Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1354091 > unrolled thread

Re: [PATCH] x86/fpu: Revert earlier patch of Disable AVX when eagerfpu is off

Started byIngo Molnar <mingo@kernel.org>
First post2016-03-09 13:50 +0100
Last post2016-03-09 20:00 +0100
Articles 4 — 3 participants

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  Re: [PATCH] x86/fpu: Revert earlier patch of Disable AVX when  eagerfpu is off Ingo Molnar <mingo@kernel.org> - 2016-03-09 13:50 +0100
    Re: [PATCH] x86/fpu: Revert earlier patch of Disable AVX when  eagerfpu is off Yu-cheng Yu <yu-cheng.yu@intel.com> - 2016-03-09 18:30 +0100
      Re: [PATCH] x86/fpu: Revert earlier patch of Disable AVX when  eagerfpu is off Ingo Molnar <mingo@kernel.org> - 2016-03-09 19:50 +0100
    Re: [PATCH] x86/fpu: Revert earlier patch of Disable AVX when  eagerfpu is off "H. Peter Anvin" <hpa@zytor.com> - 2016-03-09 20:00 +0100

#1354091 — Re: [PATCH] x86/fpu: Revert earlier patch of Disable AVX when eagerfpu is off

FromIngo Molnar <mingo@kernel.org>
Date2016-03-09 13:50 +0100
SubjectRe: [PATCH] x86/fpu: Revert earlier patch of Disable AVX when eagerfpu is off
Message-ID<raLIS-7XZ-15@gated-at.bofh.it>
* Yu-cheng Yu <yu-cheng.yu@intel.com> wrote:

> AVX was mistakenly believed to be dependent on eagerfpu switch.
> This turns out to be false. The earlier patch should be reverted.
> 
> Original patch:
>     http://git.kernel.org/tip/394db20ca240741a08d472173db13d6f6a6e5a28

So the original patch had a whole host of explanations of why that change is 
correct. This revert should explain where that argumentation was wrong.

Also note that eagerfpu=off is on the way out, on v4.6 we'll switch all CPUs to 
eagerfpu:

  58122bf1d856 x86/fpu: Default eagerfpu=on on all CPUs

and in the not so distant future, if everything goes fine with the eager mode, I'd 
like to drop the lazy FPU context switching code altogether - which will simplify 
a lot of code. At that point the 'eagerfpu' boot option will go away as well.

Thanks,

	Ingo

[toc] | [next] | [standalone]


#1354301

FromYu-cheng Yu <yu-cheng.yu@intel.com>
Date2016-03-09 18:30 +0100
Message-ID<raQ5Q-2LL-17@gated-at.bofh.it>
In reply to#1354091
On Wed, Mar 09, 2016 at 01:46:26PM +0100, Ingo Molnar wrote:
> 
> * Yu-cheng Yu <yu-cheng.yu@intel.com> wrote:
> 
> > AVX was mistakenly believed to be dependent on eagerfpu switch.
> > This turns out to be false. The earlier patch should be reverted.
> > 
> > Original patch:
> >     http://git.kernel.org/tip/394db20ca240741a08d472173db13d6f6a6e5a28
> 
> So the original patch had a whole host of explanations of why that change is 
> correct. This revert should explain where that argumentation was wrong.

I will send out another version that includes more details.

Thanks,
Yu-cheng

[toc] | [prev] | [next] | [standalone]


#1354345

FromIngo Molnar <mingo@kernel.org>
Date2016-03-09 19:50 +0100
Message-ID<raRlg-3wv-13@gated-at.bofh.it>
In reply to#1354301
* Yu-cheng Yu <yu-cheng.yu@intel.com> wrote:

> On Wed, Mar 09, 2016 at 01:46:26PM +0100, Ingo Molnar wrote:
> > 
> > * Yu-cheng Yu <yu-cheng.yu@intel.com> wrote:
> > 
> > > AVX was mistakenly believed to be dependent on eagerfpu switch.
> > > This turns out to be false. The earlier patch should be reverted.
> > > 
> > > Original patch:
> > >     http://git.kernel.org/tip/394db20ca240741a08d472173db13d6f6a6e5a28
> > 
> > So the original patch had a whole host of explanations of why that change is 
> > correct. This revert should explain where that argumentation was wrong.
> 
> I will send out another version that includes more details.

Great, thanks!

	Ingo

[toc] | [prev] | [next] | [standalone]


#1354354

From"H. Peter Anvin" <hpa@zytor.com>
Date2016-03-09 20:00 +0100
Message-ID<raRuW-3zK-15@gated-at.bofh.it>
In reply to#1354091
On 03/09/2016 04:46 AM, Ingo Molnar wrote:
> 
> * Yu-cheng Yu <yu-cheng.yu@intel.com> wrote:
> 
>> AVX was mistakenly believed to be dependent on eagerfpu switch.
>> This turns out to be false. The earlier patch should be reverted.
>>
>> Original patch:
>>     http://git.kernel.org/tip/394db20ca240741a08d472173db13d6f6a6e5a28
> 
> So the original patch had a whole host of explanations of why that change is 
> correct. This revert should explain where that argumentation was wrong.
> 
> Also note that eagerfpu=off is on the way out, on v4.6 we'll switch all CPUs to 
> eagerfpu:
> 
>   58122bf1d856 x86/fpu: Default eagerfpu=on on all CPUs
> 
> and in the not so distant future, if everything goes fine with the eager mode, I'd 
> like to drop the lazy FPU context switching code altogether - which will simplify 
> a lot of code. At that point the 'eagerfpu' boot option will go away as well.
> 

Last I heard the only use case which was seeing trouble with eagerfpu
was KVM during service of I/O events.  Might be worth checking into...

	-hpa

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web