summaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* [svn-r2559] Purpose:Thomas Radke2000-09-151-8/+8
| | | | | | | | Add the Stream VFD sources to the appropriate makefile variables. Description: Added H5FDstream.c to the LIB_SRC variable and H5FDstream.h to the PUB_HDR variable for building the Stream VFD.
* [svn-r2558] Purpose:Thomas Radke2000-09-151-0/+3
| | | | | | | | | | Define HAVE_STREAM. Description: If the Stream VFD was configured the configured script will expand this into '#define HAVE_STREAM 1' in H5config.h and '#define H5_STREAM 1' in H5pubconf.h.
* [svn-r2557] Purpose:Thomas Radke2000-09-151-0/+7
| | | | | | | Added registration of the Stream Virtual File Driver. Description: The Stream VFD is registered here if it was configured.
* [svn-r2555] Purpose:Thomas Radke2000-09-152-0/+1037
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Added source files for the Stream Virtual File Driver. Description: The Stream VFD allows users to stream complete HDF5 files via socket connections between different applications. Files which were created anew are flushed to any connected client on each H5Fflush() or H5Fclose() operation. Files which are opened as read-only will be read from a socket on a H5Fopen() call. The driver's H5FDset_fapl_stream() routine allows to pass in several parameters such as an external socket descriptor, some socket options, and flags for broadcasting a received file. If an external socket is provided the Stream VFD would use that for the socket calls. Otherwise it parses the filename argument in H5Fcreate()/H5Fopen() for a 'hostname::port' parameter. All files processed by the Stream VFD are kept in memory (same way as the core VFD does). Platforms: Tested so far under Linux, Irix 32/64bit, OSF1, Solaris, Cray Unicos, Hitachi SR8000, IBM AIX. Not tested under Windows yet.
* [svn-r2551] Purpose:Quincey Koziol2000-09-141-76/+180
| | | | | | | | | | | | | | | | Fix last couple of errors from introducing "regular" hyperslab feature into the library. Description: Code was blindly dereferencing data structures which aren't defined when operating on regular hyperslabs. Solution: Check for regular hyperslab defined and retrieve information from regular hyperslab info instead of mucking about in other hyperslab information. Platforms: Solaris 2.6
* [svn-r2548] Added #include <sys/resource.h> for the struct getrusage ↵Bill Wendling2000-09-141-24/+28
| | | | | | | declaration. Przemek Klosowski, Ph.D. <przemek@nist.gov> reported this problem on Linux RH 6.1 systems.
* [svn-r2546] Further patches to accomodate regular hyperslabs. More of the ↵Quincey Koziol2000-09-132-50/+201
| | | | | | tests are passing, but not all of them yet...
* [svn-r2544] Optimized regular hyperslab I/O routines and data structures. ↵Quincey Koziol2000-09-132-312/+1886
| | | | | | | | | | | | | On my benchmarks, they are about 4-5 times faster than before. We no longer generate "general" hyperslab data structures for regular hyperslabs, the general data structures are only generated when needed for irregular hyperslabs. Still fixing a couple of nook-and-cranny functions to understand the new information for the regular hyperslabs, so the tests aren't completely passing, but I wanted to get this checked in for Elena's benchmarks. I should have more/all tests passing later today.
* [svn-r2519] Snapshot version 1.3 release 29Albert Cheng2000-09-062-1/+3
|
* [svn-r2518] Backed out patch from user after further testing...Quincey Koziol2000-09-061-3/+4
|
* [svn-r2517] Removed the incorrect #ifdef's. Since this file is generated fromBill Wendling2000-09-061-2/+0
| | | | autoheader, it was useless to add them in here...
* [svn-r2515] Added macros around the inline #define since this is a keyword ↵Bill Wendling2000-09-061-0/+2
| | | | | | in C++ and, when adding the header to a C++ program, breaks it.
* [svn-r2511] Fix point selections so they work properly with both staticly ↵Quincey Koziol2000-09-061-4/+3
| | | | | | and dynamicly allocated memory.
* [svn-r2510] Took memory allocation out of loop and used some temporary ↵Quincey Koziol2000-09-061-8/+21
| | | | | | | | variable instead of multi-dereferenced pointer chains. This buys us another ~20% improvement in the hyperslab I/O speed. (From ~30 seconds to ~25 seconds on the h5hypers benchmark)
* [svn-r2509] Change code so that heap_size wasn't zero.Quincey Koziol2000-09-062-5/+3
|
* [svn-r2508] Purpose:Albert Cheng2000-09-064-9/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Bug fixes Description: All tests were core=dumping in IRIX64. The bug is in Generic property list creation in which malloc asked for 2*64-1 bytes due to coding bug. The object creation failed but the return code was not checked. Program eventually crashed. Solution: H5F.c: Check the return code from new file object creation and flag error accordingly. H5FL.c: H5FL_arr_free is a replacement for H5MM_xfree which accepts null value as a legal argument value. H5FL_arr_free assert on it. Since other parts of the code have been passing null value to H5MM_xfree, H5FL_arr_free must accept it too until all the calling routines are changed to not pass Null. H5P.c: some routine passes in 0 as the hashsize value which is uintn. The expression (hashsize-1) underflows to the largest unsigned int for some machines. Thus the calloc failed. Cast hashsize to unsigned int first (this assumes hashsize stays within the signed int data range. H5Smpio.c: Added the extra parameter because the H5FD_write has been redefined. Platforms tested: IRIX64 -64 and -n32
* [svn-r2505] Some of the RETURN macros were returning NULL instead of ↵Bill Wendling2000-09-051-7/+7
| | | | FAIL...Fixed.
* [svn-r2504] Revised routine to add hyperslabs to the selection, sorting the ↵Quincey Koziol2000-09-041-78/+36
| | | | | | | | arrays of hyperslab boundaries after adding them all, instead of maintaining the sorted order during each addition. This boosts performance for sub-sampled (i.e. strided) hyperslabs by about a factor of 10! :-)
* [svn-r2502] Added limits to the free lists, so they will automatically ↵Quincey Koziol2000-08-311-6/+6
| | | | | | garbage collect and not eat all the memory in the machine (by default :-)
* [svn-r2501] Added several free lists for commonly used & abused data ↵Quincey Koziol2000-08-311-120/+525
| | | | | | | | | | | | | structures. Also added code to allow metadata to be allocated out of a more contiguous block ("metadata aggregation") and also code for "catching" small metadata write calls and building a buffer of the small pieces of metadata for later writing as one, larger, block ("metadata accumulation"). These features are enabled on a per VFL driver basis with the new VFL 'query' call and both currently enabled for the sec2, family and stdio drivers. The mpio VFL driver enables only the "metadata aggregation" code, not the "metadata accumulation" code. All the other drivers have these features turned off.
* [svn-r2500] Added fields to file data structures to allow metadata ↵Quincey Koziol2000-08-311-2/+34
| | | | | | aggregation and accumulation to be enabled and tracked. Also updated some prototypes.
* [svn-r2499] Changed H5FDwrite prototype to include new 'type of data' parameterQuincey Koziol2000-08-311-1/+1
|
* [svn-r2498] Added new VFL 'query' code and added new 'type of data' ↵Quincey Koziol2000-08-319-23/+172
| | | | | | parameter to write call.
* [svn-r2497] Removing metadata aggregation & accumulation prototyping code ↵Quincey Koziol2000-08-311-64/+50
| | | | | | (it's in H5FD.c now) and updated driver to add the new VFL 'query' call.
* [svn-r2496] Changed H5F_block_write prototype & added metadata allocation ↵Quincey Koziol2000-08-311-2/+3
| | | | | | field to file access property lists.
* [svn-r2495] More type of data parameter additions for H5F_block_write.Quincey Koziol2000-08-313-3/+3
|
* [svn-r2494] Changed the default size of object headers from 16 bytes to 256 ↵Quincey Koziol2000-08-311-1/+1
| | | | bytes.
* [svn-r2493] Added a set/get pair of property modifiers to allows users to ↵Quincey Koziol2000-08-312-0/+84
| | | | | | | set/query the metadata allocation size for file access property lists. These are new API functions and should be documented for the next release.
* [svn-r2492] Added type of data parameter to H5F_block_write calls and also ↵Quincey Koziol2000-08-311-5/+7
| | | | | | added the metadata allocation size to the default file access property lists.
* [svn-r2491] Added type of data parameter to H5F_block_write calls.Quincey Koziol2000-08-316-10/+10
|
* [svn-r2477] Snapshot version 1.3 release 28HDF Admin2000-08-191-1/+1
|
* [svn-r2476] Added free lists to track various data structures in memory, to ↵Quincey Koziol2000-08-169-67/+160
| | | | | | reduce malloc abuse.
* [svn-r2475] Moved code to insert a new dataset into a group up a few blocks ↵Quincey Koziol2000-08-161-5/+23
| | | | | | of code, to get the metadata allocated in the file more tightly together.
* [svn-r2474] Updated some old variables to the proper names, added seek ↵Quincey Koziol2000-08-162-10/+81
| | | | | | logging and added test code for aggregating metadata into more localized locations in the file.
* [svn-r2472] Combine object header prefix and first object header chunk when ↵Quincey Koziol2000-08-161-20/+55
| | | | possible.
* [svn-r2471] Snapshot version 1.3 release 27HDF Admin2000-08-121-1/+1
|
* [svn-r2470] Changed parameter type from uintn to size_t to clear up some ↵Quincey Koziol2000-08-082-4/+4
| | | | warnings.
* [svn-r2469] Snapshot version 1.3 release 26HDF Admin2000-08-051-1/+1
|
* [svn-r2465] Changed nbytes to an hsize_t type since that's what it holds. ↵Bill Wendling2000-08-041-9/+9
| | | | | | Also, moved it into the if-then statement to limit it's scope.
* [svn-r2464] HGOTO_ERROR was returning NULL instead of FAIL. This was causing ↵Bill Wendling2000-08-041-14/+13
| | | | | | compiler warnings about "implicit cast of pointer to integer" things....
* [svn-r2459] Added public API function (H5set_free_list_limits) to allow ↵Quincey Koziol2000-08-044-12/+116
| | | | | | | | limits to be placed on how much memory is used by the free lists before they are garbage collected. The default is to have no limit, with garbage collection only occurring when they cannot allocate memory.
* [svn-r2454] Took out another unneeded assertion which was causing problems.Quincey Koziol2000-08-011-1/+0
|
* [svn-r2453] Made free-list code garbage collect if it runs out memory during ↵Quincey Koziol2000-08-012-83/+121
| | | | | | | a memory allocation & attempt to allocate the memory once more. Also re-named a bunch of private functions & structures to align with other function names.
* [svn-r2450] Remove asserts on non-existent variables..Quincey Koziol2000-08-011-2/+1
|
* [svn-r2448] Re-did section of hyperslab initialization code to avoid tons of ↵Quincey Koziol2000-07-282-116/+111
| | | | | | realloc() calls. Speeds user's test code up by ~25 times... :-)
* [svn-r2447] Another small calculation optimization.Quincey Koziol2000-07-281-1/+1
|
* [svn-r2446] Some small performance improvements to various allocations and ↵Quincey Koziol2000-07-282-22/+76
| | | | calculations.
* [svn-r2445] Check in initial coding effort for automatically garbage ↵Quincey Koziol2000-07-272-157/+350
| | | | | | | | collecting the free lists. Each kind of list one has hard-coded limits on when to garbage collect, which will be replaced with user-controllable knobs (through property list settings, I think) once I finish debugging some related performance problems.
* [svn-r2444] Checkpoint the Generic Property implementation. It's not active ↵Quincey Koziol2000-07-2510-340/+2642
| | | | | | yet, but it's close. This shouldn't interfere with other development work. (I think.. :-)
* [svn-r2442] Insert better error checking when memory allocation fails (in ↵Quincey Koziol2000-07-251-2/+4
| | | | | | certain obscure situations).
generated by cgit v0.12 at 2024-11-08 14:29:14 (GMT)