Vanadis - "Error has been detected in retired instruction"

[plafratt]

New Issue for sst-elements

1 - Detailed description of problem or enhancement

When simulating HPCG on Vanadis, Vanadis reports the error "Error has been detected in retired instruction".

2 - Describe how to reproduce

Edit the configuration file src/sst/elements/vanadis/tests/basic_vanadis.py to set numCpus to 2 and numThreads to 2. Change processList to set the environment variable KMP_USE_YIELD to 0 to avoid an unrecognized system call 124, as shown below.

processList = (
    ( 1, {
        "env_count" : 2,
        "env0" : "OMP_NUM_THREADS={}".format(numCpus*numThreads),
        "env1" : "KMP_USE_YIELD=0",
        "exe" : full_exe_name,
        "arg0" : exe_name,
    } ),
)

Set VANADIS_EXE to the point to the attached executable. xhpcg.zip

Note that I have run this executable natively, and it ran successfully to completion.

Set VANADIS_EXE_ARGS as shown below. Then run the simulation.

$> export VANADIS_EXE=./xhpcg
$> export VANADIS_EXE_ARGS='--rt=1 --nx=1 --ny=1 --nz=1'
$> sst ./basic_vanadis.py

Note that soon after it starts, the program will prompt you to "Press key to continue". So, you will need to press a key.

The simulation simulates about 30 ms and then fails with

[node0,Core:    0/29212366710]: Error has been detected in retired instruction. Retired register status:
SST Fatal Backtrace Information:
    0 : sst(_ZNK3SST6Output5fatalEjPKcS2_iS2_z+0x354) [0x562214]
    1 : .../lib/sst-elements-library/libvanadis.so(_ZN3SST7Vanadis21VanadisDebugComponent13performRetireEiPNS0_20VanadisCircularQueueIPNS0_18VanadisInstructionEEEm+0x63c) [0xffffa243431c]
    2 : .../lib/sst-elements-library/libvanadis.so(_ZN3SST7Vanadis21VanadisDebugComponent4tickEm+0x1e0) [0xffffa2436a20]
    3 : sst(_ZN3SST5Clock7executeEv+0x17c) [0x4dedbc]
    4 : sst(_ZN3SST15Simulation_impl3runEv+0x3e0) [0x590510]
    5 : sst() [0x4cb084]
    6 : sst(main+0x1cb0) [0x4b1a20]
    7 : /lib64/libc.so.6(__libc_start_main+0xdc) [0xffffafdd4384]
    8 : sst() [0x4c7af0]

3 - What Operating system(s) and versions

Rocky Linux 8.9

4 - What version of external libraries (Boost, MPI)

5 - Provide sha1 of all relevant sst repositories (sst-core, sst-elements, etc)

sst-elements 7e67f8f29
sst-core e952a81bcb

6 - Fill out Labels, Milestones, and Assignee fields as best possible