On Thu, 29 Oct 2020 15:28:34 -0700 John Stultz wrote:
On Thu, Oct 29, 2020 at 12:10 AM Hillf Danton hdanton@sina.com wrote:
On Thu, 29 Oct 2020 00:16:24 +0000 John Stultz wrote:
@@ -194,6 +210,9 @@ static int system_heap_mmap(struct dma_buf *dmabuf, struct vm_area_struct *vma) struct sg_page_iter piter; int ret;
if (buffer->uncached)
vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot);
Wonder why you turn back to dma_mmap_wc() and friends?
Sorry, can you expand on what you are proposing here instead? I'm not sure I see how dma_alloc/mmap/*_wc() quite fits here.
I just wondered if *_wc() could save you two minutes or three. Can you shed some light on your concerns about their unfitness?