Hi Martin,
Somehow I didn’t receive this patch email via the ovs-dev mailing list, perhaps
one of the many spam filters on the way interfered. Don't know if this response
email will be recognized by ovs patchwork.
The nsh.at lines are broken wrongly in this email, but they look OK in
patchwork
On Thu, May 6, 2021 at 7:21 PM Numan Siddique wrote:
>
>
> On Thu, May 6, 2021, 9:17 PM Han Zhou wrote:
>
>> On Thu, May 6, 2021 at 6:13 PM Numan Siddique wrote:
>>
>> >
>> >
>> > On Thu, May 6, 2021, 7:00 PM Han Zhou wrote:
>> >
>> >> On Thu, May 6, 2021 at 11:41 AM Numan Siddique wrote:
>>
On Thu, May 6, 2021 at 2:17 PM Ben Pfaff wrote:
>
> Without this patch, configure reports the OVS version like this:
> OVS version is $OVSVERSION
>
> This doesn't match the usual style for messages from configure, so this
> patch changes it so that it does, like this:
> Checking OVS versio
Hi Ihar,
Thanks for the patch, and it's great to see the significant performance
improvement!
Please see my comments below.
On Thu, May 6, 2021 at 7:35 AM Ihar Hrachyshka wrote:
>
> For allow-stateless ACLs, bypass connection tracking by avoiding
> setting ct hints for matching traffic. Avoid se
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:
WARNING: Line has trailing whitespace
#48 FILE: tutorial/northd_ddlog_test.sh:8:
if ddlog_running;
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Unexpected sign-offs from developers who are not authors or co-authors
or committers: Ben
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:
WARNING: Line is 80 characters long (recommended limit is 79)
#87 FILE: utilities/ovn-nbctl.c:823:
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:
WARNING: Line is 93 characters long (recommended limit is 79)
#677 FILE: utilities/ovn-sbctl.8.xml:
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:
WARNING: Line is 80 characters long (recommended limit is 79)
#254 FILE: utilities/ovn-dbctl.c:214:
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:
WARNING: Line is 82 characters long (recommended limit is 79)
#318 FILE: utilities/ovn-nbctl.8.xml:
From: Leonid Ryzhyk
DDlog 0.38.0 adds the `--intern-table` CLI flag to the `ovsdb2ddlog`
compiler to declare input tables coming from OVSDB as `Intern<...>`.
This is useful for tables whose records are copied around as a whole and
can therefore benefit from interning performance- and memory-wise.
This is originally from Numan Siddique. I have adapted it a bit
to run faster by using the ovn-nbctl and ovn-sbctl daemons and
combining multiple calls into just one.
I'm uncertain whether to actually commit this; I need a sign-off
from Numan to do so.
Signed-off-by: Ben Pfaff
CC: Numan Siddiqu
From: Leonid Ryzhyk
Use the `--intern-table` switch to intern `Logical_Switch_Port` records,
so that they can be copied and compared efficiently by pointer.
Signed-off-by: Leonid Ryzhyk
Signed-off-by: Ben Pfaff
---
northd/helpers.dl| 2 +-
northd/lrouter.dl| 4 ++--
northd/lswitch.d
From: Leonid Ryzhyk
Change all remaining occurrences of `Ref` to `Intern` throughout
the DDlog code base.
Signed-off-by: Leonid Ryzhyk
Signed-off-by: Ben Pfaff
---
northd/helpers.dl| 36 +-
northd/lrouter.dl| 22
northd/lswitch.dl| 61 +
From: Leonid Ryzhyk
Use the `--intern-table` switch to intern `Logical_Router_Port` records,
so that they can be copied and compared efficiently by pointer.
Signed-off-by: Leonid Ryzhyk
Signed-off-by: Ben Pfaff
---
northd/helpers.dl| 2 +-
northd/lrouter.dl| 50 ++
From: Leonid Ryzhyk
This is another instance of a performance bug when a change to a
router object forces a cascade of changes to relations that reference
the object. This time around the problem was caused by the
`Router.static_routes` field, which is copied from
`nb::Logical_Router`. Luckily,
From: Leonid Ryzhyk
This commit is analogous to 076749c99, but switches instead of routers.
`relation Switch` stores the internal representation of a logical
switch, consisting of values from the `nb::Logical_Switch` table
augmented with some additional fields. We used to do this by
copying the
From: Leonid Ryzhyk
Change the type of record in the `RouterPort` table from
`Ref` to `Intern`.
Signed-off-by: Leonid Ryzhyk
Signed-off-by: Ben Pfaff
---
northd/lrouter.dl| 37 +
northd/lswitch.dl| 4 ++--
northd/ovn_northd.dl | 14 +++---
From: Leonid Ryzhyk
We eliminate an anti-pattern in the use of smart pointers that occurred
throughout the DDlog code.
Consider relation `A` that contains field `x` wrapped in a DDlog smart
pointer (`Intern<>` or `Ref<>`):
```
relation A(x: Intern, ...)
```
Here `T` might be a complex type wit
From: Leonid Ryzhyk
Signed-off-by: Leonid Ryzhyk
Signed-off-by: Ben Pfaff
---
northd/lswitch.dl | 1 -
1 file changed, 1 deletion(-)
diff --git a/northd/lswitch.dl b/northd/lswitch.dl
index 6b2ae37a1608..eeeb968d5d97 100644
--- a/northd/lswitch.dl
+++ b/northd/lswitch.dl
@@ -23,7 +23,6 @@ imp
From: Leonid Ryzhyk
This patch is a workaround for a performance issue in the DDlog
compiler. The issue will hopefully be resolved in a future version of
DDlog, but for now we need this and possibly a few other similar fixes.
Here is one affected rule:
```
sb::Out_Port_Group(._uuid = hash128(s
From: Leonid Ryzhyk
This is the first in a series of commits that will replace the use of
the DDlog's `Ref<>` type with `Intern<>` throughout the OVN code base.
`Ref` and `Intern` are the two forms of smart pointers supported by
DDlog at the moment. `Ref` is a reference counted pointer. Copying
From: Leonid Ryzhyk
Change the type of record in the `SwitchPort` table from
`Ref` to `Intern`.
Signed-off-by: Leonid Ryzhyk
Signed-off-by: Ben Pfaff
---
northd/ipam.dl | 22 +++
northd/lswitch.dl| 64 +++-
northd/ovn_northd.dl |
From: Leonid Ryzhyk
`relation Router` stores the internal representation of a logical
router, consisting of values from the `nb::Logical_Router` table
augmented with some additional fields. We used to do this by
copying the entire `Logical_Router` record inside `Router`. This
proved highly inef
From: Leonid Ryzhyk
Change the type of record in the `Switch` table from `Ref` to
`Intern`.
Signed-off-by: Leonid Ryzhyk
Signed-off-by: Ben Pfaff
---
northd/lswitch.dl | 37 +
northd/multicast.dl | 10 +-
2 files changed, 26 insertions(+), 21 dele
In an upcoming commit, this will allow adding daemon mode to ovn-sbctl
without having a lot of duplicated code.
Signed-off-by: Ben Pfaff
---
utilities/automake.mk |5 +-
utilities/ovn-dbctl.c | 1214
utilities/ovn-dbctl.h | 60 ++
utilities/ovn-nbctl.c
Dumitru reported in #openvswitch that ovn-northd-ddlog can discard the
setting of NB_Global.options:use_logical_dp_groups at startup. I think
that this must be because it seems possible that at startup some of the
relations in the Out_NB_Global rule aren't populated yet, and yet
there is still a r
From: Leonid Ryzhyk
Upcoming commits will use a new --intern-table option of ovsdb2ddlog,
so we need to upgrade to the version of ddlog that has that feature.
To do so, we need to adapt the code to language changes in ddlog. This
commit does that for a change in 0.37 in which, when iterating ov
They do exactly the same thing but ovn-appctl has more ovn in its name.
Signed-off-by: Ben Pfaff
Suggested-by: Mark Michelson
---
utilities/ovn-nbctl.8.xml | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/utilities/ovn-nbctl.8.xml b/utilities/ovn-nbctl.8.xml
index 39f938
Also rewrite the manpage and convert it to XML for consistency with
ovn-nbctl, and add tests.
Signed-off-by: Ben Pfaff
---
NEWS | 4 +-
manpages.mk | 17 -
tests/ovn-sbctl.at| 76 +++--
utilities/automake.mk | 7 +-
utilities/ovn-dbctl.c |
Signed-off-by: Ben Pfaff
---
tests/ovn.at | 51 +++
1 file changed, 35 insertions(+), 16 deletions(-)
diff --git a/tests/ovn.at b/tests/ovn.at
index 175cce0049cd..8adaaa9fde64 100644
--- a/tests/ovn.at
+++ b/tests/ovn.at
@@ -9683,9 +9683,10 @@ wait
ovn-nbctl was retrieving a replica of the entire northbound database
regardless of what it was asked to do. This was unnecessary, but it
was not written to be smart enough to just retrieve what it needed.
This commit fixes the problem. It makes "ovn-nbctl init", for example,
run instantly on a bi
This isn't notable, since this commit frees it just before exiting, but
it cleans up the Address Sanitizer report.
Signed-off-by: Ben Pfaff
---
utilities/ovn-dbctl.c | 4
1 file changed, 4 insertions(+)
diff --git a/utilities/ovn-dbctl.c b/utilities/ovn-dbctl.c
index 28ebc6267066..d815dc5c
This rearranges the manpage into a more logical order, documents some
options that weren't documented, adds some sections such as
Environment and Exit Status that a manpage should have, puts the
headings at reasonable levels instead of all at the top level, and adds
a little more explanatory text i
Mark says that it was heavily used by ovn-kubernetes for years without
any issues.
Signed-off-by: Ben Pfaff
Suggested-by: Mark Michelson
---
NEWS | 1 +
utilities/ovn-nbctl.8.xml | 4
2 files changed, 1 insertion(+), 4 deletions(-)
diff --git a/NEWS b/NEWS
index c91bc
This isn't notable, since this commit frees it just before exiting, but
it cleans up the Address Sanitizer report.
Signed-off-by: Ben Pfaff
---
utilities/ovn-nbctl.c | 4
1 file changed, 4 insertions(+)
diff --git a/utilities/ovn-nbctl.c b/utilities/ovn-nbctl.c
index 042c210020b8..e3af0be9
I get a clean report from Address Sanitizer now.
Signed-off-by: Ben Pfaff
---
northd/ovn-northd-ddlog.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/northd/ovn-northd-ddlog.c b/northd/ovn-northd-ddlog.c
index b7d2c8a5ef8d..ca1ab325448c 100644
--- a/northd/ovn-northd-ddlog.c
+++ b/northd
This series of patches greatly improves the performance of
ovn-northd-ddlog with the benchmark added in the final patch. The
first three patches improve both the benchmark for both versions of
ovn-northd.
Here are the timings that I measure in each case. All of them
include the benefit of the fi
On Thu, May 6, 2021, 9:17 PM Han Zhou wrote:
> On Thu, May 6, 2021 at 6:13 PM Numan Siddique wrote:
>
> >
> >
> > On Thu, May 6, 2021, 7:00 PM Han Zhou wrote:
> >
> >> On Thu, May 6, 2021 at 11:41 AM Numan Siddique wrote:
> >> >
> >> > On Thu, May 6, 2021 at 2:35 AM Han Zhou wrote:
> >> > >
>
On Fri, Apr 16, 2021 at 12:05:55PM -0400, Mark Michelson wrote:
> On 4/1/21 7:20 PM, Ben Pfaff wrote:
> > Signed-off-by: Ben Pfaff
> > ---
> > tests/ovn.at | 67 ++--
> > 1 file changed, 49 insertions(+), 18 deletions(-)
> >
> > diff --git a/test
By default, OVS bridges use standalone fail-mode, which means it is
configured with a single row with the NORMAL action as its OpenFlow table.
Upon system reboot, an integration bridge with many ports and such a table
could create broadcast storms and duplicate packets. That is why
ovn-controller c
On Fri, Apr 16, 2021 at 12:05:12PM -0400, Mark Michelson wrote:
> On 4/1/21 7:20 PM, Ben Pfaff wrote:
> > +static bool
> > +check_conflicts(struct ctl_context *ctx, const char *name, char *msg)
> > +{
> > +struct sbctl_context *sbctl_ctx = sbctl_context_get(ctx);
> > if (shash_find(&sbctl
On Thu, May 6, 2021 at 6:13 PM Numan Siddique wrote:
>
>
> On Thu, May 6, 2021, 7:00 PM Han Zhou wrote:
>
>> On Thu, May 6, 2021 at 11:41 AM Numan Siddique wrote:
>> >
>> > On Thu, May 6, 2021 at 2:35 AM Han Zhou wrote:
>> > >
>> > > On Thu, Apr 22, 2021 at 6:25 AM wrote:
>> > > >
>> > > > Fr
On Thu, May 6, 2021, 7:00 PM Han Zhou wrote:
> On Thu, May 6, 2021 at 11:41 AM Numan Siddique wrote:
> >
> > On Thu, May 6, 2021 at 2:35 AM Han Zhou wrote:
> > >
> > > On Thu, Apr 22, 2021 at 6:25 AM wrote:
> > > >
> > > > From: Numan Siddique
> > > >
> > > > Presently we add 65535 priority l
On Fri, Apr 16, 2021 at 12:04:57PM -0400, Mark Michelson wrote:
> 1) The subject line here didn't include "ovn" so it didn't end up in OVN's
> patchwork. I had to find it in the OVS patchwork, where it had been marked
> as "Not Applicable".
Sorry about that.
> 2) There is some overlap in this pat
On Thu, May 6, 2021 at 11:41 AM Numan Siddique wrote:
>
> On Thu, May 6, 2021 at 2:35 AM Han Zhou wrote:
> >
> > On Thu, Apr 22, 2021 at 6:25 AM wrote:
> > >
> > > From: Numan Siddique
> > >
> > > Presently we add 65535 priority lflows in the stages -
> > > 'ls_in_acl' and 'ls_out_acl' to drop
On Thu, May 6, 2021 at 1:18 PM Mark Michelson wrote:
>
> Thanks, Han!
>
> Acked-by: Mark Michelson
Thanks Mark. I applied it to master branch.
I'd also like to backport it to branch-21.03. Do you have any concerns
before I do that? I know it is not a bug fix, but it is kind of a blocker
for larg
Without this patch, configure reports the OVS version like this:
OVS version is $OVSVERSION
This doesn't match the usual style for messages from configure, so this
patch changes it so that it does, like this:
Checking OVS version... $OVSVERSION
Signed-off-by: Ben Pfaff
---
acinclude.m4
Thank you, that is a good idea. Here is a patch for it. I will add
this to my series as well.
-8<--cut here-->8--
From: Ben Pfaff
Date: Thu, 6 May 2021 14:10:51 -0700
Subject: [PATCH ovn] configure: Check for correct ddlog version.
Signed-off-by
ACLs and logical router policies leave the user open to crafting
creative matches, such as this one:
(ip4.dst == 172.27.0.65/32) && ip4.src == $as && ip4.dst != 10.128.0.0/11
Ideally, that final part of the match would just be omitted, but in some
CMSes this kind of thing gets generated. If we ze
Thanks, Han!
Acked-by: Mark Michelson
On 5/3/21 9:33 PM, Han Zhou wrote:
For an ACL with match: outport == @PG && ip4.src == $PG_AS, given below
scale:
P: PG size
LP: number of local lports
D: number of all datapaths (logical switches)
LD: number of datapaths that contain local lports
With c
ACLs and logical router policies leave the user open to crafting
creative matches, such as this one:
(ip4.dst == 172.27.0.65/32) && ip4.src == $as && ip4.dst != 10.128.0.0/11
Ideally, that final part of the match would just be omitted, but in some
CMSes this kind of thing gets generated. If we ze
On Thu, May 6, 2021 at 2:35 AM Han Zhou wrote:
>
> On Thu, Apr 22, 2021 at 6:25 AM wrote:
> >
> > From: Numan Siddique
> >
> > Presently we add 65535 priority lflows in the stages -
> > 'ls_in_acl' and 'ls_out_acl' to drop packets which
> > match on 'ct.inv'.
> >
> > This patch provides an optio
GCC 11 warned that the prototypes and the definitions of these
functions were different: the prototypes used [], the definitions used
[IFNAMSIZ]. This is perfectly valid from a C standards perspective, but
it's confusing and providing sizes gives the compiler and the developer
a useful hint as to
The prototype said *, the definition said [CLS_MAX_TRIES]. GCC 11
complains about this (though it is perfectly valid from a C standards
perspective). It would probably be better to make them both use
[CLS_MAX_TRIES] but that's only allowed if the struct's definition is
visible at the point of the
GCC 11 pointed out that ofputil_group_to_string()'s prototype asks for
a buffer with one byte more than supplied. This fixes the problem.
This wasn't a buffer overflow because ofputil_group_to_string() honors
the buffer size passed in, which was correct. The worst that could
happen was truncatin
Patch 1 fixes a bug. Patches 2 and 3 fix warnings that were not bugs.
Ben Pfaff (3):
ofp-group: Use big-enough buffer in ofputil_format_group().
tnl-neigh-cache: Include expected array sizes in prototypes.
classifier: Make find_match_wc() prototype and definition match.
lib/classifier.c
Hello,
Could I please get a review on this patch
https://patchwork.ozlabs.org/project/ovn/patch/20210402210858.56673-1-svc.eng.git-m...@nutanix.com/
?
In the above patch I have addressed the comments from the previous two versions
https://patchwork.ozlabs.org/project/ovn/patch/20210325234059.12
On Wed, May 5, 2021 at 2:25 AM Han Zhou wrote:
>
> On Thu, Apr 29, 2021 at 9:25 AM Numan Siddique wrote:
> >
> > On Wed, Apr 28, 2021 at 3:10 PM Han Zhou wrote:
> > >
> > > On Wed, Apr 28, 2021 at 3:34 AM Numan Siddique wrote:
> > > >
> > > > On Wed, Apr 28, 2021 at 3:31 AM Han Zhou wrote:
> >
This comment-only patch could use a review.
On Wed, Apr 14, 2021 at 07:45:04PM -0700, Ben Pfaff wrote:
> Commit a14502a7c529 ("classifier: Retire partitions.") in 2015 removed
> OVS support for tags for the second time (!), so these comments about
> them should be removed too.
>
> Signed-off-by:
This documentation-only patch could use a review.
On Wed, Apr 14, 2021 at 08:34:46PM -0700, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/ovs-actions.xml | 288 +++-
> 1 file changed, 286 insertions(+), 2 deletions(-)
>
> diff --git a/lib/ovs-ac
On Fri, Apr 16, 2021 at 12:41:13PM +0200, Timothy Redaelli wrote:
> On Wed, 14 Apr 2021 13:40:20 -0700
> Ben Pfaff wrote:
>
> > The action headings were coming out all smashed together, like
> > "Theoutputaction". This fixes them so that they appear correctly, like
> > "The output action".
> >
The initial implementation of the notification mechanism that indicates
if OVS flows corresponding to a logical switch port have been installed
is not resilient enough. It didn't cover the case when transactions to
the local OVS DB or to the Southbound fail.
In order to deal with that, factor out
This new option [1] has been added to remove checks [2][3] that only
a DPDK developer cares about.
1: https://git.dpdk.org/dpdk/commit/?id=bc4617433845
2: https://git.dpdk.org/dpdk/commit/?id=d317f7ebd484
3: https://git.dpdk.org/dpdk/commit/?id=720dfda4551e
Signed-off-by: David Marchand
---
.ci
PCI_ANY_ID has been prefixed with RTE_ [1].
1: https://git.dpdk.org/dpdk/commit/?id=4d509afa7bb2
Signed-off-by: David Marchand
---
lib/netdev-dpdk.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index 9d8096668e..e67d27d4e3 100644
pyelftools is now required to build DPDK [1].
1: https://git.dpdk.org/dpdk/commit/?id=f0f93a7adfee
Signed-off-by: David Marchand
---
.ci/linux-prepare.sh | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/.ci/linux-prepare.sh b/.ci/linux-prepare.sh
index c55125cf78..67c8ecf499
A recent change [1] in DPDK bumped requirements on meson to 0.49.2.
1: https://git.dpdk.org/dpdk/commit/?id=8c1053083678
Signed-off-by: David Marchand
---
.ci/linux-prepare.sh | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/.ci/linux-prepare.sh b/.ci/linux-prepare.sh
index 6
Add Travis jobs to check compilation with DPDK experimental API enabled.
This will help us catch issues for the day we need one of them.
Note: this should not be merged to master, intended for dpdk-latest
branch only.
Signed-off-by: David Marchand
Signed-off-by: Ian Stokes
---
.ci/linux-build.
Make this branch point to current main master branch so that we can
track API breakage.
Note: this should not be merged to master, intended for dpdk-latest
branch only.
Signed-off-by: David Marchand
Acked-by: Ilya Maximets
Signed-off-by: Ian Stokes
The default branch name in DPDK is changed f
Hello Ian,
DPDK v21.05-rc2 has been tagged last night, so I spent some time today
rebasing the dpdk-latest OVS branch against OVS master branch (commit
a019868a6268 ("ofproto/ofproto-dpif-sflow: Check sflow ..."), then
fixing build against DPDK.
Patch 1: I squashed together the two patches about
> -Original Message-
> From: dev On Behalf Of Gaetan Rivet
> Sent: Sunday 25 April 2021 12:55
> To: ovs-dev@openvswitch.org
> Subject: [ovs-dev] [PATCH v3 00/28] dpif-netdev: Parallel offload processing
>
> This patch series aims to improve the performance of the management of
> hw-off
For allow-stateless ACLs, bypass connection tracking by avoiding
setting ct hints for matching traffic. Avoid sending all traffic to ct
when a stateful ACL is present.
===
Reusing an existing 'allow' verb for stateless matching would have its
drawbacks, specifically, when it comes to backwards in
From: Numan Siddique
Some smart NICs can't offload datapath flows matching on conntrack
fields. If a deployment desires to make use of such smart NICs
then it cannot configure ACLs on the logical switches. If suppose
a logical switch S1 has no ACLs configured and a logical switch S2
has ACLs co
With a big database writing snapshot could take a lot of time, for
example, on one of the systems compaction of 300MB database takes
about 10 seconds to complete. For the clustered database, 40% of this
time takes conversion of the database to the file transaction json
format, the rest of time is
FreeBSD sometimes changes the base version of python3 that is
used for packages. This affects package names. For example,
currently CI is broken, because there is no more py37- versions
of sphinx and openssl available, only py38- ones:
pkg: No packages available to install matching 'py37-opens
The devlink interface was introduced [0] in the Linux 4.6 time
frame and has since gained traction among multiple hardware
vendors.
The devlink-port [1] and devlink-info[2] interfaces are
particularly useful for managing NICs connected to multiple
distinct CPUs such as SmartNICs.
In such a topolo
90 matches
Mail list logo
