Hi.

On Mon, Jun 20, 2011 at 4:50 PM, Marek Szyprowski
[off-list ref] wrote:

 struct dma_map_ops {
-       void* (*alloc_coherent)(struct device *dev, size_t size,
-                               dma_addr_t *dma_handle, gfp_t gfp);
-       void (*free_coherent)(struct device *dev, size_t size,
-                             void *vaddr, dma_addr_t dma_handle);
+       void* (*alloc)(struct device *dev, size_t size,
+                               dma_addr_t *dma_handle, gfp_t gfp,
+                               struct dma_attrs *attrs);
+       void (*free)(struct device *dev, size_t size,
+                             void *vaddr, dma_addr_t dma_handle,
+                             struct dma_attrs *attrs);
+       int (*mmap)(struct device *, struct vm_area_struct *,
+                         void *, dma_addr_t, size_t, struct dma_attrs *attrs);
+
       dma_addr_t (*map_page)(struct device *dev, struct page *page,
                              unsigned long offset, size_t size,
                              enum dma_data_direction dir,

I still don't agree with your idea that change alloc_coherent() with alloc().
As I said before, we actually do not need dma_alloc_writecombine() anymore
because it is not different from dma_alloc_coherent() in ARM.
Most of other architectures do not have dma_alloc_writecombine().
If you want dma_alloc_coherent() to allocate user virtual address,
I believe that it is also available with mmap() you introduced.

Regards,
Cho KyongHo.