modified run_phase -> main_phase
Created by: subbu009
As per issue #1933 , I have modified run_phase
to main_phase
to be in consistent with the use of other run time phases (reset_phase
, confgure_phase
) in the base test, firmware test and firmware riscof test.
Now the problem with line 154 in uvmt_cv32e40p_firmware_test.sv is statements below are waiting on posedge of resetn. If we use run_phase
it works fine because run_phase
starts at time 0 in parallel with reset_phase
and senses the posedge of resetn whenever it happens. But when I use main_phase
instead, I observed the line 154 waits on for resetn posedge and it never happens because the reset_phase
has already completed. Due to this the test hangs and never finishes. So, to run ci_check, I have commented the line 154, and checks passed as expected. The question : is it okay if we do not wait for posedge of resetn? if we were to wait on reset, then it is better to fork off the current reset, configure and run phase tasks in only one run_phase so the reset posedge in one thread can be detected correctly by other forked thread.
On the other hand, if we would like to add on-the-fly reset feature as @MikeOpenHWGroup
explained, then it is better to use reset, configure and main phases instead of run_phase
.