[netCDF #TEQ-670685]: error during LibMesh compile

> > I don't know, but have no explanation for the appearance of the spurious
> > strings "<http://c0.nc>" and "<http://c0_4.nc>" in the middle the output
> > you sent, supposedly resulting from running the ncgen3/run_tests.sh and
> > ncgen3/run_nc4_tests.sh shell scripts.  Is it possible that the output you
> > sent was accidentally modified when you cut and pasted it into the support
> > request?
> >
> Actually I checked the config.log file (both the attachment and the
> original one) and there is no such string. I'm sending it again, please
> tell me if there is still anything weird...

I must have been unclear when I described where that string occurred.  It's 
not in your config.log file (in which I found nothing wrong), but instead
in the output from "make check" that was in the original question you sent
from which I quote:

>   While compiling libMesh I got these failed tests related to netCFD. 
>   Apparently they prevent me from using Exodus in libMesh.
> 
>   *** Testing ncgen3.
>   *** creating classic file c0.nc <http://c0.nc> from c0.cdl...
>   ../../../../../contrib/netcdf/v4/ncgen3/run_tests.sh: line 8:
>   37560 Segmentation fault: 11 ./ncgen3 -b -o c0.nc
>   <http://c0.nc> $srcdir/c0.cdl
>   FAIL: run_tests.sh
>   *** Testing ncgen3 for netCDF-4.
>   *** creating netCDF-4 file c0_4.nc <http://c0_4.nc> from c0.cdl...
>   ../../../../../contrib/netcdf/v4/ncgen3/run_nc4_tests.sh: line
>   8: 37581 Segmentation fault: 11 ./ncgen3 -k3 -b -o c0_4.nc
>   <http://c0_4.nc> $srcdir/c0.cdl
>   FAIL: run_nc4_tests.sh
>   ================================================
>   2 of 2 tests failed

I've never seen "<http://c0.nc>" or "<http://c0_4.nc>" in the output
from "make check" for netCDF, and I don't see anything in the .sh files
being run by "make check" or in your "config.log" that could explain
thoses strings.  Maybe they have nothing to do with why "make check"
failed for you, but their unexpected occurrence in the output of those
two tests made it seem that there is some relationship with the
segmentation faults you are seeing.

Do you still get those two failures from running "make check" *after*
running the configure script.  If so, do you still see the same output
as you originally reported, including those unexpected strings?

--Russ

Russ Rew                                         UCAR Unidata Program
address@hidden                      http://www.unidata.ucar.edu



Ticket Details
===================
Ticket ID: TEQ-670685
Department: Support netCDF
Priority: Normal
Status: Closed