On Wed, Dec 14, 2011 at 4:23 PM, Ben Pfaff wrote:
> On Wed, Dec 14, 2011 at 04:21:21PM -0800, Ethan Jackson wrote:
>> > ? ? ? ?4. Have the RCU callback for flow deletion send the Netlink
>> > ? ? ? ? ? broadcast message that tells userspace that the flow is gone.
>> > ? ? ? ? ? The Netlink client
On Wed, Dec 14, 2011 at 4:12 PM, Ben Pfaff wrote:
> Why do we care? Current userspace isn't really affected. At most, we
> get flow statistics wrong by one packet (I guess multiple packets is
> theoretically possible?), if and when this happens. The place where it
> keeps coming up is in situat
Since SLB bonds are fairly risky (see vswitchd/INTERNALS), it seems
prudent to require users to explicitly configure them. This patch
defaults to active-backup bonds when no other bond_mode is
explicitly specified.
Signed-off-by: Ethan Jackson
---
vswitchd/bridge.c|2 +-
vswitchd/vswitc
Fixed according to comments from Ansis.
--8<--cut here-->8--
genl_lock is not exported from older kernel. Following patch add
genl_exec() which can run any function (passed as arg) with
genl_mutex held.
Signed-off-by: Pravin B Shelar
---
datapath
On Wed, Dec 14, 2011 at 5:16 PM, Jesse Gross wrote:
> On Wed, Dec 14, 2011 at 4:48 PM, Pravin Shelar wrote:
>> On Wed, Dec 14, 2011 at 4:37 PM, Jesse Gross wrote:
>>> On Wed, Dec 14, 2011 at 3:47 PM, Pravin Shelar wrote:
On Wed, Dec 14, 2011 at 3:34 PM, Jesse Gross wrote:
> On Wed, De
On Thu, Dec 8, 2011 at 2:01 PM, Ben Pfaff wrote:
> An upcoming commit will add another user.
> ---
> lib/packets.c | 25 +
> lib/packets.h |2 ++
> ofproto/ofproto-dpif.c | 12
> 3 files changed, 31 insertions(+), 8 deletions(-)
>
>
On Wed, Dec 14, 2011 at 4:48 PM, Pravin Shelar wrote:
> On Wed, Dec 14, 2011 at 4:37 PM, Jesse Gross wrote:
>> On Wed, Dec 14, 2011 at 3:47 PM, Pravin Shelar wrote:
>>> On Wed, Dec 14, 2011 at 3:34 PM, Jesse Gross wrote:
On Wed, Dec 14, 2011 at 3:07 PM, Pravin Shelar wrote:
> On Wed,
I will be happy if you can reply
me back for i have something
very important to tell you, Ms Vannessa Nuova grafica e nuove funzionalità! Crea subito Gratis la tua nuova Casella di Posta Katamail
___
dev mailing list
dev@openvswitch.org
http://op
On Wed, Dec 14, 2011 at 4:37 PM, Jesse Gross wrote:
> On Wed, Dec 14, 2011 at 3:47 PM, Pravin Shelar wrote:
>> On Wed, Dec 14, 2011 at 3:34 PM, Jesse Gross wrote:
>>> On Wed, Dec 14, 2011 at 3:07 PM, Pravin Shelar wrote:
On Wed, Dec 14, 2011 at 2:47 PM, Ansis Atteka wrote:
> On Wed, D
On Wed, Dec 14, 2011 at 3:47 PM, Pravin Shelar wrote:
> On Wed, Dec 14, 2011 at 3:34 PM, Jesse Gross wrote:
>> On Wed, Dec 14, 2011 at 3:07 PM, Pravin Shelar wrote:
>>> On Wed, Dec 14, 2011 at 2:47 PM, Ansis Atteka wrote:
On Wed, Dec 14, 2011 at 2:08 PM, Pravin B Shelar
wrote:
>
On Wed, Dec 14, 2011 at 04:21:21PM -0800, Ethan Jackson wrote:
> > ? ? ? ?4. Have the RCU callback for flow deletion send the Netlink
> > ? ? ? ? ? broadcast message that tells userspace that the flow is gone.
> > ? ? ? ? ? The Netlink client that sent the actual deletion request
> > ? ? ? ? ? woul
> 4. Have the RCU callback for flow deletion send the Netlink
> broadcast message that tells userspace that the flow is gone.
> The Netlink client that sent the actual deletion request
> would still get a synchronous response, but the broadcast
> would
Currently, a Netlink operation that deletes a flow can send its response
to userspace before the flow is completely "gone", because another CPU
can see the flow just before the deletion becomes visible on that CPU
and still be processing its actions afterward.
I don't know if this is actually a sc
On Wed, Dec 14, 2011 at 3:34 PM, Jesse Gross wrote:
> On Wed, Dec 14, 2011 at 3:07 PM, Pravin Shelar wrote:
>> On Wed, Dec 14, 2011 at 2:47 PM, Ansis Atteka wrote:
>>> On Wed, Dec 14, 2011 at 2:08 PM, Pravin B Shelar wrote:
+int genl_exec(genl_exec_func_t func, void *data)
+{
+
On Wed, Dec 14, 2011 at 3:07 PM, Pravin Shelar wrote:
> On Wed, Dec 14, 2011 at 2:47 PM, Ansis Atteka wrote:
>> On Wed, Dec 14, 2011 at 2:08 PM, Pravin B Shelar wrote:
>>> +int genl_exec(genl_exec_func_t func, void *data)
>>> +{
>>> + struct sk_buff *skb;
>>> +
>>> + skb = genlmsg_ne
On Wed, Dec 14, 2011 at 2:47 PM, Ansis Atteka wrote:
>
> On Wed, Dec 14, 2011 at 2:08 PM, Pravin B Shelar wrote:
>>
>> genl_lock is not exported from older kernel. Following patch add
>> genl_exec() which can run any function (passed as arg) with
>> genl_lock held.
>>
>> Signed-off-by: Pravin B S
> + tag_set_add(&o->revalidate_set, e->tag);
> + mac_learning_expire (o->ml, e);
Redundant space in this function call.
As I read this, I was wondering if we could somehow expose the
all_ofprotos map in ofproto.c to providers by implementing s
Fixed according to comments from Jesse.
- fixed to make it consistent with upstream version.
--8<--cut here-->8--
Following patch introduces a timer based event to rehash flow-hash
table. It makes finding collisions difficult to for an attac
Signed-off-by: Pravin B Shelar
---
datapath/datapath.c |9 +-
datapath/linux/Modules.mk |3 +-
datapath/linux/compat/include/linux/workqueue.h | 74 ++---
datapath/linux/compat/workqueue.c | 213 +++
genl_lock is not exported from older kernel. Following patch add
genl_exec() which can run any function (passed as arg) with
genl_lock held.
Signed-off-by: Pravin B Shelar
---
datapath/datapath.c |9 ++-
datapath/linux/compat/genetlink.inc | 102 +
All your comments sound reasonable.
Ethan
On Wed, Dec 14, 2011 at 09:25, Ben Pfaff wrote:
> On Tue, Dec 13, 2011 at 03:18:17PM -0800, Ethan Jackson wrote:
>> Why are min and max args 'int's instead of 'size_t's or 'unsigned'?
>
> I was drawing an analogy to main(), which has an "int" for argc.
>
> ETH_TOTAL_MIN isn't really very useful in practice because it only
> applies to packets that go over a physical Ethernet (or some
> semi-accurate virtualization thereof). I'm not even sure that all
> physical media that we call "Ethernet" even has this requirement (is
> there any reason for it o
This makes it possible to add entries for decoding OpenFlow messages with
newer versions, e.g. OpenFlow 1.1 or 1.2. However, no actual messages for
newer versions are actually implemented yet; that will come later.
---
include/openflow/openflow.h |1 +
lib/ofp-util.c | 112 +
Most structures in this file have an "nx_" prefix, so this makes naming
more consistent.
---
include/openflow/nicira-ext.h | 10 +-
lib/ofp-print.c |4 ++--
lib/ofp-util.c|8
ofproto/ofproto.c |8
4 files changed, 15
I think that these patches are now ready for review to be checked in.
Compared to the start of the "of1.1 draft" series that I posted earlier,
"ofp-util: Extend message decoding..." is new and "Better abstract..."
now unit tests the OF1.1 error decoding support.
Ben Pfaff (3):
nicira-ext: Rename
Do you need a loan? CONTACT us with amount needed and duration___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Tue, Dec 13, 2011 at 03:41:54PM -0800, Ethan Jackson wrote:
> > + ? ?if (packet->size < ETH_HEADER_LEN) {
>
> In theory, I think we would want ETH_TOTAL_MIN here. However, that
> could potentially be annoying for it's only user ofproto/trace. So I
> think it's fine.
Thank you for the review.
On Tue, Dec 13, 2011 at 03:18:17PM -0800, Ethan Jackson wrote:
> Why are min and max args 'int's instead of 'size_t's or 'unsigned'?
I was drawing an analogy to main(), which has an "int" for argc.
> In ofproto/trace, you have a case which expects argc == 6, but you
> only allow a maximum of 4 (+
28 matches
Mail list logo
