Re: RCU WARNING on 6.4-rc2, TI AM62 and TPS65219

From: jerome Neanne
Date: Mon May 22 2023 - 08:05:23 EST




On 22/05/2023 11:27, jerome Neanne wrote:


On 19/05/2023 16:26, Francesco Dolcini wrote:
Hello all,
while testing 6.4-rc2 (2d1bcbc6cd70) on a TI K3 AM625 SoC with TPS65219
PMIC I noticed this warning.

[   80.117502] systemd-shutdown[1]: All loop devices detached.
[   80.123176] systemd-shutdown[1]: Stopping MD devices.
[   80.128700] systemd-shutdown[1]: All MD devices stopped.
[   80.134123] systemd-shutdown[1]: Detaching DM devices.
[   80.139553] systemd-shutdown[1]: All DM devices detached.
[   80.144970] systemd-shutdown[1]: All filesystems, swaps, loop devices, MD devices and DM devices detached.
[   80.162682] systemd-shutdown[1]: Syncing filesystems and block devices.
[   80.169602] systemd-shutdown[1]: Rebooting.
[   80.173817] kvm: exiting hardware virtualization
[   80.213016] reboot: Restarting system
[   80.216767] ------------[ cut here ]------------
[   80.221380] Voluntary context switch within RCU read-side critical section!
[   80.221404] WARNING: CPU: 0 PID: 1 at kernel/rcu/tree_plugin.h:318 rcu_note_context_switch+0x31c/0x390
[   80.237669] Modules linked in: 8021q garp mrp stp llc cfg80211 usb_f_ncm u_ether bluetooth ecdh_generic ecc rfkill sp
idev crct10dif_ce snd_soc_simple_card snd_soc_simple_card_utils rtc_ti_k3 sa2ul sha256_generic libsha256 authenc snd_soc
_davinci_mcasp snd_soc_ti_udma snd_soc_ti_edma snd_soc_ti_sdma ti_ads1015 ina2xx industrialio_triggered_buffer pwm_tiehr
pwm snd_soc_nau8822 tps65219_pwrbutton lm75 kfifo_buf spi_omap2_mcspi rtc_ds1307 libcomposite fuse drm ipv6
[   80.278507] CPU: 0 PID: 1 Comm: systemd-shutdow Not tainted 6.4.0-rc2-00166-gf932e7bb873b #3
[   80.286938] Hardware name: Toradex Verdin AM62 on Verdin Development Board (DT)
[   80.294238] pstate: 600000c5 (nZCv daIF -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[   80.301193] pc : rcu_note_context_switch+0x31c/0x390
[   80.306154] lr : rcu_note_context_switch+0x31c/0x390
[   80.311114] sp : ffff80000a71b4c0
[   80.314422] x29: ffff80000a71b4c0 x28: 0000000000000000 x27: ffff000000118000
[   80.321556] x26: 0000000000000000 x25: ffff000000118000 x24: ffff800008fafbd8
[   80.328689] x23: ffff000000118000 x22: 0000000000000000 x21: ffff000000118000
[   80.335822] x20: 0000000000000000 x19: ffff00003fd68cc0 x18: 0000000000000010
[   80.342955] x17: 0000000000000000 x16: 0000000000000000 x15: ffff000000118000
[   80.350087] x14: 00000000000001c5 x13: ffff000000118478 x12: 00000000ffffffea
[   80.357220] x11: 00000000ffffefff x10: 00000000ffffefff x9 : ffff80000a2a9a98
[   80.364352] x8 : 0000000000017fe8 x7 : c0000000ffffefff x6 : 000000000000bff4
[   80.371485] x5 : 0000000000057fa8 x4 : 0000000000000000 x3 : 0000000000000000
[   80.378617] x2 : ffff80000a251990 x1 : 0000000000000000 x0 : 0000000000000000
[   80.385751] Call trace:
[   80.388193]  rcu_note_context_switch+0x31c/0x390
[   80.392807]  __schedule+0x98/0xa7c
[   80.396214]  schedule+0x5c/0xc4
[   80.399354]  schedule_timeout+0x180/0x25c
[   80.403362]  wait_for_completion_timeout+0x80/0x15c
[   80.408238]  ti_sci_set_device_state+0xb4/0x1e4
[   80.412771]  ti_sci_cmd_get_device_exclusive+0x18/0x24
[   80.417907]  ti_sci_pd_power_on+0x28/0x48
[   80.421914]  _genpd_power_on+0x94/0x154
[   80.425749]  genpd_power_on.part.0+0xa4/0x174
[   80.430104]  genpd_runtime_resume+0x118/0x294
[   80.434457]  __rpm_callback+0x48/0x140
[   80.438206]  rpm_callback+0x6c/0x78
[   80.441692]  rpm_resume+0x3bc/0x59c
[   80.445179]  __pm_runtime_resume+0x4c/0x90
[   80.449272]  omap_i2c_xfer_common+0x38/0x598
[   80.453540]  omap_i2c_xfer_polling+0x14/0x20
[   80.457804]  __i2c_transfer+0x138/0x35c
[   80.461642]  i2c_transfer+0x94/0xf4
[   80.465130]  regmap_i2c_read+0x60/0xa8
[   80.468879]  _regmap_raw_read+0xf0/0x170
[   80.472799]  _regmap_bus_read+0x44/0x7c
[   80.476632]  _regmap_read+0x64/0xf4
[   80.480118]  _regmap_update_bits+0xf4/0x130
[   80.484298]  regmap_update_bits_base+0x64/0x98
[   80.488738]  tps65219_restart+0x38/0x48
[   80.492576]  atomic_notifier_call_chain+0x60/0x90
[   80.497280]  do_kernel_restart+0x24/0x30
[   80.501202]  machine_restart+0x38/0x5c
[   80.504950]  kernel_restart+0x88/0x98
[   80.508612]  __do_sys_reboot+0x1e0/0x264
[   80.512533]  __arm64_sys_reboot+0x24/0x30
[   80.516538]  invoke_syscall+0x44/0x104
[   80.520287]  el0_svc_common.constprop.0+0x44/0xec
[   80.524988]  do_el0_svc+0x38/0x98
[   80.528302]  el0_svc+0x2c/0x84
[   80.531354]  el0t_64_sync_handler+0xb8/0xbc
[   80.535534]  el0t_64_sync+0x190/0x194
[   80.539192] ---[ end trace 0000000000000000 ]---


This looks similar to what is described here [1], same issue or
something else? Any suggestion?

The issue is systematic and happens at every boot. With a TI downstream
5.10 kernel this issue was never experienced.

Francesco

[1] https://lore.kernel.org/all/20230327-tegra-pmic-reboot-v6-0-af44a4cd82e9@xxxxxxxxxxx/

I don't have the same board to test and don't experience this boot issue with the am62 board I'm using (the board version I have is a custom board not released publicly by TI unfortunately).

Notice that: https://lore.kernel.org/lkml/20230511122100.2225417-1-jneanne@xxxxxxxxxxxx/

Is not yet applied and would help handle shutdown/reboot under some conditions.

This might help for reboot handling if you are forcing a shutdown/reboot in your test.

But if the problem is your device tries to shutdown at each boot, then this is just a side effect, not the original root cause.

Regards,
Jerome
More background in thread:
Link: https://lore.kernel.org/all/7hfseqa7l0.fsf@xxxxxxxxxxxx/

This explains why downstream is different on that point.