about mmap dma-buf and sync

[Tiago Vignatti]

On 08/24/2015 03:01 PM, Tiago Vignatti wrote:
> yup, I think so. So IIUC the main changes needed for the drivers
> implement 2D sync lies in the dma_buf_sync_2d structure only. I.e.
> there's nothing really to be changed in the common code, right?

Do we have any special requirements in how we want pass the sync 
information to the drivers? I was thinking to push the whole 
responsibility for them, something like:

+int dma_buf_begin_cpu_access(struct dma_buf *dma_buf, size_t stride_bytes,
+                            size_t bytes_per_pixel, size_t num_regions,
+                            struct dma_buf_sync_region regions[], enum 
dma_data_direction dir);

Daniel Vetter mentioned about dma-buf design that should not track 
metadata but I haven't read anything about it, so do you think this 
looks alright?

Tiago