I'm attempting to run QE 6.4.1 + EPW 5.1 on a remote server. The first issue that occurred (that may or may not be relates to this error) was that this server runs under an old version of the fort compiler (13.1.3). This led to some segmentation errors, that were solved by following these instructions https://gitlab.com/QEF/q-e/wikis/Support/Patch-for-old-Intel-compilers.
After applying the patch, the compilation ended successfully and I proceeded by trying to run the eww-metal example, from test-suite. The QE part of the job runs flawlessly, but when the epw step comes in, two different error occur. Firstly, when I tried to run everything in more than one core, the job freezes after the "Running Wannier90" line on the output. This test should be light, but I waited for several hours and it didn't go beyond this point. Despite that, no error message was displayed.
For a second try I repeated the procedure in only one core, and this time the running wannier90 step was done after several minutes, and the calculation proceeded normally, until what I think is near the end, when it calculates the critical temperatures, but after this point it crashed, without no error message in the epw.out file, and with the following warning being displayed in the terminal:
Code: Select all
*** glibc detected *** /prometeu-home/fabiocosta4/qe-6.4.1/bin/epw.x: double free or corruption (!prev): 0x0000000003264730 ***
======= Backtrace: =========
/lib64/libc.so.6[0x2b9ae435caf8]
/lib64/libc.so.6(cfree+0x76)[0x2b9ae435e6e6]
/prometeu-home/fabiocosta4/qe-6.4.1/bin/epw.x[0x11a39fc]
/prometeu-home/fabiocosta4/qe-6.4.1/bin/epw.x[0x427bce]
/prometeu-home/fabiocosta4/qe-6.4.1/bin/epw.x[0x40e22d]
/prometeu-home/fabiocosta4/qe-6.4.1/bin/epw.x[0x4078f7]
/prometeu-home/fabiocosta4/qe-6.4.1/bin/epw.x[0x406b0c]
/lib64/libc.so.6(__libc_start_main+0xe6)[0x2b9ae4305436]
/prometeu-home/fabiocosta4/qe-6.4.1/bin/epw.x[0x406a09]
======= Memory map: ========
00400000-014c0000 r-xp 00000000 00:14 47800431 /prometeu-home/fabiocosta4/qe-6.4.1/EPW/src/epw.x
016c0000-016c4000 r--p 010c0000 00:14 47800431 /prometeu-home/fabiocosta4/qe-6.4.1/EPW/src/epw.x
016c4000-018d1000 rw-p 010c4000 00:14 47800431 /prometeu-home/fabiocosta4/qe-6.4.1/EPW/src/epw.x
018d1000-0341c000 rw-p 018d1000 00:00 0 [heap]
2b9ae3a4c000-2b9ae3a69000 r-xp 00000000 08:02 1439097 /lib64/ld-2.8.so
2b9ae3a69000-2b9ae3a6b000 rw-p 2b9ae3a69000 00:00 0
2b9ae3c69000-2b9ae3c6a000 r--p 0001d000 08:02 1439097 /lib64/ld-2.8.so
2b9ae3c6a000-2b9ae3c6b000 rw-p 0001e000 08:02 1439097 /lib64/ld-2.8.so
2b9ae3c6b000-2b9ae3c73000 r-xp 00000000 08:02 1439059 /lib64/librt-2.8.so
2b9ae3c73000-2b9ae3e72000 ---p 00008000 08:02 1439059 /lib64/librt-2.8.so
2b9ae3e72000-2b9ae3e73000 r--p 00007000 08:02 1439059 /lib64/librt-2.8.so
2b9ae3e73000-2b9ae3e74000 rw-p 00008000 08:02 1439059 /lib64/librt-2.8.so
2b9ae3e74000-2b9ae3e8a000 r-xp 00000000 08:02 1439061 /lib64/libpthread-2.8.so
2b9ae3e8a000-2b9ae408a000 ---p 00016000 08:02 1439061 /lib64/libpthread-2.8.so
2b9ae408a000-2b9ae408b000 r--p 00016000 08:02 1439061 /lib64/libpthread-2.8.so
2b9ae408b000-2b9ae408c000 rw-p 00017000 08:02 1439061 /lib64/libpthread-2.8.so
2b9ae408c000-2b9ae4090000 rw-p 2b9ae408c000 00:00 0
2b9ae4090000-2b9ae40e5000 r-xp 00000000 08:02 1439154 /lib64/libm-2.8.so
2b9ae40e5000-2b9ae42e4000 ---p 00055000 08:02 1439154 /lib64/libm-2.8.so
2b9ae42e4000-2b9ae42e5000 r--p 00054000 08:02 1439154 /lib64/libm-2.8.so
2b9ae42e5000-2b9ae42e6000 rw-p 00055000 08:02 1439154 /lib64/libm-2.8.so
2b9ae42e6000-2b9ae42e7000 rw-p 2b9ae42e6000 00:00 0
2b9ae42e7000-2b9ae4436000 r-xp 00000000 08:02 1439155 /lib64/libc-2.8.so
2b9ae4436000-2b9ae4636000 ---p 0014f000 08:02 1439155 /lib64/libc-2.8.so
2b9ae4636000-2b9ae463a000 r--p 0014f000 08:02 1439155 /lib64/libc-2.8.so
2b9ae463a000-2b9ae463b000 rw-p 00153000 08:02 1439155 /lib64/libc-2.8.so
2b9ae463b000-2b9ae4640000 rw-p 2b9ae463b000 00:00 0
2b9ae4640000-2b9ae4656000 r-xp 00000000 08:02 1439083 /lib64/libgcc_s.so.1
2b9ae4656000-2b9ae4855000 ---p 00016000 08:02 1439083 /lib64/libgcc_s.so.1
2b9ae4855000-2b9ae4856000 r--p 00015000 08:02 1439083 /lib64/libgcc_s.so.1
2b9ae4856000-2b9ae4857000 rw-p 00016000 08:02 1439083 /lib64/libgcc_s.so.1
2b9ae4857000-2b9ae4859000 r-xp 00000000 08:02 1439068 /lib64/libdl-2.8.so
2b9ae4859000-2b9ae4a59000 ---p 00002000 08:02 1439068 /lib64/libdl-2.8.so
2b9ae4a59000-2b9ae4a5a000 r--p 00002000 08:02 1439068 /lib64/libdl-2.8.so
2b9ae4a5a000-2b9ae4a5b000 rw-p 00003000 08:02 1439068 /lib64/libdl-2.8.so
2b9ae4a5b000-2b9ae4e5f000 rw-p 2b9ae4a5b000 00:00 0
2b9ae4e7b000-2b9ae4e85000 r-xp 00000000 08:02 1439158 /lib64/libnss_files-2.8.so
2b9ae4e85000-2b9ae5085000 ---p 0000a000 08:02 1439158 /lib64/libnss_files-2.8.so
2b9ae5085000-2b9ae5086000 r--p 0000a000 08:02 1439158 /lib64/libnss_files-2.8.so
2b9ae5086000-2b9ae5087000 rw-p 0000b000 08:02 1439158 /lib64/libnss_files-2.8.so
2b9ae5087000-2b9ae537e000 rw-p 2b9ae5087000 00:00 0
2b9ae53c2000-2b9ae53f4000 rw-p 2b9ae53c2000 00:00 0
2b9ae5438000-2b9ae5a21000 rw-p 2b9ae5438000 00:00 0
2b9ae8000000-2b9ae8021000 rw-p 2b9ae8000000 00:00 0
2b9ae8021000-2b9aec000000 ---p 2b9ae8021000 00:00 0
7fffc64a8000-7fffc705e000 rw-p 7fffff449000 00:00 0 [stack]
7fffc71fe000-7fffc7200000 r-xp 7fffc71fe000 00:00 0 [vdso]
ffffffffff600000-ffffffffff601000 r-xp 00000000 00:00 0 [vsyscall]
/prometeu-home/n1ge6/default/spool/no02/job_scripts/459: line 19: 23945 Abortado (core dumped) $QE_DIR/epw.x -i epw.in > epw.out
It is unclear to me what this error message means, and also why in more than one core the calculation won't pass further the wannier90 step. Could this be related with the old intel compiler? It is worth saying that everything seems to be working fine with the QE applications, after the patch.
Many thanks for any assistance
Fbio Costa