DPDK patches and discussions
 help / color / mirror / Atom feed
From: Jerin Jacob <jerinjacobk@gmail.com>
To: Timothy McDaniel <timothy.mcdaniel@intel.com>
Cc: Jerin Jacob <jerin.jacob@caviumnetworks.com>, dpdk-dev <dev@dpdk.org>
Subject: Re: [PATCH 1/2] event/dlb2: update rolling mask used for dequeue
Date: Tue, 22 Feb 2022 11:27:13 +0530	[thread overview]
Message-ID: <CALBAE1NU-NXNH7-YH=FJ5tYZTh=cH1cO3msF-UsHJXt86Y26Ng@mail.gmail.com> (raw)
In-Reply-To: <20220216193827.2908858-1-timothy.mcdaniel@intel.com>

On Thu, Feb 17, 2022 at 1:09 AM Timothy McDaniel
<timothy.mcdaniel@intel.com> wrote:
>
> Update the rolling mask used in dequeue operations. Fixes
> vector optimized dequeue.
>
> Fixes: 000a7b8e7582 ("event/dlb2: optimize dequeue operation")
>
> Signed-off-by: Timothy McDaniel <timothy.mcdaniel@intel.com>


Series applied to dpdk-next-net-eventdev/for-main. Thanks


> ---
>  drivers/event/dlb2/dlb2.c | 28 +++++++++++++++++++++-------
>  1 file changed, 21 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/event/dlb2/dlb2.c b/drivers/event/dlb2/dlb2.c
> index d75f12e382..09abdd1660 100644
> --- a/drivers/event/dlb2/dlb2.c
> +++ b/drivers/event/dlb2/dlb2.c
> @@ -3897,31 +3897,45 @@ dlb2_hw_dequeue_sparse(struct dlb2_eventdev *dlb2,
>         while (num < max_num) {
>                 struct dlb2_dequeue_qe qes[DLB2_NUM_QES_PER_CACHE_LINE];
>                 int num_avail;
> +
>                 if (use_scalar) {
> +                       int n_iter = 0;
> +                       uint64_t m_rshift, m_lshift, m2_rshift, m2_lshift;
> +
>                         num_avail = dlb2_recv_qe_sparse(qm_port, qes);
>                         num_avail = RTE_MIN(num_avail, max_num - num);
>                         dlb2_inc_cq_idx(qm_port, num_avail << 2);
>                         if (num_avail == DLB2_NUM_QES_PER_CACHE_LINE)
> -                               num += dlb2_process_dequeue_four_qes(ev_port,
> -                                                                 qm_port,
> -                                                                 &events[num],
> -                                                                 &qes[0]);
> +                               n_iter = dlb2_process_dequeue_four_qes(ev_port,
> +                                                               qm_port,
> +                                                               &events[num],
> +                                                               &qes[0]);
>                         else if (num_avail)
> -                               num += dlb2_process_dequeue_qes(ev_port,
> +                               n_iter = dlb2_process_dequeue_qes(ev_port,
>                                                                 qm_port,
>                                                                 &events[num],
>                                                                 &qes[0],
>                                                                 num_avail);
> +                       num += n_iter;
> +                       /* update rolling_mask for vector code support */
> +                       m_rshift = qm_port->cq_rolling_mask >> n_iter;
> +                       m_lshift = qm_port->cq_rolling_mask << (64 - n_iter);
> +                       m2_rshift = qm_port->cq_rolling_mask_2 >> n_iter;
> +                       m2_lshift = qm_port->cq_rolling_mask_2 <<
> +                                       (64 - n_iter);
> +                       qm_port->cq_rolling_mask = (m_rshift | m2_lshift);
> +                       qm_port->cq_rolling_mask_2 = (m2_rshift | m_lshift);
>                 } else { /* !use_scalar */
>                         num_avail = dlb2_recv_qe_sparse_vec(qm_port,
>                                                             &events[num],
>                                                             max_num - num);
> -                       num += num_avail;
>                         dlb2_inc_cq_idx(qm_port, num_avail << 2);
> +                       num += num_avail;
>                         DLB2_INC_STAT(ev_port->stats.traffic.rx_ok, num_avail);
>                 }
>                 if (!num_avail) {
> -                       if (num > 0)
> +                       if ((timeout == 0) || (num > 0))
> +                               /* Not waiting in any form or 1+ events recd */
>                                 break;
>                         else if (dlb2_dequeue_wait(dlb2, ev_port, qm_port,
>                                                    timeout, start_ticks))
> --
> 2.23.0
>

      parent reply	other threads:[~2022-02-22  5:57 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-02-16 19:38 Timothy McDaniel
2022-02-16 19:38 ` [PATCH 2/2] event/dlb2: poll HW CQ inflights before mapping queue Timothy McDaniel
2022-02-22  5:57 ` Jerin Jacob [this message]

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='CALBAE1NU-NXNH7-YH=FJ5tYZTh=cH1cO3msF-UsHJXt86Y26Ng@mail.gmail.com' \
    --to=jerinjacobk@gmail.com \
    --cc=dev@dpdk.org \
    --cc=jerin.jacob@caviumnetworks.com \
    --cc=timothy.mcdaniel@intel.com \
    /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).