Other Parts Discussed in Thread: ADS1015, TPS65219
Tool/software:
Hi,
The wakeup pinctrl state is optional for m_can driver. When not configuring the following error appears on poweroff:
```
[ 672.684949] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000018
[ 672.693766] Mem abort info:
[ 672.696551] ESR = 0x0000000096000004
[ 672.700297] EC = 0x25: DABT (current EL), IL = 32 bits
[ 672.705603] SET = 0, FnV = 0
[ 672.708649] EA = 0, S1PTW = 0
[ 672.711786] FSC = 0x04: level 0 translation fault
[ 672.716659] Data abort info:
[ 672.719534] ISV = 0, ISS = 0x00000004, ISS2 = 0x00000000
[ 672.725012] CM = 0, WnR = 0, TnD = 0, TagAccess = 0
[ 672.730056] GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0
[ 672.735362] user pgtable: 4k pages, 48-bit VAs, pgdp=000000008462f000
[ 672.741799] [0000000000000018] pgd=0000000000000000, p4d=0000000000000000
[ 672.748590] Internal error: Oops: 0000000096000004 [#1] PREEMPT SMP
[ 672.754844] Modules linked in: af_alg brcmfmac_cyw irq_pruss_intc pru_rproc brcmfmac brcmutil hci_uart cfg80211 btqca btbcm pwrseq_core bluetooth crct10dif_ce ecdh_generic ecc rfkill rtc_ti_k3 rti_wdt pruss omap_hw
spinlock omap_mailbox sa2ul mcrc64 tps65219_pwrbutton rtc_pcf8563 ti_ads1015 lm75 at24 industrialio_triggered_buffer kfifo_buf ti_eqep counter m_can_platform m_can spi_omap2_mcspi can_dev ramoops pwm_tiehrpwm reed_sol
omon drm drm_panel_orientation_quirks backlight fuse ipv6
[ 672.797744] CPU: 1 UID: 0 PID: 1 Comm: systemd-shutdow Not tainted 6.12.17-ti-00771-gc85877d40f8e-dirty #1
[ 672.807379] Hardware name: SYSTEC sysWORXX CTR-800 (DT)
[ 672.812591] pstate: 60000005 (nZCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[ 672.819539] pc : pinctrl_commit_state+0x68/0x20c
[ 672.824161] lr : pinctrl_commit_state+0x54/0x20c
[ 672.828769] sp : ffff8000822abba0
[ 672.832071] x29: ffff8000822abba0 x28: ffff000000508000 x27: 0000000000000000
[ 672.839199] x26: 0000000000000000 x25: 0000000000000000 x24: ffff8000820ad630
[ 672.846326] x23: ffff000008102140 x22: ffff0000083ee0c0 x21: 0000000000000000
[ 672.853452] x20: ffff000008102158 x19: ffff000008102158 x18: 0000000000000001
[ 672.860579] x17: 0000000000000000 x16: 0000000000000000 x15: ffff000037dba240
[ 672.867705] x14: 0000000000000001 x13: ffff000000508080 x12: 0000000000000001
[ 672.874832] x11: 0000000000000040 x10: ffff000000708860 x9 : ffff000000708858
[ 672.881959] x8 : ffff0000008bcdd8 x7 : 0000000000000000 x6 : 0000000000000000
[ 672.889085] x5 : ffff0000008bcdb0 x4 : 0000000000000000 x3 : 0000000000000000
[ 672.896211] x2 : 0000000000000000 x1 : 0000000000000000 x0 : ffff000000508000
[ 672.903339] Call trace:
[ 672.905776] pinctrl_commit_state+0x68/0x20c
[ 672.910039] pinctrl_select_state+0x1c/0x30
[ 672.914215] m_can_class_sysoff_handler+0x3c/0x78 [m_can]
[ 672.919614] sys_off_notify+0x40/0x6c
[ 672.923272] blocking_notifier_call_chain+0x6c/0xa0
[ 672.928140] kernel_power_off+0x50/0x80
[ 672.931968] __do_sys_reboot+0x210/0x270
[ 672.935882] __arm64_sys_reboot+0x24/0x30
[ 672.939883] invoke_syscall+0x48/0x10c
[ 672.943627] el0_svc_common.constprop.0+0x40/0xe0
[ 672.948323] do_el0_svc+0x1c/0x28
[ 672.951631] el0_svc+0x28/0x98
[ 672.954682] el0t_64_sync_handler+0x120/0x12c
[ 672.959030] el0t_64_sync+0x190/0x194
[ 672.962690] Code: eb14027f 54ffff21 aa1903f5 f90016df (f8418eb4)
[ 672.968768] ---[ end trace 0000000000000000 ]---
[ 672.980619] Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b
[ 672.988292] Kernel Offset: disabled
[ 672.991770] CPU features: 0x00,00000080,00200000,4200420b
[ 672.997158] Memory Limit: none
[ 673.007059] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b ]---
```
The error first appeared after updating to SDK version 11.00.09.04. (with Kernel 6.12)
To me it seems that the following commit introduces this issue:
I think the check of `!IS_ERR(cdev->pinctrl_state_wakeup)` is the cause. Shouldn't this be a simple null-pointer check?
Can you confirm this issue?
Regards
