Adreno gpu crash

Hi,

I don’t see this reported anywhere… and it seems a gpu driver issue:

[   53.146461] qcom-venus c00000.video-codec: Adding to iommu group 2
[   53.150099] msm 900000.mdss: bound 901000.mdp (ops mdp5_ops [msm])
[   53.162307] msm 900000.mdss: bound 9a0000.hdmi-tx (ops msm_hdmi_ops [msm])
[   53.165652] adreno b00000.gpu: supply vdd not found, using dummy regulator
[   53.173606] adreno b00000.gpu: supply vddcx not found, using dummy regulator
[   53.178546] adreno b00000.gpu: [drm:msm_gpu_init [msm]] *ERROR* Couldn't register GPU cooling device
[   53.187572] msm 900000.mdss: bound b00000.gpu (ops a3xx_ops [msm])
[   53.357953] ------------[ cut here ]------------
[   53.357996] [CRTC:58:crtc-0] must not have both a cursor plane and a cursor_set func
[   53.358053] WARNING: CPU: 0 PID: 140 at drivers/gpu/drm/drm_mode_config.c:648 drm_mode_config_validate+0x240/0x4e0 [drm]
[   53.369389] Modules linked in: venus_core msm qcom_q6v5_pas v4l2_mem2mem qcom_pil_info slim_qcom_ngd_ctrl videobuf2_v4l2 qcom_q6v5 videobuf2_common videodev qcom_sysmon pdr_interface snd_soc_apq8096 snd_soc_qcom_common qcom_common mc qcom_glink_smem qmi_helpers mdt_loader slimbus drm_kms_helper hci_uart btqca btbcm bluetooth qcom_rng ecdh_generic ecc qrtr ns qcom_spmi_temp_alarm qcom_pon socinfo rpmsg_char icc_smd_rpm rmtfs_mem fuse drm ip_tables x_tables xhci_plat_hcd xhci_hcd phy_qcom_qusb2 rtc_pm8xxx
[   53.402562] CPU: 0 PID: 140 Comm: kworker/u8:6 Not tainted 5.13.0-qcomlt-arm64 #202
[   53.424764] Hardware name: Qualcomm Technologies, Inc. DB820c (DT)
[   53.432134] Workqueue: events_unbound deferred_probe_work_func
[   53.438382] pstate: 60000005 (nZCv daif -PAN -UAO -TCO BTYPE=--)
[   53.444204] pc : drm_mode_config_validate+0x240/0x4e0 [drm]
[   53.450359] lr : drm_mode_config_validate+0x240/0x4e0 [drm]
[   53.455662] sp : ffff8000138139b0
[   53.461215] x29: ffff8000138139c0 x28: ffff000081338060 x27: 0000000000000001
[   53.464684] x26: ffff80000911fe10 x25: ffff80000911fd00 x24: ffff000081b10af8
[   53.471803] x23: 0000000000000000 x22: 0000000000000000 x21: 0000000000000001
[   53.478921] x20: ffff000081b10800 x19: ffff000081b2d800 x18: 00000000fffffffd
[   53.486047] x17: 0000000000000000 x16: 0000000000000000 x15: 0000000000000020
[   53.493164] x14: ffffffffffffffff x13: ffff800011fc3230 x12: 00000000000007ce
[   53.500272] x11: 000000000000029a x10: ffff80001201f0a0 x9 : fffffffffffea440
[   53.507387] x8 : ffff800011fc3230 x7 : ffff80001201b230 x6 : 0000000000003e70
[   53.514506] x5 : ffff0000bb3a6918 x4 : 0000000000000000 x3 : 0000000000000027
[   53.521624] x2 : 0000000000000000 x1 : 0000000000000000 x0 : ffff0000829f1d80
[   53.528743] Call trace:
[   53.535853]  drm_mode_config_validate+0x240/0x4e0 [drm]
[   53.538116]  drm_dev_register+0x180/0x210 [drm]
[   53.543322]  msm_drm_bind+0x41c/0x5c0 [msm]
[   53.547837]  try_to_bring_up_master+0x228/0x310
[   53.552001]  __component_add+0xa0/0x190
[   53.556519]  component_add+0x18/0x24
[   53.560338]  msm_hdmi_dev_probe+0x20/0x2c [msm]
[   53.564154]  platform_probe+0x6c/0xdc
[   53.568406]  really_probe+0xe8/0x550
[   53.572227]  driver_probe_device+0x100/0x170
[   53.575873]  __device_attach_driver+0x9c/0x12c
[   53.580126]  bus_for_each_drv+0x78/0xd0
[   53.584379]  __device_attach+0xdc/0x1c0
[   53.588111]  device_initial_probe+0x18/0x24
[   53.591931]  bus_probe_device+0x9c/0xa4
[   53.596102]  deferred_probe_work_func+0xa4/0xf4
[   53.599918]  process_one_work+0x1cc/0x350
[   53.604432]  worker_thread+0x13c/0x470
[   53.608601]  kthread+0x154/0x160
[   53.612242]  ret_from_fork+0x10/0x30
[   53.615631] ---[ end trace d1603db97696943b ]---

This is on a custom board based on the dragonboard820c (APQ8096) with latest linaro BSP (with updated bootloaders and Debian sid release with kernel 5.13). Any ideas?

Linux linaro-developer 5.13.0-qcomlt-arm64 #202 SMP PREEMPT Fri Aug 13 02:49:49 UTC 2021 aarch64 GNU/Linux

Thanks

Daniel