On Fri, Jul 3, 2015 at 11:03 AM, Russell King - ARM Linux
[off-list ref] wrote:

On Wed, Jul 01, 2015 at 12:12:51PM +0100, Catalin Marinas wrote:

quoted

On Tue, Jun 30, 2015 at 11:29:06PM +0200, Lorenzo Nava wrote:

quoted

This patch allows the use of CMA for DMA coherent memory allocation.
At the moment if the input parameter "is_coherent" is set to true
the allocation is not made using the CMA, which I think is not the
desired behaviour.

Signed-off-by: Lorenzo Nava <redacted>
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>

If Russell doesn't have any objections, you can send the patch to
his patch system. See here for more information:

I'm left wondering whether this patch is really want Lorenzo wants.
From my reading of it, while this has the effect of allocating from
CMA for coherent devices, it's no different from the non-coherent
case, because by calling __alloc_from_contiguous(), we end up
remapping the allocated memory, removing the cacheability status
from the allocated pages.

This brings up an interesting point: presumably, it's been tested, and
people are happy with the performance it's giving, inspite of it not
returning cacheable memory... or maybe it hasn't been tested that much?

In the weekend I've run the test once again and what I got is that:

allocation non-coherent memory (CMA) -> prot = 0x647 (bufferable, non cacheable)
allocation coherent memory (with patch, CMA) -> prot = 0x65F (writealloc)

mmap of coherent memory (without mmap patched) -> prot = 0x707
(bufferable, non cacheable) -> ERROR
mmap of coherent memory (with mmap patch) -> prot = 0x71F (writealloc) -> OK

So, it looks like the page attributes are ok from the test I ran. I
didn't repeat the performance tests on the mapped memory, but they
were ok for me as well.

I've submitted the patch to your system.
If you have any other objections or any feedback, please let me know.

Lorenzo