From: Suresh Siddha <suresh.b.siddha@intel.com>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@redhat.com>, "H. Peter Anvin" <hpa@zytor.com>,
x86@kernel.org,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v2 3/3] i387: support lazy restore of FPU state
Date: Tue, 21 Feb 2012 13:54:32 -0800 [thread overview]
Message-ID: <1329861273.29790.261.camel@sbsiddha-desk.sc.intel.com> (raw)
In-Reply-To: <alpine.LFD.2.02.1202201148100.4237@i5.linux-foundation.org>
On Mon, 2012-02-20 at 11:48 -0800, Linus Torvalds wrote:
> From: Linus Torvalds <torvalds@linux-foundation.org>
> Date: Sun, 19 Feb 2012 13:27:00 -0800
> Subject: [PATCH v2 3/3] i387: support lazy restore of FPU state
>
> This makes us recognize when we try to restore FPU state that matches
> what we already have in the FPU on this CPU, and avoids the restore
> entirely if so.
>
> To do this, we add two new data fields:
>
> - a percpu 'fpu_owner_task' variable that gets written any time we
> update the "has_fpu" field, and thus acts as a kind of back-pointer
> to the task that owns the CPU. The exception is when we save the FPU
> state as part of a context switch - if the save can keep the FPU
> state around, we leave the 'fpu_owner_task' variable pointing at the
> task whose FP state still remains on the CPU.
>
> - a per-thread 'last_cpu' field, that indicates which CPU that thread
> used its FPU on last. We update this on every context switch
> (writing an invalid CPU number if the last context switch didn't
> leave the FPU in a lazily usable state), so we know that *that*
> thread has done nothing else with the FPU since.
>
> These two fields together can be used when next switching back to the
> task to see if the CPU still matches: if 'fpu_owner_task' matches the
> task we are switching to, we know that no other task (or kernel FPU
> usage) touched the FPU on this CPU in the meantime, and if the current
> CPU number matches the 'last_cpu' field, we know that this thread did no
> other FP work on any other CPU, so the FPU state on the CPU must match
> what was saved on last context switch.
>
> In that case, we can avoid the 'f[x]rstor' entirely, and just clear the
> CR0.TS bit.
>
Reviewing this code, I think we need to set the 'last_cpu' to an invalid
number in the fpu_alloc too. Appended is the patch.
---
From: Suresh Siddha <suresh.b.siddha@intel.com>
Subject: x86, fpu: set the last_cpu in fpu_alloc() to an invalid cpu
Initialize the struct fpu's last_cpu in fpu_alloc() to an invalid cpu number,
so that the check in fpu_lazy_restore() will always return false
on the first context-switch in of this new task.
Otherwise, on a fork(), last_cpu of the new task's fpu will be copied from
the parent task and fpu_lazy_restore() can potentially return success wrongly
on the first context-switch in of this new task, leading to fpu corruption.
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
---
arch/x86/include/asm/i387.h | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/arch/x86/include/asm/i387.h b/arch/x86/include/asm/i387.h
index 2479049..58ba656 100644
--- a/arch/x86/include/asm/i387.h
+++ b/arch/x86/include/asm/i387.h
@@ -627,6 +627,7 @@ static inline int fpu_alloc(struct fpu *fpu)
if (!fpu->state)
return -ENOMEM;
WARN_ON((unsigned long)fpu->state & 15);
+ fpu->last_cpu = ~0;
return 0;
}
next prev parent reply other threads:[~2012-02-21 21:54 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-02-19 22:23 [PATCH 0/2] More i387 state save/restore work Linus Torvalds
2012-02-19 22:26 ` [PATCH 1/2] i387: use 'restore_fpu_checking()' directly in task switching code Linus Torvalds
2012-02-19 22:37 ` [PATCH 2/2] i387: support lazy restore of FPU state Linus Torvalds
2012-02-19 22:44 ` H. Peter Anvin
2012-02-19 23:18 ` H. Peter Anvin
2012-02-19 23:56 ` Linus Torvalds
2012-02-20 7:51 ` Ingo Molnar
2012-02-20 0:53 ` [PATCH 0/2] More i387 state save/restore work Michael Neuling
2012-02-20 1:03 ` Linus Torvalds
2012-02-20 1:06 ` Linus Torvalds
2012-02-20 1:11 ` Linus Torvalds
2012-03-01 11:30 ` Benjamin Herrenschmidt
2012-02-20 2:09 ` Indan Zupancic
2012-02-20 19:46 ` [PATCH v2 0/3] " Linus Torvalds
2012-02-20 19:47 ` [PATCH v2 1/3] i387: fix up some fpu_counter confusion Linus Torvalds
2012-02-20 19:48 ` [PATCH v2 2/3] i387: use 'restore_fpu_checking()' directly in task switching code Linus Torvalds
2012-02-20 19:48 ` [PATCH v2 3/3] i387: support lazy restore of FPU state Linus Torvalds
2012-02-21 1:50 ` Josh Boyer
2012-02-21 2:10 ` Linus Torvalds
2012-02-21 2:14 ` H. Peter Anvin
2012-02-21 5:27 ` Linus Torvalds
2012-02-21 5:35 ` H. Peter Anvin
2012-02-21 14:19 ` Josh Boyer
2012-02-21 17:59 ` H. Peter Anvin
2012-02-21 18:06 ` Ingo Molnar
2012-02-21 18:26 ` Linus Torvalds
2012-02-21 21:14 ` H. Peter Anvin
2012-02-21 21:39 ` [PATCH 0/2] i387: FP state interface cleanups Linus Torvalds
2012-02-21 21:40 ` [PATCH 1/2] i387: uninline the generic FP helpers that we expose to kernel modules Linus Torvalds
2012-02-21 21:41 ` [PATCH 2/2] i387: split up <asm/i387.h> into exported and internal interfaces Linus Torvalds
2012-02-21 23:50 ` [tip:x86/fpu] i387: Split " tip-bot for Linus Torvalds
2012-02-28 11:21 ` [PATCH 2/2] i387: split " Avi Kivity
2012-02-28 16:05 ` Linus Torvalds
2012-02-28 17:21 ` Avi Kivity
2012-02-28 17:37 ` Linus Torvalds
2012-02-28 18:08 ` Linus Torvalds
2012-02-28 18:29 ` Avi Kivity
2012-02-28 18:09 ` Avi Kivity
2012-02-28 18:34 ` Linus Torvalds
2012-02-28 19:06 ` Avi Kivity
2012-02-28 19:26 ` Linus Torvalds
2012-02-28 19:45 ` Avi Kivity
2012-02-21 23:49 ` [tip:x86/fpu] i387: Uninline the generic FP helpers that we expose to kernel modules tip-bot for Linus Torvalds
2012-02-21 2:18 ` [PATCH v2 3/3] i387: support lazy restore of FPU state Linus Torvalds
2012-02-21 2:32 ` H. Peter Anvin
2012-02-21 2:11 ` H. Peter Anvin
2012-02-21 21:54 ` Suresh Siddha [this message]
2012-02-21 21:57 ` Linus Torvalds
2012-02-21 22:19 ` Suresh Siddha
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=1329861273.29790.261.camel@sbsiddha-desk.sc.intel.com \
--to=suresh.b.siddha@intel.com \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--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
all inboxes | Powered by JetHome®