A typical scenario is: when XIP on PSRAM enabled, compiler optimization level is Os. Under certain binary layout, boot hangs and backtrace points to `esp_sleep_config_gpio_isolate`.
The root cause is that, during PSRAM initialization, it calls esp_gpio_reserve, which happens to place before the reported function. However, after call, there is no barrier before the clock adjustment in `mspi_timing_enter_low_speed_mode`. The clock gets changed when the cache is still fetching data, resulting in the corrupted data in the end of the cache line.
This commits add spi_flash_disable_cache as a barrier to make sure the cache transactions is finished before the clock switch.
The TLSF ROM patch code (rom_patch_tlsf.c, rom_patch_multi_heap.c),
their headers, and the per-target *.rom.heap.ld linker scripts are
semantically owned by the heap component, not esp_rom. Move them
accordingly:
- components/esp_rom/patches/esp_rom_tlsf.c
→ components/heap/rom_patches/rom_patch_tlsf.c
- components/esp_rom/patches/esp_rom_multi_heap.c
→ components/heap/rom_patches/rom_patch_multi_heap.c
- components/esp_rom/include/esp_rom_tlsf.h
→ components/heap/rom_patches/include/rom_patch_tlsf.h
- components/esp_rom/include/esp_rom_multi_heap.h
→ components/heap/rom_patches/include/rom_patch_multi_heap.h
- components/esp_rom/<target>/ld/<target>.rom.heap.ld (×8)
→ components/heap/port/<target>/ld/<target>.rom.heap.ld
Update heap/CMakeLists.txt to:
- use target_linker_script() directly for *.rom.heap.ld
- guard post-registration ROM patch setup with NOT BOOTLOADER_BUILD
- drop stale CONFIG_HEAP_TLSF_CHECK_PATCH symbol reference
- drop esp_rom_include_multi_heap_patch from TLSF_CHECK_PATCH guard
Remove the corresponding entries from esp_rom/CMakeLists.txt.
- Fix intermittent TEE stack underflow test failures
- Fix out-of-bounds access Coverity report from the attestation
component
- Add appropriate checks and asserts for TEE flash memory regions'
sizes
- Also fixed an issue where NewLib ROM APIs, when called from TEE, were
using the syscall table located in the REE SRAM. This could be abused
as an attack vector to invoke illegal functions from the TEE.
To prevent this, the syscall table is now switched to the TEE-specific
copy during every M-U mode transition.
Currently, REE SPI flash HAL operations are routed as service calls to TEE,
but the TEE implementation incorrectly uses ROM APIs instead of HAL APIs.
This leads to issues and is not the recommended approach.
- Remove GPTIMER and TIMG related definitions from soc_caps_full.h files
- Move timer peripheral definitions to appropriate HAL layer files
- Update references across components to use proper HAL abstractions
- Consolidate timer group and GPTIMER capabilities organization
- Ensure consistent timer configuration across all ESP32 variants
This refactoring improves the separation of concerns between SOC
capabilities and HAL implementations for timer-related functionality.
- Place the APM HAL into TEE IRAM when `CONFIG_SECURE_TEE_EXT_FLASH_MEMPROT_SPI1`
is enabled, as APM violations related to SPI1 can occur with the flash cache disabled.
- Also fix an issue where flash protection tests were passing due to incorrect checks
- Decreased from 32KB to 24KB, keeping in mind the current maximum TEE heap
usage and some overhead
- Make the TEE panic handler logs concise, saving some DRAM
Xiao Xufeng
Guillaume Souchere
Laukik Hase
Ashish Sharma
morris