ell.lists.linux.dev archive mirror
 help / color / mirror / Atom feed
From: Denis Kenzior <denkenz@gmail.com>
To: Andrew Zaborowski <andrew.zaborowski@intel.com>, ell@lists.linux.dev
Subject: Re: [PATCH 7/7] netconfig: Control optimistic DAD
Date: Mon, 19 Sep 2022 13:43:06 -0500	[thread overview]
Message-ID: <393c900e-f48c-d397-8388-51e5ca25050b@gmail.com> (raw)
In-Reply-To: <20220919133105.3129080-7-andrew.zaborowski@intel.com>

Hi Andrew,

On 9/19/22 08:31, Andrew Zaborowski wrote:
> Enable or disable optimistic DAD for the interface using /proc.  Add
> l_netconfig_set_optimistic_dad_enabled() for the user to request that
> RFC 4429 optimistic DAD be enabled.  As recommended in the RFC, we'll
> only actually enable optimistic DAD for the automatic address generation
> methods and disable it if a static IPv6 address is configured.  The
> default if l_netconfig_set_optimistic_dad_enabled() isn't used is to
> disable optimistic DAD always.
> 
> Optimistic DAD can shorten practical IPv6 setup time by an amount on
> the order of a second by allowing the link-local address and the global
> address to be used immediately after being generated.  For the
> link-local address this means that DHCPv6 may start sooner and
> for the global address, which is added to the kernel by the
> L_NETCONFIG_EVENT_CONFIGURE event handler, the user can start
> establishing connections sooner.  With DHCPv6 and some luck the two
> savings may compound, with SLAAC only the global address's DAD time
> should matter.
> ---
>   ell/ell.sym     |  1 +
>   ell/netconfig.c | 49 ++++++++++++++++++++++++++++++++++++++++++++++++-
>   ell/netconfig.h |  2 ++
>   3 files changed, 51 insertions(+), 1 deletion(-)
> 

I went ahead and applied all patches in this series.  Question:

> @@ -1724,7 +1737,8 @@ static void netconfig_ifaddr_ipv6_added(struct l_netconfig *nc,
>   	struct in6_addr in6;
>   	_auto_(l_free) char *ip = NULL;
>   
> -	if (ifa->ifa_flags & IFA_F_TENTATIVE)
> +	if ((ifa->ifa_flags & IFA_F_TENTATIVE) &&
> +			!(ifa->ifa_flags & IFA_F_OPTIMISTIC))
>   		return;

This implies that we set the link local address even if it is still OPTIMISTIC. 
  Do we now need to update icmp6_send_router_solicitation() logic to make sure 
we do not include SLLAO from optimistic addresses?

>   
>   	if (!nc->started)

Regards,
-Denis

  reply	other threads:[~2022-09-19 18:59 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-09-19 13:30 [PATCH 1/7] icmp6: Save SLAAC prefixes from RAs Andrew Zaborowski
2022-09-19 13:31 ` [PATCH 2/7] icmp6: Parse RDNSS and DNSSL options Andrew Zaborowski
2022-09-19 13:31 ` [PATCH 3/7] icmp6: Switch socket from AF_INET6 to AF_PACKET Andrew Zaborowski
2022-09-19 13:31 ` [PATCH 4/7] rtnl: Add l_rtnl_address_get_in_addr Andrew Zaborowski
2022-09-19 13:31 ` [PATCH 5/7] netconfig: Decouple icmp6 start from dhcp6 start Andrew Zaborowski
2022-09-19 13:31 ` [PATCH 6/7] netconfig: Create SLAAC address Andrew Zaborowski
2022-09-19 13:31 ` [PATCH 7/7] netconfig: Control optimistic DAD Andrew Zaborowski
2022-09-19 18:43   ` Denis Kenzior [this message]
2022-09-19 23:56     ` Andrew Zaborowski

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=393c900e-f48c-d397-8388-51e5ca25050b@gmail.com \
    --to=denkenz@gmail.com \
    --cc=andrew.zaborowski@intel.com \
    --cc=ell@lists.linux.dev \
    /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).