From: "H. Peter Anvin" <hpa@zytor.com>
To: Luca Barbieri <luca@luca-barbieri.com>
Cc: Andi Kleen <andi@firstfloor.org>,
mingo@elte.hu, a.p.zijlstra@chello.nl, akpm@linux-foundation.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH 09/10] x86-32: use SSE for atomic64_read/set if available
Date: Thu, 18 Feb 2010 11:07:12 -0800 [thread overview]
Message-ID: <4B7D8FE0.1040602@zytor.com> (raw)
In-Reply-To: <ff13bc9a1002181042p7414c18fu77208b6bc7aa41df@mail.gmail.com>
On 02/18/2010 10:42 AM, Luca Barbieri wrote:
>> We already do that kind of stuff, using
>> kernel_fpu_begin()..kernel_fpu_end(). We went through some pain a bit
>> ago to clean up "private hacks" that complicated things substantially.
>
> But that saves the whole FPU state on the first usage, and also
> triggers a fault when userspace attempts to use it again.
> Additionally it does a clts/stts every time which is slow for small
> algorithms (lke the atomic64 routines).
>
> The first issue can be solved by using SSE and saving only the used
> registers, and the second with lazy TS flag restoring.
>
Again, I want to see a strong use case before even *considering* making
the rules we already have any more complex.
-hpa
next prev parent reply other threads:[~2010-02-18 19:08 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-02-17 11:42 [PATCH 0/10] x86-32: improve atomic64_t functions Luca Barbieri
2010-02-17 11:42 ` [PATCH 01/10] x86: add support for multiple choice alternatives Luca Barbieri
2010-02-17 12:47 ` Avi Kivity
2010-02-18 19:46 ` H. Peter Anvin
2010-02-17 11:42 ` [PATCH 02/10] x86: add support for relative CALL and JMP in alternatives Luca Barbieri
2010-02-18 19:40 ` H. Peter Anvin
2010-02-18 23:38 ` Luca Barbieri
2010-02-18 23:54 ` H. Peter Anvin
2010-02-19 14:27 ` Masami Hiramatsu
2010-02-17 11:42 ` [PATCH 03/10] x86: add support for lock prefix " Luca Barbieri
2010-02-17 11:42 ` [PATCH 04/10] x86-32: allow UP/SMP lock replacement in cmpxchg64 Luca Barbieri
2010-02-17 11:42 ` [PATCH 05/10] lib: add self-test for atomic64_t Luca Barbieri
2010-02-17 11:42 ` [PATCH 06/10] x86-32: rewrite 32-bit atomic64 functions in assembly Luca Barbieri
2010-02-17 11:42 ` [PATCH 07/10] lib: move generic atomic64 to atomic64-impl.h Luca Barbieri
2010-02-17 11:42 ` [PATCH 08/10] x86-32: support atomic64_t on 386/486 UP/SMP Luca Barbieri
2010-02-18 10:25 ` Peter Zijlstra
2010-02-18 10:58 ` Luca Barbieri
2010-02-18 15:20 ` H. Peter Anvin
2010-02-17 11:42 ` [PATCH 09/10] x86-32: use SSE for atomic64_read/set if available Luca Barbieri
2010-02-17 22:39 ` H. Peter Anvin
2010-02-18 0:41 ` Luca Barbieri
2010-02-18 0:47 ` H. Peter Anvin
2010-02-18 9:56 ` Avi Kivity
2010-02-18 10:07 ` Luca Barbieri
2010-02-18 8:23 ` Andi Kleen
2010-02-18 9:53 ` Luca Barbieri
2010-02-18 9:56 ` Luca Barbieri
2010-02-18 10:11 ` Andi Kleen
2010-02-18 10:27 ` Luca Barbieri
2010-02-18 15:24 ` H. Peter Anvin
2010-02-18 18:14 ` Luca Barbieri
2010-02-18 18:28 ` H. Peter Anvin
2010-02-18 18:42 ` Luca Barbieri
2010-02-18 19:07 ` H. Peter Anvin [this message]
2010-02-18 20:26 ` Andi Kleen
2010-02-18 16:52 ` H. Peter Anvin
2010-02-18 18:49 ` Luca Barbieri
2010-02-18 19:06 ` H. Peter Anvin
2010-02-18 19:43 ` Luca Barbieri
2010-02-18 19:45 ` Yuhong Bao
2010-02-18 10:24 ` Peter Zijlstra
2010-02-18 10:25 ` Peter Zijlstra
2010-02-18 10:50 ` Luca Barbieri
2010-02-18 11:00 ` Peter Zijlstra
2010-02-18 12:29 ` Luca Barbieri
2010-02-18 12:32 ` Peter Zijlstra
2010-02-18 13:45 ` Luca Barbieri
2010-02-17 11:42 ` [PATCH 10/10] x86-32: panic on !CX8 && XMM Luca Barbieri
2010-02-17 22:38 ` H. Peter Anvin
2010-02-17 23:00 ` Yuhong Bao
2010-02-17 23:41 ` H. Peter Anvin
2010-02-18 1:13 ` Yuhong Bao
2010-02-25 20:24 ` Yuhong Bao
2010-02-18 0:46 ` Luca Barbieri
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=4B7D8FE0.1040602@zytor.com \
--to=hpa@zytor.com \
--cc=a.p.zijlstra@chello.nl \
--cc=akpm@linux-foundation.org \
--cc=andi@firstfloor.org \
--cc=linux-kernel@vger.kernel.org \
--cc=luca@luca-barbieri.com \
--cc=mingo@elte.hu \
/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
Powered by JetHome