“make check failed tsformat.test 8”
Added by Chris Wang almost 7 years ago
Hi guys,
I compiled cdo-1.9.2 using intel icc and ifort compilers on my Macbook, with hdf5-1.10.1, netcdf-c-4.5.0, netcdf-fortran-4.4.4 and netcdf-cxx-4.3.0, all libs were compiled correctly, all passed "make check".
I also specified the install directories when I do ./configure, when I do make check, here's the error info like that:
"FAIL: tsformat.test 8 - chaining set 1 with netCDF4" ,
all other test were passed.
Can anyone give me some info to fix it?
----Chris
Replies (2)
RE: “make check failed tsformat.test 8” - Added by Kevin Havener almost 7 years ago
Not sure I have any info to fix it, but it doesn't have anything to do with your mac/icc/ifort compilers. I get the same error with a linux/gcc/g77 box (cdo -V):
Climate Data Operators version 1.9.2 (http://mpimet.mpg.de/cdo)
Compiled: by xxxxxxx on yyyyyyyy (x86_64-unknown-linux-gnu) Jan 8 2018 09:07:14
CXX Compiler: g++ -g -O2 -fopenmp
CXX version : g++ (Ubuntu 7.2.0-8ubuntu3) 7.2.0
C Compiler: gcc -g -O2 -fopenmp
C version : gcc (Ubuntu 7.2.0-8ubuntu3) 7.2.0
F77 Compiler: f77 -g -O2
F77 version : unknown
Features: 7GB Fortran DATA PTHREADS OpenMP45 HDF5 NC4/HDF5 SSE2
Libraries: HDF5/1.10.1
Filetypes: srv ext ieg grb1 grb2 nc1 nc2 nc4 nc4c nc5
CDI library version : 1.9.2 of Jan 8 2018 09:04:04
CGRIBEX library version : 1.9.0 of Sep 29 2017 10:16:02
GRIB_API library version : 2.6.0 ##compiled with eccodes so I guess this grib_api version is right
NetCDF library version : 4.5.0 of Jan 5 2018 15:38:25 $
HDF5 library version : 1.10.1
SERVICE library version : 1.4.0 of Jan 8 2018 09:03:51
EXTRA library version : 1.4.0 of Jan 8 2018 09:03:46
IEG library version : 1.4.0 of Jan 8 2018 09:03:49
FILE library version : 1.8.3 of Jan 8 2018 09:03:46
However, I wonder if it might be a hdf5-1.10.1 problem. I suppose it could also be a netcdf-4.5.0 problem, too. I encountered the same error compiling cdo-1.8.2 on another linux machine (no C++11 compliant compiler). Here I got a core dump at least. When I get a chance, I will rebuild cdo with hdf5-1.8.latest, then perhaps netcdf-4.4.latest if needed. Here is the cdo -V and tail end of the test/tsformat.log for that machine:
limate Data Operators version 1.8.2 (http://mpimet.mpg.de/cdo)
Compiled: by xxxxxxx on yyyyyyy (x86_64-unknown-linux-gnu) Jan 8 2018 15:15:22
Compiler: gcc -std=gnu99 -fPIC -fopenmp
version: gcc (GCC) 4.4.7 20120313 (Red Hat 4.4.7-18)
Features: DATA PTHREADS OpenMP3 HDF5 NC4/HDF5 OPeNDAP SSE2
Libraries: HDF5/1.10.1
Filetypes: srv ext ieg grb1 nc1 nc2 nc4 nc4c
CDI library version : 1.8.2 of Jan 8 2018 15:14:24
CGRIBEX library version : 1.7.6 of Dec 20 2016 19:55:24
NetCDF library version : 4.5.0 of Jan 4 2018 21:00:16 $
HDF5 library version : 1.10.1
SERVICE library version : 1.4.0 of Jan 8 2018 15:14:10
EXTRA library version : 1.4.0 of Jan 8 2018 15:14:04
IEG library version : 1.4.0 of Jan 8 2018 15:14:07
FILE library version : 1.8.3 of Jan 8 2018 15:14:04
...
Running test: 8 - chaining set 1 with netCDF4
Error (cdf_put_att_double) : NetCDF: Attempt to define fill value when data already exists.
HDF5-DIAG: Error detected in HDF5 (1.10.1) thread 0:
#000: H5F.c line 636 in H5Fflush(): invalid file identifier
major: Invalid arguments to routine
minor: Inappropriate type
Error (cdf_close) : NetCDF: HDF error
../src/cdo -fldmin -timmean -select,code=130 ../test/data/pl_data(nc4) thread8_res
cdf_put_att_double : ncid = 65536 varid = 4 att = _FillValue val = -9e+33
cdo fldmin: Started child process "timmean -select,code=130 ifile3262 (pipe1.1)".
cdo(2) timmean: Started child process "select,code=130 ifile3262 (pipe2.1)".
./tsformat.test: line 59: 3424 Segmentation fault (core dumped) $CDOCOMMAND
cdo fldmin: Started child process "timmean -select,code=130 ifile3262 (pipe1.1)".
cdo(2) timmean: Started child process "select,code=130 ifile3262 (pipe2.1)".
./tsformat.test: line 59: 3428 Segmentation fault (core dumped) $CDOCOMMAND
cdo fldmin: Started child process "timmean -select,code=130 ifile3262 (pipe1.1)".
cdo(2) timmean: Started child process "select,code=130 ifile3262 (pipe2.1)".
./tsformat.test: line 59: 3432 Segmentation fault (core dumped) $CDOCOMMAND
cdo fldmin: Started child process "timmean -select,code=130 ifile3262 (pipe1.1)".
cdo(2) timmean: Started child process "select,code=130 ifile3262 (pipe2.1)".
Error (cdf_create) : thread8_res: NetCDF: HDF error
cdf_create : ncid = -1 mode = 4096 file = thread8_res
cdf_create : chunksizehint 0
./tsformat.test: line 59: 3437 Aborted (core dumped) $CDOCOMMAND
cdo fldmin: Started child process "timmean -select,code=130 ifile3262 (pipe1.1)".
cdo(2) timmean: Started child process "select,code=130 ifile3262 (pipe2.1)".
Error (cdf_get_vara_float) : ./tsformat.test: line 59: 3441 Segmentation fault (core dumped) $CDOCOMMAND
RE: “make check failed tsformat.test 8” - Added by Kevin Havener almost 7 years ago
I compiled hdf5 with --enable-threadsafe --disable-hl and the error went away. Missed the discussion on the wiki page about this somehow.