Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Existing STP and RSTP test cases only test the protocols with test
utilities. These tests test them as part of OVS using the
netdev-dummy device.
Signed-off-by: Jarno
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/rstp-state-machines.c | 22 --
ofproto/ofproto-dpif.c| 11 ++-
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Set the stp port name before enabling it, so that debugging messages
have the name to print out.
Do not treat the first state initialization as a state change. Zero
is not
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Move port's configuration options where they belong, add typing, and
correct errors.
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
vswitchd/vswitch.xml | 130
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/rstp-common.h |7 ---
1 file changed, 7 deletions(-)
diff --git a/lib/rstp-common.h
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/rstp.c | 132
++--
lib/rstp.h | 12 +++---
2 files
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Prior to this patch the default values for ports were set in three
different places. This refactors them all to one helper function.
Signed-off-by: Jarno Rajahalme
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/rstp.c | 54 ++
1 file changed, 22 insertions(+), 32
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/rstp.c | 39 ++-
ofproto/ofproto-dpif.c |2 +-
2 files
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/rstp.c | 34 --
lib/rstp.h | 42 +++---
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Current code expects there to be a single thread that is responsible
for creating rstp and creating and deleting rstp_port objects. rstp
objects are also deleted from other
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Change the RSTP send_bpdu interface so that a recursive mutex is not
needed.
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/rstp-common.h |2 +-
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/rstp-state-machines.c | 398
+
1 file changed, 182 insertions(+),
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Testing for sameness first makes the logic simpler to follow.
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/rstp-state-machines.c | 57
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
It was only used to guard against unintialized list.
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/rstp-common.h |1 -
lib/rstp-state-machines.c
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Finding a given port is faster.
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/rstp-common.h |5 +++--
lib/rstp-state-machines.c | 30
Thanks for the review!
It would be nice to have an Acked-by from you to the series. However, I
plan to squash trivial CodingStyle fixes in before pushing the series to
master. Also, I’ll add a News item stating that experimental RSTP is added,
and more compliance and interoperability
Thank You Ben for your inputs !!
Please suggest for the following as well:
1. If eviction on the switch is set on the basis of lifetime by create
table command and then the command for importance (mod-table command) is
given then what will be the basis of eviction in this case. Will in
On Tue, Sep 09, 2014 at 05:15:06PM +0530, Saloni Jain wrote:
1. If eviction on the switch is set on the basis of lifetime by
create table command and then the command for importance (mod-table
command) is given then what will be the basis of eviction in this
case. Will in this case, eviction
On Mon, Sep 8, 2014 at 9:08 PM, Mark Maglana mmagl...@gmail.com wrote:
Modify patch port name separators to use -- instead of - so that
ovs-vtep plays nice with external tools such as Mininet which uses
dashes in port names. Without this change, ovs-vtep will throw a
ValueError exception and
Here’s the updated description:
lib/dpif-netdev: Make emc_mutex recursive.
dpif_netdev_execute may be called while doing upcall processing.
Since the context of the input port is not tracked upto this point, we
use the shared dp-emc_cache for packet execution, where the
hi Sam,
Pls. find my answers inlined. Pls. feel free to repost the patches for the two
different functionalities.
On Sep 6, 2014, at 2:39 PM, Samuel Ghinet sghi...@cloudbasesolutions.com
wrote:
Yes, you were right with the TSO. I'll send a new version of the patch,
though I am not sure how
During a review, it seemed that some of the conventions were not clear.
Fixing them in this patch.
Signed-off-by: Nithin Raju nit...@vmware.com
---
datapath-windows/CodingStyle | 13 +
1 files changed, 13 insertions(+), 0 deletions(-)
diff --git a/datapath-windows/CodingStyle
During a review, it seemed that some of the conventions were not clear.
Fixing them in this patch.
Signed-off-by: Nithin Raju nit...@vmware.com
Reported-by: Samuel Ghinet sghi...@cloudbasesolutions.com
---
datapath-windows/CodingStyle | 13 +
1 files changed, 13 insertions(+), 0
On Aug 29, 2014, at 1:20 PM, Saurabh Shah ssaur...@vmware.com wrote:
Generally, we have not used '_' prefix for function parameter names. We
use them for macros etc, but not for functions. Do you prefer it this way? I
am not against it, but I don't think it is necessary. For a macro, it is
On Tue, Sep 9, 2014 at 8:04 AM, Gurucharan Shetty shet...@nicira.com
wrote:
I see that there is a bug. With your patch, would we still hit the bug
if the port name has a -- in it?
Definitely. Although I'm guessing that the probability of -- being in the
port name is negligible. I can make it
On Mon, Sep 08, 2014 at 05:06:05PM -0700, Ben Pfaff wrote:
On Fri, Sep 05, 2014 at 04:05:13PM -0700, Jarno Rajahalme wrote:
When a whole field of a key value is ignored, skip it when formatting
the key, and allow it to be left out when parsing the key from a
string. However, when the
Acked-by: Alin Gabriel Serdean aserd...@cloudbasesolutions.com
-Mesaj original-
De la: dev [mailto:dev-boun...@openvswitch.org] În numele Nithin Raju
Trimis: Tuesday, September 9, 2014 9:03 AM
Către: dev@openvswitch.org; Samuel Ghinet
Subiect: [ovs-dev] [PATCH v2] datapath-windows: update
On Sep 9, 2014, at 3:53 AM, Daniele Venturino venturino.dani...@gmail.com
wrote:
Thanks for the review!
It would be nice to have an Acked-by from you to the series. However, I plan
to squash trivial CodingStyle fixes in before pushing the series to master.
Also, I’ll add a News item
On current master, the per-thread callback event set is flushed
when ovsrcu_quiesce_start() is called or when the callback
event set is full. For threads that only call 'ovsrcu_quiesce()'
to indicate quiescient state, their callback event set will not
be flushed for execution until the set is
Hi, here are some items we would like to follow up on. Please feel free to add
anything you would like to discuss.
Outstanding review status
[1] NET_BUFFER_LIST to NET_BUFFER (Sam, Alin)
[2] Asynchronous Event user mode (Eitan)
[3] dpif_linux.c (Alin)
[4] Port Dump (Sam,
On Fri, Sep 05, 2014 at 04:05:14PM -0700, Jarno Rajahalme wrote:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
Reviewed-by: YAMAMOTO Takashi yamam...@valinux.co.jp
Acked-by: Ben Pfaff b...@nicira.com
---
v5: rebase, no other changes.
Still looks good.
Thanks for the fix, Jarno
Acked-by: Daniele Di Proietto ddiproie...@vmware.com
On 9/9/14, 8:20 AM, Jarno Rajahalme jrajaha...@nicira.com wrote:
Here¹s the updated description:
lib/dpif-netdev: Make emc_mutex recursive.
dpif_netdev_execute may be called while doing upcall
Series pushed to master, thank you for your contribution!
Jarno
On Sep 9, 2014, at 3:39 AM, Daniele Venturino venturino.dani...@gmail.com
wrote:
Acked-by: Daniele Venturino daniele.ventur...@m3s.it
2014-08-21 1:57 GMT+02:00 Jarno Rajahalme jrajaha...@nicira.com:
Finding a given port is
hi Guys,
Sorry for all the confusion generated from sending out separate patches for the
review comments. The reasons were the following:
1. I wanted to address all the review comments in the limited time I had, and
unblock others.
2. I thought it would be easier to see that the review comments
On Fri, Sep 05, 2014 at 04:05:15PM -0700, Jarno Rajahalme wrote:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
v5: Using pattern with less code duplication suggested by Ben.
odp_mask_is_exact() treats the unused bits in tun_mask-flags
differently from the unused bits in
On Mon, Sep 08, 2014 at 08:08:12PM -0700, Eitan Eliahu wrote:
We keep an outstanding, out of band, I/O request in the driver at all time.
Once an event generated the driver queues the event message, completes the
pending I/O and unblocks the calling thread through setting the event in the
Ben,
This patch is good to go as well AS AN independent patch. It was marked earlier
as [PATCH 9/9 v3], but it has since been broken off from the series, since
Ankur sent out a new series, which have been committed.
Pls. let me know if you want me to send out a fresh patch for this.
thanks,
On Fri, Sep 05, 2014 at 04:05:16PM -0700, Jarno Rajahalme wrote:
The frag member in the Netlink interface is an uint8_t enumeration
type, not a bitrfield, so it should always be either fully masked or
not masked at all.
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
The change to
On Sep 8, 2014, at 5:06 PM, Ben Pfaff b...@nicira.com wrote:
sparse says:
../lib/odp-util.c:2379:9: warning: incorrect type in argument 2 (different
base types)
../lib/odp-util.c:2379:9:expected restricted ovs_be32 [usertype] (
*key )[4]
../lib/odp-util.c:2379:9:got
On Fri, Sep 05, 2014 at 04:05:17PM -0700, Jarno Rajahalme wrote:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
Brilliant!
Acked-by: Ben Pfaff b...@nicira.com
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
Should be the same as other IPv6 address fields.
Forthcoming patches produce sparse warnings without this change.
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
datapath/linux/compat/include/linux/openvswitch.h |6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git
On Fri, Sep 05, 2014 at 04:05:18PM -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 before
I'd appreciate it if you'd resend separately, with Samuel's ack
included.
Thanks,
Ben.
On Tue, Sep 09, 2014 at 07:43:17PM +, Nithin Raju wrote:
Ben,
This patch is good to go as well AS AN independent patch. It was marked
earlier as [PATCH 9/9 v3], but it has since been broken off from
Il giorno 09/set/2014, alle ore 20:07, Jarno Rajahalme jrajaha...@nicira.com
ha scritto:
On Sep 9, 2014, at 3:53 AM, Daniele Venturino venturino.dani...@gmail.com
wrote:
Thanks for the review!
It would be nice to have an Acked-by from you to the series. However, I plan
to squash
Thanks Saurabh for clarifications,
@ Saurabh and Nithin: so you want me to re-make the patches so that #1 would be
refactor and #2 would be the bug fix? or for this one time we can go with #1
and #2 merged together?
There's some reference to this convention as follows.
On Sep 8, 2014, at 5:06 PM, Ben Pfaff b...@nicira.com wrote:
On Fri, Sep 05, 2014 at 04:05:13PM -0700, Jarno Rajahalme wrote:
When a whole field of a key value is ignored, skip it when formatting
the key, and allow it to be left out when parsing the key from a
string. However, when the
On Fri, Aug 22, 2014 at 1:16 AM, Nithin Raju nit...@vmware.com wrote:
In this patch, we add support for querying the genl family id for any
family supported by the OVS kernel datapath. On platforms that support
netlink natively, the operating system assigns a family ID, and the
OS netlink
Acked-by: Samuel Ghinet sghi...@cloudbasesolutions.com
From: Alin Serdean
Sent: Tuesday, September 09, 2014 8:43 PM
To: Nithin Raju; dev@openvswitch.org; Samuel Ghinet
Subject: RE: [ovs-dev] [PATCH v2] datapath-windows: update CodingStyle
guideline
On Tue, Sep 09, 2014 at 01:11:25PM -0700, Jarno Rajahalme wrote:
On Sep 8, 2014, at 5:06 PM, Ben Pfaff b...@nicira.com wrote:
On Fri, Sep 05, 2014 at 04:05:13PM -0700, Jarno Rajahalme wrote:
When a whole field of a key value is ignored, skip it when formatting
the key, and allow it to
Per review comment, in this patch we refactor the code to create a
OvsSetupDumpStart() which can be leveraged by dump functions in the
future. I have not refactored the code that continues the dump
operation primarily since it is not final yet. Once the netlink set
APIs are in place, we can
On Tue, Sep 09, 2014 at 01:14:31PM -0700, Nithin Raju wrote:
Per review comment, in this patch we refactor the code to create a
OvsSetupDumpStart() which can be leveraged by dump functions in the
future. I have not refactored the code that continues the dump
operation primarily since it is not
On Sep 9, 2014, at 1:09 PM, Gurucharan Shetty shet...@nicira.com
wrote:
Gives an uninitialized local variable 'sock' used. Can you take care
of it in an upcoming patch?
Sure. Thanks for point this out.
-- Nithin
___
dev mailing list
There would be one issue from me:
there was an email - reply to a vport patch time ago, where hyper-v switch
ports were separated from ovs / datapath ports.
The discussion had remained pending, but it might be important if we plan
netlink commands vport new and vport delete for the near future.
On Tue, Jul 29, 2014 at 8:23 AM, Alin Serdean
aserd...@cloudbasesolutions.com wrote:
Add two functions set_sock_pid_in_kernel and portid_next. This will allow
the channel identification for the kernel extension to send back messages.
Replace send with WriteFile equivalent and ignore
As mentioned in the comment above the function ovs_strerror(), it
should not be used to convert WINAPI error numbers to string.
Use ovs_lasterror_to_string() instead.
CC: Alin Gabriel Serdean aserd...@cloudbasesolutions.com
CC: Eitan Eliahu elia...@vmware.com
Signed-off-by: Gurucharan Shetty
When building with DPDK support, 'struct dpif_packet' won't have 'dp_hash'
member. dpif_packet_set_dp_hash() and dpif_packet_get_dp_hash() should be used.
Furthermore, the masked set action shouldn't read 'md-dp_hash' (which is
shared in a batch), but should use dpif_packet_get_dp_hash() to get
'sock' is not initialized and hence should not be un-initialized
as well in the failure path.
Signed-off-by: Nithin Raju nit...@vmware.com
Reported-by: Gurucharan Shetty shet...@nicira.com
---
lib/netlink-socket.c |2 --
1 files changed, 0 insertions(+), 2 deletions(-)
diff --git
On Tue, Sep 9, 2014 at 1:50 PM, Nithin Raju nit...@vmware.com wrote:
'sock' is not initialized and hence should not be un-initialized
as well in the failure path.
Signed-off-by: Nithin Raju nit...@vmware.com
Thanks, applied!
Reported-by: Gurucharan Shetty shet...@nicira.com
---
On Mon, Sep 08, 2014 at 02:54:13PM +0100, Thomas Graf wrote:
On 09/03/14 at 11:24am, Jiri Pirko wrote:
This patchset can be divided into 3 main sections:
- introduce switchdev api for implementing switch drivers
- add hardware acceleration bits into openvswitch datapath, This uses
Acked-by: Jarno Rajahalme jrajaha...@nicira.com
On Sep 9, 2014, at 1:32 PM, Daniele Di Proietto ddiproie...@vmware.com wrote:
When building with DPDK support, 'struct dpif_packet' won't have 'dp_hash'
member. dpif_packet_set_dp_hash() and dpif_packet_get_dp_hash() should be
used.
Windows already has a MAX_MONITORS defined in ddeml.h.
Signed-off-by: Gurucharan Shetty gshe...@nicira.com
---
lib/rconn.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/rconn.c b/lib/rconn.c
index c9b68bb..5c28806 100644
--- a/lib/rconn.c
+++ b/lib/rconn.c
@@
Thanks,
Daniele
On 9/9/14, 2:22 PM, Jarno Rajahalme jrajaha...@nicira.com wrote:
Pushed to master,
Jarno
On Sep 9, 2014, at 2:15 PM, Jarno Rajahalme jrajaha...@nicira.com wrote:
Acked-by: Jarno Rajahalme jrajaha...@nicira.com
On Sep 9, 2014, at 1:32 PM, Daniele Di Proietto
Pushed to master,
Jarno
On Sep 9, 2014, at 12:48 PM, Ben Pfaff b...@nicira.com wrote:
On Fri, Sep 05, 2014 at 04:05:17PM -0700, Jarno Rajahalme wrote:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
Brilliant!
Acked-by: Ben Pfaff b...@nicira.com
Pushed to master,
Jarno
On Sep 9, 2014, at 12:45 PM, Ben Pfaff b...@nicira.com wrote:
On Fri, Sep 05, 2014 at 04:05:16PM -0700, Jarno Rajahalme wrote:
The frag member in the Netlink interface is an uint8_t enumeration
type, not a bitrfield, so it should always be either fully masked or
Linux has an internal queue that temporarily holds packets transmitted to
certain network devices. If too many packets are transmitted to such
network devices within a single list of actions, then packets tend to get
dropped. Broadcast or flooded or multicast packets on bridges with
thousands of
I’ll try to make a habit of compiling with C=1 before posting…
Thanks for the review, fixed and pushed,
Jarno
On Sep 9, 2014, at 12:32 PM, Ben Pfaff b...@nicira.com wrote:
On Fri, Sep 05, 2014 at 04:05:15PM -0700, Jarno Rajahalme wrote:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
Applied,
Jarno
On Sep 9, 2014, at 11:29 AM, Ben Pfaff b...@nicira.com wrote:
On Fri, Sep 05, 2014 at 04:05:14PM -0700, Jarno Rajahalme wrote:
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
Reviewed-by: YAMAMOTO Takashi yamam...@valinux.co.jp
Acked-by: Ben Pfaff b...@nicira.com
---
Thank you for the review!
I did the proposed/agreed upon clarifications and pushed this to master.
Jarno
On Sep 9, 2014, at 1:14 PM, Ben Pfaff b...@nicira.com wrote:
On Tue, Sep 09, 2014 at 01:11:25PM -0700, Jarno Rajahalme wrote:
On Sep 8, 2014, at 5:06 PM, Ben Pfaff b...@nicira.com
Hi Guru,
The Windows definition refers to DDE monitors. I'm not sure you want to use
this semantics here.
Thanks,
Eitan
-Original Message-
From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Gurucharan Shetty
Sent: Tuesday, September 09, 2014 1:23 PM
To: dev@openvswitch.org
On Tue, Sep 9, 2014 at 3:20 PM, Eitan Eliahu elia...@vmware.com wrote:
Hi Guru,
The Windows definition refers to DDE monitors. I'm not sure you want to use
this semantics here.
ddeml.h gets included indirectly in our builds (likely through
windows.h?). ddeml.h already has a
MAX_MONITORS
Signed-off-by: Daniele Di Proietto ddiproie...@vmware.com
---
I didn't test the fix. If it works netdev-dpdk.o should appear in the travis
output (on master it doesn't)
Thanks,
Daniele
---
.travis.yml | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/.travis.yml b/.travis.yml
On Tue, Sep 09, 2014 at 03:17:00PM -0700, Daniele Di Proietto wrote:
Signed-off-by: Daniele Di Proietto ddiproie...@vmware.com
---
I didn't test the fix. If it works netdev-dpdk.o should appear in the travis
output (on master it doesn't)
I think that it corrects the build, but the build fails
On 9/9/14, 3:40 PM, Ben Pfaff b...@nicira.com wrote:
On Tue, Sep 09, 2014 at 03:17:00PM -0700, Daniele Di Proietto wrote:
Signed-off-by: Daniele Di Proietto ddiproie...@vmware.com
---
I didn't test the fix. If it works netdev-dpdk.o should appear in the
travis
output (on master it doesn't)
I realize that that we get a compilation error but this definition in the
ddeml.h has different semantics (and it is set to 4 rather to 8).
Perhaps you want to undefined it and have your own definition instead.
Thanks,
Eitan
-Original Message-
From: Gurucharan Shetty
OVS has had a MAX_MONITORS macro definition there forever. On
Windows, the name conflicts with something defined in a system header.
Therefore, Guru's patch renames the OVS macro to avoid a conflict.
On Tue, Sep 09, 2014 at 10:47:20PM +, Eitan Eliahu wrote:
I realize that that we get a
This reduces the need to keep counting the bits in the map.
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/classifier.c | 11 ---
lib/dpif-netdev.c | 11 +++
lib/flow.c| 36 ++--
lib/flow.h| 29
Make miniflow values 64-bit aligned, both in the starting address and
in length. This allows for 64-bit operations in some functions, such
as miniflow_hash().
Signed-off-by: Jarno Rajahalme jrajaha...@nicira.com
---
lib/classifier.c|2 +-
lib/dpif-netdev.c |7 ---
Samuel and I discussed about this patch over IRC, and we agreed that it is best
to split the patch into two:
1. Fix for handling NBL with mutiple NBs. Any refactoring of
OvsStartNBLIngress() can be done as part of this patch.
2. Updating the parsing, packet io and actions code to handle NBs
Daniele,
I just posted rebased versions of two patches on which I worked earlier in a
similar area as there two patches:
[PATCH 1/2] lib/flow: Add ‘miniflow.count’
[PATCH 2/2] lib/flow: Maintain miniflow values as 64-bit aligned.
I’d like to see if we could generalize the inlining benefits,
On Mon, Sep 8, 2014 at 8:53 PM, Joe Stringer joestrin...@nicira.com wrote:
Was there a particular tool or config flag that you used to detect this
error / generate this output?
I compile kernel with debugging options. CONFIG_PROVE_LOCKING detects
most of deadlocks in kernel, you do not need any
The upstream modules uses this_cpu_xxx APIs. Add those functions for
older kernel (3.0.0) that does not provide them.
VMware-BZ: #1319082
Signed-off-by: Andy Zhou az...@nicira.com
---
datapath/linux/compat/include/asm/percpu.h | 18 +-
1 file changed, 17 insertions(+), 1
This commit adds a new API to the 'struct netdev_class' which
allows user to query the numa node id the 'netdev' is on.
Currently, only netdev-dpdk module implements this function.
Signed-off-by: Alex Wang al...@nicira.com
---
PATCH - V2
- rebase and refactor the code.
V2 - V3:
- rebase.
---
This commit adds new variable n_txq to 'struct netdev' for recording
the number of tx queues. Correspondingly, the send_*() functions are
extended to accept queue id as input argument.
All 'netdev-*' implementation will ignore the queue id since having
multiple tx queues is not supported.
Before this commit, ovs creates one tx and one rx queue for
each dpdk interface and uses only one poll thread for handling
I/O of all dpdk interfaces. An upcoming patch will allow multiple
poll threads be created. As a preparation, this commit changes
the default to create multiple tx and rx
The previous commit makes OVS create one tx queue for each
cpu core, each pmd thread will use a separate tx queue.
Also, tx of non-pmd threads on dpdk interface is all through
'NON_PMD_THREAD_TX_QUEUE', protected by the 'nonpmd_mempool_mutex'.
Therefore, the spinlock is no longer needed. And this
Previous commit makes OVS create one tx queue for each cpu
core. An upcoming patch will allow multiple pmd threads be
created and pinned to cpu cores. So each pmd thread will use
the tx queue corresponding to its core id.
Moreover, the pmd threads running on different numa node than
the dpdk
With this commit, ovs by default will try creating 'number of
dpdk interfaces on numa node' pmd threads for each numa node
and pin the pmd threads to available cpu cores on the numa node.
NON_PMD_CORE_ID (currently 0) is used to reserve a particular
cpu core for the I/O of all non-pmd threads.
Yes, I see, my bad. MAXIMUM_MONITORS instead of MAX_MONTIORS.
Looks good.
Thanks,
Eitan
-Original Message-
From: Ben Pfaff [mailto:b...@nicira.com]
Sent: Tuesday, September 09, 2014 3:51 PM
To: Eitan Eliahu
Cc: Gurucharan Shetty; dev@openvswitch.org; Gurucharan Shetty
Subject: Re:
All NET_BUFFERs of a NET_BUFFER_LIST must go through the pipeline: extract,
find flow, execute. Previously, only the first NET_BUFFER of a NET_BUFFER_LIST
was going through this pipeline, which was erroneous.
OvsPartialCopyToMultipleNBLs is used to make each NET_BUFFER have its own
The original message was received at Wed, 10 Sep 2014 10:38:57 +0700
from openvswitch.org [64.194.198.15]
- The following addresses had permanent fatal errors -
dev@openvswitch.org
- Transcript of session follows -
... while talking to openvswitch.org.:
MAIL FROM:Returned mail
91 matches
Mail list logo
