Workaround for Intel 18 memory leak

v1.2.1

Update for realease v1.2.1.

Update to use pFUnit 4.0

• Exploited pFUnit 4.0 to allow tests that are known to break certain compilers to be "disable'd" without completely eliminating them
• Updated Travis to include testing with gfortran-9
• Various little cleanups to reduce extraneous warning messages during build

Support for external projects using Unix make

Merge pull request #73 from Goddard-Fortran-Ecosystem/develop

Develop

Recovered some important compiler workaround switches.

GFortran and Intel compilers both still require some variant implementations for upstream use cases.

Cleanup of some compiler checks

Some compiler ifdefs were determined to be obsolete.

Better support for inclusion in other projects

Probably not quite worthy of a minor version bump, but this release is now able to work being included in other projects without the cmake install step. Just add it as a subdirectory and set the appropriate include directories to include /.../gFTL/include.

update cmake to 3.0.0

Merge pull request #38 from Goddard-Fortran-Ecosystem/develop

Develop

Minor tweak to cmake metadata

v1.0.1

Merge pull request #36 from Goddard-Fortran-Ecosystem/feature/cmake-p…

Major release

Relatively little has changed since the previous release, but given the other projects that now depend on gFTL it seems timely to go beyond the v0 label.

The one significant improvement since the previous release is that pointers to contained ALLOCATABLE items should now remain valid even when the container is subsequently modified. This includes special logic for Map containers where the value (or key) is ALLOCATABLE, but the Pair itself is not. This feature was generally working for most compilers with optimization on, but until the recent changes a strict compiler could have used copies at various points which would have foiled the MOVE_ALLOC() under the hood. NAG with -C=dangling was used to expose the remaining issues.

Workaround for ifort 18 compiler issue.

v0.92

Merge pull request #14 from Goddard-Fortran-Ecosystem/hotfix/#13-ifor…