All of lore.kernel.org
 help / color / mirror / Atom feed
From: Paul Mackerras <paulus@samba.org>
To: Bharat Bhushan <r65777@freescale.com>
Cc: agraf@suse.de, scottwood@freescale.com,
	stuart.yoder@freescale.com, kvm@vger.kernel.org,
	kvm-ppc@vger.kernel.org,
	Bharat Bhushan <Bharat.Bhushan@freescale.com>
Subject: Re: [PATCH 3/4] kvm: powerpc: define a linux pte lookup function
Date: Thu, 10 Oct 2013 10:33:02 +0000	[thread overview]
Message-ID: <20131010103302.GC9906@iris.ozlabs.ibm.com> (raw)
In-Reply-To: <1381212212-29641-4-git-send-email-Bharat.Bhushan@freescale.com>

On Tue, Oct 08, 2013 at 11:33:31AM +0530, Bharat Bhushan wrote:
> We need to search linux "pte" to get "pte" attributes for
> setting TLB in KVM.
> This patch defines a linux_pte_lookup() function for same.
> 

[snip]

> +	/* wait until _PAGE_BUSY is clear */
> +	while (1) {
> +		pte = pte_val(*ptep);
> +		if (unlikely(pte & _PAGE_BUSY)) {
> +			cpu_relax();
> +			continue;
> +		}
> +	}
> +
> +	/* If pte is not present return None */
> +	if (unlikely(!(pte & _PAGE_PRESENT)))
> +		return __pte(0);

First, this looks racy to me, since nothing stops the compiler
refetching pte from memory, and it might have become busy again in the
meantime.

However, I don't think you need to do any of this, since the caller in
your next patch checks for pte_present(pte) anyway.  On book E systems
we don't use _PAGE_BUSY, so you don't need the loop for your intended
application.  I suggest you remove the entire section quoted above.

Paul.

WARNING: multiple messages have this Message-ID (diff)
From: Paul Mackerras <paulus@samba.org>
To: Bharat Bhushan <r65777@freescale.com>
Cc: agraf@suse.de, scottwood@freescale.com,
	stuart.yoder@freescale.com, kvm@vger.kernel.org,
	kvm-ppc@vger.kernel.org,
	Bharat Bhushan <Bharat.Bhushan@freescale.com>
Subject: Re: [PATCH 3/4] kvm: powerpc: define a linux pte lookup function
Date: Thu, 10 Oct 2013 21:33:02 +1100	[thread overview]
Message-ID: <20131010103302.GC9906@iris.ozlabs.ibm.com> (raw)
In-Reply-To: <1381212212-29641-4-git-send-email-Bharat.Bhushan@freescale.com>

On Tue, Oct 08, 2013 at 11:33:31AM +0530, Bharat Bhushan wrote:
> We need to search linux "pte" to get "pte" attributes for
> setting TLB in KVM.
> This patch defines a linux_pte_lookup() function for same.
> 

[snip]

> +	/* wait until _PAGE_BUSY is clear */
> +	while (1) {
> +		pte = pte_val(*ptep);
> +		if (unlikely(pte & _PAGE_BUSY)) {
> +			cpu_relax();
> +			continue;
> +		}
> +	}
> +
> +	/* If pte is not present return None */
> +	if (unlikely(!(pte & _PAGE_PRESENT)))
> +		return __pte(0);

First, this looks racy to me, since nothing stops the compiler
refetching pte from memory, and it might have become busy again in the
meantime.

However, I don't think you need to do any of this, since the caller in
your next patch checks for pte_present(pte) anyway.  On book E systems
we don't use _PAGE_BUSY, so you don't need the loop for your intended
application.  I suggest you remove the entire section quoted above.

Paul.

  parent reply	other threads:[~2013-10-10 10:33 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-10-08  6:03 [PATCH 0/4] kvm: powerpc: use cache attributes from linux pte Bharat Bhushan
2013-10-08  6:15 ` Bharat Bhushan
2013-10-08  6:03 ` [PATCH 1/4] kvm: booke: clear host tlb reference flag on guest tlb invalidation Bharat Bhushan
2013-10-08  6:15   ` Bharat Bhushan
2013-10-08  6:03 ` [PATCH 2/4] kvm: book3s: rename lookup_linux_pte() to lookup_linux_pte_and_update() Bharat Bhushan
2013-10-08  6:15   ` Bharat Bhushan
2013-10-08  6:03 ` [PATCH 3/4] kvm: powerpc: define a linux pte lookup function Bharat Bhushan
2013-10-08  6:15   ` Bharat Bhushan
2013-10-08 21:36   ` Scott Wood
2013-10-08 21:36     ` Scott Wood
2013-10-09  8:48     ` Bhushan Bharat-R65777
2013-10-09  8:48       ` Bhushan Bharat-R65777
2013-10-09 17:47       ` Scott Wood
2013-10-09 17:47         ` Scott Wood
2013-10-10 10:35         ` Paul Mackerras
2013-10-10 10:35           ` Paul Mackerras
2013-10-10 10:44           ` Bhushan Bharat-R65777
2013-10-10 10:52             ` Paul Mackerras
2013-10-10 10:52               ` Paul Mackerras
2013-10-15  5:07           ` Aneesh Kumar K.V
2013-10-15  5:19             ` Aneesh Kumar K.V
2013-10-10 10:33   ` Paul Mackerras [this message]
2013-10-10 10:33     ` Paul Mackerras
2013-10-15  5:05   ` Aneesh Kumar K.V
2013-10-15  5:17     ` Aneesh Kumar K.V
2013-10-08  6:03 ` [PATCH 4/4] kvm: powerpc: use caching attributes as per linux pte Bharat Bhushan
2013-10-08  6:15   ` Bharat Bhushan
  -- strict thread matches above, loose matches on Subject: below --
2013-10-28 10:24 [PATCH 0/4] kvm: powerpc: use cache attributes from " Bharat Bhushan
2013-10-28 10:36 ` Bharat Bhushan

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=20131010103302.GC9906@iris.ozlabs.ibm.com \
    --to=paulus@samba.org \
    --cc=Bharat.Bhushan@freescale.com \
    --cc=agraf@suse.de \
    --cc=kvm-ppc@vger.kernel.org \
    --cc=kvm@vger.kernel.org \
    --cc=r65777@freescale.com \
    --cc=scottwood@freescale.com \
    --cc=stuart.yoder@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.