[BUG 5.14] arm64/mm: dma memory mapping fails (in some cases)

[Mike Rapoport]

On Tue, Sep 21, 2021 at 10:20:07AM +0200, Christoph Hellwig wrote:
> On Mon, Sep 20, 2021 at 11:57:58AM +0100, Catalin Marinas wrote:
> > > As this WARN_ON(pfn_valid()) is only present in dma_map_resource() it's
> > > probably safe to drop it entirely. 
> > 
> > I agree, we should drop it. IIUC dma_map_resource() does not create any
> > kernel mapping to cause problems with attribute aliasing. You'd need a
> > prior devm_ioremap_resource() if you want access to that range from the
> > CPU side. For arm64 at least, the latter ends up with a
> > pfn_is_map_memory() check.
> 
> It doesn't create any new mappings.  The only real issue is that it
> does the wrong thing for RAM in a way that might not be noticed on
> simple (x86/PC) platforms.

But if the mapping request was rejected by devm_ioremap_resource() because
of an attempt to map RAM, why we would get to dma_map_resource() at all? 


-- 
Sincerely yours,
Mike.