Network stack cloning / virtualization patches

[Peter Jeremy]

On Mon, Jun 02, 2003 at 11:58:25PM +0200, Marko Zec wrote:
>There are two major possible causes for overhead increase. First, each IP
>protocol related tunable variable and most of the global symbols involved in
>network processing have been virtualized. [...]  And second, many kernel
>functions have been extended with an additional argument, typically a pointer
>to a struct vimage,

A third issue on the x86 is a lack of registers:  There are only 6
"general purpose" registers (and each of them actually has a specific
purpose).  Eating one of these registers to maintain a pointer to
a struct vimage will be a noticable performance hit.

> However, a couple of percents in overhead increase that can be
>observed only in worst case loopback tests do not present a problem
>in any real-life scenario.

Agreed.  It would be useful to get some real-world figures.

>> Julian, am I safe in assuming that you have an interest in this work?
>> If not, I may setup a p4 branch to work with and to merge these bits
>> into -CURRENT if no one else is interested.  -sc
>
>I would be really honored to see the cloning code merged in -CURRENT one day.
>However, at the moment I'm strongly opposed to such a proposal, since the code
>is simply not mature enough.

My understanding is that the p4 tree is specifically intended for this
sort of thing.  It provides a controlled environment where the code
could be nursed from its current "immature" form into something that
can be safely merged into the main CVS tree.

Peter