public inbox for [email protected]
 help / color / mirror / Atom feed
From: Pavel Begunkov <[email protected]>
To: Stefan Metzmacher <[email protected]>,
	Caleb Sander Mateos <[email protected]>
Cc: [email protected], Andres Freund <[email protected]>
Subject: Re: [PATCH 0/8] Add support for vectored registered buffers
Date: Tue, 4 Mar 2025 14:20:38 +0000	[thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>

On 3/4/25 12:12, Stefan Metzmacher wrote:
> Hi Pavel,
> 
>>> Add registered buffer support for vectored io_uring operations. That
>>> allows to pass an iovec, all entries of which must belong to and
>>> point into the same registered buffer specified by sqe->buf_index.
>>>
>>> The series covers zerocopy sendmsg and reads / writes. Reads and
>>> writes are implemented as new opcodes, while zerocopy sendmsg
>>> reuses IORING_RECVSEND_FIXED_BUF for the api.
>>>
>>> Results are aligned to what one would expect from registered buffers:
>>>
>>> t/io_uring + nullblk, single segment 16K:
>>>    34 -> 46 GiB/s
>>> examples/send-zerocopy.c default send size (64KB):
>>>    82558 -> 123855 MB/s
>>
>> Thanks for implementing this, it's great to be able to combine these 2
>> optimizations! Though I suspect many applications will want to perform
>> vectorized I/O using iovecs that come from different registered
>> buffers (e.g. separate header and data allocations). Perhaps a future
>> improvement could allow a list of buffer indices to be specified.
> 
> I'm wondering about the same. And it's not completely
> clear to me what the value of iov_base is in this case,
> is it the offset into the buffer, or the real pointer address
> that must within the range of the registered buffer?

Same as with other registered buffer requests. It's a pointer into
the initial buffer you specified when registering it, which serves
to calculate the offset.

See the io_uring_register(2) man, addr and len are iov_base and
iov_len and there are multiple of them. You can call it confusing,
and I'd agree, but that's how it was done from the very beginning,
so staying consistent here.

https://github.com/axboe/liburing/blob/master/man/io_uring_register.2#L87

  
> It might also be very useful to have some vector elements pointing
> into one of the registered buffer, while others refer to non-registered
> buffers.

See the other reply.

-- 
Pavel Begunkov


      reply	other threads:[~2025-03-04 14:19 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-03-03 15:50 [PATCH 0/8] Add support for vectored registered buffers Pavel Begunkov
2025-03-03 15:50 ` [PATCH 1/8] io_uring: introduce struct iou_vec Pavel Begunkov
2025-03-03 18:23   ` Caleb Sander Mateos
2025-03-03 15:50 ` [PATCH 2/8] io_uring: add infra for importing vectored reg buffers Pavel Begunkov
2025-03-03 20:49   ` Caleb Sander Mateos
2025-03-03 20:57     ` Keith Busch
2025-03-04 10:05     ` Pavel Begunkov
2025-03-04 15:18       ` Pavel Begunkov
2025-03-03 15:50 ` [PATCH 3/8] io_uring/rw: implement vectored registered rw Pavel Begunkov
2025-03-03 23:01   ` Caleb Sander Mateos
2025-03-03 23:37     ` Caleb Sander Mateos
2025-03-04 10:09     ` Pavel Begunkov
2025-03-03 15:50 ` [PATCH 4/8] io_uring/rw: defer reg buf vec import Pavel Begunkov
2025-03-03 23:37   ` Caleb Sander Mateos
2025-03-03 15:51 ` [PATCH 5/8] io_uring/net: combine msghdr copy Pavel Begunkov
2025-03-03 15:51 ` [PATCH 6/8] io_uring/net: pull vec alloc out of msghdr import Pavel Begunkov
2025-03-03 15:51 ` [PATCH 7/8] io_uring/net: convert to struct iou_vec Pavel Begunkov
2025-03-03 23:37   ` Caleb Sander Mateos
2025-03-03 15:51 ` [PATCH 8/8] io_uring/net: implement vectored reg bufs for zctx Pavel Begunkov
2025-03-03 21:03 ` [PATCH 0/8] Add support for vectored registered buffers Andres Freund
2025-03-04 10:21   ` Pavel Begunkov
2025-03-04  0:34 ` Caleb Sander Mateos
2025-03-04 10:26   ` Pavel Begunkov
2025-03-04 12:12   ` Stefan Metzmacher
2025-03-04 14:20     ` Pavel Begunkov [this message]

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 \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    /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