netdev - Re: [PATCHv6 bpf-next 1/3] xdp: add a new helper for dev map multicast support

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 for Android: free password hash cracker in your pocket

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

Date:   Sat, 11 Jul 2020 08:26:32 +0800
From:   Hangbin Liu <liuhangbin@...il.com>
To:     David Ahern <dsahern@...il.com>
Cc:     bpf@...r.kernel.org, netdev@...r.kernel.org,
        Toke Høiland-Jørgensen <toke@...hat.com>,
        Jiri Benc <jbenc@...hat.com>,
        Jesper Dangaard Brouer <brouer@...hat.com>,
        Eelco Chaudron <echaudro@...hat.com>, ast@...nel.org,
        Daniel Borkmann <daniel@...earbox.net>,
        Lorenzo Bianconi <lorenzo.bianconi@...hat.com>
Subject: Re: [PATCHv6 bpf-next 1/3] xdp: add a new helper for dev map
 multicast support

On Fri, Jul 10, 2020 at 07:46:18AM -0600, David Ahern wrote:
> >>
> >> I'm probably missing something fundamental, but why do you need to walk
> >> the keys? Why not just do a lookup on the device index?
> > 
> > This functions is to check if the device index is in exclude map.
> > 
> > The device indexes are stored as values in the map. The user could store
> > the values by any key number. There is no way to lookup the device index
> > directly unless loop the map and check each values we stored.
> 
> Right.
> 
> The point of DEVMAP_HASH is to allow map management where key == device
> index (vs DEVMAP which for any non-trivial use case is going to require
> key != device index). You could require the exclude map to be
> DEVMAP_HASH and the key to be the index allowing you to do a direct
> lookup. Having to roam the entire map looking for a match does not scale
> and is going to have poor performance with increasing number of entries.
> XDP is targeted at performance with expert level of control, so
> constraints like this have to be part of the deal.

Yes, if we have this constraints the performance should have some improvement.

Do you think we should do it right now or in later performance update patch.

Thanks
Hangbin