I've fallen into the trap several times of thinking that only may-block syscalls need scratch, but in fact any syscall with outparams needs to use scratch. We should ensure that this is the case. On the fence about this for 1.0; lean towards no because we're not seeing problems in the field (or #876 might be this, but I doubt it).
I've fallen into the trap several times of thinking that only may-block syscalls need scratch, but in fact any syscall with outparams needs to use scratch. We should ensure that this is the case. On the fence about this for 1.0; lean towards no because we're not seeing problems in the field (or #876 might be this, but I doubt it).