Re: Regression due to d98849aff879 (dma-direct: handle DMA_ATTR_NO_KERNEL_MAPPING in common code)

[Lucas Stach <l.stach@pengutronix.de>]

Am Dienstag, den 06.08.2019, 16:04 +0200 schrieb Christoph Hellwig:
> Ok, does this work?
> 
> --
> From 34d35f335a98f515f2516b515051e12eae744c8d Mon Sep 17 00:00:00 2001
> > From: Christoph Hellwig <hch@lst.de>
> Date: Tue, 6 Aug 2019 14:33:23 +0300
> Subject: dma-direct: fix DMA_ATTR_NO_KERNEL_MAPPING
> 
> The new DMA_ATTR_NO_KERNEL_MAPPING needs to actually assign
> a dma_addr to work.  Also skip it if the architecture needs
> forced decryption handling, as that needs a kernel virtual
> address.
> 
> Fixes: d98849aff879 (dma-direct: handle DMA_ATTR_NO_KERNEL_MAPPING in common code)
> > Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
>  kernel/dma/direct.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/kernel/dma/direct.c b/kernel/dma/direct.c
> index 59bdceea3737..b01064d884f2 100644
> --- a/kernel/dma/direct.c
> +++ b/kernel/dma/direct.c
> @@ -130,11 +130,13 @@ void *dma_direct_alloc_pages(struct device *dev, size_t size,
> >  	if (!page)
> >  		return NULL;
>  
> > -	if (attrs & DMA_ATTR_NO_KERNEL_MAPPING) {
> > +	if ((attrs & DMA_ATTR_NO_KERNEL_MAPPING) &&
> +	    !force_dma_unencrypted(dev)) {

dma_direct_free_pages() then needs the same check, as otherwise the cpu
address is treated as a cookie instead of a real address and the
encryption needs to be re-enabled.

Regards,
Lucas

>  		/* remove any dirty cache lines on the kernel alias */
> >  		if (!PageHighMem(page))
> >  			arch_dma_prep_coherent(page, size);
> >  		/* return the page pointer as the opaque cookie */
> > +		*dma_handle = phys_to_dma(dev, page_to_phys(page));
> >  		return page;
> >  	}
>