Compliments of the day to you and I hope this mail will meet you in good
health. With Due Respect, I want to introduce myself. I am Mass Anderson from
Belgium, Working for South coast Inc. South Coast Inc is an investment
sub-company under Capital One Bank U.K. South Coast Inc handles all
> >>
> >> This is potentially dangerous from the future modifications and hard to
> >> read
> for
> >> reviewer/person who tries to understand how it works.
> >>
> >> Current implementation will fail if someone will change the logic of
> >> 'DP_PACKET_BATCH_REFILL_FOR_EACH', for example.
> >>
>
OVS reads packets in batches from a given port and packets in the
batch are subjected to potentially 3 levels of lookups to identify
the datapath megaflow entry (or flow) associated with the packet.
Each megaflow entry has a dedicated buffer in which packets that match
the flow classification
Ben Pfaff writes:
> On GNU systems these macros work with arbitrary pointers, but the relevant
> standards only require IN6_IS_ADDR_MC_LINKLOCAL to work with in6_addr (and
> don't specify IN6_ARE_ADDR_EQUAL at all). Make the "sparse"
> implementations correspondingly pickier so that we catch
Thanks for the patch Yifeng
I gave the patch a spin.
Few datapoints:
1/ The new test fails for both userspace and kernel; at different points in
the test, lines 377 and 395
respectively; I later tried increasing the sleep time to 5 seconds, but
it did not seem to help.
2/ The test runs
On Wed, 27 Jun 2018 at 10:59, Qiuyu Xiao wrote:
>
> This patch adds CA-cert based authentication to the ovs-monitor-ipsec
> daemon. With CA-cert based authentication enabled, OVS approves IPsec
> tunnel if the peer has a cert signed by a trusted CA and the identity of
> the peer cert is as
Currently check-kmod's gre test is broken on certain kernel
versions where ovs's gre module conflits with kernel's gre
module. But at the same time, present gre test depends on
Linux gre module to setup gre port.
This patch repairs the gre test by completely removing the
dependancy of kernel's
Ben Pfaff writes:
> Fixes test 815 "tunnel_push_pop_ipv6 - action".
>
> CC: Aaron Conole
> Fixes: 6f231f7c3a9e ("xlate: use const struct in6_addr in linklocal check")
> Signed-off-by: Ben Pfaff
> ---
D'oh! The names killed me. Sorry :-(
Acked-by: Aaron Conole
On Tue, Jul 10, 2018 at 4:27 PM, Yifeng Sun wrote:
> Looks good to me, thanks.
>
> Reviewed-by: Yifeng Sun
>
>
>
> On Tue, Jul 10, 2018 at 10:50 AM, William Tu wrote:
>>
>> Commit 9f57c67c379d ("gre: Remove support for sharing GRE protocol hook")
>> allows only single GRE packet receiver. When
sorry, I only had a few minutes in the morning
I sent a V2, with a proper commit message.
https://patchwork.ozlabs.org/patch/942279/
Thanks Darrell
On Tue, Jul 10, 2018 at 1:27 PM, Ben Pfaff wrote:
> On Tue, Jul 10, 2018 at 09:15:09AM -0700, Darrell Ball wrote:
> > Fixes: bd5e81a0e596
When conn_update_state() returns true, conn has been freed, so skip calling
handle_ftp_ctl() with this conn and instead follow code path for new
connections.
Fixes: bd5e81a0e596 ("Userspace Datapath: Add ALG infra and FTP.")
Signed-off-by: Darrell Ball
---
Needs backporting as far back as 2.8.
Looks good to me, thanks.
Reviewed-by: Yifeng Sun
On Tue, Jul 10, 2018 at 10:50 AM, William Tu wrote:
> Commit 9f57c67c379d ("gre: Remove support for sharing GRE protocol hook")
> allows only single GRE packet receiver. When upstream kernel's gre module
> is loaded, the gre.ko exclusively
Bleep bloop. Greetings Martin Xu, I am a robot and I have tried out your patch.
Thanks for your contribution.
I encountered some error that I wasn't expecting. See the details below.
git-am:
fatal: corrupt patch at line 127
Repository lacks necessary blobs to fall back on 3-way merge.
Cannot
On Tue, Jul 10, 2018 at 03:24:17PM -0700, Ben Pfaff wrote:
> This is a partial backport of commit 396d492cfa8d ("Don't shadow
> variables.") to fix a build break due to backporting a different commit
> that depended on it.
>
> CC: Mark Michelson
> Fixes: 15fbc3baee5a ("ovn: Add router load
Bleep bloop. Greetings Ben Pfaff, I am a robot and I have tried out your patch.
Thanks for your contribution.
I encountered some error that I wasn't expecting. See the details below.
checkpatch:
ERROR: Too many signoffs; are you missing Co-authored-by lines?
Lines checked: 30, Warnings: 0,
On 7/10/2018 3:42 PM, Martin Xu wrote:
This patch only affects rhel6 spec file.
RHEL 7.4 introduced backward incompatible changes in the kernel. As
a result, prebuilt PRM packages against kernels newer than 693.17.1
will cannot be used on systems with older kernels, vice versa.
This patch
This patch only affects rhel6 spec file.
RHEL 7.4 introduced backward incompatible changes in the kernel. As
a result, prebuilt PRM packages against kernels newer than 693.17.1
will cannot be used on systems with older kernels, vice versa.
This patch allows multiple kernel version numbers
Oops. I sent a fix:
https://patchwork.ozlabs.org/patch/942271/
I forgot to add you in Reported-by but I'll do that before pushing.
On Tue, Jul 10, 2018 at 06:44:15PM -0300, Flavio Leitner wrote:
>
> Hi Ben and Mark,
>
> This patch broke 2.9 build over here, see below.
>
> On Tue, Jul
This is a partial backport of commit 396d492cfa8d ("Don't shadow
variables.") to fix a build break due to backporting a different commit
that depended on it.
CC: Mark Michelson
Fixes: 15fbc3baee5a ("ovn: Add router load balancer undnat rule for IPv6")
Signed-off-by: Ben Pfaff
---
On 7/10/2018 10:50 AM, William Tu wrote:
Commit 9f57c67c379d ("gre: Remove support for sharing GRE protocol hook")
allows only single GRE packet receiver. When upstream kernel's gre module
is loaded, the gre.ko exclusively becomes the only gre packet receiver,
preventing OVS kernel module from
On Tue, Jul 10, 2018 at 11:59:35AM +0530, Sriharsha Basavapatna via dev wrote:
> This is the first patch in the patch-set to support dynamic rebalancing
> of offloaded flows.
>
> The patch detects OOR condition on a netdev port when ENOSPC error is
> returned by TC-Flower while adding a flow
On Fri, Jun 01, 2018 at 08:17:07AM +, Rohith Basavaraja wrote:
> This patch adds unit test cases for Improved Drop statistics feature
> request raised by following
> https://patchwork.ozlabs.org/patch/918934/
>
> Signed-off-by: Rohith Basavaraja
Where is patch 1/2?
Thanks,
Ben.
On Fri, Jun 01, 2018 at 07:03:55PM +0200, Eelco Chaudron wrote:
> This patch will make sure VXLAN tunnels with and without the group
> based policy (GBP) option enabled can not coexist on the same
> destination UDP port.
>
> In theory, VXLAN tunnel with and without GBP enables can be
>
Hi Ben and Mark,
This patch broke 2.9 build over here, see below.
On Tue, Jul 10, 2018 at 02:24:46PM -0700, Ben Pfaff wrote:
> Done.
>
> On Tue, Jul 10, 2018 at 10:41:32AM -0400, Mark Michelson wrote:
> > Hi Ben,
> >
> > Thanks for applying this patch to master. Can you please apply the
On Thu, Jul 05, 2018 at 08:11:33PM -0700, Darrell Ball wrote:
> Filter out packet-ins for V6 packets as this is a V4 test.
>
> Signed-off-by: Darrell Ball
Applied to master, thanks!
___
dev mailing list
d...@openvswitch.org
On Tue, 10 Jul 2018 14:23:59 -0700
Ben Pfaff wrote:
> On Thu, Jun 14, 2018 at 01:50:16PM -0700, Ben Pfaff wrote:
> > On Thu, May 31, 2018 at 04:52:41PM +0200, Timothy Redaelli wrote:
> > > This partially reverts commit
> > > d53d2f4b84cc77868f2a55d66417fe6058d0.
> > >
> > > Python2 check
You have been a great help Guru. thank you.
On Tue, Jul 10, 2018 at 5:25 PM, Guru Shetty wrote:
> I would suggest to create a new topic with a valid subject name to attract
> the dpdk folks and post it in disc...@openvswitch.org
>
> On 10 July 2018 at 13:44, Sandeep Adapala
> wrote:
>
>> Hello
I would suggest to create a new topic with a valid subject name to attract
the dpdk folks and post it in disc...@openvswitch.org
On 10 July 2018 at 13:44, Sandeep Adapala
wrote:
> Hello Guru,
>
> Now I have installed DPDK and OVS both of them are compatible with each
> other but now I got stuck
Done.
On Tue, Jul 10, 2018 at 10:41:32AM -0400, Mark Michelson wrote:
> Hi Ben,
>
> Thanks for applying this patch to master. Can you please apply the patch to
> the 2.9 branch as well?
>
> Thank you,
> Mark
>
> On 06/26/2018 02:45 PM, Mark Michelson wrote:
> >A note: if approved, this patch
On Thu, Jun 14, 2018 at 01:50:16PM -0700, Ben Pfaff wrote:
> On Thu, May 31, 2018 at 04:52:41PM +0200, Timothy Redaelli wrote:
> > This partially reverts commit d53d2f4b84cc77868f2a55d66417fe6058d0.
> >
> > Python2 check for "six.moves.range" is not reverted.
> >
> > Signed-off-by: Timothy
On Thu, Jun 21, 2018 at 06:39:16PM +0100, Tiago Lam wrote:
> When adding a route to a bridge, by executing "$appctl ovs/route/add
> $IP/$MASK $BR", a reference to the existing netdev is taken and stored
> in an instantiated ip_dev struct which is then stored in an addr_list
> list in tnl-ports.c.
Hmm, that's interesting. What was the problem, then?
On Fri, Jun 29, 2018 at 01:14:38AM +, Linhaifeng wrote:
> Sorry.This is not ovs‘s bug.
>
> 发件人: Linhaifeng
> 发送时间: 2018年6月21日 13:54
> 收件人: 'd...@openvswitch.org'
> 主题: 答复: [bug] ovs crash when call xlate_normal_flood
>
> Or stop upcall
On Thu, Jul 05, 2018 at 01:44:33PM -0700, Ravi Kerur wrote:
> During scale flow add/delete (>10K), I am seeing ovs-vswitchd cpu usage
> spike to 100% and stay there without any sign of returning to normal cpu
> usage. It's normal OVS 2.4.1 and no DPDK involved. I am trying to get
> 'perf' working
Can you explain further? Throwing away the cache entries makes the
cache less useful.
Are you really using v2.7.0? None of the line numbers match up, either
in the backtrace or in the patch.
On Tue, Jun 26, 2018 at 10:38:50AM +, wangyunjian wrote:
> I think the function xlate_cache_clear()
Hello Guru,
Now I have installed DPDK and OVS both of them are compatible with each
other but now I got stuck at new place. I know you are not very familiar
with DPDK but can you add someone who can help me out in that.
I have configured the OVS to use dpdk library using
./configure
This was useful for testing commit 4fe080160685 ("flow: Fix buffer overread
for crafted IPv6 packets.").
Signed-off-by: Ben Pfaff
---
v1->v2: Updated commit message.
utilities/ovs-ofctl.c | 19 +++
1 file changed, 19 insertions(+)
diff --git a/utilities/ovs-ofctl.c
On Wed, Jul 11, 2018 at 12:56:39AM +0530, nusid...@redhat.com wrote:
> From: Numan Siddique
>
> The python function ovs.socket_util.check_connection_completion() uses
> select()
> (provided by python) to monitor the socket file descriptor. The select()
> returns 1 when the file descriptor
This got fixed, right?
On Fri, Jun 29, 2018 at 02:59:39PM +0200, Daniel Alvarez Sanchez wrote:
> Yes, let's hope we can get it in soon... expecting an v3 from Darrell
> apparently.
> Thanks!
>
> On Fri, Jun 29, 2018 at 2:15 PM Lucas Alvares Gomes
> wrote:
>
> > Hi,
> >
> > > this should be
Thanks Ilya and Darrell. I applied this to master and backported as far
as branch-2.8.
On Tue, Jul 10, 2018 at 09:17:42AM -0700, Darrell Ball wrote:
> Good, thanks.
>
> Acked-by: Darrell Ball
>
> This needs backporting to 2.8.
>
> I re-ran some SA tools and found some other potential use
On Tue, Jul 10, 2018 at 09:15:09AM -0700, Darrell Ball wrote:
> Fixes: bd5e81a0e596 ("Userspace Datapath: Add ALG infra and FTP.")
> Signed-off-by: Darrell Ball
This could use a little longer commit message. Is the following
accurate:
When conn_update_state() returns true, it has freed 'conn'
On Tue, Jul 10, 2018 at 11:28:16PM +0530, Numan Siddique wrote:
> On Mon, Jul 9, 2018 at 11:36 PM Ben Pfaff wrote:
>
> > On Sun, Jul 08, 2018 at 10:05:41PM +0530, nusid...@redhat.com wrote:
> > > From: Numan Siddique
> > >
> > > Calling
On Tue, Jul 10, 2018 at 07:46:55PM +0100, Ian Stokes wrote:
> This commit fixes 'maybe-uninitialized' warnings for pointers in various
> functions in ovn-nbctl when compiling with gcc 6.3.1 and -Werror.
> Pointers to structs nbrec_logical_switch, nbrec_logical_switch_port,
> nbrec_logical_router
This makes it easier to reason about the outputs, and fixes warnings for
GCC 6.3.x.
Cc: Venkata Anil
Fixes: 31114af758c7 ("ovn-nbctl: Update logical router port commands.")
Fixes: 80f408f4cffb ("ovn: Use Logical_Switch_Port in NB.")
Fixes: 36f232bca2db ("ovn: l3ha, CLI for logical router port
On 7/9/2018 5:54 PM, Ben Pfaff wrote:
On Thu, Jul 05, 2018 at 09:29:37PM +, Stokes, Ian wrote:
On Thu, Jul 05, 2018 at 09:29:12PM +0100, Ian Stokes wrote:
On 6/27/2018 6:58 PM, Qiuyu Xiao wrote:
This patch series reintroduce IPsec support for OVS tunneling and
adds new features to prepare
From: Numan Siddique
Python IDL implementation doesn't have the support to connect to the
cluster dbs. This patch adds this support. We are still missing the
support in python idl class to connect to the cluster master. That
support will be added in an upcoming patch.
This patch is similar to
From: Numan Siddique
The python function ovs.socket_util.check_connection_completion() uses select()
(provided by python) to monitor the socket file descriptor. The select()
returns 1 when the file descriptor becomes ready. For error cases like -
111 (Connection refused) and 113 (No route to
From: Numan Siddique
Python IDL library is lacking the functionality to connect to the
clustered db servers by providing multiple remotes (like -
"tcp:10.0.0.1:6641, tcp:10.0.0.2:6641, tcp:10.0.0.3:6641") in the
connection string.
This patch adds this functionality to the python idl library.
It
This commit fixes 'maybe-uninitialized' warnings for pointers in various
functions in ovn-nbctl when compiling with gcc 6.3.1 and -Werror.
Pointers to structs nbrec_logical_switch, nbrec_logical_switch_port,
nbrec_logical_router and nbrec_logical_router_port are now initialized
to NULL where
On Wed, Jul 4, 2018 at 7:23 AM, Matteo Croce wrote:
> From: Stefano Brivio
>
> Open vSwitch sends to userspace all received packets that have
> no associated flow (thus doing an "upcall"). Then the userspace
> program creates a new flow and determines the actions to apply
> based on its
On Mon, Jul 9, 2018 at 11:36 PM Ben Pfaff wrote:
> On Sun, Jul 08, 2018 at 10:05:41PM +0530, nusid...@redhat.com wrote:
> > From: Numan Siddique
> >
> > Calling ovs.stream.open_block(ovs.stream.open("tcp:127.0.0.1:6641"))
> returns
> > success even if there is no server listening on 6641. To
Commit 9f57c67c379d ("gre: Remove support for sharing GRE protocol hook")
allows only single GRE packet receiver. When upstream kernel's gre module
is loaded, the gre.ko exclusively becomes the only gre packet receiver,
preventing OVS kernel module from registering another gre receiver.
We can
On Tue, Jun 26, 2018 at 04:16:42PM -0700, Darrell Ball wrote:
> Signed-off-by: Darrell Ball
I wonder whether coverage counters would be a better alternative? They
use thread-local data so incrementing them is cheaper than atomic
increments, at least if they are accessed from multiple CPUs.
On Tue, Jul 10, 2018 at 06:09:05PM +0100, Ian Stokes wrote:
> On 7/10/2018 2:33 PM, Loftus, Ciara wrote:
> >>
> >>An entry for experimental vhost zero copy support was removed
> >>incorrectly. Re-add this entry to NEWS.
> >>
> >>Reported-by: Eelco Chaudron
> >>Cc: Ciara Loftus
> >>Fixes:
On Tue, Jul 10, 2018 at 10:14:41AM -0700, Ben Pfaff wrote:
> On Tue, Jul 10, 2018 at 11:10:17AM +0200, Bhargava Shastry wrote:
> > Hi Ben,
> >
> > Thanks a lot for the new patch. A couple of comments:
> >
> > - Could you please rename each of `tests/oss-fuzz/config/*.options` from
> >
From: Bhargava Shastry
Signed-off-by: Ben Pfaff
---
Makefile.am | 1 +
tests/automake.mk | 2 +
tests/oss-fuzz/automake.mk| 27 ++
tests/oss-fuzz/config/flow_extract_target.options | 2 +
Test 1024 PMD - stats reported different stats data during tests because of
the SMC data. This commit fix the test.
Signed-off-by: Yipeng Wang
---
tests/pmd.at | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/tests/pmd.at b/tests/pmd.at
index 60452f5..4cae6c8 100644
---
This patch adds a signature match cache (SMC) after exact match
cache (EMC). The difference between SMC and EMC is SMC only stores
a signature of a flow thus it is much more memory efficient. With
same memory space, EMC can store 8k flows while SMC can store 1M
flows. It is generally beneficial to
This patch set is the V5 implementation to combine the CD and DFC design.
Both patches intend to refactor datapath to avoid costly sequential subtable
search.
We rename the CD/DFC as signature match cache (SMC) to be more clear since the
proposed cache structure stores signatures of flows. SMC
On Tue, Jul 10, 2018 at 11:10:17AM +0200, Bhargava Shastry wrote:
> Hi Ben,
>
> Thanks a lot for the new patch. A couple of comments:
>
> - Could you please rename each of `tests/oss-fuzz/config/*.options` from
> *_fuzzer.options to *_target.options. This is because oss-fuzz infra
> requires
Fixes test 815 "tunnel_push_pop_ipv6 - action".
CC: Aaron Conole
Fixes: 6f231f7c3a9e ("xlate: use const struct in6_addr in linklocal check")
Signed-off-by: Ben Pfaff
---
ofproto/ofproto-dpif-xlate.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
On 7/10/2018 2:33 PM, Loftus, Ciara wrote:
An entry for experimental vhost zero copy support was removed
incorrectly. Re-add this entry to NEWS.
Reported-by: Eelco Chaudron
Cc: Ciara Loftus
Fixes: c3c722d2c7ee ("Documentation: document ovs-dpdk flow offload")
Signed-off-by: Ian Stokes
On Tue, Jul 10, 2018 at 01:13:31PM +0200, Jakub Sitnicki wrote:
> On Mon, 9 Jul 2018 16:34:00 -0700
> Ben Pfaff wrote:
>
> > This will be useful for daemonized ovn-nbctl.
> >
> > Signed-off-by: Ben Pfaff
> > ---
>
> Acked-by: Jakub Sitnicki
Thanks, applied to master.
Hello! My Name is Perry White,P.A to Antonia Ax:son Johnson, she made a
donation to you. Contact Email: antoniajax...@gmail.com for more details.
Thanks.
___
dev mailing list
d...@openvswitch.org
OK.
On Tue, Jul 10, 2018 at 02:58:47PM +0530, Aravind Prasad wrote:
> Hi Ben/All,
>
> If possible, Kindly hold reviewing this patch for now. Expecting an
> approval from my Org. Sorry for the inconvenience caused and thanks for the
> support.
>
> Will get back and intimate for the review as
On GNU systems these macros work with arbitrary pointers, but the relevant
standards only require IN6_IS_ADDR_MC_LINKLOCAL to work with in6_addr (and
don't specify IN6_ARE_ADDR_EQUAL at all). Make the "sparse"
implementations correspondingly pickier so that we catch any introduced
problems more
On Tue, Jul 10, 2018 at 09:34:38AM -0400, Aaron Conole wrote:
> Commit 83c2757bd16e ("xlate: Move tnl_neigh_snoop() to
> terminate_native_tunnel()") introduced a call to
> IN6_IS_ADDR_MC_LINKLOCAL() when checking neighbor discovery.
>
> The call to this assumes that the argument may be a const
Good, thanks.
Acked-by: Darrell Ball
This needs backporting to 2.8.
I re-ran some SA tools and found some other potential use after free bugs.
On Tue, Jul 10, 2018 at 4:05 AM, Ilya Maximets
wrote:
> 'alg_exp_entry' is allocated on stack memory, but could be used via
> 'alg_exp' pointer
Fixes: bd5e81a0e596 ("Userspace Datapath: Add ALG infra and FTP.")
Signed-off-by: Darrell Ball
---
Needs backporting as far back as 2.8.
lib/conntrack.c | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/lib/conntrack.c b/lib/conntrack.c
index e1c1f2e..b818584 100644
Signed-off-by: Darrell Ball
---
Needs backporting as far back as 2.8.
lib/conntrack.c | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/lib/conntrack.c b/lib/conntrack.c
index e1c1f2e..b818584 100644
--- a/lib/conntrack.c
+++ b/lib/conntrack.c
@@ -1159,8 +1159,11 @@
On 10.07.2018 17:48, Vishal Deep Ajmera wrote:
>>
>> This is potentially dangerous from the future modifications and hard to read
>> for
>> reviewer/person who tries to understand how it works.
>>
>> Current implementation will fail if someone will change the logic of
>>
>
> This is potentially dangerous from the future modifications and hard to read
> for
> reviewer/person who tries to understand how it works.
>
> Current implementation will fail if someone will change the logic of
> 'DP_PACKET_BATCH_REFILL_FOR_EACH', for example.
>
> The key point is that
Hi Ben,
Thanks for applying this patch to master. Can you please apply the patch
to the 2.9 branch as well?
Thank you,
Mark
On 06/26/2018 02:45 PM, Mark Michelson wrote:
A note: if approved, this patch will also need to go into version 2.9
On 06/26/2018 02:42 PM, Mark Michelson wrote:
On 10/07/2018 13:53, Eelco Chaudron wrote:
>
>
> On 10 Jul 2018, at 14:48, Lam, Tiago wrote:
>
[snip]
>>
>> A blatant leftover from the rebase. Thanks, Eelco.
>>
>> I'll wait for Ian's reply in patch 01/15 and take this into account
>> for v5.
>>
>> Tiago.
>
> I’m on PTO starting tomorrow
Commit 83c2757bd16e ("xlate: Move tnl_neigh_snoop() to
terminate_native_tunnel()") introduced a call to
IN6_IS_ADDR_MC_LINKLOCAL() when checking neighbor discovery.
The call to this assumes that the argument may be a const uint8_t *.
According to The Open Group Base Specifications Issue 7, 2018:
>
> An entry for experimental vhost zero copy support was removed
> incorrectly. Re-add this entry to NEWS.
>
> Reported-by: Eelco Chaudron
> Cc: Ciara Loftus
> Fixes: c3c722d2c7ee ("Documentation: document ovs-dpdk flow offload")
> Signed-off-by: Ian Stokes
Acked-by: Ciara Loftus
> ---
>
Configuring flow control at ixgbe netdev-init is throwing error in port
start.
For eg: without this fix, user cannot configure flow control on ixgbe dpdk
port as below,
"
ovs-vsctl add-port br0 dpdk0 -- set Interface dpdk0 type=dpdk \
options:dpdk-devargs=:05:00.1
An entry for experimental vhost zero copy support was removed
incorrectly. Re-add this entry to NEWS.
Reported-by: Eelco Chaudron
Cc: Ciara Loftus
Fixes: c3c722d2c7ee ("Documentation: document ovs-dpdk flow offload")
Signed-off-by: Ian Stokes
---
NEWS | 1 +
1 file changed, 1 insertion(+)
On 10 Jul 2018, at 13:06, Tiago Lam wrote:
> When configuring a mempool, in netdev_dpdk_mempool_configure(), the
> result of a call to dpdk_buf_size() is being used as the MTU. This,
> however, is not the MTU but a ROUND_UP aligned number based on the MTU,
> which could lead to the reuse of
On 10 Jul 2018, at 14:48, Lam, Tiago wrote:
On 10/07/2018 13:40, Ilya Maximets wrote:
On 10.07.2018 15:30, Eelco Chaudron wrote:
On 10 Jul 2018, at 13:06, Tiago Lam wrote:
From: Mark Kavanagh
Currently, jumbo frame support for OvS-DPDK is implemented by
increasing the size of mbufs
On 10/07/2018 13:40, Ilya Maximets wrote:
> On 10.07.2018 15:30, Eelco Chaudron wrote:
>>
>>
>> On 10 Jul 2018, at 13:06, Tiago Lam wrote:
>>
>>> From: Mark Kavanagh
>>>
>>> Currently, jumbo frame support for OvS-DPDK is implemented by
>>> increasing the size of mbufs within a mempool, such that
On 7/10/2018 1:40 PM, Ilya Maximets wrote:
On 10.07.2018 15:30, Eelco Chaudron wrote:
On 10 Jul 2018, at 13:06, Tiago Lam wrote:
From: Mark Kavanagh
Currently, jumbo frame support for OvS-DPDK is implemented by
increasing the size of mbufs within a mempool, such that each mbuf
within the
On 10.07.2018 15:30, Eelco Chaudron wrote:
>
>
> On 10 Jul 2018, at 13:06, Tiago Lam wrote:
>
>> From: Mark Kavanagh
>>
>> Currently, jumbo frame support for OvS-DPDK is implemented by
>> increasing the size of mbufs within a mempool, such that each mbuf
>> within the pool is large enough to
On 10 Jul 2018, at 13:06, Tiago Lam wrote:
From: Mark Kavanagh
Currently, jumbo frame support for OvS-DPDK is implemented by
increasing the size of mbufs within a mempool, such that each mbuf
within the pool is large enough to contain an entire jumbo frame of
a user-defined size.
I am trying to implement 8 queues at egress port in the switch and will
queue the packets based on the priority I need the current file name which
is handing the queuing for OVS .
Regards
Rakesh Kumar
___
dev mailing list
d...@openvswitch.org
On 10.07.2018 14:14, Vishal Deep Ajmera wrote:
>>> +
>>> +/* preserve the order of packet for flow batching */
>>> +index_map[packets_->count - 1] = map_cnt;
>>
>> Using the "packets_->count" without accessor + inside a refill loop is highly
>> inaccurate/potentially
> > +
> > +/* preserve the order of packet for flow batching */
> > +index_map[packets_->count - 1] = map_cnt;
>
> Using the "packets_->count" without accessor + inside a refill loop is highly
> inaccurate/potentially dangerous. Isn't it equal to "n_missed"?
>
Hi Ilya,
On Mon, 9 Jul 2018 16:34:00 -0700
Ben Pfaff wrote:
> This will be useful for daemonized ovn-nbctl.
>
> Signed-off-by: Ben Pfaff
> ---
Acked-by: Jakub Sitnicki
___
dev mailing list
d...@openvswitch.org
On 10/07/2018 12:06, Tiago Lam wrote:
> When configuring a mempool, in netdev_dpdk_mempool_configure(), the
> result of a call to dpdk_buf_size() is being used as the MTU. This,
> however, is not the MTU but a ROUND_UP aligned number based on the MTU,
> which could lead to the reuse of mempools
As it stands, OVS_DPDK_START() won't allow other configs to be set
before starting the ovs-vswitchd daemon. This is a problem since some
configs, such as the "dpdk-multi-seg-mbufs=true" for enabling the
multi-segment mbufs, need to be set prior to start OvS.
To support other options,
The following tests are added to the DPDK testsuite to add some
coverage for the multi-segment mbufs:
- Check that multi-segment mbufs are disabled by default;
- Check that providing `other_config:dpdk-multi-seg-mbufs=true` indeed
enables mbufs;
- Using a DPDK port, send a random packet out and
In order to create a minimal environment that allows the tests to get
mbufs from an existing mempool, the following approach is taken:
- EAL is initialised (by using the main dpdk_init()) and a (very) small
mempool is instantiated (mimicking the logic in dpdk_mp_create()).
This mempool
From: Mark Kavanagh
Currently, jumbo frame support for OvS-DPDK is implemented by
increasing the size of mbufs within a mempool, such that each mbuf
within the pool is large enough to contain an entire jumbo frame of
a user-defined size. Typically, for each user-defined MTU,
'requested_mtu', a
From: Mark Kavanagh
Currently, packets are only copied to a single segment in the function
dpdk_do_tx_copy(). This could be an issue in the case of jumbo frames,
particularly when multi-segment mbufs are involved.
This patch calculates the number of segments needed by a packet and
copies the
When enabled with DPDK OvS relies on mbufs allocated by mempools to
receive and output data on DPDK ports. Until now, each OvS dp_packet has
had only one mbuf associated, which is allocated with the maximum
possible size, taking the MTU into account. This approach, however,
doesn't allow us to
From: Michael Qiu
When doing packet clone, if packet source is from DPDK driver,
multi-segment must be considered, and copy the segment's data one by
one.
Also, lots of DPDK mbuf's info is missed during a copy, like packet
type, ol_flags, etc. That information is very important for DPDK to do
When configuring a mempool, in netdev_dpdk_mempool_configure(), the
result of a call to dpdk_buf_size() is being used as the MTU. This,
however, is not the MTU but a ROUND_UP aligned number based on the MTU,
which could lead to the reuse of mempools even when the real MTUs
between different ports
In its current implementation dp_packet_shift() is also unaware of
multi-seg mbufs (that holds data in memory non-contiguously) and assumes
that data exists contiguously in memory, memmove'ing data to perform the
shift.
To add support for multi-seg mbuds a new set of functions was
introduced,
From: Mark Kavanagh
dp_packets are created using xmalloc(); in the case of OvS-DPDK, it's
possible the the resultant mbuf portion of the dp_packet contains
random data. For some mbuf fields, specifically those related to
multi-segment mbufs and/or offload features, random values may cause
Most helper functions in dp-packet assume that the data held by a
dp_packet is contiguous, and perform operations such as pointer
arithmetic under that assumption. However, with the introduction of
multi-segment mbufs, where data is non-contiguous, such assumptions are
no longer possible. Some
When enabled with DPDK OvS deals with two types of packets, the ones
coming from the mempool and the ones locally created by OvS - which are
copied to mempool mbufs before output. In the latter, the space is
allocated from the system, while in the former the mbufs are allocated
from a mempool,
1 - 100 of 119 matches
Mail list logo
