netdev - Re: Ottawa and slow hash-table resize

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Wed, 25 Feb 2015 17:38:03 +0000
From:	Thomas Graf <tgraf@...g.ch>
To:	Herbert Xu <herbert@...dor.apana.org.au>
Cc:	josh@...htriplett.org, David Miller <davem@...emloft.net>,
	kaber@...sh.net, paulmck@...ux.vnet.ibm.com,
	alexei.starovoitov@...il.com, ying.xue@...driver.com,
	netdev@...r.kernel.org, netfilter-devel@...r.kernel.org
Subject: Re: Ottawa and slow hash-table resize

On 02/25/15 at 09:56pm, Herbert Xu wrote:
> On Tue, Feb 24, 2015 at 10:34:09PM +0000, Thomas Graf wrote:
> >
> > There is a side effect. We can't grow the number of bucket locks more
> > than 2x if we grow the table itself faster than 2x. So if we start
> > out with a table size of 512 and grow 4 times in a row we will end
> > up with a theoretical max bucket locks of 4K. Probably enough though.
> 
> Does this limit on bucket locks apply to my rehash scheme?

I don't think so. I assume you get rid of all the old bucket
== new bucket assumptions and introduce some form of nested
locking to account for the fact that entries in an old bucket
may map to an arbitrary number of new buckets.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html