IPCF v4.8 multi-instance affects S32G board functions(RDB2 bootup fail and RDB3 panic)

[hittzt]

Hi,

I recently test IPCF v4.8 multi-instance demos, but all failed on RDB2 and RDB3 boards.

The multi-instance firmwares are built following the release note steps.

The RDB2 test log is attached.

And the panic info on RDB3 as following:

root@nxp-s32g:~# echo 10 > /sys/kernel/ipc-shm-sample-instance0/ping
ipc-shm-sample_multi-instance: starting demo on instance 0...
ipc-shm-sample_multi-instance: INST0 ch 0 >> 20 bytes: SENDING MESSAGES: 10
ipc-shm-sample_multi-instance: INST0 ch 1 >> 32 bytes: #0 HELLO WORLD! from KERNEL
ipc-shm-sample_multi-instance: ch 1 << 32 bytes: #0 HELLO WORLD! from CORE 4
ipc-shm-sample_multi-instance: INST0 ch 2 >> 32 bytes: #1 HELLO WORLD! from KERNEL
ipc-shm-sample_multi-instance: ch 2 << 32 bytes: #1 HELLO WORLD! from CORE 4
ipc-shm-sample_multi-instance: INST0 ch 1 >> 32 bytes: #2 HELLO WORLD! from KERNEL
ipc-shm-sample_multi-instance: ch 1 << 32 bytes: #2 HELLO WORLD! from CORE 4
ipc-shm-sample_multi-instance: INST0 ch 2 >> 32 bytes: #3 HELLO WORLD! from KERNEL
ipc-shm-sample_multi-instance: ch 2 << 32 bytes: #3 HELLO WORLD! from CORE 4
ipc-shm-sample_multi-instance: INST0 ch 1 >> 32 bytes: #4 HELLO WORLD! from KERNEL
ipc-shm-sample_multi-instance: ch 1 << 32 bytes: #4 HELLO WORLD! from CORE 4
ipc-shm-sample_multi-instance: INST0 ch 2 >> 32 bytes: #5 HELLO WORLD! from KERNEL
ipc-shm-sample_multi-instance: ch 2 << 32 bytes: #5 HELLO WORLD! from CORE 4
ipc-shm-sample_multi-instance: INST0 ch 1 >> 32 bytes: #6 HELLO WORLD! from KERNEL
ipc-shm-sample_multi-instance: ch 1 << 32 bytes: #6 HELLO WORLD! from CORE 4
ipc-shm-sample_multi-instance: INST0 ch 2 >> 32 bytes: #7 HELLO WORLD! from KERNEL
ipc-shm-sample_multi-instance: ch 2 << 32 bytes: #7 HELLO WORLD! from CORE 4
ipc-shm-sample_multi-instance: INST0 ch 1 >> 32 bytes: #8 HELLO WORLD! from KERNEL
ipc-shm-sample_multi-instance: ch 1 << 32 bytes: #8 HELLO WORLD! from CORE 4
ipc-shm-sample_multi-instance: INST0 ch 2 >> 32 bytes: #9 HELLO WORLD! from KERNEL
ipc-shm-sample_multi-instance: ch 0 << 20 bytes: REPLIED MESSAGES: 10
ipc-shm-sample_multi-instance: ch 2 << 32 bytes: #9 HELLO WORLD! from CORE 4
ipc-shm-sample_multi-instance: exit demo for instance 0
root@nxp-s32g:~#
root@nxp-s32g:~#
root@nxp-s32g:~# echo 10 > /sys/kernel/ipc-shm-sample-instance1/ping
ipc-shm-sample_multi-instance: starting demo on instance 1...
ipc-shm-sample_multi-instance: INST1 ch 0 >> 20 bytes: SENDING MESSAGES: 10
ipc-shm-sample_multi-instance: INST1 ch 1 >> 32 bytes: #0 HELLO WORLD! from KERNEL
Kernel panic - not syncing: FIQ taken without a root FIQ handler
CPU: 0 PID: 0 Comm: swapper/0 Tainted: G O 6.1.35-yocto-standard #1
Hardware name: NXP S32G399A-RDB3 (DT)
Call trace:
dump_backtrace.part.0+0xc8/0xd4
show_stack+0x20/0x30
dump_stack_lvl+0x6c/0x88
dump_stack+0x18/0x34
panic+0x17c/0x338
__show_regs+0x0/0x354
call_on_irq_stack+0x24/0x54
do_interrupt_handler+0x88/0x90
el1_interrupt+0x34/0x6c
el1h_64_fiq_handler+0x18/0x24
el1h_64_fiq+0x64/0x68
arch_cpu_idle+0x18/0x2c
default_idle_call+0x74/0x190
do_idle+0x214/0x280
cpu_startup_entry+0x30/0x34
rest_init+0xf8/0x100
arch_post_acpi_subsys_init+0x0/0x28
start_kernel+0x6d4/0x714
__primary_switched+0xb4/0xbc
SMP: stopping secondary CPUs
Kernel Offset: disabled
CPU features: 0x40000,00800084,0000420b
Memory Limit: none
---[ end Kernel panic - not syncing: FIQ taken without a root FIQ handler ]---
Kernel panic - not syncing: FIQ taken without a root FIQ handler
Kernel Offset: disabled
CPU features: 0x40000,00800084,0000420b
Memory Limit: none
---[ end Kernel panic - not syncing: FIQ taken without a root FIQ handler ]---
Kernel panic - not syncing: FIQ taken without a root FIQ handler
Kernel Offset: disabled
CPU features: 0x40000,00800084,0000420b
Memory Limit: none
---[ end Kernel panic - not syncing: FIQ taken without a root FIQ handler ]---
Kernel panic - not syncing: FIQ taken without a root FIQ handler
Kernel Offset: disabled
CPU features: 0x40000,00800084,0000420b
Memory Limit: none
---[ end Kernel panic - not syncing: FIQ taken without a root FIQ handler ]---
Kernel panic - not syncing: FIQ taken without a root FIQ handler
Kernel Offset: disabled
CPU features: 0x40000,00800084,0000420b
Memory Limit: none
---[ end Kernel panic - not syncing: FIQ taken without a root FIQ handler ]---
Kernel panic - not syncing: FIQ taken without a root FIQ handler
Kernel Offset: disabled
CPU features: 0x40000,00800084,0000420b
Memory Limit: none
---[ end Kernel panic - not syncing: FIQ taken without a root FIQ handler ]---
Kernel panic - not syncing: FIQ taken without a root FIQ handler
Kernel Offset: disabled
CPU features: 0x40000,00800084,0000420b
Memory Limit: none
---[ end Kernel panic - not syncing: FIQ taken without a root FIQ handler ]---
Kernel panic - not syncing: FIQ taken without a root FIQ handler
Kernel Offset: disabled
CPU features: 0x40000,00800084,0000420b
Memory Limit: none
---[ end Kernel panic - not syncing: FIQ taken without a root FIQ handler ]---

Would you please help to check the issues above and give a fix or workaround?

Thanks,

Zhantao

[Allen_Cheng]

Hi,

I have met the same issue in RDB2.

Is the issue resolve or any workaround?

Thanks.

Allen

[Daniel-Aguirre]

Hi,

At this moment, BSP36.0 seems to have a bug on this regard. The multi-instance example does not work under this version. The recommendation from our internal team is to use BSP35.0 in the meantime.

This only applies to multi-instance, since single-instance seems to be working. We do apologize.

Please, let us know.

[Shenghao]

Hello Daniel,

Just want to confirm what's the bug of BSP36+IPCF 4.8.0.

Because we use this software combination and encoutner the IPCF forwarding problem.

Thanks.

[hittzt]

Hi,

OK, got it.

Thanks for your reply.

I notice that, SDK BSP37 has been released, is this issue fixed in BSP37 release?

Thanks,

Zhantao

[Daniel-Aguirre]

Hi,

From what the internal team has told us, BSP37 seems to still have this bug. We do apologize.

Please, let us know.

[bgarcia]

Hi Daniel

We are now at BSP39 (https://github.com/nxp-auto-linux/linux/tree/release/bsp39.0-5.15.129-rt), do you think this issue will be solved in that BSP?

Thanks