[PATCH v3] drm/msm/dpu: Capture dpu snapshot when frame_done_timer timeouts

[Paloma Arellano]

Trigger a devcoredump to dump dpu registers and capture the drm atomic
state when the frame_done_timer timeouts.

v2: Optimize the format in which frame_done_timeout_cnt is incremented
v3: Describe parameter frame_done_timeout_cnt in dpu_encoder_virt

Reported-by: kernel test robot <lkp at intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202312010225.2OJWLKmA-lkp@intel.com/
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov at linaro.org>
Signed-off-by: Paloma Arellano <quic_parellan at quicinc.com>
---
 drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c | 14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
index 1cf7ff6caff4e..1c3560e1c6252 100644
--- a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
+++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
@@ -151,6 +151,8 @@ enum dpu_enc_rc_states {
  * @crtc_frame_event_cb:	callback handler for frame event
  * @crtc_frame_event_cb_data:	callback handler private data
  * @frame_done_timeout_ms:	frame done timeout in ms
+ * @frame_done_timeout_cnt:	atomic counter tracking the number of frame
+ * 				done timeouts
  * @frame_done_timer:		watchdog timer for frame done event
  * @disp_info:			local copy of msm_display_info struct
  * @idle_pc_supported:		indicate if idle power collaps is supported
@@ -191,6 +193,7 @@ struct dpu_encoder_virt {
 	void *crtc_frame_event_cb_data;
 
 	atomic_t frame_done_timeout_ms;
+	atomic_t frame_done_timeout_cnt;
 	struct timer_list frame_done_timer;
 
 	struct msm_display_info disp_info;
@@ -1204,6 +1207,8 @@ static void dpu_encoder_virt_atomic_enable(struct drm_encoder *drm_enc,
 
 	dpu_enc->dsc = dpu_encoder_get_dsc_config(drm_enc);
 
+	atomic_set(&dpu_enc->frame_done_timeout_cnt, 0);
+
 	if (disp_info->intf_type == INTF_DP)
 		dpu_enc->wide_bus_en = msm_dp_wide_bus_available(priv->dp[index]);
 	else if (disp_info->intf_type == INTF_DSI)
@@ -2115,11 +2120,12 @@ static int _dpu_encoder_status_show(struct seq_file *s, void *data)
 	for (i = 0; i < dpu_enc->num_phys_encs; i++) {
 		struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i];
 
-		seq_printf(s, "intf:%d  wb:%d  vsync:%8d     underrun:%8d    ",
+		seq_printf(s, "intf:%d  wb:%d  vsync:%8d     underrun:%8d    frame_done_cnt:%d",
 				phys->hw_intf ? phys->hw_intf->idx - INTF_0 : -1,
 				phys->hw_wb ? phys->hw_wb->idx - WB_0 : -1,
 				atomic_read(&phys->vsync_cnt),
-				atomic_read(&phys->underrun_cnt));
+				atomic_read(&phys->underrun_cnt),
+				atomic_read(&dpu_enc->frame_done_timeout_cnt));
 
 		seq_printf(s, "mode: %s\n", dpu_encoder_helper_get_intf_type(phys->intf_mode));
 	}
@@ -2341,6 +2347,9 @@ static void dpu_encoder_frame_done_timeout(struct timer_list *t)
 
 	DPU_ERROR_ENC(dpu_enc, "frame done timeout\n");
 
+	if (atomic_inc_return(&dpu_enc->frame_done_timeout_cnt) == 1)
+		msm_disp_snapshot_state(drm_enc->dev);
+
 	event = DPU_ENCODER_FRAME_EVENT_ERROR;
 	trace_dpu_enc_frame_done_timeout(DRMID(drm_enc), event);
 	dpu_enc->crtc_frame_event_cb(dpu_enc->crtc_frame_event_cb_data, event);
@@ -2392,6 +2401,7 @@ struct drm_encoder *dpu_encoder_init(struct drm_device *dev,
 		goto fail;
 
 	atomic_set(&dpu_enc->frame_done_timeout_ms, 0);
+	atomic_set(&dpu_enc->frame_done_timeout_cnt, 0);
 	timer_setup(&dpu_enc->frame_done_timer,
 			dpu_encoder_frame_done_timeout, 0);
 
-- 
2.39.2