E3SM-Project / E3SM-Project/E3SM
ELM crashes on cori-knl with gcc/8.2.0
- Dominant language
- Fortran
- Stars
- 440
- Forks
- 481
- Avg merge
- 4d 7h
- Merged PRs (30d)
- 33
Description
Current master is crashes in the land model for gnu with following error:
```
23:
23: Program received signal SIGILL: Illegal instruction.
23:
23: Backtrace for this error:
23:
23: Program received signal SIGILL: Illegal instruction.
23:
23: Backtrace for this error:
23: #0 0x110950f in ???
23: at /home/abuild/rpmbuild/BUILD/glibc-2.26/nptl/../sysdeps/unix/sysv/linux/x86_64/sigaction.c:0
23: #1 0x981136 in ???
23: #2 0x91ee3c in ???
23: #3 0x4f03a1 in ???
23: #0 0x110950f in ???
23: at /home/abuild/rpmbuild/BUILD/glibc-2.26/nptl/../sysdeps/unix/sysv/linux/x86_64/sigaction.c:0
23: #1 0x981136 in ???
23: #2 0x91ee3c in ???
23: #3 0x4f03a1 in ???
23: #4 0x2be519d in gomp_thread_start
23: at ../../../cray-gcc-8.2.0-201811010913.df0113f60eb17/libgomp/team.c:120
23: #4 0x2be4cce in GOMP_parallel
23: at ../../../cray-gcc-8.2.0-201811010913.df0113f60eb17/libgomp/parallel.c:168
23: #5 0x4f1540 in ???
23: #6 0x4de722 in ???
23: #7 0x42989d in ???
23: #8 0x40a569 in ???
23: #9 0x4289e0 in ???
23: #10 0x2c9cf0e in __libc_start_main
23: at ../csu/libc-start.c:308
23: #11 0x4046e9 in ???
23: at ../sysdeps/x86_64/start.S:120
```
The failure can be reproduced by
`./create_test ERS.f09_g16.I1850CLM45CN.cori-knl_gnu --walltime 00:30:00 -q debug`
Contributor guide
Research direction
Start by reproducing the failure with ./create_test ERS.f09_g16.I1850CLM45CN.cori-knl_gnu --walltime 00:30:00 -q debug. Then inspect the ELM land-model run and the cori-knl_gnu build/runtime configuration around the SIGILL, using the provided backtrace as a starting point. Done means this test completes without the illegal-instruction crash.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- fortran
- Domain
- hpc
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 25/100