Hi,
I’m working with the Wi-Fi station example on the nRF54H20 using SDK v3.0.0, and I’m encountering a critical issue when trying to bring up the nRF7002EK Wi-Fi module.
The system crashes with an MPU fault right after initializing the SPI interface. Additionally, I do not observe any SPI activity on the logic analyzer, which suggests that communication never starts.
Here's the log:
[00:00:00.157,964] <inf> wifi_nrf_bus: SPIM spi@8e7000: freq = 8 MHz [00:00:00.157,987] <inf> wifi_nrf_bus: SPIM spi@8e7000: latency = 0 [00:00:00.158,024] <err> os: ***** MPU FAULT ***** [00:00:00.158,028] <err> os: Data Access Violation [00:00:00.158,032] <err> os: MMFAR Address: 0xe10463c [00:00:00.158,042] <err> os: r0/a1: 0x0e10463c r1/a2: 0x2f0467a5 r2/a3: 0x2f0467a6 [00:00:00.158,048] <err> os: r3/a4: 0x0e10463b r12/ip: 0x00000028 r14/lr: 0x0e09df6d [00:00:00.158,051] <err> os: xpsr: 0x89000000 [00:00:00.158,054] <err> os: Faulting instruction address (r15/pc): 0x0e0daad0 [00:00:00.158,067] <err> os: >>> ZEPHYR FATAL ERROR 19: Unknown error on CPU 0 [00:00:00.158,090] <err> os: Current thread: 0x2f0141c8 (unknown) [00:00:00.234,037] <err> coredump: #CD:BEGIN# [00:00:00.239,102] <err> coredump: #CD:5a4502000300050013000000 [00:00:00.245,750] <err> coredump: #CD:4102004400 [00:00:00.251,122] <err> coredump: #CD:3c46100ea567042fa667042f3b46100e280000006ddf090ed0aa0d0e00000089 [00:00:00.261,229] <err> coredump: #CD:1067042f00000000000000000000000000000000000000000000000000000000 [00:00:00.271,328] <err> coredump: #CD:00000000 [00:00:00.276,589] <err> coredump: #CD:4d0100c841012f7042012f [00:00:00.283,009] <err> coredump: #CD:e026012fac70012f000000000180000000000000000000000000000000000000 [00:00:00.293,113] <err> coredump: #CD:cf780d0e0000000000000000000000000000000000000000451300009470012f [00:00:00.303,217] <err> coredump: #CD:00000000243b012fffffffffffffffff7867042f000000002042012f2042012f [00:00:00.313,322] <err> coredump: #CD:0000000000000000000000000000000000000000000000008042042f10270000 [00:00:00.323,426] <err> coredump: #CD:100000004820012f8069042f0000000000000000000000000000000000000000 [00:00:00.333,526] <err> coredump: #CD:00000000f5ffffff [00:00:00.339,453] <err> coredump: #CD:4d01008042042f9069042f ...
I based the new code on an example that worked on the previous SDK version (v2.9.0-nRF54H20-1). In that version, it didn’t trigger a fault or crash. This suggests that the crash may be related to changes in SDK v3.0.0, rather than in the board configuration or hardware setup.
Could you help me understand what might be causing the MPU fault during SPI initialization?
Also, what other tests would you recommend to further isolate the issue? Would you suggest any changes in the source code or configuration to help debug or work around the problem?
I’ll attach the code for your reference.
Thanks in advance for your support.
Best,
Marco