stm32h7 qspi NOR flash with XiP

[FRASTM]

Following the #68607 I am trying to run the samples/application_development/code_relocation_nocopy/ on the stm32h750b disco kit.
My branch https://github.com/FRASTM/zephyr/tree/xip_h7

This target board has a Dual quad flash connected to the dual quad-spi bus at memory address 0x90000000.
The stm32qspi driver has a problem addressing the dual quad:
if configured in dual Flash Mode,

	dev_data->hqspi.Init.SampleShifting = QSPI_SAMPLE_SHIFTING_HALFCYCLE;
	dev_data->hqspi.Init.ChipSelectHighTime = QSPI_CS_HIGH_TIME_3_CYCLE;
	dev_data->hqspi.Init.DualFlash = QSPI_DUALFLASH_ENABLE;
	/* Set Dual Flash Mode only on MemoryMapped */
	dev_data->hqspi.Init.FlashID = QSPI_FLASH_ID_2;

	HAL_QSPI_Init(&dev_data->hqspi);

then the magic and SFDP is not recognized probably because it is mixing data received from both flash devices:

[00:00:00.000,000] <inf> flash_stm32_qspi: Reading SFDP
[00:00:00.000,000] <err> flash_stm32_qspi: SFDP magic 46465353 invalid

Then, just for test, skipping the SFDP, the memory examination at external flash address 0x90000000 is read correctly after setting the external flash in MemMapped mode. But a MPU fault occurs :

*** Booting Zephyr OS build zephyr-v3.5.0-5289-g043eeddf5cd2 ***
Address of main function 0x800066d
[00:00:00.000,000] <err> os: ***** MPU FAULT *****
[00:00:00.000,000] <err> os:   Instruction Access Violation
[00:00:00.000,000] <err> os: r0/a1:  0xb5084807  r1/a2:  0x49074a06  r2/a3:  0x90000001
[00:00:00.000,000] <err> os: r3/a4:  0x00000000 r12/ip:  0x00000000 r14/lr:  0x9000000d
[00:00:00.000,000] <err> os:  xpsr:  0x61000000
[00:00:00.000,000] <err> os: Faulting instruction address (r15/pc): 0xb5084806
[00:00:00.000,000] <err> os: >>> ZEPHYR FATAL ERROR 20: Unknown error on CPU 0
[00:00:00.000,000] <err> os: Current thread: 0x240007f0 (unknown)
[00:00:00.065,000] <err> os: Halting system

Note that this is more or less the same situation with the stm32h747i_disco_m7

[FRASTM]

[FRASTM]

code of the external NOR flash can be XiP on the stm32h750b disco but finally crashes
the function_in_ext_flash just print but the b.l 0x90000030 instruction does not seem to be correctly executed

[FRASTM]

building/linking/flahing the samples/application_development/code_relocation_nocopy on a disco stm32h750 (dual qspi NOR flash)
The code being executed (XiP) in the external NOR by the M7 of the stm32h750, fails on instruction bl 0x90000030 <__printk_veneer>
This instruction is ignored as well as the next b.w instruction

Error: Failed to read memory at 0x8fffffc0
   0x90000004 <function_in_ext_flash+4>:	ldr	r0, [pc, #28]	; (0x90000024 <function_in_ext_flash+36>)
   0x90000006 <function_in_ext_flash+6>:	push	{r3, lr}
=> 0x90000008 <function_in_ext_flash+8>:	bl	0x90000030 <__printk_veneer>
   0x9000000c <function_in_ext_flash+12>:	ldr	r1, [pc, #24]	; (0x90000028 <function_in_ext_flash+40>)
   0x9000000e <function_in_ext_flash+14>:	ldr	r0, [pc, #28]	; (0x9000002c <function_in_ext_flash+44>)
   0x90000010 <function_in_ext_flash+16>:	ldr	r2, [r1, #0]
   0x90000012 <function_in_ext_flash+18>:	ldmia.w	sp!, {r3, lr}
   0x90000016 <function_in_ext_flash+22>:	b.w	0x90000030 <__printk_veneer>

on instruction 0x90000008 <function_in_ext_flash+8>: bl 0x90000030
pc goes to $pc = 0x9000000a without branching on 0x90000030

   0x90000030 <__printk_veneer>:	ldr.w	pc, [pc]	; 0x90000034 <__printk_veneer+4>
   0x90000034 <__printk_veneer+4>:	str	r5, [r2, #116]	; 0x74
   0x90000036 <__printk_veneer+6>:	lsrs	r0, r0, #32