On Thu, Oct 12, 2017 at 12:39:25PM +0300, Jyri Sarha wrote:
> We are using the vrefresh to check if we are too close to vertical
> sync to update the two framebuffer DMA registers and risk a
> collision. The vrefresh is coming from user space and normally it is
> not used for anything. For instance xserver leaves vrefresh to zero
> causing a division by zero when setting the mode. We want to make sure
> the value is valid and force its recalculation in
> tilcdc_crtc_set_mode().
>
> Reported-by: Kevin Hao <kexin.hao@xxxxxxxxxxxxx>
> Fixes: 11abbc9f39e0 ("drm/tilcdc: Set framebuffer DMA address to HW only if CRTC is enabled")
> Cc: <stable@xxxxxxxxxxxxxxx> # v4.11+
> Signed-off-by: Jyri Sarha <jsarha@xxxxxx>
> ---
> drivers/gpu/drm/tilcdc/tilcdc_crtc.c | 10 ++++++++++
> 1 file changed, 10 insertions(+)
>
> diff --git a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
> index 704db24..aa5e87c 100644
> --- a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
> +++ b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
> @@ -459,6 +459,16 @@ static void tilcdc_crtc_set_mode(struct drm_crtc *crtc)
> drm_framebuffer_get(fb);
>
> crtc->hwmode = crtc->state->adjusted_mode;
> +
> + /* We are using the vrefresh to check if we are too close to
> + * vertical sync to update the two framebuffer DMA registers
> + * and risk a collision. The vrefresh is coming from user
> + * space and normally it is not used for anything. We want to
> + * make sure the value is valid and force its recalculation
> + * here.
> + */
> + crtc->hwmode.vrefresh = 0;
Why do we need this?
Thanks,
Kevin
> + crtc->hwmode.vrefresh = drm_mode_vrefresh(&crtc->hwmode);
> }
>
> static void tilcdc_crtc_enable(struct drm_crtc *crtc)
> --
> 1.9.1
>
>
> Texas Instruments Finland Oy, Porkkalankatu 22, 00180 Helsinki. Y-tunnus/Business ID: 0615521-4. Kotipaikka/Domicile: Helsinki
>
>
Attachment:
signature.asc
Description: PGP signature