Re: [PATCH 1/2] dma-mapping: add a dma_ops_bypass flag to struct device

On 24/03/2020 04:20, Christoph Hellwig wrote:

On Mon, Mar 23, 2020 at 07:58:01PM +1100, Alexey Kardashevskiy wrote:

quoted

quoted

quoted

0x100.0000.0000 .. 0x101.0000.0000

2x4G, each is 1TB aligned. And we can map directly only the first 4GB
(because of the maximum IOMMU table size) but not the other. And 1:1 on
that "pseries" is done with offset=0x0800.0000.0000.0000.

So we want to check every bus address against dev->bus_dma_limit, not
dev->coherent_dma_mask. In the example above I'd set bus_dma_limit to
0x0800.0001.0000.0000 and 1:1 mapping for the second 4GB would not be
tried. Does this sound reasonable? Thanks,

bus_dma_limit is just another limiting factor applied on top of
coherent_dma_mask or dma_mask respectively.

This is not enough for the task: in my example, I'd set bus limit to
0x0800.0001.0000.0000 but this would disable bypass for all RAM
addresses - the first and the second 4GB blocks.

So what about something like the version here:

http://git.infradead.org/users/hch/misc.git/shortlog/refs/heads/dma-bypass.3

dma_alloc_direct() and dma_map_direct() do the same thing now which is
good, did I miss anything else?

This lets us disable bypass automatically if this weird memory appears
in the system but does not let us have 1:1 after that even for normal
RAM. Thanks,


-- 
Alexey