[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CA+55aFxKLvHXRkx-x+zOkHJk__uTJc90Z0i+00W-hqDaQb_HCQ@mail.gmail.com>
Date: Mon, 20 Feb 2012 21:27:56 -0800
From: Linus Torvalds <torvalds@...ux-foundation.org>
To: "H. Peter Anvin" <hpa@...or.com>
Cc: Josh Boyer <jwboyer@...il.com>,
Jongman Heo <jongman.heo@...sung.com>,
Thomas Gleixner <tglx@...utronix.de>,
Ingo Molnar <mingo@...hat.com>, x86@...nel.org,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2 3/3] i387: support lazy restore of FPU state
On Mon, Feb 20, 2012 at 6:14 PM, H. Peter Anvin <hpa@...or.com> wrote:
> On 02/20/2012 06:10 PM, Linus Torvalds wrote:
>>
>> Although I do wonder if we should just make kernel_fpu_begin() be a
>> real function instead of inlining it. I'm not sure it makes sense to
>> inline that thing, and it might be better to export that one instead.
>> Comments?
>
> I would agree with that.
So I have a patch that does that, but it's noticeably bigger.
It uninlines a fair amount of i387.h, and moves it into i387.c. I do
think it's probably the right thing to do, though.
I did a "make allmodconfig" with this on x86-64, but it's quite
possible that x86-32 does additional cases. Does this patch work for
people?
(This is *on*top*of* the quick "let's just get it to work" patch that
just exports the new percpu variable. I already committed that and
pushed it out, since I wanted a quick fix so that people wouldn't be
held up by this)
IOW, if you can try this on top of current -git, that would be lovely.
Linus
View attachment "uninline-i387.diff" of type "text/x-patch" (5791 bytes)
Powered by blists - more mailing lists