mirror/hdf5
2
0
Fork 0
mirror of https://github.com/HDFGroup/hdf5.git synced 2024-12-15 07:40:23 +08:00
Code Issues Packages Projects Releases Wiki Activity
4c34ec0668
hdf5/test/vfd.c

1936 lines
61 KiB
C
Raw Normal View History

[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
[svn-r13253] Updated all C and C++ style source code files with the THG copyright notice. Tested platform: Kagiso only since it is only a comment block change. If it works in one machine, it should work in all, I hope. Still need to check the parallel build on copper.
2007-02-07 22:56:24 +08:00
* Copyright by The HDF Group. *
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
* Copyright by the Board of Trustees of the University of Illinois. *
* All rights reserved. *
* *
* This file is part of HDF5. The full HDF5 copyright notice, including *
* terms governing use, modification, and redistribution, is contained in *
Change copyright headers to replace url referring to file to be removed and replace it with new url for COPYING file. Fix 2 lines in java error test expected output file where messages include line numbers changed by reducing the copyright header by 2 lines.
2017-04-15 00:54:16 +08:00
* the COPYING file, which can be found at the root of the source code *
* distribution tree, or in https://support.hdfgroup.org/ftp/HDF5/releases. *
* If you do not have access to either file, you may request a copy from *
* help@hdfgroup.org. *
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
/*
* Programmer: Raymond Lu<slu@ncsa.uiuc.edu>
* Tuesday, Sept 24, 2002
*
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
* Purpose: Tests the basic features of Virtual File Drivers
*/
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
#include "h5test.h"
[svn-r13588] A support of files bigger than 2GB for STDIO driver. Configure will check if fseeko is available. Using it instead of fseek can support big files because the offset is of type off_t not long int. Also added the test for STDIO in big.c.
2007-04-05 03:59:00 +08:00
#define KB 1024U
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
#define FAMILY_NUMBER 4
#define FAMILY_SIZE (1*KB)
#define FAMILY_SIZE2 (5*KB)
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
#define MULTI_SIZE 128
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
#define CORE_INCREMENT (4*KB)
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
#define CORE_PAGE_SIZE (1024*KB)
#define CORE_DSET_NAME "core dset"
#define CORE_DSET_DIM1 1024
#define CORE_DSET_DIM2 32
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
#define DSET1_NAME "dset1"
#define DSET1_DIM1 1024
#define DSET1_DIM2 32
#define DSET3_NAME "dset3"
[svn-r12822] H5Pset_alignment seems working. Added it in test/vfd.c and made the test more complete by writing two data sets. One is aligned, and the other isn't.
2006-10-29 05:11:04 +08:00
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
/* Macros for Direct VFD */
#ifdef H5_HAVE_DIRECT
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
#define MBOUNDARY 512
#define FBSIZE (4*KB)
#define CBSIZE (8*KB)
#define THRESHOLD 1
#define DSET2_NAME "dset2"
#define DSET2_DIM 4
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
#endif /* H5_HAVE_DIRECT */
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
const char *FILENAME[] = {
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
"sec2_file", /*0*/
"core_file", /*1*/
"family_file", /*2*/
"new_family_v16_", /*3*/
"multi_file", /*4*/
"direct_file", /*5*/
"log_file", /*6*/
"stdio_file", /*7*/
"windows_file", /*8*/
[svn-r21015] Bug fix for Issue 2598 - In v1.6 library, there was EOA for the whole MULTI file saved in the super block. We took it out in v1.8 library because it's meaningless for the MULTI file. v1.8 library saves the EOA for the metadata file, instead. But this caused some backward compatibility problem. v1.8 library couldn't open the file created with v1.6 library. I fixed the problem by checking the EOA value to detect the file created with v1.6 library. Tested on jam, koala, and heiwa.
2011-06-22 23:47:57 +08:00
"new_multi_file_v16",/*9*/
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
NULL
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
};
[svn-r16779] A minor correction of previous checkin. The original files created with v1.6 library could be deleted by cleanup routine. I corrected it in this checkin. Tested on jam - simple change.
2009-04-17 23:52:38 +08:00
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
#define LOG_FILENAME "log_vfd_out.log"
[svn-r14340] Maintenance for VMS system: Skip one of the tests in flush2.c since it looks like VMS system flsuhes the file anyway (like on Winodws with DLLs) Renamed files family*.h5 that are used in backward compatibility testing to have only one '.' in a file name, otherwise VMS system becomes confused and test fails. Cleaned up the only test (Fortran multi-file test) that is failing on VMS (C counterpart passesa). Added new script to facilitate the build process on VMS Updated MANIFEST PLatforms tested: VMS server, kagiso and linew (minor changes)
2007-12-14 02:17:02 +08:00
#define COMPAT_BASENAME "family_v16_"
[svn-r21015] Bug fix for Issue 2598 - In v1.6 library, there was EOA for the whole MULTI file saved in the super block. We took it out in v1.8 library because it's meaningless for the MULTI file. v1.8 library saves the EOA for the metadata file, instead. But this caused some backward compatibility problem. v1.8 library couldn't open the file created with v1.6 library. I fixed the problem by checking the EOA value to detect the file created with v1.6 library. Tested on jam, koala, and heiwa.
2011-06-22 23:47:57 +08:00
#define MULTI_COMPAT_BASENAME "multi_file_v16"
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/*-------------------------------------------------------------------------
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
* Function: test_sec2
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
*
* Purpose: Tests the file handle interface for SEC2 driver
*
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
* Return: Success: 0
* Failure: -1
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
*
* Programmer: Raymond Lu
* Tuesday, Sept 24, 2002
*
*-------------------------------------------------------------------------
*/
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
static herr_t
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
test_sec2(void)
{
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
hid_t fid = -1; /* file ID */
hid_t fapl_id = -1; /* file access property list ID */
hid_t fapl_id_out = -1; /* from H5Fget_access_plist */
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
hid_t driver_id = -1; /* ID for this VFD */
unsigned long driver_flags = 0; /* VFD feature flags */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
char filename[1024]; /* filename */
void *os_file_handle = NULL; /* OS file handle */
hsize_t file_size; /* file size */
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
TESTING("SEC2 file driver");
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Set property list and file name for SEC2 driver. */
[svn-r30184] Converted h5_fileaccess() to H5Pcreate() in the sec2 test. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only w/ various VFDs
2016-07-14 21:55:50 +08:00
if((fapl_id = H5Pcreate(H5P_FILE_ACCESS)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5Pset_fapl_sec2(fapl_id) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
h5_fixname(FILENAME[0], fapl_id, filename, sizeof(filename));
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
/* Check that the VFD feature flags are correct */
if ((driver_id = H5Pget_driver(fapl_id)) < 0)
TEST_ERROR
if (H5FDdriver_query(driver_id, &driver_flags) < 0)
TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_ACCUMULATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_DATA_SIEVE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_SMALLDATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_POSIX_COMPAT_HANDLE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_SUPPORTS_SWMR_IO)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_DEFAULT_VFD_COMPATIBLE)) TEST_ERROR
/* Check for extra flags not accounted for above */
if(driver_flags != (H5FD_FEAT_AGGREGATE_METADATA
| H5FD_FEAT_ACCUMULATE_METADATA
| H5FD_FEAT_DATA_SIEVE
| H5FD_FEAT_AGGREGATE_SMALLDATA
| H5FD_FEAT_POSIX_COMPAT_HANDLE
| H5FD_FEAT_SUPPORTS_SWMR_IO
| H5FD_FEAT_DEFAULT_VFD_COMPATIBLE))
TEST_ERROR
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if((fid = H5Fcreate(filename, H5F_ACC_TRUNC, H5P_DEFAULT, fapl_id)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Retrieve the access property list... */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if((fapl_id_out = H5Fget_access_plist(fid)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
/* Check that the driver is correct */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5FD_SEC2 != H5Pget_driver(fapl_id_out))
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* ...and close the property list */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5Pclose(fapl_id_out) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Check that we can get an operating-system-specific handle from
* the library.
*/
if(H5Fget_vfd_handle(fid, H5P_DEFAULT, &os_file_handle) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(os_file_handle == NULL)
FAIL_PUTS_ERROR("NULL os-specific vfd/file handle was returned from H5Fget_vfd_handle");
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* There is no garantee the size of metadata in file is constant.
* Just try to check if it's reasonable.
*
* Currently it should be around 2 KB.
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
*/
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5Fget_filesize(fid, &file_size) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
if(file_size < 1 * KB || file_size > 4 * KB)
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
FAIL_PUTS_ERROR("suspicious file size obtained from H5Fget_filesize");
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
/* Close and delete the file */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5Fclose(fid) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
h5_delete_test_file(FILENAME[0], fapl_id);
/* Close the fapl */
if(H5Pclose(fapl_id) < 0)
TEST_ERROR;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
PASSED();
return 0;
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
error:
H5E_BEGIN_TRY {
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
H5Pclose(fapl_id);
H5Pclose(fapl_id_out);
H5Fclose(fid);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
} H5E_END_TRY;
return -1;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
} /* end test_sec2() */
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
/*-------------------------------------------------------------------------
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
* Function: test_core
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
*
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
* Purpose: Tests the file handle interface for CORE driver
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
*
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
* Return: Success: 0
* Failure: -1
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
*
* Programmer: Raymond Lu
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
* Tuesday, Sept 24, 2002
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
*
*-------------------------------------------------------------------------
*/
static herr_t
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
test_core(void)
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
{
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
hid_t fid = -1; /* file ID */
hid_t fapl_id = -1; /* file access property list ID */
hid_t fapl_id_out = -1; /* from H5Fget_access_plist */
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
hid_t driver_id = -1; /* ID for this VFD */
unsigned long driver_flags = 0; /* VFD feature flags */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
hid_t did = -1; /* dataset ID */
hid_t sid = -1; /* dataspace ID */
char filename[1024]; /* filename */
void *os_file_handle = NULL; /* OS file handle */
hsize_t file_size; /* file size */
size_t increment; /* core VFD increment */
hbool_t backing_store; /* use backing store? */
hbool_t use_write_tracking; /* write tracking flag */
size_t write_tracking_page_size; /* write tracking page size */
int *data_w = NULL; /* data written to the dataset */
int *data_r = NULL; /* data read from the dataset */
int val; /* data value */
int *pw = NULL, *pr = NULL; /* pointers for iterating over
data arrays (write & read) */
hsize_t dims[2]; /* dataspace dimensions */
int i, j; /* iterators */
htri_t status; /* return value from H5Lexists */
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
TESTING("CORE file driver");
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Get a file access property list and fix up the file name */
[svn-r30163] Changed a h5_fileaccess() call to H5Pcreate() in the core VFD test. Since it only tests the core VFD, there's no need to get a VFD-dependent fapl. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 vfd test w/ all VFDs tested by make check-vfd
2016-07-09 14:43:11 +08:00
if((fapl_id = H5Pcreate(H5P_FILE_ACCESS)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
h5_fixname(FILENAME[1], fapl_id, filename, sizeof(filename));
/************************************************************************
* Check that the backing store flag works by creating a file, close
* it, and ensure that the file does not exist.
************************************************************************/
/* Make sure it's not present at the start of the test */
if(HDaccess(filename, F_OK) != -1)
if(HDremove(filename) < 0)
FAIL_PUTS_ERROR("unable to remove backing store file");
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Create and close file w/ backing store off */
if(H5Pset_fapl_core(fapl_id, (size_t)CORE_INCREMENT, FALSE) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
/* Check that the VFD feature flags are correct.
* Note that the H5FDdriver_query() API call does not require a file
* so backing-store related flags will not be returned here.
*/
if ((driver_id = H5Pget_driver(fapl_id)) < 0)
TEST_ERROR
if (H5FDdriver_query(driver_id, &driver_flags) < 0)
TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_ACCUMULATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_DATA_SIEVE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_SMALLDATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_ALLOW_FILE_IMAGE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_CAN_USE_FILE_IMAGE_CALLBACKS)) TEST_ERROR
/* Check for extra flags not accounted for above */
if(driver_flags != (H5FD_FEAT_AGGREGATE_METADATA
| H5FD_FEAT_ACCUMULATE_METADATA
| H5FD_FEAT_DATA_SIEVE
| H5FD_FEAT_AGGREGATE_SMALLDATA
| H5FD_FEAT_ALLOW_FILE_IMAGE
| H5FD_FEAT_CAN_USE_FILE_IMAGE_CALLBACKS))
TEST_ERROR
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if((fid = H5Fcreate(filename, H5F_ACC_TRUNC, H5P_DEFAULT, fapl_id)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5Fclose(fid) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Check for the backing store file */
if(HDaccess(filename, F_OK) != -1)
FAIL_PUTS_ERROR("file created when backing store set to FALSE");
/************************************************************************
* Check basic core VFD operation and properties. This is done with the
* backing store on so a file will be created for later use.
************************************************************************/
/* Turn the backing store on */
if(H5Pset_fapl_core(fapl_id, (size_t)CORE_INCREMENT, TRUE) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Check that write tracking is off by default and that the default
* page size is non-zero.
*/
if(H5Pget_core_write_tracking(fapl_id, &use_write_tracking, &write_tracking_page_size) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(FALSE != use_write_tracking)
FAIL_PUTS_ERROR("write tracking should be off by default");
if(0 == write_tracking_page_size)
FAIL_PUTS_ERROR("write tracking page size should never be zero");
[svn-r12776] The internal change is to make the 3 parameters of H5Pset_fapl_direct be a structure and passed through driver info functions in H5FD_direct_open, H5FD_direct_read, and H5FD_direct_write. The external change is to add a new API function H5Pget_fapl_direct to query 3 control values, the memory boundary, file system block size, and the maximal buffer size for copying data.
2006-10-19 04:20:32 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Set core VFD properties */
if(H5Pset_core_write_tracking(fapl_id, TRUE, CORE_PAGE_SIZE) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Create the file */
if((fid = H5Fcreate(filename, H5F_ACC_TRUNC, H5P_DEFAULT, fapl_id)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r15628] Description: Remove trailing whitespace from C/C++ source files, with the following script: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Tested on: Mac OS X/32 10.5.5 (amazon) No need for h5committest, just whitespace changes...
2008-09-16 23:52:51 +08:00
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
/* Retrieve the access property list... */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if((fapl_id_out = H5Fget_access_plist(fid)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
/* Check that the driver is correct */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5FD_CORE != H5Pget_driver(fapl_id_out))
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Get the basic VFD properties from the fapl and ensure that
* they are correct.
*/
if(H5Pget_fapl_core(fapl_id_out, &increment, &backing_store) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(increment != (size_t)CORE_INCREMENT)
FAIL_PUTS_ERROR("incorrect increment from file fapl");
if(backing_store != TRUE)
FAIL_PUTS_ERROR("incorrect backing store flag from file fapl");
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
/* Check that the backing store write tracking info was saved */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* TODO: There is a bug where H5Fget_access_plist() does not return
* the write tracking properties. Until this bug is fixed, just
* test the main fapl_id.
*/
if(H5Pget_core_write_tracking(fapl_id, &use_write_tracking, &write_tracking_page_size) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if(TRUE != use_write_tracking)
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
FAIL_PUTS_ERROR("write tracking flag incorrect in fapl obtained from H5Fget_access_plist");
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if(CORE_PAGE_SIZE != write_tracking_page_size)
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
FAIL_PUTS_ERROR("write tracking page size incorrect in fapl obtained from H5Fget_access_plist");
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Close the property list */
if(H5Pclose(fapl_id_out) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Check that we can get an operating-system-specific handle from
* the library.
*/
if(H5Fget_vfd_handle(fid, H5P_DEFAULT, &os_file_handle) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(os_file_handle == NULL)
FAIL_PUTS_ERROR("NULL os-specific vfd/file handle was returned from H5Fget_vfd_handle");
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
/* There is no garantee the size of metadata in file is constant.
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
* Just try to check if it's reasonable.
*
* TODO: Needs justification of why is this is a reasonable size.
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
*/
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5Fget_filesize(fid, &file_size) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(file_size < 2 * KB || file_size > 6 * KB)
FAIL_PUTS_ERROR("suspicious file size obtained from H5Fget_filesize");
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Close the file */
if(H5Fclose(fid) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/************************************************************************
* Make changes to the file with the backing store flag OFF to ensure
* that they ARE NOT propagated.
************************************************************************/
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Open the file with backing store off for read and write.
* Changes won't be saved in file.
*/
if(H5Pset_fapl_core(fapl_id, (size_t)CORE_INCREMENT, FALSE) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if((fid = H5Fopen(filename, H5F_ACC_RDWR, fapl_id)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
/* Allocate memory for data set. */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(NULL == (data_w = (int *)HDmalloc(DSET1_DIM1 * DSET1_DIM2 * sizeof(int))))
FAIL_PUTS_ERROR("unable to allocate memory for input array");
if(NULL == (data_r = (int *)HDmalloc(DSET1_DIM1 * DSET1_DIM2 * sizeof(int))))
FAIL_PUTS_ERROR("unable to allocate memory for output array");
/* Initialize the buffers */
val = 0;
pw = data_w;
for(i = 0; i < CORE_DSET_DIM1; i++)
for(j = 0; j < CORE_DSET_DIM2; j++)
*pw++ = val++;
HDmemset(data_r, 0, DSET1_DIM1 * DSET1_DIM2 * sizeof(int));
/* Create the dataspace */
dims[0] = CORE_DSET_DIM1;
dims[1] = CORE_DSET_DIM2;
if((sid = H5Screate_simple(2, dims, NULL)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Create the dataset */
if((did = H5Dcreate2(fid, CORE_DSET_NAME, H5T_NATIVE_INT, sid, H5P_DEFAULT, H5P_DEFAULT, H5P_DEFAULT)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Write the data to the dataset */
if(H5Dwrite(did, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, data_w) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Close and reopen the dataset */
if(H5Dclose(did) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if((did = H5Dopen2(fid, CORE_DSET_NAME, H5P_DEFAULT)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r12822] H5Pset_alignment seems working. Added it in test/vfd.c and made the test more complete by writing two data sets. One is aligned, and the other isn't.
2006-10-29 05:11:04 +08:00
/* Read the data back from dset1 */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5Dread(did, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, data_r) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r12822] H5Pset_alignment seems working. Added it in test/vfd.c and made the test more complete by writing two data sets. One is aligned, and the other isn't.
2006-10-29 05:11:04 +08:00
/* Check that the values read are the same as the values written */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
pw = data_w;
pr = data_r;
for(i = 0; i < CORE_DSET_DIM1; i++)
for(j = 0; j < CORE_DSET_DIM2; j++)
if(*pr++ != *pw++) {
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
H5_FAILED();
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
printf(" Read different values than written in data set.\n");
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
printf(" At index %d,%d\n", i, j);
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
} /* end if */
[svn-r12822] H5Pset_alignment seems working. Added it in test/vfd.c and made the test more complete by writing two data sets. One is aligned, and the other isn't.
2006-10-29 05:11:04 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Close everything except the dataspace ID (needed below)*/
if(H5Dclose(did) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5Fclose(fid) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r12822] H5Pset_alignment seems working. Added it in test/vfd.c and made the test more complete by writing two data sets. One is aligned, and the other isn't.
2006-10-29 05:11:04 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Reopen the file and ensure that the dataset does not exist */
if((fid = H5Fopen(filename, H5F_ACC_RDWR, fapl_id)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
status = H5Lexists(fid, CORE_DSET_NAME, H5P_DEFAULT);
if(status < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(status > 0)
FAIL_PUTS_ERROR("core VFD dataset created in file when backing store disabled");
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Close the file */
if(H5Fclose(fid) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/************************************************************************
* Make changes to the file with the backing store flag ON to ensure
* that they ARE propagated.
************************************************************************/
[svn-r12822] H5Pset_alignment seems working. Added it in test/vfd.c and made the test more complete by writing two data sets. One is aligned, and the other isn't.
2006-10-29 05:11:04 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Open the file with backing store on for read and write.
* Changes will be saved in file.
*/
if(H5Pset_fapl_core(fapl_id, (size_t)CORE_INCREMENT, TRUE) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if((fid = H5Fopen(filename, H5F_ACC_RDWR, fapl_id)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Create the dataset */
if((did = H5Dcreate2(fid, CORE_DSET_NAME, H5T_NATIVE_INT, sid, H5P_DEFAULT, H5P_DEFAULT, H5P_DEFAULT)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Write the data to the dataset */
if(H5Dwrite(did, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, data_w) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Close everything and reopen */
if(H5Dclose(did) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5Fclose(fid) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if((fid = H5Fopen(filename, H5F_ACC_RDWR, fapl_id)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if((did = H5Dopen2(fid, CORE_DSET_NAME, H5P_DEFAULT)) < 0)
TEST_ERROR;
/* Read the data back from the dataset */
HDmemset(data_r, 0, DSET1_DIM1 * DSET1_DIM2 * sizeof(int));
if(H5Dread(did, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, data_r) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r12822] H5Pset_alignment seems working. Added it in test/vfd.c and made the test more complete by writing two data sets. One is aligned, and the other isn't.
2006-10-29 05:11:04 +08:00
/* Check that the values read are the same as the values written */
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
pw = data_w;
pr = data_r;
for(i = 0; i < CORE_DSET_DIM1; i++)
for(j = 0; j < CORE_DSET_DIM2; j++)
if(*pw++ != *pr++) {
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
H5_FAILED();
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
printf(" Read different values than written in data set.\n");
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
printf(" At index %d,%d\n", i, j);
TEST_ERROR;
} /* end if */
[svn-r12822] H5Pset_alignment seems working. Added it in test/vfd.c and made the test more complete by writing two data sets. One is aligned, and the other isn't.
2006-10-29 05:11:04 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Check file size API.
* There is no garantee the size of metadata in file is constant.
* Just try to check if it's reasonable.
*
* TODO: Needs justification of why is this is a reasonable size.
*/
if(H5Fget_filesize(fid, &file_size) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(file_size < 64 * KB || file_size > 256 * KB)
FAIL_PUTS_ERROR("suspicious file size obtained from H5Fget_filesize");
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
/* Close everything */
if(H5Sclose(sid) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(H5Dclose(did) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
HDfree(data_w);
HDfree(data_r);
[svn-r12822] H5Pset_alignment seems working. Added it in test/vfd.c and made the test more complete by writing two data sets. One is aligned, and the other isn't.
2006-10-29 05:11:04 +08:00
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
/* Close and delete the file */
if(H5Fclose(fid) < 0)
TEST_ERROR;
h5_delete_test_file(FILENAME[1], fapl_id);
/* Close the fapl */
if(H5Pclose(fapl_id) < 0)
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
PASSED();
return 0;
error:
H5E_BEGIN_TRY {
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
H5Sclose(sid);
H5Dclose(did);
H5Pclose(fapl_id_out);
H5Pclose(fapl_id);
H5Fclose(fid);
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
} H5E_END_TRY;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
if(data_w)
HDfree(data_w);
if(data_r)
HDfree(data_r);
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
return -1;
[svn-r30148] Expanded core VFD test in vfd.c. Minor tidying to sec2 VFD test also. Tested on: 64-bit Ubuntu Linux w/ gcc 5.3.1 Autotools serial
2016-07-06 09:40:31 +08:00
} /* end test_core() */
[svn-r12739] Added Direct I/O driver to VFD. It's only supported by SGI Altix (cobalt). There's a configure option --enable-direct-vfd/--disable-direct-vfd to enable/disable Direct I/O support. The default is enabled. There's a small test in test/vfd.c. Another way to test it is to set environment variable HDF5_DRIVER to "direct" and run "make check" in the test/ directory. There'll be some further improvement in the following checkin including allowing user to provide memory boundary value, file block size, and copying buffer size.
2006-10-11 04:07:16 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/*-------------------------------------------------------------------------
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
* Function: test_direct
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
*
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
* Purpose: Tests the file handle interface for DIRECT I/O driver
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
*
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
* Return: Success: 0
* Failure: -1
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
*
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
* Programmer: Raymond Lu
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
* Wednesday, 20 September 2006
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
*
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
*-------------------------------------------------------------------------
*/
static herr_t
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
test_direct(void)
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
{
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
#ifdef H5_HAVE_DIRECT
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
hid_t file=-1, fapl=-1, access_fapl = -1;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
hid_t dset1=-1, dset2=-1, space1=-1, space2=-1;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
char filename[1024];
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
int *fhandle=NULL;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
hsize_t file_size;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
hsize_t dims1[2], dims2[1];
size_t mbound;
size_t fbsize;
size_t cbsize;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
int *points = NULL, *check = NULL, *p1 = NULL, *p2 = NULL;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
int wdata2[DSET2_DIM] = {11,12,13,14};
int rdata2[DSET2_DIM];
[svn-r21366] JIRA HDFFV-2748 replace use of _WIN32 in source. Created new defines H5_HAVE_WIN32_API and H5_HAVE_VISUAL_STUDIO defines to use. These can be properly set during configurration. Tested: windows and local linux - reviewed internally
2011-09-07 00:50:32 +08:00
int i, j, n;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
#endif /*H5_HAVE_DIRECT*/
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
TESTING("DIRECT I/O file driver");
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
#ifndef H5_HAVE_DIRECT
SKIPPED();
return 0;
#else /*H5_HAVE_DIRECT*/
/* Set property list and file name for Direct driver. Set memory alignment boundary
* and file block size to 512 which is the minimum for Linux 2.6. */
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
if((fapl = H5Pcreate(H5P_FILE_ACCESS)) < 0)
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if(H5Pset_fapl_direct(fapl, MBOUNDARY, FBSIZE, CBSIZE) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
h5_fixname(FILENAME[5], fapl, filename, sizeof filename);
/* Verify the file access properties */
if(H5Pget_fapl_direct(fapl, &mbound, &fbsize, &cbsize) < 0)
TEST_ERROR;
if(mbound != MBOUNDARY || fbsize != FBSIZE || cbsize != CBSIZE)
[svn-r24993] Purpose: Adds write tracking to the core VFD, which can be configured via the H5Pset/get_core_write_tracking() API call. When enabled and writing to the backing store is enabled, this feature will track writes and only write out the changed bytes on flush/close. Tested on: 32-bit LE linux (jam) w/ Fortran and C++ 64-bit BE linux (ostrich) 64-bit Darwin 12.5.0 (kite) All were tested using the core VFD w/ paging on (core_paged in the VFD list). Make check-vfd was also tested on jam.
2014-04-09 11:35:16 +08:00
TEST_ERROR;
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if(H5Pset_alignment(fapl, (hsize_t)THRESHOLD, (hsize_t)FBSIZE) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
H5E_BEGIN_TRY {
file=H5Fcreate(filename, H5F_ACC_TRUNC, H5P_DEFAULT, fapl);
} H5E_END_TRY;
if(file<0) {
H5Pclose (fapl);
SKIPPED();
printf(" Probably the file system doesn't support Direct I/O\n");
return 0;
}
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Retrieve the access property list... */
if ((access_fapl = H5Fget_access_plist(file)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
/* Check that the driver is correct */
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if(H5FD_DIRECT != H5Pget_driver(access_fapl))
[svn-r24993] Purpose: Adds write tracking to the core VFD, which can be configured via the H5Pset/get_core_write_tracking() API call. When enabled and writing to the backing store is enabled, this feature will track writes and only write out the changed bytes on flush/close. Tested on: 32-bit LE linux (jam) w/ Fortran and C++ 64-bit BE linux (ostrich) 64-bit Darwin 12.5.0 (kite) All were tested using the core VFD w/ paging on (core_paged in the VFD list). Make check-vfd was also tested on jam.
2014-04-09 11:35:16 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* ...and close the property list */
if (H5Pclose(access_fapl) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
/* Check file handle API */
if(H5Fget_vfd_handle(file, H5P_DEFAULT, (void **)&fhandle) < 0)
TEST_ERROR;
if(*fhandle<0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Check file size API */
if(H5Fget_filesize(file, &file_size) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
/* There is no guarantee of the number of metadata allocations, but it's
* 4 currently and the size of the file should be between 3 & 4 file buffer
* sizes..
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
*/
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if(file_size < (FBSIZE * 3) || file_size >= (FBSIZE * 4))
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
/* Allocate aligned memory for data set 1. For data set 1, everything is aligned including
* memory address, size of data, and file address. */
if(0 != HDposix_memalign(&points, (size_t)FBSIZE, (size_t)(DSET1_DIM1 * DSET1_DIM2 * sizeof(int))))
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if(0 != HDposix_memalign(&check, (size_t)FBSIZE, (size_t)(DSET1_DIM1 * DSET1_DIM2 * sizeof(int))))
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
TEST_ERROR;
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
/* Initialize the dset1 */
p1 = points;
[svn-r14193] Description: Make H5Dopen versioned and change all internal usage to use H5Dopen2 Add simple regression test for H5Dopen1 Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-09 03:59:36 +08:00
for(i = n = 0; i < DSET1_DIM1; i++)
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
for(j = 0; j < DSET1_DIM2; j++)
*p1++ = n++;
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
/* Create the data space1 */
dims1[0] = DSET1_DIM1;
dims1[1] = DSET1_DIM2;
[svn-r14193] Description: Make H5Dopen versioned and change all internal usage to use H5Dopen2 Add simple regression test for H5Dopen1 Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-09 03:59:36 +08:00
if((space1 = H5Screate_simple(2, dims1, NULL)) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
/* Create the dset1 */
[svn-r14199] Description: Add H5Dcreate to API versioned routines, replacing internal usage with H5Dcreate2 Fix thread-safe error stack initialization for API versioned error stack printing routines. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-12 00:24:11 +08:00
if((dset1 = H5Dcreate2(file, DSET1_NAME, H5T_NATIVE_INT, space1, H5P_DEFAULT, H5P_DEFAULT, H5P_DEFAULT)) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
/* Write the data to the dset1 */
[svn-r14193] Description: Make H5Dopen versioned and change all internal usage to use H5Dopen2 Add simple regression test for H5Dopen1 Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-09 03:59:36 +08:00
if(H5Dwrite(dset1, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, points) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Dclose(dset1) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
[svn-r14193] Description: Make H5Dopen versioned and change all internal usage to use H5Dopen2 Add simple regression test for H5Dopen1 Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-09 03:59:36 +08:00
if((dset1 = H5Dopen2(file, DSET1_NAME, H5P_DEFAULT)) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
/* Read the data back from dset1 */
[svn-r14193] Description: Make H5Dopen versioned and change all internal usage to use H5Dopen2 Add simple regression test for H5Dopen1 Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-09 03:59:36 +08:00
if(H5Dread(dset1, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, check) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
/* Check that the values read are the same as the values written */
p1 = points;
p2 = check;
[svn-r14193] Description: Make H5Dopen versioned and change all internal usage to use H5Dopen2 Add simple regression test for H5Dopen1 Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-09 03:59:36 +08:00
for(i = 0; i < DSET1_DIM1; i++)
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
for(j = 0; j < DSET1_DIM2; j++)
if(*p1++ != *p2++) {
H5_FAILED();
printf(" Read different values than written in data set 1.\n");
printf(" At index %d,%d\n", i, j);
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
} /* end if */
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
/* Create the data space2. For data set 2, memory address and data size are not aligned. */
dims2[0] = DSET2_DIM;
if((space2 = H5Screate_simple(1, dims2, NULL)) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
/* Create the dset2 */
if((dset2 = H5Dcreate2(file, DSET2_NAME, H5T_NATIVE_INT, space2, H5P_DEFAULT, H5P_DEFAULT, H5P_DEFAULT)) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
/* Write the data to the dset1 */
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if(H5Dwrite(dset2, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, wdata2) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if(H5Dclose(dset2) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if((dset2 = H5Dopen2(file, DSET2_NAME, H5P_DEFAULT)) < 0)
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
TEST_ERROR;
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
/* Read the data back from dset1 */
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if(H5Dread(dset2, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, rdata2) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
/* Check that the values read are the same as the values written */
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
for(i = 0; i < DSET2_DIM; i++)
if(wdata2[i] != rdata2[i]) {
H5_FAILED();
printf(" Read different values than written in data set 2.\n");
printf(" At index %d\n", i);
TEST_ERROR;
} /* end if */
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Sclose(space1) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Dclose(dset1) < 0)
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
TEST_ERROR;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
if(H5Sclose(space2) < 0)
TEST_ERROR;
if(H5Dclose(dset2) < 0)
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
HDfree(points);
HDfree(check);
[svn-r13001] Enabled CORE file driver to read or write an existing file depending the setting of backing store through H5Pset_core_fapl and file open flag. Added test cases in vfd.c.
2006-12-01 04:49:01 +08:00
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
/* Close and delete the file */
if(H5Fclose(file) < 0)
TEST_ERROR;
h5_delete_test_file(FILENAME[5], fapl);
/* Close the fapl */
if(H5Pclose(fapl) < 0)
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
PASSED();
return 0;
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
error:
H5E_BEGIN_TRY {
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
H5Pclose(fapl);
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
H5Sclose(space1);
H5Dclose(dset1);
H5Sclose(space2);
H5Dclose(dset2);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
H5Fclose(file);
} H5E_END_TRY;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if(points)
HDfree(points);
if(check)
HDfree(check);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
return -1;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
#endif /*H5_HAVE_DIRECT*/
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
}
/*-------------------------------------------------------------------------
* Function: test_family_opens
*
* Purpose: Private function for test_family() to tests wrong ways of
* reopening family file.
*
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
* Return: Success: 0
* Failure: -1
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
*
* Programmer: Raymond Lu
* Thursday, May 19, 2005
*
*-------------------------------------------------------------------------
*/
static herr_t
test_family_opens(char *fname, hid_t fa_pl)
{
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
hid_t file=-1;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
char first_name[1024];
char wrong_name[1024];
int i;
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
/* Case 1: reopen file with 1st member file name and default property list */
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
HDsnprintf(first_name, sizeof(first_name), fname, 0);
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
H5E_BEGIN_TRY {
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
file = H5Fopen(first_name, H5F_ACC_RDWR, H5P_DEFAULT);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
} H5E_END_TRY;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if(file >= 0)
TEST_ERROR
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Case 2: reopen file with correct name template but default property list */
H5E_BEGIN_TRY {
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
file = H5Fopen(fname, H5F_ACC_RDWR, H5P_DEFAULT);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
} H5E_END_TRY;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if(file >= 0)
TEST_ERROR
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Case 3: reopen file with wrong member size */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pset_fapl_family(fa_pl, (hsize_t)128, H5P_DEFAULT) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
H5E_BEGIN_TRY {
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
file = H5Fopen(fname, H5F_ACC_RDWR, fa_pl);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
} H5E_END_TRY;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if(file >= 0)
TEST_ERROR
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Case 4: reopen file with wrong name template */
[svn-r12705] Description: Fix file handle destroy routine to not attempt to flush out partially initialized file handles (when opening a file fails). Tested on: (until they finished testing, this time... :-) FreeBSD/32 4.11 (sleipnir) w/threadsafe Linux/64 2.4 (mir) w/1.6 compat Linux/32 2.4 (heping) w/FORTRAN & C++ Mac OSX/32 10.4.8 (amazon)
2006-10-03 02:32:46 +08:00
HDstrcpy(wrong_name, fname);
for(i = 0; i < 1024; i++)
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
if(wrong_name[i] == '5') {
wrong_name[i] = '4';
break;
}
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pset_fapl_family(fa_pl, (hsize_t)FAMILY_SIZE, H5P_DEFAULT) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
H5E_BEGIN_TRY {
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
file = H5Fopen(wrong_name, H5F_ACC_RDWR, fa_pl);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
} H5E_END_TRY;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if(file >= 0)
TEST_ERROR
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
return 0;
[svn-r12705] Description: Fix file handle destroy routine to not attempt to flush out partially initialized file handles (when opening a file fails). Tested on: (until they finished testing, this time... :-) FreeBSD/32 4.11 (sleipnir) w/threadsafe Linux/64 2.4 (mir) w/1.6 compat Linux/32 2.4 (heping) w/FORTRAN & C++ Mac OSX/32 10.4.8 (amazon)
2006-10-03 02:32:46 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
error:
return -1;
[svn-r12705] Description: Fix file handle destroy routine to not attempt to flush out partially initialized file handles (when opening a file fails). Tested on: (until they finished testing, this time... :-) FreeBSD/32 4.11 (sleipnir) w/threadsafe Linux/64 2.4 (mir) w/1.6 compat Linux/32 2.4 (heping) w/FORTRAN & C++ Mac OSX/32 10.4.8 (amazon)
2006-10-03 02:32:46 +08:00
} /* end test_family_opens() */
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/*-------------------------------------------------------------------------
* Function: test_family
*
* Purpose: Tests the file handle interface for FAMILY driver
*
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
* Return: Success: 0
* Failure: -1
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
*
* Programmer: Raymond Lu
* Tuesday, Sept 24, 2002
*
*-------------------------------------------------------------------------
*/
static herr_t
test_family(void)
{
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
hid_t file=-1, fapl=-1, fapl2=-1, space=-1, dset=-1;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
hid_t access_fapl = -1;
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
hid_t driver_id = -1; /* ID for this VFD */
unsigned long driver_flags = 0; /* VFD feature flags */
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
char filename[1024];
char dname[]="dataset";
[svn-r13588] A support of files bigger than 2GB for STDIO driver. Configure will check if fseeko is available. Using it instead of fseek can support big files because the offset is of type off_t not long int. Also added the test for STDIO in big.c.
2007-04-05 03:59:00 +08:00
unsigned int i, j;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
int *fhandle=NULL, *fhandle2=NULL;
int buf[FAMILY_NUMBER][FAMILY_SIZE];
hsize_t dims[2]={FAMILY_NUMBER, FAMILY_SIZE};
hsize_t file_size;
TESTING("FAMILY file driver");
/* Set property list and file name for FAMILY driver */
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
if((fapl = H5Pcreate(H5P_FILE_ACCESS)) < 0)
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pset_fapl_family(fapl, (hsize_t)FAMILY_SIZE, H5P_DEFAULT) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
h5_fixname(FILENAME[2], fapl, filename, sizeof(filename));
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
/* Check that the VFD feature flags are correct */
if ((driver_id = H5Pget_driver(fapl)) < 0)
TEST_ERROR
if (H5FDdriver_query(driver_id, &driver_flags) < 0)
TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_ACCUMULATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_DATA_SIEVE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_SMALLDATA)) TEST_ERROR
/* Check for extra flags not accounted for above */
if(driver_flags != (H5FD_FEAT_AGGREGATE_METADATA
| H5FD_FEAT_ACCUMULATE_METADATA
| H5FD_FEAT_DATA_SIEVE
| H5FD_FEAT_AGGREGATE_SMALLDATA))
TEST_ERROR
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if((file=H5Fcreate(filename, H5F_ACC_TRUNC, H5P_DEFAULT, fapl)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Fclose(file) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
/* Test different wrong ways to reopen family files where there's only
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
* one member file existing. */
[svn-r12705] Description: Fix file handle destroy routine to not attempt to flush out partially initialized file handles (when opening a file fails). Tested on: (until they finished testing, this time... :-) FreeBSD/32 4.11 (sleipnir) w/threadsafe Linux/64 2.4 (mir) w/1.6 compat Linux/32 2.4 (heping) w/FORTRAN & C++ Mac OSX/32 10.4.8 (amazon)
2006-10-03 02:32:46 +08:00
if(test_family_opens(filename, fapl) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Reopen the file with default member file size */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pset_fapl_family(fapl, (hsize_t)H5F_FAMILY_DEFAULT, H5P_DEFAULT) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if((file=H5Fopen(filename, H5F_ACC_RDWR, fapl)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Check file size API */
if(H5Fget_filesize(file, &file_size) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r12705] Description: Fix file handle destroy routine to not attempt to flush out partially initialized file handles (when opening a file fails). Tested on: (until they finished testing, this time... :-) FreeBSD/32 4.11 (sleipnir) w/threadsafe Linux/64 2.4 (mir) w/1.6 compat Linux/32 2.4 (heping) w/FORTRAN & C++ Mac OSX/32 10.4.8 (amazon)
2006-10-03 02:32:46 +08:00
/* The file size is supposed to be about 800 bytes right now. */
if(file_size < (KB / 2) || file_size > KB)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Create and write dataset */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if((space=H5Screate_simple(2, dims, NULL)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Retrieve the access property list... */
if ((access_fapl = H5Fget_access_plist(file)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
/* Check that the driver is correct */
if(H5FD_FAMILY != H5Pget_driver(access_fapl))
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* ...and close the property list */
if (H5Pclose(access_fapl) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r14199] Description: Add H5Dcreate to API versioned routines, replacing internal usage with H5Dcreate2 Fix thread-safe error stack initialization for API versioned error stack printing routines. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-12 00:24:11 +08:00
if((dset=H5Dcreate2(file, dname, H5T_NATIVE_INT, space, H5P_DEFAULT, H5P_DEFAULT, H5P_DEFAULT)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
for(i = 0; i < FAMILY_NUMBER; i++)
for(j = 0; j < FAMILY_SIZE; j++)
buf[i][j] = (int)((i * 10000) + j);
[svn-r13588] A support of files bigger than 2GB for STDIO driver. Configure will check if fseeko is available. Using it instead of fseek can support big files because the offset is of type off_t not long int. Also added the test for STDIO in big.c.
2007-04-05 03:59:00 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Dwrite(dset, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, buf) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* check file handle API */
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if((fapl2 = H5Pcreate(H5P_FILE_ACCESS)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pset_family_offset(fapl2, (hsize_t)0) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Fget_vfd_handle(file, fapl2, (void **)&fhandle) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if(*fhandle < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pset_family_offset(fapl2, (hsize_t)(FAMILY_SIZE*2)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Fget_vfd_handle(file, fapl2, (void **)&fhandle2) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if(*fhandle2 < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Check file size API */
if(H5Fget_filesize(file, &file_size) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
/* Some data has been written. The file size should be bigger (18KB+976)
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
* bytes if int size is 4 bytes) now. */
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
#if H5_SIZEOF_INT <= 4
if(file_size < (18 * KB) || file_size > (20 * KB))
TEST_ERROR;
#elif H5_SIZEOF_INT >= 8
if(file_size < (32 * KB) || file_size > (40 * KB))
TEST_ERROR;
#endif
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Sclose(space) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Dclose(dset) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pclose(fapl2) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Fclose(file) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Test different wrong ways to reopen family files when there're multiple
* member files existing. */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(test_family_opens(filename, fapl) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Reopen the file with correct member file size. */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pset_fapl_family(fapl, (hsize_t)FAMILY_SIZE, H5P_DEFAULT) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if((file = H5Fopen(filename, H5F_ACC_RDWR, fapl)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
/* Close and delete the file */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Fclose(file) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
h5_delete_test_file(FILENAME[2], fapl);
/* Close the fapl */
if(H5Pclose(fapl) < 0)
TEST_ERROR;
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
PASSED();
return 0;
error:
H5E_BEGIN_TRY {
H5Sclose(space);
H5Dclose(dset);
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
H5Pclose(fapl);
H5Pclose(fapl2);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
H5Fclose(file);
} H5E_END_TRY;
return -1;
}
/*-------------------------------------------------------------------------
* Function: test_family_compat
*
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
* Purpose: Tests the backward compatibility for FAMILY driver.
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
* See if we can open files created with v1.6 library.
* The source file was created by the test/file_handle.c
* of the v1.6 library. Then tools/misc/h5repart.c was
* used to concantenated. The command was "h5repart -m 5k
[svn-r14340] Maintenance for VMS system: Skip one of the tests in flush2.c since it looks like VMS system flsuhes the file anyway (like on Winodws with DLLs) Renamed files family*.h5 that are used in backward compatibility testing to have only one '.' in a file name, otherwise VMS system becomes confused and test fails. Cleaned up the only test (Fortran multi-file test) that is failing on VMS (C counterpart passesa). Added new script to facilitate the build process on VMS Updated MANIFEST PLatforms tested: VMS server, kagiso and linew (minor changes)
2007-12-14 02:17:02 +08:00
* family_file%05d.h5 family_v16_%05d.h5".
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
*
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
* Return: Success: 0
* Failure: -1
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
*
* Programmer: Raymond Lu
* June 3, 2005
*
*-------------------------------------------------------------------------
*/
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
static herr_t
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
test_family_compat(void)
{
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
hid_t file = -1, fapl = -1;
hid_t dset = -1;
[svn-r16609] Bug fix (#1161): When a family file created with v1.6 library is opened and closed with v1.8 library, v1.8 library writes the driver info block in the superblock. But v1.6 doesn't write it. This caused the data after the superblock to be overwritten. The solution is to use a flag to indicate when the original file doesn't have the driver info, v1.8 library doesn't write it either. Tested on jam. I have some trouble to access smirom and linew using h5committest.
2009-03-25 06:06:07 +08:00
char dname[]="dataset";
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
char filename[1024];
[svn-r16759] The test of family file compatibility with v1.6 used the original family files under the library source directory. Since the test opens the files with write mode, it's safer to make copies of the original files. I added the copying step in test_family_compat in vfd.c. I changed the Makefile.am to clean up the copied files. I created a new family_v16_00000.h5 with v1.6 library. Tested on jam - supposed to be simple change.
2009-04-16 05:42:02 +08:00
char pathname[1024], pathname_individual[1024];
char newname[1024], newname_individual[1024];
int counter = 0;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
TESTING("FAMILY file driver backward compatibility");
/* Set property list and file name for FAMILY driver */
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
if((fapl = H5Pcreate(H5P_FILE_ACCESS)) < 0)
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pset_fapl_family(fapl, (hsize_t)FAMILY_SIZE2, H5P_DEFAULT) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
h5_fixname(COMPAT_BASENAME, fapl, filename, sizeof(filename));
h5_fixname(FILENAME[3], fapl, newname, sizeof(newname));
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r11562] Purpose: Avoid bug for family file driver backward compatibility with 1.6. Description: Because there is no srcdir on windows, string pathname should be initialized as NULL before using strcat. Solution: Changed the code as above. Platforms tested: too minor to test. Misc. update:
2005-10-15 03:16:54 +08:00
pathname[0] = '\0';
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
HDstrcat(pathname, filename);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r16759] The test of family file compatibility with v1.6 used the original family files under the library source directory. Since the test opens the files with write mode, it's safer to make copies of the original files. I added the copying step in test_family_compat in vfd.c. I changed the Makefile.am to clean up the copied files. I created a new family_v16_00000.h5 with v1.6 library. Tested on jam - supposed to be simple change.
2009-04-16 05:42:02 +08:00
/* The following code makes the copies of the family files in the source directory.
* Since we're going to open the files with write mode, this protects the original
* files.
*/
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
HDsnprintf(newname_individual, sizeof(newname_individual), newname, counter);
HDsnprintf(pathname_individual, sizeof(pathname_individual), pathname, counter);
[svn-r16759] The test of family file compatibility with v1.6 used the original family files under the library source directory. Since the test opens the files with write mode, it's safer to make copies of the original files. I added the copying step in test_family_compat in vfd.c. I changed the Makefile.am to clean up the copied files. I created a new family_v16_00000.h5 with v1.6 library. Tested on jam - supposed to be simple change.
2009-04-16 05:42:02 +08:00
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
while(h5_make_local_copy(pathname_individual, newname_individual) >= 0) {
[svn-r16759] The test of family file compatibility with v1.6 used the original family files under the library source directory. Since the test opens the files with write mode, it's safer to make copies of the original files. I added the copying step in test_family_compat in vfd.c. I changed the Makefile.am to clean up the copied files. I created a new family_v16_00000.h5 with v1.6 library. Tested on jam - supposed to be simple change.
2009-04-16 05:42:02 +08:00
counter++;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
HDsnprintf(newname_individual, sizeof(newname_individual), newname, counter);
HDsnprintf(pathname_individual, sizeof(pathname_individual), pathname, counter);
Fixed a few memory problems in test/cache_logging.c and test/h5test.c (which show up in test/vfd.c).
2016-11-23 09:55:03 +08:00
} /* end while */
[svn-r16759] The test of family file compatibility with v1.6 used the original family files under the library source directory. Since the test opens the files with write mode, it's safer to make copies of the original files. I added the copying step in test_family_compat in vfd.c. I changed the Makefile.am to clean up the copied files. I created a new family_v16_00000.h5 with v1.6 library. Tested on jam - supposed to be simple change.
2009-04-16 05:42:02 +08:00
[svn-r16609] Bug fix (#1161): When a family file created with v1.6 library is opened and closed with v1.8 library, v1.8 library writes the driver info block in the superblock. But v1.6 doesn't write it. This caused the data after the superblock to be overwritten. The solution is to use a flag to indicate when the original file doesn't have the driver info, v1.8 library doesn't write it either. Tested on jam. I have some trouble to access smirom and linew using h5committest.
2009-03-25 06:06:07 +08:00
/* Make sure we can open the file. Use the read and write mode to flush the
* superblock. */
[svn-r16759] The test of family file compatibility with v1.6 used the original family files under the library source directory. Since the test opens the files with write mode, it's safer to make copies of the original files. I added the copying step in test_family_compat in vfd.c. I changed the Makefile.am to clean up the copied files. I created a new family_v16_00000.h5 with v1.6 library. Tested on jam - supposed to be simple change.
2009-04-16 05:42:02 +08:00
if((file = H5Fopen(newname, H5F_ACC_RDWR, fapl)) < 0)
[svn-r16609] Bug fix (#1161): When a family file created with v1.6 library is opened and closed with v1.8 library, v1.8 library writes the driver info block in the superblock. But v1.6 doesn't write it. This caused the data after the superblock to be overwritten. The solution is to use a flag to indicate when the original file doesn't have the driver info, v1.8 library doesn't write it either. Tested on jam. I have some trouble to access smirom and linew using h5committest.
2009-03-25 06:06:07 +08:00
TEST_ERROR;
[svn-r16612] I changed the API function H5Dopen to H5Dopen2, which I introduced in recent checkin. Tested on jam - simple change.
2009-03-26 03:28:01 +08:00
if((dset = H5Dopen2(file, dname, H5P_DEFAULT)) < 0)
[svn-r16609] Bug fix (#1161): When a family file created with v1.6 library is opened and closed with v1.8 library, v1.8 library writes the driver info block in the superblock. But v1.6 doesn't write it. This caused the data after the superblock to be overwritten. The solution is to use a flag to indicate when the original file doesn't have the driver info, v1.8 library doesn't write it either. Tested on jam. I have some trouble to access smirom and linew using h5committest.
2009-03-25 06:06:07 +08:00
TEST_ERROR;
if(H5Dclose(dset) < 0)
TEST_ERROR;
if(H5Fclose(file) < 0)
TEST_ERROR;
/* Open the file again to make sure it isn't corrupted. */
[svn-r16759] The test of family file compatibility with v1.6 used the original family files under the library source directory. Since the test opens the files with write mode, it's safer to make copies of the original files. I added the copying step in test_family_compat in vfd.c. I changed the Makefile.am to clean up the copied files. I created a new family_v16_00000.h5 with v1.6 library. Tested on jam - supposed to be simple change.
2009-04-16 05:42:02 +08:00
if((file = H5Fopen(newname, H5F_ACC_RDWR, fapl)) < 0)
[svn-r16609] Bug fix (#1161): When a family file created with v1.6 library is opened and closed with v1.8 library, v1.8 library writes the driver info block in the superblock. But v1.6 doesn't write it. This caused the data after the superblock to be overwritten. The solution is to use a flag to indicate when the original file doesn't have the driver info, v1.8 library doesn't write it either. Tested on jam. I have some trouble to access smirom and linew using h5committest.
2009-03-25 06:06:07 +08:00
TEST_ERROR;
[svn-r16612] I changed the API function H5Dopen to H5Dopen2, which I introduced in recent checkin. Tested on jam - simple change.
2009-03-26 03:28:01 +08:00
if((dset = H5Dopen2(file, dname, H5P_DEFAULT)) < 0)
[svn-r16609] Bug fix (#1161): When a family file created with v1.6 library is opened and closed with v1.8 library, v1.8 library writes the driver info block in the superblock. But v1.6 doesn't write it. This caused the data after the superblock to be overwritten. The solution is to use a flag to indicate when the original file doesn't have the driver info, v1.8 library doesn't write it either. Tested on jam. I have some trouble to access smirom and linew using h5committest.
2009-03-25 06:06:07 +08:00
TEST_ERROR;
if(H5Dclose(dset) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
/* Close and delete the file */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Fclose(file) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
h5_delete_test_file(FILENAME[3], fapl);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
/* Close the fapl */
if(H5Pclose(fapl) < 0)
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
PASSED();
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
return 0;
error:
H5E_BEGIN_TRY {
H5Fclose(file);
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
H5Pclose(fapl);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
} H5E_END_TRY;
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
return -1;
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
} /* end test_family_compat() */
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/*-------------------------------------------------------------------------
* Function: test_multi_opens
*
* Purpose: Private function for test_multi() to tests wrong ways of
* reopening multi file.
*
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
* Return: Success: 0
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
* Failure: -1
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
*
* Programmer: Raymond Lu
* Thursday, May 19, 2005
*
*-------------------------------------------------------------------------
*/
static herr_t
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
test_multi_opens(char *fname)
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
{
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
hid_t file=-1;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
char super_name[1024]; /*name string "%%s-s.h5"*/
char sf_name[1024]; /*name string "multi_file-s.h5"*/
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
/* Case: reopen with the name of super file and default property list */
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
HDsnprintf(super_name, sizeof(super_name), "%%s-%c.h5", 's');
HDsnprintf(sf_name, sizeof(sf_name), super_name, fname);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
H5E_BEGIN_TRY {
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
file = H5Fopen(sf_name, H5F_ACC_RDWR, H5P_DEFAULT);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
} H5E_END_TRY;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
return(file >= 0 ? -1 : 0);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
}
/*-------------------------------------------------------------------------
* Function: test_multi
*
* Purpose: Tests the file handle interface for MUTLI driver
*
[svn-r14210] Description: Enable the backward compatibility file format test for the family VFD in all situations. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-19 01:47:43 +08:00
* Return: Success: 0
* Failure: -1
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
*
* Programmer: Raymond Lu
* Tuesday, Sept 24, 2002
*
*-------------------------------------------------------------------------
*/
static herr_t
test_multi(void)
{
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
hid_t file=-1, fapl=-1, fapl2=-1, dset=-1, space=-1;
hid_t root=-1, attr=-1, aspace=-1, atype=-1;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
hid_t access_fapl = -1;
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
hid_t driver_id = -1; /* ID for this VFD */
unsigned long driver_flags = 0; /* VFD feature flags */
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
char filename[1024];
int *fhandle2=NULL, *fhandle=NULL;
hsize_t file_size;
H5FD_mem_t mt, memb_map[H5FD_MEM_NTYPES];
hid_t memb_fapl[H5FD_MEM_NTYPES];
haddr_t memb_addr[H5FD_MEM_NTYPES];
const char *memb_name[H5FD_MEM_NTYPES];
char sv[H5FD_MEM_NTYPES][32];
hsize_t dims[2]={MULTI_SIZE, MULTI_SIZE};
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
hsize_t adims[1]={1};
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
char dname[]="dataset";
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
char meta[] = "this is some metadata on this file";
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
int i, j;
int buf[MULTI_SIZE][MULTI_SIZE];
TESTING("MULTI file driver");
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Set file access property list for MULTI driver */
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
if((fapl = H5Pcreate(H5P_FILE_ACCESS)) < 0)
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
HDmemset(memb_map, 0, sizeof(memb_map));
HDmemset(memb_fapl, 0, sizeof(memb_fapl));
HDmemset(memb_name, 0, sizeof(memb_name));
HDmemset(memb_addr, 0, sizeof(memb_addr));
HDmemset(sv, 0, sizeof(sv));
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
for(mt=H5FD_MEM_DEFAULT; mt<H5FD_MEM_NTYPES; H5_INC_ENUM(H5FD_mem_t,mt)) {
memb_fapl[mt] = H5P_DEFAULT;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
memb_map[mt] = H5FD_MEM_SUPER;
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
}
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
memb_map[H5FD_MEM_DRAW] = H5FD_MEM_DRAW;
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
memb_map[H5FD_MEM_BTREE] = H5FD_MEM_BTREE;
memb_map[H5FD_MEM_GHEAP] = H5FD_MEM_GHEAP;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
sprintf(sv[H5FD_MEM_SUPER], "%%s-%c.h5", 's');
memb_name[H5FD_MEM_SUPER] = sv[H5FD_MEM_SUPER];
memb_addr[H5FD_MEM_SUPER] = 0;
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
sprintf(sv[H5FD_MEM_BTREE], "%%s-%c.h5", 'b');
memb_name[H5FD_MEM_BTREE] = sv[H5FD_MEM_BTREE];
memb_addr[H5FD_MEM_BTREE] = HADDR_MAX/4;
[svn-r15628] Description: Remove trailing whitespace from C/C++ source files, with the following script: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Tested on: Mac OS X/32 10.5.5 (amazon) No need for h5committest, just whitespace changes...
2008-09-16 23:52:51 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
sprintf(sv[H5FD_MEM_DRAW], "%%s-%c.h5", 'r');
memb_name[H5FD_MEM_DRAW] = sv[H5FD_MEM_DRAW];
memb_addr[H5FD_MEM_DRAW] = HADDR_MAX/2;
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
sprintf(sv[H5FD_MEM_GHEAP], "%%s-%c.h5", 'g');
memb_name[H5FD_MEM_GHEAP] = sv[H5FD_MEM_GHEAP];
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
memb_addr[H5FD_MEM_GHEAP] = (HADDR_MAX/4)*3;
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pset_fapl_multi(fapl, memb_map, memb_fapl, memb_name, memb_addr, TRUE) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r16779] A minor correction of previous checkin. The original files created with v1.6 library could be deleted by cleanup routine. I corrected it in this checkin. Tested on jam - simple change.
2009-04-17 23:52:38 +08:00
h5_fixname(FILENAME[4], fapl, filename, sizeof filename);
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
/* Check that the VFD feature flags are correct */
if ((driver_id = H5Pget_driver(fapl)) < 0)
TEST_ERROR
if (H5FDdriver_query(driver_id, &driver_flags) < 0)
TEST_ERROR
if(!(driver_flags & H5FD_FEAT_DATA_SIEVE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_SMALLDATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_USE_ALLOC_SIZE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_PAGED_AGGR)) TEST_ERROR
/* Check for extra flags not accounted for above */
if(driver_flags != (H5FD_FEAT_DATA_SIEVE
| H5FD_FEAT_AGGREGATE_SMALLDATA
| H5FD_FEAT_USE_ALLOC_SIZE
| H5FD_FEAT_PAGED_AGGR))
TEST_ERROR
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if((file=H5Fcreate(filename, H5F_ACC_TRUNC, H5P_DEFAULT, fapl)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Fclose(file) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Test wrong ways to reopen multi files */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(test_multi_opens(filename) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Reopen the file */
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if((file = H5Fopen(filename, H5F_ACC_RDWR, fapl)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Create and write data set */
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if((space = H5Screate_simple(2, dims, NULL)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Retrieve the access property list... */
if ((access_fapl = H5Fget_access_plist(file)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
/* Check that the driver is correct */
if(H5FD_MULTI != H5Pget_driver(access_fapl))
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* ...and close the property list */
if (H5Pclose(access_fapl) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Check file size API */
if(H5Fget_filesize(file, &file_size) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Before any data is written, the raw data file is empty. So
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
* the file size is only the size of b-tree + HADDR_MAX/4.
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
*/
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
if(file_size < HADDR_MAX/4 || file_size > HADDR_MAX/2)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r14199] Description: Add H5Dcreate to API versioned routines, replacing internal usage with H5Dcreate2 Fix thread-safe error stack initialization for API versioned error stack printing routines. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-12 00:24:11 +08:00
if((dset=H5Dcreate2(file, dname, H5T_NATIVE_INT, space, H5P_DEFAULT, H5P_DEFAULT, H5P_DEFAULT)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
for(i=0; i<MULTI_SIZE; i++)
for(j=0; j<MULTI_SIZE; j++)
buf[i][j] = i*10000+j;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Dwrite(dset, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, buf) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if((fapl2=H5Pcreate(H5P_FILE_ACCESS)) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pset_multi_type(fapl2, H5FD_MEM_SUPER) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Fget_vfd_handle(file, fapl2, (void **)&fhandle) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
if(*fhandle<0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pset_multi_type(fapl2, H5FD_MEM_DRAW) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Fget_vfd_handle(file, fapl2, (void **)&fhandle2) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
if(*fhandle2<0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/* Check file size API */
if(H5Fget_filesize(file, &file_size) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
/* After the data is written, the file size is huge because the
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
* beginning of raw data file is set at HADDR_MAX/2. It's supposed
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
* to be (HADDR_MAX/2 + 128*128*4)
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
*/
if(file_size < HADDR_MAX/2 || file_size > HADDR_MAX)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Sclose(space) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Dclose(dset) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Pclose(fapl2) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
/* Create and write attribute for the root group. */
[svn-r14115] Description: Remove all plain calls to H5Gopen() from source, replacing them with either H5Gopen2(). Add test for H5Gopen1(). Reformatted several pieces of code, to clean them up. Tested on: FreeBSD/32 6.2 (duty) FreeBSD/64 6.2 (liberty) Linux/32 2.6 (kagiso) Linux/64 2.6 (smirom) Solaris/32 5.10 (linew) Mac OS X/32 10.4.10 (amazon)
2007-08-28 23:02:54 +08:00
if((root = H5Gopen2(file, "/", H5P_DEFAULT)) < 0)
[svn-r30297] Changed FAIL_STACK_ERROR to TEST_ERROR in test/vfd.c. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial, vfd test only
2016-08-18 22:20:03 +08:00
TEST_ERROR
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
/* Attribute string. */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if((atype = H5Tcopy(H5T_C_S1)) < 0)
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Tset_size(atype, strlen(meta) + 1) < 0)
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Tset_strpad(atype, H5T_STR_NULLTERM) < 0)
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
TEST_ERROR;
/* Create and write attribute */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if((aspace = H5Screate_simple(1, adims, NULL)) < 0)
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
TEST_ERROR;
[svn-r14218] Description: Changed H5Acreate2 -> H5Acreate_by_name, to be more consistent with other new API routines. Re-added simpler form of H5Acreate2, which creates attributes directly on an object. Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-31 02:13:48 +08:00
if((attr = H5Acreate2(root, "Metadata", atype, aspace, H5P_DEFAULT, H5P_DEFAULT)) < 0)
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Awrite(attr, atype, meta) < 0)
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
TEST_ERROR;
/* Close IDs */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Tclose(atype) < 0)
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Sclose(aspace) < 0)
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
TEST_ERROR;
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Aclose(attr) < 0)
[svn-r13139] Fixed the MULTI problem (Bug #731) that corrupted the data. The problem came from the EOA for the whole MULTI file. It's taken out because it's meaningless for MULTI file. Instead, each individual file has its EOA.
2007-01-13 04:29:43 +08:00
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
/* Close and delete the file */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(H5Fclose(file) < 0)
[svn-r11712] Purpose: New feature Description: Check in baseline for compact group revisions, which radically revises the source code for managing groups and object headers. WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! This initiates the "unstable" phase of the 1.7.x branch, leading up to the 1.8.0 release. Please test this code, but do _NOT_ keep files created with it - the format will change again before the release and you will not be able to read your old files!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! Solution: There's too many changes to really describe them all, but some of them include: - Stop abusing the H5G_entry_t structure and split it into two separate structures for non-symbol table node use within the library: H5O_loc_t for object locations in a file and H5G_name_t to store the path to an opened object. H5G_entry_t is now only used for storing symbol table entries on disk. - Retire H5G_namei() in favor of a more general mechanism for traversing group paths and issuing callbacks on objects located. This gets us out of the business of hacking H5G_namei() for new features, generally. - Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t - Lots more... Platforms tested: h5committested and maybe another dozen configurations.... :-)
2005-11-15 10:55:39 +08:00
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
h5_delete_test_file(FILENAME[4], fapl);
/* Close the fapl */
if(H5Pclose(fapl) < 0)
TEST_ERROR;
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
PASSED();
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
return 0;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
error:
H5E_BEGIN_TRY {
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
H5Sclose(space);
H5Dclose(dset);
H5Pclose(fapl);
H5Pclose(fapl2);
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
H5Fclose(file);
} H5E_END_TRY;
return -1;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
}
[svn-r21015] Bug fix for Issue 2598 - In v1.6 library, there was EOA for the whole MULTI file saved in the super block. We took it out in v1.8 library because it's meaningless for the MULTI file. v1.8 library saves the EOA for the metadata file, instead. But this caused some backward compatibility problem. v1.8 library couldn't open the file created with v1.6 library. I fixed the problem by checking the EOA value to detect the file created with v1.6 library. Tested on jam, koala, and heiwa.
2011-06-22 23:47:57 +08:00
/*-------------------------------------------------------------------------
* Function: test_multi_compat
*
* Purpose: Tests the backward compatibility for MULTI driver.
* See if we can open files created with v1.6 library.
* The source file was created by the test/file_handle.c
* of the v1.6 library. This test verifies the fix for
* Issue 2598. In v1.6 library, there was EOA for the whole
* MULTI file saved in the super block. We took it out in
* v1.8 library because it's meaningless for the MULTI file.
* v1.8 library saves the EOA for the metadata file, instead.
*
* Return: Success: 0
* Failure: -1
*
* Programmer: Raymond Lu
* 21 June 2011
*
*-------------------------------------------------------------------------
*/
static herr_t
test_multi_compat(void)
{
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
hid_t file=-1, fapl=-1, dset=-1, space=-1;
[svn-r22411] reduce compiler warnings by removing unreferenced variables and adding cast Tested: local linux
2012-05-26 01:39:07 +08:00
char newname[1024];
[svn-r21015] Bug fix for Issue 2598 - In v1.6 library, there was EOA for the whole MULTI file saved in the super block. We took it out in v1.8 library because it's meaningless for the MULTI file. v1.8 library saves the EOA for the metadata file, instead. But this caused some backward compatibility problem. v1.8 library couldn't open the file created with v1.6 library. I fixed the problem by checking the EOA value to detect the file created with v1.6 library. Tested on jam, koala, and heiwa.
2011-06-22 23:47:57 +08:00
char filename_s[1024], newname_s[1024];
char filename_r[1024], newname_r[1024];
H5FD_mem_t mt, memb_map[H5FD_MEM_NTYPES];
hid_t memb_fapl[H5FD_MEM_NTYPES];
haddr_t memb_addr[H5FD_MEM_NTYPES];
const char *memb_name[H5FD_MEM_NTYPES];
char sv[H5FD_MEM_NTYPES][32];
hsize_t dims[2]={MULTI_SIZE, MULTI_SIZE};
int i, j;
int buf[MULTI_SIZE][MULTI_SIZE];
TESTING("MULTI file driver backward compatibility");
/* Set file access property list for MULTI driver */
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
if((fapl = H5Pcreate(H5P_FILE_ACCESS)) < 0)
TEST_ERROR;
[svn-r21015] Bug fix for Issue 2598 - In v1.6 library, there was EOA for the whole MULTI file saved in the super block. We took it out in v1.8 library because it's meaningless for the MULTI file. v1.8 library saves the EOA for the metadata file, instead. But this caused some backward compatibility problem. v1.8 library couldn't open the file created with v1.6 library. I fixed the problem by checking the EOA value to detect the file created with v1.6 library. Tested on jam, koala, and heiwa.
2011-06-22 23:47:57 +08:00
HDmemset(memb_map, 0, sizeof memb_map);
HDmemset(memb_fapl, 0, sizeof memb_fapl);
HDmemset(memb_name, 0, sizeof memb_name);
HDmemset(memb_addr, 0, sizeof memb_addr);
HDmemset(sv, 0, sizeof sv);
for(mt=H5FD_MEM_DEFAULT; mt<H5FD_MEM_NTYPES; H5_INC_ENUM(H5FD_mem_t,mt))
memb_map[mt] = H5FD_MEM_SUPER;
memb_map[H5FD_MEM_DRAW] = H5FD_MEM_DRAW;
memb_fapl[H5FD_MEM_SUPER] = H5P_DEFAULT;
sprintf(sv[H5FD_MEM_SUPER], "%%s-%c.h5", 's');
memb_name[H5FD_MEM_SUPER] = sv[H5FD_MEM_SUPER];
memb_addr[H5FD_MEM_SUPER] = 0;
memb_fapl[H5FD_MEM_DRAW] = H5P_DEFAULT;
sprintf(sv[H5FD_MEM_DRAW], "%%s-%c.h5", 'r');
memb_name[H5FD_MEM_DRAW] = sv[H5FD_MEM_DRAW];
memb_addr[H5FD_MEM_DRAW] = HADDR_MAX/2;
if(H5Pset_fapl_multi(fapl, memb_map, memb_fapl, memb_name, memb_addr, TRUE)<0)
TEST_ERROR;
h5_fixname(FILENAME[9], fapl, newname, sizeof newname);
/* Make copy for the data file in the build directory, to protect the
* original file in the source directory */
sprintf(filename_s, "%s-%c.h5", MULTI_COMPAT_BASENAME, 's');
sprintf(newname_s, "%s-%c.h5", FILENAME[9], 's');
h5_make_local_copy(filename_s, newname_s);
sprintf(filename_r, "%s-%c.h5", MULTI_COMPAT_BASENAME, 'r');
sprintf(newname_r, "%s-%c.h5", FILENAME[9], 'r');
h5_make_local_copy(filename_r, newname_r);
/* Reopen the file for read only. Verify 1.8 library can open file
* created with 1.6 library. */
if((file=H5Fopen(newname, H5F_ACC_RDONLY, fapl)) < 0)
TEST_ERROR;
if((dset = H5Dopen2(file, DSET1_NAME, H5P_DEFAULT)) < 0)
TEST_ERROR;
if(H5Dclose(dset) < 0)
TEST_ERROR;
if(H5Fclose(file) < 0)
TEST_ERROR;
/* Make sure we can reopen the file for read and write */
if((file=H5Fopen(newname, H5F_ACC_RDWR, fapl)) < 0)
TEST_ERROR;
if((dset = H5Dopen2(file, DSET1_NAME, H5P_DEFAULT)) < 0)
TEST_ERROR;
if(H5Dclose(dset) < 0)
TEST_ERROR;
if(H5Fclose(file) < 0)
TEST_ERROR;
/* Reopen the file for adding another dataset. The new EOA for metadata file
* should be written to the file */
if((file=H5Fopen(newname, H5F_ACC_RDWR, fapl)) < 0)
TEST_ERROR;
/* Create and write data set */
if((space=H5Screate_simple(2, dims, NULL)) < 0)
TEST_ERROR;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if((dset=H5Dcreate2(file, DSET3_NAME, H5T_NATIVE_INT, space, H5P_DEFAULT, H5P_DEFAULT, H5P_DEFAULT)) < 0)
[svn-r21015] Bug fix for Issue 2598 - In v1.6 library, there was EOA for the whole MULTI file saved in the super block. We took it out in v1.8 library because it's meaningless for the MULTI file. v1.8 library saves the EOA for the metadata file, instead. But this caused some backward compatibility problem. v1.8 library couldn't open the file created with v1.6 library. I fixed the problem by checking the EOA value to detect the file created with v1.6 library. Tested on jam, koala, and heiwa.
2011-06-22 23:47:57 +08:00
TEST_ERROR;
for(i=0; i<MULTI_SIZE; i++)
for(j=0; j<MULTI_SIZE; j++)
buf[i][j] = i*10000+j;
if(H5Dwrite(dset, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, buf) < 0)
TEST_ERROR;
if(H5Dclose(dset) < 0)
TEST_ERROR;
if(H5Sclose(space) < 0)
TEST_ERROR;
if(H5Fclose(file) < 0)
TEST_ERROR;
/* Reopen the file for read only again. Verify the library can handle
* the EOA correctly */
if((file=H5Fopen(newname, H5F_ACC_RDONLY, fapl)) < 0)
TEST_ERROR;
if((dset = H5Dopen2(file, DSET1_NAME, H5P_DEFAULT)) < 0)
TEST_ERROR;
if(H5Dclose(dset) < 0)
TEST_ERROR;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if((dset = H5Dopen2(file, DSET3_NAME, H5P_DEFAULT)) < 0)
[svn-r21015] Bug fix for Issue 2598 - In v1.6 library, there was EOA for the whole MULTI file saved in the super block. We took it out in v1.8 library because it's meaningless for the MULTI file. v1.8 library saves the EOA for the metadata file, instead. But this caused some backward compatibility problem. v1.8 library couldn't open the file created with v1.6 library. I fixed the problem by checking the EOA value to detect the file created with v1.6 library. Tested on jam, koala, and heiwa.
2011-06-22 23:47:57 +08:00
TEST_ERROR;
if(H5Dclose(dset) < 0)
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
/* Close and delete the file */
[svn-r21015] Bug fix for Issue 2598 - In v1.6 library, there was EOA for the whole MULTI file saved in the super block. We took it out in v1.8 library because it's meaningless for the MULTI file. v1.8 library saves the EOA for the metadata file, instead. But this caused some backward compatibility problem. v1.8 library couldn't open the file created with v1.6 library. I fixed the problem by checking the EOA value to detect the file created with v1.6 library. Tested on jam, koala, and heiwa.
2011-06-22 23:47:57 +08:00
if(H5Fclose(file) < 0)
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
h5_delete_test_file(FILENAME[9], fapl);
/* Close the fapl */
if(H5Pclose(fapl) < 0)
TEST_ERROR;
[svn-r21015] Bug fix for Issue 2598 - In v1.6 library, there was EOA for the whole MULTI file saved in the super block. We took it out in v1.8 library because it's meaningless for the MULTI file. v1.8 library saves the EOA for the metadata file, instead. But this caused some backward compatibility problem. v1.8 library couldn't open the file created with v1.6 library. I fixed the problem by checking the EOA value to detect the file created with v1.6 library. Tested on jam, koala, and heiwa.
2011-06-22 23:47:57 +08:00
PASSED();
return 0;
error:
H5E_BEGIN_TRY {
H5Sclose(space);
H5Dclose(dset);
H5Pclose(fapl);
H5Fclose(file);
} H5E_END_TRY;
return -1;
}
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
/*-------------------------------------------------------------------------
* Function: test_log
*
* Purpose: Tests the file handle interface for log driver
*
* Return: Success: 0
* Failure: -1
*
* Programmer: Dana Robinson
* Tuesday, March 22, 2011
*
*-------------------------------------------------------------------------
*/
static herr_t
test_log(void)
{
hid_t file = -1;
hid_t fapl = -1;
hid_t access_fapl = -1;
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
hid_t driver_id = -1; /* ID for this VFD */
unsigned long driver_flags = 0; /* VFD feature flags */
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
char filename[1024];
int *fhandle = NULL;
hsize_t file_size = 0;
unsigned int flags = H5FD_LOG_ALL;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
size_t buf_size = 4 * KB;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
TESTING("LOG file driver");
/* Set property list and file name for log driver. */
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
if((fapl = H5Pcreate(H5P_FILE_ACCESS)) < 0)
TEST_ERROR;
[svn-r23219] Description: Bring reviewed changes from Coverity branch back to trunk (QK & JK): r20457: Coverity issue 691: return of H5duo could be negative. Fixed by using STDOUT_FILENO and redesign parse_command_line and main to cleanup file allocations. The output_file var is null when using stdout. In cleanup do not close output_file if NULL. r20510: Initialize ufid = -1 and predicate HDclose call on ufid != -1 r20511: Purpose: Fix coverity issue 1715 Description: Free "file" and nested data on failure in H5FD_core_open. r20512: Initialize ifid = -1 and predicate HDclose call on ifid != -1 r20514: Initialize h5fid = -1 and predicate HDclose call on h5fid != -1 r20516: Added else branch to the if (ret_value < 0) check. r20522: Addressed coverity issues 930-933, 850, 836, 835, 1307. All minor potential buffer overwrite bugs, or coverity errors. Fixed by replacing strcpy and sprintf with strncpy and snprintf. r20523: fixed coverity issues 68, 1120, 1116i r20524: Check H5Z_SZIP->encoder_present < 1 assuming 0 represents absence. r20601: Purpose: Fix coverity issues 1703-1705 Description: Modified the cleanup code in test_free in accum.c to reset allocated buffers to NULL after they are freed, and modified the error cleanup code to check if these buffers are NULL before freeing them. Also fixed some unrelated warnings in accum.c. r20602: Use HDsnprintf and HDstrncat r20603: Purpose: Fix coverity issues 808-809 Description: Modified test_core in vfd.c to check the returns from malloc, and keep track of whether points and check are allocated by setting them to NULL when they are not. Added code to free points and check on error if they are not NULL. Also fixed unrelated warnings in vfd.c. r20604: Use HDstrncpy. r20605: Use HDstrncpy and HDstrncat. r20606: Purpose: Fix coverity issue 807 Description: Modified long_compact in stab.c to keep track of whether objname is allocated by setting it to NULL when it is not. Added code to free objname on error if it is not NULL. r20607: Changed string function calls to use versions that specify the string length to fix coverity issues 832 and 839. Tested on: Mac OSX/64 10.8.2 (amazon) (Too minor to require h5committest)
2013-02-02 09:53:32 +08:00
if(H5Pset_fapl_log(fapl, LOG_FILENAME, flags, buf_size) < 0)
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
TEST_ERROR;
h5_fixname(FILENAME[6], fapl, filename, sizeof filename);
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
/* Check that the VFD feature flags are correct */
if ((driver_id = H5Pget_driver(fapl)) < 0)
TEST_ERROR
if (H5FDdriver_query(driver_id, &driver_flags) < 0)
TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_ACCUMULATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_DATA_SIEVE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_SMALLDATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_POSIX_COMPAT_HANDLE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_SUPPORTS_SWMR_IO)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_DEFAULT_VFD_COMPATIBLE)) TEST_ERROR
/* Check for extra flags not accounted for above */
if(driver_flags != (H5FD_FEAT_AGGREGATE_METADATA
| H5FD_FEAT_ACCUMULATE_METADATA
| H5FD_FEAT_DATA_SIEVE
| H5FD_FEAT_AGGREGATE_SMALLDATA
| H5FD_FEAT_POSIX_COMPAT_HANDLE
| H5FD_FEAT_SUPPORTS_SWMR_IO
| H5FD_FEAT_DEFAULT_VFD_COMPATIBLE))
TEST_ERROR
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
/* Create the test file */
if((file = H5Fcreate(filename, H5F_ACC_TRUNC, H5P_DEFAULT, fapl)) < 0)
TEST_ERROR;
/* Retrieve the access property list... */
if((access_fapl = H5Fget_access_plist(file)) < 0)
TEST_ERROR;
/* Check that the driver is correct */
if(H5FD_LOG != H5Pget_driver(access_fapl))
TEST_ERROR;
/* ...and close the property list */
if(H5Pclose(access_fapl) < 0)
TEST_ERROR;
/* Check file handle API */
if(H5Fget_vfd_handle(file, H5P_DEFAULT, (void **)&fhandle) < 0)
TEST_ERROR;
if(*fhandle < 0)
TEST_ERROR;
/* Check file size API */
if(H5Fget_filesize(file, &file_size) < 0)
TEST_ERROR;
/* There is no guarantee the size of metadata in file is constant.
* Just try to check if it's reasonable. It's 2KB right now.
*/
if(file_size < 1 * KB || file_size > 4 * KB)
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
/* Close and delete the file */
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
if(H5Fclose(file) < 0)
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
h5_delete_test_file(FILENAME[6], fapl);
/* Close the fapl */
if(H5Pclose(fapl) < 0)
TEST_ERROR;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
PASSED();
return 0;
error:
H5E_BEGIN_TRY {
H5Pclose(fapl);
H5Fclose(file);
} H5E_END_TRY;
return -1;
}
/*-------------------------------------------------------------------------
* Function: test_stdio
*
* Purpose: Tests the file handle interface for STDIO driver
*
* Return: Success: 0
* Failure: -1
*
* Programmer: Dana Robinson
* Tuesday, March 22, 2011
*
*-------------------------------------------------------------------------
*/
static herr_t
test_stdio(void)
{
hid_t file = -1;
hid_t fapl = -1;
hid_t access_fapl = -1;
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
hid_t driver_id = -1; /* ID for this VFD */
unsigned long driver_flags = 0; /* VFD feature flags */
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
char filename[1024];
[svn-r20551] Description: Correct 'FILE **' back to 'FILE *' for H5Fget_vfd_handle call. Tested on: FreeBSD/32 8.2 (loyalty) w/gcc 4.6 & debug FreeBSD/64 8.2 (freedom) w/gcc 4.6 & debug Solaris 2.6 (linew)
2011-04-19 19:34:51 +08:00
FILE *fhandle = NULL;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
hsize_t file_size = 0;
TESTING("STDIO file driver");
/* Set property list and file name for STDIO driver. */
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
if((fapl = H5Pcreate(H5P_FILE_ACCESS)) < 0)
TEST_ERROR;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
if(H5Pset_fapl_stdio(fapl) < 0)
TEST_ERROR;
h5_fixname(FILENAME[7], fapl, filename, sizeof filename);
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
/* Check that the VFD feature flags are correct */
if ((driver_id = H5Pget_driver(fapl)) < 0)
TEST_ERROR
if (H5FDdriver_query(driver_id, &driver_flags) < 0)
TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_ACCUMULATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_DATA_SIEVE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_SMALLDATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_DEFAULT_VFD_COMPATIBLE)) TEST_ERROR
/* Check for extra flags not accounted for above */
if(driver_flags != (H5FD_FEAT_AGGREGATE_METADATA
| H5FD_FEAT_ACCUMULATE_METADATA
| H5FD_FEAT_DATA_SIEVE
| H5FD_FEAT_AGGREGATE_SMALLDATA
| H5FD_FEAT_DEFAULT_VFD_COMPATIBLE))
TEST_ERROR
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
if((file = H5Fcreate(filename, H5F_ACC_TRUNC, H5P_DEFAULT, fapl)) < 0)
TEST_ERROR;
/* Retrieve the access property list... */
if((access_fapl = H5Fget_access_plist(file)) < 0)
TEST_ERROR;
/* Check that the driver is correct */
if(H5FD_STDIO != H5Pget_driver(access_fapl))
TEST_ERROR;
/* ...and close the property list */
if(H5Pclose(access_fapl) < 0)
TEST_ERROR;
/* Check file handle API */
if(H5Fget_vfd_handle(file, H5P_DEFAULT, (void **)&fhandle) < 0)
TEST_ERROR;
[svn-r20551] Description: Correct 'FILE **' back to 'FILE *' for H5Fget_vfd_handle call. Tested on: FreeBSD/32 8.2 (loyalty) w/gcc 4.6 & debug FreeBSD/64 8.2 (freedom) w/gcc 4.6 & debug Solaris 2.6 (linew)
2011-04-19 19:34:51 +08:00
if(NULL == fhandle)
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
TEST_ERROR;
/* Check file size API */
if(H5Fget_filesize(file, &file_size) < 0)
TEST_ERROR;
/* There is no guarantee the size of metadata in file is constant.
* Just try to check if it's reasonable. It's 2KB right now.
*/
if(file_size < 1 * KB || file_size > 4 * KB)
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
/* Close and delete the file */
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
if(H5Fclose(file) < 0)
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
h5_delete_test_file(FILENAME[7], fapl);
/* Close the fapl */
if(H5Pclose(fapl) < 0)
TEST_ERROR;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
PASSED();
return 0;
error:
H5E_BEGIN_TRY {
H5Pclose(fapl);
H5Fclose(file);
} H5E_END_TRY;
return -1;
}
/*-------------------------------------------------------------------------
* Function: test_windows
*
* Purpose: Tests the file handle interface for WINDOWS driver
*
* Return: Success: 0
* Failure: -1
*
* Programmer: Dana Robinson
* Tuesday, March 22, 2011
*
*-------------------------------------------------------------------------
*/
static herr_t
test_windows(void)
{
[svn-r21366] JIRA HDFFV-2748 replace use of _WIN32 in source. Created new defines H5_HAVE_WIN32_API and H5_HAVE_VISUAL_STUDIO defines to use. These can be properly set during configurration. Tested: windows and local linux - reviewed internally
2011-09-07 00:50:32 +08:00
#ifdef H5_HAVE_WINDOWS
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
hid_t file = -1;
hid_t fapl = -1;
hid_t access_fapl = -1;
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
hid_t driver_id = -1; /* ID for this VFD */
unsigned long driver_flags = 0; /* VFD feature flags */
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
char filename[1024];
int *fhandle = NULL;
hsize_t file_size = 0;
[svn-r21366] JIRA HDFFV-2748 replace use of _WIN32 in source. Created new defines H5_HAVE_WIN32_API and H5_HAVE_VISUAL_STUDIO defines to use. These can be properly set during configurration. Tested: windows and local linux - reviewed internally
2011-09-07 00:50:32 +08:00
#endif /*H5_HAVE_WINDOWS*/
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
TESTING("WINDOWS file driver");
[svn-r21366] JIRA HDFFV-2748 replace use of _WIN32 in source. Created new defines H5_HAVE_WIN32_API and H5_HAVE_VISUAL_STUDIO defines to use. These can be properly set during configurration. Tested: windows and local linux - reviewed internally
2011-09-07 00:50:32 +08:00
#ifndef H5_HAVE_WINDOWS
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
SKIPPED();
return 0;
[svn-r21366] JIRA HDFFV-2748 replace use of _WIN32 in source. Created new defines H5_HAVE_WIN32_API and H5_HAVE_VISUAL_STUDIO defines to use. These can be properly set during configurration. Tested: windows and local linux - reviewed internally
2011-09-07 00:50:32 +08:00
#else /* H5_HAVE_WINDOWS */
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
/* Set property list and file name for WINDOWS driver. */
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
if((fapl = H5Pcreate(H5P_FILE_ACCESS)) < 0)
TEST_ERROR;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
if(H5Pset_fapl_windows(fapl) < 0)
TEST_ERROR;
h5_fixname(FILENAME[8], fapl, filename, sizeof filename);
Merge of recent develop work to 1.10 branch: * H5FD_FEAT_DEFAULT_VFD_COMPATIBLE VFD feature flag (HDFFV-10214) * H5FDdriver_query() API call (HDFFV-10215) * Fix to skip SWMR tests that when the current VFD is not compatible with the default VFD (HDFFV-10197) * Removed obsolete compiler versions from config/gnu-flags (HDFFV-9937) * Many warning fixes
2017-06-01 05:45:41 +08:00
/* Check that the VFD feature flags are correct */
if ((driver_id = H5Pget_driver(fapl)) < 0)
TEST_ERROR
if (H5FDdriver_query(driver_id, &driver_flags) < 0)
TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_ACCUMULATE_METADATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_DATA_SIEVE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_AGGREGATE_SMALLDATA)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_POSIX_COMPAT_HANDLE)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_SUPPORTS_SWMR_IO)) TEST_ERROR
if(!(driver_flags & H5FD_FEAT_DEFAULT_VFD_COMPATIBLE)) TEST_ERROR
/* Check for extra flags not accounted for above */
if(driver_flags != (H5FD_FEAT_AGGREGATE_METADATA
| H5FD_FEAT_ACCUMULATE_METADATA
| H5FD_FEAT_DATA_SIEVE
| H5FD_FEAT_AGGREGATE_SMALLDATA
| H5FD_FEAT_POSIX_COMPAT_HANDLE
| H5FD_FEAT_SUPPORTS_SWMR_IO
| H5FD_FEAT_DEFAULT_VFD_COMPATIBLE))
TEST_ERROR
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
if((file = H5Fcreate(filename, H5F_ACC_TRUNC, H5P_DEFAULT, fapl)) < 0)
TEST_ERROR;
/* Retrieve the access property list... */
if((access_fapl = H5Fget_access_plist(file)) < 0)
TEST_ERROR;
/* Check that the driver is correct */
if(H5FD_WINDOWS!= H5Pget_driver(access_fapl))
TEST_ERROR;
/* ...and close the property list */
if(H5Pclose(access_fapl) < 0)
TEST_ERROR;
/* Check file handle API */
if(H5Fget_vfd_handle(file, H5P_DEFAULT, (void **)&fhandle) < 0)
TEST_ERROR;
if(*fhandle < 0)
TEST_ERROR;
/* Check file size API */
if(H5Fget_filesize(file, &file_size) < 0)
TEST_ERROR;
/* There is no guarantee the size of metadata in file is constant.
* Just try to check if it's reasonable. It's 2KB right now.
*/
if(file_size < 1 * KB || file_size > 4 * KB)
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
/* Close and delete the file */
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
if(H5Fclose(file) < 0)
TEST_ERROR;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
h5_delete_test_file(FILENAME[8], fapl);
/* Close the fapl */
if(H5Pclose(fapl) < 0)
TEST_ERROR;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
PASSED();
return 0;
error:
H5E_BEGIN_TRY {
H5Pclose(fapl);
H5Fclose(file);
} H5E_END_TRY;
return -1;
[svn-r21366] JIRA HDFFV-2748 replace use of _WIN32 in source. Created new defines H5_HAVE_WIN32_API and H5_HAVE_VISUAL_STUDIO defines to use. These can be properly set during configurration. Tested: windows and local linux - reviewed internally
2011-09-07 00:50:32 +08:00
#endif /* H5_HAVE_WINDOWS */
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
} /* end test_windows() */
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
/*-------------------------------------------------------------------------
* Function: main
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
*
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
* Purpose: Tests the basic features of Virtual File Drivers
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
*
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
* Return: Success: 0
* Failure: 1
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
*
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
* Programmer: Raymond Lu
* Tuesday, Sept 24, 2002
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
*
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
*-------------------------------------------------------------------------
*/
int
main(void)
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
{
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
int nerrors = 0;
[svn-r11245] Purpose: Code cleanup Description: Trim trailing whitespace, which is making 'diff'ing the two branches difficult. Solution: Ran this script in each directory: foreach f (*.[ch] *.cpp) sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f end Platforms tested: FreeBSD 4.11 (sleipnir) Too minor to require h5committest
2005-08-14 04:53:35 +08:00
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
h5_reset();
HDprintf("Testing basic Virtual File Driver functionality.\n");
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
nerrors += test_sec2() < 0 ? 1 : 0;
nerrors += test_core() < 0 ? 1 : 0;
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
nerrors += test_direct() < 0 ? 1 : 0;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
nerrors += test_family() < 0 ? 1 : 0;
nerrors += test_family_compat() < 0 ? 1 : 0;
nerrors += test_multi() < 0 ? 1 : 0;
[svn-r21015] Bug fix for Issue 2598 - In v1.6 library, there was EOA for the whole MULTI file saved in the super block. We took it out in v1.8 library because it's meaningless for the MULTI file. v1.8 library saves the EOA for the metadata file, instead. But this caused some backward compatibility problem. v1.8 library couldn't open the file created with v1.6 library. I fixed the problem by checking the EOA value to detect the file created with v1.6 library. Tested on jam, koala, and heiwa.
2011-06-22 23:47:57 +08:00
nerrors += test_multi_compat() < 0 ? 1 : 0;
[svn-r20459] Purpose: Progress on bug 2138 (bugzilla) / HDFFV-4298 (JIRA) Description: - Added basic VFD tests (i.e. cloned the SEC2 test) for the Windows, STDIO and log VFDs. - Added H5Pget_driver() sanity check to all VFD's tests. - Added log output file to list of files to clean up in Makefile.am. Tested on: - h5committest script platforms (jam, amani, heiwa) - Windows 7 (32- and 64-bit)
2011-04-09 05:31:31 +08:00
nerrors += test_log() < 0 ? 1 : 0;
nerrors += test_stdio() < 0 ? 1 : 0;
nerrors += test_windows() < 0 ? 1 : 0;
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
if(nerrors) {
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
HDprintf("***** %d Virtual File Driver TEST%s FAILED! *****\n",
[svn-r27768] Description: Complete revamp of package initialization/shutdown mechanism in the library. Each package now has a single init/term routine. This new way should avoid packages being re-initialized during library shutdown and is also be _much_ more proactive about giving feedback for resource leaks internal to the library. Introduces a new "module" header file for packages in the library (e.g src/H5Fmodule.h) which sets up some necessary package configuration macros for the FUNC_ENTER/LEAVE macros. (The VFL drivers have their own slightly modified version of this header, src/H5FDdrvr_module.h) Also cleaned up a bunch of resources leaks all across the library and tests, along with addressing many warnings, as I encountered them. Tested on: MacOSX/64 10.10.5 (amazon) w/serial & parallel Linux/64 3.10.x (kituo) w/serial & parallel Linux/64 2.6.x (ostrich) w/serial
2015-09-14 11:58:59 +08:00
nerrors, nerrors > 1 ? "S" : "");
return 1;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
} /* end if */
HDprintf("All Virtual File Driver tests passed.\n");
[svn-r10979] Purpose: Renamed file_handle.c as vfd.c since its tests have evloved from merely file handles tests to other basic features of VFD's. Did some minor cosmetic changes in vfd.c. Updated Makefile.am with the new name. Makefile.in is regenerated. Platforms tested: In heping only since the changes are trivial. Misc. update:
2005-06-24 22:32:05 +08:00
return 0;
[svn-r30306] Updated vfd test to be better about fapls and h5_reset(). Prevents the test from segfaulting in other sub-tests when a previous test fails. Tested on: 64-bit Ubuntu Linux 16.04 LTS w/ gcc 5.4.0 Autotools serial + direct VFD (vfd test only) Also tested with a forced error to ensure that the test no longer segfaults and confirmed that all test files are cleaned up.
2016-08-20 00:30:05 +08:00
} /* end main() */
[svn-r14187] Description: Put H5Acreate() under API versioning, with all internal usage shifted to H5Acreate2(). Add regression tests for H5Acreate1(). Tested on: FreeBSD/32 6.2 (duty) in debug mode FreeBSD/64 6.2 (liberty) w/C++ & FORTRAN, in debug mode Linux/32 2.6 (kagiso) w/PGI compilers, w/C++ & FORTRAN, w/threadsafe, in debug mode Linux/64-amd64 2.6 (smirom) w/default API=1.6.x, w/C++ & FORTRAN, in production mode Linux/64-ia64 2.6 (cobalt) w/Intel compilers, w/C++ & FORTRAN, in production mode Solaris/32 2.10 (linew) w/deprecated symbols disabled, w/C++ & FORTRAN, w/szip filter, in production mode Mac OS X/32 10.4.10 (amazon) in debug mode
2007-10-05 06:19:07 +08:00
Reference in New Issue Copy Permalink