AM62A7: USB fail in reboot test

Hi Bin Liu,

After applying the patch we get the below prints

am65-cpsw-nuss 8000000.ethernet: ALE Table size 512
[    1.401848] pps pps0: new PPS source ptp0
[    1.406124] am65-cpsw-nuss 8000000.ethernet: CPTS ver 0x4e8a010c, freq:500000000, add_val:1 pps:1
[    1.416314] am65-cpsw-nuss 8000000.ethernet: set new flow-id-base 19
[    1.424296] dwc3-am62 f900000.dwc3-usb: invalid resource
[    1.429631] dwc3-am62 f900000.dwc3-usb: can't map IOMEM resource 1
[    1.435826] dwc3-am62: probe of f900000.dwc3-usb failed with error -22
[    1.442775] dwc3-am62 f910000.dwc3-usb: invalid resource
[    1.448089] dwc3-am62 f910000.dwc3-usb: can't map IOMEM resource 1
[    1.454278] dwc3-am62: probe of f910000.dwc3-usb failed with error -22

Regards,
Sushmitha K

Hi Sushmitha,

Sushmitha K said:

[    1.424296] dwc3-am62 f900000.dwc3-usb: invalid resource
[    1.429631] dwc3-am62 f900000.dwc3-usb: can't map IOMEM resource 1

The first section of the patch is for arch/arm64/boot/dts/ti/k3-am62-main.dtsi.

The kernel log indicates you either didn't apply the patch correctly, or you didn't update the new board dtb file to your Linux root filesystem.

Sorry, I just realized you use AM62A devices, but the referred patch is for AM62. Can you please refer to the patch to k3-am62-main.dtsi, and manually modify the same in k3-am62a-main.dtsi? This should solve the kernel dwc3 probe error.

Hi Bin Liu,

Usb fail issue got solved but below mentioned crash prints are seen in SW reboot test kept for 12 hours.

Can you please try keeping SW reboot test continuously  for 12 hours in your board and share your observation?


[   31.395660] rcu: INFO: rcu_preempt self-detected stall on CPU
[   31.401413] rcu:       0-....: (5249 ticks this GP) idle=ed04/0/0x3 softirq=2276/2276 fqs=2600
[   31.409578]    (t=5253 jiffies g=1125 q=263 ncpus=4)
[   31.414361] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G           O       6.1.46-g247b2535b2 #24
[   31.422869] Hardware name: iW-RainboW-G55M-TI-AM62A7 OSM (DT)
[   31.428600] pstate: 20000005 (nzCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[   31.435548] pc : _raw_spin_unlock_irqrestore+0xc/0x50
[   31.440602] lr : vblank_disable_fn+0x84/0xa0
[   31.444865] sp : ffff800008003e30
[   31.448167] x29: ffff800008003e30 x28: 0000000000000005 x27: 0000000000000020
[   31.455291] x26: ffff80000863d5c0 x25: ffff8000092879c0 x24: ffff00007fc3c1a8
[   31.462415] x23: 0000000000000000 x22: ffff000000d1e880 x21: ffff0000016e3000
[   31.469538] x20: ffff0000016e3144 x19: 0000000000000000 x18: ffff800009413000
[   31.476662] x17: ffff800076b2e000 x16: ffff800008000000 x15: 000000000000003c
[   31.483786] x14: ffffffffffffffff x13: 0000000000000001 x12: 000000000000003c
[   31.490909] x11: 0000000000000901 x10: 0000000000000002 x9 : 0000000000000031
[   31.498033] x8 : 000000000000002b x7 : ffff800009413948 x6 : 0000044dc64ada2d
[   31.505156] x5 : 03ffffffffffffff x4 : ffff0000002901c0 x3 : 0000000000000000
[   31.512279] x2 : 0000000000000000 x1 : 0000000000000000 x0 : ffff0000016e3144
[   31.519402] Call trace:
[   31.521839]  _raw_spin_unlock_irqrestore+0xc/0x50
[   31.526533]  call_timer_fn.constprop.0+0x24/0x80
[   31.531142]  __run_timers.part.0+0x1f4/0x234
[   31.535401]  run_timer_softirq+0x3c/0x7c
[   31.539313]  _stext+0x124/0x2a4
[   31.542446]  ____do_softirq+0x10/0x20
[   31.546100]  call_on_irq_stack+0x24/0x4c
[   31.550013]  do_softirq_own_stack+0x1c/0x30
[   31.554184]  __irq_exit_rcu+0xcc/0xf4
[   31.557839]  irq_exit_rcu+0x10/0x20
[   31.561318]  el1_interrupt+0x38/0x70
[   31.564887]  el1h_64_irq_handler+0x18/0x2c
[   31.568975]  el1h_64_irq+0x64/0x68
[   31.572366]  arch_cpu_idle+0x18/0x2c
[   31.575931]  default_idle_call+0x30/0x6c
[   31.579844]  do_idle+0x248/0x2c0
[   31.583067]  cpu_startup_entry+0x28/0x30
[   31.586980]  kernel_init+0x0/0x130
[   31.590373]  arch_post_acpi_subsys_init+0x0/0x18
[   31.594982]  start_kernel+0x650/0x694
[   31.598635]  __primary_switched+0xbc/0xc4
[   94.611657] rcu: INFO: rcu_preempt self-detected stall on CPU
[   94.617393] rcu:       0-....: (21003 ticks this GP) idle=ed04/0/0x3 softirq=2276/2276 fqs=10425
[   94.625730]    (t=21057 jiffies g=1125 q=263 ncpus=4)
[   94.630597] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G           O       6.1.46-g247b2535b2 #24
[   94.639103] Hardware name: iW-RainboW-G55M-TI-AM62A7 OSM (DT)
[   94.644833] pstate: 20000005 (nzCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[   94.651780] pc : _raw_spin_unlock_irqrestore+0xc/0x50
[   94.656824] lr : vblank_disable_fn+0x84/0xa0
[   94.661087] sp : ffff800008003e30
[   94.664388] x29: ffff800008003e30 x28: 0000000000000005 x27: 0000000000000020
[   94.671512] x26: ffff80000863d5c0 x25: ffff8000092879c0 x24: ffff00007fc3c1a8
[   94.678636] x23: 0000000000000000 x22: ffff000000d1e880 x21: ffff0000016e3000
[   94.685759] x20: ffff0000016e3144 x19: 0000000000000000 x18: ffff800009413000
[   94.692882] x17: ffff800076b2e000 x16: ffff800008000000 x15: 000000000000003c
[   94.700006] x14: ffffffffffffffff x13: 0000000000000001 x12: 000000000000003c
[   94.707130] x11: 0000000000000901 x10: 0000000000000002 x9 : 0000000000000031
[   94.714254] x8 : 000000000000002b x7 : ffff800009413948 x6 : 0000044dc64ada2d
[   94.721377] x5 : 03ffffffffffffff x4 : ffff0000002901c0 x3 : 0000000000000000
[   94.728501] x2 : 0000000000000000 x1 : 0000000000000000 x0 : ffff0000016e3144
[   94.735624] Call trace:
[   94.738060]  _raw_spin_unlock_irqrestore+0xc/0x50
[   94.742755]  call_timer_fn.constprop.0+0x24/0x80
[   94.747361]  __run_timers.part.0+0x1f4/0x234
[   94.751619]  run_timer_softirq+0x3c/0x7c
[   94.755532]  _stext+0x124/0x2a4
[   94.758663]  ____do_softirq+0x10/0x20
[   94.762316]  call_on_irq_stack+0x24/0x4c
[   94.766229]  do_softirq_own_stack+0x1c/0x30
[   94.770401]  __irq_exit_rcu+0xcc/0xf4
[   94.774053]  irq_exit_rcu+0x10/0x20
[   94.777531]  el1_interrupt+0x38/0x70
[   94.781100]  el1h_64_irq_handler+0x18/0x2c
[   94.785188]  el1h_64_irq+0x64/0x68
[   94.788579]  arch_cpu_idle+0x18/0x2c
[   94.792143]  default_idle_call+0x30/0x6c
[   94.796057]  do_idle+0x248/0x2c0
[   94.799278]  cpu_startup_entry+0x28/0x30
[   94.803192]  kernel_init+0x0/0x130
[   94.806584]  arch_post_acpi_subsys_init+0x0/0x18
[   94.811190]  start_kernel+0x650/0x694
[   94.814842]  __primary_switched+0xbc/0xc4

Regards,
Sushmitha K

Hi Sushmitha,

Is this on AM62A EVM or your custom board?

Can you please describe the details of the reboot test you are doing?

Hi Bin Liu,

The issue is in our custom board.

We are running the below script and running it as a service.

vi /etc/systemd/system/reboot-automate.service

[Unit]
Description=Reboot after boot up

[Service]
ExecStart=/sbin/reboot

[Install]
WantedBy=multi-user.target

systemctl enable reboot-automate.service
systemctl start reboot-automate.service

Regards,
Sushmitha K

Hi Sushmitha,

Thanks for the details. Does your board support any other boot mode, for example boot from SD card? Does the crash issue also happen in other boot mode?

Hi Bin Liu,

We have tried with emmc boot mode and issue is seen

Regards,
Sushmitha K

Okay, let me see if I can reproduce the issue on my board.

Hi Bin Liu,

What is the rootcause for the below error prints. Can you explain?

[    1.438265] xhci-hcd xhci-hcd.0.auto: new USB bus registered, assigned bus number 1
[   16.020848] xhci-hcd xhci-hcd.0.auto: can't setup: -110
[   16.026083] xhci-hcd xhci-hcd.0.auto: USB bus 1 deregistered
[   16.031751] xhci-hcd: probe of xhci-hcd.0.auto failed with error -110

Regards,
Sushmitha K

Hi Shshmitha,

This is due to Errata Advisory i2409.

Hi Sushmitha,

Sushmitha K said:

Usb fail issue got solved but below mentioned crash prints are seen in SW reboot test kept for 12 hours.

Can you please try keeping SW reboot test continuously  for 12 hours in your board and share your observation?

Should the kernel have the USB patch applied to see this rcu stall issue?

I have been running the SDK 9.1 default SD card WIC image without this USB patch, the test has been running for 15+ hour now but the issue doesn't happen.