Re: [PATCH 1/2] dma-mapping: fix page attributes for dma_mmap_*

[Shawn Anastasio <shawn@anastas.io>]

On 8/5/19 10:01 AM, Christoph Hellwig wrote:
> diff --git a/include/linux/dma-noncoherent.h b/include/linux/dma-noncoherent.h
> index 3813211a9aad..9ae5cee543c4 100644
> --- a/include/linux/dma-noncoherent.h
> +++ b/include/linux/dma-noncoherent.h
> @@ -42,13 +42,8 @@ void arch_dma_free(struct device *dev, size_t size, void *cpu_addr,
>   		dma_addr_t dma_addr, unsigned long attrs);
>   long arch_dma_coherent_to_pfn(struct device *dev, void *cpu_addr,
>   		dma_addr_t dma_addr);
> -
> -#ifdef CONFIG_ARCH_HAS_DMA_MMAP_PGPROT
>   pgprot_t arch_dma_mmap_pgprot(struct device *dev, pgprot_t prot,
>   		unsigned long attrs);
> -#else
> -# define arch_dma_mmap_pgprot(dev, prot, attrs)	pgprot_noncached(prot)
> -#endif

Nit, but maybe the prototype should still be ifdef'd here? It at least
could prevent a reader from incorrectly thinking that the function is
always present.

Also, like Will mentioned earlier, the function name isn't entirely
accurate anymore. I second the suggestion of using something like
arch_dma_noncoherent_pgprot(). As for your idea of defining
pgprot_dmacoherent for all architectures as

#ifndef pgprot_dmacoherent
#define pgprot_dmacoherent pgprot_noncached
#endif

I think that the name here is kind of misleading too, since this
definition will only be used when there is no support for proper
DMA coherency.