On Fri, Nov 9, 2012 at 10:56 AM, Inki Dae <inki.dae@xxxxxxxxxxx> wrote: > > > 2012/11/9 Rob Clark <robdclark@xxxxxxxxx> >> >> On Fri, Nov 9, 2012 at 1:39 AM, Inki Dae <inki.dae@xxxxxxxxxxx> wrote: >> > This patch fixes access issue to invalid memory region. >> > >> > crtc had only one drm_framebuffer object so when framebuffer >> > cleanup was requested after page flip, it'd try to disable >> > hardware overlay to current crtc. >> > But if current crtc points to another drm_framebuffer, >> > This may induce invalid memory access. >> >> btw, this should instead be done by holding a ref to the GEM >> object(s).. or these days you can increment the reference count on >> the fb and let the fb hold ref's to the GEM object(s) (which makes it >> a bit easier to deal with multi-planar formats) >> > > Rob, let's discuss that again after you read my latest comment. Please, see > my latest comment. oh, I think our emails crossed paths. But I think the crtc (or plane) holding a ref to the fb until scanout stops should solve the issue for you. BR, -R > Thanks, > Inki Dae > _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/dri-devel
