* gromacs: fix fftw dependency
Only depend on fftw+mpi when gromacs is built with mpi,
and depend on fftw~mpi otherwise.
* gromacs: fix cmake dependency
master branch depends on cmake 3.11 (as specified in CMakeLists.txt
cmake dependency is also bumped to 3.11 when fj compilers are used
in order to fix OpenMP detection.
* Some minor fixes to set_permissions() in file_permissions.py
The set_permissions() routine claims to prevent users from creating
world writable suid binaries. However, it seems to only be checking
for/preventing group writable suid binaries.
This patch modifies the routine to check for both world and group
writable suid binaries, and complain appropriately.
* permissions.py: Add test to check blocks world writable SUID files
The original test_chmod_rejects_group_writable_suid tested
that the set_permissions() function in
lib/spack/spack/util/file_permissions.py
would raise an exception if changed permission on a file with
both SUID and SGID plus sticky bits is chmod-ed to g+rwx and o+rwx.
I have modified so that more narrowly tests a file with SUID
(and no SGID or sticky bit) set is chmod-ed to g+w.
I have added a second test test_chmod_rejects_world_writable_suid
that checks that exception is raised if an SUID file is chmod-ed
to o+w
* file_permissions.py: Raise exception when try to make sgid file world writable
Updated set_permissions() in file_permissions.py to also raise
an exception if try to make an SGID file world writable. And
added corresponding unit test as well.
* Remove debugging prints from permissions.py
* Module index should not be unconditionally overwritten
Uncovered after we switched our CI to generate modules for packages
one-by-one rather than in bulk. This overwrote a complete module index
with an index with a single entry, and broke our downstream Spack
instances that needed the upstream module index.
I get the following error message, if I do not use editline from the system.
```
>> 3090 Undefined symbols for architecture x86_64:
3091 "_tgetent", referenced from:
3092 _terminal_set in libedit.a(terminal.c.o)
3093 "_tgetflag", referenced from:
3094 _terminal_set in libedit.a(terminal.c.o)
3095 "_tgetnum", referenced from:
3096 _terminal_set in libedit.a(terminal.c.o)
...
3110 _terminal_insertwrite in libedit.a(terminal.c.o)
3111 _terminal_clear_EOL in libedit.a(terminal.c.o)
3112 _terminal_clear_screen in libedit.a(terminal.c.o)
3113 _terminal_beep in libedit.a(terminal.c.o)
3114 ...
3115 ld: symbol(s) not found for architecture x86_64
```
* Added unit tests to Github Actions
* Set user e-mail and name for git tests to succeed
* Simplify setup.sh logic
* Replicate Travis script on Github Actions
* Update flags since '.' is not allowed
* Added badge, simplified workflow
* Remove pinning of coverage
* Remove unit tests run on Github Actions from Travis
Bug fix release:
4.0.4 -- June, 2020
-----------------------
- Fix a memory patcher issue intercepting shmat and shmdt. This was
observed on RHEL 8.x ppc64le (see README for more info).
- Fix an illegal access issue caught using gcc's address sanitizer.
Thanks to Georg Geiser for reporting.
- Add checks to avoid conflicts with a libevent library shipped with LSF.
- Switch to linking against libevent_core rather than libevent, if present.
- Add improved support for UCX 1.9 and later.
- Fix an ABI compatibility issue with the Fortran 2008 bindings.
Thanks to Alastair McKinstry for reporting.
- Fix an issue with rpath of /usr/lib64 when building OMPI on
systems with Lustre. Thanks to David Shrader for reporting.
- Fix a memory leak occurring with certain MPI RMA operations.
- Fix an issue with ORTE's mapping of MPI processes to resources.
Thanks to Alex Margolin for reporting and providing a fix.
- Correct a problem with incorrect error codes being returned
by OMPI MPI_T functions.
- Fix an issue with debugger tools not being able to attach
to mpirun more than once. Thanks to Gregory Lee for reporting.
- Fix an issue with the Fortran compiler wrappers when using
NAG compilers. Thanks to Peter Brady for reporting.
- Fix an issue with the ORTE ssh based process launcher at scale.
Thanks to Benjamín Hernández for reporting.
- Address an issue when using shared MPI I/O operations. OMPIO will
now successfully return from the file open statement but will
raise an error if the file system does not supported shared I/O
operations. Thanks to Romain Hild for reporting.
- Fix an issue with MPI_WIN_DETACH. Thanks to Thomas Naughton for reporting.
Signed-off-by: Howard Pritchard <howardp@lanl.gov>
* Fix how the Conduit detects that the MPI compiler is the same as the
CC compiler and is more careful when it sets the MPI compilers to be
the Cray PE system compilers.
* Remove unnecessary push of the MPI compilers to the C compilers for Hydrogen.
