On 7/24/24 17:28, Mike Pattrick wrote:
> On Mon, Jul 22, 2024 at 11:21 AM Ilya Maximets wrote:
>>
>> Pre-commit must not abort the transaction, otherwise the upper layers
>> may crash accessing it. E.g. ovsdb_trigger_try() checks the state of
>> the transactio
On 8/8/24 20:00, Ilya Maximets wrote:
> MLD flows are added to pipelines unconditionally in order to avoid
> sending such traffic through conntrack. The problem is that these
> matches turn into matches on ip6.dst that end up as exact matches in
> datapath flows. This means a separ
ticast for some reason, so
it didn't cause issues for IPv4 traffic. But let's fix it as well.
Tests were using incorrect multicast addresses, fixed now.
Fixes: 677a3ba4d66b ("ovn: Add MLD support.")
Signed-off-by: Ilya Maximets
---
lib/logical-fields.c | 7
On 8/8/24 13:53, Ales Musil wrote:
>
>
> On Thu, Aug 8, 2024 at 1:26 PM Ilya Maximets <mailto:i.maxim...@ovn.org>> wrote:
>
> On 8/8/24 13:18, Ales Musil wrote:
> >
> >
> > On Thu, Aug 8, 2024 at 12:55 PM Ilya Maximets <mailto:i
On 8/8/24 13:18, Ales Musil wrote:
>
>
> On Thu, Aug 8, 2024 at 12:55 PM Ilya Maximets <mailto:i.maxim...@ovn.org>> wrote:
>
> On 8/8/24 12:38, Ales Musil wrote:
> > The test was checking if there was any recompute after all operations.
> > T
On 8/7/24 17:48, Mike Pattrick wrote:
> Currently it is possible to call netdev-dpdk/get-mempool-info before a
> mempool as been created. This can happen because a device is added to
> the netdev_shash before a mempool is allocated for it, which results in
> a segmentation fault.
>
> Now we check
t
> +dp_packet_get_total_length(struct dp_packet *p)
> +{
> +struct rte_ether_hdr *eth_hdr;
> +struct rte_vlan_hdr *vlan_hdr;
> +struct rte_ipv4_hdr *ipv4_hdr;
> +struct rte_ipv6_hdr *ipv6_hdr;
And we should not use DPDK-specific structures and functions is there are
gen
hile not
actually testing anything useful.
There could be a few ways to avoid that:
1. Collect recompute stats per failed node, so we can explicitly check
for a specific reason. May still be a little flaky though.
2. Add at least one positive check that ensures that log we're looking
for
i, Michael. Looks like some ports ended up not cleaned up after previous
tests. And that breaks tests on new patches.
Could you, please, take a look?
It seems the last test before failures started was killed with SIGTERM:
https://mail.openvswitch.org/pipermail/ovs-build/2024-August/040693.html
Thanks.
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
prefer a different date we can
> reconsider this.
FWIW, I'll likely be off for two weeks around or starting that date.
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
icitly instead.
Best regards, Ilya Maximets.
> Now, hlen is uniformly
> modified to packet->l4_ofs, which includes the total length of both
> Layer 2 and Layer 3 headers, thereby correctly removing the tunnel
> header.
>
> Signed-off-by: Sunyang Wu
> ---
> lib/n
On 8/6/24 18:55, Mike Pattrick wrote:
> On Tue, Aug 6, 2024 at 9:29 AM Ilya Maximets wrote:
>>
>> On 7/24/24 20:07, Mike Pattrick wrote:
>>> Currently it is possible to call netdev-dpdk/get-mempool-info before a
>>> mempool as been created. This can happen bec
On 7/24/24 20:07, Mike Pattrick wrote:
> Currently it is possible to call netdev-dpdk/get-mempool-info before a
> mempool as been created. This can happen because a device is added to
> the netdev_shash before a mempool is allocated for it, which results in
> a segmentation fault.
>
> Now we check
On 7/19/24 20:31, Mike Pattrick wrote:
> Add a new test for mirroring packets over a tunnel, which isn't covered
> by other existing tests.
Thanks, Mike. I think it's a good to have a test covering this functionality.
See some comments below.
Best regards, Ilya Maximets.
&g
izing this! I'll revert back to 8 bits (255
> values) for the Sample_Collector.ID in v7. I'll post a new version
> later today.
The math here makes me think that something is terribly wrong with the
design of this feature. Can't put a finger on what it is though yet.
B
ndeed is not zero. Though it is
concerning
that we're receiving more reports about hardware switches adding non-zero
padding
to packets.
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
f not, I can consider applying these patches myself, but I'll be fully
relying on your testing, which may be OK, since you're the main and possibly
the only users of the Windows datapath. But it's still not the best option
since we don't have any kind of CI, except for chec
_CHECK([ovs-vsctl set Interface int-br options:ol_tcp_csum=true])
+
+dnl Break the IP checksum with --bad-csum and replace the correct TCP
+dnl checksum '0xb106' with incorrect '0xabcd'.
+bad_packet=$(ovs-ofctl compose-packet --bare --bad-csum 'TCP_PKT' '' \
+| sed 's/b106/abcd/' )
+echo bad_packet: ${bad_packet}
+AT_CHECK([ovs-appctl netdev-dummy/receive int-br "${bad_packet}${padding}"])
+OVS_WAIT_UNTIL([test $(ovs-pcap p0.pcap | grep -c "${encap}${packet}$") -eq 3])
+
+OVS_VSWITCHD_STOP
+AT_CLEANUP
---
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
;
> +"from %s to %s (now carrying %"PRIu64"kB and "
> +"%"PRIu64"kB load, respectively)",
> + bond->name, delta, hash - bond->hash,
> +from->name, to->name,
> +
On 6/6/24 18:02, Ihar Hrachyshka wrote:
> On Thu, Jun 6, 2024 at 11:12 AM Ihar Hrachyshka <mailto:ihrac...@redhat.com>> wrote:
>
> On Wed, Jun 5, 2024 at 11:17 AM Brian Haley <mailto:haleyb@gmail.com>> wrote:
>
> Hi Ilya,
>
>
On 8/1/24 15:09, Ilya Maximets wrote:
> On 8/1/24 14:26, Dumitru Ceara wrote:
>> On 8/1/24 13:58, Ilya Maximets wrote:
>>> On 7/31/24 19:38, Dumitru Ceara wrote:
>>>> On 7/31/24 18:39, Ilya Maximets wrote:
>>>>> On 7/31/24 18:17, Ilya Maximets wrote:
On 8/1/24 14:26, Dumitru Ceara wrote:
> On 8/1/24 13:58, Ilya Maximets wrote:
>> On 7/31/24 19:38, Dumitru Ceara wrote:
>>> On 7/31/24 18:39, Ilya Maximets wrote:
>>>> On 7/31/24 18:17, Ilya Maximets wrote:
>>>>> On 7/31/24 11:05, Du
On 8/1/24 13:58, Ilya Maximets wrote:
> On 7/31/24 19:38, Dumitru Ceara wrote:
>> On 7/31/24 18:39, Ilya Maximets wrote:
>>> On 7/31/24 18:17, Ilya Maximets wrote:
>>>> On 7/31/24 11:05, Dumitru Ceara wrote:
>>>>> From: Adrian Moreno
>>>>&
On 7/31/24 19:38, Dumitru Ceara wrote:
> On 7/31/24 18:39, Ilya Maximets wrote:
>> On 7/31/24 18:17, Ilya Maximets wrote:
>>> On 7/31/24 11:05, Dumitru Ceara wrote:
>>>> From: Adrian Moreno
>>>>
>>>> Introduce a new table called Sample wh
On 7/31/24 18:17, Ilya Maximets wrote:
> On 7/31/24 11:05, Dumitru Ceara wrote:
>> From: Adrian Moreno
>>
>> Introduce a new table called Sample where per-flow IPFIX configuration
>> can be specified.
>> Also, reference rows from such table from the ACL table to e
reno
> Co-authored-by: Ales Musil
> Signed-off-by: Ales Musil
> Co-authored-by: Dumitru Ceara
> Signed-off-by: Dumitru Ceara
> ---
> V4:
> - added explicit sampling stages
> - reduced set_id max supported value
I don't get that. Does it end up in the observation domain/point
somehow? Or in conntrack mark/label? Sounds strange. If it's only
in logical flow and OpenFlow actions, then it shouldn't matter what
the ID is. Or am I missing something?
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
On 7/31/24 17:05, Dumitru Ceara wrote:
> On 7/31/24 16:40, Ilya Maximets wrote:
>> On 7/31/24 16:17, Dumitru Ceara wrote:
>>> On 7/31/24 16:07, Ilya Maximets wrote:
>>>> On 7/31/24 16:04, Ilya Maximets wrote:
>>>>> On 7/31/24 11:05, Dumitru Ceara wrot
On 7/31/24 16:17, Dumitru Ceara wrote:
> On 7/31/24 16:07, Ilya Maximets wrote:
>> On 7/31/24 16:04, Ilya Maximets wrote:
>>> On 7/31/24 11:05, Dumitru Ceara wrote:
>>>> This will represent a unified place to store IPFIX observation domain ID
>>>>
On 7/31/24 16:04, Ilya Maximets wrote:
> On 7/31/24 11:05, Dumitru Ceara wrote:
>> This will represent a unified place to store IPFIX observation domain ID
>> configurations for sampling applications (currently only drop sampling
>> is supported as application but following
On 7/31/24 11:05, Dumitru Ceara wrote:
> This will represent a unified place to store IPFIX observation domain ID
> configurations for sampling applications (currently only drop sampling
> is supported as application but following commits will add more).
>
> Acked-by: Mark Michelson
> Signed-off-
On 7/31/24 15:25, Ales Musil wrote:
>
>
> On Wed, Jul 31, 2024 at 3:03 PM Ilya Maximets <mailto:i.maxim...@ovn.org>> wrote:
>
> On 7/31/24 11:05, Dumitru Ceara wrote:
> > From: Ales Musil mailto:amu...@redhat.com>>
> >
> > Curren
at b/tests/system-ovn.at
> index 52b2b84d7b..a9cb54b452 100644
> --- a/tests/system-ovn.at
> +++ b/tests/system-ovn.at
> @@ -7724,7 +7724,7 @@ NS_CHECK_EXEC([sw0-p3], [ping -q -c 10 -i 0.3 -w 15
> 10.0.0.2 | FORMAT_PING], \
> AT_CHECK([ovs-appctl dpc
On 7/31/24 14:40, Dumitru Ceara wrote:
> On 7/31/24 14:31, Ilya Maximets wrote:
>> On 7/31/24 11:05, Dumitru Ceara wrote:
>>> From: Adrian Moreno
>>>
>>> Introduce a new table called Sample where per-flow IPFIX configuration
>>> can be specified.
>
ected that they're both enabled together.
Not a full review, but this is a user-visible behavior, so it should be
documented. In the description of both features. Note in a commit message
is not enough.
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
On 7/31/24 14:14, Dumitru Ceara wrote:
> On 7/31/24 12:45, Ilya Maximets wrote:
>> On 7/31/24 12:38, Ilya Maximets wrote:
>>> On 7/31/24 11:05, Dumitru Ceara wrote:
>>>> From: Ales Musil
>>>>
>>>> Add detection for sample action that allow
On 7/31/24 12:38, Ilya Maximets wrote:
> On 7/31/24 11:05, Dumitru Ceara wrote:
>> From: Ales Musil
>>
>> Add detection for sample action that allows to configure
>> obs_domain_id and obs_point_id via registers. This feature
>> is available only from OvS versi
OVS_CT_TUPLE_FLUSH_SUPPORT = (1 << OVS_CT_TUPLE_FLUSH_BIT),
> OVS_DP_HASH_L4_SYM_SUPPORT = (1 << OVS_DP_HASH_L4_SYM_BIT),
> OVS_DP_GROUP_SUPPORT = (1 << OVS_DP_GROUP_SUPPORT_BIT),
> +OVS_DP_SAMPLE_REG_SUPPORT = (1 << OVS_DP_SAMPLE_REG_SUPPORT_BIT),
Same here. The 'DP' part makes no sense as this is feature has
nothing to do with the datapath implementation.
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
t; +OVS_DP_GROUP_SUPPORT_BIT,
Not a full review, but this is not a good name for a feature bit.
Unlike meters, support for OF groups has nothing to do with the
datapath implementation. DP_GROUP is also confusing as this name
overlaps with datapath groups in Sb DB.
t;/>
>>
>> +Other Options
>> +> xmlns:xi="http://www.w3.org/2003/XInclude"/>
>> +
>
> This seems weird.. should it actually be part of common.xml?
It's a separator between sections, so it shouldn't be part of the section.
Our x
On 7/30/24 02:42, Numan Siddique wrote:
> On Mon, Jul 29, 2024 at 6:44 PM Ilya Maximets wrote:
>>
>> Just released sphinx 8.0 requires Pygments >= 2.17, but it fails to
>> build with it in practice. This breaks our CI completely:
>>
>> Traceback (most recent
rsion for now to unblock CI.
Can be reverted once https://github.com/sphinx-doc/sphinx/issues/12711
is fixed.
Signed-off-by: Ilya Maximets
---
utilities/containers/py-requirements.txt | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/utilities/containers/py-requirements.txt
On 7/29/24 14:39, Ilya Maximets wrote:
> It is annoying to have scheduled workflows running on your own fork,
> especially if there are flaky tests or the workflow is broken and
> permanently failing for an extended period of time generating an email
> for every failure every week.
r the
purpose of testing code before sending patches upstream. Real forks
have code changes already, and they are free to modify the workflows
if they want the scheduled ones.
Signed-off-by: Ilya Maximets
---
.github/workflows/containers.yml | 1 +
.github/workflows/ovn-fake-mult
Ilya Maximets (2):
github: Don't run scheduled workflows on forks.
github: Fix running multi-node tests on forks.
.github/workflows/containers.yml | 1 +
.github/workflows/ovn-fake-multinode-tests.yml | 8 +---
.github/workflows/ovn-kubernetes.yml | 1 +
.g
Forks may not have branch-22.03 in them. Even if they do, the branch
may not be up to date. Use the upstream one instead.
Signed-off-by: Ilya Maximets
---
.github/workflows/ovn-fake-multinode-tests.yml | 7 ---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/.github
alformed header.
We may still calculate a wrong checksum if the header is wrong, but within
the range, but we can't really do anything about that.
Another test for a short packet would be also good to have.
> }
>
> /* Returns 'true' if the packet 'p
On 7/25/24 07:28, Ales Musil wrote:
>
>
> On Wed, Jul 24, 2024 at 10:44 PM Ilya Maximets <mailto:i.maxim...@ovn.org>> wrote:
>
> On 7/24/24 08:23, Ales Musil wrote:
> > Fix wrong IPv6 notation which could result in test failures when
> > the DNS
On 7/25/24 00:45, Mike Pattrick wrote:
> On Wed, Jul 24, 2024 at 3:56 PM Ilya Maximets wrote:
>>
>> On 7/12/24 23:30, Mike Pattrick wrote:
>>> When prepending a tunnel header to a packet marked for segmentation, we
>>> need to adjust the segment size. Failu
sip=ae80::ea2aeafffe2800aa
> +sip=ae80::ea2a:eaff:fe28:00aa
> test_ipv6 ${i}3 f0:00:00:00:0${i}:${i}3 f0:00:00:00:00:21 $sip $tip
> done
>
Hmm. Does it return the IP address after DNS lookup fails?
If not, why the test doesn't fail in this case?
Bets regard
uot;), new_nbsp->name)) {
> +add_op_to_northd_tracked_ports(&trk_lsps->updated,
> + op_v);
Is there a chance that these ports end up in both updated and created?
Can this cause issues later?
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
e already small enough. Or am I missing something?
Best regards, Ilya Maximets.
>
> Fixes: 084c8087292c ("userspace: Support VXLAN and GENEVE TSO.")
> Signed-off-by: Mike Pattrick
> ---
> lib/netdev-native-tnl.c | 11 +++
> 1 file changed, 11 insertions(+
On 7/19/24 14:35, Phelan, Michael wrote:
>
>> -Original Message-
>> From: Ilya Maximets
>> Sent: Thursday, July 18, 2024 9:09 PM
>> To: Phelan, Michael
>> Cc: i.maxim...@ovn.org; ovs-dev
>> Subject: Re: [ovs-build] |fail| pw1960612 [ovs-dev,
wever, the code is still wrong and
a bit misleading, so should be fixed.
Fixes: 53178986d7fc ("ovsdb: Add support for online schema conversion.")
Signed-off-by: Ilya Maximets
---
ovsdb/transaction.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/ovsdb/transaction.c b/ovsdb/transacti
On 7/18/24 22:45, Mike Pattrick wrote:
> On Thu, Jul 18, 2024 at 3:12 PM Ilya Maximets wrote:
>>
>> There are two types of netlink notifier callbacks: nln_notify_func and
>> rtnetlink_notify_func. The rtnetlink_notify_func is only registered
>> via rtnetlink_notifi
On 7/19/24 07:04, Ales Musil wrote:
>
>
> On Thu, Jul 18, 2024 at 9:40 PM Ilya Maximets <mailto:i.maxim...@ovn.org>> wrote:
>
> This tests constructs a very large operation that frequently takes more
> than a second on busy CI systems. As a result ovn-nbctl
On 7/17/24 17:31, Phelan, Michael wrote:
>
>> -Original Message-
>> From: Ilya Maximets
>> Sent: Wednesday, July 17, 2024 10:43 AM
>> To: Phelan, Michael
>> Cc: i.maxim...@ovn.org; ovs-dev
>> Subject: Re: [ovs-build] |fail| pw1960612 [ovs-dev,
makes the test a bit slower, but
much more reliable.
Fixes: 7e99500e60bf ("northd: Add support for disabling vxlan mode.")
Signed-off-by: Ilya Maximets
---
tests/ovn-northd.at | 8 ++--
1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/tests/ovn-northd.at b/tests/ovn
in _start (vswitchd/ovs-vswitchd+0x42dfd4)
Turn off function sanitizing for nln_report() the same as we do for
RCU callbacks to avoid runtime errors with UBsan enabled.
Reproduced with OVN test suite running multiple tests in parallel.
Signed-off-by: Ilya Maximets
---
lib/netlink-notifier.c | 2
On 7/18/24 17:27, Ilya Maximets wrote:
> On 7/18/24 15:48, Simon Horman wrote:
>> The current QEMU documentation URLs, which point to a presumably
>> old Git repository, appear to time out.
>> Update it with a new links under https://www.qemu.org/docs
>>
>
>
>
> -.. _QEMU documentation:
> http://git.qemu-project.org/?p=qemu.git;a=blob;f=docs/specs/vhost-user.txt;h=7890d7169;hb=HEAD
> +.. _QEMU documentation:
> https://www.qemu.org/docs/master/interop/vhost-user.html
>
> Jumbo Frames
>
&g
hri vipul.as...@ericsson.com
> Vishal Swarankarvishal.swarn...@gmail.com
> Vjekoslav Brajkovic bal...@cs.washington.edu
> Voravit T. vora...@kth.se
Thanks, Simon! I was applying some other patches and picked
th
On 7/17/24 16:18, Mike Pattrick wrote:
> On Wed, Jul 17, 2024 at 8:01 AM Ilya Maximets wrote:
>>
>> GCC 14.1.1 of Fedora 41 thinks that 'i' can be in a full range and
>> so 8 bytes is not enough to print it.
>>
>> lib/match.c: In funct
On 7/17/24 15:40, Mike Pattrick wrote:
> On Wed, Jul 17, 2024 at 8:01 AM Ilya Maximets wrote:
>>
>> For some reason GCC 14.1.1 on Fedora 41 assumes that format can
>> be NULL and emits a warning:
>>
>> lib/util.c: In function 'xvasprintf':
>&
On 7/17/24 15:38, Mike Pattrick wrote:
> On Wed, Jul 17, 2024 at 8:01 AM Ilya Maximets wrote:
>>
>> For some reason gcc 14.1.1 from Fedora 41 thinks that the variable
>> may end up not initialized:
>>
>> ofproto/ofproto-dpif-xlate.c: In function 'compose_sa
On 7/17/24 13:55, Simon Horman wrote:
> On Wed, Jul 17, 2024 at 12:55:02PM +0200, Ilya Maximets wrote:
>> Read the Docs was always mangling the conf.py during the build to
>> inject custom domains configured in the project settings and some
>> other stuff. But they will st
On 7/16/24 15:12, Mike Pattrick wrote:
> On Tue, Jul 16, 2024 at 8:47 AM Ilya Maximets wrote:
>>
>> 'wc' can't be NULL there and if it can we'd already crash a few lines
>> before setting up vlan flags.
>>
>> The check is misleading as it m
346:5
>> 17 0x53eed7 in main /workspace/ovn/ovs/vswitchd/ovs-vswitchd.c:130:9
>>
>> Signed-off-by: Ales Musil
>
> Acked-by: Simon Horman
>
Thanks, Ales and Simon!
Applied and backported down to 2.17.
It is a legit issue, but for some reason I was not able to reproduc
On 7/16/24 15:40, Simon Horman wrote:
> On Tue, Jul 16, 2024 at 12:48:06PM +0200, Ilya Maximets wrote:
>> Found by Coverity:
>>
>> CID 397544: Incorrect expression (COPY_PASTE_ERROR)
>> "obs_point_src" in "(*os).obs_point_src.n_bits" looks
&g
ogical router.
Hi, Numan. Thanks for the change!
It seems like you missed the update for system tests. A few ECMP
system tests are failing in CI.
Also, not sure how big of a problem that is, but we may still spray
out uni-directional traffic. For example,
https://www.qemu.org/docs/master/interop/vhost-user.html
In the source tree it is in docs/interop/vhost-user.rst.
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
on removing the check entirely.
I think, we should keep the code as long as there is still one
supported version (2.17) that technically ships the modules.
We'll need to do a larger cleanup after 3.5 release when 2.17
will go EoL.
Also, I wonder which interface windows tunnels are using...
B
.
We could define a common reply_to_string() function for both dbctl
and appctl, but these programs beahve differently in regerads to exit
codes and we'll need to do more refactoring to handle the difference.
Will move to v3.4.0 once it is available.
Signed-off-by: Ilya Maximets
---
contr
xim...@ovn.org/
Though I definitely missed a few places and so the patch doesn't work
in its current form.
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
^~~
In practice that value can't be larger than 2, but it's not a
performance critical code, so let's just increase the size to
a maximum 12.
Signed-off-by: Ilya Maximets
---
lib/match.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-
~~~
I didn't find any users where this can be true. Adding an
assertion to silence the warning. In the worst case we'll
find out where it is being called incorrectly.
Signed-off-by: Ilya Maximets
---
lib/util.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/lib/util.c b/lib/ut
e an assertion in the code to ensure that at least one of
the actions is present (userspace or psample), so the variable
should actually be always initialized.
Initialize explicitly just to silence the warning.
Fixes: 516569d31fbf ("ofproto: xlate: Make sampled drops explicit.")
Sign
GCC 14.1.1 on Fedora 41 fails to build reporting some warnings when
building with '-fsanitize=address,undefined'.
Ilya Maximets (3):
ofproto-dpif-xlate: Initialize observe_offset for sample actions.
util: Add non-NULL format assertion to xvasprintf.
match: Fix false-positive snp
://about.readthedocs.com/blog/2024/07/addons-by-default/
Signed-off-by: Ilya Maximets
---
Documentation/conf.py | 8
1 file changed, 8 insertions(+)
diff --git a/Documentation/conf.py b/Documentation/conf.py
index 15785605a..2364405ad 100644
--- a/Documentation/conf.py
+++ b/Documentation/conf.py
@@ -12,6
On 7/17/24 10:36, Phelan, Michael wrote:
>
>> -Original Message-
>> From: Ilya Maximets
>> Sent: Tuesday, July 16, 2024 12:25 PM
>> To: Phelan, Michael
>> Cc: i.maxim...@ovn.org; ovs-dev
>> Subject: Re: [ovs-build] |fail| pw1960612 [ovs-dev,
(FORWARD_NULL)
25. var_deref_op: Dereferencing null pointer ctx->wc.
14. var_compare_op: Comparing ctx->wc to null implies that ctx->wc
might be null
Remove the check.
Fixes: 3b1882261c8b ("ofproto-dpif-mirror: Add support for pre-selection
filter.")
Signed-off-by: Ilya Maxi
pool ok
Hi, Michael. Could you, please, check the reason why these tests are failing?
The logs are truncated a little too much, so it's hard to tell what went wrong.
> 2024-07-15T19:14:11Z|7|dpdk|INFO|Using DPDK 23.11.0
I remember that there were some vhost issues in
mple action.")
Signed-off-by: Ilya Maximets
---
lib/ofp-actions.c| 2 +-
tests/ofp-actions.at | 2 ++
2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/lib/ofp-actions.c b/lib/ofp-actions.c
index 2a1f5c3c4..fe6a17b6d 100644
--- a/lib/ofp-actions.c
+++ b/lib/ofp-actions.c
@@ -
On 7/15/24 13:23, Ilya Maximets wrote:
> The plan is to branch by the end of today. There is still a couple of
> patch sets on the list that can / will be applied before that.
>
> Ilya Maximets (2):
> Prepare for 3.4.0.
> Prepare for post-3.4.0 (3.4.90).
>
> Docume
Hi, everyone.
branch-3.4 was just created. Please, test it and report issues!
Official release date, according to our release process, should be
on Thursday, August 15th (1 month from now).
Best regards, Ilya Maximets.
___
dev mailing list
d
_vlans)
> && mirror->out == out
> && mirror->out_vlan == out_vlan
> -&& mirror->snaplen == ms->snaplen)
> +&& mirror->snaplen == ms->snaplen
> +&& nullable_string_is_
o/ofproto-dpif-xlate-cache.c | 34
> ofproto/ofproto-dpif-xlate-cache.h | 2 +
> ofproto/ofproto-provider.h | 2 +
> ofproto/ofproto.c | 11 +--
> tests/ofproto-dpif.at | 56
> tests/ofproto-macros.at|
On 7/15/24 10:48, Timothy Redaelli wrote:
> On Fri, 12 Jul 2024 18:14:28 +0200
> Ilya Maximets wrote:
>
>> On 7/10/24 13:06, Timothy Redaelli wrote:
>>> Since on CentOS/RHEL the builds are based on stable branches and not on
>>> tags for debugging purpose i
Signed-off-by: Ilya Maximets
---
NEWS | 4
configure.ac | 2 +-
debian/changelog | 6 ++
3 files changed, 11 insertions(+), 1 deletion(-)
diff --git a/NEWS b/NEWS
index 507d91721..4ca9ae692 100644
--- a/NEWS
+++ b/NEWS
@@ -1,3 +1,7 @@
+Post-v3.4.0
The plan is to branch by the end of today. There is still a couple of
patch sets on the list that can / will be applied before that.
Ilya Maximets (2):
Prepare for 3.4.0.
Prepare for post-3.4.0 (3.4.90).
Documentation/faq/releases.rst | 1 +
NEWS | 6
Signed-off-by: Ilya Maximets
---
Documentation/faq/releases.rst | 1 +
NEWS | 2 +-
configure.ac | 2 +-
debian/changelog | 4 ++--
debian/rules | 4 ++--
5 files changed, 7 insertions(+), 6 deletions(-)
diff --git a
tem-traffic.at | 385
> tests/test-psample.c| 288 +
> vswitchd/bridge.c | 82 +-
> vswitchd/vswitch.ovsschema | 9 +-
> vswitchd/vswitch.xml| 70 -
> 42 files changed, 2462 insertions(+), 166 deletions(-)
> create mode 100644 include/linux/psample.h
> create mode 100644 ofproto/ofproto-dpif-lsample.c
> create mode 100644 ofproto/ofproto-dpif-lsample.h
> create mode 100644 tests/test-psample.c
>
Thanks, Adrian! I fixed a few small style issues and applied the set.
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
in table L loaded into it.
'''
So, you learn your rules into the table L setting some register, then in table
B + 1 you match on that register and perform any complex actions you like:
table A:
actions=learn(load:XXX->reg1)
table L:
actions=load:XXX->reg1
table B:
actions
v2 -> v3
> - Added a knob to configure explicit sampled drops.
> - Addressed Eelco comments.
> - Rebased.
I went thought the set. Most of the comments are simple nits that I
could even fix myself on commit, but I'm not sure what to do with
patches 7 and 8 as I don't think we need to report incorrect stats
and there seem to be no real way to get correct ones.
Let me know what do you think.
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
On 7/12/24 19:06, Adrian Moreno wrote:
> When sample action gets used as a way of sampling traffic with
> controller-generated metadata (i.e: obs_domain_id and obs_point_id),
> the controller will have to increase the number of flows to ensure each
> part of the pipeline contains the right metadata
On 7/12/24 19:06, Adrian Moreno wrote:
> When the flow translation results in a datapath action list whose last
> action is an "observational" action, i.e: one generated for IPFIX,
> sFlow or local sampling applications, the packet is actually going to be
> dropped (and observed).
>
> In that case
On 7/12/24 19:06, Adrian Moreno wrote:
> Test simultaneous IPFIX and local sampling including slow-path.
>
> Signed-off-by: Adrian Moreno
> ---
> tests/system-common-macros.at | 4 +
> tests/system-traffic.at | 358 ++
> 2 files changed, 362 insertions(+)
get the actual number of sampled packets and bytes.
Should we just drop patches 7 and 8 and remove the lsample/show calls from
the tests?
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
On 7/12/24 19:06, Adrian Moreno wrote:
> Use the newly added psample action to implement OpenFlow sample() actions
> with local sampling configuration if possible.
>
> A bit of refactoring in compose_sample_actions arguments helps make it a
> bit more readable.
>
> Acked-by: Eelco Chaudron
> Sig
Nit: I'd add something like:
See 'local-group-id' column in Flow_Sample_Collector_Set table.
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
On 7/12/24 19:06, Adrian Moreno wrote:
> Add a new resource in ofproto-dpif and the corresponding API in
> ofproto_provider.h to represent and local sampling configuration.
>
> Acked-by: Eelco Chaudron
> Signed-off-by: Adrian Moreno
> ---
> ofproto/automake.mk| 2 +
> ofproto/ofpr
101 - 200 of 1021 matches
Mail list logo