public inbox for netdev@vger.kernel.org
 help / color / mirror / Atom feed
From: Alexander Duyck <alexander.h.duyck@redhat.com>
To: netdev@vger.kernel.org
Subject: [RFC PATCH 00/29] Phase 2 of fib_trie updates
Date: Tue, 24 Feb 2015 12:47:55 -0800	[thread overview]
Message-ID: <20150224202837.26106.87623.stgit@ahduyck-vm-fedora20> (raw)

This patch series implements the second phase of the fib_trie changes.  I
presented on these and the previous changes at Netdev01 and netconf.  The
slides for the Netdev01 presentation can be found at
https://www.netdev01.org/docs/duyck-fib-trie.pdf.

I'm currently debating if I should just submit the entire patch-set as-is
or if I should hold off on submitting the last 10 patches as they currently
have a potential performance impact in the case of a large number of
entries placed in the local table.  Specifically I have seen that removing
an interface in the case of 8K local subnets being configured on it
resulted in the time for a dummy interface being removed increasing from
about .6 seconds to 2.4 seconds.  I am not sure how common of a use-case
something like this would be.  I have not seen the same issue if I assign
8K routes to the interface as I believe the fib_table_flush aggregates them
all in to one resize action.

The entire series reduces the total look-up time by another 20-35% versus
what is currently in the 4.0-rc1 kernel.  So for example a set of routing
look-ups which took 140ns in the 4.0-rc1 kernel will now only take about
105ns after these patches.

---

Alexander Duyck (29):
      fib_trie: Convert fib_alias to hlist from list
      fib_trie: Replace plen with slen in leaf_info
      fib_trie: Add slen to fib alias
      fib_trie: Remove leaf_info
      fib_trie: Only resize N/2 times instead N * log(N) times in fib_table_flush
      fib_trie: Fib walk rcu should take a tnode and key instead of a trie and a leaf
      fib_trie: Fib find node should return parent
      fib_trie: Update insert and delete to make use of tp from find_node
      fib_trie: Make fib_table rcu safe
      fib_trie: Return pointer to tnode pointer in resize/inflate/halve
      fib_trie: Rename tnode to key_vector
      fib_trie: move leaf and tnode to occupy the same spot in the key vector
      fib_trie: replace tnode_get_child functions with get_child macros
      fib_trie: Rename tnode_child_length to child_length
      fib_trie: Add tnode struct as a container for fields not needed in key_vector
      fib_trie: Move rcu from key_vector to tnode, add accessors.
      fib_trie: Pull empty_children and full_children into tnode
      fib_trie: Move parent from key_vector to tnode
      fib_trie: Add key vector to root, return parent key_vector in resize
      fib_trie: Push net pointer down into fib_trie insert/delete/flush calls
      fib_trie: Rewrite handling of RCU to include parent in replacement
      fib_trie: Allocate tnode as array of key_vectors instead of key_vector as array of tnode pointers
      fib_trie: Add leaf_init
      fib_trie: Update tnode_new to drop use of put_child_root
      fib_trie: Add function for dropping children from trie
      fib_trie: Use put_child to only copy key_vectors instead of pointers
      fib_trie: Move key and pos into key_vector from tnode
      fib_trie: Move slen from tnode to key vector
      fib_trie: Push bits up one level, and move leaves up into parent key_vector array


 include/net/ip_fib.h     |   80 +-
 include/net/netns/ipv4.h |    7 
 net/ipv4/fib_frontend.c  |   89 ++
 net/ipv4/fib_lookup.h    |    3 
 net/ipv4/fib_semantics.c |    4 
 net/ipv4/fib_trie.c      | 1974 ++++++++++++++++++++++++----------------------
 6 files changed, 1152 insertions(+), 1005 deletions(-)

--

             reply	other threads:[~2015-02-24 20:47 UTC|newest]

Thread overview: 39+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-02-24 20:47 Alexander Duyck [this message]
2015-02-24 20:48 ` [RFC PATCH 01/29] fib_trie: Convert fib_alias to hlist from list Alexander Duyck
2015-02-24 21:51   ` Or Gerlitz
2015-02-24 21:52     ` Or Gerlitz
2015-02-24 22:08     ` David Miller
2015-02-24 22:14       ` Alexander Duyck
2015-02-24 22:47   ` Julian Anastasov
2015-02-24 23:09   ` Julian Anastasov
2015-02-24 20:48 ` [RFC PATCH 02/29] fib_trie: Replace plen with slen in leaf_info Alexander Duyck
2015-02-24 20:48 ` [RFC PATCH 03/29] fib_trie: Add slen to fib alias Alexander Duyck
2015-02-24 20:48 ` [RFC PATCH 04/29] fib_trie: Remove leaf_info Alexander Duyck
2015-02-24 20:48 ` [RFC PATCH 05/29] fib_trie: Only resize N/2 times instead N * log(N) times in fib_table_flush Alexander Duyck
2015-02-24 20:48 ` [RFC PATCH 06/29] fib_trie: Fib walk rcu should take a tnode and key instead of a trie and a leaf Alexander Duyck
2015-02-24 20:48 ` [RFC PATCH 07/29] fib_trie: Fib find node should return parent Alexander Duyck
2015-02-24 20:48 ` [RFC PATCH 08/29] fib_trie: Update insert and delete to make use of tp from find_node Alexander Duyck
2015-02-24 20:48 ` [RFC PATCH 09/29] fib_trie: Make fib_table rcu safe Alexander Duyck
2015-02-24 20:49 ` [RFC PATCH 10/29] fib_trie: Return pointer to tnode pointer in resize/inflate/halve Alexander Duyck
2015-02-24 20:49 ` [RFC PATCH 11/29] fib_trie: Rename tnode to key_vector Alexander Duyck
2015-02-24 20:49 ` [RFC PATCH 12/29] fib_trie: move leaf and tnode to occupy the same spot in the key vector Alexander Duyck
2015-02-24 20:49 ` [RFC PATCH 13/29] fib_trie: replace tnode_get_child functions with get_child macros Alexander Duyck
2015-02-24 20:49 ` [RFC PATCH 14/29] fib_trie: Rename tnode_child_length to child_length Alexander Duyck
2015-02-24 20:49 ` [RFC PATCH 15/29] fib_trie: Add tnode struct as a container for fields not needed in key_vector Alexander Duyck
2015-02-24 20:49 ` [RFC PATCH 16/29] fib_trie: Move rcu from key_vector to tnode, add accessors Alexander Duyck
2015-02-24 20:49 ` [RFC PATCH 17/29] fib_trie: Pull empty_children and full_children into tnode Alexander Duyck
2015-02-24 20:49 ` [RFC PATCH 18/29] fib_trie: Move parent from key_vector to tnode Alexander Duyck
2015-02-24 20:50 ` [RFC PATCH 19/29] fib_trie: Add key vector to root, return parent key_vector in resize Alexander Duyck
2015-02-24 20:50 ` [RFC PATCH 20/29] fib_trie: Push net pointer down into fib_trie insert/delete/flush calls Alexander Duyck
2015-02-24 20:50 ` [RFC PATCH 21/29] fib_trie: Rewrite handling of RCU to include parent in replacement Alexander Duyck
2015-02-24 20:50 ` [RFC PATCH 22/29] fib_trie: Allocate tnode as array of key_vectors instead of key_vector as array of tnode pointers Alexander Duyck
2015-02-24 20:50 ` [RFC PATCH 23/29] fib_trie: Add leaf_init Alexander Duyck
2015-02-24 20:50 ` [RFC PATCH 24/29] fib_trie: Update tnode_new to drop use of put_child_root Alexander Duyck
2015-02-24 20:50 ` [RFC PATCH 25/29] fib_trie: Add function for dropping children from trie Alexander Duyck
2015-02-24 20:50 ` [RFC PATCH 26/29] fib_trie: Use put_child to only copy key_vectors instead of pointers Alexander Duyck
2015-02-24 20:50 ` [RFC PATCH 27/29] fib_trie: Move key and pos into key_vector from tnode Alexander Duyck
2015-02-24 20:51 ` [RFC PATCH 28/29] fib_trie: Move slen from tnode to key vector Alexander Duyck
2015-02-24 20:51 ` [RFC PATCH 29/29] fib_trie: Push bits up one level, and move leaves up into parent key_vector array Alexander Duyck
2015-02-25  3:53 ` [RFC PATCH 00/29] Phase 2 of fib_trie updates David Miller
2015-02-25  5:12   ` Alexander Duyck
2015-02-27 21:01     ` David Miller

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=20150224202837.26106.87623.stgit@ahduyck-vm-fedora20 \
    --to=alexander.h.duyck@redhat.com \
    --cc=netdev@vger.kernel.org \
    /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