mkannwischer
commented
11 months ago Awesome! I tested it and it gives me very similar results to the previous benchmarks. Itt resolves my problems with the stack measurements for some schemes. Thanks @rpls! https://github.com/mupq/mupq/pull/122
Use the new stack spraying/checking API of the HAL. This adds a small test in the
boardtest.elf, which for me returns an accurate result (400/800 bytes alocated on stack, measurement returns 400/800 bytes). Some random spot-checks of schemes return very similar values for stack usage (few bytes difference, might also be explained by newer compiler).The implementations checks entire words instead of bytes, so the result will only be on word granularity, i.e., 4 bytes.