In the Windows datapath, Tx tunneling functionality is implemented by checking
if the outport in the action is a tunnel port. If so, the packet is
encapsulated and sent out on the PIF bridge for as second flow lookup.
Basically, we don't use the hypervisor's IP stack to send out a packet, and
short
-1 was used to indicate both nested and variable length attributes.
This patch introduces OVS_KEY_LEN_NESTED and OVS_KEY_LEN_VARIABLE to
tell them apart.
Refactor nlattr_set() to support ovs netlink key attributes in a more
general way.
Signed-off-by: Andy Zhou
Acked-by: Daniele Di Proietto
--
I have a couple comments not really related to this
specific patch. See below.
On Tue, Aug 12, 2014 at 10:43:35AM -0700, Daniele Di Proietto wrote:
> With this commit we move our DPDK support to 1.7.0.
> DPDK binaries (starting with dpdk 1.7.0) should be linked with --whole-archive
> to include
I've noticed that the directory eventfd_link and its
Makefile.in are in the tarball generated by 'make dist'.
However, the eventfd_link.c and eventfd_link.h are missing.
I haven't looked at anything else in this patch.
Thanks,
fbl
On Fri, Aug 08, 2014 at 01:28:55PM +0100, maryam.tahhan wrote:
>
Will do, tomorrow :-)
Jarno
> On Aug 12, 2014, at 1:58 PM, Ben Pfaff wrote:
>
> I'm happy enough with these patches. Jarno, do you want to apply
> them?
>
>> On Fri, Aug 08, 2014 at 09:13:42AM -0700, Jarno Rajahalme wrote:
>> Acked-by: Jarno Rajahalme
>>
>> Sent from my iPhone
>>
>>> On
This cleans up the dpif interface to make it more consistent with the
other dpif operations, and allows flows to be fetched in batches.
Signed-off-by: Joe Stringer
---
v2: Require callers to pass down an initialized buffer.
Remove NLM_F_ECHO flag from get request.
Rebase against master.
R
On Tue, Jul 22, 2014 at 5:06 PM, Daniele Di Proietto
wrote:
> ofpbuf_rss() is introduced to get the RSS hash from the NIC. It works only for
> packets received by DPDK (otherwise it returns 0, which tells the caller to
> compute the hash in another way)
>
> This commit also configure DPDK devices
On Tue, Jul 22, 2014 at 5:06 PM, Daniele Di Proietto
wrote:
> struct dp_netdev_flow used to have a reference counter. It has been replaced
> by
> RCU. Unfortunately RCU is not enough if we plan to hold a reference to the
> dp_netdev_flow for a long time. So this commit reintroduces reference coun
On Tue, Aug 12, 2014 at 10:43 AM, Daniele Di Proietto
wrote:
> With this commit we move our DPDK support to 1.7.0.
> DPDK binaries (starting with dpdk 1.7.0) should be linked with --whole-archive
> to include pmd drivers
>
I updated INSTALL.DPDK file and pushed patch to master.
> Signed-off-by: D
On Tue, Aug 12, 2014 at 10:43 AM, Daniele Di Proietto
wrote:
> rte_eth_tx_burst() _should_ transmit every packet that it is passed unless the
> queue is full. Nontheless some implementation of rte_eth_tx_burst (e.g.
> ixgbe_xmit_pkts_vec()) does not transmit more than a fixed number (32) of
> pack
On Tue, Aug 05, 2014 at 04:38:54PM -0700, Jarno Rajahalme wrote:
> Make the packet out and trace processing perform the same actions
> checks as flow mod processing does.
>
> This used to be the case before, but at some point these have diverged
> to perform different combinations of checks.
>
>
On Tue, Aug 05, 2014 at 04:38:53PM -0700, Jarno Rajahalme wrote:
> Signed-off-by: Jarno Rajahalme
> ---
> utilities/ovs-ofctl.8.in |2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/utilities/ovs-ofctl.8.in b/utilities/ovs-ofctl.8.in
> index 5251c53..77eddb4 100644
> ---
On Tue, Aug 05, 2014 at 04:38:52PM -0700, Jarno Rajahalme wrote:
> Meter requests should use dump/stats transaction, instead of
> transact_noreply, which caused the output to go to stderr and an error
> exit.
>
> Signed-off-by: Jarno Rajahalme
Acked-by: Ben Pfaff
___
On Mon, Aug 11, 2014 at 09:15:01AM -0700, Jarno Rajahalme wrote:
> Masked set action allows more megaflow wildcarding. Masked set action
> is now supported for all writeable key types, except for the tunnel
> key.
>
> The set tunnel action is an exception as any input tunnel info is
> cleared bef
On Mon, Aug 11, 2014 at 09:15:00AM -0700, Jarno Rajahalme wrote:
> Signed-off-by: Jarno Rajahalme
The amount of redundancy in the code at this point is starting to give
me a headache. How about a pattern like this (provided as an
incremental)?
diff --git a/lib/odp-util.c b/lib/odp-util.c
index
Hi Andy,
Two comments inline
Other than those, LGTM, but review by others might be useful :)
Acked-by: Daniele Di Proietto
On 8/12/14, 1:45 PM, "Andy Zhou" wrote:
>-1 was used to indicate both nested and variable length attributes.
>This patch introduces OVS_KEY_LEN_NESTED and OVS_KEY_LEN_V
On Mon, Aug 11, 2014 at 09:14:59AM -0700, Jarno Rajahalme wrote:
> Signed-off-by: Jarno Rajahalme
> Reviewed-by: YAMAMOTO Takashi
Acked-by: Ben Pfaff
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Mon, Aug 11, 2014 at 09:14:58AM -0700, Jarno Rajahalme wrote:
> Add a new action type OVS_ACTION_ATTR_SET_MASKED, and support for
> parsing, printing, and committing them.
>
> Masked set actions add a mask, immediately following the netlink
> attribute data, within the netlink attribute itself.
From: Christoph Jaeger
Date: Tue, 12 Aug 2014 09:27:57 +0200
> ovs_vport_alloc() bails out without freeing the memory 'vport' points to.
>
> Picked up by Coverity - CID 1230503.
>
> Fixes: 5cd667b0a4 ("openvswitch: Allow each vport to have an array of
> 'port_id's.")
> Signed-off-by: Christoph
I'm happy enough with these patches. Jarno, do you want to apply
them?
On Fri, Aug 08, 2014 at 09:13:42AM -0700, Jarno Rajahalme wrote:
> Acked-by: Jarno Rajahalme
>
> Sent from my iPhone
>
> > On Aug 7, 2014, at 6:15 PM, Daniele Di Proietto
> > wrote:
> >
> > This commit introduces the BUI
Thank you. Applied.
On Tue, Aug 12, 2014 at 11:49:03AM -0700, Alex Wang wrote:
> Looks good to me,
>
>
> On Mon, Aug 11, 2014 at 2:36 PM, Ben Pfaff wrote:
>
> > These error codes are proposed in the ONF extensibility working group as an
> > OpenFlow 1.3 extension.
> >
> > Error codes are also
-1 was used to indicate both nested and variable length attributes.
This patch introduces OVS_KEY_LEN_NESTED and OVS_KEY_LEN_VARIABLE to
tell them apart.
Refactor nlattr_set() to more more generally all ovs netlink key
attributes.
Signed-off-by: Andy Zhou
---
datapath/flow_netlink.c | 88 ++
Applied, thanks.
On Tue, Aug 12, 2014 at 08:27:51PM +, Alin Serdean wrote:
> Hi Ben,
>
> Also remember to add them under .gitignore:
> Untracked files:
> (use "git add ..." to include in what will be committed)
>
> lib/ofp-actions.inc1
> lib/ofp-actions.inc2
>
> diff --git a/l
Hi Ben,
Also remember to add them under .gitignore:
Untracked files:
(use "git add ..." to include in what will be committed)
lib/ofp-actions.inc1
lib/ofp-actions.inc2
diff --git a/lib/.gitignore b/lib/.gitignore
index 2d07244..32c1e4c 100644
--- a/lib/.gitignore
+++ b/lib/.git
Applied, thanks!
On Tue, Aug 12, 2014 at 11:36:25AM -0700, Alex Wang wrote:
> Looks good to me,
>
> Acked-by: Alex Wang
>
>
> On Mon, Aug 4, 2014 at 1:02 PM, Ben Pfaff wrote:
>
> > On Wed, Jul 16, 2014 at 03:00:15PM -0700, Ben Pfaff wrote:
> > > The controller setup in my personal test envir
On Tue, Aug 12, 2014 at 11:36:27AM -0700, Ankur Sharma wrote:
> Added saurabh's fix to not to include some header files
> in lib/netlink-protocol.h not needed by windows driver.
>
> This is a temporary fix to make sure that windows driver can compile.
>
> We are working in parallel on adding netl
Thank you for the review. I'll apply this to master in a minute.
On Tue, Aug 12, 2014 at 06:50:21PM +, Ankur Sharma wrote:
> Hi Ben,
>
> Changes look good to me.
>
> Regards,
> Ankur
>
> From: Ben Pfaff
> Sent: Tuesday, August 12, 2014 11:45 AM
> To:
Thank you for the reviews. I'll apply this to master in a minute.
On Tue, Aug 12, 2014 at 07:01:53PM +, Alin Serdean wrote:
> Thanks for the quick fix Ben, I also tested it just to make sure everything
> is in place.
>
> Alin.
>
> -Mesaj original-
> De la: Ben Pfaff [mailto:bpf...@
Acked-by : Alin Serdean
-Mesaj original-
De la: Ben Pfaff [mailto:b...@nicira.com]
Trimis: Tuesday, August 12, 2014 9:44 PM
Către: dev@openvswitch.org
Cc: Ben Pfaff; Alin Serdean
Subiect: [PATCH 1/2] ofp-actions: Make struct ofpact constant size across
implementations.
Before commit c2
Thanks for the quick fix Ben, I also tested it just to make sure everything is
in place.
Alin.
-Mesaj original-
De la: Ben Pfaff [mailto:bpf...@vmware.com]
Trimis: Tuesday, August 12, 2014 9:45 PM
Către: Alin Serdean
Cc: dev@openvswitch.org
Subiect: Re: WINDOWS userspace build broken
O
Hi Ben,
Changes look good to me.
Regards,
Ankur
From: Ben Pfaff
Sent: Tuesday, August 12, 2014 11:45 AM
To: Ankur Sharma
Cc: Alin Serdean; dev@openvswitch.org
Subject: Re: WINDOWS userspace build broken
On Tue, Aug 12, 2014 at 10:45:30AM -0700, Ankur Shar
Looks good to me,
On Mon, Aug 11, 2014 at 2:36 PM, Ben Pfaff wrote:
> These error codes are proposed in the ONF extensibility working group as an
> OpenFlow 1.3 extension.
>
> Error codes are also proposed for the other three "bad property" error
> codes, but those already have standardized Ope
On Tue, Aug 12, 2014 at 10:45:30AM -0700, Ankur Sharma wrote:
> Just to add to it.
> make distcheck is also failing with following error
>
> ERROR: files left in build directory after distclean:
> ./lib/ofp-actions.inc1
> ./lib/ofp-actions.inc2
Thanks for the report. I sent out a fix:
ht
On Tue, Aug 12, 2014 at 05:39:21PM +, Alin Serdean wrote:
> c:\temp\12_08_2014\ovs\lib\ofp-actions.h(465) : error C2229: struct
> 'ofpact_nest
> ' has an illegal zero-sized array
> make[2]: *** [lib/bundle.lo] Error 1
> make[2]: Leaving directory `/c/temp/12_08_2014/ovs'
> make[1]: *** [all-re
Before commit c2d936a44fa6 (ofp-actions: Centralize all OpenFlow action
code for maintainability.), struct ofpact was 4 bytes with GCC and Clang,
and 12 bytes with other compilers. That commit changed struct ofpact so
that it remained 4 bytes with GCC and Clang but shrank to 8 bytes on other
compi
Fixes a failure with make distcheck.
Reported-by: Ankur Sharma
Signed-off-by: Ben Pfaff
---
lib/automake.mk |1 +
1 file changed, 1 insertion(+)
diff --git a/lib/automake.mk b/lib/automake.mk
index f4426f2..e95191d 100644
--- a/lib/automake.mk
+++ b/lib/automake.mk
@@ -438,6 +438,7 @@ lib/
Looks good to me,
Acked-by: Alex Wang
On Mon, Aug 4, 2014 at 1:02 PM, Ben Pfaff wrote:
> On Wed, Jul 16, 2014 at 03:00:15PM -0700, Ben Pfaff wrote:
> > The controller setup in my personal test environment has been broken for
> a
> > while. I figured that it wasn't anything important, though,
Added saurabh's fix to not to include some header files
in lib/netlink-protocol.h not needed by windows driver.
This is a temporary fix to make sure that windows driver can compile.
We are working in parallel on adding netlink support in windows driver
(https://github.com/openvswitch/ovs-issues/i
On Wed, Aug 06, 2014 at 07:40:25PM -0700, Ethan Jackson wrote:
> This patch avoids the relatively inefficient miss handling processes
> dictated by the dpif process, by calling into ofproto-dpif directly
> through a callback.
>
> Signed-off-by: Ethan Jackson
In dp_netdev_input(), the variable na
With this commit we move our DPDK support to 1.7.0.
DPDK binaries (starting with dpdk 1.7.0) should be linked with --whole-archive
to include pmd drivers
Signed-off-by: Daniele Di Proietto
---
INSTALL.DPDK | 12 ++--
acinclude.m4 | 13 -
lib/
rte_eth_tx_burst() _should_ transmit every packet that it is passed unless the
queue is full. Nontheless some implementation of rte_eth_tx_burst (e.g.
ixgbe_xmit_pkts_vec()) does not transmit more than a fixed number (32) of
packets at a time.
With this commit we assume that there's an error only
Hi,
Just to add to it.
make distcheck is also failing with following error
ERROR: files left in build directory after distclean:
./lib/ofp-actions.inc1
./lib/ofp-actions.inc2
Regards,
Ankur
From: dev on behalf of Alin Serdean
Sent: Tuesday, August 12,
On Wed, Aug 06, 2014 at 07:40:24PM -0700, Ethan Jackson wrote:
> This patch reorganizes ofproto-dpif in preparation for future patches
> which allow direct upcall processing from dpif-netdev. The main goals
> are to share as much code as possible between the dpif-linux and
> dpif-netdev upcall pat
Hi Ben,
I noticed today the windows build was broken.
Error is:
source='lib/bundle.c' object='lib/bundle.lo' libtool=yes \
DEPDIR=.deps depmode=none /bin/sh ./build-aux/depcomp \
/bin/sh ./libtool --tag=CC --mode=compile ./build-aux/cccl -DHAVE_CON
FIG_H -I. -I ./incl
Thanks Pravin and Jarno for the review. Pushed to master.
On Tue, Aug 12, 2014 at 9:47 AM, Pravin Shelar wrote:
> On Mon, Aug 11, 2014 at 3:36 PM, Andy Zhou wrote:
>> When flow key becomes invalid due to push or pop actions, current
>> implementation leaves it as invalid, only rebuild the flow
On Thu, Aug 07, 2014 at 02:25:42PM +1200, Joe Stringer wrote:
> On 16 July 2014 11:09, Ben Pfaff wrote:
>
> > The dpif provider 'operate' call duplicates all of the features available
> > from the 'flow_put', 'flow_del', and 'execute' calls, yielding redundant
> > code in providers that support b
On Mon, Aug 11, 2014 at 3:36 PM, Andy Zhou wrote:
> When flow key becomes invalid due to push or pop actions, current
> implementation leaves it as invalid, only rebuild the flow key used
> for recirculation.
>
> This works, but is less efficient in case of multiple recirc
> actions. Each recirc a
On 11/08/14 14:28, Ben Pfaff wrote:
> On Mon, Aug 11, 2014 at 08:17:19PM +0200, Dariusz Dwornikowski wrote:
> > -BEGIN PGP SIGNED MESSAGE-
> > Hash: SHA256
> >
> > On 11.08.2014 17:38, Ben Pfaff wrote:
> > > On Mon, Aug 11, 2014 at 09:53:09AM +0200, Dariusz Dwornikowski
> > > wrote:
> > >>
ovs_vport_alloc() bails out without freeing the memory 'vport' points to.
Picked up by Coverity - CID 1230503.
Fixes: 5cd667b0a4 ("openvswitch: Allow each vport to have an array of
'port_id's.")
Signed-off-by: Christoph Jaeger
---
net/openvswitch/vport.c | 4 +++-
1 file changed, 3 insertions(
49 matches
Mail list logo
