summaryrefslogtreecommitdiffstats
path: root/src/hdf5.h
diff options
context:
space:
mode:
authorRobb Matzke <matzke@llnl.gov>1998-03-17 01:29:54 (GMT)
committerRobb Matzke <matzke@llnl.gov>1998-03-17 01:29:54 (GMT)
commit31a709a6b24f4cf80f9cd99a3e55f56e81cf3066 (patch)
treec38f19508a87b990f911f8fe2d526b0cb1f2a2b3 /src/hdf5.h
parent48e075110624e0039b1918a7187316366a20462b (diff)
downloadhdf5-31a709a6b24f4cf80f9cd99a3e55f56e81cf3066.zip
hdf5-31a709a6b24f4cf80f9cd99a3e55f56e81cf3066.tar.gz
hdf5-31a709a6b24f4cf80f9cd99a3e55f56e81cf3066.tar.bz2
[svn-r322] Changes since 19980313
---------------------- ./configure.in ./configure ./test/iopipe.c Added check for system(3) ./config/freebsd2.2.1 ./config/linux Added -DH5D_DEBUG to the debug flags. ./src/H5D.c ./src/H5Dprivate.h ./src/H5P.c ./src/H5Ppublic.h ./src/H5Sprivate.h ./src/H5Ssimp.c ./html/Datasets.html Finally implemented strip mining in the I/O pipeline, placing a user-defined upper bound on the amount of temporary memory used by the pipeline. The default is 1MB type conversion and background buffers allocated/freed on demand. However, the size can be changed and/or application-allocated buffers supplied with H5Pset_buffers() called on the data transfer property list passed to H5Dread() and H5Dwrite(). Minor optimizations to H5Dread() and H5Dwrite(). More coming later... ./test/dsets.c Added calls to H5Pset_buffer() to test application-defined temporary buffers in the I/O pipeline. ./test/Makefile.in Removed `iopipe' from the list of confidence tests. Saying `make timings' in the test directory runs timing tests. I did this because (1) they don't really test anything new, and (2) they can take a long time to run.
Diffstat (limited to 'src/hdf5.h')
0 files changed, 0 insertions, 0 deletions