Re: [PATCH] drm/amd/display: fix flickering caused by S/G mode

From: Hamza Mahfooz
Date: Mon Apr 17 2023 - 12:31:57 EST


On 4/17/23 11:41, Wu, Hersen wrote:
[AMD Official Use Only - General]

Hi,

The change applies to all AMD GPU ASIC.
Please communicate with issue reporter to see if the issue could be reproduced older ASIC, like Mendocino, CZN.

From the community reports, it can be reproduced on as far back as the
Ryzen 4800H (which I guess is Renoir).


Thanks!
Hersen


-----Original Message-----
From: Mahfooz, Hamza <Hamza.Mahfooz@xxxxxxx>
Sent: Monday, April 17, 2023 11:26 AM
To: Koenig, Christian <Christian.Koenig@xxxxxxx>; amd-gfx@xxxxxxxxxxxxxxxxxxxxx
Cc: stable@xxxxxxxxxxxxxxx; Wentland, Harry <Harry.Wentland@xxxxxxx>; Li, Sun peng (Leo) <Sunpeng.Li@xxxxxxx>; Siqueira, Rodrigo <Rodrigo.Siqueira@xxxxxxx>; Deucher, Alexander <Alexander.Deucher@xxxxxxx>; Pan, Xinhui <Xinhui.Pan@xxxxxxx>; David Airlie <airlied@xxxxxxxxx>; Daniel Vetter <daniel@xxxxxxxx>; Pillai, Aurabindo <Aurabindo.Pillai@xxxxxxx>; Zhuo, Qingqing (Lillian) <Qingqing.Zhuo@xxxxxxx>; Hans de Goede <hdegoede@xxxxxxxxxx>; Wu, Hersen <hersenxs.wu@xxxxxxx>; Wang, Chao-kai (Stylon) <Stylon.Wang@xxxxxxx>; Tuikov, Luben <Luben.Tuikov@xxxxxxx>; dri-devel@xxxxxxxxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx
Subject: Re: [PATCH] drm/amd/display: fix flickering caused by S/G mode


On 4/17/23 11:03, Christian König wrote:
Am 17.04.23 um 16:51 schrieb Hamza Mahfooz:

On 4/17/23 01:59, Christian König wrote:
Am 14.04.23 um 21:33 schrieb Hamza Mahfooz:
Currently, we allow the framebuffer for a given plane to move
between memory domains, however when that happens it causes the
screen to flicker, it is even possible for the framebuffer to
change memory domains on every plane update (causing a continuous flicker effect).
So,
to fix this, don't perform an immediate flip in the aforementioned
case.

That sounds strongly like you just forget to wait for the move to
finish!

What is the order of things done here? E.g. who calls
amdgpu_bo_pin() and who waits for fences for finish signaling? Is
that maybe just in the wrong order?

The pinning logic is in dm_plane_helper_prepare_fb(). Also, it seems
like we wait for the fences in amdgpu_dm_atomic_commit_tail(), using
drm_atomic_helper_wait_for_fences(). The ordering should be fine as
well, since prepare_fb() is always called before atomic_commit_tail().

Ok, then why is there any flickering?

BTW reserving a fence slot is completely unnecessary. That looks like
you copy&pasted the code from somewhere else without checking what it
actually does.

It seemed like it was necessary to read `tbo.resource` since the documentation for `struct ttm_buffer_object` makes mention of a "bo::resv::reserved" lock.


Regards,
Christian.



Regards,
Christian.


Cc: stable@xxxxxxxxxxxxxxx
Fixes: 81d0bcf99009 ("drm/amdgpu: make display pinning more
flexible
(v2)")
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@xxxxxxx>
---
  .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 41
++++++++++++++++++-
  1 file changed, 39 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index da3045fdcb6d..9a4e7408384a 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -7897,6 +7897,34 @@ static void amdgpu_dm_commit_cursors(struct
drm_atomic_state *state)
              amdgpu_dm_plane_handle_cursor_update(plane,
old_plane_state);
  }
+static inline uint32_t get_mem_type(struct amdgpu_device *adev,
+                    struct drm_gem_object *obj,
+                    bool check_domain) {
+    struct amdgpu_bo *abo = gem_to_amdgpu_bo(obj);
+    uint32_t mem_type;
+
+    if (unlikely(amdgpu_bo_reserve(abo, true)))
+        return 0;
+
+    if (unlikely(dma_resv_reserve_fences(abo->tbo.base.resv, 1)))
+        goto err;
+
+    if (check_domain &&
+        amdgpu_display_supported_domains(adev, abo->flags) !=
+        (AMDGPU_GEM_DOMAIN_VRAM | AMDGPU_GEM_DOMAIN_GTT))
+        goto err;
+
+    mem_type = abo->tbo.resource->mem_type;
+    amdgpu_bo_unreserve(abo);
+
+    return mem_type;
+
+err:
+    amdgpu_bo_unreserve(abo);
+    return 0;
+}
+
  static void amdgpu_dm_commit_planes(struct drm_atomic_state
*state,
                      struct dc_state *dc_state,
                      struct drm_device *dev, @@ -7916,6 +7944,7 @@
static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
to_dm_crtc_state(drm_atomic_get_old_crtc_state(state, pcrtc));
      int planes_count = 0, vpos, hpos;
      unsigned long flags;
+    uint32_t mem_type;
      u32 target_vblank, last_flip_vblank;
      bool vrr_active = amdgpu_dm_crtc_vrr_active(acrtc_state);
      bool cursor_update = false;
@@ -8035,13 +8064,21 @@ static void amdgpu_dm_commit_planes(struct
drm_atomic_state *state,
              }
          }
+        mem_type = get_mem_type(dm->adev,
+old_plane_state->fb->obj[0],
+                    true);
+
          /*
           * Only allow immediate flips for fast updates that don't
-         * change FB pitch, DCC state, rotation or mirroing.
+         * change memory domain, FB pitch, DCC state, rotation or
+         * mirroring.
           */
          bundle->flip_addrs[planes_count].flip_immediate =
              crtc->state->async_flip &&
-            acrtc_state->update_type == UPDATE_TYPE_FAST;
+            acrtc_state->update_type == UPDATE_TYPE_FAST &&
+            (!mem_type || (mem_type && get_mem_type(dm->adev,
+                                fb->obj[0],
+                                false) ==
+                       mem_type));
          timestamp_ns = ktime_get_ns();
bundle->flip_addrs[planes_count].flip_timestamp_in_us =
div_u64(timestamp_ns, 1000);


--
Hamza
--
Hamza