Greentea fails for GCC_ARM in TMPM066

[ghost]

Description

• Type: Bug
• Priority: Major | Minor

---

Bug

Target
Toshiba TMPM066

Toolchain:
GCC_ARM

Expected behavior
Greentea testsuites (features-tests-filesystem-mbr_block_device & tests-mbedmicro-rtos-mbed-threads) should pass without change in main stack

Actual behavior
Greentea testsuites (features-tests-filesystem-mbr_block_device & tests-mbedmicro-rtos-mbed-threads) fails due to not enough memory (Operator new[] out of memory).

Steps to reproduce

1. Download master from https://github.com/ganesh-ramachandran/mbed-os (since Toshiba TMPM066 is yet to be merged)
2. mbed test -t GCC_ARM -m TMPM066 -v -n features-tests-filesystem-mbr_block_device,tests-mbedmicro-rtos-mbed-threads

---

Details

Not enough memory is the message from the log - "Operator new[] out of memory" - Line numbers 191 & 289 from the attached log "M066_GCC_ARM_GreenTea_MainStack_4K.txt". Did not face this memory issue with ARM and IAR.

M066 Memory : RAM - 16K and Flash - 128K.

If main stack memory is reduced to 2K for GCC_ARM in mbed_rtx.h, then Greentea tests are passed (16af199).

Attached below the GCC_ARM logs with 4K and 2K main stack size.
M066_GCC_ARM_GreenTea_MainStack_4K.txt
M066_GCC_ARM_GreenTea_MainStack_2K.txt

Related Discussion are in #4840

[0xc0170]

cc @bulislaw @c1728p9 @geky

[bulislaw]

@0xc0170 do we have any 16k boards passing our testing?

[0xc0170]

@0xc0170 do we have any 16k boards passing our testing?

KL25Z has the same flash/ram sizes (not now plugged in CI though but it is released for mbed OS 5). NRF51 has even lower ram - 16k parts (16k - softdevice use) and it's passing. The tested nrf51dk has around 20k of RAM. We can look at those and analyze

[cmonr]

Closing since fix looks to be merged in and issue inactivity.