Re: [v1][PATCH 1/1] KVM: PPC: disable preemption when using hard_irq_disable()

All of lore.kernel.org
 help / color / mirror / Atom feed

From: "tiejun.chen" <tiejun.chen@windriver.com>
To: Scott Wood <scottwood@freescale.com>
Cc: agraf@suse.de, kvm-ppc@vger.kernel.org, kvm@vger.kernel.org,
	linuxppc-dev@lists.ozlabs.org
Subject: Re: [v1][PATCH 1/1] KVM: PPC: disable preemption when using hard_irq_disable()
Date: Thu, 11 Jul 2013 02:59:56 +0000	[thread overview]
Message-ID: <51DE1FD1.90609@windriver.com> (raw)
In-Reply-To: <1373483740.8183.223@snotra>

On 07/11/2013 03:15 AM, Scott Wood wrote:
> On 07/10/2013 01:02:19 AM, Tiejun Chen wrote:
>> We should ensure the preemption cannot occur while calling get_paca()
>> insdide hard_irq_disable(), otherwise the paca_struct may be the
>> wrong one just after. And btw, we may update timing stats in this case.
>
> The soft-ee mechanism depends on accessing the PACA directly via r13 to avoid
> this.  We probably should be using inline asm to read was_enabled rather than

Yes.

> hoping the compiler doesn't do anything silly.

Do you recommend I should directly replace get_paca() with local_paca inside 
hard_irq_disable()?

#define hard_irq_disable()      do {                    \
         u8 _was_enabled = get_paca()->soft_enabled;     \

->	u8 _was_enabled = local_paca->soft_enabled;

But is this safe for all scenarios?

Tiejun

WARNING: multiple messages have this Message-ID (diff)

From: "tiejun.chen" <tiejun.chen@windriver.com>
To: Scott Wood <scottwood@freescale.com>
Cc: linuxppc-dev@lists.ozlabs.org, agraf@suse.de,
	kvm-ppc@vger.kernel.org, kvm@vger.kernel.org
Subject: Re: [v1][PATCH 1/1] KVM: PPC: disable preemption when using hard_irq_disable()
Date: Thu, 11 Jul 2013 11:00:33 +0800	[thread overview]
Message-ID: <51DE1FD1.90609@windriver.com> (raw)
In-Reply-To: <1373483740.8183.223@snotra>

On 07/11/2013 03:15 AM, Scott Wood wrote:
> On 07/10/2013 01:02:19 AM, Tiejun Chen wrote:
>> We should ensure the preemption cannot occur while calling get_paca()
>> insdide hard_irq_disable(), otherwise the paca_struct may be the
>> wrong one just after. And btw, we may update timing stats in this case.
>
> The soft-ee mechanism depends on accessing the PACA directly via r13 to avoid
> this.  We probably should be using inline asm to read was_enabled rather than

Yes.

> hoping the compiler doesn't do anything silly.

Do you recommend I should directly replace get_paca() with local_paca inside 
hard_irq_disable()?

#define hard_irq_disable()      do {                    \
         u8 _was_enabled = get_paca()->soft_enabled;     \

->	u8 _was_enabled = local_paca->soft_enabled;

But is this safe for all scenarios?

Tiejun

WARNING: multiple messages have this Message-ID (diff)

From: "tiejun.chen" <tiejun.chen@windriver.com>
To: Scott Wood <scottwood@freescale.com>
Cc: <agraf@suse.de>, <kvm-ppc@vger.kernel.org>, <kvm@vger.kernel.org>,
	<linuxppc-dev@lists.ozlabs.org>
Subject: Re: [v1][PATCH 1/1] KVM: PPC: disable preemption when using hard_irq_disable()
Date: Thu, 11 Jul 2013 11:00:33 +0800	[thread overview]
Message-ID: <51DE1FD1.90609@windriver.com> (raw)
In-Reply-To: <1373483740.8183.223@snotra>

On 07/11/2013 03:15 AM, Scott Wood wrote:
> On 07/10/2013 01:02:19 AM, Tiejun Chen wrote:
>> We should ensure the preemption cannot occur while calling get_paca()
>> insdide hard_irq_disable(), otherwise the paca_struct may be the
>> wrong one just after. And btw, we may update timing stats in this case.
>
> The soft-ee mechanism depends on accessing the PACA directly via r13 to avoid
> this.  We probably should be using inline asm to read was_enabled rather than

Yes.

> hoping the compiler doesn't do anything silly.

Do you recommend I should directly replace get_paca() with local_paca inside 
hard_irq_disable()?

#define hard_irq_disable()      do {                    \
         u8 _was_enabled = get_paca()->soft_enabled;     \

->	u8 _was_enabled = local_paca->soft_enabled;

But is this safe for all scenarios?

Tiejun

next prev parent reply	other threads:[~2013-07-11  2:59 UTC|newest]

Thread overview: 52+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-07-10  6:02 [v1][PATCH 1/1] KVM: PPC: disable preemption when using hard_irq_disable() Tiejun Chen
2013-07-10  6:02 ` Tiejun Chen
2013-07-10  9:49 ` Alexander Graf
2013-07-10  9:49   ` Alexander Graf
2013-07-11  2:48   ` tiejun.chen
2013-07-11  2:48     ` tiejun.chen
2013-07-11  9:49     ` Alexander Graf
2013-07-11  9:49       ` Alexander Graf
2013-07-11 12:28       ` Benjamin Herrenschmidt
2013-07-11 12:28         ` Benjamin Herrenschmidt
2013-07-11 12:47         ` Alexander Graf
2013-07-11 12:47           ` Alexander Graf
2013-07-11 12:54           ` Benjamin Herrenschmidt
2013-07-11 12:54             ` Benjamin Herrenschmidt
2013-07-11 13:07             ` Alexander Graf
2013-07-11 13:07               ` Alexander Graf
2013-07-12  0:19               ` Benjamin Herrenschmidt
2013-07-12  0:19                 ` Benjamin Herrenschmidt
2013-07-12  2:13                 ` tiejun.chen
2013-07-12  2:13                   ` tiejun.chen
2013-07-12  3:57                   ` Benjamin Herrenschmidt
2013-07-12  3:57                     ` Benjamin Herrenschmidt
2013-07-12  4:54                     ` tiejun.chen
2013-07-12  4:54                       ` tiejun.chen
2013-07-14  4:13                       ` Benjamin Herrenschmidt
2013-07-14  4:13                         ` Benjamin Herrenschmidt
2013-07-15  3:04                         ` tiejun.chen
2013-07-15  3:04                           ` tiejun.chen
2013-07-10 19:15 ` Scott Wood
2013-07-10 19:15   ` Scott Wood
2013-07-10 19:15   ` Scott Wood
2013-07-11  2:59   ` tiejun.chen [this message]
2013-07-11  3:00     ` tiejun.chen
2013-07-11  3:00     ` tiejun.chen
2013-07-11 14:13     ` Scott Wood
2013-07-11 14:13       ` Scott Wood
     [not found] <1373559480.8183.258@snotra>
2013-07-12  0:30 ` Benjamin Herrenschmidt
2013-07-12  0:30   ` Benjamin Herrenschmidt
     [not found] ` <FB21594A-C233-4A97-8503-E2A1275F8F17@suse.de>
     [not found]   ` <1373560585.8183.261@snotra>
2013-07-12  3:22     ` tiejun.chen
2013-07-12  3:22       ` tiejun.chen
     [not found] <1373651433.8183.276@snotra>
2013-07-12 23:05 ` Benjamin Herrenschmidt
2013-07-12 23:05   ` Benjamin Herrenschmidt
2013-07-15  2:20   ` tiejun.chen
2013-07-15  2:20     ` tiejun.chen
2013-07-15  2:47     ` Benjamin Herrenschmidt
2013-07-15  2:47       ` Benjamin Herrenschmidt
2013-07-15  3:03       ` tiejun.chen
2013-07-15  3:03         ` tiejun.chen
     [not found]     ` <1373909248.8183.303@snotra>
2013-07-16  2:15       ` tiejun.chen
2013-07-16  2:15         ` tiejun.chen
2013-07-15  2:25 ` tiejun.chen
2013-07-15  2:25   ` tiejun.chen

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=51DE1FD1.90609@windriver.com \
    --to=tiejun.chen@windriver.com \
    --cc=agraf@suse.de \
    --cc=kvm-ppc@vger.kernel.org \
    --cc=kvm@vger.kernel.org \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=scottwood@freescale.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.