From: Mathieu Desnoyers via lttng-dev <lttng-dev@lists.lttng.org>
To: "Kienan Stewart" <kstewart@efficios.com>, 方锴 <fangkaihit@163.com>,
lttng-dev@lists.lttng.org
Subject: Re: need help for avoiding system calls while using URCU (Userspace RCU)
Date: Thu, 6 Mar 2025 11:58:21 -0500 [thread overview]
Message-ID: <7d25653c-5082-4632-8bc8-a387d41760e4@efficios.com> (raw)
In-Reply-To: <ab5ea7b4-abd6-430e-bc2c-52762654301a@efficios.com>
On 2025-03-06 11:17, Kienan Stewart via lttng-dev wrote:
> Hi 方锴,
>
> On 3/4/25 3:09 AM, 方锴 via lttng-dev wrote:
>
>> I have referenced your project, URCU (Userspace RCU), on GitHub, which
>> is based on the QSBR model. However, I
> noticed that when using it in user mode, it makes system calls to enter
> the kernel. Could you please provide guidance on how to address this issue?
>
> Are there specific parts of the API that you need to be user-space only?
>
> > I aim to achieve a pure user-mode implementation. Specifically, I
> would appreciate any insights or suggestions on avoiding system calls
> while still maintaining the functionality of the QSBR model in user
> space. Thank you very much for your time and assistance.
>
> Currently the locking is done using futexs, which are system calls.
Clarification: the wait/wakeup between read-side and synchronize is done
using the futex system call. There is no locking on the read-side
per-se.
Thanks,
Mathieu
> Those could potentially be replaced with a poll/retry system either in a
> different flavour of URCU or at build-time.
>
> thanks,
> kienan
--
Mathieu Desnoyers
EfficiOS Inc.
https://www.efficios.com
next prev parent reply other threads:[~2025-03-06 16:58 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-04 8:09 方锴 via lttng-dev
2025-03-06 16:17 ` Kienan Stewart via lttng-dev
2025-03-06 16:58 ` Mathieu Desnoyers via lttng-dev [this message]
2025-03-08 9:18 ` 方锴 via lttng-dev
2025-03-08 9:16 ` 方锴 via lttng-dev
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=7d25653c-5082-4632-8bc8-a387d41760e4@efficios.com \
--to=lttng-dev@lists.lttng.org \
--cc=fangkaihit@163.com \
--cc=kstewart@efficios.com \
--cc=mathieu.desnoyers@efficios.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