Is there a reason for the changes made to the first few lines in the sw/support/reset.S compared to sw.old/support/reset.S? Without initializing r0 to 0 as was done in the old reset.S, the RTL testbench fails because all registers remain at their XXXXXXXXX... values. It's a simple fix of adding the following three lines to reset.S. But I was just wondering why the change was made and if it has anything to do with running the benchmarks on the or1ksim. .org 0x100 _reset_vector: l.nop l.nop l.movhi r0, 0x0 l.slli r0,r0,16 l.addi r1,r0,0x0