Re: [PATCH 1/4] mm, rmap: do easy-job first in anon_vma_fork

[Johannes Weiner]

On Tue, Aug 06, 2013 at 05:43:37PM +0900, Joonsoo Kim wrote:
> If we fail due to some errorous situation, it is better to quit
> without doing heavy work. So changing order of execution.
> 
> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
> 
> diff --git a/mm/rmap.c b/mm/rmap.c
> index a149e3a..c2f51cb 100644
> --- a/mm/rmap.c
> +++ b/mm/rmap.c
> @@ -278,19 +278,19 @@ int anon_vma_fork(struct vm_area_struct *vma, struct vm_area_struct *pvma)
>  	if (!pvma->anon_vma)
>  		return 0;
>  
> +	/* First, allocate required objects */
> +	avc = anon_vma_chain_alloc(GFP_KERNEL);
> +	if (!avc)
> +		goto out_error;
> +	anon_vma = anon_vma_alloc();
> +	if (!anon_vma)
> +		goto out_error_free_avc;
> +
>  	/*
> -	 * First, attach the new VMA to the parent VMA's anon_vmas,
> +	 * Then attach the new VMA to the parent VMA's anon_vmas,
>  	 * so rmap can find non-COWed pages in child processes.
>  	 */
>  	if (anon_vma_clone(vma, pvma))
> -		return -ENOMEM;
> -
> -	/* Then add our own anon_vma. */
> -	anon_vma = anon_vma_alloc();
> -	if (!anon_vma)
> -		goto out_error;
> -	avc = anon_vma_chain_alloc(GFP_KERNEL);
> -	if (!avc)
>  		goto out_error_free_anon_vma;

Which heavy work?  anon_vma_clone() is anon_vma_chain_alloc() in a
loop.

Optimizing error paths only makes sense if they are common and you
actually could save something by reordering.  This matches neither.
--
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/