linux-arm-msm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH 0/2] drm: Fix atomic helper dirtyfb stalls
@ 2021-05-08 19:56 Rob Clark
  2021-05-08 19:56 ` [PATCH 2/2] drm/msm/dpu: Wire up needs_dirtyfb Rob Clark
  0 siblings, 1 reply; 5+ messages in thread
From: Rob Clark @ 2021-05-08 19:56 UTC (permalink / raw)
  To: dri-devel
  Cc: Rob Clark, Abhinav Kumar, Bernard,
	open list:DRM DRIVER FOR MSM ADRENO GPU, Hongbo Yao,
	Kalyan Thota, open list:DRM DRIVER FOR MSM ADRENO GPU, open list,
	Maxime Ripard, Qinglang Miao, Stephen Boyd

From: Rob Clark <robdclark@chromium.org>

Someone on IRC once asked an innocent enough sounding question:  Why
with xf86-video-modesetting is es2gears limited at 120fps.

So I broke out the perfetto tracing mesa MR and took a look.  It turns
out the problem was drm_atomic_helper_dirtyfb(), which would end up
waiting for vblank.. es2gears would rapidly push two frames to Xorg,
which would blit them to screen and in idle hook (I assume) call the
DIRTYFB ioctl.  Which in turn would do an atomic update to flush the
dirty rects, which would stall until the next vblank.  And then the
whole process would repeat.

But this is a bit silly, we only need dirtyfb for command mode DSI
panels.  So lets just skip it otherwise.

Rob Clark (2):
  drm: Fix dirtyfb stalls
  drm/msm/dpu: Wire up needs_dirtyfb

 drivers/gpu/drm/drm_damage_helper.c      |  8 ++++++++
 drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c | 14 ++++++++++++++
 include/drm/drm_modeset_helper_vtables.h | 14 ++++++++++++++
 3 files changed, 36 insertions(+)

-- 
2.30.2


^ permalink raw reply	[flat|nested] 5+ messages in thread

* [PATCH 2/2] drm/msm/dpu: Wire up needs_dirtyfb
  2021-05-08 19:56 [PATCH 0/2] drm: Fix atomic helper dirtyfb stalls Rob Clark
@ 2021-05-08 19:56 ` Rob Clark
  2021-05-09 15:38   ` Tested houdek.ryan
  2021-05-09 16:28   ` [PATCH 2/2] drm/msm/dpu: Wire up needs_dirtyfb Rob Clark
  0 siblings, 2 replies; 5+ messages in thread
From: Rob Clark @ 2021-05-08 19:56 UTC (permalink / raw)
  To: dri-devel
  Cc: Rob Clark, Rob Clark, Sean Paul, David Airlie, Daniel Vetter,
	Abhinav Kumar, Stephen Boyd, Maxime Ripard, Kalyan Thota,
	Qinglang Miao, Hongbo Yao,
	open list:DRM DRIVER FOR MSM ADRENO GPU,
	open list:DRM DRIVER FOR MSM ADRENO GPU, open list

From: Rob Clark <robdclark@chromium.org>

Signed-off-by: Rob Clark <robdclark@chromium.org>
---
 drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c
index 5a74f93e29da..868ee6136438 100644
--- a/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c
+++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c
@@ -143,6 +143,19 @@ static bool dpu_crtc_get_scanout_position(struct drm_crtc *crtc,
 	return true;
 }
 
+static bool dpu_crtc_needs_dirtyfb(struct drm_crtc *crtc)
+{
+	struct drm_encoder *encoder;
+
+	drm_for_each_encoder_mask (encoder, crtc->dev, crtc->state->encoder_mask) {
+		if (dpu_encoder_get_intf_mode(encoder) == INTF_MODE_CMD) {
+			return true;
+		}
+	}
+
+	return false;
+}
+
 static void _dpu_crtc_setup_blend_cfg(struct dpu_crtc_mixer *mixer,
 		struct dpu_plane_state *pstate, struct dpu_format *format)
 {
@@ -1343,6 +1356,7 @@ static const struct drm_crtc_helper_funcs dpu_crtc_helper_funcs = {
 	.atomic_begin = dpu_crtc_atomic_begin,
 	.atomic_flush = dpu_crtc_atomic_flush,
 	.get_scanout_position = dpu_crtc_get_scanout_position,
+	.needs_dirtyfb = dpu_crtc_needs_dirtyfb,
 };
 
 /* initialize crtc */
-- 
2.30.2


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* Tested
  2021-05-08 19:56 ` [PATCH 2/2] drm/msm/dpu: Wire up needs_dirtyfb Rob Clark
@ 2021-05-09 15:38   ` houdek.ryan
  2021-05-10 15:26     ` Tested Alex Deucher
  2021-05-09 16:28   ` [PATCH 2/2] drm/msm/dpu: Wire up needs_dirtyfb Rob Clark
  1 sibling, 1 reply; 5+ messages in thread
From: houdek.ryan @ 2021-05-09 15:38 UTC (permalink / raw)
  To: robdclark
  Cc: abhinavk, airlied, dri-devel, freedreno, kalyan_t, linux-arm-msm,
	linux-kernel, maxime, miaoqinglang, robdclark, sean, swboyd,
	yaohongbo, Ryan Houdek

I have tested this on my end and it resolves the 120hz problem.

Tested-By: Ryan Houdek <Houdek.Ryan@fex-emu.org>

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH 2/2] drm/msm/dpu: Wire up needs_dirtyfb
  2021-05-08 19:56 ` [PATCH 2/2] drm/msm/dpu: Wire up needs_dirtyfb Rob Clark
  2021-05-09 15:38   ` Tested houdek.ryan
@ 2021-05-09 16:28   ` Rob Clark
  1 sibling, 0 replies; 5+ messages in thread
From: Rob Clark @ 2021-05-09 16:28 UTC (permalink / raw)
  To: dri-devel
  Cc: Rob Clark, Sean Paul, David Airlie, Daniel Vetter, Abhinav Kumar,
	Stephen Boyd, Maxime Ripard, Kalyan Thota, Qinglang Miao,
	Hongbo Yao, open list:DRM DRIVER FOR MSM ADRENO GPU,
	open list:DRM DRIVER FOR MSM ADRENO GPU, open list, Ryan Houdek

On Sat, May 8, 2021 at 12:53 PM Rob Clark <robdclark@gmail.com> wrote:
>
> From: Rob Clark <robdclark@chromium.org>
>
> Signed-off-by: Rob Clark <robdclark@chromium.org>

From Ryan (sending this for him because mailing list workflow is lame):

I have tested this on my end and it resolves the 120hz problem.

Tested-By: Ryan Houdek <Houdek.Ryan@fex-emu.org>

> ---
>  drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)
>
> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c
> index 5a74f93e29da..868ee6136438 100644
> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c
> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c
> @@ -143,6 +143,19 @@ static bool dpu_crtc_get_scanout_position(struct drm_crtc *crtc,
>         return true;
>  }
>
> +static bool dpu_crtc_needs_dirtyfb(struct drm_crtc *crtc)
> +{
> +       struct drm_encoder *encoder;
> +
> +       drm_for_each_encoder_mask (encoder, crtc->dev, crtc->state->encoder_mask) {
> +               if (dpu_encoder_get_intf_mode(encoder) == INTF_MODE_CMD) {
> +                       return true;
> +               }
> +       }
> +
> +       return false;
> +}
> +
>  static void _dpu_crtc_setup_blend_cfg(struct dpu_crtc_mixer *mixer,
>                 struct dpu_plane_state *pstate, struct dpu_format *format)
>  {
> @@ -1343,6 +1356,7 @@ static const struct drm_crtc_helper_funcs dpu_crtc_helper_funcs = {
>         .atomic_begin = dpu_crtc_atomic_begin,
>         .atomic_flush = dpu_crtc_atomic_flush,
>         .get_scanout_position = dpu_crtc_get_scanout_position,
> +       .needs_dirtyfb = dpu_crtc_needs_dirtyfb,
>  };
>
>  /* initialize crtc */
> --
> 2.30.2
>

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: Tested
  2021-05-09 15:38   ` Tested houdek.ryan
@ 2021-05-10 15:26     ` Alex Deucher
  0 siblings, 0 replies; 5+ messages in thread
From: Alex Deucher @ 2021-05-10 15:26 UTC (permalink / raw)
  To: houdek.ryan
  Cc: Rob Clark, robdclark, Dave Airlie, linux-arm-msm, yaohongbo,
	LKML, Maling list - DRI developers, swboyd, Sean Paul,
	Qinglang Miao, abhinavk, kalyan_t, freedreno, Maxime Ripard

Sorry, what patch are you referring to?

Alex

On Mon, May 10, 2021 at 4:04 AM <houdek.ryan@fex-emu.org> wrote:
>
> I have tested this on my end and it resolves the 120hz problem.
>
> Tested-By: Ryan Houdek <Houdek.Ryan@fex-emu.org>

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2021-05-10 15:31 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-05-08 19:56 [PATCH 0/2] drm: Fix atomic helper dirtyfb stalls Rob Clark
2021-05-08 19:56 ` [PATCH 2/2] drm/msm/dpu: Wire up needs_dirtyfb Rob Clark
2021-05-09 15:38   ` Tested houdek.ryan
2021-05-10 15:26     ` Tested Alex Deucher
2021-05-09 16:28   ` [PATCH 2/2] drm/msm/dpu: Wire up needs_dirtyfb Rob Clark

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).