dev <dev@dpdk.org>, Marcin Wojtas <mw@semihalf.com>,
gtzalik@amazon.com, evgenys@amazon.com, matua@amazon.com,
igorch@amazon.com, "Burakov,
Anatoly" <anatoly.burakov@intel.com>,
dpdk stable <stable@dpdk.org>
Subject: Re: [dpdk-stable] [PATCH v2] net/ena: fix assigning NUMA node to IO queue
Date: Tue, 4 Jun 2019 12:47:46 +0200 [thread overview]
Message-ID: <CAJFAV8y35DoCprPPgqLtqg90r_8i0vMnXENXkndsYmEoZqKRJw@mail.gmail.com> (raw)
In-Reply-To: <1559644955-29088-1-git-send-email-mk@semihalf.com>
On Tue, Jun 4, 2019 at 12:43 PM Michal Krawczyk <mk@semihalf.com> wrote:
> Previous solution was using memzones in invalid way in hope to assign
> IO queue to the appropriate NUMA zone.
>
> The right way is to use socket_id from the rx/tx queue setup function
> and then pass it to the IO queue.
>
> Fixes: 3d3edc265fc8 ("net/ena: make coherent memory allocation NUMA-aware")
> Cc: stable@dpdk.org
>
> Signed-off-by: Michal Krawczyk <mk@semihalf.com>
> ---
> v2:
> * Remove gerrit change id
> * Add dpdk-stable to cc
>
> drivers/net/ena/ena_ethdev.c | 24 +++++-------------------
> drivers/net/ena/ena_ethdev.h | 2 ++
> 2 files changed, 7 insertions(+), 19 deletions(-)
>
> diff --git a/drivers/net/ena/ena_ethdev.c b/drivers/net/ena/ena_ethdev.c
> index b6651fc..e9ccb04 100644
> --- a/drivers/net/ena/ena_ethdev.c
> +++ b/drivers/net/ena/ena_ethdev.c
> @@ -272,22 +272,6 @@ static const struct eth_dev_ops ena_dev_ops = {
> .reta_query = ena_rss_reta_query,
> };
>
> -#define NUMA_NO_NODE SOCKET_ID_ANY
> -
> -static inline int ena_cpu_to_node(int cpu)
> -{
> - struct rte_config *config = rte_eal_get_configuration();
> - struct rte_fbarray *arr = &config->mem_config->memzones;
> - const struct rte_memzone *mz;
> -
> - if (unlikely(cpu >= RTE_MAX_MEMZONE))
> - return NUMA_NO_NODE;
> -
> - mz = rte_fbarray_get(arr, cpu);
> -
> - return mz->socket_id;
> -}
> -
>
I suppose you can remove the rte_eal_memconfig.h header inclusion now.
static inline void ena_rx_mbuf_prepare(struct rte_mbuf *mbuf,
> struct ena_com_rx_ctx *ena_rx_ctx)
> {
> @@ -1126,7 +1110,7 @@ static int ena_create_io_queue(struct ena_ring *ring)
> }
> ctx.qid = ena_qid;
> ctx.msix_vector = -1; /* interrupts not used */
> - ctx.numa_node = ena_cpu_to_node(ring->id);
> + ctx.numa_node = ring->numa_socket_id;
>
> rc = ena_com_create_io_queue(ena_dev, &ctx);
> if (rc) {
> @@ -1224,7 +1208,7 @@ static int ena_queue_start(struct ena_ring *ring)
> static int ena_tx_queue_setup(struct rte_eth_dev *dev,
> uint16_t queue_idx,
> uint16_t nb_desc,
> - __rte_unused unsigned int socket_id,
> + unsigned int socket_id,
> const struct rte_eth_txconf *tx_conf)
> {
> struct ena_ring *txq = NULL;
> @@ -1262,6 +1246,7 @@ static int ena_tx_queue_setup(struct rte_eth_dev
> *dev,
> txq->next_to_clean = 0;
> txq->next_to_use = 0;
> txq->ring_size = nb_desc;
> + txq->numa_socket_id = socket_id;
>
> txq->tx_buffer_info = rte_zmalloc("txq->tx_buffer_info",
> sizeof(struct ena_tx_buffer) *
> @@ -1309,7 +1294,7 @@ static int ena_tx_queue_setup(struct rte_eth_dev
> *dev,
> static int ena_rx_queue_setup(struct rte_eth_dev *dev,
> uint16_t queue_idx,
> uint16_t nb_desc,
> - __rte_unused unsigned int socket_id,
> + unsigned int socket_id,
> __rte_unused const struct rte_eth_rxconf
> *rx_conf,
> struct rte_mempool *mp)
> {
> @@ -1347,6 +1332,7 @@ static int ena_rx_queue_setup(struct rte_eth_dev
> *dev,
> rxq->next_to_clean = 0;
> rxq->next_to_use = 0;
> rxq->ring_size = nb_desc;
> + rxq->numa_socket_id = socket_id;
> rxq->mb_pool = mp;
>
> rxq->rx_buffer_info = rte_zmalloc("rxq->buffer_info",
> diff --git a/drivers/net/ena/ena_ethdev.h b/drivers/net/ena/ena_ethdev.h
> index dcc8690..9067e90 100644
> --- a/drivers/net/ena/ena_ethdev.h
> +++ b/drivers/net/ena/ena_ethdev.h
> @@ -143,6 +143,8 @@ struct ena_ring {
> struct ena_stats_rx rx_stats;
> struct ena_stats_tx tx_stats;
> };
> +
> + unsigned int numa_socket_id;
> } __rte_cache_aligned;
>
> enum ena_adapter_state {
> --
> 2.7.4
>
>
Reviewed-by: David Marchand <david.marchand@redhat.com>
--
David Marchand
next prev parent reply other threads:[~2019-06-04 10:48 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <1559644299-28575-1-git-send-email-mk@semihalf.com>
2019-06-04 10:42 ` Michal Krawczyk
2019-06-04 10:47 ` David Marchand [this message]
2019-06-04 10:59 ` [dpdk-stable] [PATCH v3] " Michal Krawczyk
2019-06-10 17:47 ` Ferruh Yigit
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CAJFAV8y35DoCprPPgqLtqg90r_8i0vMnXENXkndsYmEoZqKRJw@mail.gmail.com \
--to=david.marchand@redhat.com \
--cc=anatoly.burakov@intel.com \
--cc=dev@dpdk.org \
--cc=evgenys@amazon.com \
--cc=gtzalik@amazon.com \
--cc=igorch@amazon.com \
--cc=matua@amazon.com \
--cc=mk@semihalf.com \
--cc=mw@semihalf.com \
--cc=stable@dpdk.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).