From: "Yang, Qiming" <qiming.yang@intel.com>
To: "Wang, YuanX" <yuanx.wang@intel.com>,
"Zhang, Qi Z" <qi.z.zhang@intel.com>
Cc: "dev@dpdk.org" <dev@dpdk.org>,
"Wang, YuanX" <yuanx.wang@intel.com>,
"stable@dpdk.org" <stable@dpdk.org>
Subject: Re: [dpdk-stable] [DPDK] net/ice: fix flow rule checking
Date: Thu, 28 Jan 2021 08:13:00 +0000 [thread overview]
Message-ID: <BN6PR11MB0017AF720A5AFF7C286F63D7E5BA9@BN6PR11MB0017.namprd11.prod.outlook.com> (raw)
In-Reply-To: <20210127094236.44186-1-yuanx.wang@intel.com>
> -----Original Message-----
> From: yuanx.wang <yuanx.wang@intel.com>
> Sent: Wednesday, January 27, 2021 17:43
> To: Yang, Qiming <qiming.yang@intel.com>; Zhang, Qi Z
> <qi.z.zhang@intel.com>
> Cc: dev@dpdk.org; Wang, YuanX <yuanx.wang@intel.com>; stable@dpdk.org
> Subject: [DPDK] net/ice: fix flow rule checking
>
> Currently, when the RSS rule set to l4-src-only and l4-dst-only, the hash value is
> different while the ingress packets have same L4 parameters.
>
> This patch keeps the hash value consistent in above situation.
>
> Fixes: 38d632cbdc88 ("net/ice: refactor PF RSS")
> Fixes: 0b952714e9c1 ("net/ice: refactor PF hash flow")
> Cc: stable@dpdk.org
>
> Signed-off-by: yuanx.wang <yuanx.wang@intel.com>
> ---
> drivers/net/ice/ice_hash.c | 66 +++++++++++++++++++++-----------------
> 1 file changed, 36 insertions(+), 30 deletions(-)
>
> diff --git a/drivers/net/ice/ice_hash.c b/drivers/net/ice/ice_hash.c index
> 2d23c8dd5..8092c1acc 100644
> --- a/drivers/net/ice/ice_hash.c
> +++ b/drivers/net/ice/ice_hash.c
> @@ -712,14 +712,18 @@ ice_refine_hash_cfg_l234(struct ice_rss_hash_cfg
> *hash_cfg,
> if (rss_type &
> (ETH_RSS_NONFRAG_IPV4_UDP |
> ETH_RSS_NONFRAG_IPV6_UDP)) {
> - if (rss_type & ETH_RSS_L4_SRC_ONLY)
> - *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_UDP_DST_PORT));
> - else if (rss_type & ETH_RSS_L4_DST_ONLY)
> - *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_UDP_SRC_PORT));
> - else if (rss_type &
> - (ETH_RSS_L3_SRC_ONLY |
> - ETH_RSS_L3_DST_ONLY))
> - *hash_flds &= ~ICE_FLOW_HASH_UDP_PORT;
> + if ((rss_type &
> + (ETH_RSS_L4_SRC_ONLY |
> ETH_RSS_L4_DST_ONLY)) !=
> + (ETH_RSS_L4_SRC_ONLY | ETH_RSS_L4_DST_ONLY))
Do you want to avoid set ETH_RSS_L4_SRC_ONLY and ETH_RSS_L4_DST_ONLY at the same time?
> {
> + if (rss_type & ETH_RSS_L4_SRC_ONLY)
> + *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_UDP_DST_PORT));
> + else if (rss_type & ETH_RSS_L4_DST_ONLY)
> + *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_UDP_SRC_PORT));
> + else if (rss_type &
> + (ETH_RSS_L3_SRC_ONLY |
> + ETH_RSS_L3_DST_ONLY))
> + *hash_flds &=
> ~ICE_FLOW_HASH_UDP_PORT;
> + }
> } else {
> *hash_flds &= ~ICE_FLOW_HASH_UDP_PORT;
> }
> @@ -729,14 +733,18 @@ ice_refine_hash_cfg_l234(struct ice_rss_hash_cfg
> *hash_cfg,
> if (rss_type &
> (ETH_RSS_NONFRAG_IPV4_TCP |
> ETH_RSS_NONFRAG_IPV6_TCP)) {
> - if (rss_type & ETH_RSS_L4_SRC_ONLY)
> - *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_TCP_DST_PORT));
> - else if (rss_type & ETH_RSS_L4_DST_ONLY)
> - *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_TCP_SRC_PORT));
> - else if (rss_type &
> - (ETH_RSS_L3_SRC_ONLY |
> - ETH_RSS_L3_DST_ONLY))
> - *hash_flds &= ~ICE_FLOW_HASH_TCP_PORT;
> + if ((rss_type &
> + (ETH_RSS_L4_SRC_ONLY |
> ETH_RSS_L4_DST_ONLY)) !=
> + (ETH_RSS_L4_SRC_ONLY | ETH_RSS_L4_DST_ONLY))
> {
> + if (rss_type & ETH_RSS_L4_SRC_ONLY)
> + *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_TCP_DST_PORT));
> + else if (rss_type & ETH_RSS_L4_DST_ONLY)
> + *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_TCP_SRC_PORT));
> + else if (rss_type &
> + (ETH_RSS_L3_SRC_ONLY |
> + ETH_RSS_L3_DST_ONLY))
> + *hash_flds &=
> ~ICE_FLOW_HASH_TCP_PORT;
> + }
> } else {
> *hash_flds &= ~ICE_FLOW_HASH_TCP_PORT;
> }
> @@ -746,14 +754,18 @@ ice_refine_hash_cfg_l234(struct ice_rss_hash_cfg
> *hash_cfg,
> if (rss_type &
> (ETH_RSS_NONFRAG_IPV4_SCTP |
> ETH_RSS_NONFRAG_IPV6_SCTP)) {
> - if (rss_type & ETH_RSS_L4_SRC_ONLY)
> - *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_SCTP_DST_PORT));
> - else if (rss_type & ETH_RSS_L4_DST_ONLY)
> - *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_SCTP_SRC_PORT));
> - else if (rss_type &
> - (ETH_RSS_L3_SRC_ONLY |
> - ETH_RSS_L3_DST_ONLY))
> - *hash_flds &= ~ICE_FLOW_HASH_SCTP_PORT;
> + if ((rss_type &
> + (ETH_RSS_L4_SRC_ONLY |
> ETH_RSS_L4_DST_ONLY)) !=
> + (ETH_RSS_L4_SRC_ONLY | ETH_RSS_L4_DST_ONLY))
> {
> + if (rss_type & ETH_RSS_L4_SRC_ONLY)
> + *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_SCTP_DST_PORT));
> + else if (rss_type & ETH_RSS_L4_DST_ONLY)
> + *hash_flds &=
> ~(BIT_ULL(ICE_FLOW_FIELD_IDX_SCTP_SRC_PORT));
> + else if (rss_type &
> + (ETH_RSS_L3_SRC_ONLY |
> + ETH_RSS_L3_DST_ONLY))
> + *hash_flds &=
> ~ICE_FLOW_HASH_SCTP_PORT;
> + }
> } else {
> *hash_flds &= ~ICE_FLOW_HASH_SCTP_PORT;
> }
> @@ -946,12 +958,6 @@ ice_hash_parse_action(struct
> ice_pattern_match_item *pattern_match_item,
> RTE_FLOW_ERROR_TYPE_ACTION,
> action,
> "a non-NULL RSS queue is not
> supported");
>
> - /**
> - * Check simultaneous use of SRC_ONLY and
> DST_ONLY
> - * of the same level.
> - */
> - rss_type = rte_eth_rss_hf_refine(rss_type);
> -
> if (ice_any_invalid_rss_type(rss->func, rss_type,
> pattern_match_item-
> >input_set_mask))
> return rte_flow_error_set(error, ENOTSUP,
> --
> 2.25.1
next prev parent reply other threads:[~2021-01-28 8:13 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-27 9:42 yuanx.wang
2021-01-28 8:13 ` Yang, Qiming [this message]
-- strict thread matches above, loose matches on Subject: below --
2021-01-27 7:42 yuanx.wang
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=BN6PR11MB0017AF720A5AFF7C286F63D7E5BA9@BN6PR11MB0017.namprd11.prod.outlook.com \
--to=qiming.yang@intel.com \
--cc=dev@dpdk.org \
--cc=qi.z.zhang@intel.com \
--cc=stable@dpdk.org \
--cc=yuanx.wang@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).