On Mon, Oct 19, 2015 at 03:05:46PM -0700, yu.dai@xxxxxxxxx wrote:
> From: Alex Dai <yu.dai@xxxxxxxxx>
>
> There is a memory leak warning message from i915_gem_context_clean
> when GuC submission is enabled. The reason is that the request (so
> the LRC associated with it) is freed early than moving the vma list
> to inactive. When retire a gem object, this patch moves its vma
> list to inactive first to avoid the false alert of memory leak.
>
> We are not seeing this in ExecList (non-GuC) mode because the gem
> request is moved to execlist_retired_req_list queue. The management
> of this queue, therefore free of LRC, happens after retire of vma
> list (i915_gem_retire_requests_ring).
Instead of hacking up the core active tracking code can we just fix lrc
context object tracking instead? This patch here seems to be supremely
fragile, and I really don't want it.
-Daniel
>
> Signed-off-by: Alex Dai <yu.dai@xxxxxxxxx>
> ---
> drivers/gpu/drm/i915/i915_gem.c | 29 ++++++++++++++++-------------
> 1 file changed, 16 insertions(+), 13 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> index 7d6b0c8..a903d45 100644
> --- a/drivers/gpu/drm/i915/i915_gem.c
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -2377,28 +2377,31 @@ i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring)
> RQ_BUG_ON(obj->last_read_req[ring] == NULL);
> RQ_BUG_ON(!(obj->active & (1 << ring)));
>
> + obj->active &= ~(1 << ring);
> + if (!obj->active) {
> + /* Bump our place on the bound list to keep it roughly in LRU
> + * order so that we don't steal from recently used but inactive
> + * objects (unless we are forced to ofc!)
> + */
> + list_move_tail(&obj->global_list,
> + &to_i915(obj->base.dev)->mm.bound_list);
> +
> + list_for_each_entry(vma, &obj->vma_list, vma_link) {
> + if (!list_empty(&vma->mm_list))
> + list_move_tail(&vma->mm_list,
> + &vma->vm->inactive_list);
> + }
> + }
> +
> list_del_init(&obj->ring_list[ring]);
> i915_gem_request_assign(&obj->last_read_req[ring], NULL);
>
> if (obj->last_write_req && obj->last_write_req->ring->id == ring)
> i915_gem_object_retire__write(obj);
>
> - obj->active &= ~(1 << ring);
> if (obj->active)
> return;
>
> - /* Bump our place on the bound list to keep it roughly in LRU order
> - * so that we don't steal from recently used but inactive objects
> - * (unless we are forced to ofc!)
> - */
> - list_move_tail(&obj->global_list,
> - &to_i915(obj->base.dev)->mm.bound_list);
> -
> - list_for_each_entry(vma, &obj->vma_list, vma_link) {
> - if (!list_empty(&vma->mm_list))
> - list_move_tail(&vma->mm_list, &vma->vm->inactive_list);
> - }
> -
> i915_gem_request_assign(&obj->last_fenced_req, NULL);
> drm_gem_object_unreference(&obj->base);
> }
> --
> 1.9.1
>
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx
--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
