summaryrefslogtreecommitdiffstats
path: root/src/H5D.c
Commit message (Collapse)AuthorAgeFilesLines
* [svn-r114] Mostly changes to add new parameter to the H5Ainit_group call. ↵Quincey Koziol1997-10-081-42/+66
| | | | | | The H5D, H5P & H5T interfaces now share datatypes & dataspaces properly also.
* [svn-r111] Changed hdf5_file_t to H5F_t, split data struct ito two halves, fixedRobb Matzke1997-09-241-3/+3
| | | | problems when opening the same file more than once.
* [svn-r108] Changed all hatom_t types to hid_t. (Isn't 'sed' nifty.. :-)Quincey Koziol1997-09-241-25/+25
|
* [svn-r106] Fixed some symbol table bugs. Fixed a free memory read in H5Osdim.c.Robb Matzke1997-09-221-47/+7
|
* [svn-r102] Removed ifdef now that code is working.Quincey Koziol1997-09-191-2/+0
|
* [svn-r100] Interim checkin for purify. Also contains H5G shadow stuff, changesRobb Matzke1997-09-191-95/+93
| | | | | | to int64 encode/decode to fix purify on 32-bit systems, changes to H5O_modify, changes to H5D.c to use H5F_open/close(), etc. Documentation will be mailed shortly...
* [svn-r96] Added "atexit" routines to each interface to free buffers ↵Quincey Koziol1997-09-181-3/+34
| | | | | | allocated during runtime. Isolated but can't figure out how to fix bug reported with purify.
* [svn-r95] Tracked down memory bug.Quincey Koziol1997-09-181-2/+4
|
* [svn-r91] Bug fixes for multiple datasets in the file. Interim checkin, ↵Quincey Koziol1997-09-161-1/+1
| | | | | | theres still a problem with the simple dimensionality caching.
* [svn-r89] ./src/H5Apublic.hRobb Matzke1997-09-161-249/+191
| | | | | | | | | | | | | | Removed H5_OID. ./src/H5D.c Fixed a few things to work better with symbol tables. Combined the H5D_oid_t and H5D_dataset_t structs. ./src/H5Dprivate.c Combined the H5D_oid_t and H5D_dataset_t types. ./src/H5M.c Removed the callback list for H5_OID.
* [svn-r81] Finished basic data-type conversions, files are now portably ↵Quincey Koziol1997-09-151-6/+35
| | | | | | written on most "normal" architectures.
* [svn-r78] Checkpointing dataset code. Everything is currently working, ↵Quincey Koziol1997-09-151-9/+152
| | | | | | except writing a second dataset out to the file seems to loose the first one.
* [svn-r73] Lots of added code for dataset I/O. Its now writing out datasets ↵Quincey Koziol1997-09-121-34/+165
| | | | | | | to the disk correctly, but only in "native" format. I'm adding the data-type conversions and checking the reading later today.
* [svn-r70] Added clarifying comments from code review meeting to H5Mflush and ↵Quincey Koziol1997-09-051-1/+2
| | | | H5D_flush.
* [svn-r53] Added datatype and dimensionality messages to dataset header, ↵Quincey Koziol1997-08-291-28/+48
| | | | | | cleaned up more code, etc.
* [svn-r52] ./src/H5O.cRobb Matzke1997-08-291-22/+43
| | | | | | | | | | | | If the symbol table entry doesn't change then the ent_modified argument to H5O_modify() doesn't change. This allows us to set ent_modified to false and then call H5O_modify() a whole bunch of times and easily determine if any of the calls modified the symbol table entry. ./src/H5D.c Fixed caching of things in the symbol table entry using the scheme we came up with a couple months ago.
* [svn-r49] ./src/H5D.cRobb Matzke1997-08-281-57/+72
| | | | | | | | | | | Changed H5D_flush() to work with the new symbol table functions. I just remembered that caching data in the symbol table entries still needs some work, so there's a line in there that invalidates the symbol table entry cache. Also, there are some comments in there with the word `Quincey' for you to read about inserting the messages in the dataset object header.
* [svn-r47] Finished flashing out the H5M* functions calls. Cleaned up ↵Quincey Koziol1997-08-281-13/+4
| | | | | | non-compiling source. Finished adding in remainder of H5Osdtyp.c functions.
* [svn-r44] ./src/H5D.cRobb Matzke1997-08-281-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | I commented out a bunch of code in H5D_flush() so I could compile and link the test cases. If you commit your changes to that file, whatever they may be, I'll change H5D_flush() to use the new directory functions described below... ./src/H5G.c H5G_namei() ----------- Understands files without directories. The root object (directory or not) can be found by asking for `/' (or any equivalent absolute or relative name like `/.//.///' or `.//.///' (if the cwd is the root)) or by name like `/foo' (or any other equivalent absolute or relative name) if the object has a name message with the value `foo'. This function is used by most of the other H5G functions so they now understand directory-less files too. H5G_new() --------- Creates the root directory implicitly if it doesn't already exist, moving any previous non-directory root object into the new root directory before creating the requested directory within the root. The creation of the root directory happens first, regardless of whether the requested directory creation succeeds. Once the root directory is created it never disappears even if the file later has only one object. This "feature" can be used to force a file to have a root directory by saying something like: H5G_new (..., "/", ...); H5ECLEAR; /*we don't care about the H5E_EXISTS failure*/ The H5G_new() in the previous example fails because the root directory is created implicitly before we attempt to create the requested "/" directory. H5G_find() ---------- Understands files without directories. If there is no root object then this function is guaranteed to fail even if the request is for `/'. As mentioned for H5G_namei(), the root object can be retrieved with `/' or `/foo' (or equivalent absolute or relative names). H5G_insert() ------------ Understands files without directories and attempts to create such files when there is only one object. The root symbol table is created implicitly when necessary. A root object can be given a name message by the caller and then inserted with the name `/' or it can be inserted with the name `/foo' in which case H5G_insert() will set the name message to `foo'. H5G_modify() ------------ Understands files without directories. ./src/H5Fprivate.h Removed the root_type field from hdf5_file_t. ./src/H5Fprivate.h ./src/H5F.c Removed H5F_root_type() ./src/H5Fpublic.h Removed H5F_root_symtype_t. ./src/H5Oprivate.h ./src/H5O.c Added H5O_remove() which is currently a no-op.
* [svn-r42] Minor tweaks to correct parameters for H5O_newQuincey Koziol1997-08-261-1/+1
|
* [svn-r41] Interim checkpoint of dataset code for Robb to look at.Quincey Koziol1997-08-261-11/+54
|
* [svn-r35] ./src/H5ACproto.hRobb Matzke1997-08-151-4/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ./src/H5Aproto.h ./src/H5Bproto.h ./src/H5Cproto.h ./src/H5Dproto.h ./src/H5Eproto.h ./src/H5Fproto.h ./src/H5Gproto.h ./src/H5Hproto.h ./src/H5MFproto.h ./src/H5MMproto.h ./src/H5Mproto.h ./src/H5Oproto.h ./src/H5Pproto.h ./src/H5Tproto.h ./src/H5proto.h These files were removed from the library and renamed by changing `proto' to `public'. ./src/H5ACpublic.h NEW ./src/H5Apublic.h NEW ./src/H5Bpublic.h NEW ./src/H5Cpublic.h NEW ./src/H5Dpublic.h NEW ./src/H5Epublic.h NEW ./src/H5Fpublic.h NEW ./src/H5Gpublic.h NEW ./src/H5Hpublic.h NEW ./src/H5MFpublic.h NEW ./src/H5MMpublic.h NEW ./src/H5Mpublic.h NEW ./src/H5Opublic.h NEW ./src/H5Ppublic.h NEW ./src/H5Tpublic.h NEW ./src/H5public.h NEW These files came from the old H5*proto.h files. ./src/Makefile ./src/Makefile.in NEW Removed. Now generated automatically from Makefile.in by running configure. ./src/h5oplat.h ./src/hdf5fort.h ./src/hdf5gen.h ./src/hdf5glob.h ./src/hdf5lims.h ./src/hdf5meta.h ./src/hdf5pabl.h ./src/hdf5plat.h ./src/hdf5port.h ./src/hdf5type.h Removed. The contents of these files has moved to other header files or source files depending on it's nature. ./src/H5.c ./src/H5A.c ./src/H5B.c ./src/H5C.c ./src/H5D.c ./src/H5E.c ./src/H5F.c ./src/H5G.c ./src/H5H.c ./src/H5M.c ./src/H5MF.c ./src/H5MM.c ./src/H5O.c ./src/H5Ocont.c ./src/H5Oname.c ./src/H5Onull.c ./src/H5Ostab.c ./src/H5P.c ./src/H5T.c Fixed include files. Moved some things from old headers into these files. ./src/H5ACprivate.h ./src/H5Aprivate.h ./src/H5Bprivate.h ./src/H5Cprivate.h ./src/H5Dprivate.h ./src/H5Eprivate.h ./src/H5Fprivate.h ./src/H5Gprivate.h ./src/H5Hprivate.h ./src/H5MFprivate.h ./src/H5MMprivate.h ./src/H5Mprivate.h ./src/H5Oprivate.h ./src/H5Pprivate.h ./src/H5Tprivate.h ./src/H5private.h ./src/debug.c Fixed include files. ./src/hdf5.h This is now the top-level *PUBLIC* include file. It should never appear in the library *.c files.
* [svn-r31] Added code for H5Mget_file & H5Mflush and re-targeted some of the ↵Quincey Koziol1997-08-151-3/+3
| | | | | | H5D calls to use them.
* [svn-r30] Intermediate checkin of (non-working) dataset code to allow for ↵Quincey Koziol1997-08-151-6/+75
| | | | | | source code rearrangement.
* [svn-r24] ./src/H5B.cRobb Matzke1997-08-111-1/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ./src/H5Bprivate.h ./src/H5G.c H5B_new() only takes two arguments. ./src/H5C.c Fixed initializer for default_file_create to be the same order as it's definition. ./src/H5D.c Changed commented-out code to #if 0/#endif. #if'd out a variable that was unused because of #if'd out code. ./src/H5Dproto.c Added a prototype for H5Dwrite(). ./src/H5F.c Changed parentheses to fix precedence problems from: flags & (~H5AC_OVERWRITE) != 0 to (flags & ~H5AC_OVERWRITE) != 0 ./src/H5H.c Heap reallocation doubles the heap size instead of tripling it. ./src/H5O.c Got rid of a statement with no effect. ./src/hdf5port.h Indented a few #defines. Added a defn for MAX3().
* [svn-r17] ./src/H5AC.cRobb Matzke1997-08-091-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Renamed H5AC_find() to H5AC_find_f() which is invoked from the H5AC_find() macro. Changed HASH() to H5AC_HASH(). ./src/H5ACprivate.h Increased the number of cache slots from 1033 to 10330 to see how it affects performance. This should probably be changeable on a per-file basis. ./src/H5B.c Fixed some bugs now that symbol tables are actually using the stuff. Improved debugging a little. Fixed uninitialized memory appearing in the file. ./src/H5D.c Changed a FUNC_ENTER() argument from H5Dset_info to H5Dwrite. ./src/H5F.c Plugged a memory leak in H5F_dest(). ./src/H5G.c ./src/H5Gprivate.h Many of these functions take a symbol table entry which describes the symbol table rather than just the symbol table address. Moved some functions to make room for the ones that understand directory names: H5G_new() -> H5G_stab_new() H5G_find() -> H5G_stab_find() H5G_modify() -> H5G_stab_modify() H5G_insert() -> H5G_stab_insert() H5G_list() -> H5G_stab_list() ./src/H5Gnode.c Added more assertions. Zero new memory so junk doesn't appear in the data file. This is a problem when one tries to declassify a classified data file. H5G_node_debug() can take an extra argument which is the address of the heap for the symbol table. If supplied, the symbol names are printed along with the heap offsets. ./src/H5H.c ./src/H5Hprivate.h Fixed a bug with the free blocks by forcing things to align on even boundaries. It's still possible to lose heap memory if the hole is smaller than the free list header. The alternative is to align heap objects on 12 or 20 byte boundaries, but this tends to waste to much space. Zero new memory like with H5Gnode.c. Fixed a "mispeling" of NDEBUG. ./src/H5Ocont.c ./src/H5Onull.c Added the pablo mask defn. ./src/H5Ostab.c Added the pablo mask defn. Fixed a read from uninitialized memory. ./src/H5Oprivate.h Tuned some parameters to see how they affect performance. ./src/debug.c An optional third argument can be supplied which is the address of a heap to use to display names in a symbol table. ./src/hdf5pabl.h Changed `int' to `intn'.
* [svn-r14] Preliminary checkin of dataset code, to allow the interface & code ↵Quincey Koziol1997-08-061-4/+128
| | | | | | | to be shared. Everything except the H5Dwrite should be working (or have sections of code commented about features to implement).
* [svn-r13] ./src/H5.cRobb Matzke1997-08-051-17/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Added an `_g' to the end of library_initialize, thread_initialize, and interface_initialize to abide by the naming convention. Removed setting of these variables from the various initialization functions since it happens in the FUNC_ENTER() macro now. Defined PABLO_MASK. Removed `CONSTR(FUNC,"function_name")' everywhere since it's handled by the various FUNC_ENTER macros. Fixed calls to FUNC_ENTER(), FUNC_LEAVE(), and HRETURN_ERROR() so they don't need so many arguments. Changed PABLO_TRACE_ON() to FUNC_ENTER() since there is no longer any danger of infinite recursion. H5_term_library() now returns SUCCEED/FAIL and uses FUNC_ENTER/FUNC_EXIT macros. ./src/H5A.c Changes similar to H5.c. Most (all?) of the functions called PABLO_TRACE_ON() and the package doesn't have an interface initializer that I can see, so the second argument to FUNC_ENTER() is always NULL. H5A_release_atom_node() returns SUCCEED/FAIL. ./src/H5AC.c Added error handling. Arguments for internal functions are checked with assert(). ./src/H5C.c Changes similar to H5.c Fixed the FUNC variable in H5C_get_default_atom() since it was initialized to the wrong name. ./src/H5D.c Changes similar to H5.c ./src/H5E.c Changes similar to H5.c Changed the pablo mask from H5_mask to H5E_mask in H5E_init_interface(). H5Eclear(), H5E_store(), and H5Epush() return SUCCEED/FAIL. Changed PABLO_TRACE_OFF() calls to FUNC_LEAVE() calls in the same functions. ./src/H5Eprivate.h ./src/H5Eproto.h Added additional error symbols for the H5AC package. Changed prototypes for H5Eclear() and H5Epush(). Changes to HRETURN_ERROR() and HGOTO_ERROR() to reduce the number of arguments. ./src/H5F.c Changes similr to H5.c Changed the pablo mask from H5_mask to H5F_mask for the FUNC_LEAVE() call of H5F_init_interface(). Added FUNC_ENTER() and FUNC_LEAVE() calls to some functions that didn't have them. ./src/H5M.c Changes similar to H5.c Fixed the FUNC variable in H5M_init_interface() since it was initialized to the wrong name. ./src/H5P.c Changes similar to H5.c ./src/H5T.c Changes similar to H5.c ./src/hdf5gen.c Changes to FUNC_ENTER() and FUNC_EXIT() to reduce the number of arguments. FUNC_ENTER() is now safe from infinite recursion since it updates the library_initialize_g, thread_initialize_g, or interface_initialize_g variables before calling the appropriate initialize function.
* [svn-r9] Added framework code for creating datasets and hooked them into theQuincey Koziol1997-08-011-0/+162
"meta-object" interface
generated by cgit v0.12 at 2025-02-22 06:25:10 (GMT)