Why? When taking into account the cpu cache, branch mispredictions, etc, I bet it would be slower than just zeroing it, besides it wouldn't be secure at all, im

Why? When taking into account the cpu cache, branch mispredictions, etc, I bet it would be slower than just zeroing it, besides it wouldn't be secure at all, imagine a process that stores a secret key, and then releases the memory, if another process can trigger the first to generate and release the key memory multiple times, they would be able to read it.