On Wed, Nov 04, 2020 at 02:00:21PM +0100, Christian König wrote:
> This is deprecated.
>
> Signed-off-by: Christian König <christian.koenig@xxxxxxx>
So I tried to prove to myself that ttm doesn't access ->pages for these
cases, and kinda couldn't. We still seem to allocate the pages array and
all that, and there's lots of code using ->pages all over. And between
ttm_bo_type_sg and TTM_PAGE_FLAG_SG I didn't manage to chase a whole lot
of paths to their full conclusion.
So I reduced my ambitions and wanted to prove that at least for dma-buf
imports aka ttm_bo_type_sg, we're guaranteed that we don't try to mmap
these to userspace. And also failed to find that check.
btw this is across all drivers, mostly ttm code, not radeon specific.
So conclusion, still a mess here that at least I can't see throug clearly
:-/ here = ttm_tt and the entire backing storage handling and everything
that ties into it. Probably the area that still has the most midlayer feel
to ttm with all the refactoring in-flight in still.
tldr; tried to review patches 1-3, gave up.
Cheers, Daniel
> ---
> drivers/gpu/drm/radeon/radeon_ttm.c | 9 +++++----
> 1 file changed, 5 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
> index 95038ac3382e..f41fcee35f70 100644
> --- a/drivers/gpu/drm/radeon/radeon_ttm.c
> +++ b/drivers/gpu/drm/radeon/radeon_ttm.c
> @@ -494,8 +494,8 @@ static int radeon_ttm_tt_pin_userptr(struct ttm_bo_device *bdev, struct ttm_tt *
> if (r)
> goto release_sg;
>
> - drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages,
> - gtt->ttm.dma_address, ttm->num_pages);
> + drm_prime_sg_to_page_addr_arrays(ttm->sg, NULL, gtt->ttm.dma_address,
> + ttm->num_pages);
>
> return 0;
>
> @@ -673,8 +673,9 @@ static int radeon_ttm_tt_populate(struct ttm_bo_device *bdev,
> }
>
> if (slave && ttm->sg) {
> - drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages,
> - gtt->ttm.dma_address, ttm->num_pages);
> + drm_prime_sg_to_page_addr_arrays(ttm->sg, NULL,
> + gtt->ttm.dma_address,
> + ttm->num_pages);
> return 0;
> }
>
> --
> 2.25.1
>
> _______________________________________________
> dri-devel mailing list
> dri-devel@xxxxxxxxxxxxxxxxxxxxx
> https://lists.freedesktop.org/mailman/listinfo/dri-devel
--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
dri-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/dri-devel
