Date: Fri, 21 May 2004 01:26:52 -0400 From: Dmitri Denissov <ddenissov@sandvine.com> To: 'Julian Elischer' <julian@elischer.org>, freebsd-net@freebsd.org Subject: RE: protecting netgraph calls from outside of the network context Message-ID: <FE045D4D9F7AED4CBFF1B3B813C85337012C0D5D@mail.sandvine.com>
next in thread | raw e-mail | index | archive | help
On Thu, 20 May 2004, Julian Elischer wrote: > > This is kind of a bridge, connected to ng_ether interface nodes. > > Sometimes it queues received packets and later /on a timer > call or a call > > from the user space/ > > it re-injects the packets using ng_send_data. > > > > reinjects it to where? > To the adjacent ng_ether node connected via the lower hook. So the question is how to protect the ng_send_data() call originated from timer or user space context. > > > > > > From: Julian Elischer [mailto:julian@elischer.org] > > > Sent: Thursday, May 20, 2004 5:34 PM > > > > > > > > > Ha! funny you should ask that exactly now.. > > > I was just discussing this with Robert Watson.. > > > > > > The answer is "it depends on what you want to do". > > > > > > What DO you want to do and what does your node do? > > > netgraph has internal locking in 5.x that you need to > > > interact with but it should be pretty transparrent > > > depending on what you are doing.. > > > > > > > > > On Thu, 20 May 2004, Dmitri Denissov wrote: > > > > > > > Hi, > > > > > > > > Currently netgraph code uses splnet/splx to protect > timeout calls. > > > > This doesn't work with 5.2 SMP kernel. What is the > proper method > > > > here for a custom netgraph node? Is the Giant lock only the way? > > > > > > > > Thanks > > > > > > > > -- > > > > Dmitri > > >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?FE045D4D9F7AED4CBFF1B3B813C85337012C0D5D>