From: Thomas Graf <tgraf@suug.ch>
To: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Patrick McHardy <kaber@trash.net>,
davem@davemloft.net, paulmck@linux.vnet.ibm.com,
ying.xue@windriver.com, netdev@vger.kernel.org,
netfilter-devel@vger.kernel.org
Subject: Re: [PATCH 3/3] netlink: Lock out table resizes while dumping Netlink sockets
Date: Wed, 21 Jan 2015 09:37:22 +0000 [thread overview]
Message-ID: <20150121093722.GM20315@casper.infradead.org> (raw)
In-Reply-To: <20150121050819.GA23062@gondor.apana.org.au>
On 01/21/15 at 04:08pm, Herbert Xu wrote:
> On Tue, Jan 20, 2015 at 03:35:56PM +0000, Thomas Graf wrote:
> > On 01/20/15 at 03:21pm, Patrick McHardy wrote:
> > > I think its preferrable to make the need to handle NETLINK_F_DUMP_INTR
> > > as noticable as possible and not hide it. Silent failure is the worst
> > > kind of failure.
> >
> > I agree to that. The point here is to avoid unnecessary use of
> > NETLINK_F_DUMP_INTR if all entries fit into a single message buffer.
>
> OK I think I have a solution for you guys. But first you'll need to
> wait for me to undo the nulls stuff so I can steal that bit which
> is central to my solution.
Without having seen your code, can we make it configurable on what
the bit is used for? Use of nulls marker is a strict requirement for
some targeted users of rhashtable.
> Essentially I need a bit to indicate an entry in the bucket chain
> should be skipped, either because it has just been removed or that
> it is a walker entry (see xfrm_state_walk).
>
> The way it'll work then is exactly the same as xfrm_state_walk,
> except that the linked list is broken up into individual buckets.
>
> Of course we'll still need to postpone resizes (and rehashes which
> is what my work is about) during a walk but I think that's a fair
> price to pay.
If I understand this correctly we also need to block out parallel
walkers and we need to start taking bucket locks while walking to
modify the walker mark bit in peace.
> This also means handling insertion failures but I think that
> should be acceptable if we make it based on a configurable maximum
> chain length along with forced resize/rehash where possible.
>
> Note that this can be made optional, i.e., if the user can afford
> memory to do their own walking (e.g., xfrm_state), then none of
> this needs to happen and it'll just work as it does now. IOW if
> you don't use this special rhashtable walk function then you're
> not affected.
That sounds like the best option to me.
next prev parent reply other threads:[~2015-01-21 9:37 UTC|newest]
Thread overview: 79+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-01-20 13:20 [PATCH 0/3 net-next] rhashtable: Notify on resize to allow signaling interrupted dumps Thomas Graf
2015-01-20 13:20 ` [PATCH 1/3] rhashtable: Provide notifier for deferred resizes Thomas Graf
2015-01-20 13:20 ` [PATCH 2/3] netlink: Mark dumps as inconsistent which have been interrupted by a resize Thomas Graf
2015-01-21 8:13 ` Ying Xue
2015-01-21 12:17 ` Thomas Graf
2015-01-22 8:49 ` Herbert Xu
2015-01-22 8:56 ` Patrick McHardy
2015-01-22 9:22 ` Herbert Xu
2015-01-22 10:07 ` Patrick McHardy
2015-01-25 23:20 ` [PATCH 0/2] rhashtable: Add walk iterator primitives and use them in netlink Herbert Xu
2015-01-25 23:21 ` [PATCH 1/2] rhashtable: Introduce rhashtable_walk_* Herbert Xu
2015-01-26 8:20 ` Thomas Graf
2015-01-26 22:21 ` Herbert Xu
2015-01-26 10:09 ` David Laight
2015-01-26 22:23 ` Herbert Xu
2015-01-26 22:36 ` David Miller
2015-01-26 22:42 ` Herbert Xu
2015-01-26 23:31 ` Herbert Xu
2015-01-27 9:45 ` Thomas Graf
2015-01-27 9:54 ` Herbert Xu
2015-01-27 10:15 ` Thomas Graf
2015-01-27 10:24 ` Herbert Xu
2015-01-27 11:16 ` Thomas Graf
2015-01-27 11:23 ` Herbert Xu
2015-01-27 11:40 ` Thomas Graf
2015-01-27 20:39 ` Herbert Xu
2015-01-27 22:10 ` David Miller
2015-01-27 23:16 ` Herbert Xu
2015-01-27 13:09 ` Patrick McHardy
2015-01-27 20:36 ` Herbert Xu
2015-01-28 19:07 ` Patrick McHardy
2015-01-30 5:58 ` Herbert Xu
2015-01-30 8:10 ` Patrick McHardy
2015-01-27 10:09 ` David Laight
2015-01-27 10:12 ` Herbert Xu
2015-01-25 23:21 ` [PATCH 2/2] netlink: Use rhashtable walk iterator Herbert Xu
2015-01-27 23:19 ` [PATCH 0/2] rhashtable: Add walk iterator primitives and use them in netlink Herbert Xu
2015-01-27 23:20 ` [PATCH 1/2] rhashtable: Introduce rhashtable_walk_* Herbert Xu
2015-01-29 22:26 ` Thomas Graf
2015-01-27 23:20 ` [PATCH 2/2] netlink: Use rhashtable walk iterator Herbert Xu
2015-01-29 22:27 ` Thomas Graf
2015-01-29 22:42 ` [PATCH 0/2] rhashtable: Add walk iterator primitives and use them in netlink David Miller
2015-01-31 3:13 ` Herbert Xu
2015-01-31 3:14 ` [PATCH 1/2] rhashtable: Introduce rhashtable_walk_* Herbert Xu
2015-01-31 3:14 ` [PATCH 2/2] netlink: Use rhashtable walk iterator Herbert Xu
2015-01-31 4:31 ` netfilter: " Herbert Xu
2015-02-01 7:45 ` Patrick McHardy
2015-02-03 3:19 ` David Miller
2015-02-03 3:19 ` [PATCH 0/2] rhashtable: Add walk iterator primitives and use them in netlink David Miller
2015-01-20 13:20 ` [PATCH 3/3] netlink: Lock out table resizes while dumping Netlink sockets Thomas Graf
2015-01-20 14:31 ` Patrick McHardy
2015-01-20 14:55 ` Thomas Graf
2015-01-20 15:21 ` Patrick McHardy
2015-01-20 15:35 ` Thomas Graf
2015-01-21 5:08 ` Herbert Xu
2015-01-21 5:15 ` Herbert Xu
2015-01-21 9:14 ` Herbert Xu
2015-01-21 9:56 ` Thomas Graf
2015-01-21 9:59 ` Herbert Xu
2015-01-21 10:00 ` Patrick McHardy
2015-01-21 9:37 ` Thomas Graf [this message]
2015-01-21 9:38 ` Herbert Xu
2015-01-21 9:49 ` Thomas Graf
2015-01-21 9:58 ` Herbert Xu
2015-01-21 10:23 ` Thomas Graf
2015-01-22 6:35 ` Herbert Xu
2015-01-22 7:20 ` Herbert Xu
2015-01-22 9:05 ` Thomas Graf
2015-01-22 9:50 ` Herbert Xu
2015-01-21 10:34 ` Thomas Graf
2015-01-21 10:40 ` Patrick McHardy
2015-01-21 11:37 ` Thomas Graf
2015-01-21 11:59 ` Patrick McHardy
2015-01-21 12:07 ` Thomas Graf
2015-01-21 12:09 ` Patrick McHardy
2015-01-21 10:36 ` David Laight
2015-01-20 15:00 ` David Laight
2015-01-20 15:05 ` Thomas Graf
2015-01-21 5:11 ` Herbert Xu
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=20150121093722.GM20315@casper.infradead.org \
--to=tgraf@suug.ch \
--cc=davem@davemloft.net \
--cc=herbert@gondor.apana.org.au \
--cc=kaber@trash.net \
--cc=netdev@vger.kernel.org \
--cc=netfilter-devel@vger.kernel.org \
--cc=paulmck@linux.vnet.ibm.com \
--cc=ying.xue@windriver.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).