Mark Melvin

This is with the latest version. Maybe I am mixing two issues here. We have also worked around the potential lockup and some segfaulting issues by wrapping cocotb.coroutine with an...

Hi @ktbarrett - I tried your suggestion of yielding a NullTrigger() first and that did not do anything. Segfault still occurs. It is a bit annoying because the segfaulting process...

Yeah, it gives no info. It looks like it is probably segfaulting in vcs...probably in the code that accesses the dut?? The current stacktrace is: ``` *** Error in `sim_build/simv':...

Got a core dump - a little more info: ``` Core was generated by `sim_build/simv +define+COCOTB_SIM=1 +vcs+grwavesoff +vpdignore +vcs+dumpvarsoff'. Program terminated with signal 11, Segmentation fault. #0 0x00002aaab2b362c7 in raise...

Setting `DEBUG=1` gives a bit more info for this particular issue: ``` --- Stack trace follows: Dumping VCS Annotated Stack: #0 0x00002aaab2cdebbc in waitpid () from /lib64/libc.so.6 #1 0x00002aaab2c5cea2 in...

Actually ignore that previous thing. I believe that error happened due to me trying to attach with a debugger. That is probably not the root cause. I seem to get...

> The above problem is still real though, right - even if its not the one you were originally debugging? Perhaps. This occurred when the sim process was relaunching itself...

OK, I just experienced an exception in our code (seen with Icarus) that undoubtedly would have triggered a segfault in VCS. The problem is when VCS segfaults, I believe it...

@eric-wieser - thanks, but this didn't work. I can see my logging info printed for normal test failures, but when a segfault occurs there is no additional information. :(

Thanks @ktbarrett . So some interesting behaviour...if I reproduce the test case as I originally posted in this issue, my debug prints in `cocotb.outcomes.Error.__init__` do actually get printed before the...