This patch adds multiple internal ports support to the windows datapath.
All tunnels types have been updated to accommodate this new functionality.
Signed-off-by: Alin Gabriel Serdean
Co-authored-by: Sorin Vinturis
Acked-by: Paul Boca
---
v3: Add acked
v2: Rebase
---
datapath-windows/ovsext/Ac
+Ben +Guru
Thanks a lot for the review!
To be short the subject of interest is the sanitization of the input during
queries.
You need to have root access or know how the remote management was created and
have access to it, thus you could do worse things.
WQL is based on DMTF(https://www.dmtf.
Thanks for the review! I agree with the comments and will update the
documentation/respin a new series once the Patch 2/3 is reviewed.
Thanks,
Alin.
> Sai:
[Alin Gabriel Serdean: ] Did you also have a comment here?
> >+In our example, this is how Œovs-vsctl show¹ looks like with the NICs
> >sep
Thanks!
De la: Guru Shetty [mailto:g...@ovn.org]
Trimis: Wednesday, August 3, 2016 5:27 PM
Către: Alin Serdean
Cc: dev@openvswitch.org
Subiect: Re: [ovs-dev] [PATCH v7] Windows: Local named pipe implementation
On 2 August 2016 at 11:19, Alin Serdean
mailto:aserd...@cloudbasesolutions.com
Hi Sai,
Thanks for the update!
Comments inlined.
Thanks,
Alin.
> -Mesaj original-
> De la: Sairam Venugopal [mailto:vsai...@vmware.com]
> Trimis: Saturday, August 6, 2016 12:46 AM
> Către: Alin Serdean ;
> dev@openvswitch.org
> Subiect: Re: [ovs-dev] [PATCH v3 2/3
> > Is there really anyone that can justify the way socket paths are
> > created should contain the PID? You clearly saw the problem with it in
> > the Windows code base, but yet pushed on to keep it in the Linux one.
> > I simply do not see why you're not taking this patch as a way to
> > encourag
> > While it's an alternative, I feel like going down that path is hasty
> > and for my use case I will probably just patch our local fork of
> > OpenvSwitch.
> >
> > Why so hostile towards this change? Is there a better forum I can
> > explain why I believe this problem should be reconsidered?
>
Small nit: maybe rename the macro from OVS_SKIP_NON_ADMIN to
OVS_SKIP_NON_ADMIN_WIN
Acked-by: Alin Gabriel Serdean
Tested-by: Alin Gabriel Serdean
> +dnl OVS_SKIP_NON_ADMIN()
> +dnl
> +dnl Checks if we have enough rights to create a service
> +m4_define([OVS_SKIP_NON_ADMIN],
> + [
> + AT_SK
From windows point of view, it looks good, because we need to close file
(handle) before we try to move it.
In my opinion we should not change anything on the Linux side, so either ifdef
it or create a new macro for this particular case (closing a log file) which
can be used later on.
Maybe th
That will disable the test if it is under
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Friday, June 3, 2016 4:07 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH] tests: Skip vlog close and set if no python is
> installed
>
From the commit message it is hard to understand what are you trying to fix or
why.
Are you trying to get the absolute path if "name" is missing ":" ?
Please state what are you trying to achieve and why, in the commit message.
Thanks,
Alin.
> -Mesaj original-
> De la: dev [mailto:dev-b
Hi Lance,
Thanks a lot for the patch just one small nit on the windows side, please see
inlined comment.
Thanks,
Alin.
> diff --git a/tests/ovs-macros.at b/tests/ovs-macros.at index
> e5710a0..fadd2be 100644
> --- a/tests/ovs-macros.at
> +++ b/tests/ovs-macros.at
> @@ -132,20 +132,22 @@ m4_defi
> > >+curNbl = *newNbl;
> > >+curNb = NET_BUFFER_LIST_FIRST_NB(curNbl);
> >
> > The reason we were doing the partial copy earlier is to make sure that
> > ETH
> > + IP + GRE headers are all in the first MDL. The code access these
> > + headers
> > assuming they are contiguous in memory. Is
>
> Hi guys!
>
> Thanks for the review!
> Now I see that I didn't fully understood what AT_CAPTURE_FILE macro does
> when I wrote the patch (this was my first contact with autoconf testsuites).
[Alin Gabriel Serdean: ] It is a tough cookie at first :).
You are doing well!
> I will send another on
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Tuesday, June 7, 2016 11:01 AM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH V3 2/4] ovs-ofctl: Fixed PID file naming on
> windows
>
> If an absol
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Wednesday, June 8, 2016 11:41 AM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH V4 2/4] ovs-ofctl: Fixed PID file naming on
> windows
>
> On Window
This patch introduces GRE checksum computation if the userspace requires
it on Tx. On Rx we verify the GRE checksum if the checksum bit was
specified and also inform the userspace about it.
Also fix the GRE header length as specified by the GRE flags not the
tunnel flags.
Signed-off-by: Alin Gabr
Using the --detach parameter the child does not propagate the first
message to the parent.
Proposed change use the log file instead of the stderr.
Signed-off-by: Alin Gabriel Serdean
---
tests/vlog.at | 10 ++
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/tests/vlog.at b
This patch removes the code which disables Python tests to be run on
Windows.
Signed-off-by: Alin Gabriel Serdean
---
tests/atlocal.in | 7 ---
1 file changed, 7 deletions(-)
diff --git a/tests/atlocal.in b/tests/atlocal.in
index f174061..410199f 100644
--- a/tests/atlocal.in
+++ b/tests/at
.
> -Mesaj original-
> De la: Ben Pfaff [mailto:b...@ovn.org]
> Trimis: Wednesday, June 8, 2016 6:07 PM
> Către: Alin Serdean
> Cc: dev@openvswitch.org
> Subiect: Re: [ovs-dev] [PATCH] at tests: Allow Python tests to be run on
> Windows
>
> On Wed, Jun 08, 201
Thanks for the review answer inlined.
Thanks,
Alin.
> -Mesaj original-
> De la: Nithin Raju [mailto:nit...@vmware.com]
> Trimis: Wednesday, June 8, 2016 10:10 PM
> Către: Alin Serdean ;
> dev@openvswitch.org
> Subiect: Re: [ovs-dev] [PATCH v3] datapath-windows
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Sorin Vinturis
> Trimis: Wednesday, June 1, 2016 6:50 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH v2] datapath-windows: Sample action support.
>
> This patch add
Hi Nithin,
Thanks for the patch. Beside a few small nits regarding whitespace and extra
comments, please see considerations to GRE tunnels in inlined comments.
Alin.
> -if (tunnelVport) {
> +break;
> +case OVS_VPORT_TYPE_VXLAN:
> ovsAction
Hi Nithin,
Thanks for the patch. Beside a few small nits regarding whitespace and extra
comments, please see considerations to GRE tunnels in inlined comments.
Alin.
> -if (tunnelVport) {
> +break;
> +case OVS_VPORT_TYPE_VXLAN:
> ovsAction
Please disregard this review I it was intended for the V2
(https://patchwork.ozlabs.org/patch/629912/)
Thanks,
Alin.
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Alin Serdean
> Trimis: Tuesday, June 14, 2016 7:01 AM
> Către: Nithi
Hi Sai,
Thanks for incorporating the comments so far. It looks good but we need to
treat corner cases like no valid resource allocations and NULL checks for
parameter.
A few small nits inlined.
Thanks,
Alin.
<---cut>
> ctx->key.src.port = flowKey->ipKey.l4.tpSrc;
>
Hi Yin,
Thanks a lot for the patch, and welcome to the team :).
Please try to limit lines to 79 characters per coding style
(https://github.com/openvswitch/ovs/blob/master/datapath-windows/CodingStyle).
Your patch is dependent on a patch that Nithin sent
http://openvswitch.org/pipermail/dev/20
I looked briefly over stt.c and there is a list of issues to be fixed if you
want to make part of this patch:
NdisGetDataBuffer can be null and should be checked
(https://msdn.microsoft.com/en-us/library/windows/hardware/ff562631(v=vs.85).aspx
)
" The return value can also be NULL due to a low
> >greater than the maximum header size which is configurable. In the case
> >of VXLAN / STT we can be in the case above, because of the UDP/TCP like
> >packet
>
> [Nithin]: Yes, in fact even for NICs that support header-split (mostly Intel),
[Alin Gabriel Serdean: ] I looked over some Intel specs
I'll take a look over it tomorrow :).
I saw that you had troubles compiling/linking the project on Windows. What
Visual Studio version did you use to compile it?
Thanks,
Alin.
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Lance
> Richardson
> Trimis: Fri
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Nithin Raju
> Trimis: Thursday, June 16, 2016 8:17 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH] datapath-windows: comment cleanup and
> indentation
>
> Signed-of
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Sairam
> Venugopal
> Trimis: Thursday, June 16, 2016 8:54 PM
> Către: Sairam Venugopal ; dev@openvswitch.org
> Subiect: Re: [ovs-dev] [PATCH v6] datapath-windows: Add support for UDP
> and ICMP to Conntrack Mod
Maybe I was a bit unclear in my previous comment. See inlined comment.
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Nithin Raju
> Trimis: Wednesday, June 15, 2016 1:11 AM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH v3] datapath-windows: use ip
>
> [Nithin]: The code is simpler this way. No? I can add a comment to clarify and
> an ASSERT as well. BTW, we already make assumptions in the code w.r.t the
> L4 port number for GRE ports. Pls. have a look at the following code:
[Alin Gabriel Serdean: ] The problem is not we make an assumption o
This patch introduces GRE checksum computation if the userspace requires
it on Tx. On Rx we verify the GRE checksum if the checksum bit was
specified and also inform the userspace about it.
Also fix the GRE header length as specified by the GRE flags not the
tunnel flags.
Signed-off-by: Alin Gabr
Thanks for incorporating the comments:
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Nithin Raju
> Trimis: Friday, June 17, 2016 8:52 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH v4] datapath-windows: use ip pr
Hi Lance,
I applied your patch and testing it out right now. I'll let you know the result
as soon as possible.
> Initially I tried using 2015, but ran into multiple compilation issues (the
> first
> due to a type collision in pthread.h, easily fixed, I don't recall what I ran
> into
> next), t
suite provides a good starting point. Its output may
be found below `tests/testsuite.dir'.
As a first run. Analyzing things in more detail.
> -Mesaj original-----
> De la: Alin Serdean
> Trimis: Wednesday, June 22, 2016 1:58 AM
> Către: 'Lance Richardson'
> Cc
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Sairam
> Venugopal
> Trimis: Wednesday, June 22, 2016 1:24 AM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH] datapath-windows: Remove unused headers in
> Event.c
>
>
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Sairam
> Venugopal
> Trimis: Wednesday, June 22, 2016 2:54 AM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH] datapath-windows: Handle memory allocation
> failure for event creation
>
> Release the l
There still are some NULL pointer checks that need to be done, but we can add
them later.
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Sairam
> Venugopal
> Trimis: Tuesday, June 21, 2016 4:15 AM
> Către: dev@openvswitch.o
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Sunday, June 26, 2016 3:12 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH V4 17/20] ovs-ofctl.at: Prevent msys from getting
> confused with ipv6
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Sunday, June 26, 2016 3:12 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH V4 01/20] tests: Fixed PMD tests on Windows
>
> CHECK_CPU_DISCOVERED c
This patch makes automake skip the 'datapath' folder under Windows.
The main reason is that msys does not have an implementation of rev
used by:
https://github.com/openvswitch/ovs/blob/master/datapath/Makefile.am#L34
and can be seen on:
https://ci.appveyor.com/project/blp/ovs/build/1.0.1851#L322
Also paths in the test do not exist :) i.e. :
" mkdir etc etc/init.d etc/racoon etc/racoon/certs
mkdir usr usr/sbin"
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Friday, July 1, 2016 7:27 PM
> Către: de
> De la: Ben Pfaff [mailto:b...@ovn.org]
> Trimis: Friday, July 1, 2016 11:31 PM
> Către: Alin Serdean
> Cc: dev@openvswitch.org
> Subiect: Re: [ovs-dev] [PATCH] build: Skip datapath subdir on Windows
>
> I think that this will make "make dist" on Windows omit th
This patch makes automake skip the function 'check-export-symbol' under
Windows.
The main reason is that msys does not have an implementation of rev
used by:
https://github.com/openvswitch/ovs/blob/master/datapath/Makefile.am#L34
and can be seen on:
https://ci.appveyor.com/project/blp/ovs/build/1
Acked-by: Alin Gabriel Serdean
Thanks,
Alin.
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Monday, July 4, 2016 1:01 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH V2] ovn test: Skip "send gratuitous arp on
> localnet" if
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Friday, July 1, 2016 7:27 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH V5 16/17] tests: Skip vlog tests that try to move
> opened file
>
> m4_define([VLOG_REOPEN_PYN],
>
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Friday, July 1, 2016 7:28 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH V5] windows: Added lockf function and lock PID
> file
>
> If the PID file isn't locked then appctl.py det
> > I used only shared lock, in order to be compatible with Python tests,
> > which try to acquire the lock exclusively. On Windows if the exclusive
> > lock is used, than the read access is denied too for other instances of this
> file.
> >
> > Signed-off-by: Paul-Daniel Boca
>
> I'm still not s
> > I'm still not sure that real locking is actually needed for Windows.
> > I believe that, on Windows, the functionality of a lock file only
> > requires holding the file open. For a program to detect whether the
> > lock file is locked, it can just try to delete it; on Windows, I
> > believe th
This patch needs to be applied last.
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: Paul Boca
> Trimis: Wednesday, July 6, 2016 3:39 PM
> Către: dev@openvswitch.org
> Cc: Alin Serdean
> Subiect: [PATCH V6 13/17] at tests: Allow Python tests to be run on
As a personal preference I think the comment:
> + # This test won't work as-is on Windows because Windows doesn't allow
> + # files that are open to be renamed.
is sufficient.
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În nume
Just out of curiosity why is this test is skipped on Windows?
Since this patch was skipped thus far I will acknowledge the patch and we can
fix the test if we can in another.
Acked-by: Alin Gabriel Serdean
Thanks,
Alin.
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.o
The code is sound but looks a bit messy. What do you think about?
@@ -15,6 +15,7 @@
import atexit
import os
import signal
+import sys
import ovs.vlog
@@ -129,8 +130,12 @@ def _init():
if not _inited:
_inited = True
-for signr in (signal.SIGTERM, signal.SIGINT,
-
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Wednesday, July 6, 2016 3:38 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH V6 03/17] python tests: Fixed ctl file name for
> Windows
>
> On Windows the CTL file doesn't contain
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Wednesday, July 6, 2016 3:38 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH V6 04/17] python tests: Fixed unixctl python tests
> for Windows
>
>
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Wednesday, July 6, 2016 3:38 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH V6 05/17] python tests: Added fcntl module for
> Windows
>
> This is needed for lockf function used to
We discussed this issue offline root cause seems to be:
https://bugs.python.org/issue6926
https://github.com/openvswitch/ovs/blob/master/python/ovs/socket_util.py#L309
Needs a commit message update.
Otherwise:
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-b
+ AT_CHECK([$3 $srcdir/test-unixctl.py --log-file=`pwd`/log --pidfile \
+ --detach 2>/dev/null], [0], [], [ignore])
Either ignore stderr or 2>/dev/null. In case [ignore] doesn't work please add a
comment so others will be aware of the problem
Alin.
> -Mesaj original-
> De la: dev
It's ok with me to skip the tests for the moment, but please add a comment why
are you skipping them i.e. "echo $! gives shell pid instead of parent process"
Otherwise:
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Bo
Just add a comment in the code so people know.
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca
> Trimis: Wednesday, July 6, 2016 3:38 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH V6 10/17] python tes
Multiple internal / external ports support for the Windows
datapath.
Alin Gabriel Serdean (4):
Add internal port per OVS bridge
datapath - windows: Add multiple internal ports support
Update Windows documentation
Forwarding team extension
INSTALL.Windows.md | 146 +++--
This patch updates the following commands in the vswitch:
ovs-vsctl add-br br-test
ovs-vsctl del-br br-test
ovs-vsctl add-br br-test:
This command will now create an internal port on the MSFT virtual switch
using the WMI interface from Msvm_VirtualEthernetSwitchManagementService
leveraging
Multiple internal / external ports support for the Windows datapath.
Alin Gabriel Serdean (4):
Add internal port per OVS bridge
datapath - windows: Add multiple internal ports support
Update Windows documentation
Forwarding team extension
INSTALL.Windows.md | 146 +++--
This patch updates the documentation to be in sync with the new internal
port enhancements.
Signed-off-by: Alin Gabriel Serdean
---
INSTALL.Windows.md | 54 +++---
1 file changed, 23 insertions(+), 31 deletions(-)
diff --git a/INSTALL.Windows.md b
This patch updates the documentation in how to set up OVS with multiple
NICs.
Signed-off-by: Alin Gabriel Serdean
---
INSTALL.Windows.md | 70 +++---
1 file changed, 67 insertions(+), 3 deletions(-)
diff --git a/INSTALL.Windows.md b/INSTALL.Window
This patch adds multiple internal ports support to the windows datapath.
All tunnels types have been updated to accommodate this new functionality.
Signed-off-by: Alin Gabriel Serdean
Co-authored-by: Sorin Vinturis
---
datapath-windows/ovsext/Actions.c | 32 +-
datapath-windows/ovsext/Gre.c
This patch adds the Visual Studio runtime to the binaries by adding
the code generation compile flags: MT (release version) and MTd (debug
version).
More on the subject can be found under:
https://msdn.microsoft.com/en-us/library/2kzt1wy3%28v=vs.120%29.aspx
This helps shipping binaries and guaran
A PnP(plug and play) event will be triggered before trying to disable
the extension. We could use this PnP event to prepare for detaching
the datapath.
This patch sets the switch into a paused state so no more net buffers
are queued.
Also clean some commentaries.
Signed-off-by: Alin Gabriel Serd
Found by inspection.
Signed-off-by: Alin Gabriel Serdean
---
lib/netlink-socket.c | 16 +---
1 file changed, 1 insertion(+), 15 deletions(-)
diff --git a/lib/netlink-socket.c b/lib/netlink-socket.c
index 5ef4b15..42d475b 100644
--- a/lib/netlink-socket.c
+++ b/lib/netlink-socket.c
@
Small nit can you please add a default to the switch and put an assert.
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Sairam
> Venugopal
> Trimis: Friday, February 26, 2016 1:07 AM
> Către: dev@openvswitch.org
> Subiect: [o
You are right Sai, the flags need to be set before the checksum is computed.
I acked you patch.
Alin.
> -Mesaj original-
> De la: Sairam Venugopal [mailto:vsai...@vmware.com]
> Trimis: Friday, February 26, 2016 12:56 AM
> Către: Sorin Vinturis ;
> dev@openvswitch.org
>
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Sairam
> Venugopal
> Trimis: Monday, February 29, 2016 9:42 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH v2] datapath-windows: Support for IPv6 in TCP
> segmentati
@vmware.com]
> Trimis: Tuesday, March 1, 2016 1:17 AM
> Către: Alin Serdean ;
> dev@openvswitch.org
> Subiect: Re: [ovs-dev] [PATCH] Build Windows include runtime
>
> Alin,
> Did you notice any change in the size of the binary? Eg. Ovs-vswitchd.exe?
>
> With or withou
Thanks for letting us know.
I will also miss the meeting as I have to tag along to another.
Thanks,
Alin.
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Ben Pfaff
> Trimis: Tuesday, March 1, 2016 6:24 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] mi
I will respin it later on tomorrow.
Alin.
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Sairam
> Venugopal
> Trimis: Monday, February 29, 2016 9:35 PM
> Către: Sorin Vinturis ;
> dev@openvswitch.org
> Subiect: Re: [ovs-dev] [PATCH v3 1/6] datapath-windows:
Comments inlined.
> -Mesaj original-
> De la: Nithin Raju [mailto:nit...@vmware.com]
> Trimis: Monday, February 29, 2016 10:43 PM
> Către: Alin Serdean ;
> dev@openvswitch.org
> Subiect: Re: [ovs-dev] [PATCH] datapath-windows: Pause switch state
>
>
This patch series can be dropped.
I will squash patch 2/3 and also change the tiles of the patches.
Thanks,
Alin.
> -Mesaj original-
> De la: Alin Serdean
> Trimis: Friday, February 26, 2016 7:37 AM
> Către: dev@openvswitch.org
> Cc: Alessandro Pilotti ; Alin Serde
Unless I am reading wrong: OvsAddConfiguredSwitchPorts and
OvsInitConfiguredSwitchNics only fail if we could not allocate memory or could
not issue an OID request.
I am not in favor of reporting back to NDIS gracefully if any of the above
conditions was broken even once.
Thanks,
Alin.
> -
Recirculation support for the OVS extension.
Tested using PING and iperf with Driver Verifier enabled.
Signed-off-by: Sorin Vinturis
Co-authored-by: Alin Gabriel Serdean
Reported-by: Sorin Vinturis
Reported-at: https://github.com/openvswitch/ovs-issues/issues/104
---
v2: Initialize flow key be
avoid the extension to be run on two or more switches, and also
check the elementnames of the VIFs and inform the user if the activation failed
and why.
Alin.
> -Mesaj original-
> De la: Nithin Raju [mailto:nit...@vmware.com]
> Trimis: Tuesday, March 8, 2016 3:20 AM
> Către:
h and resend it. Do you think I should respin the rest of
the patches also?
Thanks,
Alin.
> -Mesaj original-
> De la: Sairam Venugopal [mailto:vsai...@vmware.com]
> Trimis: Thursday, March 10, 2016 1:16 AM
> Către: Alin Serdean ;
> dev@openvswitch.org
> Subiect: Re:
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Sairam
> Venugopal
> Trimis: Monday, February 29, 2016 9:42 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH v2] datapath-windows: Support for IPv6 in TCP
> segmentati
A PnP(plug and play) event will be triggered before trying to disable
the extension. We could use this PnP event to prepare for detaching
the datapath.
This patch sets the switch into a paused state so no more net buffers
are queued.
Also clean some commentaries.
Signed-off-by: Alin Gabriel Serd
Revisit the mapping of an IPv4 key to netlink key and add the according
transformation.
Also add support for OVS_KEY_ATTR_ICMP to the windows datapath.
Signed-off-by: Alin Gabriel Serdean
---
datapath-windows/ovsext/Flow.c | 16 +++-
1 file changed, 11 insertions(+), 5 deletions(-)
This patch applies the conversions needed for the ICMPv6 type and code.
Signed-off-by: Alin Gabriel Serdean
---
datapath-windows/ovsext/Flow.c | 9 -
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git a/datapath-windows/ovsext/Flow.c b/datapath-windows/ovsext/Flow.c
index 5eec51
Recirculation support for the OVS extension.
Tested using PING and iperf with Driver Verifier enabled.
Signed-off-by: Sorin Vinturis
Co-authored-by: Alin Gabriel Serdean
Reported-by: Sorin Vinturis
Reported-at: https://github.com/openvswitch/ovs-issues/issues/104
---
v2: Initialize flow key be
This patch updates the following commands in the vswitch:
ovs-vsctl add-br br-test
ovs-vsctl del-br br-test
ovs-vsctl add-br br-test:
This command will now create an internal port on the MSFT virtual switch
using the WMI interface from Msvm_VirtualEthernetSwitchManagementService
leveraging
This patch updates the documentation in how to set up OVS with multiple
NICs.
Signed-off-by: Alin Gabriel Serdean
---
v2: Change title
---
---
INSTALL.Windows.md | 70 +++---
1 file changed, 67 insertions(+), 3 deletions(-)
diff --git a/INSTALL.Wi
This patch adds multiple internal ports support to the windows datapath.
All tunnels types have been updated to accommodate this new functionality.
Signed-off-by: Alin Gabriel Serdean
Co-authored-by: Sorin Vinturis
---
v2: change title
---
---
datapath-windows/ovsext/Actions.c | 35 +-
datapa
Ty!
> -Mesaj original-
> De la: Ben Pfaff [mailto:b...@ovn.org]
> Trimis: Monday, March 14, 2016 7:25 PM
> Către: Alin Serdean
> Cc: dev@openvswitch.org
> Subiect: Re: [ovs-dev] [PATCH] Build Windows include runtime
>
> On Fri, Feb 26, 2016 at 01:59:10PM +
Currently the datapath reports the tick counter to the userspace.
The userspace uses KeQueryPerformanceCounter as a monotonic clock.
This patch changes the flow stats to be reported in a monotonic format, while
also decaying the time between the flow actual usage and the flow report usage.
This p
This patch updates the appveyor dowload link for OpenSSL prebuilt binaries.
Signed-off-by: Alin Gabriel Serdean
---
appveyor.yml | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/appveyor.yml b/appveyor.yml
index 24a2814..422c4af 100644
--- a/appveyor.yml
+++ b/appveyor.y
Acked-by: Alin Gabriel Serdean
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Sorin Vinturis
> Trimis: Friday, March 25, 2016 5:56 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH] datapath-windows: Revert unrelated change to
> ovs-atomic.h
>
>
Comments inlined.
Thanks,
Alin.
> -Mesaj original-
> De la: dev [mailto:dev-boun...@openvswitch.org] În numele Sairam
> Venugopal
> Trimis: Friday, March 25, 2016 8:07 PM
> Către: dev@openvswitch.org
> Subiect: [ovs-dev] [PATCH v2] datapath-windows: Update Recirculation to
> use the right
Please move:
destKey->l2.keyLen = OVS_WIN_TUNNEL_KEY_SIZE + OVS_L2_KEY_SIZE
- destKey->l2.offset;
After
_MapTunAttrToFlowPut(keyAttrs, tunnelAttrs, destKey);
:
$ git diff
diff --git a/datapath-windows/ovsext/Flow.c b/datapath-windows/ovsext/Flow.c
index 02c41b7..49abb8
If we don't move that upwards it will reset the keylen after recircid/hash has
been set to update it.
> -Mesaj original-
> De la: Nithin Raju [mailto:nit...@vmware.com]
> Trimis: Tuesday, March 29, 2016 5:45 PM
> Către: Alin Serdean ; Sairam Venugopal
> ; dev@openv
Again please move the keylen reset upwards before the keylen gets updated.
We could do as you propose Nithin. Although hash is applicable from my
knowledge only in the case of bonding.
We could simply just align the structure "OvsFlowKey" to 8 and make sure we
initialize the members properly th
201 - 300 of 799 matches
Mail list logo
