StanfordAHA/lassen

Lassen Bug Tracker

Closed this issue · 8 comments

rdaly525 commented

FP_Mult Error mode issue
-Problem: RTL mismatched FPVector and Halide due to wrapper
-Solution: Use DW FP IP instead
-Verification: Lassen RTL tests
-Can be resolved today

SLT/SGT Bug
-problem: SMT cannot infer SLT and SGT. But the random tests are passing
-Solution: Still debugging (Caleb)
-Verification: Automapper finding SLT and SGT

SubExp Bug
-problem: A small portion of random tests (~5%) are failing the SubExp micro op RTL
-solution: Still debugging (Lenny/anyone else wants to help?)
-verification: Lassen RTL tests

leonardt commented

RE: SubExp Bug

Can you point to an example failure? It would be useful to get the specific inputs causing the failure so we can debug it. Is this on master?

rdaly525 commented

actions = [Poke(WrappedPE.inst, 42081637117181558932), Poke(WrappedPE.CLK, 0), Poke(WrappedPE.data0, 22666), Poke(WrappedPE.data1, 31920), Eval(), Expect(O0, 7050)]

Got : 0x3ddf
Expected : 0xbddf
i : 5
Port : WrappedPE.O0

leonardt commented

RE: SubExp Bug we believe this was because an old version of fault was not generating the sigend values properly, we will assume it's fixed unless it pops again either locally for someone or on buildkite/travis

rdaly525 commented

SLT/SGT bug is due to hwtypes issue here: leonardt/hwtypes#52

leonardt commented

I believe these have all been resolved, please re open if that's not the case

cdonovick commented

leonardt/hwtypes#52 was never merged

leonardt commented

merged