Re: [PATCH] swiotlb-xen: override common mmap and get_sgtable dma ops

[Boris Ostrovsky]

On 6/11/21 5:55 AM, Roman Skakun wrote:
>  
> +static int
> +xen_swiotlb_dma_mmap(struct device *dev, struct vm_area_struct *vma,
> +		void *cpu_addr, dma_addr_t dma_addr, size_t size,
> +		unsigned long attrs)
> +{
> +	unsigned long user_count = vma_pages(vma);
> +	unsigned long count = PAGE_ALIGN(size) >> PAGE_SHIFT;
> +	unsigned long off = vma->vm_pgoff;
> +	struct page *page;
> +
> +	if (is_vmalloc_addr(cpu_addr))
> +		page = vmalloc_to_page(cpu_addr);
> +	else
> +		page = virt_to_page(cpu_addr);
> +
> +	vma->vm_page_prot = dma_pgprot(dev, vma->vm_page_prot, attrs);
> +
> +	if (dma_mmap_from_dev_coherent(dev, vma, cpu_addr, size, &ret))
> +		return -ENXIO;
> +
> +	if (off >= count || user_count > count - off)
> +		return -ENXIO;
> +
> +	return remap_pfn_range(vma, vma->vm_start,
> +			page_to_pfn(page) + vma->vm_pgoff,
> +			user_count << PAGE_SHIFT, vma->vm_page_prot);
> +}


I suggest you create a helper for computing page value and then revert 922659ea771b3fd728149262c5ea15608fab9719 and pass result of the helper instead of cpu_addr. Here and in xen_swiotlb_dma_get_sgtable().


And use this new helper in xen_swiotlb_free_coherent() too. I am curious though why this was not a problem when Stefano was looking at the problem that introduced this vmalloc check (i.e. 8b1e868f66076490189a36d984fcce286cdd6295). Stefano?


-boris