From: "Burakov, Anatoly" <anatoly.burakov@intel.com>
To: "Tan, Jianfeng" <jianfeng.tan@intel.com>, "dev@dpdk.org" <dev@dpdk.org>
Cc: "Ananyev, Konstantin" <konstantin.ananyev@intel.com>,
"Richardson, Bruce" <bruce.richardson@intel.com>,
"Pattan, Reshma" <reshma.pattan@intel.com>,
"viktorin@rehivetech.com" <viktorin@rehivetech.com>
Subject: Re: [dpdk-dev] [PATCH v2 07/12] eal: add channel for primary/secondary communication
Date: Mon, 2 Oct 2017 11:08:58 +0100 [thread overview]
Message-ID: <9ebf5003-ee17-df40-2344-fb6366a94b9a@intel.com> (raw)
In-Reply-To: <a82f1a15-8081-c383-27a0-55eda58d3e36@intel.com>
On 30-Sep-17 5:07 AM, Tan, Jianfeng wrote:
>
>
> On 9/29/2017 6:00 PM, Burakov, Anatoly wrote:
>> On 29-Sep-17 2:03 AM, Tan, Jianfeng wrote:
>>> + Reshma and Jan.
>>>
>>>> -----Original Message-----
>>>> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Burakov, Anatoly
>>>> Sent: Thursday, September 28, 2017 11:30 PM
>>>> To: dev@dpdk.org
>>>> Subject: Re: [dpdk-dev] [PATCH v2 07/12] eal: add channel for
>>>> primary/secondary communication
>>>>
>>>> On 28-Sep-17 4:01 PM, Ananyev, Konstantin wrote:
>>>>> Hi Jianfeng,
>>>>>
>>>>>
>>>>>> -----Original Message-----
>>>>>> From: Tan, Jianfeng
>>>>>> Sent: Thursday, September 28, 2017 2:56 PM
>>>>>> To: dev@dpdk.org
>>>>>> Cc: Richardson, Bruce <bruce.richardson@intel.com>; Ananyev,
>>>> Konstantin <konstantin.ananyev@intel.com>; De Lara Guarch, Pablo
>>>>>> <pablo.de.lara.guarch@intel.com>; thomas@monjalon.net;
>>>> yliu@fridaylinux.org; maxime.coquelin@redhat.com; mtetsuyah@gmail.com;
>>>>>> Yigit, Ferruh <ferruh.yigit@intel.com>; Tan, Jianfeng
>>>> <jianfeng.tan@intel.com>
>>>>>> Subject: [PATCH v2 07/12] eal: add channel for primary/secondary
>>>> communication
>>>>>>
>>>>>> Previouly, there is only one way for primary/secondary to exchange
>>>>>> messages, that is, primary process writes info into some predefind
>>>>>> file, and secondary process reads info out. That cannot address
>>>>>> the requirements:
>>>>>> a. Secondary wants to send info to primary, for example,
>>>>>> secondary
>>>>>> would like to send request (about some specific vdev to
>>>>>> primary).
>>>>>> b. Sending info at any time, instead of just initialization time.
>>>>>> c. Share FDs with the other side, for vdev like vhost, related
>>>>>> FDs
>>>>>> (memory region, kick) should be shared.
>>>>>>
>>>>>> This patch proposes to create a communication channel, as an unix
>>>>>> socket connection, for above requirements. Primary will listen on
>>>>>> the unix socket; secondary will connect this socket to talk.
>>>>>>
>>>>>> Three new APIs are added:
>>>>>>
>>>>>> 1. rte_eal_mp_action_register is used to register an action,
>>>>>> indexed by a string; if the calling component wants to
>>>>>> response the messages from the corresponding component in
>>>>>> its primary process or secondary processes.
>>>>>> 2. rte_eal_mp_action_unregister is used to unregister the action
>>>>>> if the calling component does not want to response the
>>>>>> messages.
>>>>>> 3. rte_eal_mp_sendmsg is used to send a message.
>>>>>
>>>>> I think we already have similar channel in librte_pdump().
>>>>> Also it seems like eal_vfio also has it's own socket to communicate
>>>> between mp/sp.
>>>>> Could we probably make it generic - so same code (and socket) be
>>>>> used by
>>>> all such places.
>>>>> Konstantin
>>>>>
>>>>
>>>> Agreed, however looking at this, it's already a generic-enough
>>>> solution,
>>>> and other places could be fixed to use this in later releases.
>>>
>>> Yes, to provide a generic communication way, instead of one channel
>>> for each subsystem, is the goal of this patch.
>>>
>>> Reshma and Jan, can I ask comment from you to have a look if the way
>>> of this patch is generic enough to cover pdump and vfio-sync's
>>> requirement?
>>>
>>> Possible limitation of this patch (so far) is that it only provides
>>> an async way for request/response, do we need synchronous way?
>>>
>>>> That said, i believe this particular part of the patchset should go
>>>> in as a
>>>> separate patchset and more design consideration and input from others.
>>>
>>> OK, let's collect more info here, and then take out this patch out as
>>> a separate patch.
>>>
>>> Thanks,
>>> Jianfeng
>>>
>> Hi Jianfeng,
>>
>> Yes, i believe VFIO does need synchronous communcation, because it
>> relies on passing fd's from primary to secondary, on request. It could
>> be rewritten to be asynchronous, similarly to how you handle vdevs
>> here, but as it stands, it assumes synchronous communication.
>>
>
> Good to know, thanks Anatoly. Even it can be rewritten to do in asyn
> way, do we need to propose sync way now?
>
> Thanks,
> Jianfeng
>
I believe that we do, because we can't assume that everything can be
rewritten to be asynchronous. I'm open to other opinions though :)
--
Thanks,
Anatoly
next prev parent reply other threads:[~2017-10-02 10:09 UTC|newest]
Thread overview: 158+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-08-25 9:40 [dpdk-dev] [PATCH 00/12] support to run vdev in the secondary process Jianfeng Tan
2017-08-25 9:40 ` [dpdk-dev] [PATCH 01/12] cryptodev: remove crypto vdev init Jianfeng Tan
2017-09-18 11:48 ` De Lara Guarch, Pablo
2017-08-25 9:40 ` [dpdk-dev] [PATCH 02/12] eal: avoid calling rte_vdev_init() Jianfeng Tan
2017-08-29 12:50 ` Gaëtan Rivet
2017-08-29 22:25 ` Tan, Jianfeng
2017-08-25 9:40 ` [dpdk-dev] [PATCH 03/12] crypto: move vdev helper functions into dedicated file Jianfeng Tan
2017-09-18 11:51 ` De Lara Guarch, Pablo
2017-08-25 9:40 ` [dpdk-dev] [PATCH 04/12] vdev: move to drivers/bus Jianfeng Tan
2017-08-29 13:04 ` Gaëtan Rivet
2017-08-29 22:47 ` Tan, Jianfeng
2017-09-18 11:47 ` De Lara Guarch, Pablo
2017-09-19 6:01 ` Tan, Jianfeng
2017-08-25 9:40 ` [dpdk-dev] [PATCH 05/12] bus/vdev: change log type from EAL to PMD Jianfeng Tan
2017-08-29 12:54 ` Gaëtan Rivet
2017-08-29 22:27 ` Tan, Jianfeng
2017-08-25 9:40 ` [dpdk-dev] [PATCH 06/12] eal: add channel for primary/secondary communication Jianfeng Tan
2017-09-18 13:49 ` Jiayu Hu
2017-09-21 6:11 ` Tan, Jianfeng
2017-09-20 3:00 ` Jiayu Hu
2017-09-21 6:53 ` Tan, Jianfeng
2017-09-27 12:19 ` Yuanhan Liu
2017-09-28 13:50 ` Tan, Jianfeng
2017-09-29 1:24 ` Yuanhan Liu
2017-09-29 10:09 ` Burakov, Anatoly
2017-09-29 10:25 ` Yuanhan Liu
2017-08-25 9:40 ` [dpdk-dev] [PATCH 07/12] bus/vdev: scan and probe vdev in secondary processes Jianfeng Tan
2017-08-25 9:40 ` [dpdk-dev] [PATCH 08/12] ethdev: support attach vdev in secondary process Jianfeng Tan
2017-08-25 9:40 ` [dpdk-dev] [PATCH 09/12] vhost: allocate virtio_net in memzone Jianfeng Tan
2017-08-25 9:40 ` [dpdk-dev] [PATCH 10/12] vhost: support to kick in secondary process Jianfeng Tan
2017-09-21 3:33 ` Yuanhan Liu
2017-09-21 7:04 ` Tan, Jianfeng
2017-09-21 9:17 ` Yuanhan Liu
2017-09-22 2:30 ` Tan, Jianfeng
2017-09-27 9:36 ` Yuanhan Liu
2017-09-28 5:10 ` Tan, Jianfeng
2017-09-28 8:09 ` Tan, Jianfeng
2017-09-30 8:18 ` Yuanhan Liu
2017-09-30 10:50 ` Tan, Jianfeng
2017-08-25 9:40 ` [dpdk-dev] [PATCH 11/12] net/vhost: support to run in the " Jianfeng Tan
2017-09-21 4:29 ` Yuanhan Liu
2017-08-25 9:40 ` [dpdk-dev] [PATCH 12/12] examples/helloworld: do not exit automatically Jianfeng Tan
2017-09-18 11:44 ` De Lara Guarch, Pablo
2017-09-19 5:07 ` Tan, Jianfeng
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 00/12] support to run vdev in the secondary process Jianfeng Tan
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 01/12] cryptodev: remove crypto vdev init API Jianfeng Tan
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 02/12] eal: avoid calling rte_vdev_init() Jianfeng Tan
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 03/12] cryptodev: avoid dependency on rte_vdev.h Jianfeng Tan
2017-10-05 13:13 ` Jan Blunck
2017-10-09 1:04 ` Tan, Jianfeng
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 04/12] bus/fslmc: introduce RTE_LOGTYPE_BUS for bus drivers Jianfeng Tan
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 05/12] bus/vdev: move to vdev bus to drivers/bus Jianfeng Tan
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 06/12] bus/vdev: normalize log type Jianfeng Tan
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 07/12] eal: add channel for primary/secondary communication Jianfeng Tan
2017-09-28 15:01 ` Ananyev, Konstantin
2017-09-28 15:29 ` Burakov, Anatoly
2017-09-29 1:03 ` Tan, Jianfeng
2017-09-29 10:00 ` Burakov, Anatoly
2017-09-30 4:07 ` Tan, Jianfeng
2017-10-02 10:08 ` Burakov, Anatoly [this message]
2017-10-05 12:01 ` Jan Blunck
2017-10-09 1:27 ` Tan, Jianfeng
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 08/12] bus/vdev: scan and probe vdev in secondary processes Jianfeng Tan
2017-10-05 13:04 ` Jan Blunck
2017-10-09 1:08 ` Tan, Jianfeng
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 09/12] ethdev: support attach vdev in secondary process Jianfeng Tan
2017-10-05 14:26 ` Jan Blunck
2017-10-09 0:56 ` Tan, Jianfeng
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 10/12] vhost: allocate virtio_net in memzone Jianfeng Tan
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 11/12] vhost: support to kick in secondary process Jianfeng Tan
2017-09-28 13:55 ` [dpdk-dev] [PATCH v2 12/12] net/vhost: support to run in the " Jianfeng Tan
2017-09-29 8:28 ` Yuanhan Liu
2017-09-30 4:03 ` Tan, Jianfeng
2017-09-30 8:16 ` Yuanhan Liu
2017-09-30 10:06 ` Tan, Jianfeng
2017-09-30 11:49 ` Yuanhan Liu
2017-10-01 23:48 ` Tan, Jianfeng
2017-09-30 8:23 ` Yuanhan Liu
2017-09-30 10:53 ` Tan, Jianfeng
2017-09-30 11:34 ` Yuanhan Liu
2017-10-01 23:46 ` Tan, Jianfeng
2017-10-09 3:20 ` [dpdk-dev] [PATCH v3 0/5] move vdev into drivers/bus Jianfeng Tan
2017-10-09 3:20 ` [dpdk-dev] [PATCH v3 1/5] cryptodev: remove crypto vdev init API Jianfeng Tan
2017-10-09 3:20 ` [dpdk-dev] [PATCH v3 2/5] eal: avoid calling rte_vdev_init() Jianfeng Tan
2017-10-09 3:20 ` [dpdk-dev] [PATCH v3 3/5] bus: introduce RTE_LOGTYPE_BUS for bus drivers Jianfeng Tan
2017-10-09 3:20 ` [dpdk-dev] [PATCH v3 4/5] bus/vdev: move to vdev bus to drivers/bus Jianfeng Tan
2017-10-09 3:20 ` [dpdk-dev] [PATCH v3 5/5] bus/vdev: normalize log type Jianfeng Tan
2017-10-09 10:55 ` [dpdk-dev] [PATCH v4 0/5] move vdev into drivers/bus Jianfeng Tan
2017-10-09 10:55 ` [dpdk-dev] [PATCH v4 1/5] bus/vdev: scan and probe vdev in secondary processes Jianfeng Tan
2017-10-09 10:55 ` [dpdk-dev] [PATCH v4 2/5] ethdev: support attach vdev in secondary process Jianfeng Tan
2017-10-09 10:55 ` [dpdk-dev] [PATCH v4 3/5] vhost: allocate virtio_net in memzone Jianfeng Tan
2017-10-09 10:55 ` [dpdk-dev] [PATCH v4 4/5] vhost: support to kick in secondary process Jianfeng Tan
2017-10-09 10:55 ` [dpdk-dev] [PATCH v4 5/5] net/vhost: support to run in the " Jianfeng Tan
2017-10-09 11:08 ` [dpdk-dev] [PATCH v4 0/5] move vdev into drivers/bus Tan, Jianfeng
2017-10-09 11:27 ` [dpdk-dev] [PATCH v5 " Jianfeng Tan
2017-10-09 11:27 ` [dpdk-dev] [PATCH v5 1/5] cryptodev: remove crypto vdev init API Jianfeng Tan
2017-10-09 11:27 ` [dpdk-dev] [PATCH v5 2/5] eal: remove dependency on vdev Jianfeng Tan
2017-10-09 11:27 ` [dpdk-dev] [PATCH v5 3/5] bus: introduce new log type for bus drivers Jianfeng Tan
2017-10-11 6:54 ` Shreyansh Jain
2017-10-11 10:42 ` Tan, Jianfeng
2017-10-11 11:20 ` Shreyansh Jain
2017-10-12 2:14 ` Tan, Jianfeng
2017-10-09 11:27 ` [dpdk-dev] [PATCH v5 4/5] bus/vdev: move to vdev bus to drivers/bus Jianfeng Tan
2017-10-09 11:27 ` [dpdk-dev] [PATCH v5 5/5] bus/vdev: normalize log type Jianfeng Tan
2017-10-12 8:46 ` [dpdk-dev] [PATCH v6 0/4] move vdev into drivers/bus Jianfeng Tan
2017-10-12 8:46 ` [dpdk-dev] [PATCH v6 1/4] cryptodev: remove crypto vdev init API Jianfeng Tan
2017-10-12 10:06 ` Thomas Monjalon
2017-10-12 8:46 ` [dpdk-dev] [PATCH v6 2/4] eal: remove dependency on vdev Jianfeng Tan
2017-10-12 8:46 ` [dpdk-dev] [PATCH v6 3/4] bus/vdev: move to vdev bus to drivers/bus Jianfeng Tan
2017-10-12 8:46 ` [dpdk-dev] [PATCH v6 4/4] bus/vdev: change log type Jianfeng Tan
2017-10-13 2:04 ` [dpdk-dev] [PATCH v7 0/4] move vdev into drivers/bus Jianfeng Tan
2017-10-13 2:04 ` [dpdk-dev] [PATCH v7 1/4] ethdev: support attach vdev in secondary process Jianfeng Tan
2017-10-13 2:04 ` [dpdk-dev] [PATCH v7 2/4] vhost: allocate virtio_net in memzone Jianfeng Tan
2017-10-13 2:04 ` [dpdk-dev] [PATCH v7 3/4] vhost: support to kick in secondary process Jianfeng Tan
2017-10-13 2:04 ` [dpdk-dev] [PATCH v7 4/4] net/vhost: support to run in the " Jianfeng Tan
2017-10-13 8:26 ` [dpdk-dev] [PATCH v7 0/4] move vdev into drivers/bus Thomas Monjalon
2017-10-13 11:49 ` Tan, Jianfeng
2017-10-13 11:51 ` Jianfeng Tan
2017-10-13 11:51 ` [dpdk-dev] [PATCH v7 1/4] cryptodev: remove crypto vdev init API Jianfeng Tan
2017-10-23 10:06 ` De Lara Guarch, Pablo
2017-10-13 11:51 ` [dpdk-dev] [PATCH v7 2/4] eal: remove dependency on vdev Jianfeng Tan
2017-10-13 11:52 ` [dpdk-dev] [PATCH v7 3/4] bus/vdev: move to vdev bus to drivers/bus Jianfeng Tan
2017-10-23 14:34 ` De Lara Guarch, Pablo
2017-10-13 11:52 ` [dpdk-dev] [PATCH v7 4/4] bus/vdev: change log type Jianfeng Tan
2017-10-25 17:10 ` [dpdk-dev] [PATCH v8 0/4] move vdev into drivers/bus Jianfeng Tan
2017-10-25 17:10 ` [dpdk-dev] [PATCH v8 1/4] cryptodev: remove crypto vdev init API Jianfeng Tan
2017-10-25 17:10 ` [dpdk-dev] [PATCH v8 2/4] eal: remove dependency on vdev Jianfeng Tan
2017-10-25 17:10 ` [dpdk-dev] [PATCH v8 3/4] bus/vdev: move to vdev bus to drivers/bus Jianfeng Tan
2017-10-25 21:32 ` De Lara Guarch, Pablo
2017-10-25 23:03 ` Gaëtan Rivet
2017-10-25 17:10 ` [dpdk-dev] [PATCH v8 4/4] bus/vdev: change log type Jianfeng Tan
2017-10-27 1:06 ` [dpdk-dev] [PATCH v0 0/4] move vdev into drivers/bus Jianfeng Tan
2017-10-27 1:06 ` [dpdk-dev] [PATCH v9 1/4] cryptodev: remove crypto vdev init API Jianfeng Tan
2017-10-27 1:06 ` [dpdk-dev] [PATCH v9 2/4] eal: remove dependency on vdev Jianfeng Tan
2017-10-27 1:06 ` [dpdk-dev] [PATCH v9 3/4] bus/vdev: move to vdev bus to drivers/bus Jianfeng Tan
2017-10-27 7:56 ` Thomas Monjalon
2017-10-27 8:19 ` Tan, Jianfeng
2017-10-27 8:53 ` Thomas Monjalon
2017-10-27 16:57 ` Gaëtan Rivet
2017-10-27 1:06 ` [dpdk-dev] [PATCH v9 4/4] bus/vdev: change log type Jianfeng Tan
2017-10-27 3:23 ` [dpdk-dev] [PATCH v10 0/4] move vdev into drivers/bus Jianfeng Tan
2017-10-27 3:23 ` [dpdk-dev] [PATCH v10 1/4] cryptodev: remove crypto vdev init API Jianfeng Tan
2017-10-27 3:23 ` [dpdk-dev] [PATCH v10 2/4] eal: remove dependency on vdev Jianfeng Tan
2017-10-27 3:23 ` [dpdk-dev] [PATCH v10 3/4] bus/vdev: move to vdev bus to drivers/bus Jianfeng Tan
2017-10-27 3:23 ` [dpdk-dev] [PATCH v10 4/4] bus/vdev: change log type Jianfeng Tan
2017-10-30 8:28 ` [dpdk-dev] [PATCH v11 0/4] move vdev into drivers/bus Jianfeng Tan
2017-10-30 8:28 ` [dpdk-dev] [PATCH v11 1/4] cryptodev: remove crypto vdev init API Jianfeng Tan
2017-10-30 8:28 ` [dpdk-dev] [PATCH v11 2/4] eal: remove dependency on vdev Jianfeng Tan
2017-10-30 8:28 ` [dpdk-dev] [PATCH v11 3/4] bus/vdev: move to vdev bus to drivers/bus Jianfeng Tan
2017-11-07 2:43 ` Thomas Monjalon
2017-11-07 6:21 ` Tan, Jianfeng
2017-10-30 8:28 ` [dpdk-dev] [PATCH v11 4/4] bus/vdev: change log type Jianfeng Tan
2017-11-07 6:54 ` [dpdk-dev] [PATCH v12 0/4] move vdev into drivers/bus Jianfeng Tan
2017-11-07 6:54 ` [dpdk-dev] [PATCH v12 1/4] cryptodev: remove crypto vdev init API Jianfeng Tan
2017-11-07 6:54 ` [dpdk-dev] [PATCH v12 2/4] eal: remove dependency on vdev Jianfeng Tan
2017-11-07 6:54 ` [dpdk-dev] [PATCH v12 3/4] bus/vdev: move to vdev bus to drivers/bus Jianfeng Tan
2017-11-07 6:54 ` [dpdk-dev] [PATCH v12 4/4] bus/vdev: change log type Jianfeng Tan
2017-11-07 15:43 ` [dpdk-dev] [PATCH v12 0/4] move vdev into drivers/bus Thomas Monjalon
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=9ebf5003-ee17-df40-2344-fb6366a94b9a@intel.com \
--to=anatoly.burakov@intel.com \
--cc=bruce.richardson@intel.com \
--cc=dev@dpdk.org \
--cc=jianfeng.tan@intel.com \
--cc=konstantin.ananyev@intel.com \
--cc=reshma.pattan@intel.com \
--cc=viktorin@rehivetech.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).