Add OVS_PACKET_ATTR_KEY_INFO to relieve userspace from re-computing
data already computed within the kernel datapath. In the typical
case of an upcall with perfect key fitness between kernel and
userspace this eliminates flow_extract() and flow_hash() calls in
handle_miss_upcalls().
Additional bo
Please have a look at this version. The other version had a problem in the
unit tests which could caused them to fail when the hash implementation
changed. In this function, I remove the dependency that caused the problem.
---
This commit moves responsibility for implementing patch ports from
t
On Tue, Jan 22, 2013 at 12:51:45PM -0800, Ben Pfaff wrote:
> On Tue, Jan 22, 2013 at 09:35:01AM +0900, Simon Horman wrote:
> > On Thu, Jan 17, 2013 at 09:42:29AM +0900, Simon Horman wrote:
> > > On Wed, Jan 16, 2013 at 01:59:21PM -0800, Ben Pfaff wrote:
> > > > On Tue, Jan 15, 2013 at 01:20:57PM +0
This commit moves responsibility for implementing patch ports from
the datapath to ofproto-dpif. There are two main reasons to do
this.
The first is a matter of design: ofproto-dpif both has more
information than the datapath, and is better suited to handle the
complexity required to implement p
> I think we discussed this at one point but I don't remember the
> conclusion: did you consider making these per-dpif_backer rather than
> global? It is pretty rare to have more than one backer, so maybe that
> is the logic.
I don't think it matters too much either way. In general, upon reflect
Thanks, I pushed this.
Ethan
On Tue, Jan 22, 2013 at 3:51 PM, Ben Pfaff wrote:
> On Tue, Jan 22, 2013 at 03:44:21PM -0800, Ethan Jackson wrote:
>> Found by inspection.
>>
>> Signed-off-by: Ethan Jackson
>
> Looks good, thanks.
___
dev mailing list
dev
On Tue, Jan 22, 2013 at 03:44:21PM -0800, Ethan Jackson wrote:
> Found by inspection.
>
> Signed-off-by: Ethan Jackson
Looks good, thanks.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Thu, Jan 17, 2013 at 02:26:06PM +0900, Simon Horman wrote:
> > On Tue, Jan 15, 2013 at 11:05:45PM -0800, Ben Pfaff wrote:
> > > parse_l3_onward() could use flow_innermost_dl_type() since that's what
> > > it's effectively calculating as 'dl_type' (maybe it should be
> > > 'inner_dl_type' or 'inn
Found by inspection.
Signed-off-by: Ethan Jackson
---
ofproto/ofproto-dpif.c |1 +
1 file changed, 1 insertion(+)
diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.c
index b37b482..2c216fe 100644
--- a/ofproto/ofproto-dpif.c
+++ b/ofproto/ofproto-dpif.c
@@ -1022,6 +1022,7 @@ open_d
On Tue, Jan 22, 2013 at 10:47:17AM -0800, Ben Pfaff wrote:
> On Fri, Jan 18, 2013 at 09:27:55AM +0900, Simon Horman wrote:
> > This patch implements use-space datapath and non-datapath code
> > to match and use the datapath API set out in Leo Alterman's patch
> > "user-space datapath: Add basic MPL
On Tue, Jan 22, 2013 at 02:02:04PM -0800, Justin Pettit wrote:
> On Jan 22, 2013, at 1:04 PM, Ben Pfaff wrote:
>
> > On Fri, Jan 18, 2013 at 01:18:08PM -0800, Justin Pettit wrote:
> >> Signed-off-by: Justin Pettit
> >
> > The only problem I see is that we never released 1.7.2 as far as I can
>
On Jan 22, 2013, at 1:04 PM, Ben Pfaff wrote:
> On Fri, Jan 18, 2013 at 01:18:08PM -0800, Justin Pettit wrote:
>> Signed-off-by: Justin Pettit
>
> The only problem I see is that we never released 1.7.2 as far as I can
> tell. Should we just release 1.7.2 instead of skipping a number?
As we di
On Jan 22, 2013, at 1:54 PM, Ben Pfaff wrote:
> On Tue, Jan 22, 2013 at 01:53:49PM -0800, Justin Pettit wrote:
>> Signed-off-by: Justin Pettit
>
> Looks good, thanks.
Thanks. I pushed it.
--Justin
___
dev mailing list
dev@openvswitch.org
http://o
On Tue, Jan 22, 2013 at 01:53:49PM -0800, Justin Pettit wrote:
> Signed-off-by: Justin Pettit
Looks good, thanks.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
Applied to branch-1.9. (I thought we'd had murmurhash around for
longer, but it wasn't in branch-1.8.)
On Tue, Jan 22, 2013 at 01:47:51PM -0800, Ethan Jackson wrote:
> It seems low risk to me. I'd like it to be there in case we end up
> backporting something important which relies on mhash_finis
Signed-off-by: Justin Pettit
---
NEWS |4
configure.ac |2 +-
debian/changelog |6 ++
3 files changed, 11 insertions(+), 1 deletions(-)
diff --git a/NEWS b/NEWS
index a170814..a318b15 100644
--- a/NEWS
+++ b/NEWS
@@ -1,3 +1,7 @@
+v1.4.5 - 22 Jan 2013
+---
It seems low risk to me. I'd like it to be there in case we end up
backporting something important which relies on mhash_finish().
Ethan
On Tue, Jan 22, 2013 at 1:44 PM, Ben Pfaff wrote:
> Thanks. I pushed these to master.
>
> Do you think I should apply "hash: Correct implementation of
> mhas
Thanks. I pushed these to master.
Do you think I should apply "hash: Correct implementation of
mhash_finish()." to older branches?
On Tue, Jan 22, 2013 at 01:37:07PM -0800, Ethan Jackson wrote:
> Acked-by: Ethan Jackson
>
> Sorry it took a while to get around to this. Looks good, thanks.
>
>
Acked-by: Ethan Jackson
Sorry it took a while to get around to this. Looks good, thanks.
Ethan
On Fri, Dec 14, 2012 at 4:33 PM, Ben Pfaff wrote:
> murmurhash includes an xor with the number of bytes hashed in its finishing
> step. Until now, we've only had murmurhash for full words, but an u
On Tue, Jan 22, 2013 at 07:39:01PM +0100, Markus Schuster wrote:
> Open vSwitch sends out the multicast frames on ALL physical interfaces
> belonging to the SLB bond. That causes a lot of confusion on the physical
> switches that XCP hosts are connected to (VM MAC addresses jumping between
> por
On Fri, Jan 18, 2013 at 01:18:08PM -0800, Justin Pettit wrote:
> Signed-off-by: Justin Pettit
The only problem I see is that we never released 1.7.2 as far as I can
tell. Should we just release 1.7.2 instead of skipping a number?
___
dev mailing list
d
On Tue, Jan 22, 2013 at 09:35:01AM +0900, Simon Horman wrote:
> On Thu, Jan 17, 2013 at 09:42:29AM +0900, Simon Horman wrote:
> > On Wed, Jan 16, 2013 at 01:59:21PM -0800, Ben Pfaff wrote:
> > > On Tue, Jan 15, 2013 at 01:20:57PM +0900, Simon Horman wrote:
> > > > Optimise OpenFlow flow expiry by p
Signed-off-by: Justin Pettit
---
NEWS |4
configure.ac |2 +-
debian/changelog |7 +++
3 files changed, 12 insertions(+), 1 deletions(-)
diff --git a/NEWS b/NEWS
index 9e1e740..76d52c6 100644
--- a/NEWS
+++ b/NEWS
@@ -1,3 +1,7 @@
+v1.7.3 - 22 Jan 2013
+--
On Tue, Jan 22, 2013 at 09:27:14AM +0900, Simon Horman wrote:
> On Thu, Jan 17, 2013 at 02:26:06PM +0900, Simon Horman wrote:
> > On Wed, Jan 16, 2013 at 05:38:08PM +0900, Simon Horman wrote:
> > > On Tue, Jan 15, 2013 at 11:05:45PM -0800, Ben Pfaff wrote:
>
> [snip]
>
> > > > parse_l3_onward() c
Hi everybody,
we're using Open vSwitch as it's the default software switch/bridge in XCP
1.6 (Xen Cloud Platform). As far as I know, it's version 1.4.2.
I've already posted this problem to the XCP mailinglist but got no reply, so
I'm trying again here :)
The following is a quote of my mail to
On Fri, Jan 18, 2013 at 09:27:55AM +0900, Simon Horman wrote:
> This patch implements use-space datapath and non-datapath code
> to match and use the datapath API set out in Leo Alterman's patch
> "user-space datapath: Add basic MPLS support to kernel".
I'm resuming my review starting from packets
From: Lorand Jakab
LISP is an experimental layer 3 tunneling protocol. This patch adds
support for LISP tunneling. Since LISP encapsulated packets do not
carry an Ethernet header, it is removed before encapsulation, and added
with hardcoded source and destination MAC addresses after decapsulati
Add support to the tunneling code for a "pre_tunnel"
function. This allows the tunneling code to perform
operations on the packet before the outer IP header is
added.
A tunneling protocol such as LISP may require this, as it needs
to remove the MAC header before applying the LISP header.
Signed-o
The following two patches provide support for the LISP tunneling protocol into
Open vSwitch. See the latest IETF draft for LISP here:
http://tools.ietf.org/html/draft-ietf-lisp-24
Kyle Mestery (1):
Add support to the tunneling code for a "pre_tunnel" function.
This allows the tunneling
Thanks. I think I ought to test it before I commit it, so it probably
won't go in before afternoon.
On Tue, Jan 22, 2013 at 10:17:16AM -0800, Justin Pettit wrote:
> Looks good. Thanks!
>
> --Justin
>
>
> On Jan 22, 2013, at 10:15 AM, Ben Pfaff wrote:
>
> > On Tue, Jan 22, 2013 at 10:06:12AM
On Mon, Jan 21, 2013 at 03:10:02PM +0900, Simon Horman wrote:
> Given the current implementation it is not clear to me that the result
> calculated by ofproto_enumerate_types() can ever change.
>
> By caching the lookup made by ofproto_enumerate_types() calls to
> shash_add() and shash_destroy() d
Looks good. Thanks!
--Justin
On Jan 22, 2013, at 10:15 AM, Ben Pfaff wrote:
> On Tue, Jan 22, 2013 at 10:06:12AM -0800, Justin Pettit wrote:
>> I hope you had this section queued up for review, and not that you
>> whipped it up in the minute since you responded to that QoS question
>> on ovs-
On Tue, Jan 22, 2013 at 10:06:12AM -0800, Justin Pettit wrote:
> I hope you had this section queued up for review, and not that you
> whipped it up in the minute since you responded to that QoS question
> on ovs-discuss. I already feel inadequate enough.
It took me about half an hour to write it
On Mon, Jan 21, 2013 at 03:10:01PM +0900, Simon Horman wrote:
> It is not clear to me that the result calculated by dp_enumerate_type()
> can ever change.
>
> By caching the lookup made by dp_enumerate_types() calls to
> shash_add() and shash_destroy() destroy are avoided reducing the use of
> mal
I hope you had this section queued up for review, and not that you whipped it
up in the minute since you responded to that QoS question on ovs-discuss. I
already feel inadequate enough.
> +A: Suppose that you want to set up bridge br0 connected to physical
> + Ethernet port eth0 (a 1 Gbps dev
On Mon, Jan 21, 2013 at 03:10:00PM +0900, Simon Horman wrote:
> In general the network devices don't change very often.
> By caching the lookup made by netdev_linux_miimon_run()
> calls to shash_add() and shash_destroy() destroy are avoided
> reducing the use of malloc(), free() and hash_bytes().
>
On Fri, Jan 18, 2013 at 08:09:55PM -0500, Pavithra Ramesh wrote:
> If socket path specified in the set-controller command is
> relative to ovs_rundir(), append the directory name to target.
>
> Bug#14029
> Signed-off-by: Pavithra Ramesh
This code doesn't own c->target (it belongs to the database
Signed-off-by: Ben Pfaff
---
I haven't actually tested the vsctl or ofctl commands here yet
FAQ | 104 +++
1 file changed, 104 insertions(+)
diff --git a/FAQ b/FAQ
index ab1c1cc..7fecc52 100644
--- a/FAQ
+++ b/FAQ
@@ -455,6 +455,1
HELLO MY DEAR.
Happy new year...
Hello, My name is Edda, I am a woman
can we be friends? I am sorry for disturbing you,if yes, please ,i will
like to hear from you, upon your reply i will send you my photos and
tell you more about me.
39 matches
Mail list logo
