From: Parav Pandit <parav@nvidia.com>
To: zhenwei pi <pizhenwei@bytedance.com>,
Stefan Hajnoczi <stefanha@redhat.com>
Cc: "mst@redhat.com" <mst@redhat.com>,
"jasowang@redhat.com" <jasowang@redhat.com>,
"virtio-comment@lists.oasis-open.org"
<virtio-comment@lists.oasis-open.org>,
"houp@yusur.tech" <houp@yusur.tech>,
"helei.sig11@bytedance.com" <helei.sig11@bytedance.com>,
"xinhao.kong@duke.edu" <xinhao.kong@duke.edu>
Subject: RE: [virtio-comment] RE: RE: Re: Re: Re: [PATCH v2 06/11] transport-fabrics: introduce command set
Date: Sun, 11 Jun 2023 20:56:12 +0000 [thread overview]
Message-ID: <PH0PR12MB5481B15C36FA7600FB9CFD73DC57A@PH0PR12MB5481.namprd12.prod.outlook.com> (raw)
In-Reply-To: <36cf507b-2dd8-2d77-c6a9-b561e5ba491e@bytedance.com>
> From: zhenwei pi <pizhenwei@bytedance.com>
> Sent: Thursday, June 8, 2023 11:55 PM
> I tried to understand your proposal, please correct me if I misunderstand...
>
> Define data structure like:
>
> struct virtio_of_keyed_desc {
> le64 addr;
> le32 length;
> le32 key;
> };
>
> struct virtio_of_command_vq {
> le16 opcode;
> le16 command_id;
> le32 out_length;
> le32 in_length;
> union {
> struct virtio_of_keyed {
> le32 out_offset;
> };
>
> struct virtio_of_stream {
> u8 rsvd[4];
> };
> };
> };
>
> struct virtio_of_completion {
> le16 status;
> le16 command_id;
> u8 rsvd[4];
> union {
> le64 value;
> struct virtio_of_vq_completion {
> le32 in_length;
> le32 len;
> };
> }
> };
>
>
> For stream(Ex TCP/IP), the request PDU includes [struct virtio_of_command_vq
> + data], the response PDU includes [struct virtio_of_completion + data].
>
> For keyed(Ex RDMA), the request PDU includes [struct virtio_of_command_vq +
> struct virtio_of_keyed_desc], there are 2 opcodes for keyed transmission:
> 1, opcode virtio_of_op_vq: (basic and required command) the initiator prepares
> a buffer of [out_length + in_length], the target recv a 32B command, and reads
> the remote memory [addr, addr+out_length) by RDMA READ, then writes the
> remote memory [addr+out_length,
> addr+out_length+in_length) by RDMA WRITE, finally sends completion by
> RDMA SEND.
>
Maybe we can switch to the 64B format which has two benefits.
1. separate RDMA buffer for in, out xfer as each can have different DMA attributes.
2. ability to have one or more inline descs
A good way is to negotiate the max_cmd_size minimum being 32, maximum being a finite reasonable number of 64 or 128.
> 2, opcode virtio_of_op_vq_write_inline: (optional command)
> the initiator gets a remote buffer of target(Ex, 128K) after feature
> negotiation.
>
> The initiator selects a region of target remote memory(Ex, 4k - 12k),
> and writes payload by RDMA WRITE, then sends a 32B command by RDMA
> SEND(out_offset is 4K, ).
> The target handles command, writes the remote memory [addr,
> addr+in_length), finally sends completion by RDMA SEND.
Yes.
next prev parent reply other threads:[~2023-06-11 20:56 UTC|newest]
Thread overview: 74+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-04 8:18 [virtio-comment] [PATCH v2 00/11] Introduce Virtio Over Fabrics zhenwei pi
2023-05-04 8:19 ` [virtio-comment] [PATCH v2 01/11] transport-fabrics: introduce Virtio Over Fabrics overview zhenwei pi
2023-05-04 8:57 ` David Hildenbrand
2023-05-04 9:46 ` zhenwei pi
2023-05-04 10:05 ` Michael S. Tsirkin
2023-05-04 10:12 ` David Hildenbrand
2023-05-04 10:50 ` Re: " zhenwei pi
2023-05-31 14:00 ` [virtio-comment] " Stefan Hajnoczi
2023-06-02 1:17 ` [virtio-comment] " zhenwei pi
2023-06-05 2:39 ` [virtio-comment] " Parav Pandit
2023-06-05 2:39 ` Parav Pandit
2023-05-04 8:19 ` [virtio-comment] [PATCH v2 02/11] transport-fabrics: introduce Virtio Qualified Name zhenwei pi
2023-05-31 14:06 ` Stefan Hajnoczi
2023-06-02 1:50 ` zhenwei pi
2023-06-05 2:40 ` Parav Pandit
2023-06-05 7:57 ` zhenwei pi
2023-06-05 17:05 ` Stefan Hajnoczi
2023-05-04 8:19 ` [virtio-comment] [PATCH v2 03/11] transport-fabircs: introduce Segment Descriptor Definition zhenwei pi
2023-05-31 14:23 ` Stefan Hajnoczi
2023-06-02 3:08 ` zhenwei pi
2023-06-05 2:40 ` [virtio-comment] " Parav Pandit
2023-05-04 8:19 ` [virtio-comment] [PATCH v2 04/11] transport-fabrics: introduce Stream Transmission zhenwei pi
2023-05-31 15:20 ` Stefan Hajnoczi
2023-06-02 2:26 ` zhenwei pi
2023-06-05 16:11 ` Stefan Hajnoczi
2023-06-06 3:13 ` zhenwei pi
2023-06-06 13:09 ` Stefan Hajnoczi
2023-05-04 8:19 ` [virtio-comment] [PATCH v2 05/11] transport-fabrics: introduce Keyed Transmission zhenwei pi
2023-05-31 16:20 ` [virtio-comment] " Stefan Hajnoczi
2023-06-01 9:02 ` zhenwei pi
2023-06-01 11:33 ` Stefan Hajnoczi
2023-06-01 13:09 ` zhenwei pi
2023-06-01 19:13 ` Stefan Hajnoczi
2023-06-01 21:23 ` Stefan Hajnoczi
2023-06-02 0:55 ` zhenwei pi
2023-06-05 17:21 ` Stefan Hajnoczi
2023-06-05 2:41 ` Parav Pandit
2023-06-05 8:41 ` zhenwei pi
2023-06-05 11:45 ` Parav Pandit
2023-06-05 12:50 ` zhenwei pi
2023-06-05 13:12 ` Parav Pandit
2023-06-06 7:13 ` zhenwei pi
2023-06-06 21:52 ` Parav Pandit
2023-05-04 8:19 ` [virtio-comment] [PATCH v2 06/11] transport-fabrics: introduce command set zhenwei pi
2023-05-31 17:10 ` [virtio-comment] " Stefan Hajnoczi
2023-06-02 5:15 ` [virtio-comment] " zhenwei pi
2023-06-05 16:30 ` Stefan Hajnoczi
2023-06-06 1:31 ` [virtio-comment] " zhenwei pi
2023-06-06 13:34 ` Stefan Hajnoczi
2023-06-07 2:58 ` [virtio-comment] " zhenwei pi
2023-06-08 16:41 ` Stefan Hajnoczi
2023-06-08 17:01 ` [virtio-comment] " Parav Pandit
2023-06-09 1:39 ` [virtio-comment] " zhenwei pi
2023-06-09 2:06 ` [virtio-comment] " Parav Pandit
2023-06-09 3:55 ` zhenwei pi
2023-06-11 20:56 ` Parav Pandit [this message]
2023-06-06 2:02 ` [virtio-comment] " zhenwei pi
2023-06-06 13:44 ` Stefan Hajnoczi
2023-06-07 2:03 ` [virtio-comment] " zhenwei pi
2023-05-04 8:19 ` [virtio-comment] [PATCH v2 07/11] transport-fabrics: introduce opcodes zhenwei pi
2023-05-31 17:11 ` [virtio-comment] " Stefan Hajnoczi
[not found] ` <20230531205508.GA1509630@fedora>
2023-06-02 8:39 ` [virtio-comment] " zhenwei pi
2023-06-05 16:46 ` Stefan Hajnoczi
2023-05-04 8:19 ` [virtio-comment] [PATCH v2 08/11] transport-fabrics: introduce status of completion zhenwei pi
2023-05-04 8:19 ` [virtio-comment] [PATCH v2 09/11] transport-fabrics: add TCP&RDMA binding zhenwei pi
[not found] ` <20230531210255.GC1509630@fedora>
2023-06-02 9:07 ` [virtio-comment] Re: " zhenwei pi
2023-06-05 16:57 ` Stefan Hajnoczi
2023-06-06 1:41 ` [virtio-comment] " zhenwei pi
2023-06-06 13:51 ` Stefan Hajnoczi
2023-06-07 2:15 ` zhenwei pi
2023-05-04 8:19 ` [virtio-comment] [PATCH v2 10/11] transport-fabrics: add device initialization zhenwei pi
[not found] ` <20230531210925.GD1509630@fedora>
2023-06-02 9:11 ` zhenwei pi
2023-05-04 8:19 ` [virtio-comment] [PATCH v2 11/11] transport-fabrics: support inline data for keyed transmission zhenwei pi
2023-05-29 0:56 ` [virtio-comment] PING: [PATCH v2 00/11] Introduce Virtio Over Fabrics zhenwei pi
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=PH0PR12MB5481B15C36FA7600FB9CFD73DC57A@PH0PR12MB5481.namprd12.prod.outlook.com \
--to=parav@nvidia.com \
--cc=helei.sig11@bytedance.com \
--cc=houp@yusur.tech \
--cc=jasowang@redhat.com \
--cc=mst@redhat.com \
--cc=pizhenwei@bytedance.com \
--cc=stefanha@redhat.com \
--cc=virtio-comment@lists.oasis-open.org \
--cc=xinhao.kong@duke.edu \
/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 read-only IMAP folder(s) and NNTP newsgroup(s).