Can not run correctly on NXP MIMXRT1061 CVL5A.

[beiszhihao]

Describe the bug
I am currently trying to run Zephyr7.0 on a development board equipped with MCU: NXP MIMXRT1061 CVL5A.
I simply compiled a sample: \samples\basic\blinky, but it couldn't run correctly.
At first I thought it was a problem with the XIP format that caused Zephyr to not be booted correctly, but then I used SWD to debug it and found that it was booted correctly.
But when calling: /zephyr/arch/arm/core/aarch32/prep_c.c: z_bss_zero(); function Zephyr went wrong

Compilation process
The board I use is mimxrt1060_evk, theoretically there is no problem, because 1061 is based on 1060

1. west build -p auto -b mimxrt1060_evk .\samples\basic\blinky\
2. west flash

Debugging process
The first debugging was broken at z_interrupt_stacks

z_arm_reset () at C:/Users/zhihao3x/work/zephyrproject/zephyr/arch/arm/core/aarch32/cortex_m\reset.S:105
105         msr BASEPRI, r0
(gdb) n
134         ldr r0, =z_interrupt_stacks
(gdb) n
135         ldr r1, =CONFIG_ISR_STACK_SIZE + MPU_GUARD_ALIGN_AND_SIZE
(gdb) n
136         adds r0, r0, r1
(gdb) n
137         msr PSP, r0
(gdb)
138         mrs r0, CONTROL
(gdb)
139         movs r1, #2
(gdb)
140         orrs r0, r1 /* CONTROL_SPSEL_Msk */
(gdb)
141         msr CONTROL, r0
(gdb)
147         isb
(gdb)
154         bl z_arm_prep_c
(gdb)
134         ldr r0, =z_interrupt_stacks
(gdb)

Program received signal SIGTRAP, Trace/breakpoint trap.
(gdb)

Later, I used single-step debugging and found that Zephyr reached z_bss_zero

(gdb)
z_arm_floating_point_init ()
    at C:/Users/zhihao3x/work/zephyrproject/modules/hal/cmsis/CMSIS/Core/Include/cmsis_gcc.h:1003
1003      __ASM volatile ("MSR control, %0" : : "r" (control) : "memory");
(gdb)
163             __set_CONTROL(__get_CONTROL() & (~(CONTROL_FPCA_Msk)));
(gdb)
0x6000305c in __set_CONTROL (control=3758157056)
    at C:/Users/zhihao3x/work/zephyrproject/modules/hal/cmsis/CMSIS/Core/Include/cmsis_gcc.h:1003
1003      __ASM volatile ("MSR control, %0" : : "r" (control) : "memory");
(gdb)
1004      __ISB();
(gdb)
__ISB () at C:/Users/zhihao3x/work/zephyrproject/modules/hal/cmsis/CMSIS/Core/Include/cmsis_gcc.h:260
260       __ASM volatile ("isb 0xF":::"memory");
(gdb)
z_arm_prep_c () at C:/Users/zhihao3x/work/zephyrproject/zephyr/arch/arm/core/aarch32/prep_c.c:183
183             z_bss_zero();
(gdb) n

Program received signal SIGTRAP, Trace/breakpoint trap.
0x62652dc0 in ?? ()

Eventually I located the problem in the memset function，when this function is called, it will fall into an infinite loop and then my program will crash

z_arm_prep_c () at C:/Users/zhihao3x/work/zephyrproject/zephyr/arch/arm/core/aarch32/prep_c.c:183
183             z_bss_zero();
(gdb)
z_bss_zero () at C:/Users/zhihao3x/work/zephyrproject/zephyr/kernel/init.c:89
89              (void)memset(__bss_start, 0, __bss_end - __bss_start);
(gdb)
memset (buf=0x80000030 <z_idle_threads>, c=c@entry=0, n=428)
    at C:/Users/zhihao3x/work/zephyrproject/zephyr/lib/libc/minimal/source/string/string.c:355

(gdb)

Program received signal SIGTRAP, Trace/breakpoint trap.
0x63f5fbf8 in ?? ()

Later I tried to debug the memset function and found that the dest address did not change during the loop
Very strange, I do not specify whether it is a gdb problem or a zephyr problem. The address is 0x80000030 before the increment, and 0x80000031 after the increment, and it changes back to 0x80000030 when the next cycle is repeated.

357             unsigned char c_byte = (unsigned char)c;
(gdb)
389             while (n > 0) {
(gdb)
390                     *(d_byte++) = c_byte;
392                     n--;
(gdb) p d_byte
$1 = (unsigned char *) 0x80000030 <z_idle_threads> "\b"

I tried to change the z_bss_zero function

(void)memset(__bss_start, 0, __bss_end - __bss_start);

Changed the size length

(void)memset(__bss_start, 0, 3);

But when I was debugging with gdb, I found that the while loop inside memset exceeded three times without stopping, and it entered an endless loop.
I speculate that there is a problem with the incremented code after these two lines of code, because when I use the gdb print command to print the value of the variable after each increment, the value of the variable does not change

*(d_byte++) = c_byte;
		n--;

Another weird place is that when I use gdb to debug, I find that there is a pointer. Gdb will not break to this side, but skip this code and execute it. I doubt that the compiler is doing it to me. Is the code optimized?

More details
When I use next to execute z_bss_zero, gdb will break to an indeterminate address at 0xdeadbeee. I guess it may be that Zephyr destroyed the stack when initializing the memory.

(gdb)
z_arm_prep_c () at C:/Users/zhihao3x/work/zephyrproject/zephyr/arch/arm/core/aarch32/prep_c.c:183
183             z_bss_zero();
(gdb) n

Program received signal SIGTRAP, Trace/breakpoint trap.
0xdeadbeee in ?? ()

At the same time, the Jlink debugger also output an error log

ERROR: Cannot read register 15 (R15) while CPU is running
Reading all registers
ERROR: Cannot read register 0 (R0) while CPU is running
ERROR: Cannot read register 1 (R1) while CPU is running
ERROR: Cannot read register 2 (R2) while CPU is running
ERROR: Cannot read register 3 (R3) while CPU is running
ERROR: Cannot read register 4 (R4) while CPU is running
ERROR: Cannot read register 5 (R5) while CPU is running
ERROR: Cannot read register 6 (R6) while CPU is running
ERROR: Cannot read register 7 (R7) while CPU is running
ERROR: Cannot read register 8 (R8) while CPU is running
ERROR: Cannot read register 9 (R9) while CPU is running
ERROR: Cannot read register 10 (R10) while CPU is running
ERROR: Cannot read register 11 (R11) while CPU is running
ERROR: Cannot read register 12 (R12) while CPU is running
ERROR: Cannot read register 13 (R13) while CPU is running
ERROR: Cannot read register 14 (R14) while CPU is running
ERROR: Cannot read register 15 (R15) while CPU is running
ERROR: Cannot read register 16 (XPSR) while CPU is running
ERROR: Cannot read register 17 (MSP) while CPU is running
ERROR: Cannot read register 18 (PSP) while CPU is running
ERROR: Cannot read register 24 (PRIMASK) while CPU is running
ERROR: Cannot read register 25 (BASEPRI) while CPU is running
ERROR: Cannot read register 26 (FAULTMASK) while CPU is running
ERROR: Cannot read register 27 (CONTROL) while CPU is running
ERROR: Cannot read register 32 (FPSCR) while CPU is running
ERROR: Cannot read register 33 (FPS0) while CPU is running
ERROR: Cannot read register 34 (FPS1) while CPU is running
ERROR: Cannot read register 35 (FPS2) while CPU is running
ERROR: Cannot read register 36 (FPS3) while CPU is running
ERROR: Cannot read register 37 (FPS4) while CPU is running
ERROR: Cannot read register 38 (FPS5) while CPU is running
ERROR: Cannot read register 39 (FPS6) while CPU is running
ERROR: Cannot read register 40 (FPS7) while CPU is running
ERROR: Cannot read register 41 (FPS8) while CPU is running
ERROR: Cannot read register 42 (FPS9) while CPU is running
ERROR: Cannot read register 43 (FPS10) while CPU is running
ERROR: Cannot read register 44 (FPS11) while CPU is running
ERROR: Cannot read register 45 (FPS12) while CPU is running
ERROR: Cannot read register 46 (FPS13) while CPU is running
ERROR: Cannot read register 47 (FPS14) while CPU is running
ERROR: Cannot read register 48 (FPS15) while CPU is running
ERROR: Cannot read register 49 (FPS16) while CPU is running
ERROR: Cannot read register 50 (FPS17) while CPU is running
ERROR: Cannot read register 51 (FPS18) while CPU is running
ERROR: Cannot read register 52 (FPS19) while CPU is running
ERROR: Cannot read register 53 (FPS20) while CPU is running
ERROR: Cannot read register 54 (FPS21) while CPU is running
ERROR: Cannot read register 55 (FPS22) while CPU is running
ERROR: Cannot read register 56 (FPS23) while CPU is running
ERROR: Cannot read register 57 (FPS24) while CPU is running
ERROR: Cannot read register 58 (FPS25) while CPU is running
ERROR: Cannot read register 59 (FPS26) while CPU is running
ERROR: Cannot read register 60 (FPS27) while CPU is running
ERROR: Cannot read register 61 (FPS28) while CPU is running
ERROR: Cannot read register 62 (FPS29) while CPU is running
ERROR: Cannot read register 63 (FPS30) while CPU is running
ERROR: Cannot read register 64 (FPS31) while CPU is running
ERROR: Cannot read register 33 (FPS0) while CPU is running
ERROR: Cannot read register 34 (FPS1) while CPU is running
ERROR: Cannot read register 35 (FPS2) while CPU is running
ERROR: Cannot read register 36 (FPS3) while CPU is running
ERROR: Cannot read register 37 (FPS4) while CPU is running
ERROR: Cannot read register 38 (FPS5) while CPU is running
ERROR: Cannot read register 39 (FPS6) while CPU is running
ERROR: Cannot read register 40 (FPS7) while CPU is running
ERROR: Cannot read register 41 (FPS8) while CPU is running
ERROR: Cannot read register 42 (FPS9) while CPU is running
ERROR: Cannot read register 43 (FPS10) while CPU is running
ERROR: Cannot read register 44 (FPS11) while CPU is running
ERROR: Cannot read register 45 (FPS12) while CPU is running
ERROR: Cannot read register 46 (FPS13) while CPU is running
ERROR: Cannot read register 47 (FPS14) while CPU is running
ERROR: Cannot read register 48 (FPS15) while CPU is running
ERROR: Cannot read register 49 (FPS16) while CPU is running
ERROR: Cannot read register 50 (FPS17) while CPU is running
ERROR: Cannot read register 51 (FPS18) while CPU is running
ERROR: Cannot read register 52 (FPS19) while CPU is running
ERROR: Cannot read register 53 (FPS20) while CPU is running
ERROR: Cannot read register 54 (FPS21) while CPU is running
ERROR: Cannot read register 55 (FPS22) while CPU is running
ERROR: Cannot read register 56 (FPS23) while CPU is running
ERROR: Cannot read register 57 (FPS24) while CPU is running
ERROR: Cannot read register 58 (FPS25) while CPU is running
ERROR: Cannot read register 59 (FPS26) while CPU is running
ERROR: Cannot read register 60 (FPS27) while CPU is running
ERROR: Cannot read register 61 (FPS28) while CPU is running
ERROR: Cannot read register 62 (FPS29) while CPU is running
ERROR: Cannot read register 63 (FPS30) while CPU is running
ERROR: Cannot read register 64 (FPS31) while CPU is running
Removing breakpoint @ address 0x60003068, Size = 2
WARNING: Failed to read memory @ address 0xDEADBEEE

Jlink output
The following is the MCU information output by Jlink, I am not sure if this version is supported by Zephyr

-----GDB Server start settings-----
GDBInit file:                  none
GDB Server Listening port:     2331
SWO raw output listening port: 2332
Terminal I/O port:             2333
Accept remote connection:      localhost only
Generate logfile:              off
Verify download:               off
Init regs on start:            off
Silent mode:                   on
Single run mode:               on
Target connection timeout:     5000 ms
------J-Link related settings------
J-Link Host interface:         USB
J-Link script:                 none
J-Link settings file:          none
------Target related settings------
Target device:                 MIMXRT1062xxx6A
Target interface:              SWD
Target interface speed:        auto
Target endian:                 little

[dleach02]

@beiszhihao this sample is working on my MIMXRT1060-EVK board (Rev A2). Is your board a custom board?

When using 'west debug' on your target, try running 'monitor reset' at the beginning of your debug session. I'm still investigating but I've found some boards require me to do the 'monitor reset' when starting a debugging session.

[beiszhihao]

@beiszhihao this sample is working on my MIMXRT1060-EVK board (Rev A2). Is your board a custom board?

When using 'west debug' on your target, try running 'monitor reset' at the beginning of your debug session. I'm still investigating but I've found some boards require me to do the 'monitor reset' when starting a debugging session.

Thank you very much for your help. This problem has been solved. The sdram of my board is off-chip, not on-chip, so I modified the address