netfilter-devel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Pablo Neira Ayuso <pablo@netfilter.org>
To: Phil Sutter <phil@nwl.cc>
Cc: netfilter-devel@vger.kernel.org
Subject: Re: [iptables PATCH v3 04/11] nft-cache: Introduce cache levels
Date: Wed, 9 Oct 2019 11:37:23 +0200	[thread overview]
Message-ID: <20191009093723.snbyd6xvtd5gpnto@salvia> (raw)
In-Reply-To: <20191008161447.6595-5-phil@nwl.cc>

Hi Phil,

On Tue, Oct 08, 2019 at 06:14:40PM +0200, Phil Sutter wrote:
> Replace the simple have_cache boolean by a cache level indicator
> defining how complete the cache is. Since have_cache indicated full
> cache (including rules), make code depending on it check for cache level
> NFT_CL_RULES.
> 
> Core cache fetching routine __nft_build_cache() accepts a new level via
> parameter and raises cache completeness to that level.
> 
> Signed-off-by: Phil Sutter <phil@nwl.cc>
> ---
>  iptables/nft-cache.c | 51 +++++++++++++++++++++++++++++++-------------
>  iptables/nft.h       |  9 +++++++-
>  2 files changed, 44 insertions(+), 16 deletions(-)
> 
> diff --git a/iptables/nft-cache.c b/iptables/nft-cache.c
> index 5444419a5cc3b..22a87e94efd76 100644
> --- a/iptables/nft-cache.c
> +++ b/iptables/nft-cache.c
> @@ -224,30 +224,49 @@ static int fetch_rule_cache(struct nft_handle *h)
>  	return 0;
>  }
>  
> -static void __nft_build_cache(struct nft_handle *h)
> +static void __nft_build_cache(struct nft_handle *h, enum nft_cache_level level)
>  {
>  	uint32_t genid_start, genid_stop;
>  
> +	if (level <= h->cache_level)
> +		return;
>  retry:
>  	mnl_genid_get(h, &genid_start);
> -	fetch_table_cache(h);
> -	fetch_chain_cache(h);
> -	fetch_rule_cache(h);
> -	h->have_cache = true;
> -	mnl_genid_get(h, &genid_stop);
>  
> +	switch (h->cache_level) {
> +	case NFT_CL_NONE:
> +		fetch_table_cache(h);
> +		if (level == NFT_CL_TABLES)
> +			break;
> +		/* fall through */
> +	case NFT_CL_TABLES:

If the existing level is TABLES and use wants chains, then you have to
invalidate the existing table cache, then fetch the tables and chains
to make sure cache is consistent. I mean, extending an existing cache
might lead to inconsistencies.

Am I missing anything?

> +		fetch_chain_cache(h);
> +		if (level == NFT_CL_CHAINS)
> +			break;
> +		/* fall through */
> +	case NFT_CL_CHAINS:
> +		fetch_rule_cache(h);
> +		if (level == NFT_CL_RULES)
> +			break;
> +		/* fall through */
> +	case NFT_CL_RULES:
> +		break;
> +	}
> +
> +	mnl_genid_get(h, &genid_stop);
>  	if (genid_start != genid_stop) {
>  		flush_chain_cache(h, NULL);
>  		goto retry;
>  	}
>  
> +	h->cache_level = level;
>  	h->nft_genid = genid_start;
>  }

  reply	other threads:[~2019-10-09  9:37 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-08 16:14 [iptables PATCH v3 00/11] Improve iptables-nft performance with large rulesets Phil Sutter
2019-10-08 16:14 ` [iptables PATCH v3 01/11] nft: Pass nft_handle to flush_cache() Phil Sutter
2019-10-09  9:30   ` Pablo Neira Ayuso
2019-10-08 16:14 ` [iptables PATCH v3 02/11] nft: Avoid nested cache fetching Phil Sutter
2019-10-09  9:30   ` Pablo Neira Ayuso
2019-10-08 16:14 ` [iptables PATCH v3 03/11] nft: Extract cache routines into nft-cache.c Phil Sutter
2019-10-09  9:32   ` Pablo Neira Ayuso
2019-10-08 16:14 ` [iptables PATCH v3 04/11] nft-cache: Introduce cache levels Phil Sutter
2019-10-09  9:37   ` Pablo Neira Ayuso [this message]
2019-10-09 10:29     ` Pablo Neira Ayuso
2019-10-10 22:09       ` Phil Sutter
2019-10-11  9:28         ` Pablo Neira Ayuso
2019-10-11 11:24           ` Phil Sutter
2019-10-14 10:00             ` Pablo Neira Ayuso
2019-10-11 10:20         ` Pablo Neira Ayuso
2019-10-08 16:14 ` [iptables PATCH v3 05/11] nft-cache: Fetch only chains in nft_chain_list_get() Phil Sutter
2019-10-08 16:14 ` [iptables PATCH v3 06/11] nft-cache: Cover for multiple fetcher invocation Phil Sutter
2019-10-08 16:14 ` [iptables PATCH v3 07/11] nft-cache: Support partial cache per table Phil Sutter
2019-10-08 16:14 ` [iptables PATCH v3 08/11] nft-cache: Support partial rule cache per chain Phil Sutter
2019-10-08 16:14 ` [iptables PATCH v3 09/11] nft: Reduce cache overhead of nft_chain_builtin_init() Phil Sutter
2019-10-08 16:14 ` [iptables PATCH v3 10/11] nft: Support nft_is_table_compatible() per chain Phil Sutter
2019-10-08 16:14 ` [iptables PATCH v3 11/11] nft: Optimize flushing all chains of a table Phil Sutter

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=20191009093723.snbyd6xvtd5gpnto@salvia \
    --to=pablo@netfilter.org \
    --cc=netfilter-devel@vger.kernel.org \
    --cc=phil@nwl.cc \
    /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).