On 05/07/2020 09:41 AM, Huacai Chen wrote:
Hi, Christoph,
On Wed, May 6, 2020 at 10:44 PM Christoph Hellwig <[email protected]> wrote:
On Wed, May 06, 2020 at 04:47:30PM +0800, Huacai Chen wrote:
For the above reasons, I think what you are concerned is not a
big deal.
I don't think so, this is obviously a regression. If we can accept a
regression of RS780E, why we still maintain Loongson-2EF rather than
simply drop them?
While I much prefer to use the default, regression an otherwise
working platform seems like a bad idea. I don't really know much
about the Loongson platforms, do they all boot using the same kernel
image?
All Loongson-3 machines (with LS7A bridge or RS780 bridge) use the
same kernel image.
Hi Christoph, Huacai and Jiaxun,
Thank you very much for your reviews and discussions.
If you agree to use the platform dependent implementation
of __phys_to_dma() and __dma_to_phys(), I will make a slight
modification based on the v4 patch [1] to put ls7a things
before rs780e things, and then send v7 as soon as possible.
If anyone has any objections, please let me know.
[1] https://lore.kernel.org/patchwork/patch/1220010/
Thanks,
Tiezhu Yang
