Message ID | 20220314074239.28507-2-yongqiang.niu@mediatek.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | [v1,1/1] drm/mediatek: fixup crtc event null pointer issue | expand |
On Mon, 2022-03-14 at 15:42 +0800, Yongqiang Niu wrote: > if crtc event is null pointer, do not send vblank event > > Signed-off-by: Yongqiang Niu <yongqiang.niu@mediatek.com> > --- > drivers/gpu/drm/mediatek/mtk_drm_crtc.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > index d661edf7e0fe..265fed446628 100644 > --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > @@ -92,6 +92,9 @@ static void mtk_drm_crtc_finish_page_flip(struct > mtk_drm_crtc *mtk_crtc) > struct drm_crtc *crtc = &mtk_crtc->base; > unsigned long flags; > > + if (!mtk_crtc->event) > + return; > + Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.com> > spin_lock_irqsave(&crtc->dev->event_lock, flags); > drm_crtc_send_vblank_event(crtc, mtk_crtc->event); > drm_crtc_vblank_put(crtc);
Hi, Yongqiang: On Mon, 2022-03-14 at 15:42 +0800, Yongqiang Niu wrote: > if crtc event is null pointer, do not send vblank event This is a bug-fix, so add a Fixes tag here. > > Signed-off-by: Yongqiang Niu <yongqiang.niu@mediatek.com> > --- > drivers/gpu/drm/mediatek/mtk_drm_crtc.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > index d661edf7e0fe..265fed446628 100644 > --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > @@ -92,6 +92,9 @@ static void mtk_drm_crtc_finish_page_flip(struct > mtk_drm_crtc *mtk_crtc) > struct drm_crtc *crtc = &mtk_crtc->base; > unsigned long flags; > > + if (!mtk_crtc->event) > + return; > + > spin_lock_irqsave(&crtc->dev->event_lock, flags); > drm_crtc_send_vblank_event(crtc, mtk_crtc->event); I think pending_needs_vblank is used to protect this situation. It seems that pending_needs_vblank should be protected by critical section. Regards, CK > drm_crtc_vblank_put(crtc);
On Mon, 2022-03-28 at 17:34 +0800, CK Hu wrote: > Hi, Yongqiang: > > On Mon, 2022-03-14 at 15:42 +0800, Yongqiang Niu wrote: > > if crtc event is null pointer, do not send vblank event > > This is a bug-fix, so add a Fixes tag here. Fixes tag will be added in next version. > > > > > Signed-off-by: Yongqiang Niu <yongqiang.niu@mediatek.com> > > --- > > drivers/gpu/drm/mediatek/mtk_drm_crtc.c | 3 +++ > > 1 file changed, 3 insertions(+) > > > > diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > > b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > > index d661edf7e0fe..265fed446628 100644 > > --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > > +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > > @@ -92,6 +92,9 @@ static void mtk_drm_crtc_finish_page_flip(struct > > mtk_drm_crtc *mtk_crtc) > > struct drm_crtc *crtc = &mtk_crtc->base; > > unsigned long flags; > > > > + if (!mtk_crtc->event) > > + return; > > + > > spin_lock_irqsave(&crtc->dev->event_lock, flags); > > drm_crtc_send_vblank_event(crtc, mtk_crtc->event); > > I think pending_needs_vblank is used to protect this situation. It > seems that pending_needs_vblank should be protected by critical > section. > > Regards, > CK actually, in the NG case pending_needs_vblank is true, but crtc->event is null. > > > drm_crtc_vblank_put(crtc); > >
Hi, Yongqiang: On Wed, 2022-05-11 at 18:08 +0800, yongqiang.niu wrote: > On Mon, 2022-03-28 at 17:34 +0800, CK Hu wrote: > > Hi, Yongqiang: > > > > On Mon, 2022-03-14 at 15:42 +0800, Yongqiang Niu wrote: > > > if crtc event is null pointer, do not send vblank event > > > > This is a bug-fix, so add a Fixes tag here. > > Fixes tag will be added in next version. > > > > > > > > Signed-off-by: Yongqiang Niu <yongqiang.niu@mediatek.com> > > > --- > > > drivers/gpu/drm/mediatek/mtk_drm_crtc.c | 3 +++ > > > 1 file changed, 3 insertions(+) > > > > > > diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > > > b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > > > index d661edf7e0fe..265fed446628 100644 > > > --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > > > +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c > > > @@ -92,6 +92,9 @@ static void > > > mtk_drm_crtc_finish_page_flip(struct > > > mtk_drm_crtc *mtk_crtc) > > > struct drm_crtc *crtc = &mtk_crtc->base; > > > unsigned long flags; > > > > > > + if (!mtk_crtc->event) > > > + return; > > > + > > > spin_lock_irqsave(&crtc->dev->event_lock, flags); > > > drm_crtc_send_vblank_event(crtc, mtk_crtc->event); > > > > I think pending_needs_vblank is used to protect this situation. It > > seems that pending_needs_vblank should be protected by critical > > section. > > > > Regards, > > CK > > actually, in the NG case pending_needs_vblank is true, but crtc- > >event > is null. It looks like that other driver access an invalid address which store mtk_crtc->event, so mediatek drm driver has no bug and we should fix the bug from other driver. This work around does not fix bug but just hide the bug. So try to fix this bug in other driver. Regards, CK > > > > > > drm_crtc_vblank_put(crtc); > > > > > >
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c index d661edf7e0fe..265fed446628 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c @@ -92,6 +92,9 @@ static void mtk_drm_crtc_finish_page_flip(struct mtk_drm_crtc *mtk_crtc) struct drm_crtc *crtc = &mtk_crtc->base; unsigned long flags; + if (!mtk_crtc->event) + return; + spin_lock_irqsave(&crtc->dev->event_lock, flags); drm_crtc_send_vblank_event(crtc, mtk_crtc->event); drm_crtc_vblank_put(crtc);
if crtc event is null pointer, do not send vblank event Signed-off-by: Yongqiang Niu <yongqiang.niu@mediatek.com> --- drivers/gpu/drm/mediatek/mtk_drm_crtc.c | 3 +++ 1 file changed, 3 insertions(+)