[Intel-gfx] [PATCH v2 4/4] drm/vgem: use shmem helpers

[Daniel Vetter]

On Thu, Jun 03, 2021 at 09:42:00PM +0300, Andi Shyti wrote:
> Hi Daniel,
> 
> > +/*
> > + * This just sets wc mode for shmem helpers. vgem doesn't have any begin/end cpu
> > + * access ioctls, there must use coherent memory or dma-buf sharing just wont
> > + * work.
> > + */
> > +static struct drm_gem_object *vgem_gem_create_object(struct drm_device *dev, size_t size)
> >  {
> > -	struct drm_vgem_gem_object *obj;
> > -	int npages;
> > +	struct drm_gem_shmem_object *obj;
> >  
> > -	obj = __vgem_gem_create(dev, attach->dmabuf->size);
> > -	if (IS_ERR(obj))
> > -		return ERR_CAST(obj);
> > -
> > -	npages = PAGE_ALIGN(attach->dmabuf->size) / PAGE_SIZE;
> > +	obj = kzalloc(sizeof(*obj), GFP_KERNEL);
> > +	if (!obj)
> > +		return NULL;
> >  
> > -	obj->table = sg;
> > -	obj->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
> > -	if (!obj->pages) {
> > -		__vgem_gem_destroy(obj);
> > -		return ERR_PTR(-ENOMEM);
> > -	}
> > +	obj->base.funcs = &drm_gem_shmem_funcs;
> > +	obj->map_wc = true;
> >  
> > -	obj->pages_pin_count++; /* perma-pinned */
> > -	drm_prime_sg_to_page_array(obj->table, obj->pages, npages);
> >  	return &obj->base;
> 
> here you are allocating a bigger object than what you are
> returning, in size. How does it get freed?

We're using the drm_gem_shmem_helper.c helpers, which set up all the shmem
functions for us, including an appropriate free callback.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch