diff options
| author | Quincey Koziol <koziol@hdfgroup.org> | 2000-11-15 17:12:03 (GMT) |
|---|---|---|
| committer | Quincey Koziol <koziol@hdfgroup.org> | 2000-11-15 17:12:03 (GMT) |
| commit | 825e9ff8ff4eee0373c8a22804d528d6c73d793c (patch) | |
| tree | 61d94322f46aa186eeeb1b9063139eea7a822bc3 /tools/h5tools.c | |
| parent | 710bf78e217510783c6b2419086531feffb42c31 (diff) | |
| download | hdf5-825e9ff8ff4eee0373c8a22804d528d6c73d793c.zip hdf5-825e9ff8ff4eee0373c8a22804d528d6c73d793c.tar.gz hdf5-825e9ff8ff4eee0373c8a22804d528d6c73d793c.tar.bz2 | |
[svn-r2922] Purpose:
Bug fix
Description:
[Most] SGIs failed on dumping VL data during the daily tests they
seem to be initializing variables differently, exposing a bug in dumping
datasets with scalar dataspaces.
Also, clean up code to get rid of compiler warnings.
Solution:
Initilialize variable correctly.
Platforms tested:
SGI IRIX 6.5 (paz)
Diffstat (limited to 'tools/h5tools.c')
| -rw-r--r-- | tools/h5tools.c | 8 |
1 files changed, 6 insertions, 2 deletions
diff --git a/tools/h5tools.c b/tools/h5tools.c index 8d06467..2408470 100644 --- a/tools/h5tools.c +++ b/tools/h5tools.c @@ -1534,8 +1534,8 @@ h5dump_simple_dset(FILE *stream, const h5dump_t *info, hid_t dset, return -1; /* Print the data */ - flags = (((elmtno == 0) ? START_OF_DATA : 0) | - (((elmtno + hs_nelmts) >= p_nelmts) ? END_OF_DATA : 0)); + flags = (elmtno == 0) ? START_OF_DATA : 0; + flags |= ((elmtno + hs_nelmts) >= p_nelmts) ? END_OF_DATA : 0; h5dump_simple_data(stream, info, dset, &ctx, flags, hs_nelmts, p_type, sm_buf); @@ -2036,6 +2036,10 @@ recheck: /* Create a simple memory space so that we can read in the hvl_t object */ mem_space = H5Screate_simple(0, NULL, NULL); + /* Fake dims[0] for ndim==0, FIX! */ + if(ndims==0) + dims[0]=1; + for (i = 0; i < dims[0]; i++) { herr_t ret; hvl_t vldata; |