On Sat, Dec 19, 2015 at 3:07 AM, William Tu wrote:
> Hi Ben,
>
> Thank you for the feedback.
> About the wrong free in tun-metadata.c, which one do you mean?
It was this one (from the original patch):
> --- a/lib/tun-metadata.c
> +++ b/lib/tun-metadata.c
> @@ -596,6 +596,7 @@ tun_metadata_add_e
On Sun, Dec 20, 2015 at 1:21 AM, Pravin B Shelar wrote:
> The bug was reported by Joe Stringer.
>
> Signed-off-by: Pravin B Shelar
Acked-by: Jesse Gross
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Sat, Dec 19, 2015 at 10:19 PM, Pravin B Shelar wrote:
> STT device is added to the device list at device create time. and
> the dev socket is initialized when dev is UP. So avoid accessing
> stt socket while searching a device.
>
> ---8<---
> IP: [] nf_ip_hook+0xfd/0x180 [openvswitch]
> Oops: 0
On Sun, Dec 20, 2015 at 3:57 AM, Wanlong Gao wrote:
> Backport from the upstream commit, this commit is necessary for using
> 2.3.x on RHEL with updated kernel, and absolutely more sane than the kernel
> version adjudgement:
> commit 572e54faff70760f4bfe4a7cffc9553e6743199e
Thanks, cherry
On Sun, Dec 20, 2015 at 1:21 AM, Pravin B Shelar wrote:
> Keep devices in UP state in stt-list. So that devices in UP state
> would be searched on a packet receive in STT. This way devices in
> DOWN state would not accept any packets.
>
> Signed-off-by: Pravin B Shelar
I think this works but it
On Sun, Dec 20, 2015 at 5:48 PM, Pravin Shelar wrote:
> On Sun, Dec 20, 2015 at 10:31 AM, Jesse Gross wrote:
>> One other thing that I noticed is that the STT fragment hash table
>> commingles fragments that were addressed to different destination
>> ports. This is unlike
es on packet receive.
>
> Signed-off-by: Pravin B Shelar
Acked-by: Jesse Gross
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Tue, Dec 22, 2015 at 12:44 PM, William Tu wrote:
> Found by valgrind, test case 643.
>
> Signed-off-by: William Tu
Thanks, applied to master and branch-2.5.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Mon, Dec 21, 2015 at 6:21 PM, Ben Pfaff wrote:
> On Mon, Dec 14, 2015 at 05:23:36PM -0800, Daniele Di Proietto wrote:
>> In the ODP context an empty mask netlink attribute usually means that
>> the flow should be an exact match.
>>
>> odp_flow_key_to_mask() instead returns a struct flow_wildcar
On Tue, Dec 22, 2015 at 7:13 PM, Daniele Di Proietto
wrote:
>
>
> On 17/12/2015 16:28, "Jesse Gross" wrote:
>
>>When handling an upcall with the userspace datapath, it's currently
>>possible for a flow from a packet with no tunnel options to come back
+mask->masks.in_port.odp_port = u32_to_odp(UINT32_MAX);
My guess is that this is no longer necessary (and it's gone on
master). However, I would understand if you want to keep it for the
backport since there is a possibility that it has some other side
effect that
On Wed, Jan 6, 2016 at 12:59 PM, David Miller wrote:
> From: David Wragg
> Date: Wed, 6 Jan 2016 13:33:04 +
>
>> Prior to 4.3, openvswitch vxlan vports could transmit vxlan packets of
>> any size, constrained only by the ability to transmit the resulting
>> UDP packets. 4.3 introduced vxlan
On Wed, Jan 6, 2016 at 3:25 PM, David Wragg wrote:
> David Miller writes:
>>> Prior to 4.3, openvswitch vxlan vports could transmit vxlan packets of
>>> any size, constrained only by the ability to transmit the resulting
>>> UDP packets. 4.3 introduced vxlan netdevs corresponding to vxlan
>>> vp
On Wed, Jan 6, 2016 at 4:14 PM, Hannes Frederic Sowa
wrote:
> Hi,
>
>
> On 07.01.2016 00:57, Jesse Gross wrote:
>>
>> On Wed, Jan 6, 2016 at 3:25 PM, David Wragg wrote:
>>>
>>> David Miller writes:
>>>>>
>>>>> Prior to 4.3,
On Wed, Jan 6, 2016 at 4:29 PM, David Wragg wrote:
> Jesse Gross writes:
>> On Wed, Jan 6, 2016 at 3:25 PM, David Wragg wrote:
>>> I'm certainly open to suggestions of better ways to solve the problem.
>>
>> One option is to simply set the MTU on the device f
On Tue, Jan 12, 2016 at 11:12 PM, Alexandru Ardelean
wrote:
> On Tue, Jan 12, 2016 at 9:25 PM, Alexandru Ardelean
> wrote:
>
>> On Tue, Jan 12, 2016 at 6:23 PM, Ben Pfaff wrote:
>>
>>> On Tue, Jan 12, 2016 at 09:26:35AM +0200, Alexandru Ardelean wrote:
>>> > But as it turns out, OVS is getting b
On Wed, Jan 13, 2016 at 3:51 PM, Ramu Ramamurthy
wrote:
>
> Problem:
> -
>
> We are running centos 7.2 (ovs-kernel module from distribution) with OVS
> 2.3.1.
>
> Large tcp frames (> 30Kbytes gro'd) are coming into a vxlan-tunnel port and
> switched by an ovs-bridge onto a veth port
>
> vx
The most recent version of the Geneve draft included an option
class assignment for OVN:
https://tools.ietf.org/html/draft-ietf-nvo3-geneve-01
As a result, we can stop using the experimental class and switch to
the allocated one (0x0102).
Signed-off-by: Jesse Gross
---
I plan on applying this
On Thu, Jan 14, 2016 at 2:35 PM, Ben Pfaff wrote:
> On Thu, Jan 14, 2016 at 02:31:14PM -0800, Jesse Gross wrote:
>> The most recent version of the Geneve draft included an option
>> class assignment for OVN:
>> https://tools.ietf.org/html/draft-ietf-nvo3-geneve-01
>>
On Thu, Jan 14, 2016 at 3:29 AM, Haggai Eran wrote:
> These two patches allow passing tunnelled flows without a match on the
> outer header TTL to the datapath. The first patch simply modifies ovs-dpctl
> to pass along fields that don't have a mask set. The second patch changes
> the behavior of o
On Sun, Jan 10, 2016 at 11:18 PM, Pravin B Shelar wrote:
> new file mode 100644
> index 000..755b6c1
> --- /dev/null
> +++ b/lib/netdev-vport-private.h
[...]
> +struct vport_class {
> +const char *dpif_port;
> +struct netdev_class netdev_class;
> +};
Is this used anywhere outside of n
On Sun, Jan 10, 2016 at 11:18 PM, Pravin B Shelar wrote:
> diff --git a/lib/packets.c b/lib/packets.c
> index d82341d..a910f50 100644
> --- a/lib/packets.c
> +++ b/lib/packets.c
> +void
> +IP_ECN_set_ce(struct dp_packet *pkt, bool is_ipv6)
> +{
> +if (is_ipv6) {
> +struct ovs_16aligned
On Sun, Jan 10, 2016 at 11:18 PM, Pravin B Shelar wrote:
> Used by STT tunnel.
>
> Signed-off-by: Pravin B Shelar
Acked-by: Jesse Gross
In general, I think it would be good to expand the commit messages in
this series. Once this is checked in, "Used by STT tunnel" w
On Thu, Jan 21, 2016 at 1:45 PM, pravin shelar wrote:
> On Thu, Jan 21, 2016 at 12:56 PM, Jesse Gross wrote:
>> On Sun, Jan 10, 2016 at 11:18 PM, Pravin B Shelar wrote:
>>> new file mode 100644
>>> index 000..755b6c1
>>> --- /dev/null
On Thu, Jan 21, 2016 at 1:50 PM, pravin shelar wrote:
> On Thu, Jan 21, 2016 at 1:29 PM, Jesse Gross wrote:
>> What about checking to see if the packet is originally ECT or even IP?
>> It doesn't inherently have to be done in this function but skipping
>> ahead to the
On Sun, Jan 10, 2016 at 11:18 PM, Pravin B Shelar wrote:
> diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
> index c746cc2..215e9b6 100644
> --- a/lib/dpif-netdev.c
> +++ b/lib/dpif-netdev.c
> @@ -3529,8 +3528,12 @@ dp_execute_cb(void *aux_, struct dp_packet **packets,
> int cnt,
>
On Sun, Jan 10, 2016 at 11:18 PM, Pravin B Shelar wrote:
> Used by STT.
>
> Signed-off-by: Pravin B Shelar
This looks fine but I wonder if it is better to do any necessary
initialization at the time of the first port creation. The STT data
structures are only used in the event that userspace STT
On Sun, Jan 10, 2016 at 11:18 PM, Pravin B Shelar wrote:
> This scratchpad can be used by any layer to keep private data.
> STT will use it for TCP reassembly state.
>
> Signed-off-by: Pravin B Shelar
Acked-by: Jesse Gross
___
dev mail
On Thu, Jan 21, 2016 at 11:05 PM, pravin shelar wrote:
> On Thu, Jan 21, 2016 at 3:41 PM, Jesse Gross wrote:
>> On Sun, Jan 10, 2016 at 11:18 PM, Pravin B Shelar wrote:
>>> diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
>>> index c746cc2..215e9b6 100644
>>&g
On Sun, Jan 10, 2016 at 11:18 PM, Pravin B Shelar wrote:
> diff --git a/lib/tnl-ports.c b/lib/tnl-ports.c
> index e7f2066..13d114d 100644
> --- a/lib/tnl-ports.c
> +++ b/lib/tnl-ports.c
> @@ -95,24 +96,28 @@ tnl_port_init_flow(struct flow *flow, struct eth_addr mac,
> -if (udp_port) {
> -
On Sun, Jan 10, 2016 at 11:18 PM, Pravin B Shelar wrote:
> diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
> index cd72e62..c746cc2 100644
> --- a/lib/dpif-netdev.c
> +++ b/lib/dpif-netdev.c
> +static inline unsigned long long
> +cycles_counter(void)
> +{
> +#ifdef DPDK_NETDEV
> +return rte
On Sun, Jan 10, 2016 at 11:18 PM, Pravin B Shelar wrote:
> This patch used userpsace tunneling mechanism for implementing
> STT tunneling protocol.
>
> Signed-off-by: Pravin B Shelar
It looks like this doesn't support segmentation either on transmit
(which would only happen here if the incoming
On Sun, Jan 24, 2016 at 9:24 PM, Han Zhou wrote:
> I went into the "bad key length" in below datapath flow in a test
> environment:
>
> recirc_id(0x109),in_port(5),ct_state(-new+est-rel-inv+trk),eth(src=fa:16:3e:00:49:66,dst=fa:16:3e:ca:4a:20),eth_type(0x0800),ipv4(src=
> 22.22.22.4/255.255.255.25
On Tue, Jan 19, 2016 at 10:15 PM, Simon Horman
wrote:
> Add support for layer 3 GRE vports (non-tap aka non-VTEP).
>
> This makes use of a separate vport type for GRE, rather than a new mode for
> the existing (tap/VTEP) GRE vports as this fits more naturally with the
> kernel where implementation
On Wed, Jan 27, 2016 at 4:36 PM, pravin shelar wrote:
> On Mon, Jan 25, 2016 at 9:34 AM, Jesse Gross wrote:
>> I'm also not sure that it is really safe to use the raw TSC in all
>> cases. It's possible that DPDK support is compiled in but we aren't
>> usin
On Thu, Jan 28, 2016 at 2:10 PM, pravin shelar wrote:
> On Wed, Jan 27, 2016 at 6:06 PM, Jesse Gross wrote:
>> On Wed, Jan 27, 2016 at 4:36 PM, pravin shelar wrote:
>>> On Mon, Jan 25, 2016 at 9:34 AM, Jesse Gross wrote:
>> Direct use of the TSC has a lot of nasty cases
On Mon, Feb 1, 2016 at 8:02 PM, zhangyali (D) wrote:
> Hi all,
>
> I have noticed that OVS support multiple tunneling protocols, such as, GRE,
> VXLAN, etc. I am not sure if the MPLS has been also supported now (could not
> find it ...). Since the openflow protocol has support the corresponding
On Mon, Feb 1, 2016 at 8:05 AM, Simon Horman wrote:
> On Tue, Jan 26, 2016 at 10:09:28PM -0800, Jesse Gross wrote:
>> On Tue, Jan 19, 2016 at 10:15 PM, Simon Horman
>> wrote:
>> > Add support for layer 3 GRE vports (non-tap aka non-VTEP).
>> >
>> > This
OpenFlow
> actions=pop_mpls:0x0800,push_vlan:0x8100,set_field:4100->vlan_vid,set_field:b6:ee:28:59:97:18->eth_src,set_field:fa:16:3e:a3:03:b8->eth_dst,output:2
>
> The rule corresponding to the last action, is:
>
> # ovs-ofctl dump-flows br-mpls |grep mo
On Thu, Feb 4, 2016 at 6:22 PM, Daniele Di Proietto
wrote:
> The init script 'ovs-ctl' starts both the database and vswitchd.
> The DPDK parameters should be written by the user in the database
> with ovs-vsctl, after the database is started, but before vswitchd is.
> Currently there's no chance t
On Thu, Feb 4, 2016 at 7:55 PM, Pravin B Shelar wrote:
> There are multiple issues in IPv6 userspace tunnel
> implementation. There is not enough time to get all
> fixes in branch-2.5. So it make sense to disable the
> support on 2.5.
>
> Signed-off-by: Pravin B Shelar
This will also affect the
On Wed, Feb 10, 2016 at 12:41 PM, David Wragg wrote:
> Tom Herbert writes:
>> The correct thing to do is determine the maximum amount of
>> encapsulation overhead that can ever be set in a packet and use for
>> setting the MTU. For instance, when RCO is enable in GUE, the size of
>> the option is
On Wed, Feb 10, 2016 at 3:21 PM, Tom Herbert wrote:
> On Wed, Feb 10, 2016 at 12:59 PM, Jesse Gross wrote:
>> On Wed, Feb 10, 2016 at 12:41 PM, David Wragg wrote:
>>> Tom Herbert writes:
>>>> The correct thing to do is determine the maximum amount of
>>>
On Tue, Feb 16, 2016 at 4:33 AM, David Wragg wrote:
> diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
> index 028e387..f57f9bc 100644
> --- a/drivers/net/geneve.c
> +++ b/drivers/net/geneve.c
> @@ -37,6 +37,9 @@ MODULE_PARM_DESC(log_ecn_error, "Log packets received with
> corrupted ECN")
On Tue, Feb 16, 2016 at 10:43 AM, Joe Stringer wrote:
> diff --git a/datapath/linux/compat/lisp.c b/datapath/linux/compat/lisp.c
> index e5a6a7fe00a4..0c81dd52ecab 100644
> --- a/datapath/linux/compat/lisp.c
> +++ b/datapath/linux/compat/lisp.c
> @@ -114,6 +114,7 @@ struct lisphdr {
> };
>
> #de
On Tue, Feb 16, 2016 at 10:43 AM, Joe Stringer wrote:
> diff --git a/datapath/linux/compat/include/net/stt.h
> b/datapath/linux/compat/include/net/stt.h
> index 28d4dc53c061..aa55c9a0a2a5 100644
> --- a/datapath/linux/compat/include/net/stt.h
> +++ b/datapath/linux/compat/include/net/stt.h
[...]
to disable the support on 2.5.
>
> Signed-off-by: Pravin B Shelar
Acked-by: Jesse Gross
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Fri, Feb 12, 2016 at 11:25 AM, Simon Horman
wrote:
> If an skb was not MPLS initially then it may be GSO and in that case if it
> became MPLS then GSO can't be performed because both MPLS and tunnels make
> use of the inner_protocol field of struct skbuff in order to allow GSO to
> be performed
In addition, there have been numerous revisions of a previous patch
series that implements this. I believe that this is the most recent
one:
http://openvswitch.org/pipermail/dev/2015-November/061861.html
Please work with that author instead of starting from scratch.
On Wed, Feb 17, 2016 at 6:49 A
On Wed, Feb 17, 2016 at 10:26 AM, Joe Stringer wrote:
> On 16 February 2016 at 12:35, Joe Stringer wrote:
>> On 16 February 2016 at 11:43, Jesse Gross wrote:
>>> On Tue, Feb 16, 2016 at 10:43 AM, Joe Stringer wrote:
>>>> diff --git a/datapath/linux/compat/lisp.c
ip_tunnel_change_mtu
> - Exclude GENEVE options from max MTU calculation
>
> Signed-off-by: David Wragg
Acked-by: Jesse Gross
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Thu, Feb 18, 2016 at 11:59 PM, Simon Horman
wrote:
> On Tue, Feb 16, 2016 at 02:53:39PM -0800, Jesse Gross wrote:
>> On Fri, Feb 12, 2016 at 11:25 AM, Simon Horman
>> wrote:
>> > If an skb was not MPLS initially then it may be GSO and in that case if it
>> &
in practice they will
> handle any size packet that the underlying devices support sending.
Acked-by: Jesse Gross
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Mon, Mar 9, 2015 at 3:12 PM, Pravin B Shelar wrote:
> openvswitch_headers = \
> diff --git a/datapath/linux/.gitignore b/datapath/linux/.gitignore
> index 69d6658..1d1aae7 100644
> --- a/datapath/linux/.gitignore
> +++ b/datapath/linux/.gitignore
> @@ -50,6 +50,7 @@
> /vport-lisp.c
> /vport-
point in
time and the VXLAN spec (where the main problem lies) recommends
against them. Therefore, this is considered to be an advanced user
feature and we settle for just documenting the issue.
Signed-off-by: Jesse Gross
---
FAQ.md | 9 +
NEWS | 1 +
lib/n
On Tue, Mar 24, 2015 at 1:21 PM, Pravin Shelar wrote:
> On Mon, Mar 23, 2015 at 12:23 PM, Jesse Gross wrote:
>> On Mon, Mar 9, 2015 at 3:12 PM, Pravin B Shelar wrote:
>>> diff --git a/datapath/linux/compat/stt.c b/datapath/linux/compat/stt.c
>>> new file mode 100644
Thanks for the quick review; I applied this to master.
On Tue, Mar 24, 2015 at 2:56 PM, Pritesh Kothari (pritkoth)
wrote:
> LGTM
>
> Acked-by: Pritesh Kothari
>
> Regards,
> Pritesh
>
> On Mar 24, 2015, at 12:38 PM, Jesse Gross wrote:
>
> The kernel module
e commit) was in Linux before
> 2.6.32, so it seems unlikely that we need any compatibility code there.
>
> CC: Jonathan Vestin
> Signed-off-by: Ben Pfaff
Acked-by: Jesse Gross
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Thu, Mar 26, 2015 at 6:35 PM, Pravin Shelar wrote:
> On Tue, Mar 24, 2015 at 12:58 PM, Jesse Gross wrote:
>> On Tue, Mar 24, 2015 at 1:21 PM, Pravin Shelar wrote:
>>> On Mon, Mar 23, 2015 at 12:23 PM, Jesse Gross wrote:
>>>> On Mon, Mar 9, 2015 at 3
sock_create (owned by udp_tunnel)
>
> Fixes: eb6eebd28 ("datapath: Account for "udp: Add udp_sock_create for UDP
> tunnels to open listener socket")
> Signed-off-by: Thomas Graf
> Cc: Jesse Gross
Acked-by: Jesse Gross
__
On Sun, Mar 29, 2015 at 9:27 PM, yinpeijun wrote:
> From: caochengrong
>
> tcp flows with gso between two VMs in diffrent host,
> go through vxlan tunnel, cause kernel crash.
>
> Signed-off-by: caochengrong
> Signed-off-by: Arika Chen
What OVS and host kernel version is this for? I don't think
decide whether to calculate the
checksum.
Signed-off-by: Jesse Gross
---
lib/netdev-vport.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/netdev-vport.c b/lib/netdev-vport.c
index a9639d3..0e0d791 100644
--- a/lib/netdev-vport.c
+++ b/lib/netdev-vport.c
@@ -1040,7
The IP TTL is currently omitted in the extracted tunnel information
that is stored in the flow for userspace tunneling. This includes it
so that the same logic used by the kernel also applies.
Signed-off-by: Jesse Gross
---
lib/netdev-vport.c | 1 +
1 file changed, 1 insertion(+)
diff --git a
: Jesse Gross
---
lib/netdev-vport.c | 6 ++
lib/odp-util.c | 8
2 files changed, 6 insertions(+), 8 deletions(-)
diff --git a/lib/netdev-vport.c b/lib/netdev-vport.c
index 1ee68bc..a9639d3 100644
--- a/lib/netdev-vport.c
+++ b/lib/netdev-vport.c
@@ -996,10 +996,8
As OVS adds userspace support for being the endpoint in protocols
like tunnels, it will need to be able to calculate pseudoheaders
as part of the checksum calculation.
Signed-off-by: Jesse Gross
---
lib/packets.c | 14 ++
lib/packets.h | 1 +
tests/library.at | 2
Currently, the userspace VXLAN implementation contains the code
for generating and parsing both the UDP and VXLAN headers. This
pulls out the UDP portion for better layering and to make it
easier to support additional UDP based tunnels and features.
Signed-off-by: Jesse Gross
---
lib/netdev
well.
Signed-off-by: Jesse Gross
---
NEWS | 4 +-
lib/netdev-vport.c | 114 ++-
lib/odp-util.c | 28 ++--
lib/packets.h| 19
tests/odp.at | 1 +
tests/tunnel-push-pop.at
Kernel based OVS recently added the ability to support checksums
for UDP based tunnels (Geneve and VXLAN). This adds similar support
for the userspace datapath to bring feature parity.
Signed-off-by: Jesse Gross
---
lib/netdev-vport.c | 37 ++---
lib/odp
On receive, the userspace GRE code doesn't check the protocol
field. Since OVS only understands Ethernet packets, this adds a
check that the inner protocol is Ethernet and discards other types
of packets.
Signed-off-by: Jesse Gross
---
lib/netdev-vport.c | 4
1 file changed, 4 inser
Currently when printing a userspace tunnel action for VXLAN, the
VNI is treated as a 32 bit field rather than 24 bit. Even if this
is the representation that we use internally, we should still show
the right VNI to avoid confusing people.
Signed-off-by: Jesse Gross
---
lib/odp-util.c
r=0, crc=0
+])
+
OVS_VSWITCHD_STOP
AT_CLEANUP
On Mon, Mar 30, 2015 at 3:42 PM, Pritesh Kothari (pritkoth)
wrote:
> should a test case be added to catch this ?
>
> Regards,
> Pritesh
>
>> On Mar 30, 2015, at 3:14 PM, Jesse Gross wrote:
>&g
inner protocol fields that we have backported.
Without this, a kernel crash can be triggered with tunnel
packets that do not have enough headroom and need to be
segmented.
Reported-by: Yinpeijun
Signed-off-by: Jesse Gross
---
datapath/linux/compat/gso.h | 21 +
1 file
On Mon, Mar 30, 2015 at 7:57 PM, Yinpeijun wrote:
>>> tcp flows with gso between two VMs in diffrent host, go through vxlan
>>> tunnel, cause kernel crash.
>>>
>>> Signed-off-by: caochengrong
>>> Signed-off-by: Arika Chen
>>
>>What OVS and host kernel version is this for? I don't think this shou
On Tue, Mar 31, 2015 at 11:02 AM, Pravin Shelar wrote:
> On Tue, Mar 31, 2015 at 9:34 AM, Jesse Gross wrote:
>> skb protocol offsets are relative to the beginning of the
>> buffer and therefore must be updated if the buffer size is
>> expanded. Kernel functions do this autom
.
Reported-by: Yinpeijun
Signed-off-by: Jesse Gross
---
v2: Use existing backported skb offset APIs to avoid introducing
additional version checks.
---
datapath/linux/compat/gso.h | 40
1 file changed, 40 insertions(+)
diff --git a/datapath/linux/compat/gso.h b
On Tue, Mar 31, 2015 at 1:12 PM, Pravin Shelar wrote:
> On Tue, Mar 31, 2015 at 11:44 AM, Jesse Gross wrote:
>> On Tue, Mar 31, 2015 at 11:02 AM, Pravin Shelar wrote:
>>> On Tue, Mar 31, 2015 at 9:34 AM, Jesse Gross wrote:
>>>> skb protocol offsets are re
On Thu, Mar 26, 2015 at 4:41 PM, Pravin B Shelar wrote:
> diff --git a/datapath/datapath.c b/datapath/datapath.c
> index 7f431ed..ea9c6ae 100644
> --- a/datapath/datapath.c
> +++ b/datapath/datapath.c
> @@ -2192,6 +2192,7 @@ static int __net_init ovs_init_net(struct net *net)
>
> INIT_LIST
On Tue, Mar 31, 2015 at 4:21 PM, Ben Pfaff wrote:
> On Mon, Mar 30, 2015 at 03:14:46PM -0700, Jesse Gross wrote:
>> The GRE checksum is a 16 bit field stored in a 32 bit option (the
>> rest is reserved). The current code treats the checksum as a 32-bit
>> field and places
On Tue, Mar 31, 2015 at 4:34 PM, Pravin Shelar wrote:
> On Tue, Mar 31, 2015 at 1:18 PM, Jesse Gross wrote:
>> On Tue, Mar 31, 2015 at 1:12 PM, Pravin Shelar wrote:
>>> On Tue, Mar 31, 2015 at 11:44 AM, Jesse Gross wrote:
>>>> On Tue, Mar 31, 2015 at 11:02 AM,
.
Reported-by: Yinpeijun
Signed-off-by: Jesse Gross
---
v2: Use existing backported skb offset APIs to avoid introducing
additional version checks.
v3: Move macro into skbuff.h to ensure that it rewrites all calls
of pskb_expand_head().
---
datapath/linux/compat/gso.h | 17
On Thu, Apr 2, 2015 at 12:18 PM, Pravin Shelar wrote:
> On Tue, Mar 31, 2015 at 5:07 PM, Jesse Gross wrote:
>> On Thu, Mar 26, 2015 at 4:41 PM, Pravin B Shelar wrote:
>>> diff --git a/datapath/datapath.c b/datapath/datapath.c
>>> index 7f431ed..ea9c6ae 100644
&g
On Wed, Apr 1, 2015 at 10:14 PM, Neil McKee wrote:
> I've been looking at filling in the sFlow structures to report on tunnel
> encap and other transformations. sFlow sampling is best done on ingress
> only, so I can't use the egress-sampling action that the IPFIX
> implementation uses to get t
On Sat, Apr 4, 2015 at 8:45 AM, Neil McKee wrote:
>
> On Fri, Apr 3, 2015 at 3:14 PM, Jesse Gross wrote:
>>
>> On Wed, Apr 1, 2015 at 10:14 PM, Neil McKee wrote:
>> > I've been looking at filling in the sFlow structures to report on tunnel
>> > encap an
On Tue, Apr 7, 2015 at 6:56 AM, 马啸 wrote:
> Hi,all
>
>
> I know in normal vxlan support,vteps can be in different subnets.
> But when I config the vxlan tunnel of openvswitch with the local ip and the
> remote ip in different subnets,
> I find the flow can not be matched.
Yes, it should work.
I
On Tue, Apr 7, 2015 at 8:29 AM, Thomas Graf wrote:
> On 04/07/15 at 10:20am, Russell Bryant wrote:
>> IIRC, the proposal was actually quite explicit that the tag is a VLAN
>> ID. It's not a hidden implementation detail because something (not OVN)
>> has to set up ovs inside the VM with all of the
On Tue, Apr 7, 2015 at 11:14 AM, Thomas Graf wrote:
> On 04/07/15 at 10:49am, Jesse Gross wrote:
>> On Tue, Apr 7, 2015 at 8:29 AM, Thomas Graf wrote:
>> > I remember this discussion. One alternative that comes to mind is to
>> > simply push a Geneve header in front of
On Tue, Apr 7, 2015 at 4:22 PM, Pravin Shelar wrote:
> On Mon, Mar 30, 2015 at 3:14 PM, Jesse Gross wrote:
>> diff --git a/lib/netdev-vport.c b/lib/netdev-vport.c
>> index 0c9f5a4..ef96862 100644
>> --- a/lib/netdev-vport.c
>> +++ b/lib/netdev-vport.c
>>
eyor.com/project/blp/ovs/build/1.0.140
>
> On Tue, Apr 7, 2015 at 4:22 PM, Pravin Shelar wrote:
>> On Mon, Mar 30, 2015 at 3:14 PM, Jesse Gross wrote:
>>> Kernel based OVS recently added the ability to support checksums
>>> for UDP based tunnels (Geneve and VXLAN). T
)
>
> On Tue, Apr 7, 2015 at 6:40 PM, Gurucharan Shetty wrote:
>> On Tue, Apr 7, 2015 at 6:01 PM, Jesse Gross wrote:
>>> Thanks, I pushed a fix to avoid this - nothing was using the member at
>>> the moment, so I just replaced it with a comment.
>> Thank you fo
issue that was preexisting but is
intermittently exercised by one of the new unit tests. I'm
working on a fix for that as well.
Jesse Gross (2):
packet: Avoid array of struct with zero length member.
geneve: Zero header before parsing userspace tunneling action.
lib/odp-util.c | 1 +
lib
different.
Signed-off-by: Jesse Gross
---
lib/odp-util.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/lib/odp-util.c b/lib/odp-util.c
index 7725a09..8a81f6b 100644
--- a/lib/odp-util.c
+++ b/lib/odp-util.c
@@ -921,6 +921,7 @@ ovs_parse_tnl_push(const char *s, struct
ovs_action_push_tnl
ff-by: Jesse Gross
---
lib/packets.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/packets.h b/lib/packets.h
index 29ea54f..b146a50 100644
--- a/lib/packets.h
+++ b/lib/packets.h
@@ -730,7 +730,7 @@ struct geneve_opt {
uint8_t r2:1;
uint8_t r1:1;
#
e packet captured should match the flow, but the n_packets not
> be increased.
>
>
> The VxLAN UDP can be captured , so I think 10.10.24.69 and 10.11.24.64 can
> be reached from each other.
>
> Can you give me more help? Thanks!
>
>
>
>
>
> At 2015-04-07 23:39:06, &q
ll of the packet data. This guarantees that the
packet gets dropped during the next round of processing.
Signed-off-by: Jesse Gross
---
lib/netdev.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/lib/netdev.c b/lib/netdev.c
index aed5289..790bab9 100644
--- a/lib/netdev.c
operate on single packets to enable
better code reuse.
Signed-off-by: Jesse Gross
---
lib/netdev-provider.h | 9 ++--
lib/netdev-vport.c| 140 +++---
lib/netdev.c | 34 +---
3 files changed, 62 insertions(+), 121 deletions
On Thu, Apr 9, 2015 at 11:00 PM, Pravin B Shelar wrote:
> diff --git a/NEWS b/NEWS
> index 87460a7..6e5e070 100644
> --- a/NEWS
> +++ b/NEWS
> @@ -77,6 +77,7 @@ Post-v2.3.0
> numbers. OpenFlow is 6653 and OVSDB is 6640.
> - Support for DPDK vHost.
> - Support for outer UDP checksums
On Thu, Apr 9, 2015 at 1:02 PM, Pravin Shelar wrote:
> On Wed, Apr 8, 2015 at 1:19 PM, Jesse Gross wrote:
>> The userspace tunneling API for pushing and popping tunnel headers
>> is currently based on processing batches of packets. However, there
>> is no obvious way
On Thu, Apr 9, 2015 at 2:29 PM, Pravin Shelar wrote:
> On Thu, Apr 9, 2015 at 12:22 PM, Jesse Gross wrote:
>> Upstream commit:
>> udptunnels: Call handle_offloads after inserting vlan tag.
>>
>> handle_offloads() calls skb_reset_inner_headers() to store
>&
changes the MAC header for the encapsulated
packet but it is not reflected in skb->inner_mac_header, which
breaks GSO and drivers which attempt to use this for encapsulation
offloads.
Fixes: b736a623 ("vxlan: Add tx-vlan offload support.")
Signed-off-by: Jesse Gross
.
Reported-by: Yinpeijun
Signed-off-by: Jesse Gross
---
v2: Use existing backported skb offset APIs to avoid introducing
additional version checks.
v3: Move macro into skbuff.h to ensure that it rewrites all calls
of pskb_expand_head().
v4: Rebase and add EXPORT_SYMBOL for vport module use.
Offsets
701 - 800 of 3414 matches
Mail list logo
