HDF5Examples/JAVA/H5D/JavaDatasetExample.sh.in | 2 1 + 1 - 0 !
HDF5Examples/JAVA/H5G/JavaGroupExample.sh.in | 2 1 + 1 - 0 !
HDF5Examples/JAVA/H5T/JavaDatatypeExample.sh.in | 2 1 + 1 - 0 !
HDF5Examples/JAVA/TUTR/runExample.sh.in | 2 1 + 1 - 0 !
bin/Makefile.am | 4 4 + 0 - 0 !
bin/h5cc.in | 13 7 + 6 - 0 !
c++/src/Makefile.am | 39 33 + 6 - 0 !
c++/src/h5c++.in | 17 9 + 8 - 0 !
config/commence.am | 13 7 + 6 - 0 !
configure.ac | 20 20 + 0 - 0 !
fortran/src/Makefile.am | 24 20 + 4 - 0 !
fortran/src/h5fc.in | 15 8 + 7 - 0 !
hl/c++/src/Makefile.am | 20 16 + 4 - 0 !
hl/fortran/src/Makefile.am | 21 17 + 4 - 0 !
hl/src/Makefile.am | 20 16 + 4 - 0 !
java/test/junit.sh.in | 2 1 + 1 - 0 !
src/H5build_settings.autotools.c.in | 1 1 + 0 - 0 !
src/Makefile.am | 48 36 + 12 - 0 !
src/libhdf5.settings.autotools.in | 1 1 + 0 - 0 !
test/test_libinfo.sh.in | 6 3 + 3 - 0 !
20 files changed, 203 insertions(+), 69 deletions(-)

 add option --with-flavor=<flavor-name> to configure
 so that '_<flavor-name>' is appended to each built library, .e.g.
 libhdf5_<flavor-name>.so.
 This we can avoid library name collisions when installing both serial
 and $mpi flavors of the libraries:
 * serial is built with no flavor name
 * $mpi parallel versions use --with-flavor=$mpi

src/H5Fint.c | 18 1 + 17 - 0 !
1 file changed, 1 insertion(+), 17 deletions(-)

 avoid the issue of path_max not being available on hurd.
 Call realpath() the POSIX way, with NULL as second parameter so that
 the returned string is dynamically allocated. This way there is no
 more need for the strdup() call.

src/H5T.c | 2 0 + 2 - 0 !
src/H5Tconv_integer.c | 2 0 + 2 - 0 !
2 files changed, 4 deletions(-)

 force the ullong conversion anywhere.

configure.ac | 6 3 + 3 - 0 !
1 file changed, 3 insertions(+), 3 deletions(-)

---

bin/h5cc.in | 2 1 + 1 - 0 !
c++/src/h5c++.in | 2 1 + 1 - 0 !
fortran/src/h5fc.in | 2 1 + 1 - 0 !
3 files changed, 3 insertions(+), 3 deletions(-)

 drop rpath flags from h5cc et al commands
 See: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1023446#22

java/test/TestH5Arw.java | 2 1 + 1 - 0 !
1 file changed, 1 insertion(+), 1 deletion(-)

 fix array overflow
 DIM_X=8
 DIM128_Y=128
 The dataset has 8x128 values, which matches the array's firt two
 dimensions.
 Each value is a 128 bit float.
 Because 128 bits = 16 bytes, the third dimension of the array should be 16.

HDF5Examples/JAVA/H5D/JavaDatasetExample.sh.in | 24 0 + 24 - 0 !
HDF5Examples/JAVA/H5G/JavaGroupExample.sh.in | 24 0 + 24 - 0 !
HDF5Examples/JAVA/H5T/JavaDatatypeExample.sh.in | 24 0 + 24 - 0 !
java/src/Makefile.am | 2 1 + 1 - 0 !
java/test/Makefile.am | 2 1 + 1 - 0 !
java/test/junit.sh.in | 36 6 + 30 - 0 !
6 files changed, 8 insertions(+), 104 deletions(-)

---

configure.ac | 1 1 + 0 - 0 !
1 file changed, 1 insertion(+)

 [patch] re-add call to ac_check_sizeof([__float128]). fixes #4916.
 (#4924)

fortran/test/tH5R.F90 | 2 1 + 1 - 0 !
1 file changed, 1 insertion(+), 1 deletion(-)

 temporary patch to cheat the fortranlib_test which fails
 badly on 32 bit architectures.
 This is upstream issue https://github.com/HDFGroup/hdf5/issues/4926.

config/cmake/ConfigureChecks.cmake | 1 1 + 0 - 0 !
configure.ac | 1 1 + 0 - 0 !
fortran/src/H5_ff.F90 | 9 9 + 0 - 0 !
fortran/src/H5config_f.inc.cmake | 6 6 + 0 - 0 !
fortran/src/H5config_f.inc.in | 3 3 + 0 - 0 !
5 files changed, 20 insertions(+)

---