Re: [PATCH 10/18] mm, hugetlb: call vma_has_reserve() before enteringalloc_huge_page()

[Naoya Horiguchi]

On Mon, Jul 29, 2013 at 02:32:01PM +0900, Joonsoo Kim wrote:
> To implement a graceful failure handling, we need to know whether
> allocation request is for reserved pool or not, on higher level.
> In this patch, we just move up vma_has_reseve() to caller function
> in order to know it. There is no functional change.
> 
> Following patches implement a grace failure handling and remove
> a hugetlb_instantiation_mutex.
> 
> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
> 
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> index a66226e..5f31ca5 100644
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -1123,12 +1123,12 @@ static void vma_commit_reservation(struct hstate *h,
>  }
>  
>  static struct page *alloc_huge_page(struct vm_area_struct *vma,
> -				    unsigned long addr, int avoid_reserve)
> +				    unsigned long addr, int use_reserve)
>  {
>  	struct hugepage_subpool *spool = subpool_vma(vma);
>  	struct hstate *h = hstate_vma(vma);
>  	struct page *page;
> -	int ret, idx, use_reserve;
> +	int ret, idx;
>  	struct hugetlb_cgroup *h_cg;
>  
>  	idx = hstate_index(h);
> @@ -1140,11 +1140,6 @@ static struct page *alloc_huge_page(struct vm_area_struct *vma,
>  	 * need pages and subpool limit allocated allocated if no reserve
>  	 * mapping overlaps.
>  	 */
> -	use_reserve = vma_has_reserves(h, vma, addr);
> -	if (use_reserve < 0)
> -		return ERR_PTR(-ENOMEM);
> -
> -	use_reserve = use_reserve && !avoid_reserve;
>  	if (!use_reserve && (hugepage_subpool_get_pages(spool, 1) < 0))
>  			return ERR_PTR(-ENOSPC);
>  
> @@ -2520,7 +2515,7 @@ static int hugetlb_cow(struct mm_struct *mm, struct vm_area_struct *vma,
>  {
>  	struct hstate *h = hstate_vma(vma);
>  	struct page *old_page, *new_page;
> -	int outside_reserve = 0;
> +	int use_reserve, outside_reserve = 0;
>  	unsigned long mmun_start;	/* For mmu_notifiers */
>  	unsigned long mmun_end;		/* For mmu_notifiers */
>  
> @@ -2553,7 +2548,18 @@ retry_avoidcopy:
>  
>  	/* Drop page_table_lock as buddy allocator may be called */
>  	spin_unlock(&mm->page_table_lock);
> -	new_page = alloc_huge_page(vma, address, outside_reserve);
> +
> +	use_reserve = vma_has_reserves(h, vma, address);
> +	if (use_reserve == -ENOMEM) {
> +		page_cache_release(old_page);
> +
> +		/* Caller expects lock to be held */
> +		spin_lock(&mm->page_table_lock);
> +		return VM_FAULT_OOM;
> +	}
> +	use_reserve = use_reserve && !outside_reserve;

When outside_reserve is true, we don't have to call vma_has_reserves
because then use_reserve is always false. So something like:

  use_reserve = 0;
  if (!outside_reserve) {
          use_reserve = vma_has_reserves(...);
          ...
  }

looks better to me.
Or if you expect vma_has_reserves to change resv_map implicitly,
could you add a comment about it.

Thanks,
Naoya Horiguchi

> +
> +	new_page = alloc_huge_page(vma, address, use_reserve);
>  
>  	if (IS_ERR(new_page)) {
>  		long err = PTR_ERR(new_page);
> @@ -2679,6 +2685,7 @@ static int hugetlb_no_page(struct mm_struct *mm, struct vm_area_struct *vma,
>  	struct page *page;
>  	struct address_space *mapping;
>  	pte_t new_pte;
> +	int use_reserve = 0;
>  
>  	/*
>  	 * Currently, we are forced to kill the process in the event the
> @@ -2704,7 +2711,14 @@ retry:
>  		size = i_size_read(mapping->host) >> huge_page_shift(h);
>  		if (idx >= size)
>  			goto out;
> -		page = alloc_huge_page(vma, address, 0);
> +
> +		use_reserve = vma_has_reserves(h, vma, address);
> +		if (use_reserve == -ENOMEM) {
> +			ret = VM_FAULT_OOM;
> +			goto out;
> +		}
> +
> +		page = alloc_huge_page(vma, address, use_reserve);
>  		if (IS_ERR(page)) {
>  			ret = PTR_ERR(page);
>  			if (ret == -ENOMEM)
> -- 
> 1.7.9.5
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/