From: Jerin Jacob <jerinjacobk@gmail.com>
To: Nithin Kumar Dabilpuram <ndabilpuram@marvell.com>
Cc: Jerin Jacob <jerinj@marvell.com>,
Kiran Kumar K <kirankumark@marvell.com>,
Sunil Kumar Kori <skori@marvell.com>,
Satha Rao <skoteshwar@marvell.com>, dpdk-dev <dev@dpdk.org>
Subject: Re: [PATCH 03/20] common/cnxk: adjust shaper rates to lower boundaries
Date: Tue, 22 Feb 2022 23:51:58 +0530 [thread overview]
Message-ID: <CALBAE1PcsvF6nx6GoV6=BBaq6M6GmVsajL3y=mFsOw+Ammskvw@mail.gmail.com> (raw)
In-Reply-To: <b9d5dacf-c9dc-a382-53f8-2f89de62d650@marvell.com>
On Tue, Feb 22, 2022 at 11:49 PM Nithin Kumar Dabilpuram
<ndabilpuram@marvell.com> wrote:
>
> Please see inline.
>
> On 2/17/22 6:50 PM, Jerin Jacob wrote:
> > On Mon, Feb 7, 2022 at 1:00 PM Nithin Dabilpuram
> > <ndabilpuram@marvell.com> wrote:
> >>
> >> From: Satha Rao <skoteshwar@marvell.com>
> >>
> >> New api to get floor values for a requested shaper rate, which can assure
> >
> > Since it is internal API, no need to mention in the comment log
> >
> >> packets should never be transmitted at a rate higher than configured.
> >>
> >> Keep the old api to get HW suggested values.
> >> And introduce new parameter to select appropriate api.
> >
> > api -> API
> >
> >>
> >> Signed-off-by: Satha Rao <skoteshwar@marvell.com>
> >> ---
> >> +static uint64_t
> >> +nix_tm_shaper_rate_conv_floor(uint64_t value, uint64_t *exponent_p,
> >> + uint64_t *mantissa_p, uint64_t *div_exp_p)
> >> +{
> >> + uint64_t div_exp, exponent, mantissa;
> >> +
> >> + /* Boundary checks */
> >> + if (value < NIX_TM_MIN_SHAPER_RATE || value > NIX_TM_MAX_SHAPER_RATE)
> >> + return 0;
> >> +
> >> + if (value <= NIX_TM_SHAPER_RATE(0, 0, 0)) {
> >> + /* Calculate rate div_exp and mantissa using
> >> + * the following formula:
> >> + *
> >> + * value = (2E6 * (256 + mantissa)
> >> + * / ((1 << div_exp) * 256))
> >> + */
> >> + div_exp = 0;
> >> + exponent = 0;
> >> + mantissa = NIX_TM_MAX_RATE_MANTISSA;
> >> +
> >> + while (value <= (NIX_TM_SHAPER_RATE_CONST / (1 << div_exp)))
> >> + div_exp += 1;
> >> +
> >> + while (value <= ((NIX_TM_SHAPER_RATE_CONST * (256 + mantissa)) /
> >> + ((1 << div_exp) * 256)))
> >> + mantissa -= 1;
> >
> > Please move this as another static function.
> This is not same as the non-floor function though it looks same. The
> while loops terminate in <= in this function. Do you still need sub
> functions for these ?
No. It is fine.
> >
> >
> >> + } else {
> >> + /* Calculate rate exponent and mantissa using
> >> + * the following formula:
> >> + *
> >> + * value = (2E6 * ((256 + mantissa) << exponent)) / 256
> >> + *
> >> + */
> >> + div_exp = 0;
> >> + exponent = NIX_TM_MAX_RATE_EXPONENT;
> >> + mantissa = NIX_TM_MAX_RATE_MANTISSA;
> >> +
> >> + while (value <= (NIX_TM_SHAPER_RATE_CONST * (1 << exponent)))
> >> + exponent -= 1;
> >> +
> >> + while (value <= ((NIX_TM_SHAPER_RATE_CONST *
> >> + ((256 + mantissa) << exponent)) /
> >> + 256))
> >> + mantissa -= 1;
> >
> > Please move this as another static function.
> Same comment as above.
> >
> >> + }
> >> +
> >> + if (div_exp > NIX_TM_MAX_RATE_DIV_EXP ||
> >> + exponent > NIX_TM_MAX_RATE_EXPONENT ||
> >> + mantissa > NIX_TM_MAX_RATE_MANTISSA)
> >> + return 0;
> >> +
> >> + if (div_exp_p)
> >> + *div_exp_p = div_exp;
> >> + if (exponent_p)
> >> + *exponent_p = exponent;
> >> + if (mantissa_p)
> >> + *mantissa_p = mantissa;
> >> +
> >> + /* Calculate real rate value */
> >> + return NIX_TM_SHAPER_RATE(exponent, mantissa, div_exp);
> >> +}
> >> +
> >> +static uint64_t
> >> +nix_tm_shaper_rate_conv_exact(uint64_t value, uint64_t *exponent_p,
> >> + uint64_t *mantissa_p, uint64_t *div_exp_p)
> >> {
> >> uint64_t div_exp, exponent, mantissa;
> >>
> >> @@ -188,6 +251,23 @@ nix_tm_shaper_rate_conv(uint64_t value, uint64_t *exponent_p,
> >> return NIX_TM_SHAPER_RATE(exponent, mantissa, div_exp);
> >> }
> >>
> >> +/* With zero accuracy we will tune parameters as defined by HW,
> >> + * non zero accuracy will keep the parameters close to lower values
> >> + * and make sure long term shaper rate will not exceed requested rate.
> >
> > long-term
> > requested -> the requested
next prev parent reply other threads:[~2022-02-22 18:22 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-07 7:29 [PATCH 01/20] common/cnxk: increase resource count for bitmap alloc Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 02/20] common/cnxk: realloc inline device XAQ AURA Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 03/20] common/cnxk: adjust shaper rates to lower boundaries Nithin Dabilpuram
2022-02-17 13:20 ` Jerin Jacob
2022-02-22 18:19 ` Nithin Kumar Dabilpuram
2022-02-22 18:21 ` Jerin Jacob [this message]
2022-02-07 7:29 ` [PATCH 04/20] common/cnxk: support inline device API without ROC NIX Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 05/20] common/cnxk: use common SA init API for default options Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 06/20] common/cnxk: enable l3hdr write back in SA Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 07/20] common/cnxk: support to enable aura tail drop for RQ Nithin Dabilpuram
2022-02-17 13:24 ` Jerin Jacob
2022-02-07 7:29 ` [PATCH 08/20] common/cnxk: use SSO time counter threshold for IRQ Nithin Dabilpuram
2022-02-17 13:25 ` Jerin Jacob
2022-02-07 7:29 ` [PATCH 09/20] common/cnxk: allow force use of SSO pffunc for outb inline Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 10/20] net/cnxk: added Rx metadata negotiate operation Nithin Dabilpuram
2022-02-17 13:33 ` Jerin Jacob
2022-02-22 18:31 ` Nithin Kumar Dabilpuram
2022-02-07 7:29 ` [PATCH 11/20] common/cnxk: removed tracking of mark actions Nithin Dabilpuram
2022-02-17 13:36 ` Jerin Jacob
2022-02-07 7:29 ` [PATCH 12/20] net/cnxk: fix inline device RQ tag mask Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 13/20] net/cnxk: register callback early to handle initial packets Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 14/20] net/cnxk: realloc inline dev XAQ for security Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 15/20] net/cnxk: use raw mbuf free on inline sec err Nithin Dabilpuram
2022-02-17 13:45 ` Jerin Jacob
2022-02-07 7:29 ` [PATCH 16/20] net/cnxk: use NPA batch burst free for meta buffers Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 17/20] net/cnxk: enable packet pool tail drop Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 18/20] net/cnxk: enable flow control by default on device configure Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 19/20] net/cnxk: add dev args for min-max spi Nithin Dabilpuram
2022-02-07 7:29 ` [PATCH 20/20] net/cnxk: add option to override outbound inline sa iv Nithin Dabilpuram
2022-02-17 13:54 ` Jerin Jacob
2022-02-17 13:11 ` [PATCH 01/20] common/cnxk: increase resource count for bitmap alloc Jerin Jacob
2022-02-17 13:13 ` Jerin Jacob
2022-02-22 19:34 ` [PATCH v2 01/21] common/cnxk: increase SMQ resource count Nithin Dabilpuram
2022-02-22 19:34 ` [PATCH v2 02/21] common/cnxk: realloc inline device XAQ AURA Nithin Dabilpuram
2022-02-22 19:34 ` [PATCH v2 03/21] common/cnxk: adjust shaper rates to lower boundaries Nithin Dabilpuram
2022-02-22 19:34 ` [PATCH v2 04/21] common/cnxk: support inline device API without ROC NIX Nithin Dabilpuram
2022-02-22 19:34 ` [PATCH v2 05/21] common/cnxk: use common SA init API for default options Nithin Dabilpuram
2022-02-22 19:34 ` [PATCH v2 06/21] common/cnxk: enable l3hdr write back in SA Nithin Dabilpuram
2022-02-22 19:34 ` [PATCH v2 07/21] common/cnxk: support to enable AURA tail drop for RQ Nithin Dabilpuram
2022-02-22 19:34 ` [PATCH v2 08/21] common/cnxk: use SSO time counter threshold for IRQ Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 09/21] common/cnxk: allow force use of SSO pffunc for outb inline Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 10/21] net/cnxk: added Rx metadata negotiate operation Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 11/21] common/cnxk: remove tracking of mark actions Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 12/21] net/cnxk: fix inline device RQ tag mask Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 13/21] net/cnxk: register callback early to handle initial packets Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 14/21] net/cnxk: realloc inline dev XAQ for security Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 15/21] net/cnxk: fix inline IPsec security error handling Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 16/21] net/cnxk: use NPA batch burst free for meta buffers Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 17/21] net/cnxk: enable packet pool tail drop Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 18/21] net/cnxk: enable flow control by default on device configure Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 19/21] net/cnxk: add dev args for min-max spi Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 20/21] net/cnxk: add option to override outbound inline SA IV Nithin Dabilpuram
2022-02-22 19:35 ` [PATCH v2 21/21] doc: add table for environment variables used by cnxk Nithin Dabilpuram
2022-02-26 9:22 ` Thomas Monjalon
2022-02-26 9:37 ` Jerin Jacob
2022-02-26 13:31 ` Thomas Monjalon
2022-02-23 16:45 ` [PATCH v2 01/21] common/cnxk: increase SMQ resource count Jerin Jacob
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='CALBAE1PcsvF6nx6GoV6=BBaq6M6GmVsajL3y=mFsOw+Ammskvw@mail.gmail.com' \
--to=jerinjacobk@gmail.com \
--cc=dev@dpdk.org \
--cc=jerinj@marvell.com \
--cc=kirankumark@marvell.com \
--cc=ndabilpuram@marvell.com \
--cc=skori@marvell.com \
--cc=skoteshwar@marvell.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).