On 31.01.2019 11:48, Lilijun wrote:
> This patch fix the dead lock when using dpdk userspace datapath. The
> problem is described as follows:
> 1) when add or delete port, the main thread will call
> reconfigure_datapath() in the function dpif_netdev_run()
> 2) Here the dp->port_mutex is locked.
DPDK 18.11 uses libverbs and libmlx4 when MLX4 PMD is enabled.
This commit makes OVS to link to libverbs and libmlx4 when MLX4 PMD is
enabled on DPDK.
Signed-off-by: Eli Britstein
Reviewed-by: Asaf Penso
---
acinclude.m4 | 15 +++
1 file changed, 15 insertions(+)
diff --git
This patch set automatically links with the necessary libraries for
MLX4/5 DPDK PMDs.
Patch 1 automatically links with the necessary libraries for MLX5 PMD
Patch 2 automatically links with the necessary libraries for MLX4 PMD
Patch 3 removes unnecessary define previously done for automatically
Commit fc3b425fa02f ("acinclude: Include libmnl when needed") added
unnecessary include of DPDK_MNL. Omit it.
Fixes: fc3b425fa02f ("acinclude: Include libmnl when needed")
Signed-off-by: Eli Britstein
---
acinclude.m4 | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git
DPDK 18.11 uses libverbs and libmlx5 when MLX5 PMD is enabled.
This commit makes OVS to link to libverbs and libmlx5 when MLX5 PMD is
enabled on DPDK.
Signed-off-by: Eli Britstein
Reviewed-by: Shahaf Shuler
Reviewed-by: Asaf Penso
---
acinclude.m4 | 24
1 file
On 2/5/2019 10:23 PM, Ben Pfaff wrote:
> On Tue, Feb 05, 2019 at 10:22:09AM -0800, Gregory Rose wrote:
>> On 2/5/2019 4:02 AM, Eli Britstein wrote:
>>> On 2/4/2019 10:07 PM, David Miller wrote:
From: Yi-Hung Wei
Date: Mon, 4 Feb 2019 10:47:18 -0800
> For example, to see how
On Wed, Feb 6, 2019 at 1:22 AM David Marchand
wrote:
> On Mon, Feb 4, 2019 at 6:56 PM Ben Pfaff wrote:
>
>> On Thu, Jan 31, 2019 at 11:35:41PM -0800, Darrell Ball wrote:
>> > The call to inet_ntop() in repl_ftp_v6_addr() is 1 short to handle
>> > the maximum possible V6 address size for v4
Fragmentation handling is added for supporting conntrack.
Both v4 and v6 are supported.
After discussion with several people, I decided to not store
configuration state in the database to be more consistent with
the kernel in future, similarity with other conntrack configuration
which will not be
Signed-off-by: Darrell Ball
---
lib/ovs-atomic.h | 36
1 file changed, 36 insertions(+)
diff --git a/lib/ovs-atomic.h b/lib/ovs-atomic.h
index 21e230e..11fa192 100644
--- a/lib/ovs-atomic.h
+++ b/lib/ovs-atomic.h
@@ -479,6 +479,42 @@
Use 'must' instead of 'should'.
Suggested-by: Justin Pettit
Signed-off-by: Darrell Ball
---
lib/conntrack.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/conntrack.c b/lib/conntrack.c
index bd5b7a6..b92dee5 100644
--- a/lib/conntrack.c
+++ b/lib/conntrack.c
@@ -1297,7
Acked-by: Justin Pettit
Signed-off-by: Darrell Ball
---
tests/system-traffic.at | 2 ++
1 file changed, 2 insertions(+)
diff --git a/tests/system-traffic.at b/tests/system-traffic.at
index 3a62e17..6e66946 100644
--- a/tests/system-traffic.at
+++ b/tests/system-traffic.at
@@ -2978,6 +2978,7 @@
Acked-by: Justin Pettit
Signed-off-by: Darrell Ball
---
lib/conntrack.c | 4 ++--
lib/flow.c | 44 ++--
lib/flow.h | 3 ++-
3 files changed, 38 insertions(+), 13 deletions(-)
diff --git a/lib/conntrack.c b/lib/conntrack.c
index
Acked-by: Justin Pettit
Signed-off-by: Darrell Ball
---
lib/dp-packet.h | 16
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/lib/dp-packet.h b/lib/dp-packet.h
index 7b85dd9..a6e2a4f 100644
--- a/lib/dp-packet.h
+++ b/lib/dp-packet.h
@@ -509,28 +509,28 @@
Fragmentation support for userspace datapath conntrack is added; both
v4 and v6 are supported. See the patches for additional details.
Fragmentation tests for the userspace datapath are enabled
by the patches and other test enhancements are added.
v10: Addressed Ben's review comments.
Thanks very much for the thorough review
sorry; I made the changes last year and then ran into vacation and internal
priorities.
On Tue, Dec 11, 2018 at 8:15 AM Ben Pfaff wrote:
> On Mon, Nov 19, 2018 at 11:09:25AM -0800, Darrell Ball wrote:
> > Fragmentation handling is added for supporting
I'd like to hear some kind of overall use case here. Sure, you can use
it to identify an OVN ACL, or a security group, or anything else. How
does that contribute to a larger system? There should be a hint to the
reader about how and why to use it.
On Wed, Feb 06, 2019 at 10:06:46PM +,
On Tue, Feb 5, 2019 at 4:41 PM Ben Pfaff wrote:
> On Tue, Feb 05, 2019 at 04:26:56PM -0800, Ashish Varma wrote:
> > In the parse_flow_monitor_request(), usable_protocols is an out
> parameter.
> > (and is set in parse_flow_monitor_request() only if a match field is
> > provided. )
> >
Hi Ben,
Thanks a lot for review.
Sure, V2 will have all the comments addressed.
Reason for using 128 bits:
a. Connection tracker has only 2 fields for metadata, ct.mark(32 bits) and
ct.label(128 bits).
b. Purpose of this series is to ensure that we use smaller field, i.e ct.mark
for flags and
Hi Ben,
Thanks for the review.
a. My bad on the Signed off by, will take care of it in V2.
b. Sure, I will add the test case as well.
c. I will update the ovn-sb.xml as well.
d. Yes, openflow does not restrict the usage of registers for connection
tracker, I think probably because of lack of use
Hi Ben,
Thanks a lot for the review.
Sure, I will add comments in logical-fields.c explaining the reason for
retaining ct_label.blocked.
I will rename ct_mark.blocked to ct.blocked as well.
V2 will have all these changes.
Thanks again for review.
Regards,
Ankur
-Original Message-
"Stokes, Ian" writes:
>> DPDK 18.11 builds using the more modern meson build system no more provide
>> the -ldpdk linker script. Instead it is expected to use pkgconfig for
>> linker options as well.
>>
>> This change will set DPDK_LIB from pkg-config (if pkg-config was
>> available) and since
Christian Ehrhardt writes:
> DPDK 18.11 builds using the more modern meson build system no more
> provide the -ldpdk linker script. Instead it is expected to use
> pkgconfig for linker options as well.
>
> This change will set DPDK_LIB from pkg-config (if pkg-config was
> available) and since
It looks like we fell off ovs-dev somehow. I've added it back.
On Wed, Feb 06, 2019 at 05:50:39PM +, Venugopal Iyer wrote:
> Thanks, Ben!
>
> I missed and {} for an if; have fixed it; thanks for catching it!
>
> As for the TODOS:
> - ovn/controller/chassis.c was something I just wanted to
On 06.02.2019 18:05, Aaron Conole wrote:
> Ilya Maximets writes:
>
>> On 05.02.2019 23:19, Flavio Leitner wrote:
>>> On Tue, Feb 05, 2019 at 06:43:08PM +0300, Ilya Maximets wrote:
Hi Flavio.
Thanks for taking a look.
On 05.02.2019 15:38, Flavio Leitner wrote:
>
> Hi
Having a single structure allows to simplify the code path and
clear all the items at once (probably faster). This does not
increase stack memory usage because all the L4 related items
grouped in a union.
Changes:
- Memsets combined.
- 'ipv4_next_proto_mask' dropped as we already know the
On 1/31/2019 2:47 AM, Li RongQing wrote:
netdev_dpdk_filter_packet_len() does not need to be protected
by tx_q[].tx_lock, and tx_q[].tx_lock can not protect it too,
same to netdev_dpdk_qos_run
so move them out of this lock to improve the scalability
Thanks for the Patch Li, can you give more
On 06.02.2019 18:16, Ophir Munk wrote:
> Hi,
> Please find comments inline.
>
>> -Original Message-
>> From: ovs-dev-boun...@openvswitch.org > boun...@openvswitch.org> On Behalf Of Ilya Maximets
>> Sent: Wednesday, February 6, 2019 3:19 PM
>> To: ovs-dev@openvswitch.org; Ian Stokes
>>
Hi.
See comments inline.
Best regards, Ilya Maximets.
On 06.02.2019 7:01, Anju Thomas wrote:
>
> Hi Ben/Ilya,
>
> I have addressed the comments in the below patch. Can you tell me if this is
> fin>
> Regards
> Anju
> -Original Message-
> From: Anju Thomas
Hi,
Please find comments inline.
> -Original Message-
> From: ovs-dev-boun...@openvswitch.org boun...@openvswitch.org> On Behalf Of Ilya Maximets
> Sent: Wednesday, February 6, 2019 3:19 PM
> To: ovs-dev@openvswitch.org; Ian Stokes
> Cc: Ilya Maximets ; Roni Bar Yanai
>
> Subject:
Ilya Maximets writes:
> On 05.02.2019 23:19, Flavio Leitner wrote:
>> On Tue, Feb 05, 2019 at 06:43:08PM +0300, Ilya Maximets wrote:
>>> Hi Flavio.
>>> Thanks for taking a look.
>>>
>>> On 05.02.2019 15:38, Flavio Leitner wrote:
Hi Ilya,
Thanks for the patch and I think we
Having a single structure allows to simplify the code path and
clear all the items at once (probably faster). This does not
increase stack memory usage because all the L4 related items
grouped in a union.
Changes:
- Memsets combined.
- 'ipv4_next_proto_mask' calculated at the top as we
On 2/5/2019 1:18 PM, Ilya Maximets wrote:
On 05.02.2019 16:09, Asaf Penso wrote:
Regards,
Asaf Penso
-Original Message-
From: Ilya Maximets
Sent: Tuesday, February 5, 2019 1:46 PM
To: Asaf Penso ; ovs-dev@openvswitch.org
Cc: Roni Bar Yanai ; Stokes, Ian
; Finn Christensen
Subject:
On Mon, Feb 4, 2019 at 6:56 PM Ben Pfaff wrote:
> On Thu, Jan 31, 2019 at 11:35:41PM -0800, Darrell Ball wrote:
> > The call to inet_ntop() in repl_ftp_v6_addr() is 1 short to handle
> > the maximum possible V6 address size for v4 mapping case.
> >
> > Found by inspection.
> >
> > Fixes:
On 05.02.2019 23:19, Flavio Leitner wrote:
> On Tue, Feb 05, 2019 at 06:43:08PM +0300, Ilya Maximets wrote:
>> Hi Flavio.
>> Thanks for taking a look.
>>
>> On 05.02.2019 15:38, Flavio Leitner wrote:
>>>
>>> Hi Ilya,
>>>
>>> Thanks for the patch and I think we knew about that pain when we
>>>
34 matches
Mail list logo
