On Tue, Feb 03, 2015 at 10:24:18PM -0800, Shu Shen wrote:
> This patch adds support for OFPR_ACTION_SET as the packet-in reason when
> a Packet-In message is triggered by an output action within an
> action-set. By default reason code OFPR_ACTION_SET is enabled for async
> messages when Openflow 1.
On Sat, Feb 07, 2015 at 07:11:13AM +, Jason Kölker wrote:
> On Sat, Feb 7, 2015 at 6:58 AM, Ben Pfaff wrote:
> > I'm pretty sure that this version will not parse "ssl:1.2.3.4:5"
> > correctly; won't it return "1.2.3.4:5" instead of just "1.2.3.4"?
>
> It handles bolth cases. The split tuple w
On Tue, Feb 03, 2015 at 11:53:29PM -0800, Shu Shen wrote:
> The previous workflow is to convert dot diagrams into .pic format and
> embed into manpages; double borders and arrows were not used in dot but
> introduced in .pic conversion; edge routing in .pic was also worse than
> dot.
>
> The updat
On Sat, Feb 7, 2015 at 6:58 AM, Ben Pfaff wrote:
> I'm pretty sure that this version will not parse "ssl:1.2.3.4:5"
> correctly; won't it return "1.2.3.4:5" instead of just "1.2.3.4"?
It handles bolth cases. The split tuple will be ('ssl', '1.2.3.4',
'5') and it will return the middle value, else
On Fri, Feb 06, 2015 at 04:31:34PM -0800, Andy Zhou wrote:
> Executing clean-up commands in the reverse order of their addition
> seems to be better for most of the cleanup situations. For example,
> in kmod tests, we should remove name spaces before remove kernel
> modules.
>
> Signed-off-by: And
On Fri, Feb 06, 2015 at 04:31:35PM -0800, Andy Zhou wrote:
> To reduce the complexity of writing kmod tests.
>
> Suggested-by: Joe Stringer
> Signed-off-by: Andy Zhou
Acked-by: Ben Pfaff
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org
On Sat, Feb 07, 2015 at 06:47:51AM +, Jason Kölker wrote:
> With XenServer only 1 manager is configured in the pool, which may not
> be the first manager returned from `get-manager` as it returns in
> lexicographical order.
>
> v2: Fixes vswitchCurrentControllers() to always return a list
> v3
On Sat, Feb 07, 2015 at 06:33:08AM +, Jason Kölker wrote:
> On Sat, Feb 7, 2015 at 5:34 AM, Ben Pfaff wrote:
> > On Thu, Feb 05, 2015 at 04:32:57AM +, Jason Kölker wrote:
> >> With XenServer only 1 manager is configured in the pool, which may not
> >> be the first manager returned from `ge
With XenServer only 1 manager is configured in the pool, which may not
be the first manager returned from `get-manager` as it returns in
lexicographical order.
v2: Fixes vswitchCurrentControllers() to always return a list
v3: Use str.replace instead of lstrip for dns name saftey
v4: Use filter fun
With XenServer only 1 manager is configured in the pool, which may not
be the first manager returned from `get-manager` as it returns in
lexicographical order.
v2: Fixes vswitchCurrentControllers() to always return a list
v3: Use str.replace instead of lstrip for dns name saftey
Signed-off-by: Ja
On Sat, Feb 7, 2015 at 5:34 AM, Ben Pfaff wrote:
> On Thu, Feb 05, 2015 at 04:32:57AM +, Jason Kölker wrote:
>> With XenServer only 1 manager is configured in the pool, which may not
>> be the first manager returned from `get-manager` as it returns in
>> lexicographical order.
>>
>> V2: Fixes
On Wed, Feb 04, 2015 at 11:24:58AM +0100, Thomas Graf wrote:
> On 02/03/15 at 10:09am, Pravin Shelar wrote:
> > When I tried this patch series against mainline 3.19-rc7, I got
> > following error on insert.
> >
> > openvswitch: Unknown symbol rpl_vxlan_xmit_skb (err 0)
> > openvswitch: Unknown sym
You could use "timeout 10s modprobe -r ...". The "timeout" command is
specific to GNU coreutils, but "modprobe" is Linux-specific, so it's
probably OK (although it's also a newer utility, introduced in 2008).
Alternatively it's probably possible to build an equivalent to timeout
with shell constru
On Thu, Feb 05, 2015 at 04:32:57AM +, Jason Kölker wrote:
> With XenServer only 1 manager is configured in the pool, which may not
> be the first manager returned from `get-manager` as it returns in
> lexicographical order.
>
> V2: Fixes vswitchCurrentControllers() to always return a list
>
>
On Wed, Feb 04, 2015 at 11:12:30PM +, Jason Kölker wrote:
> Signed-off-by: Jason Kölker
Applied, thanks.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
To reduce the complexity of writing kmod tests.
Suggested-by: Joe Stringer
Signed-off-by: Andy Zhou
---
tests/kmod-macros.at | 1 +
tests/kmod-traffic.at | 1 -
2 files changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/kmod-macros.at b/tests/kmod-macros.at
index 1281e25..5824940 10064
Executing clean-up commands in the reverse order of their addition
seems to be better for most of the cleanup situations. For example,
in kmod tests, we should remove name spaces before remove kernel
modules.
Signed-off-by: Andy Zhou
---
tests/ovs-macros.at | 5 ++---
1 file changed, 2 insertion
On Fri, Feb 6, 2015 at 3:18 PM, Joe Stringer wrote:
> On 5 February 2015 at 16:06, Andy Zhou wrote:
>> # commands to delete any warnings that are actually expected, e.g.:
>> #
>> # OVS_KMOD_VSWITCHD_STOP(["/expected error/d"])
>> +#
>> +# 'extra_cmds' are shell commands to be executed afte O
I was hoping that some people with more understanding of how linux
modules work in general could give some context :-)
The difference between the error path and normal exit is that it's
more likely for the kernel to be in a bad state if it's the error
path.
Maybe the earlier patches to cleanup na
On 5 February 2015 at 16:06, Andy Zhou wrote:
> # commands to delete any warnings that are actually expected, e.g.:
> #
> # OVS_KMOD_VSWITCHD_STOP(["/expected error/d"])
> +#
> +# 'extra_cmds' are shell commands to be executed afte OVS_VSWITCHD_STOP() is
> +# invoked. They can be used to perf
I am seeing 4 tests failed with following assertion failed which is
introduced by the patch.
"assertion v <= OFPBUF_DATA_MAX failed in ofpbuf_set_size()"
I have couple of comments inline.
On Fri, Feb 6, 2015 at 8:43 AM, Mark Kavanagh wrote:
> DPDK v1.8.0 makes significant changes to struct rte_m
This is a good idea. Will do with a follow up patch.
On Fri, Feb 6, 2015 at 2:36 PM, Joe Stringer wrote:
> On 5 February 2015 at 16:06, Andy Zhou wrote:
>> Improves test failure handling. Remove name spaces that are no longer
>> needed so they to linger around before the next test starts.
>>
>>
I don't have a solution to this. Any suggestions?
The patch just add a module removal for the error path. The problem is
also there for the normal test exit.
On Fri, Feb 6, 2015 at 2:40 PM, Joe Stringer wrote:
> On 5 February 2015 at 16:06, Andy Zhou wrote:
>> Improve test failure handling.
>>
On 5 February 2015 at 16:06, Andy Zhou wrote:
> Improve test failure handling.
>
> Signed-off-by: Andy Zhou
What happens if "modprobe -r" hangs? (I occasionally notice this when
working with experimental kernel code, but I don't know what causes
it)
__
On 5 February 2015 at 16:06, Andy Zhou wrote:
> The ping output can be useful. There is no down side in saving it.
>
> Signed-off-by: Andy Zhou
Acked-by: Joe Stringer
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On 6 February 2015 at 14:33, Ben Pfaff wrote:
> On Fri, Feb 06, 2015 at 02:31:00PM -0800, Joe Stringer wrote:
>> On 5 February 2015 at 16:06, Andy Zhou wrote:
>> If we apply the same to the OVS test case here, then
>> this means that the namespaces will not be cleaned up if the test
>> fails. Wha
On 5 February 2015 at 16:06, Andy Zhou wrote:
> Improves test failure handling. Remove name spaces that are no longer
> needed so they to linger around before the next test starts.
>
> Signed-off-by: Andy Zhou
I wonder if we could hide this in ADD_NAMESPACES? I realise that
creating then deletin
On 5 February 2015 at 16:06, Andy Zhou wrote:
> This is mostly a style fix. The macro is used in the next patch to
> add commands to the 'cleanup" file. This fix makes the 'cleanup' file
> easier to read.
>
> Signed-off-by: Andy Zhou
> ---
> tests/kmod-macros.at | 2 +-
> 1 file changed, 1 inser
On Fri, Feb 06, 2015 at 02:31:00PM -0800, Joe Stringer wrote:
> On 5 February 2015 at 16:06, Andy Zhou wrote:
> If we apply the same to the OVS test case here, then
> this means that the namespaces will not be cleaned up if the test
> fails. What I've done in private testing is to modify the
> ADD
On 5 February 2015 at 16:06, Andy Zhou wrote:
> Those are not necessary and makes the test output file harder to read.
>
> Signed-off-by: Andy Zhou
If the latter case for my feedback on the previous patch is addressed,
then I think this is reasonable.
Acked-by: Joe Stringer
___
On 5 February 2015 at 16:06, Andy Zhou wrote:
> Remove namespaces also removes the ports in them, thus may
> causing vswitch to generate warning log messages about not being
> able find the port before it exits.
>
> Remove namespaces after ovs-vswitchd exits improves test reliability.
>
> Signed-o
On Fri, Feb 06, 2015 at 04:43:26PM +, Mark Kavanagh wrote:
> DPDK v1.8.0 makes significant changes to struct rte_mbuf, including
> removal of the 'pkt' and 'data' fields. The latter, formally a
> pointer, is now calculated via an offset from the start of the
> segment buffer. These fields are r
On 02/06/15 at 11:13am, Pravin Shelar wrote:
> On Wed, Feb 4, 2015 at 7:45 AM, Thomas Graf wrote:
> > Signed-off-by: Thomas Graf
> Acked-by: Pravin B Shelar
Thanks. I pushed patch 1-5 to master.
___
dev mailing list
dev@openvswitch.org
http://openvswi
On Wed, Feb 4, 2015 at 7:45 AM, Thomas Graf wrote:
> Signed-off-by: Thomas Graf
Acked-by: Pravin B Shelar
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Wed, Feb 4, 2015 at 7:45 AM, Thomas Graf wrote:
> Upstream commit:
> openvswitch: Support VXLAN Group Policy extension
>
> Introduces support for the group policy extension to the VXLAN virtual
> port. The extension is disabled by default and only enabled if the user
> has provi
On Thu, Feb 5, 2015 at 1:40 PM, Jarno Rajahalme wrote:
> OVS userspace already probes the openvswitch kernel module for
> OVS_ACTION_ATTR_SET_MASKED support. This patch adds the kernel module
> implementation of masked set actions.
>
> The existing set action sets many fields at once. When only
On 4 February 2015 at 14:49, Andy Zhou wrote:
> Added a eBPF program loader in ofproto-dpif layer. It only loades
> the OVS programs.
>
> Signed-off-by: Andy Zhou
> ---
A little more precise: It only loads BPF programs with sections that
begin with the 'ovs' prefix.
There's a few scattered OVS
On 4 February 2015 at 14:49, Andy Zhou wrote:
> Add a configuration option to allow specifing the LLVM backend module
> for compiling BPF C source code into BPF instruction sets.
>
> See INSTALL.BPF.md for usgae information.
>
> Signed-off-by: Joe Stringer
> Signed-off-by: Andy Zhou
> ---
> INS
Hi Nithin,
No worries go for it.
Thanks,
Alin.
-Mesaj original-
De la: Nithin Raju [mailto:nit...@vmware.com]
Trimis: Friday, February 6, 2015 2:41 AM
Către: Alin Serdean
Cc: Gurucharan Shetty; dev@openvswitch.org; Gurucharan Shetty
Subiect: Re: [ovs-dev] [PATCH 2/2] appveyor: Provide a
On Fri, Feb 6, 2015 at 3:09 AM, Thomas Graf wrote:
> On 02/05/15 at 07:53pm, Pravin Shelar wrote:
>> On Thu, Feb 5, 2015 at 6:11 PM, Jesse Gross wrote:
>> > We should also think about how to apply this in a consistent manner to
>> > other protocols that don't have this issue like Geneve.
>
> As a
DPDK v1.8.0 makes significant changes to struct rte_mbuf, including
removal of the 'pkt' and 'data' fields. The latter, formally a
pointer, is now calculated via an offset from the start of the
segment buffer. These fields are referenced by OVS when accessing
the data section of an ofpbuf.
The fol
Guru, driver build is very short (less than a min), It uses precompiled
headers. Let's have the debug build until we will have it as an option.
For user mode to take advantage on precompiled header, there is work to be dome
in arranging the header files.
Thank you for taking the time!
Eitan
> Would be nice to have an option for Release or debug build. User mode code
> too.
Right now, we are not deploying the created executables anywhere.
Appveyor does have that option. We can probably enhance it once we
have a in-tree windows installer.
On the release/debug question, the only option
On Thu, Feb 5, 2015 at 10:38 PM, Nithin Raju wrote:
>> +- C:\MinGW\msys\1.0\bin\bash -lc "cd /c/openvswitch && ./configure
>> CC=build-aux/cccl LD=\"`which link`\" LIBS=-lws2_32
>> --with-pthread=C:/pthreads-win32/Pre-built.2 --with-openssl=C:/OpenSSL-Win32
>> --with-vstudioddk=\"Win8.1 Release
On Thu, Feb 5, 2015 at 10:22 PM, Nithin Raju wrote:
> In a previous commit, support had been added to support building
> the kernel datapath from command line. In this patch, we document
> it.
>
> Signed-off-by: Nithin Raju
Thank you, applied!
> ---
> INSTALL.Windows.md | 28 ++
On 02/05/15 at 07:53pm, Pravin Shelar wrote:
> On Thu, Feb 5, 2015 at 6:11 PM, Jesse Gross wrote:
> > On Thu, Feb 5, 2015 at 6:02 PM, Pravin Shelar wrote:
> >> Since it should be possible to configure vxlan tunnel with only
> >> REMCSUM_RX or only REMCSUM_TX, I do not think we can store REMCSUM_T
46 matches
Mail list logo
