Re: [patch][rfc] nommu: reverse mappings for nommu to solve get_user_pages problem

[Nick Piggin <nickpiggin@yahoo.com.au>]

Nick Piggin wrote:

>  int get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
>  	unsigned long start, int len, int write, int force,
>  	struct page **pages, struct vm_area_struct **vmas)
>  {
>  	int i;
> -	static struct vm_area_struct dummy_vma;
> +	struct page *__page;
> +	static struct vm_area_struct *__vma;
> +	unsigned long addr = start;
>  
>  	for (i = 0; i < len; i++) {
> +		__vma = find_vma(mm, addr);
> +		if (!__vma)
> +			goto out_failed;
> +
> +		__page = virt_to_page(addr);
> +		if (!__page)
> +			goto out_failed;
> +
> +		BUG_ON(page_vma(__page) != __vma);
> +

Actually this check is leftover from a previous version. I think it
needs to be removed.

>  		if (pages) {
> -			pages[i] = virt_to_page(start);
> -			if (pages[i])
> -				page_cache_get(pages[i]);
> +			if (!__page->mapping) {
> +				printk(KERN_INFO "get_user_pages on unaligned"
> +						"anonymous page unsupported\n");				dump_stack();
> +				goto out_failed;
> +			}
> +

And this could trigger for file-backed pages that have been truncated meanwhile
I think. It wouldn't be a problem for a simple test-run, but does need to be
reworked slightly in order to be correct. Sub-page anonymous mappings cause a
lot of headaches :)

--
SUSE Labs, Novell Inc.
Send instant messages to your online friends http://au.messenger.yahoo.com 
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/