[Top] [All Lists]


To: Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx>
Subject: Re: [PATCH] NETLINK_UESTABLISHED notifier event
From: Dmitry Yusupov <dmitry_yus@xxxxxxxxx>
Date: Fri, 08 Apr 2005 08:30:14 -0700
Cc: Mike Christie <michaelc@xxxxxxxxxxx>, netdev@xxxxxxxxxxx, davem@xxxxxxxxxxxxx
In-reply-to: <20050408113654.GA26095@xxxxxxxxxxxxxxxxxxx>
References: <E1DJ0YI-0003PR-00@xxxxxxxxxxxxxxxxxxxxxxxx> <42540CF3.7070501@xxxxxxxxxxx> <20050406212906.GA24775@xxxxxxxxxxxxxxxxxxx> <1112823442.16753.68.camel@beastie> <20050406220417.GA4443@xxxxxxxxxxxxxxxxxxx> <1112826385.16753.99.camel@beastie> <20050407213231.GA28738@xxxxxxxxxxxxxxxxxxx> <1112917001.3893.77.camel@beastie> <20050408113654.GA26095@xxxxxxxxxxxxxxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx
On Fri, 2005-04-08 at 21:36 +1000, Herbert Xu wrote:
> On Thu, Apr 07, 2005 at 04:36:41PM -0700, Dmitry Yusupov wrote:
> > 
> > This actually even better since we will guarantee "down" call delivery.
> > My only concern is that it is not very generic. Do you know clean way to
> > implement it?
> How generic do you want this? Do you need this for socket types other
> than netlink?

as discussed on kernel-summit-discuss. yes. we need generic OOM-safety

> For a one-packet version, we can pre-allocate an skb/page in response to
> a setsockopt and store it in sk_send_head/sk_sndmsg_page.
> This can then be used at sendmsg time.  Obviously subsequent messages
> will have to use alloc_skb until that skb is released.
> Alternatively, we can let the socket allocate skb's from an emergency
> pool similar to what was discussed in the thread
> "Summary of 2005 Kernel Summit Proposed Topics".

This is ideally how it should be for sendmsg paths. socket applications
like iscsi, nbd, etc will use it for TCP/IP type of socket. iscsi could
re-use the same generic "emergency pool" code for netlink.

> Again this could enabled on a setsockopt.


> Cheers,

<Prev in Thread] Current Thread [Next in Thread>