Re: [patch 3/8] x86 PAT: Add follow_pfnmp_pte routine to help tracking pfnmap pages

From: Nick Piggin <npiggin@suse.de>
To: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Cc: Ingo Molnar <mingo@elte.hu>, Thomas Gleixner <tglx@linutronix.de>,
	"H.Peter Anvin" <hpa@zytor.com>, Hugh Dickins <hugh@veritas.com>,
	Roland Dreier <rdreier@cisco.com>,
	Jesse Barnes <jbarnes@virtuousgeek.org>,
	Jeremy Fitzhardinge <jeremy@goop.org>,
	Arjan van de Ven <arjan@infradead.com>,
	linux-kernel@vger.kernel.org,
	Suresh Siddha <suresh.b.siddha@intel.com>
Subject: Re: [patch 3/8] x86 PAT: Add follow_pfnmp_pte routine to help tracking pfnmap pages
Date: Thu, 13 Nov 2008 00:27:06 +0100	[thread overview]
Message-ID: <20081112232706.GC29363@wotan.suse.de> (raw)
In-Reply-To: <20081112212900.291460000@intel.com>

On Wed, Nov 12, 2008 at 01:26:50PM -0800, Venkatesh Pallipadi wrote:
> 
> Add a generic interface to follow pfn in a pfnmap vma range. This is used by
> one of the subsequent x86 PAT related patch to keep track of memory types
> for vma regions across vma copy and free.
> 
> Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
> Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
> 
> ---
>  include/linux/mm.h |    3 +++
>  mm/memory.c        |   43 +++++++++++++++++++++++++++++++++++++++++++
>  2 files changed, 46 insertions(+)
> 
> Index: tip/include/linux/mm.h
> ===================================================================
> --- tip.orig/include/linux/mm.h	2008-11-06 15:41:43.000000000 -0800
> +++ tip/include/linux/mm.h	2008-11-10 09:44:45.000000000 -0800
> @@ -1223,6 +1223,9 @@ struct page *follow_page(struct vm_area_
>  #define FOLL_GET	0x04	/* do get_page on page */
>  #define FOLL_ANON	0x08	/* give ZERO_PAGE if no pgtable */
>  
> +unsigned long follow_pfnmap_pte(struct vm_area_struct *vma,
> +				unsigned long address, pte_t *ret_ptep);
> +
>  typedef int (*pte_fn_t)(pte_t *pte, pgtable_t token, unsigned long addr,
>  			void *data);
>  extern int apply_to_page_range(struct mm_struct *mm, unsigned long address,
> Index: tip/mm/memory.c
> ===================================================================
> --- tip.orig/mm/memory.c	2008-11-06 15:41:43.000000000 -0800
> +++ tip/mm/memory.c	2008-11-10 09:44:45.000000000 -0800
> @@ -1111,6 +1111,49 @@ no_page_table:
>  	return page;
>  }
>  
> +unsigned long follow_pfnmap_pte(struct vm_area_struct *vma,
> +			unsigned long address, pte_t *ret_ptep)
> +{

I think we'd typically return an int for functions like this. It
probably wouldn't hurt to return more descriptive errors either.
-EINVAL, perhaps, seems to fit best with what existing code there
is doing.

> +	pgd_t *pgd;
> +	pud_t *pud;
> +	pmd_t *pmd;
> +	pte_t *ptep, pte;
> +	spinlock_t *ptl;
> +	struct page *page;
> +	struct mm_struct *mm = vma->vm_mm;
> +
> +	if (!is_pfn_mapping(vma))
> +		goto err;
> +
> +	page = NULL;
> +	pgd = pgd_offset(mm, address);
> +	if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd)))
> +		goto err;
> +
> +	pud = pud_offset(pgd, address);
> +	if (pud_none(*pud) || unlikely(pud_bad(*pud)))
> +		goto err;
> +
> +	pmd = pmd_offset(pud, address);
> +	if (pmd_none(*pmd) || unlikely(pmd_bad(*pmd)))
> +		goto err;
> +
> +	ptep = pte_offset_map_lock(mm, pmd, address, &ptl);
> +
> +	pte = *ptep;
> +	if (!pte_present(pte))
> +		goto err_unlock;
> +
> +	*ret_ptep = pte;
> +	pte_unmap_unlock(ptep, ptl);
> +	return 0;
> +
> +err_unlock:
> +	pte_unmap_unlock(ptep, ptl);
> +err:
> +	return -1;
> +}
> +
>  /* Can we do the FOLL_ANON optimization? */
>  static inline int use_zero_page(struct vm_area_struct *vma)
>  {
> 
> --