autotools build not working on current master

[edhartnett]

I just checked out master and tried to build, but got a failure:

autoreconf -i && CC=mpicc CPPFLAGS=-I/usr/local/netcdf-c-4.6.2_mpich-3.2/include LDFLAGS=-L/usr/local/netcdf-c-4.6.2_mpich-3.2/lib ./configure && make check
and I get:

FAIL: run_tests.sh

Running the test itself I see the problem is test_pioc:

mpiexec -n 4 ./test_pioc

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   PID 4448 RUNNING AT mikado
=   EXIT CODE: 139
=   CLEANING UP REMAINING PROCESSES
=   YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
===================================================================================
YOUR APPLICATION TERMINATED WITH THE EXIT STRING: Segmentation fault (signal 11)
This typically refers to a problem with your application.
Please see the FAQ page for debugging suggestions
ed@mikado:~/tmp/ParallelIO/tests/cunit$ 

[jedwards4b]

Can you provide more information? Compiler version, os, etc - typically this error (segfault with no other output) indicates a problem with mpi and not with the test itself.

[edhartnett]

This turns out to be a bug in netCDF 4.6.2. I am fixing now...

[edhartnett]

Well it's also a PIO issue. I now have to change the PIO code to not rely on NULL being a valid stride (standing for all 1's).

An annoyance, but required if PIO is to work with netCDF 4.6.2. So I will reopen this issue and submit a PR with those changes.

[jedwards4b]

I just want to make sure I'm understanding this - netcdf 4.6.2 is broken and this will be fixed in the next netcdf release but we still want to put a hack in pio so that it works with that version. Is that right?

[edhartnett]

Yes, otherwise we will have to constantly answer support questions about why PIO is broken with netcdf 4.6.2. ;-)

It's a simple hack that won't affect performance.

[jedwards4b]

Hi Ed - what is the ETA on this change?

[edhartnett]

Will try to get to it next week...

…

On Fri, Jan 18, 2019 at 9:48 AM jedwards4b ***@***.***> wrote: Hi Ed - what is the ETA on this change? — You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub <#1320 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEUr3OmW3hqSiq8itfTsf9B5YK9m8xbbks5vEfrigaJpZM4ZlZRr> .

[edhartnett]

Working on this now...

This is happening here:

0x0000555555569c53 in PIOc_put_vars_tc (ncid=<optimized out>, ncid@entry=52, varid=<o
    start=start@entry=0x555555882360, count=count@entry=0x555555a09e40, stride=stride@ent
    xtype=<optimized out>, xtype@entry=4, buf=0x7fffffffd7a0) at pio_getput_int.c:1158
1158                    ierr = nc_put_vars_int(file->fh, varid, (size_t *)start, (size_t