From: Wilco Dijkstra <Wilco.Dijkstra@arm.com>
To: Noah Goldstein <goldstein.w.n@gmail.com>,
"H.J. Lu" <hjl.tools@gmail.com>
Cc: Adhemerval Zanella Netto <adhemerval.zanella@linaro.org>,
GNU C Library <libc-alpha@sourceware.org>
Subject: Re: [PATCH 3/7] Remove atomic_bit_set/bit_test_set
Date: Tue, 12 Jul 2022 20:39:16 +0000 [thread overview]
Message-ID: <AM5PR0801MB1668822744D550957A803AA483869@AM5PR0801MB1668.eurprd08.prod.outlook.com> (raw)
In-Reply-To: <CAFUsyfJfYGGczz_G9RWnoo9rPpnkfiQBAh5bYJy_pfDotdPMLw@mail.gmail.com>
Hi,
>> > More importantly, it raises a question for locking: we currently use a compare-exchange
>> > for each lock and unlock. If compare-exchange has higher overheads than other atomics
>> > then we should investigate more efficient locking sequences.
>>
>> compare-exchange can be very expensive under contention.
>>
>
> Which locks should we look into dropping CAS? pthread_mutex? Anything else?
There are several internal locking schemes plus uses of CAS spread all over.
There is also the odd define ATOMIC_EXCHANGE_USES_CAS which almost no target
seems to know what to set to - it is used in pthread_spin_lock in a way that makes no
sense. I believe the answer should be 0 for all since exchange is typically simpler and
faster than compare-exchange.
Cheers,
Wilco
prev parent reply other threads:[~2022-07-12 20:39 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-07-06 15:14 Wilco Dijkstra
2022-07-06 16:14 ` Noah Goldstein
2022-07-06 18:25 ` Noah Goldstein
2022-07-06 18:59 ` Wilco Dijkstra
2022-07-06 19:14 ` Noah Goldstein
2022-07-06 19:30 ` H.J. Lu
2022-07-06 19:36 ` Wilco Dijkstra
2022-07-06 19:51 ` Noah Goldstein
2022-07-06 19:56 ` Noah Goldstein
2022-07-06 20:14 ` Wilco Dijkstra
2022-07-06 20:30 ` Noah Goldstein
2022-07-06 20:56 ` H.J. Lu
2022-07-12 17:47 ` Adhemerval Zanella Netto
2022-07-12 18:09 ` H.J. Lu
2022-07-12 18:44 ` Wilco Dijkstra
2022-07-12 19:11 ` H.J. Lu
2022-07-12 19:18 ` Noah Goldstein
2022-07-12 20:39 ` Wilco Dijkstra [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 \
--in-reply-to=AM5PR0801MB1668822744D550957A803AA483869@AM5PR0801MB1668.eurprd08.prod.outlook.com \
--to=wilco.dijkstra@arm.com \
--cc=adhemerval.zanella@linaro.org \
--cc=goldstein.w.n@gmail.com \
--cc=hjl.tools@gmail.com \
--cc=libc-alpha@sourceware.org \
/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).