Message ID | 1452825753-36364-1-git-send-email-puck.chen@hisilicon.com |
---|---|
State | New |
Headers | show |
On 2016/1/16 7:23, Russell King - ARM Linux wrote: > On Fri, Jan 15, 2016 at 03:03:42PM -0800, Laura Abbott wrote: >> (adding linux-arm and a few people) >> >> On 01/14/2016 06:42 PM, Chen Feng wrote: >>> The page is already alloc at ion_alloc function, >>> ion_mmap map the alloced pages to user-space. >>> >>> The default prot can be PTE_RDONLY. Take a look at >>> here: >>> set_pte_at() >>> arch/arm64/include/asm: >>> if (pte_dirty(pte) && pte_write(pte)) >>> pte_val(pte) &= ~PTE_RDONLY; >>> else >>> pte_val(pte) |= PTE_RDONLY; >>> >>> So with the dirty bit,it can improve the efficiency >>> and donnot need to handle memory fault when use access. >>> >>> Signed-off-by: Chen Feng <puck.chen@hisilicon.com> >>> Signed-off-by: Wei Dong <weidong2@hisilicon.com> >>> Reviewed-by: Zhuangluan Su <suzhuangluan@hisilicon.com> >>> --- >>> drivers/staging/android/ion/ion.c | 3 +++ >>> 1 file changed, 3 insertions(+) >>> >>> diff --git a/drivers/staging/android/ion/ion.c b/drivers/staging/android/ion/ion.c >>> index e237e9f..dba5942 100644 >>> --- a/drivers/staging/android/ion/ion.c >>> +++ b/drivers/staging/android/ion/ion.c >>> @@ -1026,6 +1026,9 @@ static int ion_mmap(struct dma_buf *dmabuf, struct vm_area_struct *vma) >>> if (!(buffer->flags & ION_FLAG_CACHED)) >>> vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot); >>> >>> + /*Default writeable*/ >>> + vma->vm_page_prot = pte_mkdirty(vma->vm_page_prot); >>> + >>> mutex_lock(&buffer->lock); >>> /* now map it to userspace */ >>> ret = buffer->heap->ops->map_user(buffer->heap, buffer, vma); >>> >> >> The extra fault is unfortunate but I'm skeptical about just setting >> pte_mkdirty. >> >> Catalin/Will, do you have any thoughts? Right now it seems like any >> range mapped with remap_pfn_range will have this extra fault >> behavior. Is marking the range dirty the best solution? Laura Abbott, I agree with you, it seems all the remap_pfn_range have this fault behavior. > > What happens if the mapping requested was read only - at the very > least, I don't think this should be done unconditionally. > Russell, I am not sure about doing this unconditionally, but it can waste memory&time while handling page fault with ion alloced page. And the page can be used directly.
diff --git a/drivers/staging/android/ion/ion.c b/drivers/staging/android/ion/ion.c index e237e9f..dba5942 100644 --- a/drivers/staging/android/ion/ion.c +++ b/drivers/staging/android/ion/ion.c @@ -1026,6 +1026,9 @@ static int ion_mmap(struct dma_buf *dmabuf, struct vm_area_struct *vma) if (!(buffer->flags & ION_FLAG_CACHED)) vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot); + /*Default writeable*/ + vma->vm_page_prot = pte_mkdirty(vma->vm_page_prot); + mutex_lock(&buffer->lock); /* now map it to userspace */ ret = buffer->heap->ops->map_user(buffer->heap, buffer, vma);