DPDK patches and discussions
 help / color / mirror / Atom feed
From: Thomas Monjalon <thomas@monjalon.net>
To: Laurent Hardy <laurent.hardy@6wind.com>
Cc: Zijie Pan <zijie.pan@6wind.com>, dev <dev@dpdk.org>,
	"beilei.xing" <beilei.xing@intel.com>,
	"qi.z.zhang" <qi.z.zhang@intel.com>,
	"helin.zhang" <helin.zhang@intel.com>
Subject: Re: [dpdk-dev] [PATCH] net/i40e: add a specific API to control the LLDP agent
Date: Mon, 21 May 2018 12:10:07 +0200	[thread overview]
Message-ID: <15428557.l5iPiH4DZL@xps> (raw)
In-Reply-To: <6fc1b039-f79b-e82a-0e22-26f02a196697@6wind.com>

21/05/2018 10:58, Laurent Hardy:
> Hello Thomas,
> 
> On 05/21/2018 10:39 AM, Thomas Monjalon wrote:
> > 21/05/2018 05:35, Zijie Pan:
> >>>> Add a new API rte_pmd_i40e_set_lldp_cmd to control LLDP agent for i40e.
> >>>> It supports the following i40e debug lldp commands:
> >>>>      - start/stop of the LLDP agent.
> >>>>      - get local/remote of the LLDP MIB (Management Information Base).
> >>>>
> >>>> Signed-off-by: Laurent Hardy <laurent.hardy@6wind.com>
> >>>> Signed-off-by: Zijie Pan <zijie.pan@6wind.com>
> >>> Does it mean that this feature is enabled by default in i40e devices?
> >>> Should it be disabled by default?
> >> The new API is just an interface to control the LLDP agent. No functions of i40e call it by default. It is only used to debug LLDP of i40e devices. For example we can add a new command for test-pmd to debug lldp of i40e by calling rte_pmd_i40e_set_lldp_cmd().
> > There is a command to stop the LLDP agent.
> > My question is to know whether this agent is running by default
> > and should be disabled?
> This agent is running by default.
> Beside, disabling LLDP agent when using old fw (NVM 4.53 or 5.x) could 
> have some side effects.
> 
> https://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/xl710-10-40-controller-spec-update.pdf
> 
> Extracted from x710 intel spec:
> 
> "70. LLDP Disable Can Result in Incorrect Configuration of the Receive Packet Buffer
> 
> Problem:
> ---------
> The default operation of the device is LLDP enabled.
> LLDP can be disabled by an AQ command or by a custom NVM obtained from Intel.
> Disabling LLDP when using NVM 4.53 or 5.x can result in the device firmware not configuring the Receive Packet Buffer according to the link mode and flow control settings.
> Implication:
> ------------
> If flow control is disabled, a hang of the receive data path could occur wherein all received packets are dropped.
> If flow control is enabled, the incorrect settings could cause packet drops despite the flow control, making the flow control ineffective.
> Also, throughput might be sub-optimal in some cases.
> Workaround:
> -----------
> When using NVM 4.53 or 5.x, do one of the following:
> •  Do not disable LLDP or DCBx. Leave the device in default operation mode.
> •  Use the Set Local LLDP MIB admin queue command to apply the (default) DCB configuration during driver initialization and following any link-up event.
> When using NVM 6.01 or later, the receive data path does not hang even if the FW LLDP is disabled."

OK, interesting, thank you.

I think you should add a comment in the code to explain why it is not disabled.
And later, we could consider disabling it with new firmwares.


> > I mean i40e is generating LLDP packets by default?

So the answer is yes, i40e generates packets silently :)

  reply	other threads:[~2018-05-21 10:10 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-05-18 10:14 Zijie Pan
2018-05-18 10:27 ` Thomas Monjalon
2018-05-21  3:35   ` Zijie Pan
2018-05-21  8:39     ` Thomas Monjalon
2018-05-21  8:58       ` Laurent Hardy
2018-05-21 10:10         ` Thomas Monjalon [this message]
2018-05-21 13:11 ` Zhang, Qi Z
2018-05-22  9:47   ` Laurent Hardy
2018-05-22 13:00 ` Zhang, Qi Z
2018-05-24  1:32 ` [dpdk-dev] [PATCH v2] " Zijie Pan
2018-05-29  5:54   ` Zhang, Helin
2018-06-14 15:58   ` Ferruh Yigit
2018-05-24  1:45 [dpdk-dev] [PATCH] " Zijie Pan

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=15428557.l5iPiH4DZL@xps \
    --to=thomas@monjalon.net \
    --cc=beilei.xing@intel.com \
    --cc=dev@dpdk.org \
    --cc=helin.zhang@intel.com \
    --cc=laurent.hardy@6wind.com \
    --cc=qi.z.zhang@intel.com \
    --cc=zijie.pan@6wind.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).