|
On Saturday, October 30, 2010 08:53:03 Simon Horman wrote:
> On Fri, Oct 29, 2010 at 02:22:00PM +0200, Hans Schillstrom wrote:
> > Add pack/unpack of IPv6 address in sync message
> > in binary form similar to the ASCII ::
> >
> > A packed IPv6 address constists of -
> > first byte
> > high nibble first segment leng in bytes
> > low nibble possition of last segment.
> > then
> > First segemnt i.e. left side of ::
> > Last segment right side of ::
> >
> > Examle FC00::2
> > unpacked FC00 0000 0000 0000 0000 0000 0000 0002
> > packed 1F FC02
> > Example 2 2003::2:100
> > unpacked 2003 0000 0000 0000 0000 0000 0002 0100
> > packed 2D 20 03 02 01 00
>
> Hi Hans,
>
> are you sure the space saved is worth the additional complexity
> that this patch introduces?
> --
Hello Simon
I can see that my first mail 0/1 is still in my outbox.....
that mail contains just the question above and some calculations
My conclusion was that if you have 100M interface and a x64 CPU with a more
optimized version
than it might be worth to pack the IPv6 Address.
On a 32bit CPU with 1G interfce the packing consume the same time that you
might earn
The patch is just an example of how to do it,
Sorry for the confusion that I created without the [RFC PATCH 0/1]
The intension was to ask if anyone thinks that packing of IPv6 address should
be there ?
I would say, much pain for little gain
Regards
Hans
--
To unsubscribe from this list: send the line "unsubscribe lvs-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
|
