Hello,
We created a custom board around nrf7002-DK (which includes nrf5340 and nrf7002 for wifi). We are using TF-M (including a custom secure partition), BT (child image) and WIFI.
The problem with flash space raised and we started to lower the partitions for everything since not all of them are correctly optimized.
For example, in TF-M case we reduced 50% of it by using the config entry CONFIG_PM_PARTITION_SIZE_TFM and everything is fine here.
The problem is with the BLE child image. Since there is no config entry for this, from what I researched in documentation, there is need for a static configuration file for partition manager for child image, which I created under the board directory with the name pm_static_<custom_name>_nrf5340_cpunet.yml and I see in the build flow that this file is taken into account. The problem is that the generated partition yml that I see in the build/hci_ipc/partitions_CPUNET.yml is wrong and the end result is of course wrong.
The generated yml takes into account the lowered app entry from 0x40000 to 0x20000 but the hci_ipc entry is still 0x40000. And what about the regions ? How we change that? If I put those also in the same paritition manager static config file the script cannot parse them.
The defined static partition manager config file for child image:
app: address: 0x1000000 end_address: 0x1020000 orig_span: &id001 - hci_ipc region: flash_primary size: 0x20000 span: *id001 hci_ipc: address: 0x1000000 end_address: 0x1020000 region: flash_primary size: 0x20000 sram_primary: address: 0x21000000 end_address: 0x21010000 region: sram_primary size: 0x10000
The generated partitions_CPUNET.yml
app: address: 0x1000000 end_address: 0x1020000 orig_span: &id001 - hci_ipc region: flash_primary size: 0x20000 span: *id001 hci_ipc: address: 0x1000000 end_address: 0x1040000 region: flash_primary size: 0x40000 sram_primary: address: 0x21000000 end_address: 0x21010000 region: sram_primary size: 0x10000
