On Tue, Nov 17, 2015 at 11:53:28AM +0900, Inki Dae wrote:
> 
> 
> 2015년 11월 17일 01:23에 Daniel Vetter 이(가) 쓴 글:
> > On Mon, Nov 16, 2015 at 05:22:42PM +0100, Daniel Vetter wrote:
> >> On Tue, Jul 28, 2015 at 05:53:23PM +0900, Joonyoung Shim wrote:
> >>> Don't create a fake mmap offset in exynos_drm_gem_dumb_map_offset. If
> >>> not, it will call drm_gem_create_mmap_offset whenever user requests
> >>> DRM_IOCTL_MODE_MAP_DUMB ioctl.
> >>>
> >>> Signed-off-by: Joonyoung Shim <jy0922.shim at samsung.com>
> >>> ---
> >>>  drivers/gpu/drm/exynos/exynos_drm_gem.c | 12 +++++++-----
> >>>  1 file changed, 7 insertions(+), 5 deletions(-)
> >>>
> >>> diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c 
> >>> b/drivers/gpu/drm/exynos/exynos_drm_gem.c
> >>> index 550d267..c76aa8a 100644
> >>> --- a/drivers/gpu/drm/exynos/exynos_drm_gem.c
> >>> +++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c
> >>> @@ -151,6 +151,13 @@ struct exynos_drm_gem_obj 
> >>> *exynos_drm_gem_init(struct drm_device *dev,
> >>>           return ERR_PTR(ret);
> >>>   }
> >>>  
> >>> + ret = drm_gem_create_mmap_offset(obj);
> >>> + if (ret < 0) {
> >>> +         drm_gem_object_release(obj);
> >>> +         kfree(exynos_gem_obj);
> >>> +         return ERR_PTR(ret);
> >>> + }
> >>> +
> >>>   DRM_DEBUG_KMS("created file object = 0x%x\n", (unsigned int)obj->filp);
> >>>  
> >>>   return exynos_gem_obj;
> >>> @@ -521,14 +528,9 @@ int exynos_drm_gem_dumb_map_offset(struct drm_file 
> >>> *file_priv,
> >>>           goto unlock;
> >>>   }
> >>>  
> >>> - ret = drm_gem_create_mmap_offset(obj);
> >>
> >> drm_gem_create_mmap_offset internally checks whether it's been already
> >> (protected by locks), so this code is perfectly fine. I don't see any
> >> justification for this change (but only noticed it because rockchip
> >> cargo-culted this change).
> > 
> > I think it'd be good to revert this to stay consistent with cma helpers
> > and other drivers.
> 
> At least, seems cma halper doesn't also call drm_gem_create_mmap_offset 
> function
> at drm_gem_cma_dumb_map_offset function and calls it at cma creation instead.
> So I think now Exynos drm keeps consistent with cma helper.

Indeed, common style is to create the offset at alloc time - I checked a
few other drivers too. Never mind my comment, sorry for the noise.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch

Reply via email to