I changed 'netdev' to 'dev' in netdev_dpdk_init(), added Jan to AUTHORS and
pushed this to master.
Thanks!
Daniele
2016-06-02 5:42 GMT-07:00 Zoltán Balogh :
> Hi Daniele,
>
> I fixed the patch based on your comments:
>
> OVS reports that link state of a vhost-user
On 02/06/2016 13:16, "Ben Pfaff" <b...@ovn.org> wrote:
>On Wed, Jun 01, 2016 at 07:22:48PM -0700, Daniele Di Proietto wrote:
>> In the run() function of ofproto-dpif we call rule_expire() for every
>> possible flow that has a timeout and rule_expire() calls ti
linux/in.h:28:3: error: got 0
/usr/include/linux/in.h:84:16: error: redefinition of struct in_addr
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
include/sparse/netinet/in.h | 40
1 file changed, 20 insertions(+), 20 deletions(-)
diff
in run(), to reduce the workload.
Keeping the flows ordered by expiration in some kind of heap or timing
wheel data structure could help make this process more efficient, if
rule_expire() turns out to be a bottleneck.
VMware-BZ: #1655122
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.
()'?
The "instant" part makes sense compared to the current code, but that
code is removed.
Acked-by: Daniele Di Proietto <diproiet...@vmware.com>
If there are no objection I can push this separately from the rest of
the series.
On 24/05/2016 06:34, "Ilya Maximets" <i.maxim...@
Hi Ilya,
apologies for the delay.
I didn't take a extremely detailed look at this series, but I have
a few high level comments.
Thanks for adding a command to configure the rxq affinity. Have
you thought about using the database instead? I think it will
be easier to use because it survives
On 25/05/2016 04:03, "Ilya Maximets" <i.maxim...@samsung.com> wrote:
>On 23.05.2016 17:55, Traynor, Kevin wrote:
>>> -Original Message-
>>> From: Ilya Maximets [mailto:i.maxim...@samsung.com]
>>> Sent: Tuesday, May 17, 2016 4:09 PM
&g
Thanks for the patch!
I'm not sure how to best handle the libnuma dependency. Question:
Is it still useful to move the device to a PMD thread on the appropriate
numa socket, even if DPDK is compiled without
CONFIG_RTE_LIBRTE_VHOST_NUMA? If it's useful, I'm fine with the
approach followed by this
t;
>> -Original Message-
>> From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Daniele
>> Di Proietto
>> Sent: Tuesday, May 17, 2016 1:56 AM
>> To: dev@openvswitch.org
>> Subject: [ovs-dev] [PATCH v3 00/16] Userspace (DPDK) connection
>>
Thanks for the patch!
I have a few comments inline, otherwise this looks good to me
2016-05-12 9:13 GMT-07:00 Zoltán Balogh :
> Hi,
>
> OVS reports that link state of a vhost-user port (type=dpdkvhostuser) is
> DOWN, even when traffic is running through the port
On 25/05/2016 07:19, "Fischetti, Antonio" <antonio.fische...@intel.com> wrote:
>
>
>> -Original Message-
>> From: Joe Stringer [mailto:j...@ovn.org]
>> Sent: Tuesday, May 24, 2016 7:26 PM
>> To: Fischetti, Antonio <antonio.fische..
On 24/05/2016 08:52, "Fischetti, Antonio" <antonio.fische...@intel.com> wrote:
>Hi Daniele, one question below.
>
>Thanks,
>Antonio
>
>
>> -Original Message-
>> From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Daniele
>>
Thanks for the patch!
I've added your name to AUTHORS and applied this to master
2016-05-25 8:03 GMT-07:00 wisd0me :
> When we're adding the port into ovs bridge, its mtu is updated to the
> minimal
> mtu of the included port. But when the port is getting removed, no
Thanks for the patch!
I realized that we also need to delete the policer in netdev_dpdk_destruct,
so I folded in the following:
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index 900ec64..01da275 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/netdev-dpdk.c
@@ -929,6 +929,8 @@
On 23/05/2016 08:07, "Fischetti, Antonio" <antonio.fische...@intel.com> wrote:
>Hi Daniele,
>I added one comment below.
>
>Regards,
>Antonio
>
>> -Original Message-
>> From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Daniele
&
On 23/05/2016 20:16, "Ben Pfaff" <b...@ovn.org> wrote:
>On Mon, May 23, 2016 at 05:49:04PM -0700, Daniele Di Proietto wrote:
>> Instead of hardcoding 'kill `cat pid`' on every call to AT_CHECK is
>> simpler to use on_exit.
>>
>> This makes sure
Hi Aaron,
I'm still a little bit nervous about calling chown on a (partially) user
controlled file name.
Before moving forward I wanted to discuss a couple of other options:
* Ansis (in CC) suggested using -runas parameter in qemu. This way qemu can
open the socket as root and drop
On 23/05/2016 14:14, "Ben Pfaff" <b...@ovn.org> wrote:
>On Mon, May 23, 2016 at 09:07:25PM +, Daniele Di Proietto wrote:
>> On 23/05/2016 14:00, "Ben Pfaff" <b...@ovn.org> wrote:
>>
>> >On Mon, May 23, 2016 at 01:43:47PM
ed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
tests/ovsdb-idl.at | 91 ++
1 file changed, 51 insertions(+), 40 deletions(-)
diff --git a/tests/ovsdb-idl.at b/tests/ovsdb-idl.at
index 1369a6f..057d09f 100644
--- a/tests/ovsdb-idl.at
Thanks for the patch, it's simpler than I expected (all the infrastructure
was already there).
Minor comments inline, otherwise this looks good to me.
If you agree with the comments, would you mind sending another version?
Daniele
2016-05-10 2:19 GMT-07:00 Ian Stokes :
>
I would drop the "__" suffix from the function names.
Otherwise this looks good to me
Thanks,
Daniele
2016-05-10 2:19 GMT-07:00 Ian Stokes :
> Add generic policer functions to avoid code duplication.
>
> Policing can be implemented on both egress and ingress paths.
>
On 23/05/2016 14:00, "Ben Pfaff" <b...@ovn.org> wrote:
>On Mon, May 23, 2016 at 01:43:47PM -0700, Daniele Di Proietto wrote:
>> Without the last OVSDB_SERVER_SHUTDOWN, the test hangs on travis,
>> because ovsdb-server is still running.
>>
>> Teste
Without the last OVSDB_SERVER_SHUTDOWN, the test hangs on travis,
because ovsdb-server is still running.
Tested-at: https://travis-ci.org/ddiproietto/ovs/builds/132343426
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
tests/ovsdb-idl.at | 1 +
1 file changed, 1 ins
Thanks for spotting this, fixed
On 15/05/2016 23:22, "Ilya Maximets" <i.maxim...@samsung.com> wrote:
>One typo inline.
>
>Best regards, Ilya Maximets.
>
>On 23.04.2016 04:03, Daniele Di Proietto wrote:
>> This introduces in dpif-netdev and netdev-dpdk the
On 16/05/2016 13:31, "Ben Pfaff" <b...@ovn.org> wrote:
>On Fri, Apr 22, 2016 at 06:03:02PM -0700, Daniele Di Proietto wrote:
>> ovs_mutex_cond_wait() is used in many functions in dpif-netdev to
>> synchronize with pmd threads, but we can't guarantee that
t;Acked-by: Ilya Maximets <i.maxim...@samsung.com>
>
>On 23.04.2016 04:02, Daniele Di Proietto wrote:
>> Currently we treat set_multiq() calls specially in netdev and dpif-netdev:
>> every pmd thread must be stopped and set_multiq() is allowed to destroy
Thanks for the series.
I fixed a warning with gcc-4.8:
../lib/netdev-dpdk.c: In function 'dpdk_init__':
../lib/netdev-dpdk.c:3197:27: error: 'err' may be used uninitialized in
this function [-Werror=maybe-uninitialized]
if (auto_determine && !err) {
^
cc1: all
tches the error at " 2032:
>ovn.at:1230 <http://ovn.at:1230> ovn -- 3 HVs, 3 LS, 3 lports/LS, 1 LR" is
>resolved.
>
>
>
>The ukey issue also disappears.
>some comments inline.
>
>
>On Tue, May 17, 2016 at 10:31 PM, Daniele Di Proietto
><diproie
We only stress the same code path in testcase "ovn -- 3 HVs, 3 LS,
3 lports/LS, 1 LR", which is slow to execute under valgrind.
It's probably worth adding a separate case.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
tests/tunnel-
Otherwise the tests can fail under heavy load (or with valgrind).
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
tests/tunnel-push-pop.at | 5 +
1 file changed, 5 insertions(+)
diff --git a/tests/tunnel-push-pop.at b/tests/tunnel-push-pop.at
index 64a0cbe..4695bb7
On 19/05/2016 20:59, "Ben Pfaff" <b...@ovn.org> wrote:
>On Tue, May 17, 2016 at 10:31:23PM -0700, Daniele Di Proietto wrote:
>> Almost every caller expects [mini]flow_hash_5tuple() to be able to deal
>> with all kinds of flows, not only TCP and UDP.
>>
&
On 19/05/2016 21:02, "Ben Pfaff" <b...@ovn.org> wrote:
>On Tue, May 17, 2016 at 10:31:25PM -0700, Daniele Di Proietto wrote:
>> The datapath code expects the RSS hash to always be initialized. This
>> is enforced by checking in emc_processing() that the h
On 19/05/2016 14:04, "William Tu" <u9012...@gmail.com> wrote:
>
>
>
>Reported-by: William Tu <u9012...@gmail.com>
>Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
>---
> lib/dpif-netdev.c | 5 +
> 1 file changed, 5 insert
Hi Flavio,
thanks for the patch, it looks good to me
Some minor comments inline
2016-05-04 11:49 GMT-07:00 Flavio Leitner :
> The PMD thread needs to keep processing RX queues in order
> to achieve maximum throughput. It also needs to sweep emc
> cache and quiesce which use
Hi William,
Could you see if this series fixes it?
http://openvswitch.org/pipermail/dev/2016-May/071190.html
Thanks,
Daniele
2016-05-17 9:04 GMT-07:00 William Tu :
> Hi Joe,
>
> I did a quick try and the error is still there.
>
> Regards,
> William
>
> On Mon, May 16,
(). This caused an uninitialized memory read, and a
random value to be assigned to the outer tunnel header source port.
Reported-by: William Tu <u9012...@gmail.com>
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/dpif-netdev.c | 5 +
1 file changed, 5 insertions(+)
of reading uninitialized memory.
Found by valgrind.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/flow.c | 41 -
1 file changed, 32 insertions(+), 9 deletions(-)
diff --git a/lib/flow.c b/lib/flow.c
index fc8843e..d09bde2 100644
---
in the userspace datapath to generate the RSS hash by
a following commit
* It can be used by the userspace connection tracker to avoid extracting
the l3 type again.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/dpif.c| 4 +++-
lib/
We only stress the same code path in testcase "ovn -- 3 HVs, 3 LS,
3 lports/LS, 1 LR", which is slow to execute under valgrind.
This makes it easier to reproduce a valgrind error, which is fixed by a
later commit.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
t
and to allow the pickup of already
established connections.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
COPYING | 1 +
debian/copyright.in | 4 +
include/openvswitch/types.h | 4 +
lib/automake.mk | 5 +
lib/conntrack-o
isn't supported yet (algo=ftp, for instance, or
>because it's an IP fragment), but otherwise I haven't noticed any
>other issues.
>
>Thanks,
>fbl
>
>
>On Fri, Apr 15, 2016 at 05:02:32PM -0700, Daniele Di Proietto wrote:
>> This series aims to implement the ct() action for
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
tests/system-traffic.at | 86 +
1 file changed, 86 insertions(+)
diff --git a/tests/system-traffic.at b/tests/system-traffic.at
index c3b8a54..ffb9d2e 100644
--- a/tests/
This is useful to test the connection tracker, which performs checksum
verification.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/flow.c| 46 +++-
tests/ofproto-dpif.at | 198 +-
2 files change
While the system testsuite already has connection tracking tests, it
will be still useful to add some to the standard testsuite because:
* They're run more often by developers.
* Some of them are more interesting for the userspace datapath.
Signed-off-by: Daniele Di Proietto <dipro
It will be used by connection tracking tests.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/flow.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/lib/flow.c b/lib/flow.c
index fbe754a..1417803 100644
--- a/lib/flow.c
+++ b/lib/flow.c
@@ -2206,6 +
Reported-by: Subramani Paramasivam <subramani.paramasi...@wipro.com>
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/automake.mk | 1 +
lib/conntrack-icmp.c| 105
lib/conntrack-private.h | 11 -
lib/con
New functions are implemented in the conntrack module to support this.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
Acked-by: Flavio Leitner <f...@sysclose.org>
---
lib/conntrack.c | 23 +++
lib/conntrack.h | 2 ++
lib/dpif-netdev.c | 10 +
different batching and locking strategies.
E.g. the line:
`./test/ovstest test-conntrack benchmark 1 1488 32`
starts 1 thread that will send 1488 packets to the connection
tracker, 32 at a time. It will print the time taken to process them.
Signed-off-by: Daniele Di Proietto <dipro
New functions are implemented in the conntrack module to support this.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
Acked-by: Flavio Leitner <f...@sysclose.org>
---
lib/conntrack-private.h | 3 ++
lib/conntrack-tcp.c | 34 +
lib/conntrack.c
The function simply returns the ethernet type of the packet (after
eventually discarding the VLAN tag). It will be used by a following
commit.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
Acked-by: Flavio Leitner <f...@sysclose.org>
---
lib/flow.c | 14 --
datapath checks for the IPv4 header checksum, so
fix those in the hardcoded packets.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
Acked-by: Joe Stringer <j...@ovn.org>
Acked-by: Flavio Leitner <f...@sysclose.org>
---
tests/system-kmod-macros.at | 28 +
This will be used by a future commit.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
Acked-by: Joe Stringer <j...@ovn.org>
Acked-by: Flavio Leitner <f...@sysclose.org>
---
lib/flow.c | 140 ++---
lib/flow.h
-off-by: Daniele Di Proietto <diproiet...@vmware.com>
Acked-by: Flavio Leitner <f...@sysclose.org>
---
tests/test-conntrack.c | 66 ++
1 file changed, 66 insertions(+)
diff --git a/tests/test-conntrack.c b/tests/test-conntrack.c
index de18
of the
corresponding 'timeout' list. This ways, the list is always ordered by
'expiration'.
When the cleanup thread iterates through the lists for expired
connections, it can stop at the first non expired connection.
Suggested-by: Joe Stringer <j...@ovn.org>
Signed-off-by: Daniele Di Pr
Linux and FreeBSD have slightly different names for these constants.
Windows doesn't define them. It is simpler to redefine them from
scratch for OVS. The new names are different than those used in Linux
and FreeBSD.
These definitions will be used by a future commit.
Signed-off-by: Daniele Di
This commit implements the OVS_ACTION_ATTR_CT action in dpif-netdev.
To allow ofproto-dpif to detect the conntrack feature, flow_put will not
discard anymore flows with ct_* fields set. We still shouldn't allow
flows with NAT bits set, since there is no support for NAT.
Signed-off-by: Daniele Di
ackets in the same batch.
* Added an ovs-test module to throw pcap files at the connection tracker.
* Added a workaround for the userspace testsuite on new kernels and a tcp
non-conntrack test.
Daniele Di Proietto (16):
packets: Define ICMP types.
flow: Export parse_ipv6_ext_hdrs().
flow: Int
2016-05-12 13:40 GMT-07:00 pravin shelar :
> On Thu, May 12, 2016 at 12:59 PM, Jesse Gross wrote:
> > On Thu, May 12, 2016 at 11:18 AM, pravin shelar wrote:
> >> On Tue, May 10, 2016 at 6:31 PM, Jesse Gross wrote:
> >>> I'm
Thanks for fixing this!
Acked-by: Daniele Di Proietto <diproiet...@vmware.com>
2016-05-10 15:50 GMT-07:00 Joe Stringer <j...@ovn.org>:
> Clang complains:
> lib/netdev-dpdk.c:1860:1: error: mutex 'dev->mutex' is not locked on every
> path
> through here [-Werr
Thanks for the review! Applied to master and branch-2.5
On 04/05/2016 11:28, "Jarno Rajahalme" <ja...@ovn.org> wrote:
>Acked-by: Jarno Rajahalme <jrajaha...@nicira.com>
>
>> On May 3, 2016, at 6:10 PM, Daniele Di Proietto <diproiet...@vmware.com>
>
Hi Christian,
Thanks for reporting this. We don't usually backport features to older
branches, but in this case I think in make sense to backport the whole commit,
given that it's pretty much self contained.
I applied the commit to branch-2.5.
Thanks,
Daniele
On 04/05/2016 07:49,
ix/sysv/linux/x86_64/clone.S:111
Fixes: 361d808dd9e4("flow: Split miniflow's map.")
CC: Jarno Rajahalme <ja...@ovn.org>
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/dpif-netdev.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/lib/dpif-netdev.c b/li
On 02/05/2016 13:11, "Jarno Rajahalme" <ja...@ovn.org> wrote:
>
>> On Apr 29, 2016, at 5:38 PM, Daniele Di Proietto <diproiet...@vmware.com>
>> wrote:
>>
>> flow_wc_map() should include 'tp_src' and 'tp_dst' for ICMPv6 packe
flow_wc_map() should include 'tp_src' and 'tp_dst' for ICMPv6 packet,
since they're used for ICMPv6 code and type.
This caused installed flows in the userspace datapath to always have
ICMPv6 code and type wildcarded (there are no other users of this
function).
Signed-off-by: Daniele Di Proietto
Hi Ciara,
thanks for doing this. I really think this has the potential to clean up
the netdev-dpdk code.
The clang thread safety analyzer reports some warnings:
CC lib/netdev-dpdk.lo
../lib/netdev-dpdk.c:882:1: error: mutex 'dpdk_mutex' is not held on every
path through here
en to the following
>folks (in alphabetical order):
>* Ben Pfaff (Reviews, vhost-sock-dir escape suggestion)
>* Christian Erhardt (Testing)
>* Daniele Di Proietto (Reviews, general suggestions)
>* Flavio Leitner(Original efforts, reviews)
>* Kevin Trayn
Hi Aaron, one more comment inline
The rest of the series looks good to me
Thanks,
Daniele
On 26/04/2016 12:42, "Aaron Conole" wrote:
>A previous patch introduced the ability to pass arbitrary EAL command
>line options via the dpdk_extras database entry. This commit
If vhost-sock-dir is empty, vhost_sock_dir will be NULL, because
process_vhost_flags() will return 0.
On 26/04/2016 12:42, "Aaron Conole" wrote:
>Since the vhost-user sockets directory now comes from the database, it is
>possible for any user with database access to
Hi Aaron,
thanks (again!) for this patch. Few comments:
* As I mentioned on my previous round of review, I don't think it's necessary
to pass the whole Open_vSwitch table to dpdk_init(). I.e., instead of doing
dpdk_init();
I'd prefer
if (cfg) {
dpdk_init(>other_config);
}
Thanks, pushed this to master!
On 20/04/2016 21:06, "Sairam Venugopal" <vsai...@vmware.com> wrote:
>Acked-by: Sairam Venugopal <vsai...@vmware.com>
>
>
>On 4/15/16, 5:04 PM, "Daniele Di Proietto" <diproiet...@vmware.com> wrote:
>
>>T
On 26/04/2016 16:38, "Flavio Leitner" <f...@sysclose.org> wrote:
>On Fri, Apr 15, 2016 at 05:02:45PM -0700, Daniele Di Proietto wrote:
>> The system userspace testsuite uses the userspace datapath with
>> netdev-linux devices, connected to veth
On 26/04/2016 08:24, "Ben Pfaff" <b...@ovn.org> wrote:
>On Fri, Apr 22, 2016 at 06:02:58PM -0700, Daniele Di Proietto wrote:
>> Makes popping each member of the hmap a bit easier.
>>
>> Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
On 26/04/2016 16:38, "Flavio Leitner" <f...@sysclose.org> wrote:
>On Fri, Apr 15, 2016 at 05:02:46PM -0700, Daniele Di Proietto wrote:
>> Useful to test the datapath ability to forward tcp packets without the
>> complexity of connection tracking.
>>
On 26/04/2016 16:37, "Flavio Leitner" <f...@sysclose.org> wrote:
>On Fri, Apr 15, 2016 at 05:02:41PM -0700, Daniele Di Proietto wrote:
>> This commit implements the OVS_ACTION_ATTR_CT action in dpif-netdev.
>>
>> To allow ofproto-dpif to d
++---+
>
>
>
>ALWAYS_INLINE
>=
>I've also checked what we already discussed with the patch set v1.
>I applied patches one by one and measured the performance, it seems
>that ALWAYS_INLINE in parse_vlan() and parse_ethertype() has fixed
>
Thank you for your comments, Flavio!
On 26/04/2016 16:35, "Flavio Leitner" <f...@sysclose.org> wrote:
>On Fri, Apr 15, 2016 at 05:02:36PM -0700, Daniele Di Proietto wrote:
>> This commit adds the conntrack module.
>>
>> It is a connection tracker tha
On 26/04/2016 16:38, "Flavio Leitner" <f...@sysclose.org> wrote:
>On Fri, Apr 15, 2016 at 05:02:43PM -0700, Daniele Di Proietto wrote:
>> Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
>
>This could be squashed with "[07/15] con
On 26/04/2016 16:36, "Flavio Leitner" <f...@sysclose.org> wrote:
>On Fri, Apr 15, 2016 at 05:02:37PM -0700, Daniele Di Proietto wrote:
>> This introduces a very limited but simple benchmark for
>> conntrack_execute(). It just sends repeatedly the s
ssage-
>> From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Daniele Di
>> Proietto
>> Sent: Saturday, April 16, 2016 1:03 AM
>> To: dev@openvswitch.org
>> Subject: [ovs-dev] [PATCH v2 04/15] conntrack: New userspace connection
>> tracker.
>>
>&g
Thanks for the detailed review Joe, replies inline
On 19/04/2016 14:36, "Joe Stringer" <j...@ovn.org> wrote:
>On 15 April 2016 at 17:02, Daniele Di Proietto <diproiet...@vmware.com> wrote:
>> This commit adds the conntrack module.
>>
>> It is
On 19/04/2016 15:15, "Joe Stringer" <j...@ovn.org> wrote:
>On 15 April 2016 at 17:02, Daniele Di Proietto <diproiet...@vmware.com> wrote:
>> The system userspace testsuite uses the userspace datapath with
>> netdev-linux devices, connected to
On 19/04/2016 17:24, "Joe Stringer" <j...@ovn.org> wrote:
>On 19 April 2016 at 16:53, Joe Stringer <j...@ovn.org> wrote:
>> On 15 April 2016 at 17:02, Daniele Di Proietto <diproiet...@vmware.com>
>> wrote:
>>> The userspace connection trac
On 26/04/2016 16:35, "Flavio Leitner" <f...@sysclose.org> wrote:
>On Fri, Apr 15, 2016 at 05:02:35PM -0700, Daniele Di Proietto wrote:
>> The function simply returns the ethernet type of the packet (after
>> eventually discarding the VLAN tag). It will b
On 19/04/2016 13:03, "Joe Stringer" <j...@ovn.org> wrote:
>On 15 April 2016 at 17:02, Daniele Di Proietto <diproiet...@vmware.com> wrote:
>> This will be used by a future commit.
>>
>> Signed-off-by: Daniele Di Proietto <dipro
On 19/04/2016 14:52, "Joe Stringer" <j...@ovn.org> wrote:
>On 15 April 2016 at 17:02, Daniele Di Proietto <diproiet...@vmware.com> wrote:
>> This commit implements the OVS_ACTION_ATTR_CT action in dpif-netdev.
>>
>> To allow ofproto-dpi
On 19/04/2016 14:54, "Joe Stringer" <j...@ovn.org> wrote:
>On 15 April 2016 at 17:02, Daniele Di Proietto <diproiet...@vmware.com> wrote:
>> Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
>
>Could be combined with the earlier patch that
On 19/04/2016 14:49, "Joe Stringer" <j...@ovn.org> wrote:
>On 15 April 2016 at 17:02, Daniele Di Proietto <diproiet...@vmware.com> wrote:
>> Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
>
>
>
>> +static void
>
Good idea, I'll combine them
Thanks for the review!
On 19/04/2016 14:40, "Joe Stringer" <j...@ovn.org> wrote:
>On 15 April 2016 at 17:02, Daniele Di Proietto <diproiet...@vmware.com> wrote:
>> Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com&g
to the proper NUMA socket.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/dpif-netdev.c | 69 +-
lib/netdev-bsd.c | 2 +-
lib/netdev-dpdk.c | 195 ++
lib/netdev-dummy.c| 2 +-
lib/netdev-l
}
>+dp_netdev_add_port_to_pmds__(dp, port, _reload);
> }
>
> HMAPX_FOR_EACH (node, _reload) {
>--
>
>Best regards, Ilya Maximets.
>
>On 20.04.2016 01:28, diproiettod at vmware.com (Daniele Di Proietto)
>wrote:
>> A future commit
eds to synchronize with
every pmd thread.
Among the advantages, keeping a per thread port mapping could allow
greater control over the txq assigment.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/dpif-netdev.c | 348 ++
1 file cha
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/dpif-netdev.c | 144 ++--
lib/dpif-provider.h | 3 +-
2 files changed, 84 insertions(+), 63 deletions(-)
diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
index 5
to the RCU model we use in userspace.
This commit changes the port container from cmap to hmap. 'port_mutex'
must be held by readers and writers. This shouldn't have performace
impact, as readers in the fast path use a thread local cache.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.
An upcoming commit will make dpif_poll_threads_set() record the
requested configuration and dpif_run() apply it, so it makes sense to
change the order.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
Tested-by: Ilya Maximets <i.maxim...@samsung.com>
Acked-by: Ilya Maxim
state, so explicit calls to
ovsrcu_quiesce_start() and ovsrcu_quiesce_end() are added there.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/ovs-rcu.h | 3 +--
lib/ovs-thread.c| 2 --
vswitchd/system-stats.c | 6 ++
3 files changed, 7 insertions
This commit introduces some functions to add/remove rxqs from pmd
threads without reloading them. They will be used by next commits.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/dpif-netdev.c | 77 ---
1 file chang
Errors returned by netdev_set_multiq() and netdev_rxq_open() weren't
handled properly in reconfigure_pmd_threads(). In case of error now we
remove the port from the datapath.
Also, part of the code is moved in a new function, port_reconfigure().
Signed-off-by: Daniele Di Proietto <dipro
Makes popping each member of the hmap a bit easier.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/cfm.c| 5 ++---
lib/hmap.h | 20
lib/id-pool.c| 5 ++---
lib/learning-switch.c
rx and tx queues in
netdev-dpdk.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
Tested-by: Ilya Maximets <i.maxim...@samsung.com>
Acked-by: Ilya Maximets <i.maxim...@samsung.com>
Acked-by: Mark Kavanagh <mark.b.kavan...@intel.com>
---
lib/netdev-bsd.c |
The interface will be more similar to the cmap.
Signed-off-by: Daniele Di Proietto <diproiet...@vmware.com>
---
lib/hmap.c | 26 --
lib/hmap.h | 7 ++-
lib/sset.c | 12 +---
lib/sset.h | 7 ++-
o
501 - 600 of 1432 matches
Mail list logo