[PATCH] staging: zsmalloc: fix memory leak

Dan Magenheimer dan.magenheimer at oracle.com
Wed Apr 4 16:03:22 UTC 2012


> From: Seth Jennings [mailto:sjenning at linux.vnet.ibm.com]
> Sent: Monday, April 02, 2012 8:14 AM
> To: Greg Kroah-Hartman
> Cc: Nitin Gupta; Dan Magenheimer; Konrad Rzeszutek Wilk; Robert Jennings; Seth Jennings;
> devel at driverdev.osuosl.org; linux-kernel at vger.kernel.org; linux-mm at kvack.org
> Subject: [PATCH] staging: zsmalloc: fix memory leak
> 
> From: Nitin Gupta <ngupta at vflare.org>
> 
> This patch fixes a memory leak in zsmalloc where the first
> subpage of each zspage is leaked when the zspage is freed.
> 
> Based on 3.4-rc1.
> 
> Signed-off-by: Nitin Gupta <ngupta at vflare.org>
> Acked-by: Seth Jennings <sjenning at linux.vnet.ibm.com>

This is a rather severe memory leak and will affect most
benchmarking anyone does to evaluate zcache in 3.4 (e.g. as
to whether zcache is suitable for promotion), so t'would be nice
to get this patch in for -rc2.  (Note it fixes a "regression"
since it affects zcache only in 3.4+ because the fix is to
the new zsmalloc allocator... so no change to stable trees.)

Acked-by: Dan Magenheimer <dan.magenheimer at oracle.com>

> ---
>  drivers/staging/zsmalloc/zsmalloc-main.c |   30 ++++++++++++++++++------------
>  1 files changed, 18 insertions(+), 12 deletions(-)
> 
> diff --git a/drivers/staging/zsmalloc/zsmalloc-main.c b/drivers/staging/zsmalloc/zsmalloc-main.c
> index 09caa4f..917461c 100644
> --- a/drivers/staging/zsmalloc/zsmalloc-main.c
> +++ b/drivers/staging/zsmalloc/zsmalloc-main.c
> @@ -267,33 +267,39 @@ static unsigned long obj_idx_to_offset(struct page *page,
>  	return off + obj_idx * class_size;
>  }
> 
> +static void reset_page(struct page *page)
> +{
> +	clear_bit(PG_private, &page->flags);
> +	clear_bit(PG_private_2, &page->flags);
> +	set_page_private(page, 0);
> +	page->mapping = NULL;
> +	page->freelist = NULL;
> +	reset_page_mapcount(page);
> +}
> +
>  static void free_zspage(struct page *first_page)
>  {
> -	struct page *nextp, *tmp;
> +	struct page *nextp, *tmp, *head_extra;
> 
>  	BUG_ON(!is_first_page(first_page));
>  	BUG_ON(first_page->inuse);
> 
> -	nextp = (struct page *)page_private(first_page);
> +	head_extra = (struct page *)page_private(first_page);
> 
> -	clear_bit(PG_private, &first_page->flags);
> -	clear_bit(PG_private_2, &first_page->flags);
> -	set_page_private(first_page, 0);
> -	first_page->mapping = NULL;
> -	first_page->freelist = NULL;
> -	reset_page_mapcount(first_page);
> +	reset_page(first_page);
>  	__free_page(first_page);
> 
>  	/* zspage with only 1 system page */
> -	if (!nextp)
> +	if (!head_extra)
>  		return;
> 
> -	list_for_each_entry_safe(nextp, tmp, &nextp->lru, lru) {
> +	list_for_each_entry_safe(nextp, tmp, &head_extra->lru, lru) {
>  		list_del(&nextp->lru);
> -		clear_bit(PG_private_2, &nextp->flags);
> -		nextp->index = 0;
> +		reset_page(nextp);
>  		__free_page(nextp);
>  	}
> +	reset_page(head_extra);
> +	__free_page(head_extra);
>  }
> 
>  /* Initialize a newly allocated zspage */
> --
> 1.7.5.4



More information about the devel mailing list