Double zeroes in execution trace
A Java based, pMars compatible, CoreWars simulator with GA
Status: Beta
Brought to you by:
miha_vitorovic
{
myLW: {96, 96, 0, 0, 96},
victory: 1,
myOW: {96, 96, 96, 0, 96},
cycles: 79931,
size: 37,
id: 96
}
{
victory: 1,
cycles: 80000,
size: 37,
id: 99
}
Two problems:
- double zero in the execution trace of the first warrior (unless there were multiple processes - check)
- the battle obviously run the the execution limit (80000), but the engine identified it as a "draw", meaning that both had processes active, but the first one only ran for 79931 cycles. Either the cycle counting is flawed, or process die detection if flawed.
Diff:
The discrepancy in the number of cycles probably comes by not counting the execution of the DAT directive(s).
Execution counting should be fixed by d203bfe49c1a948805852657f7e08dd3a6606dcd.
0 in both traces (but not on the last place) is may be caused by SPL (into uninitialized memory). Will research later.