-邮件原件-
发件人: Mike Pattrick
发送时间: 2023年9月19日 23:58
收件人: Dexia Li
抄送: David Marchand ; ovs-dev@openvswitch.org;
i.maxim...@ovn.org; Eelco Chaudron ; Qingmin Liu
; Joey Xing
主题: Re: [ovs-dev] [PATCH] netdev-dpdk: fix tso bug
On Tue, Sep 19, 2023 at 7:51 AM Dexia Li wrote:
>
> Hi,
Thanks Xavier,
Acked-by: Mark Michelson
On 9/18/23 12:46, Xavier Simonart wrote:
When ha-chassis-group is updated, causing a new hv to become master,
port move to a different chassis, and flows get created in table
ls_in_dhcp_options.
If pb->chassis is reported empty by sb (because set to []
Thank you Ihar and Ales.
I pushed the change to main and all branches back to 22.03.
The speedup is amazing even with the daemon restarting for each test.
Out of curiosity, I tried to determine how we could run the scapy daemon
globally for the entire testsuite run. In my (very brief) look at
Thanks Ales,
Acked-by: Mark Michelson
On 9/15/23 08:15, Ales Musil wrote:
When client sends RS we reply with RA to the source address,
however the client is allowed to send RS with unspecified
source address ("::"). In that case we would send the RA to
that empty address which is not correct.
On Tue, Sep 19, 2023 at 7:51 AM Dexia Li wrote:
>
> Hi, David
>
> Sorry, I am missing something too...
>
> Firstly, RTE_MBUF_F_TX_TCP_SEG flag in mbuf->ol_flags implies
> RTE_MBUF_F_TX_TCP_CKSUM in last letter.
> Some drivers such as ice can use RTE_MBUF_F_TX_TCP_SEG with
>
Thanks for the patch set Xavier.
With the exception of patch 13:
Acked-by: Mark Michelson
I'm acking the patch set, but the nature of these changes worries me. I
could easily see another patch series like this needed in the future
because it's too easy to write flaky tests. Here are some
Of the 15 patches in this set, this is the only one that I disagree with.
The problem with this change is that the test writer likely has a
specific amount of recomputes they expect to see based on the changes
that are applied. With the "-le" change, it means that the test could
pass in a
On 26 Aug 2023, at 8:01, mit...@outlook.com wrote:
From: Lin Huang
OvS has supported packet-per-second policer which can be set at
ingress
and egress side in kernel datapath. But the userspace datapath dosen't
support for ingress and egress packet-per-second policing now.
So, this patch
On 26 Aug 2023, at 8:01, mit...@outlook.com wrote:
From: Lin Huang
OvS has supported packet-per-second policer which can be set at
ingress
and egress side in kernel datapath. But the userspace datapath doesn't
support for ingress and egress packet-per-second policing now.
So, this patch
On 26 Aug 2023, at 8:01, mit...@outlook.com wrote:
From: Lin Huang
Currently srtcm_policer free packet one by one, if packets are exceed
rate limit.
That is a inefficient way to free memory which we have to call
rte_pktmbuf_free()
pkt_cnt times.
To improve this, we can use
On 26 Aug 2023, at 8:01, mit...@outlook.com wrote:
From: Lin Huang
Now, token-bucket 'last_fill' is updated by token_bucket_withdraw()
itself.
Add a new function parameter 'now' to update timestamp by caller.
Signed-off-by: Lin Huang
Thanks Lin for following up, and sorry for the late
On 9/12/23 11:53, Ales Musil wrote:
> From: Ihar Hrachyshka
>
> This function will later be used to handle port binding updates for
> postponed (throttled) bindings.
>
> Conflicts:
> controller/binding.c
>
> Signed-off-by: Ihar Hrachyshka
> Acked-by: Mark Michelson
> Signed-off-by:
On 9/18/23 14:48, Ales Musil wrote:
> To avoid any warning spam in the northd.log remove
> the "hosting-chassis" status only if it was previously
> specified.
>
> Fixes: 19164b030404 ("Expose distributed gateway port information in NB DB")
> Reported-at: https://issues.redhat.com/browse/FDP-54
>
On 9/12/23 11:53, Ales Musil wrote:
> From: Ihar Hrachyshka
>
> When multiple chassis are fighting for the same port (requested-chassis
> is not set, e.g. for gateway ports), they may produce an unreasonable
> number of chassis field updates in a very short time frame (hundreds of
> updates in
On Fri, Sep 8, 2023 at 12:04 AM Naveen Yerramneni
wrote:
>
> Hi Team,
>
> Could someone please look into this ?
>
> Thanks,
> Naveen
>
> > On 23-Aug-2023, at 8:06 AM, Naveen Yerramneni
> > wrote:
> >
> > This patch contains changes to enable DHCP Proxy Agent support for overlay
> > subnets.
>
Ilya Maximets, Sep 19, 2023 at 13:47:
> With flexibility of appctl comes absolutely no guarantee for API
> stability. But as soon as we have structured output, someone will
> expect it. If we can agree that users cannot rely on the structure
> of that structured output, then it's fine.
Hi, David
Sorry, I am missing something too...
Firstly, RTE_MBUF_F_TX_TCP_SEG flag in mbuf->ol_flags implies
RTE_MBUF_F_TX_TCP_CKSUM in last letter.
Some drivers such as ice can use RTE_MBUF_F_TX_TCP_SEG with
RTE_MBUF_F_TX_TCP_CKSUM but
Some drivers such as Iavf will result in driver hang
On 9/19/23 13:27, Eelco Chaudron wrote:
>
>
> On 19 Sep 2023, at 10:42, Robin Jarry wrote:
>
>> Adrian Moreno, Sep 19, 2023 at 09:18:
Both OVSDB and appctl are literally JSON-RPC protocols. There is no
need to re-invent anything.
>>>
>>> Right. Isn't appctl simpler in this case?
On 19 Sep 2023, at 13:25, Simon Horman wrote:
> On Thu, Sep 14, 2023 at 03:52:50PM +0200, Eelco Chaudron wrote:
>>
>>
>> On 14 Sep 2023, at 15:44, Eelco Chaudron wrote:
>>
>>> This patch adds WARNINGs for the subject line length and the format,
>>> i.e., the sentence should start with a capital
On 19 Sep 2023, at 10:42, Robin Jarry wrote:
> Adrian Moreno, Sep 19, 2023 at 09:18:
>>> Both OVSDB and appctl are literally JSON-RPC protocols. There is no
>>> need to re-invent anything.
>>
>> Right. Isn't appctl simpler in this case? IIUC, it would still satisfy
>> the requirements: client
On Thu, Sep 14, 2023 at 03:52:50PM +0200, Eelco Chaudron wrote:
>
>
> On 14 Sep 2023, at 15:44, Eelco Chaudron wrote:
>
> > This patch adds WARNINGs for the subject line length and the format,
> > i.e., the sentence should start with a capital and end with a dot.
> >
> > Acked-by: Simon Horman
Hello,
Please don't top post.
On Mon, Sep 18, 2023 at 11:47 AM Dexia Li wrote:
>
> Hi,david
>
> In dpdk 22.11 rte_mbuf_core.h file, there are notes as follows:
>
> /**
> * TCP segmentation offload. To enable this offload feature for a
> * packet to be transmitted on hardware supporting TSO:
>
Adrian Moreno, Sep 19, 2023 at 09:18:
> > Both OVSDB and appctl are literally JSON-RPC protocols. There is no
> > need to re-invent anything.
>
> Right. Isn't appctl simpler in this case? IIUC, it would still satisfy
> the requirements: client decides update interval, flexible schema, etc
> with
On 9/18/23 20:24, Ilya Maximets wrote:
On 9/12/23 15:47, Eelco Chaudron wrote:
On 12 Sep 2023, at 15:19, Robin Jarry wrote:
Eelco Chaudron, Sep 12, 2023 at 09:17:
I feel like if we do need another way of getting (real time)
statistics out of OVS, we should use the same communication
24 matches
Mail list logo