TDA4VM-Q1: patch for rcu_preempt self-detected stall on CPU

Hi expert:
              We are using TDA4VE platform, not VM, sdk version j721s2 sdk0806, I see the patches you provided are all in the A core drivers/gpu/drm/tidss directory, I need to confirm some details, the changes you provided are in the DSS driver framework, right? DSS can be placed in both A core and R core, right? In our project, the node in the device tree of dss: dss@4a00000 is disabled, and our DSS driver is placed in mcu20, so the patch you provided is useless, and it will not call the API in this directory.

Gaston yan said:

I need to confirm some details, the changes you provided are in the DSS driver framework, right? DSS can be placed in both A core and R core, right? In our project, the node in the device tree of dss: dss@4a00000 is disabled, and our DSS driver is placed in mcu20, so the patch you provided is useless, and it will not call the API in this directory

Then the above patches should not matter. Yes agreed. For the TDA4VE we do not have any patches.

Thanks as well

Hi expert:

             In our TDA4VE sdk0806 platform, we did encounter the CPU stall problem. How can we debug and solve this problem?The log is as follows

[ 9559.015016] rcu: INFO: rcu_preempt detected expedited stalls on CPUs/tasks: { 0-... } 592913 jiffies s: 93 root: 0x1/.
[ 9559.026024] rcu: blocking rcu_node structures:
[ 9559.030629] Task dump for CPU 0:
[ 9559.033898] task:uartserver.bin state:R running task stack: 0 pid: 362 ppid: 1 flags:0x00000022
[ 9559.043996] Call trace:
[ 9559.046503] __switch_to+0xc0/0x118
[ 9559.050033] try_to_del_timer_sync+0x2c/0x90
[ 9559.054450] 0x1

Hello, 

We will need more details:

• What use case is running? 
• After how long you see the issue?
• Is this reproducible on the EVM?
• Do you see this every boot?

Best Regards,

Keerthy