From: Peter Zijlstra <peterz@infradead.org>
To: Thiago Macieira <thiago.macieira@intel.com>
Cc: fweimer@redhat.com, "Enrico Weigelt,
metux IT consult" <lkml@metux.net>,
hjl.tools@gmail.com, libc-alpha@sourceware.org,
linux-api@vger.kernel.org, linux-arch@vger.kernel.org,
x86@kernel.org
Subject: Re: x86 CPU features detection for applications (and AMX)
Date: Mon, 28 Jun 2021 21:08:16 +0200 [thread overview]
Message-ID: <20210628190816.GC13401@worktop.programming.kicks-ass.net> (raw)
In-Reply-To: <7968759.t5oPhm3B3B@tjmaciei-mobl1>
On Mon, Jun 28, 2021 at 10:23:47AM -0700, Thiago Macieira wrote:
> On Monday, 28 June 2021 10:11:16 PDT Peter Zijlstra wrote:
> > > Consequence: CPU feature checking is done *very* early, often before
> > > main().
> > For the linker based ones, yes. IIRC the ifunc() attribute is
> > particularly useful here.
>
> Exactly. ifunc was designed for this exact purpose. And hence the fact that
> CPUID initialisation will be done very, very early.
>
> Anyway, if the AMX state is a sticky "set once per process", it's likely going
> to get set early for every process that *may* use AMX. And this is assuming we
> do the library right and only set it if has AMX code at all, instead of all
> the time.
This, AFAIU. If the ifunc() resolver finds we haz AMX it can do the
prctl() and on success pick the AMX routine.
Assuming of course, that if a program links with a library that supports
AMX, it will actually end up using it.
next prev parent reply other threads:[~2021-06-28 19:09 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-06-23 15:04 Florian Weimer
2021-06-23 15:32 ` Dave Hansen
2021-07-08 6:05 ` Florian Weimer
2021-07-08 14:19 ` Dave Hansen
2021-07-08 14:31 ` Florian Weimer
2021-07-08 14:36 ` Dave Hansen
2021-07-08 14:41 ` Florian Weimer
2021-06-25 23:31 ` Thiago Macieira
[not found] ` <3c5c29e2-1b52-3576-eda2-018fb1e58ff9@metux.net>
2021-06-28 13:20 ` Peter Zijlstra
[not found] ` <534d0171-2cc5-cd0a-904f-cd3c499b55af@metux.net>
2021-06-30 15:36 ` Thiago Macieira
2021-06-28 15:08 ` Thiago Macieira
2021-06-28 15:27 ` Peter Zijlstra
2021-06-28 16:13 ` Thiago Macieira
2021-06-28 17:11 ` Peter Zijlstra
2021-06-28 17:23 ` Thiago Macieira
2021-06-28 19:08 ` Peter Zijlstra [this message]
2021-06-28 19:26 ` Thiago Macieira
2021-06-28 17:43 ` Peter Zijlstra
2021-06-28 19:05 ` Thiago Macieira
[not found] ` <e07294c9-b02a-e1c5-3620-7fae7269fdf1@metux.net>
2021-06-30 14:34 ` Florian Weimer
[not found] ` <030f1462-2bf9-39bc-d620-6d9fbe454a27@metux.net>
2021-06-30 15:38 ` Florian Weimer
[not found] ` <4ba30cb7-6854-0691-fad6-4ca9ce674ac2@metux.net>
2021-07-01 8:21 ` Florian Weimer
[not found] ` <034dcf9b-1f8c-23ee-86a6-791122bc0f8c@metux.net>
2021-07-06 12:57 ` Florian Weimer
2021-06-30 15:29 ` Thiago Macieira
2021-07-08 7:08 ` Florian Weimer
2021-07-08 15:13 ` Thiago Macieira
2021-07-08 17:56 ` Mark Brown
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=20210628190816.GC13401@worktop.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=fweimer@redhat.com \
--cc=hjl.tools@gmail.com \
--cc=libc-alpha@sourceware.org \
--cc=linux-api@vger.kernel.org \
--cc=linux-arch@vger.kernel.org \
--cc=lkml@metux.net \
--cc=thiago.macieira@intel.com \
--cc=x86@kernel.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).