From: Florian Weimer <fweimer@redhat.com>
To: Noah Goldstein <goldstein.w.n@gmail.com>
Cc: Noah Goldstein via Libc-alpha <libc-alpha@sourceware.org>,
hjl.tools@gmail.com, carlos@systemhalted.org
Subject: Re: [PATCH v1] x86: Use 64MB as nt-store threshold if no cacheinfo [BZ #30429]
Date: Thu, 25 May 2023 10:11:33 +0200 [thread overview]
Message-ID: <87lehcomoa.fsf@oldenburg.str.redhat.com> (raw)
In-Reply-To: <CAFUsyf+Fykb89nT6C3UM796YA_O0v-82=FzLtiSUkJ_pJc-JzA@mail.gmail.com> (Noah Goldstein's message of "Tue, 9 May 2023 11:01:42 -0500")
* Noah Goldstein:
> On Tue, May 9, 2023 at 3:56 AM Florian Weimer <fweimer@redhat.com> wrote:
>>
>> * Noah Goldstein via Libc-alpha:
>>
>> > If `non_temporal_threshold` is below `minimum_non_temporal_threshold`,
>> > it almost certainly means we failed to read the systems cache info.
>> >
>> > In this case, rather than defaulting the minimum correct value, we
>> > should default to a value that gets at least reasonable
>> > performance. 64MB is chosen conservatively to be at the very high
>> > end. This should never cause non-temporal stores when, if we had read
>> > cache info, we wouldn't have otherwise.
>>
>> I think that's quite surprising for GLIBC_TUNABLES. Maybe that logic
>> should only activate if the default was set from cache sizes?
>>
> I don't quite understand what you mean by "only active if the default
> was set from cache sizes"?
>
> This logic only triggers if total_cachesize / 8 < ~16kb. I think this should
> only ever really happen if we failed to read cache info. What is the
> surprise?
Right, my mistake. The patch should work as-is.
Reviewed-by: Florian Weimer <fweimer@redhat.com>
Thanks,
Florian
next prev parent reply other threads:[~2023-05-25 8:11 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-09 3:10 Noah Goldstein
2023-05-09 8:56 ` Florian Weimer
2023-05-09 16:01 ` Noah Goldstein
2023-05-11 18:15 ` Noah Goldstein
2023-05-25 8:11 ` Florian Weimer [this message]
2023-05-25 18:36 ` H.J. Lu
2023-06-05 17:14 ` Noah Goldstein
2023-06-05 17:45 ` H.J. Lu
2023-06-05 20:28 ` Noah Goldstein
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=87lehcomoa.fsf@oldenburg.str.redhat.com \
--to=fweimer@redhat.com \
--cc=carlos@systemhalted.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).