On Mon, Dec 22, 2014 at 6:24 PM, Pravin B Shelar wrote:
> Today vport-send has complex error handling because it involves
> freeing skb and updating stats depending on return value from
> vport send implementation.
> This can be simplified by delegating responsibility of freeing
> skb to the vport
On Mon, Dec 22, 2014 at 6:24 PM, Pravin B Shelar wrote:
> handle offload code is replicated for different tunneling protocols
> define compat function to simplify the code.
>
> Signed-off-by: Pravin B Shelar
Acked-by: Jesse Gross
___
dev mailing list
After commit 0c7812e5e (recirculation: Do not drop packet when
there is no match from internal table.), if flow keys are modified
before the recirculation action (e.g. set vlan ID), the miss
handling of recirc'ed packets may not reach the intended
'ofproto_dpif' which has rules looking up the 'reci
Thx for the review, Jarno,
I find it conceivable that post recirculation flows would be able to cross
> a later patch port in the pipeline. Therefore, rather than stating that
> “post recirculation flows are not allowed to cross a patch port”, I think
> we should say that recirculated flows must b
Assuming the notes below are addressed:
Acked-by: Jarno Rajahalme
On Dec 21, 2014, at 9:58 AM, Alex Wang wrote:
> After commit 0c7812e5e (recirculation: Do not drop packet when
> there is no match from internal table.), if flow keys are modified
> before the recirculation action (e.g. set vlan
Today vport-send has complex error handling because it involves
freeing skb and updating stats depending on return value from
vport send implementation.
This can be simplified by delegating responsibility of freeing
skb to the vport implementation for all cases. So that
vport-send needs just update
handle offload code is replicated for different tunneling protocols
define compat function to simplify the code.
Signed-off-by: Pravin B Shelar
---
Fixed according to comments from Jesse.
---
datapath/linux/compat/gre.c | 32 +
datapath/linux/compat/gso.c | 51 +++
On Mon, Dec 22, 2014 at 1:30 PM, Jesse Gross wrote:
> On Mon, Dec 22, 2014 at 3:08 PM, Pravin Shelar wrote:
>> On Mon, Dec 22, 2014 at 11:56 AM, Jesse Gross wrote:
>>> On Mon, Dec 22, 2014 at 1:25 PM, Pravin Shelar wrote:
On Sun, Dec 21, 2014 at 9:39 PM, Jesse Gross wrote:
> On Sun, D
On Mon, Dec 22, 2014 at 3:08 PM, Pravin Shelar wrote:
> On Mon, Dec 22, 2014 at 11:56 AM, Jesse Gross wrote:
>> On Mon, Dec 22, 2014 at 1:25 PM, Pravin Shelar wrote:
>>> On Sun, Dec 21, 2014 at 9:39 PM, Jesse Gross wrote:
On Sun, Dec 21, 2014 at 6:17 PM, Jesse Gross wrote:
> On Fri, D
Thx for the review,
Forgot to mention, that I'll add (both in code and in commit log) that this
will not fix the case where the recv_bridge is not the same as
recirc_bridge, and rules matching in_port are installed on recirc_bridge.
(since in that case, the in_port is set to OFPP_NONE)
Thanks,
Al
On Dec 11, 2014, at 6:40 PM, Ben Pfaff wrote:
> On Mon, Dec 08, 2014 at 02:01:34PM -0800, Jarno Rajahalme wrote:
>> OpenFlow bundle messages should be decoded and validated at the time
>> they are added to the bundle. This commit does this for flow mod and
>> port mod messages.
>>
>> Signed-of
On Dec 11, 2014, at 5:52 PM, Ben Pfaff wrote:
> On Mon, Dec 08, 2014 at 02:01:33PM -0800, Jarno Rajahalme wrote:
>> This fixes ofp_bundle_commit() in two ways:
>>
>> - Always remove the bundle regardless of the error condition.
>> - A commit of an empty bundle must be successful.
>>
>> The act
On Dec 11, 2014, at 5:51 PM, Ben Pfaff wrote:
> On Mon, Dec 08, 2014 at 02:01:32PM -0800, Jarno Rajahalme wrote:
>> Slave controllers are rejected for flow and port mod messages, so they
>> should also be rejected for bundles (that require support at least for
>> flow and port mod messages).
>>
On Dec 11, 2014, at 5:49 PM, Ben Pfaff wrote:
> On Mon, Dec 08, 2014 at 02:01:31PM -0800, Jarno Rajahalme wrote:
>> Reject bundle add messages containing messages that should not be bundled.
>>
>> Signed-off-by: Jarno Rajahalme
>
> Thanks for the patch!
>
> In ofputil_is_bundlable(), there's
On Dec 11, 2014, at 5:45 PM, Ben Pfaff wrote:
> On Mon, Dec 08, 2014 at 02:01:30PM -0800, Jarno Rajahalme wrote:
>> OpenFlow 1.4 states that the switch must reject a bundle add message
>> whose flags are different from the ones in the bundle. Do it.
>>
>> Signed-off-by: Jarno Rajahalme
>
> A
On Dec 11, 2014, at 5:45 PM, Ben Pfaff wrote:
> On Mon, Dec 08, 2014 at 02:01:29PM -0800, Jarno Rajahalme wrote:
>> A switch may optionally verify that the 'xid' of an added message is
>> the same as the 'xid' of the bundle add message itself. Do it.
>>
>> Signed-off-by: Jarno Rajahalme
>
>
On Dec 11, 2014, at 5:43 PM, Ben Pfaff wrote:
> On Mon, Dec 08, 2014 at 02:01:28PM -0800, Jarno Rajahalme wrote:
>> Use of OF 1.4 bundle messages by a controller should indicate that the
>> controller has decided to use the switch, hence make is_admitted_msg()
>> return 'true' for them.
>>
>> S
On Dec 11, 2014, at 5:43 PM, Ben Pfaff wrote:
> On Mon, Dec 08, 2014 at 02:01:27PM -0800, Jarno Rajahalme wrote:
>> Signed-off-by: Jarno Rajahalme
>
> I see 'bundles' seems basically unused so I guess this doesn't fix a
> memory leak.
>
Right.
> Acked-by: Ben Pfaff
Thanks for the review,
On Mon, Dec 22, 2014 at 11:56 AM, Jesse Gross wrote:
> On Mon, Dec 22, 2014 at 1:25 PM, Pravin Shelar wrote:
>> On Sun, Dec 21, 2014 at 9:39 PM, Jesse Gross wrote:
>>> On Sun, Dec 21, 2014 at 6:17 PM, Jesse Gross wrote:
On Fri, Dec 19, 2014 at 7:25 PM, Pravin B Shelar
wrote:
> T
On Mon, Dec 22, 2014 at 1:06 PM, Pravin Shelar wrote:
> On Sun, Dec 21, 2014 at 2:56 PM, Jesse Gross wrote:
>> On Fri, Dec 19, 2014 at 7:24 PM, Pravin B Shelar wrote:
>>> diff --git a/datapath/linux/compat/ip_tunnels_core.c
>>> b/datapath/linux/compat/ip_tunnels_core.c
>>> index e71ba4e..7606ad
On Mon, Dec 22, 2014 at 1:25 PM, Pravin Shelar wrote:
> On Sun, Dec 21, 2014 at 9:39 PM, Jesse Gross wrote:
>> On Sun, Dec 21, 2014 at 6:17 PM, Jesse Gross wrote:
>>> On Fri, Dec 19, 2014 at 7:25 PM, Pravin B Shelar wrote:
Today vport-send has complex error handling because it involves
>>>
Acked-by: Andy Zhou
On Sun, Dec 21, 2014 at 9:58 AM, Alex Wang wrote:
> After commit 0c7812e5e (recirculation: Do not drop packet when
> there is no match from internal table.), if flow keys are modified
> before the recirculation action (e.g. set vlan ID), the miss
> handling of recirc'ed packe
On Sun, Dec 21, 2014 at 9:39 PM, Jesse Gross wrote:
> On Sun, Dec 21, 2014 at 6:17 PM, Jesse Gross wrote:
>> On Fri, Dec 19, 2014 at 7:25 PM, Pravin B Shelar wrote:
>>> Today vport-send has complex error handling because it involves
>>> freeing skb and updating stats depending on return value fr
On Sun, Dec 21, 2014 at 2:56 PM, Jesse Gross wrote:
> On Fri, Dec 19, 2014 at 7:24 PM, Pravin B Shelar wrote:
>> handle offload code is replicated for different tunneling protocols
>> define compat function to simplify the code.
>>
>> Signed-off-by: Pravin B Shelar
>
> I got some compiler errors
On Sun, Dec 21, 2014 at 9:31 AM, Traynor, Kevin wrote:
> Hi,
>
> I'd like to get some feedback about using RCU on the virtio_dev structure as
> per the comment below. Comments inline.
>
> Thanks,
> Kevin.
>
>> -Original Message-
>> From: dev [mailto:dev-boun...@openvswitch.org] On Behalf
Filter and sublayer WFP objects are created when the filter attach routine
is called and the tunnel filter is initialized, and they are released
in the uninitialize part. But the release part is done indirectly, through
the FwpmEngineClose function call which releases the latter objects when
the tu
This patch was enforced by the WHCK logo testing. In order to pass the
Windows Filtering Platform tests we need to add a persistent system
provider.
Signed-off-by: Sorin Vinturis
Reported-by: Sorin Vinturis
Reported-at: https://github.com/openvswitch/ovs-issues/issues/65
---
datapath-windows/ov
Removed duplicate include of "Switch.h" header.
Removed useless usage of UNREFERENCED_PARAMETER macro for a
referenced parameter.
Added debug logs for dropped packets.
Signed-off-by: Sorin Vinturis
---
datapath-windows/ovsext/Actions.c | 8 +---
datapath-windows/ovsext/Datapath.c | 1 -
dat
Sorin Vinturis (3):
[PATCH v2 1/3] datapath-windows: Cleanup OVS
[PATCH v2 2/3] datapath-windows: Release WFP allocated objects
[PATCH v2 3/3] datapath-windows: Add a WFP system provider
___
dev mailing list
dev@openvswitch.org
http://openvswitch.or
puÊ0ÞoÙ¤P]v{MôîF< NK*VâÚÓ
qî±KÀDm¾m9CJR.c6_Ùpòýܹ¹,.K¤ÕbõbúA6,Y`g/-GÑjw¿ê-¥¾!
dp ^è#êú'?ßÐûtäVÌaâK^î[¢ÃØÍHÐ)¤ÒW
ÉW¡uH¤æýLбX¿îfJ©ôÐ'!òúÁõðtóig%.
K#åM$1§R~*Ñ
t!NØ^-¸E©Y
[nã±*±jEË
ãiC3¶^q êTíîÁh¸I$e^Ì9~fQY®[
PLûü¶![2c.
z0`5
1<æÞñedØÔ~B¿ÄÞgi¬SDî.^ÑNHáýë
Hi,
* Thomas Goirand [141125 14:15]:
> On 11/25/2014 12:32 AM, Jonathan Dupart wrote:
> > Hi,
> >
> > I am looking for a sponsor for the package openvswitch to correct bug
> > #768095.
> >
> > This bug prevents building the openvswitch kernel module with the last
> > stable kernel.
> >
> > As
Thanks a lot for sharing these minutes.
On 12/19/14 at 06:49pm, Andy Zhou wrote:
> Possible use cases of BPF in OVS Linux kernel datapath
> ===
>
> [...]
>
> 4. Using BPF to implement overall OVS kernel module functionality
>
>Alexei likes this approach
Купить, продать или сдать в аренду
готовый бизнес - ЛЕГКО! Компания "Скания
Инвест" окажет квалифицированную помощь в поиске покупателей на Ваш
бизнес. * Мы предоставим Вам правовую поддержку при оформлении сделки
33 matches
Mail list logo
